diff options
author | Ton Voon <tonvoon@users.sourceforge.net> | 2007-01-24 22:47:25 +0000 |
---|---|---|
committer | Ton Voon <tonvoon@users.sourceforge.net> | 2007-01-24 22:47:25 +0000 |
commit | fe856aa957978504137c1d425815d4ed8a22be40 (patch) | |
tree | a5bb46ce0e64b2056f75700eadbf27aba7c39418 /gl | |
parent | 210f39bc84cfbb21cd72dc054e43f13815ee0616 (diff) | |
download | monitoring-plugins-fe856aa957978504137c1d425815d4ed8a22be40.tar.gz |
Sync with gnulib - lots of extraneous code removed in preference to GNU code
git-svn-id: https://nagiosplug.svn.sourceforge.net/svnroot/nagiosplug/nagiosplug/trunk@1580 f882894a-f735-0410-b71e-b25c423dba1c
Diffstat (limited to 'gl')
190 files changed, 31850 insertions, 0 deletions
diff --git a/gl/Makefile.am b/gl/Makefile.am new file mode 100644 index 00000000..67c9872d --- /dev/null +++ b/gl/Makefile.am @@ -0,0 +1,607 @@ +## DO NOT EDIT! GENERATED AUTOMATICALLY! +## Process this file with automake to produce Makefile.in. +# Copyright (C) 2004-2007 Free Software Foundation, Inc. +# +# This file is free software, distributed under the terms of the GNU +# General Public License. As a special exception to the GNU General +# Public License, this file may be distributed as part of a program +# that contains a configuration script generated by Autoconf, under +# the same distribution terms as the rest of that program. +# +# Generated by gnulib-tool. +# Reproduce by: gnulib-tool --import --dir=. --lib=libgnu --source-base=gl --m4-base=gl/m4 --doc-base=doc --aux-dir=. --no-libtool --macro-prefix=gl dirname fsusage getaddrinfo gethostname getloadavg getopt gettext mountlist regex vasprintf vsnprintf + +AUTOMAKE_OPTIONS = 1.5 gnits + +noinst_HEADERS = +noinst_LIBRARIES = +noinst_LTLIBRARIES = +EXTRA_DIST = +BUILT_SOURCES = +SUFFIXES = +MOSTLYCLEANFILES = core *.stackdump +MOSTLYCLEANDIRS = +CLEANFILES = +DISTCLEANFILES = +MAINTAINERCLEANFILES = + +AM_CPPFLAGS = + +noinst_LIBRARIES += libgnu.a + +libgnu_a_SOURCES = +libgnu_a_LIBADD = $(gl_LIBOBJS) +libgnu_a_DEPENDENCIES = $(gl_LIBOBJS) +EXTRA_libgnu_a_SOURCES = + +## begin gnulib module alloca + + +EXTRA_DIST += alloca.c + +EXTRA_libgnu_a_SOURCES += alloca.c + +libgnu_a_LIBADD += @ALLOCA@ +libgnu_a_DEPENDENCIES += @ALLOCA@ +## end gnulib module alloca + +## begin gnulib module alloca-opt + +BUILT_SOURCES += $(ALLOCA_H) + +# We need the following in order to create <alloca.h> when the system +# doesn't have one that works with the given compiler. +alloca.h: alloca_.h + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + cat $(srcdir)/alloca_.h; \ + } > $@-t + mv -f $@-t $@ +MOSTLYCLEANFILES += alloca.h alloca.h-t + +EXTRA_DIST += alloca_.h + +## end gnulib module alloca-opt + +## begin gnulib module arpa_inet + +BUILT_SOURCES += $(ARPA_INET_H) + +# We need the following in order to create <arpa/inet.h> when the system +# doesn't have one. +arpa/inet.h: + test -d arpa || mkdir arpa + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + echo '#include <sys/socket.h>'; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += arpa/inet.h arpa/inet.h-t +MOSTLYCLEANDIRS += arpa + +## end gnulib module arpa_inet + +## begin gnulib module c-strtod + + +EXTRA_DIST += c-strtod.c c-strtod.h + +EXTRA_libgnu_a_SOURCES += c-strtod.c + +## end gnulib module c-strtod + +## begin gnulib module cloexec + + +EXTRA_DIST += cloexec.c cloexec.h + +EXTRA_libgnu_a_SOURCES += cloexec.c + +## end gnulib module cloexec + +## begin gnulib module dirname + + +EXTRA_DIST += basename.c dirname.c dirname.h stripslash.c + +EXTRA_libgnu_a_SOURCES += basename.c dirname.c stripslash.c + +## end gnulib module dirname + +## begin gnulib module error + + +EXTRA_DIST += error.c error.h + +EXTRA_libgnu_a_SOURCES += error.c + +## end gnulib module error + +## begin gnulib module exit + +libgnu_a_SOURCES += exit.h + +## end gnulib module exit + +## begin gnulib module exitfail + + +EXTRA_DIST += exitfail.c exitfail.h + +EXTRA_libgnu_a_SOURCES += exitfail.c + +## end gnulib module exitfail + +## begin gnulib module fcntl-safer + + +EXTRA_DIST += creat-safer.c fcntl--.h fcntl-safer.h open-safer.c + +EXTRA_libgnu_a_SOURCES += creat-safer.c open-safer.c + +## end gnulib module fcntl-safer + +## begin gnulib module fsusage + + +EXTRA_DIST += fsusage.c fsusage.h + +EXTRA_libgnu_a_SOURCES += fsusage.c + +## end gnulib module fsusage + +## begin gnulib module full-read + +libgnu_a_SOURCES += full-read.h full-read.c + +## end gnulib module full-read + +## begin gnulib module full-write + +libgnu_a_SOURCES += full-write.h full-write.c + +## end gnulib module full-write + +## begin gnulib module getaddrinfo + + +EXTRA_DIST += gai_strerror.c getaddrinfo.c getaddrinfo.h + +EXTRA_libgnu_a_SOURCES += gai_strerror.c getaddrinfo.c + +## end gnulib module getaddrinfo + +## begin gnulib module gethostname + + +EXTRA_DIST += gethostname.c + +EXTRA_libgnu_a_SOURCES += gethostname.c + +## end gnulib module gethostname + +## begin gnulib module getloadavg + + +EXTRA_DIST += getloadavg.c + +EXTRA_libgnu_a_SOURCES += getloadavg.c + +## end gnulib module getloadavg + +## begin gnulib module getopt + +BUILT_SOURCES += $(GETOPT_H) + +# We need the following in order to create <getopt.h> when the system +# doesn't have one that works with the given compiler. +getopt.h: getopt_.h + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + cat $(srcdir)/getopt_.h; \ + } > $@-t + mv -f $@-t $@ +MOSTLYCLEANFILES += getopt.h getopt.h-t + +EXTRA_DIST += getopt.c getopt1.c getopt_.h getopt_int.h + +EXTRA_libgnu_a_SOURCES += getopt.c getopt1.c + +## end gnulib module getopt + +## begin gnulib module gettext + +# This is for those projects which use "gettextize --intl" to put a source-code +# copy of libintl into their package. In such projects, every Makefile.am needs +# -I$(top_builddir)/intl, so that <libintl.h> can be found in this directory. +# For the Makefile.ams in other directories it is the maintainer's +# responsibility; for the one from gnulib we do it here. +# This option has no effect when the user disables NLS (because then the intl +# directory contains no libintl.h file) or when the project does not use +# "gettextize --intl". +AM_CPPFLAGS += -I$(top_builddir)/intl + +## end gnulib module gettext + +## begin gnulib module gettext-h + +libgnu_a_SOURCES += gettext.h + +## end gnulib module gettext-h + +## begin gnulib module inet_ntop + + +EXTRA_DIST += inet_ntop.c inet_ntop.h + +EXTRA_libgnu_a_SOURCES += inet_ntop.c + +## end gnulib module inet_ntop + +## begin gnulib module intprops + + +EXTRA_DIST += intprops.h + +## end gnulib module intprops + +## begin gnulib module malloc + + +EXTRA_DIST += malloc.c + +EXTRA_libgnu_a_SOURCES += malloc.c + +## end gnulib module malloc + +## begin gnulib module mbchar + + +EXTRA_DIST += mbchar.c mbchar.h + +EXTRA_libgnu_a_SOURCES += mbchar.c + +## end gnulib module mbchar + +## begin gnulib module mbuiter + +libgnu_a_SOURCES += mbuiter.h + +## end gnulib module mbuiter + +## begin gnulib module memchr + + +EXTRA_DIST += memchr.c + +EXTRA_libgnu_a_SOURCES += memchr.c + +## end gnulib module memchr + +## begin gnulib module minmax + +libgnu_a_SOURCES += minmax.h + +## end gnulib module minmax + +## begin gnulib module mountlist + + +EXTRA_DIST += mountlist.c mountlist.h + +EXTRA_libgnu_a_SOURCES += mountlist.c + +## end gnulib module mountlist + +## begin gnulib module netinet_in + +BUILT_SOURCES += $(NETINET_IN_H) + +# We need the following in order to create <netinet/in.h> when the system +# doesn't have one. +netinet/in.h: + test -d netinet || mkdir netinet + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + echo '#include <sys/socket.h>'; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += netinet/in.h netinet/in.h-t +MOSTLYCLEANDIRS += netinet + +## end gnulib module netinet_in + +## begin gnulib module regex + + +EXTRA_DIST += regcomp.c regex.c regex.h regex_internal.c regex_internal.h regexec.c + +EXTRA_libgnu_a_SOURCES += regcomp.c regex.c regex_internal.c regexec.c + +## end gnulib module regex + +## begin gnulib module safe-read + + +EXTRA_DIST += safe-read.c safe-read.h + +EXTRA_libgnu_a_SOURCES += safe-read.c + +## end gnulib module safe-read + +## begin gnulib module safe-write + + +EXTRA_DIST += safe-write.c safe-write.h + +EXTRA_libgnu_a_SOURCES += safe-write.c + +## end gnulib module safe-write + +## begin gnulib module size_max + +libgnu_a_SOURCES += size_max.h + +## end gnulib module size_max + +## begin gnulib module snprintf + + +EXTRA_DIST += snprintf.c snprintf.h + +EXTRA_libgnu_a_SOURCES += snprintf.c + +## end gnulib module snprintf + +## begin gnulib module stdbool + +BUILT_SOURCES += $(STDBOOL_H) + +# We need the following in order to create <stdbool.h> when the system +# doesn't have one that works. +stdbool.h: stdbool_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's/@''HAVE__BOOL''@/$(HAVE__BOOL)/g' < $(srcdir)/stdbool_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += stdbool.h stdbool.h-t + +EXTRA_DIST += stdbool_.h + +## end gnulib module stdbool + +## begin gnulib module stdint + +BUILT_SOURCES += $(STDINT_H) + +# We need the following in order to create <stdint.h> when the system +# doesn't have one that works with the given compiler. +stdint.h: stdint_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's/@''HAVE_STDINT_H''@/$(HAVE_STDINT_H)/g' \ + -e 's|@''ABSOLUTE_STDINT_H''@|$(ABSOLUTE_STDINT_H)|g' \ + -e 's/@''HAVE_SYS_TYPES_H''@/$(HAVE_SYS_TYPES_H)/g' \ + -e 's/@''HAVE_INTTYPES_H''@/$(HAVE_INTTYPES_H)/g' \ + -e 's/@''HAVE_SYS_INTTYPES_H''@/$(HAVE_SYS_INTTYPES_H)/g' \ + -e 's/@''HAVE_SYS_BITYPES_H''@/$(HAVE_SYS_BITYPES_H)/g' \ + -e 's/@''HAVE_LONG_LONG_INT''@/$(HAVE_LONG_LONG_INT)/g' \ + -e 's/@''HAVE_UNSIGNED_LONG_LONG_INT''@/$(HAVE_UNSIGNED_LONG_LONG_INT)/g' \ + -e 's/@''BITSIZEOF_PTRDIFF_T''@/$(BITSIZEOF_PTRDIFF_T)/g' \ + -e 's/@''PTRDIFF_T_SUFFIX''@/$(PTRDIFF_T_SUFFIX)/g' \ + -e 's/@''BITSIZEOF_SIG_ATOMIC_T''@/$(BITSIZEOF_SIG_ATOMIC_T)/g' \ + -e 's/@''HAVE_SIGNED_SIG_ATOMIC_T''@/$(HAVE_SIGNED_SIG_ATOMIC_T)/g' \ + -e 's/@''SIG_ATOMIC_T_SUFFIX''@/$(SIG_ATOMIC_T_SUFFIX)/g' \ + -e 's/@''BITSIZEOF_SIZE_T''@/$(BITSIZEOF_SIZE_T)/g' \ + -e 's/@''SIZE_T_SUFFIX''@/$(SIZE_T_SUFFIX)/g' \ + -e 's/@''BITSIZEOF_WCHAR_T''@/$(BITSIZEOF_WCHAR_T)/g' \ + -e 's/@''HAVE_SIGNED_WCHAR_T''@/$(HAVE_SIGNED_WCHAR_T)/g' \ + -e 's/@''WCHAR_T_SUFFIX''@/$(WCHAR_T_SUFFIX)/g' \ + -e 's/@''BITSIZEOF_WINT_T''@/$(BITSIZEOF_WINT_T)/g' \ + -e 's/@''HAVE_SIGNED_WINT_T''@/$(HAVE_SIGNED_WINT_T)/g' \ + -e 's/@''WINT_T_SUFFIX''@/$(WINT_T_SUFFIX)/g' \ + < $(srcdir)/stdint_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += stdint.h stdint.h-t + +EXTRA_DIST += stdint_.h + +## end gnulib module stdint + +## begin gnulib module strcase + + +EXTRA_DIST += strcase.h strcasecmp.c strncasecmp.c + +EXTRA_libgnu_a_SOURCES += strcasecmp.c strncasecmp.c + +## end gnulib module strcase + +## begin gnulib module strdup + + +EXTRA_DIST += strdup.c strdup.h + +EXTRA_libgnu_a_SOURCES += strdup.c + +## end gnulib module strdup + +## begin gnulib module strndup + + +EXTRA_DIST += strndup.c strndup.h + +EXTRA_libgnu_a_SOURCES += strndup.c + +## end gnulib module strndup + +## begin gnulib module strnlen + + +EXTRA_DIST += strnlen.c strnlen.h + +EXTRA_libgnu_a_SOURCES += strnlen.c + +## end gnulib module strnlen + +## begin gnulib module strnlen1 + +libgnu_a_SOURCES += strnlen1.h strnlen1.c + +## end gnulib module strnlen1 + +## begin gnulib module sys_socket + +BUILT_SOURCES += $(SYS_SOCKET_H) + +# We need the following in order to create <sys/socket.h> when the system +# doesn't have one that works with the given compiler. +sys/socket.h: socket_.h + @MKDIR_P@ sys + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + cat $(srcdir)/socket_.h; \ + } > $@-t + mv -f $@-t $@ +MOSTLYCLEANFILES += sys/socket.h sys/socket.h-t +MOSTLYCLEANDIRS += sys + +EXTRA_DIST += socket_.h + +## end gnulib module sys_socket + +## begin gnulib module unistd + +BUILT_SOURCES += $(UNISTD_H) + +# We need the following in order to create an empty placeholder for +# <unistd.h> when the system doesn't have one. +unistd.h: unistd_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's|@''ABSOLUTE_UNISTD_H''@|$(ABSOLUTE_UNISTD_H)|g' \ + < $(srcdir)/unistd_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += unistd.h unistd.h-t + +EXTRA_DIST += unistd_.h + +## end gnulib module unistd + +## begin gnulib module unistd-safer + + +EXTRA_DIST += dup-safer.c fd-safer.c pipe-safer.c unistd--.h unistd-safer.h + +EXTRA_libgnu_a_SOURCES += dup-safer.c fd-safer.c pipe-safer.c + +## end gnulib module unistd-safer + +## begin gnulib module vasnprintf + + +EXTRA_DIST += asnprintf.c printf-args.c printf-args.h printf-parse.c printf-parse.h vasnprintf.c vasnprintf.h + +EXTRA_libgnu_a_SOURCES += asnprintf.c printf-args.c printf-parse.c vasnprintf.c + +## end gnulib module vasnprintf + +## begin gnulib module vasprintf + + +EXTRA_DIST += asprintf.c vasprintf.c vasprintf.h + +EXTRA_libgnu_a_SOURCES += asprintf.c vasprintf.c + +## end gnulib module vasprintf + +## begin gnulib module vsnprintf + + +EXTRA_DIST += vsnprintf.c vsnprintf.h + +EXTRA_libgnu_a_SOURCES += vsnprintf.c + +## end gnulib module vsnprintf + +## begin gnulib module wchar + +BUILT_SOURCES += $(WCHAR_H) + +# We need the following in order to create <wchar.h> when the system +# version does not work standalone. +wchar.h: wchar_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's|@''ABSOLUTE_WCHAR_H''@|$(ABSOLUTE_WCHAR_H)|g' \ + < $(srcdir)/wchar_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += wchar.h wchar.h-t + +EXTRA_DIST += wchar_.h + +## end gnulib module wchar + +## begin gnulib module wctype + +BUILT_SOURCES += $(WCTYPE_H) + +# We need the following in order to create <wctype.h> when the system +# doesn't have one that works with the given compiler. +wctype.h: wctype_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's/@''HAVE_WCTYPE_H''@/$(HAVE_WCTYPE_H)/g' \ + -e 's|@''ABSOLUTE_WCTYPE_H''@|$(ABSOLUTE_WCTYPE_H)|g' \ + -e 's/@''HAVE_WCTYPE_CTMP_BUG''@/$(HAVE_WCTYPE_CTMP_BUG)/g' \ + -e 's/@''HAVE_WINT_T''@/$(HAVE_WINT_T)/g' \ + < $(srcdir)/wctype_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += wctype.h wctype.h-t + +EXTRA_DIST += wctype_.h + +## end gnulib module wctype + +## begin gnulib module wcwidth + +libgnu_a_SOURCES += wcwidth.h + +## end gnulib module wcwidth + +## begin gnulib module xalloc + + +EXTRA_DIST += xalloc.h xmalloc.c + +EXTRA_libgnu_a_SOURCES += xmalloc.c + +## end gnulib module xalloc + +## begin gnulib module xalloc-die + +libgnu_a_SOURCES += xalloc-die.c + +## end gnulib module xalloc-die + +## begin gnulib module xsize + +libgnu_a_SOURCES += xsize.h + +## end gnulib module xsize + +## begin gnulib module xstrndup + +libgnu_a_SOURCES += xstrndup.h xstrndup.c + +## end gnulib module xstrndup + + +mostlyclean-local: mostlyclean-generic + @for dir in '' $(MOSTLYCLEANDIRS); do \ + if test -n "$$dir" && test -d $$dir; then \ + echo "rmdir $$dir"; rmdir $$dir; \ + fi; \ + done diff --git a/gl/alloca.c b/gl/alloca.c new file mode 100644 index 00000000..3a1f4e27 --- /dev/null +++ b/gl/alloca.c @@ -0,0 +1,489 @@ +/* alloca.c -- allocate automatically reclaimed memory + (Mostly) portable public-domain implementation -- D A Gwyn + + This implementation of the PWB library alloca function, + which is used to allocate space off the run-time stack so + that it is automatically reclaimed upon procedure exit, + was inspired by discussions with J. Q. Johnson of Cornell. + J.Otto Tennant <jot@cray.com> contributed the Cray support. + + There are some preprocessor constants that can + be defined when compiling for your specific system, for + improved efficiency; however, the defaults should be okay. + + The general concept of this implementation is to keep + track of all alloca-allocated blocks, and reclaim any + that are found to be deeper in the stack than the current + invocation. This heuristic does not reclaim storage as + soon as it becomes invalid, but it will do so eventually. + + As a special case, alloca(0) reclaims storage without + allocating any. It is a good idea to use alloca(0) in + your main control loop, etc. to force garbage collection. */ + +#include <config.h> + +#include <alloca.h> + +#include <string.h> +#include <stdlib.h> + +#ifdef emacs +# include "lisp.h" +# include "blockinput.h" +# ifdef EMACS_FREE +# undef free +# define free EMACS_FREE +# endif +#else +# define memory_full() abort () +#endif + +/* If compiling with GCC 2, this file's not needed. */ +#if !defined (__GNUC__) || __GNUC__ < 2 + +/* If someone has defined alloca as a macro, + there must be some other way alloca is supposed to work. */ +# ifndef alloca + +# ifdef emacs +# ifdef static +/* actually, only want this if static is defined as "" + -- this is for usg, in which emacs must undefine static + in order to make unexec workable + */ +# ifndef STACK_DIRECTION +you +lose +-- must know STACK_DIRECTION at compile-time +/* Using #error here is not wise since this file should work for + old and obscure compilers. */ +# endif /* STACK_DIRECTION undefined */ +# endif /* static */ +# endif /* emacs */ + +/* If your stack is a linked list of frames, you have to + provide an "address metric" ADDRESS_FUNCTION macro. */ + +# if defined (CRAY) && defined (CRAY_STACKSEG_END) +long i00afunc (); +# define ADDRESS_FUNCTION(arg) (char *) i00afunc (&(arg)) +# else +# define ADDRESS_FUNCTION(arg) &(arg) +# endif + +/* Define STACK_DIRECTION if you know the direction of stack + growth for your system; otherwise it will be automatically + deduced at run-time. + + STACK_DIRECTION > 0 => grows toward higher addresses + STACK_DIRECTION < 0 => grows toward lower addresses + STACK_DIRECTION = 0 => direction of growth unknown */ + +# ifndef STACK_DIRECTION +# define STACK_DIRECTION 0 /* Direction unknown. */ +# endif + +# if STACK_DIRECTION != 0 + +# define STACK_DIR STACK_DIRECTION /* Known at compile-time. */ + +# else /* STACK_DIRECTION == 0; need run-time code. */ + +static int stack_dir; /* 1 or -1 once known. */ +# define STACK_DIR stack_dir + +static void +find_stack_direction (void) +{ + static char *addr = NULL; /* Address of first `dummy', once known. */ + auto char dummy; /* To get stack address. */ + + if (addr == NULL) + { /* Initial entry. */ + addr = ADDRESS_FUNCTION (dummy); + + find_stack_direction (); /* Recurse once. */ + } + else + { + /* Second entry. */ + if (ADDRESS_FUNCTION (dummy) > addr) + stack_dir = 1; /* Stack grew upward. */ + else + stack_dir = -1; /* Stack grew downward. */ + } +} + +# endif /* STACK_DIRECTION == 0 */ + +/* An "alloca header" is used to: + (a) chain together all alloca'ed blocks; + (b) keep track of stack depth. + + It is very important that sizeof(header) agree with malloc + alignment chunk size. The following default should work okay. */ + +# ifndef ALIGN_SIZE +# define ALIGN_SIZE sizeof(double) +# endif + +typedef union hdr +{ + char align[ALIGN_SIZE]; /* To force sizeof(header). */ + struct + { + union hdr *next; /* For chaining headers. */ + char *deep; /* For stack depth measure. */ + } h; +} header; + +static header *last_alloca_header = NULL; /* -> last alloca header. */ + +/* Return a pointer to at least SIZE bytes of storage, + which will be automatically reclaimed upon exit from + the procedure that called alloca. Originally, this space + was supposed to be taken from the current stack frame of the + caller, but that method cannot be made to work for some + implementations of C, for example under Gould's UTX/32. */ + +void * +alloca (size_t size) +{ + auto char probe; /* Probes stack depth: */ + register char *depth = ADDRESS_FUNCTION (probe); + +# if STACK_DIRECTION == 0 + if (STACK_DIR == 0) /* Unknown growth direction. */ + find_stack_direction (); +# endif + + /* Reclaim garbage, defined as all alloca'd storage that + was allocated from deeper in the stack than currently. */ + + { + register header *hp; /* Traverses linked list. */ + +# ifdef emacs + BLOCK_INPUT; +# endif + + for (hp = last_alloca_header; hp != NULL;) + if ((STACK_DIR > 0 && hp->h.deep > depth) + || (STACK_DIR < 0 && hp->h.deep < depth)) + { + register header *np = hp->h.next; + + free (hp); /* Collect garbage. */ + + hp = np; /* -> next header. */ + } + else + break; /* Rest are not deeper. */ + + last_alloca_header = hp; /* -> last valid storage. */ + +# ifdef emacs + UNBLOCK_INPUT; +# endif + } + + if (size == 0) + return NULL; /* No allocation required. */ + + /* Allocate combined header + user data storage. */ + + { + /* Address of header. */ + register header *new; + + size_t combined_size = sizeof (header) + size; + if (combined_size < sizeof (header)) + memory_full (); + + new = malloc (combined_size); + + if (! new) + memory_full (); + + new->h.next = last_alloca_header; + new->h.deep = depth; + + last_alloca_header = new; + + /* User storage begins just after header. */ + + return (void *) (new + 1); + } +} + +# if defined (CRAY) && defined (CRAY_STACKSEG_END) + +# ifdef DEBUG_I00AFUNC +# include <stdio.h> +# endif + +# ifndef CRAY_STACK +# define CRAY_STACK +# ifndef CRAY2 +/* Stack structures for CRAY-1, CRAY X-MP, and CRAY Y-MP */ +struct stack_control_header + { + long shgrow:32; /* Number of times stack has grown. */ + long shaseg:32; /* Size of increments to stack. */ + long shhwm:32; /* High water mark of stack. */ + long shsize:32; /* Current size of stack (all segments). */ + }; + +/* The stack segment linkage control information occurs at + the high-address end of a stack segment. (The stack + grows from low addresses to high addresses.) The initial + part of the stack segment linkage control information is + 0200 (octal) words. This provides for register storage + for the routine which overflows the stack. */ + +struct stack_segment_linkage + { + long ss[0200]; /* 0200 overflow words. */ + long sssize:32; /* Number of words in this segment. */ + long ssbase:32; /* Offset to stack base. */ + long:32; + long sspseg:32; /* Offset to linkage control of previous + segment of stack. */ + long:32; + long sstcpt:32; /* Pointer to task common address block. */ + long sscsnm; /* Private control structure number for + microtasking. */ + long ssusr1; /* Reserved for user. */ + long ssusr2; /* Reserved for user. */ + long sstpid; /* Process ID for pid based multi-tasking. */ + long ssgvup; /* Pointer to multitasking thread giveup. */ + long sscray[7]; /* Reserved for Cray Research. */ + long ssa0; + long ssa1; + long ssa2; + long ssa3; + long ssa4; + long ssa5; + long ssa6; + long ssa7; + long sss0; + long sss1; + long sss2; + long sss3; + long sss4; + long sss5; + long sss6; + long sss7; + }; + +# else /* CRAY2 */ +/* The following structure defines the vector of words + returned by the STKSTAT library routine. */ +struct stk_stat + { + long now; /* Current total stack size. */ + long maxc; /* Amount of contiguous space which would + be required to satisfy the maximum + stack demand to date. */ + long high_water; /* Stack high-water mark. */ + long overflows; /* Number of stack overflow ($STKOFEN) calls. */ + long hits; /* Number of internal buffer hits. */ + long extends; /* Number of block extensions. */ + long stko_mallocs; /* Block allocations by $STKOFEN. */ + long underflows; /* Number of stack underflow calls ($STKRETN). */ + long stko_free; /* Number of deallocations by $STKRETN. */ + long stkm_free; /* Number of deallocations by $STKMRET. */ + long segments; /* Current number of stack segments. */ + long maxs; /* Maximum number of stack segments so far. */ + long pad_size; /* Stack pad size. */ + long current_address; /* Current stack segment address. */ + long current_size; /* Current stack segment size. This + number is actually corrupted by STKSTAT to + include the fifteen word trailer area. */ + long initial_address; /* Address of initial segment. */ + long initial_size; /* Size of initial segment. */ + }; + +/* The following structure describes the data structure which trails + any stack segment. I think that the description in 'asdef' is + out of date. I only describe the parts that I am sure about. */ + +struct stk_trailer + { + long this_address; /* Address of this block. */ + long this_size; /* Size of this block (does not include + this trailer). */ + long unknown2; + long unknown3; + long link; /* Address of trailer block of previous + segment. */ + long unknown5; + long unknown6; + long unknown7; + long unknown8; + long unknown9; + long unknown10; + long unknown11; + long unknown12; + long unknown13; + long unknown14; + }; + +# endif /* CRAY2 */ +# endif /* not CRAY_STACK */ + +# ifdef CRAY2 +/* Determine a "stack measure" for an arbitrary ADDRESS. + I doubt that "lint" will like this much. */ + +static long +i00afunc (long *address) +{ + struct stk_stat status; + struct stk_trailer *trailer; + long *block, size; + long result = 0; + + /* We want to iterate through all of the segments. The first + step is to get the stack status structure. We could do this + more quickly and more directly, perhaps, by referencing the + $LM00 common block, but I know that this works. */ + + STKSTAT (&status); + + /* Set up the iteration. */ + + trailer = (struct stk_trailer *) (status.current_address + + status.current_size + - 15); + + /* There must be at least one stack segment. Therefore it is + a fatal error if "trailer" is null. */ + + if (trailer == 0) + abort (); + + /* Discard segments that do not contain our argument address. */ + + while (trailer != 0) + { + block = (long *) trailer->this_address; + size = trailer->this_size; + if (block == 0 || size == 0) + abort (); + trailer = (struct stk_trailer *) trailer->link; + if ((block <= address) && (address < (block + size))) + break; + } + + /* Set the result to the offset in this segment and add the sizes + of all predecessor segments. */ + + result = address - block; + + if (trailer == 0) + { + return result; + } + + do + { + if (trailer->this_size <= 0) + abort (); + result += trailer->this_size; + trailer = (struct stk_trailer *) trailer->link; + } + while (trailer != 0); + + /* We are done. Note that if you present a bogus address (one + not in any segment), you will get a different number back, formed + from subtracting the address of the first block. This is probably + not what you want. */ + + return (result); +} + +# else /* not CRAY2 */ +/* Stack address function for a CRAY-1, CRAY X-MP, or CRAY Y-MP. + Determine the number of the cell within the stack, + given the address of the cell. The purpose of this + routine is to linearize, in some sense, stack addresses + for alloca. */ + +static long +i00afunc (long address) +{ + long stkl = 0; + + long size, pseg, this_segment, stack; + long result = 0; + + struct stack_segment_linkage *ssptr; + + /* Register B67 contains the address of the end of the + current stack segment. If you (as a subprogram) store + your registers on the stack and find that you are past + the contents of B67, you have overflowed the segment. + + B67 also points to the stack segment linkage control + area, which is what we are really interested in. */ + + stkl = CRAY_STACKSEG_END (); + ssptr = (struct stack_segment_linkage *) stkl; + + /* If one subtracts 'size' from the end of the segment, + one has the address of the first word of the segment. + + If this is not the first segment, 'pseg' will be + nonzero. */ + + pseg = ssptr->sspseg; + size = ssptr->sssize; + + this_segment = stkl - size; + + /* It is possible that calling this routine itself caused + a stack overflow. Discard stack segments which do not + contain the target address. */ + + while (!(this_segment <= address && address <= stkl)) + { +# ifdef DEBUG_I00AFUNC + fprintf (stderr, "%011o %011o %011o\n", this_segment, address, stkl); +# endif + if (pseg == 0) + break; + stkl = stkl - pseg; + ssptr = (struct stack_segment_linkage *) stkl; + size = ssptr->sssize; + pseg = ssptr->sspseg; + this_segment = stkl - size; + } + + result = address - this_segment; + + /* If you subtract pseg from the current end of the stack, + you get the address of the previous stack segment's end. + This seems a little convoluted to me, but I'll bet you save + a cycle somewhere. */ + + while (pseg != 0) + { +# ifdef DEBUG_I00AFUNC + fprintf (stderr, "%011o %011o\n", pseg, size); +# endif + stkl = stkl - pseg; + ssptr = (struct stack_segment_linkage *) stkl; + size = ssptr->sssize; + pseg = ssptr->sspseg; + result += size; + } + return (result); +} + +# endif /* not CRAY2 */ +# endif /* CRAY */ + +# endif /* no alloca */ +#endif /* not GCC version 2 */ diff --git a/gl/alloca_.h b/gl/alloca_.h new file mode 100644 index 00000000..dd0b3e98 --- /dev/null +++ b/gl/alloca_.h @@ -0,0 +1,54 @@ +/* Memory allocation on the stack. + + Copyright (C) 1995, 1999, 2001, 2002, 2003, 2004, 2006 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published + by the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + General Public License for more details. + + You should have received a copy of the GNU General Public + License along with this program; if not, write to the Free Software + Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, + USA. */ + +/* Avoid using the symbol _ALLOCA_H here, as Bison assumes _ALLOCA_H + means there is a real alloca function. */ +#ifndef _GNULIB_ALLOCA_H +# define _GNULIB_ALLOCA_H + +/* alloca (N) returns a pointer to N bytes of memory + allocated on the stack, which will last until the function returns. + Use of alloca should be avoided: + - inside arguments of function calls - undefined behaviour, + - in inline functions - the allocation may actually last until the + calling function returns, + - for huge N (say, N >= 65536) - you never know how large (or small) + the stack is, and when the stack cannot fulfill the memory allocation + request, the program just crashes. + */ + +#ifndef alloca +# ifdef __GNUC__ +# define alloca __builtin_alloca +# elif defined _AIX +# define alloca __alloca +# elif defined _MSC_VER +# include <malloc.h> +# define alloca _alloca +# else +# include <stddef.h> +# ifdef __cplusplus +extern "C" +# endif +void *alloca (size_t); +# endif +#endif + +#endif /* _GNULIB_ALLOCA_H */ diff --git a/gl/asnprintf.c b/gl/asnprintf.c new file mode 100644 index 00000000..26c3988b --- /dev/null +++ b/gl/asnprintf.c @@ -0,0 +1,35 @@ +/* Formatted output to strings. + Copyright (C) 1999, 2002, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include "vasnprintf.h" + +#include <stdarg.h> + +char * +asnprintf (char *resultbuf, size_t *lengthp, const char *format, ...) +{ + va_list args; + char *result; + + va_start (args, format); + result = vasnprintf (resultbuf, lengthp, format, args); + va_end (args); + return result; +} diff --git a/gl/asprintf.c b/gl/asprintf.c new file mode 100644 index 00000000..29ac6cf2 --- /dev/null +++ b/gl/asprintf.c @@ -0,0 +1,35 @@ +/* Formatted output to strings. + Copyright (C) 1999, 2002, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include "vasprintf.h" + +#include <stdarg.h> + +int +asprintf (char **resultp, const char *format, ...) +{ + va_list args; + int result; + + va_start (args, format); + result = vasprintf (resultp, format, args); + va_end (args); + return result; +} diff --git a/gl/basename.c b/gl/basename.c new file mode 100644 index 00000000..fbe17ff9 --- /dev/null +++ b/gl/basename.c @@ -0,0 +1,129 @@ +/* basename.c -- return the last element in a file name + + Copyright (C) 1990, 1998, 1999, 2000, 2001, 2003, 2004, 2005, 2006 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "dirname.h" + +#include <string.h> +#include "xalloc.h" +#include "xstrndup.h" + +/* Return the address of the last file name component of NAME. If + NAME has no relative file name components because it is a file + system root, return the empty string. */ + +char * +last_component (char const *name) +{ + char const *base = name + FILE_SYSTEM_PREFIX_LEN (name); + char const *p; + bool saw_slash = false; + + while (ISSLASH (*base)) + base++; + + for (p = base; *p; p++) + { + if (ISSLASH (*p)) + saw_slash = true; + else if (saw_slash) + { + base = p; + saw_slash = false; + } + } + + return (char *) base; +} + + +/* In general, we can't use the builtin `basename' function if available, + since it has different meanings in different environments. + In some environments the builtin `basename' modifies its argument. + + Return the last file name component of NAME, allocated with + xmalloc. On systems with drive letters, a leading "./" + distinguishes relative names that would otherwise look like a drive + letter. Unlike POSIX basename(), NAME cannot be NULL, + base_name("") returns "", and the first trailing slash is not + stripped. + + If lstat (NAME) would succeed, then { chdir (dir_name (NAME)); + lstat (base_name (NAME)); } will access the same file. Likewise, + if the sequence { chdir (dir_name (NAME)); + rename (base_name (NAME), "foo"); } succeeds, you have renamed NAME + to "foo" in the same directory NAME was in. */ + +char * +base_name (char const *name) +{ + char const *base = last_component (name); + size_t length; + + /* If there is no last component, then name is a file system root or the + empty string. */ + if (! *base) + return xstrndup (name, base_len (name)); + + /* Collapse a sequence of trailing slashes into one. */ + length = base_len (base); + if (ISSLASH (base[length])) + length++; + + /* On systems with drive letters, `a/b:c' must return `./b:c' rather + than `b:c' to avoid confusion with a drive letter. On systems + with pure POSIX semantics, this is not an issue. */ + if (FILE_SYSTEM_PREFIX_LEN (base)) + { + char *p = xmalloc (length + 3); + p[0] = '.'; + p[1] = '/'; + memcpy (p + 2, base, length); + p[length + 2] = '\0'; + return p; + } + + /* Finally, copy the basename. */ + return xstrndup (base, length); +} + +/* Return the length of the basename NAME. Typically NAME is the + value returned by base_name or last_component. Act like strlen + (NAME), except omit all trailing slashes. */ + +size_t +base_len (char const *name) +{ + size_t len; + size_t prefix_len = FILE_SYSTEM_PREFIX_LEN (name); + + for (len = strlen (name); 1 < len && ISSLASH (name[len - 1]); len--) + continue; + + if (DOUBLE_SLASH_IS_DISTINCT_ROOT && len == 1 + && ISSLASH (name[0]) && ISSLASH (name[1]) && ! name[2]) + return 2; + + if (FILE_SYSTEM_DRIVE_PREFIX_CAN_BE_RELATIVE && prefix_len + && len == prefix_len && ISSLASH (name[prefix_len])) + return prefix_len + 1; + + return len; +} diff --git a/gl/c-strtod.c b/gl/c-strtod.c new file mode 100644 index 00000000..2234ed0f --- /dev/null +++ b/gl/c-strtod.c @@ -0,0 +1,79 @@ +/* Convert string to double, using the C locale. + + Copyright (C) 2003, 2004, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <config.h> + +#include "c-strtod.h" + +#include <locale.h> +#include <stdlib.h> + +#include "xalloc.h" + +#if LONG +# define C_STRTOD c_strtold +# define DOUBLE long double +# define STRTOD_L strtold_l +#else +# define C_STRTOD c_strtod +# define DOUBLE double +# define STRTOD_L strtod_l +#endif + +/* c_strtold falls back on strtod if strtold doesn't conform to C99. */ +#if LONG && HAVE_C99_STRTOLD +# define STRTOD strtold +#else +# define STRTOD strtod +#endif + +DOUBLE +C_STRTOD (char const *nptr, char **endptr) +{ + DOUBLE r; + +#ifdef LC_ALL_MASK + + locale_t c_locale = newlocale (LC_ALL_MASK, "C", 0); + r = STRTOD_L (nptr, endptr, c_locale); + freelocale (c_locale); + +#else + + char *saved_locale = setlocale (LC_NUMERIC, NULL); + + if (saved_locale) + { + saved_locale = xstrdup (saved_locale); + setlocale (LC_NUMERIC, "C"); + } + + r = STRTOD (nptr, endptr); + + if (saved_locale) + { + setlocale (LC_NUMERIC, saved_locale); + free (saved_locale); + } + +#endif + + return r; +} diff --git a/gl/c-strtod.h b/gl/c-strtod.h new file mode 100644 index 00000000..ca9a9e7c --- /dev/null +++ b/gl/c-strtod.h @@ -0,0 +1,2 @@ +double c_strtod (char const *, char **); +long double c_strtold (char const *, char **); diff --git a/gl/cloexec.c b/gl/cloexec.c new file mode 100644 index 00000000..6480006a --- /dev/null +++ b/gl/cloexec.c @@ -0,0 +1,59 @@ +/* closexec.c - set or clear the close-on-exec descriptor flag + + Copyright (C) 1991, 2004, 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + + The code is taken from glibc/manual/llio.texi */ + +#include <config.h> + +#include "cloexec.h" + +#include <unistd.h> +#include <fcntl.h> + +#ifndef FD_CLOEXEC +# define FD_CLOEXEC 1 +#endif + +/* Set the `FD_CLOEXEC' flag of DESC if VALUE is true, + or clear the flag if VALUE is false. + Return 0 on success, or -1 on error with `errno' set. */ + +int +set_cloexec_flag (int desc, bool value) +{ +#if defined F_GETFD && defined F_SETFD + + int flags = fcntl (desc, F_GETFD, 0); + + if (0 <= flags) + { + int newflags = (value ? flags | FD_CLOEXEC : flags & ~FD_CLOEXEC); + + if (flags == newflags + || fcntl (desc, F_SETFD, newflags) != -1) + return 0; + } + + return -1; + +#else + + return 0; + +#endif +} diff --git a/gl/cloexec.h b/gl/cloexec.h new file mode 100644 index 00000000..c25921d6 --- /dev/null +++ b/gl/cloexec.h @@ -0,0 +1,2 @@ +#include <stdbool.h> +int set_cloexec_flag (int desc, bool value); diff --git a/gl/creat-safer.c b/gl/creat-safer.c new file mode 100644 index 00000000..f4a2e59d --- /dev/null +++ b/gl/creat-safer.c @@ -0,0 +1,32 @@ +/* Invoke creat, but avoid some glitches. + + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Jim Meyering. */ + +#include <config.h> + +#include "fcntl-safer.h" + +#include <fcntl.h> +#include "unistd-safer.h" + +int +creat_safer (char const *file, mode_t mode) +{ + return fd_safer (creat (file, mode)); +} diff --git a/gl/dirname.c b/gl/dirname.c new file mode 100644 index 00000000..16552c64 --- /dev/null +++ b/gl/dirname.c @@ -0,0 +1,85 @@ +/* dirname.c -- return all but the last element in a file name + + Copyright (C) 1990, 1998, 2000, 2001, 2003, 2004, 2005, 2006 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "dirname.h" + +#include <string.h> +#include "xalloc.h" + +/* Return the length of the prefix of FILE that will be used by + dir_name. If FILE is in the working directory, this returns zero + even though `dir_name (FILE)' will return ".". Works properly even + if there are trailing slashes (by effectively ignoring them). */ + +size_t +dir_len (char const *file) +{ + size_t prefix_length = FILE_SYSTEM_PREFIX_LEN (file); + size_t length; + + /* Advance prefix_length beyond important leading slashes. */ + prefix_length += (prefix_length != 0 + ? (FILE_SYSTEM_DRIVE_PREFIX_CAN_BE_RELATIVE + && ISSLASH (file[prefix_length])) + : (ISSLASH (file[0]) + ? ((DOUBLE_SLASH_IS_DISTINCT_ROOT + && ISSLASH (file[1]) && ! ISSLASH (file[2]) + ? 2 : 1)) + : 0)); + + /* Strip the basename and any redundant slashes before it. */ + for (length = last_component (file) - file; + prefix_length < length; length--) + if (! ISSLASH (file[length - 1])) + break; + return length; +} + + +/* In general, we can't use the builtin `dirname' function if available, + since it has different meanings in different environments. + In some environments the builtin `dirname' modifies its argument. + + Return the leading directories part of FILE, allocated with xmalloc. + Works properly even if there are trailing slashes (by effectively + ignoring them). Unlike POSIX dirname(), FILE cannot be NULL. + + If lstat (FILE) would succeed, then { chdir (dir_name (FILE)); + lstat (base_name (FILE)); } will access the same file. Likewise, + if the sequence { chdir (dir_name (FILE)); + rename (base_name (FILE), "foo"); } succeeds, you have renamed FILE + to "foo" in the same directory FILE was in. */ + +char * +dir_name (char const *file) +{ + size_t length = dir_len (file); + bool append_dot = (length == 0 + || (FILE_SYSTEM_DRIVE_PREFIX_CAN_BE_RELATIVE + && length == FILE_SYSTEM_PREFIX_LEN (file) + && file[2] != '\0' && ! ISSLASH (file[2]))); + char *dir = xmalloc (length + append_dot + 1); + memcpy (dir, file, length); + if (append_dot) + dir[length++] = '.'; + dir[length] = '\0'; + return dir; +} diff --git a/gl/dirname.h b/gl/dirname.h new file mode 100644 index 00000000..91e7ed33 --- /dev/null +++ b/gl/dirname.h @@ -0,0 +1,70 @@ +/* Take file names apart into directory and base names. + + Copyright (C) 1998, 2001, 2003-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef DIRNAME_H_ +# define DIRNAME_H_ 1 + +# include <stdbool.h> +# include <stddef.h> + +# ifndef DIRECTORY_SEPARATOR +# define DIRECTORY_SEPARATOR '/' +# endif + +# ifndef ISSLASH +# define ISSLASH(C) ((C) == DIRECTORY_SEPARATOR) +# endif + +# ifndef FILE_SYSTEM_PREFIX_LEN +# if FILE_SYSTEM_ACCEPTS_DRIVE_LETTER_PREFIX + /* This internal macro assumes ASCII, but all hosts that support drive + letters use ASCII. */ +# define _IS_DRIVE_LETTER(c) (((unsigned int) (c) | ('a' - 'A')) - 'a' \ + <= 'z' - 'a') +# define FILE_SYSTEM_PREFIX_LEN(Filename) \ + (_IS_DRIVE_LETTER ((Filename)[0]) && (Filename)[1] == ':' ? 2 : 0) +# else +# define FILE_SYSTEM_PREFIX_LEN(Filename) 0 +# endif +# endif + +# ifndef FILE_SYSTEM_DRIVE_PREFIX_CAN_BE_RELATIVE +# define FILE_SYSTEM_DRIVE_PREFIX_CAN_BE_RELATIVE 0 +# endif + +# ifndef DOUBLE_SLASH_IS_DISTINCT_ROOT +# define DOUBLE_SLASH_IS_DISTINCT_ROOT 0 +# endif + +# if FILE_SYSTEM_DRIVE_PREFIX_CAN_BE_RELATIVE +# define IS_ABSOLUTE_FILE_NAME(F) ISSLASH ((F)[FILE_SYSTEM_PREFIX_LEN (F)]) +# else +# define IS_ABSOLUTE_FILE_NAME(F) \ + (ISSLASH ((F)[0]) || 0 < FILE_SYSTEM_PREFIX_LEN (F)) +# endif +# define IS_RELATIVE_FILE_NAME(F) (! IS_ABSOLUTE_FILE_NAME (F)) + +char *base_name (char const *file); +char *dir_name (char const *file); +size_t base_len (char const *file); +size_t dir_len (char const *file); +char *last_component (char const *file); + +bool strip_trailing_slashes (char *file); + +#endif /* not DIRNAME_H_ */ diff --git a/gl/dup-safer.c b/gl/dup-safer.c new file mode 100644 index 00000000..7b12b615 --- /dev/null +++ b/gl/dup-safer.c @@ -0,0 +1,45 @@ +/* Invoke dup, but avoid some glitches. + + Copyright (C) 2001, 2004, 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <config.h> + +#include "unistd-safer.h" + +#include <fcntl.h> + +#include <unistd.h> +#ifndef STDERR_FILENO +# define STDERR_FILENO 2 +#endif + +/* Like dup, but do not return STDIN_FILENO, STDOUT_FILENO, or + STDERR_FILENO. */ + +int +dup_safer (int fd) +{ +#if defined F_DUPFD && !defined FCHDIR_REPLACEMENT + return fcntl (fd, F_DUPFD, STDERR_FILENO + 1); +#else + /* fd_safer calls us back, but eventually the recursion unwinds and + does the right thing. */ + return fd_safer (dup (fd)); +#endif +} diff --git a/gl/error.c b/gl/error.c new file mode 100644 index 00000000..cf863433 --- /dev/null +++ b/gl/error.c @@ -0,0 +1,338 @@ +/* Error handler for noninteractive utilities + Copyright (C) 1990-1998, 2000-2005, 2006 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by David MacKenzie <djm@gnu.ai.mit.edu>. */ + +#if !_LIBC +# include <config.h> +#endif + +#include "error.h" + +#include <stdarg.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +#if !_LIBC && ENABLE_NLS +# include "gettext.h" +#endif + +#ifdef _LIBC +# include <libintl.h> +# include <stdbool.h> +# include <stdint.h> +# include <wchar.h> +# define mbsrtowcs __mbsrtowcs +#endif + +#if USE_UNLOCKED_IO +# include "unlocked-io.h" +#endif + +#ifndef _ +# define _(String) String +#endif + +/* If NULL, error will flush stdout, then print on stderr the program + name, a colon and a space. Otherwise, error will call this + function without parameters instead. */ +void (*error_print_progname) (void); + +/* This variable is incremented each time `error' is called. */ +unsigned int error_message_count; + +#ifdef _LIBC +/* In the GNU C library, there is a predefined variable for this. */ + +# define program_name program_invocation_name +# include <errno.h> +# include <limits.h> +# include <libio/libioP.h> + +/* In GNU libc we want do not want to use the common name `error' directly. + Instead make it a weak alias. */ +extern void __error (int status, int errnum, const char *message, ...) + __attribute__ ((__format__ (__printf__, 3, 4))); +extern void __error_at_line (int status, int errnum, const char *file_name, + unsigned int line_number, const char *message, + ...) + __attribute__ ((__format__ (__printf__, 5, 6)));; +# define error __error +# define error_at_line __error_at_line + +# include <libio/iolibio.h> +# define fflush(s) INTUSE(_IO_fflush) (s) +# undef putc +# define putc(c, fp) INTUSE(_IO_putc) (c, fp) + +# include <bits/libc-lock.h> + +#else /* not _LIBC */ + +# if !HAVE_DECL_STRERROR_R && STRERROR_R_CHAR_P +# ifndef HAVE_DECL_STRERROR_R +"this configure-time declaration test was not run" +# endif +char *strerror_r (); +# endif + +/* The calling program should define program_name and set it to the + name of the executing program. */ +extern char *program_name; + +# if HAVE_STRERROR_R || defined strerror_r +# define __strerror_r strerror_r +# endif /* HAVE_STRERROR_R || defined strerror_r */ +#endif /* not _LIBC */ + +static void +print_errno_message (int errnum) +{ + char const *s; + +#if defined HAVE_STRERROR_R || _LIBC + char errbuf[1024]; +# if STRERROR_R_CHAR_P || _LIBC + s = __strerror_r (errnum, errbuf, sizeof errbuf); +# else + if (__strerror_r (errnum, errbuf, sizeof errbuf) == 0) + s = errbuf; + else + s = 0; +# endif +#else + s = strerror (errnum); +#endif + +#if !_LIBC + if (! s) + s = _("Unknown system error"); +#endif + +#if _LIBC + __fxprintf (NULL, ": %s", s); +#else + fprintf (stderr, ": %s", s); +#endif +} + +static void +error_tail (int status, int errnum, const char *message, va_list args) +{ +#if _LIBC + if (_IO_fwide (stderr, 0) > 0) + { +# define ALLOCA_LIMIT 2000 + size_t len = strlen (message) + 1; + wchar_t *wmessage = NULL; + mbstate_t st; + size_t res; + const char *tmp; + bool use_malloc = false; + + while (1) + { + if (__libc_use_alloca (len * sizeof (wchar_t))) + wmessage = (wchar_t *) alloca (len * sizeof (wchar_t)); + else + { + if (!use_malloc) + wmessage = NULL; + + wchar_t *p = (wchar_t *) realloc (wmessage, + len * sizeof (wchar_t)); + if (p == NULL) + { + free (wmessage); + fputws_unlocked (L"out of memory\n", stderr); + return; + } + wmessage = p; + use_malloc = true; + } + + memset (&st, '\0', sizeof (st)); + tmp = message; + + res = mbsrtowcs (wmessage, &tmp, len, &st); + if (res != len) + break; + + if (__builtin_expect (len >= SIZE_MAX / 2, 0)) + { + /* This really should not happen if everything is fine. */ + res = (size_t) -1; + break; + } + + len *= 2; + } + + if (res == (size_t) -1) + { + /* The string cannot be converted. */ + if (use_malloc) + { + free (wmessage); + use_malloc = false; + } + wmessage = (wchar_t *) L"???"; + } + + __vfwprintf (stderr, wmessage, args); + + if (use_malloc) + free (wmessage); + } + else +#endif + vfprintf (stderr, message, args); + va_end (args); + + ++error_message_count; + if (errnum) + print_errno_message (errnum); +#if _LIBC + __fxprintf (NULL, "\n"); +#else + putc ('\n', stderr); +#endif + fflush (stderr); + if (status) + exit (status); +} + + +/* Print the program name and error message MESSAGE, which is a printf-style + format string with optional args. + If ERRNUM is nonzero, print its corresponding system error message. + Exit with status STATUS if it is nonzero. */ +void +error (int status, int errnum, const char *message, ...) +{ + va_list args; + +#if defined _LIBC && defined __libc_ptf_call + /* We do not want this call to be cut short by a thread + cancellation. Therefore disable cancellation for now. */ + int state = PTHREAD_CANCEL_ENABLE; + __libc_ptf_call (pthread_setcancelstate, (PTHREAD_CANCEL_DISABLE, &state), + 0); +#endif + + fflush (stdout); +#ifdef _LIBC + _IO_flockfile (stderr); +#endif + if (error_print_progname) + (*error_print_progname) (); + else + { +#if _LIBC + __fxprintf (NULL, "%s: ", program_name); +#else + fprintf (stderr, "%s: ", program_name); +#endif + } + + va_start (args, message); + error_tail (status, errnum, message, args); + +#ifdef _LIBC + _IO_funlockfile (stderr); +# ifdef __libc_ptf_call + __libc_ptf_call (pthread_setcancelstate, (state, NULL), 0); +# endif +#endif +} + +/* Sometimes we want to have at most one error per line. This + variable controls whether this mode is selected or not. */ +int error_one_per_line; + +void +error_at_line (int status, int errnum, const char *file_name, + unsigned int line_number, const char *message, ...) +{ + va_list args; + + if (error_one_per_line) + { + static const char *old_file_name; + static unsigned int old_line_number; + + if (old_line_number == line_number + && (file_name == old_file_name + || strcmp (old_file_name, file_name) == 0)) + /* Simply return and print nothing. */ + return; + + old_file_name = file_name; + old_line_number = line_number; + } + +#if defined _LIBC && defined __libc_ptf_call + /* We do not want this call to be cut short by a thread + cancellation. Therefore disable cancellation for now. */ + int state = PTHREAD_CANCEL_ENABLE; + __libc_ptf_call (pthread_setcancelstate, (PTHREAD_CANCEL_DISABLE, &state), + 0); +#endif + + fflush (stdout); +#ifdef _LIBC + _IO_flockfile (stderr); +#endif + if (error_print_progname) + (*error_print_progname) (); + else + { +#if _LIBC + __fxprintf (NULL, "%s:", program_name); +#else + fprintf (stderr, "%s:", program_name); +#endif + } + +#if _LIBC + __fxprintf (NULL, file_name != NULL ? "%s:%d: " : " ", + file_name, line_number); +#else + fprintf (stderr, file_name != NULL ? "%s:%d: " : " ", + file_name, line_number); +#endif + + va_start (args, message); + error_tail (status, errnum, message, args); + +#ifdef _LIBC + _IO_funlockfile (stderr); +# ifdef __libc_ptf_call + __libc_ptf_call (pthread_setcancelstate, (state, NULL), 0); +# endif +#endif +} + +#ifdef _LIBC +/* Make the weak alias. */ +# undef error +# undef error_at_line +weak_alias (__error, error) +weak_alias (__error_at_line, error_at_line) +#endif diff --git a/gl/error.h b/gl/error.h new file mode 100644 index 00000000..5a5f2476 --- /dev/null +++ b/gl/error.h @@ -0,0 +1,66 @@ +/* Declaration for error-reporting function + Copyright (C) 1995, 1996, 1997, 2003, 2006 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _ERROR_H +#define _ERROR_H 1 + +#ifndef __attribute__ +/* This feature is available in gcc versions 2.5 and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 5) || __STRICT_ANSI__ +# define __attribute__(Spec) /* empty */ +# endif +/* The __-protected variants of `format' and `printf' attributes + are accepted by gcc versions 2.6.4 (effectively 2.7) and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 7) +# define __format__ format +# define __printf__ printf +# endif +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +/* Print a message with `fprintf (stderr, FORMAT, ...)'; + if ERRNUM is nonzero, follow it with ": " and strerror (ERRNUM). + If STATUS is nonzero, terminate the program with `exit (STATUS)'. */ + +extern void error (int __status, int __errnum, const char *__format, ...) + __attribute__ ((__format__ (__printf__, 3, 4))); + +extern void error_at_line (int __status, int __errnum, const char *__fname, + unsigned int __lineno, const char *__format, ...) + __attribute__ ((__format__ (__printf__, 5, 6))); + +/* If NULL, error will flush stdout, then print on stderr the program + name, a colon and a space. Otherwise, error will call this + function without parameters instead. */ +extern void (*error_print_progname) (void); + +/* This variable is incremented each time `error' is called. */ +extern unsigned int error_message_count; + +/* Sometimes we want to have at most one error per line. This + variable controls whether this mode is selected or not. */ +extern int error_one_per_line; + +#ifdef __cplusplus +} +#endif + +#endif /* error.h */ diff --git a/gl/exit.h b/gl/exit.h new file mode 100644 index 00000000..e8f77388 --- /dev/null +++ b/gl/exit.h @@ -0,0 +1,32 @@ +/* exit() function. + Copyright (C) 1995, 2001 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _EXIT_H +#define _EXIT_H + +/* Get exit() declaration. */ +#include <stdlib.h> + +/* Some systems do not define EXIT_*, despite otherwise supporting C89. */ +#ifndef EXIT_SUCCESS +# define EXIT_SUCCESS 0 +#endif +#ifndef EXIT_FAILURE +# define EXIT_FAILURE 1 +#endif + +#endif /* _EXIT_H */ diff --git a/gl/exitfail.c b/gl/exitfail.c new file mode 100644 index 00000000..97abc674 --- /dev/null +++ b/gl/exitfail.c @@ -0,0 +1,25 @@ +/* Failure exit status + + Copyright (C) 2002, 2003, 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "exitfail.h" +#include "exit.h" + +int volatile exit_failure = EXIT_FAILURE; diff --git a/gl/exitfail.h b/gl/exitfail.h new file mode 100644 index 00000000..e46cf9c1 --- /dev/null +++ b/gl/exitfail.h @@ -0,0 +1,20 @@ +/* Failure exit status + + Copyright (C) 2002 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +extern int volatile exit_failure; diff --git a/gl/fcntl--.h b/gl/fcntl--.h new file mode 100644 index 00000000..51b869e6 --- /dev/null +++ b/gl/fcntl--.h @@ -0,0 +1,28 @@ +/* Like fcntl.h, but redefine some names to avoid glitches. + + Copyright (C) 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <fcntl.h> +#include "fcntl-safer.h" + +#undef open +#define open open_safer + +#undef creat +#define creat creat_safer diff --git a/gl/fcntl-safer.h b/gl/fcntl-safer.h new file mode 100644 index 00000000..cab6aab1 --- /dev/null +++ b/gl/fcntl-safer.h @@ -0,0 +1,24 @@ +/* Invoke fcntl-like functions, but avoid some glitches. + + Copyright (C) 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <sys/types.h> + +int open_safer (char const *, int, ...); +int creat_safer (char const *, mode_t); diff --git a/gl/fd-safer.c b/gl/fd-safer.c new file mode 100644 index 00000000..256bfa4a --- /dev/null +++ b/gl/fd-safer.c @@ -0,0 +1,57 @@ +/* Return a safer copy of a file descriptor. + + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <config.h> + +#include "unistd-safer.h" + +#include <errno.h> + +#include <unistd.h> +#ifndef STDIN_FILENO +# define STDIN_FILENO 0 +#endif +#ifndef STDERR_FILENO +# define STDERR_FILENO 2 +#endif + +/* Return FD, unless FD would be a copy of standard input, output, or + error; in that case, return a duplicate of FD, closing FD. On + failure to duplicate, close FD, set errno, and return -1. Preserve + errno if FD is negative, so that the caller can always inspect + errno when the returned value is negative. + + This function is usefully wrapped around functions that return file + descriptors, e.g., fd_safer (open ("file", O_RDONLY)). */ + +int +fd_safer (int fd) +{ + if (STDIN_FILENO <= fd && fd <= STDERR_FILENO) + { + int f = dup_safer (fd); + int e = errno; + close (fd); + errno = e; + fd = f; + } + + return fd; +} diff --git a/gl/fsusage.c b/gl/fsusage.c new file mode 100644 index 00000000..337bf531 --- /dev/null +++ b/gl/fsusage.c @@ -0,0 +1,264 @@ +/* fsusage.c -- return space usage of mounted file systems + + Copyright (C) 1991, 1992, 1996, 1998, 1999, 2002, 2003, 2004, 2005, 2006 + Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "fsusage.h" + +#include <limits.h> +#include <sys/types.h> + +#if STAT_STATVFS /* POSIX 1003.1-2001 (and later) with XSI */ +# include <sys/statvfs.h> +#else +/* Don't include backward-compatibility files unless they're needed. + Eventually we'd like to remove all this cruft. */ +# include <fcntl.h> +# include <unistd.h> +# include <sys/stat.h> +# if HAVE_SYS_PARAM_H +# include <sys/param.h> +# endif +# if HAVE_SYS_MOUNT_H +# include <sys/mount.h> +# endif +# if HAVE_SYS_VFS_H +# include <sys/vfs.h> +# endif +# if HAVE_SYS_FS_S5PARAM_H /* Fujitsu UXP/V */ +# include <sys/fs/s5param.h> +# endif +# if defined HAVE_SYS_FILSYS_H && !defined _CRAY +# include <sys/filsys.h> /* SVR2 */ +# endif +# if HAVE_SYS_STATFS_H +# include <sys/statfs.h> +# endif +# if HAVE_DUSTAT_H /* AIX PS/2 */ +# include <sys/dustat.h> +# endif +# include "full-read.h" +#endif + +/* The results of open() in this file are not used with fchdir, + therefore save some unnecessary work in fchdir.c. */ +#undef open +#undef close + +/* Many space usage primitives use all 1 bits to denote a value that is + not applicable or unknown. Propagate this information by returning + a uintmax_t value that is all 1 bits if X is all 1 bits, even if X + is unsigned and narrower than uintmax_t. */ +#define PROPAGATE_ALL_ONES(x) \ + ((sizeof (x) < sizeof (uintmax_t) \ + && (~ (x) == (sizeof (x) < sizeof (int) \ + ? - (1 << (sizeof (x) * CHAR_BIT)) \ + : 0))) \ + ? UINTMAX_MAX : (uintmax_t) (x)) + +/* Extract the top bit of X as an uintmax_t value. */ +#define EXTRACT_TOP_BIT(x) ((x) \ + & ((uintmax_t) 1 << (sizeof (x) * CHAR_BIT - 1))) + +/* If a value is negative, many space usage primitives store it into an + integer variable by assignment, even if the variable's type is unsigned. + So, if a space usage variable X's top bit is set, convert X to the + uintmax_t value V such that (- (uintmax_t) V) is the negative of + the original value. If X's top bit is clear, just yield X. + Use PROPAGATE_TOP_BIT if the original value might be negative; + otherwise, use PROPAGATE_ALL_ONES. */ +#define PROPAGATE_TOP_BIT(x) ((x) | ~ (EXTRACT_TOP_BIT (x) - 1)) + +/* Fill in the fields of FSP with information about space usage for + the file system on which FILE resides. + DISK is the device on which FILE is mounted, for space-getting + methods that need to know it. + Return 0 if successful, -1 if not. When returning -1, ensure that + ERRNO is either a system error value, or zero if DISK is NULL + on a system that requires a non-NULL value. */ +int +get_fs_usage (char const *file, char const *disk, struct fs_usage *fsp) +{ +#if defined STAT_STATVFS /* POSIX */ + + struct statvfs fsd; + + if (statvfs (file, &fsd) < 0) + return -1; + + /* f_frsize isn't guaranteed to be supported. */ + fsp->fsu_blocksize = (fsd.f_frsize + ? PROPAGATE_ALL_ONES (fsd.f_frsize) + : PROPAGATE_ALL_ONES (fsd.f_bsize)); + +#elif defined STAT_STATFS2_FS_DATA /* Ultrix */ + + struct fs_data fsd; + + if (statfs (file, &fsd) != 1) + return -1; + + fsp->fsu_blocksize = 1024; + fsp->fsu_blocks = PROPAGATE_ALL_ONES (fsd.fd_req.btot); + fsp->fsu_bfree = PROPAGATE_ALL_ONES (fsd.fd_req.bfree); + fsp->fsu_bavail = PROPAGATE_TOP_BIT (fsd.fd_req.bfreen); + fsp->fsu_bavail_top_bit_set = EXTRACT_TOP_BIT (fsd.fd_req.bfreen) != 0; + fsp->fsu_files = PROPAGATE_ALL_ONES (fsd.fd_req.gtot); + fsp->fsu_ffree = PROPAGATE_ALL_ONES (fsd.fd_req.gfree); + +#elif defined STAT_READ_FILSYS /* SVR2 */ +# ifndef SUPERBOFF +# define SUPERBOFF (SUPERB * 512) +# endif + + struct filsys fsd; + int fd; + + if (! disk) + { + errno = 0; + return -1; + } + + fd = open (disk, O_RDONLY); + if (fd < 0) + return -1; + lseek (fd, (off_t) SUPERBOFF, 0); + if (full_read (fd, (char *) &fsd, sizeof fsd) != sizeof fsd) + { + close (fd); + return -1; + } + close (fd); + + fsp->fsu_blocksize = (fsd.s_type == Fs2b ? 1024 : 512); + fsp->fsu_blocks = PROPAGATE_ALL_ONES (fsd.s_fsize); + fsp->fsu_bfree = PROPAGATE_ALL_ONES (fsd.s_tfree); + fsp->fsu_bavail = PROPAGATE_TOP_BIT (fsd.s_tfree); + fsp->fsu_bavail_top_bit_set = EXTRACT_TOP_BIT (fsd.s_tfree) != 0; + fsp->fsu_files = (fsd.s_isize == -1 + ? UINTMAX_MAX + : (fsd.s_isize - 2) * INOPB * (fsd.s_type == Fs2b ? 2 : 1)); + fsp->fsu_ffree = PROPAGATE_ALL_ONES (fsd.s_tinode); + +#elif defined STAT_STATFS3_OSF1 + + struct statfs fsd; + + if (statfs (file, &fsd, sizeof (struct statfs)) != 0) + return -1; + + fsp->fsu_blocksize = PROPAGATE_ALL_ONES (fsd.f_fsize); + +#elif defined STAT_STATFS2_BSIZE /* 4.3BSD, SunOS 4, HP-UX, AIX */ + + struct statfs fsd; + + if (statfs (file, &fsd) < 0) + return -1; + + fsp->fsu_blocksize = PROPAGATE_ALL_ONES (fsd.f_bsize); + +# ifdef STATFS_TRUNCATES_BLOCK_COUNTS + + /* In SunOS 4.1.2, 4.1.3, and 4.1.3_U1, the block counts in the + struct statfs are truncated to 2GB. These conditions detect that + truncation, presumably without botching the 4.1.1 case, in which + the values are not truncated. The correct counts are stored in + undocumented spare fields. */ + if (fsd.f_blocks == 0x7fffffff / fsd.f_bsize && fsd.f_spare[0] > 0) + { + fsd.f_blocks = fsd.f_spare[0]; + fsd.f_bfree = fsd.f_spare[1]; + fsd.f_bavail = fsd.f_spare[2]; + } +# endif /* STATFS_TRUNCATES_BLOCK_COUNTS */ + +#elif defined STAT_STATFS2_FSIZE /* 4.4BSD */ + + struct statfs fsd; + + if (statfs (file, &fsd) < 0) + return -1; + + fsp->fsu_blocksize = PROPAGATE_ALL_ONES (fsd.f_fsize); + +#elif defined STAT_STATFS4 /* SVR3, Dynix, Irix, AIX */ + +# if !_AIX && !defined _SEQUENT_ && !defined DOLPHIN +# define f_bavail f_bfree +# endif + + struct statfs fsd; + + if (statfs (file, &fsd, sizeof fsd, 0) < 0) + return -1; + + /* Empirically, the block counts on most SVR3 and SVR3-derived + systems seem to always be in terms of 512-byte blocks, + no matter what value f_bsize has. */ +# if _AIX || defined _CRAY + fsp->fsu_blocksize = PROPAGATE_ALL_ONES (fsd.f_bsize); +# else + fsp->fsu_blocksize = 512; +# endif + +#endif + +#if (defined STAT_STATVFS \ + || (!defined STAT_STATFS2_FS_DATA && !defined STAT_READ_FILSYS)) + + fsp->fsu_blocks = PROPAGATE_ALL_ONES (fsd.f_blocks); + fsp->fsu_bfree = PROPAGATE_ALL_ONES (fsd.f_bfree); + fsp->fsu_bavail = PROPAGATE_TOP_BIT (fsd.f_bavail); + fsp->fsu_bavail_top_bit_set = EXTRACT_TOP_BIT (fsd.f_bavail) != 0; + fsp->fsu_files = PROPAGATE_ALL_ONES (fsd.f_files); + fsp->fsu_ffree = PROPAGATE_ALL_ONES (fsd.f_ffree); + +#endif + + return 0; +} + +#if defined _AIX && defined _I386 +/* AIX PS/2 does not supply statfs. */ + +int +statfs (char *file, struct statfs *fsb) +{ + struct stat stats; + struct dustat fsd; + + if (stat (file, &stats) != 0) + return -1; + if (dustat (stats.st_dev, 0, &fsd, sizeof (fsd))) + return -1; + fsb->f_type = 0; + fsb->f_bsize = fsd.du_bsize; + fsb->f_blocks = fsd.du_fsize - fsd.du_isize; + fsb->f_bfree = fsd.du_tfree; + fsb->f_bavail = fsd.du_tfree; + fsb->f_files = (fsd.du_isize - 2) * fsd.du_inopb; + fsb->f_ffree = fsd.du_tinode; + fsb->f_fsid.val[0] = fsd.du_site; + fsb->f_fsid.val[1] = fsd.du_pckno; + return 0; +} + +#endif /* _AIX && _I386 */ diff --git a/gl/fsusage.h b/gl/fsusage.h new file mode 100644 index 00000000..7fa9f8d6 --- /dev/null +++ b/gl/fsusage.h @@ -0,0 +1,41 @@ +/* fsusage.h -- declarations for file system space usage info + + Copyright (C) 1991, 1992, 1997, 2003, 2004, 2005, 2006 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Space usage statistics for a file system. Blocks are 512-byte. */ + +#if !defined FSUSAGE_H_ +# define FSUSAGE_H_ + +# include <stdint.h> +# include <stdbool.h> + +struct fs_usage +{ + uintmax_t fsu_blocksize; /* Size of a block. */ + uintmax_t fsu_blocks; /* Total blocks. */ + uintmax_t fsu_bfree; /* Free blocks available to superuser. */ + uintmax_t fsu_bavail; /* Free blocks available to non-superuser. */ + bool fsu_bavail_top_bit_set; /* 1 if fsu_bavail represents a value < 0. */ + uintmax_t fsu_files; /* Total file nodes. */ + uintmax_t fsu_ffree; /* Free file nodes. */ +}; + +int get_fs_usage (char const *file, char const *disk, struct fs_usage *fsp); + +#endif diff --git a/gl/full-read.c b/gl/full-read.c new file mode 100644 index 00000000..8c3472a4 --- /dev/null +++ b/gl/full-read.c @@ -0,0 +1,19 @@ +/* An interface to read that retries after partial reads and interrupts. + Copyright (C) 2002, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#define FULL_READ +#include "full-write.c" diff --git a/gl/full-read.h b/gl/full-read.h new file mode 100644 index 00000000..05d83a76 --- /dev/null +++ b/gl/full-read.h @@ -0,0 +1,24 @@ +/* An interface to read() that reads all it is asked to read. + + Copyright (C) 2002 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, read to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <stddef.h> + +/* Read COUNT bytes at BUF to descriptor FD, retrying if interrupted + or if partial reads occur. Return the number of bytes successfully + read, setting errno if that is less than COUNT. errno = 0 means EOF. */ +extern size_t full_read (int fd, void *buf, size_t count); diff --git a/gl/full-write.c b/gl/full-write.c new file mode 100644 index 00000000..cc168720 --- /dev/null +++ b/gl/full-write.c @@ -0,0 +1,81 @@ +/* An interface to read and write that retries (if necessary) until complete. + + Copyright (C) 1993, 1994, 1997, 1998, 1999, 2000, 2001, 2002, 2003, + 2004, 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#ifdef FULL_READ +# include "full-read.h" +#else +# include "full-write.h" +#endif + +#include <errno.h> + +#ifdef FULL_READ +# include "safe-read.h" +# define safe_rw safe_read +# define full_rw full_read +# undef const +# define const /* empty */ +#else +# include "safe-write.h" +# define safe_rw safe_write +# define full_rw full_write +#endif + +#ifdef FULL_READ +/* Set errno to zero upon EOF. */ +# define ZERO_BYTE_TRANSFER_ERRNO 0 +#else +/* Some buggy drivers return 0 when one tries to write beyond + a device's end. (Example: Linux 1.2.13 on /dev/fd0.) + Set errno to ENOSPC so they get a sensible diagnostic. */ +# define ZERO_BYTE_TRANSFER_ERRNO ENOSPC +#endif + +/* Write(read) COUNT bytes at BUF to(from) descriptor FD, retrying if + interrupted or if a partial write(read) occurs. Return the number + of bytes transferred. + When writing, set errno if fewer than COUNT bytes are written. + When reading, if fewer than COUNT bytes are read, you must examine + errno to distinguish failure from EOF (errno == 0). */ +size_t +full_rw (int fd, const void *buf, size_t count) +{ + size_t total = 0; + const char *ptr = (const char *) buf; + + while (count > 0) + { + size_t n_rw = safe_rw (fd, ptr, count); + if (n_rw == (size_t) -1) + break; + if (n_rw == 0) + { + errno = ZERO_BYTE_TRANSFER_ERRNO; + break; + } + total += n_rw; + ptr += n_rw; + count -= n_rw; + } + + return total; +} diff --git a/gl/full-write.h b/gl/full-write.h new file mode 100644 index 00000000..d20d2fe4 --- /dev/null +++ b/gl/full-write.h @@ -0,0 +1,35 @@ +/* An interface to write() that writes all it is asked to write. + + Copyright (C) 2002-2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <stddef.h> + + +#ifdef __cplusplus +extern "C" { +#endif + + +/* Write COUNT bytes at BUF to descriptor FD, retrying if interrupted + or if partial writes occur. Return the number of bytes successfully + written, setting errno if that is less than COUNT. */ +extern size_t full_write (int fd, const void *buf, size_t count); + + +#ifdef __cplusplus +} +#endif diff --git a/gl/gai_strerror.c b/gl/gai_strerror.c new file mode 100644 index 00000000..7f0e034e --- /dev/null +++ b/gl/gai_strerror.c @@ -0,0 +1,77 @@ +/* Copyright (C) 1997, 2001, 2002, 2004, 2005, 2006 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Philip Blundell <pjb27@cam.ac.uk>, 1997. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _LIBC +# include <config.h> +# include "getaddrinfo.h" +#endif + +#include <stdio.h> +#ifdef HAVE_NETDB_H +# include <netdb.h> +#endif + +#ifdef _LIBC +# include <libintl.h> +#else +# include "gettext.h" +# define _(String) gettext (String) +# define N_(String) String +#endif + +static struct + { + int code; + const char *msg; + } +values[] = + { + { EAI_ADDRFAMILY, N_("Address family for hostname not supported") }, + { EAI_AGAIN, N_("Temporary failure in name resolution") }, + { EAI_BADFLAGS, N_("Bad value for ai_flags") }, + { EAI_FAIL, N_("Non-recoverable failure in name resolution") }, + { EAI_FAMILY, N_("ai_family not supported") }, + { EAI_MEMORY, N_("Memory allocation failure") }, + { EAI_NODATA, N_("No address associated with hostname") }, + { EAI_NONAME, N_("Name or service not known") }, + { EAI_SERVICE, N_("Servname not supported for ai_socktype") }, + { EAI_SOCKTYPE, N_("ai_socktype not supported") }, + { EAI_SYSTEM, N_("System error") }, +#ifdef __USE_GNU + { EAI_INPROGRESS, N_("Processing request in progress") }, + { EAI_CANCELED, N_("Request canceled") }, + { EAI_NOTCANCELED, N_("Request not canceled") }, + { EAI_ALLDONE, N_("All requests done") }, + { EAI_INTR, N_("Interrupted by a signal") }, + { EAI_IDN_ENCODE, N_("Parameter string not correctly encoded") } +#endif + }; + +const char * +gai_strerror (int code) +{ + size_t i; + for (i = 0; i < sizeof (values) / sizeof (values[0]); ++i) + if (values[i].code == code) + return _(values[i].msg); + + return _("Unknown error"); +} +#ifdef _LIBC +libc_hidden_def (gai_strerror) +#endif diff --git a/gl/getaddrinfo.c b/gl/getaddrinfo.c new file mode 100644 index 00000000..f523f765 --- /dev/null +++ b/gl/getaddrinfo.c @@ -0,0 +1,417 @@ +/* Get address information (partial implementation). + Copyright (C) 1997, 2001, 2002, 2004, 2005, 2006 Free Software + Foundation, Inc. + Contributed by Simon Josefsson <simon@josefsson.org>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "getaddrinfo.h" + +#if HAVE_NETINET_IN_H +# include <netinet/in.h> +#endif + +/* Get calloc. */ +#include <stdlib.h> + +/* Get memcpy. */ +#include <string.h> + +#include <stdbool.h> + +#include "gettext.h" +#define _(String) gettext (String) +#define N_(String) String + +#include "inet_ntop.h" +#include "snprintf.h" +#include "strdup.h" + +/* BeOS has AF_INET, but not PF_INET. */ +#ifndef PF_INET +# define PF_INET AF_INET +#endif +/* BeOS also lacks PF_UNSPEC. */ +#ifndef PF_UNSPEC +# define PF_UNSPEC 0 +#endif + +#if defined _WIN32 || defined __WIN32__ +# define WIN32_NATIVE +#endif + +#ifdef WIN32_NATIVE +typedef int (WSAAPI *getaddrinfo_func) (const char*, const char*, + const struct addrinfo*, + struct addrinfo**); +typedef void (WSAAPI *freeaddrinfo_func) (struct addrinfo*); +typedef int (WSAAPI *getnameinfo_func) (const struct sockaddr*, + socklen_t, char*, DWORD, + char*, DWORD, int); + +static getaddrinfo_func getaddrinfo_ptr = NULL; +static freeaddrinfo_func freeaddrinfo_ptr = NULL; +static getnameinfo_func getnameinfo_ptr = NULL; + +static int +use_win32_p (void) +{ + static int done = 0; + HMODULE h; + + if (done) + return getaddrinfo_ptr ? 1 : 0; + + done = 1; + + h = GetModuleHandle ("ws2_32.dll"); + + if (h) + { + getaddrinfo_ptr = (getaddrinfo_func) GetProcAddress (h, "getaddrinfo"); + freeaddrinfo_ptr = (freeaddrinfo_func) GetProcAddress (h, "freeaddrinfo"); + getnameinfo_ptr = (getnameinfo_func) GetProcAddress (h, "getnameinfo"); + } + + /* If either is missing, something is odd. */ + if (!getaddrinfo_ptr || !freeaddrinfo_ptr || !getnameinfo_ptr) + { + getaddrinfo_ptr = NULL; + freeaddrinfo_ptr = NULL; + getnameinfo_ptr = NULL; + return 0; + } + + return 1; +} +#endif + +static inline bool +validate_family (int family) +{ + /* FIXME: Support more families. */ +#if HAVE_IPV4 + if (family == PF_INET) + return true; +#endif +#if HAVE_IPV6 + if (family == PF_INET6) + return true; +#endif + if (family == PF_UNSPEC) + return true; + return false; +} + +/* Translate name of a service location and/or a service name to set of + socket addresses. */ +int +getaddrinfo (const char *restrict nodename, + const char *restrict servname, + const struct addrinfo *restrict hints, + struct addrinfo **restrict res) +{ + struct addrinfo *tmp; + int port = 0; + struct hostent *he; + void *storage; + size_t size; +#if HAVE_IPV6 + struct v6_pair { + struct addrinfo addrinfo; + struct sockaddr_in6 sockaddr_in6; + }; +#endif +#if HAVE_IPV4 + struct v4_pair { + struct addrinfo addrinfo; + struct sockaddr_in sockaddr_in; + }; +#endif + +#ifdef WIN32_NATIVE + if (use_win32_p ()) + return getaddrinfo_ptr (nodename, servname, hints, res); +#endif + + if (hints && (hints->ai_flags & ~(AI_CANONNAME|AI_PASSIVE))) + /* FIXME: Support more flags. */ + return EAI_BADFLAGS; + + if (hints && !validate_family (hints->ai_family)) + return EAI_FAMILY; + + if (hints && + hints->ai_socktype != SOCK_STREAM && hints->ai_socktype != SOCK_DGRAM) + /* FIXME: Support other socktype. */ + return EAI_SOCKTYPE; /* FIXME: Better return code? */ + + if (!nodename) + { + if (!(hints->ai_flags & AI_PASSIVE)) + return EAI_NONAME; + +#ifdef HAVE_IPV6 + nodename = (hints->ai_family == AF_INET6) ? "::" : "0.0.0.0"; +#else + nodename = "0.0.0.0"; +#endif + } + + if (servname) + { + struct servent *se = NULL; + const char *proto = + (hints && hints->ai_socktype == SOCK_DGRAM) ? "udp" : "tcp"; + + if (!(hints->ai_flags & AI_NUMERICSERV)) + /* FIXME: Use getservbyname_r if available. */ + se = getservbyname (servname, proto); + + if (!se) + { + char *c; + if (!(*servname >= '0' && *servname <= '9')) + return EAI_NONAME; + port = strtoul (servname, &c, 10); + if (*c || port > 0xffff) + return EAI_NONAME; + port = htons (port); + } + else + port = se->s_port; + } + + /* FIXME: Use gethostbyname_r if available. */ + he = gethostbyname (nodename); + if (!he || he->h_addr_list[0] == NULL) + return EAI_NONAME; + + switch (he->h_addrtype) + { +#if HAVE_IPV6 + case PF_INET6: + size = sizeof (struct v6_pair); + break; +#endif + +#if HAVE_IPV4 + case PF_INET: + size = sizeof (struct v4_pair); + break; +#endif + + default: + return EAI_NODATA; + } + + storage = calloc (1, size); + if (!storage) + return EAI_MEMORY; + + switch (he->h_addrtype) + { +#if HAVE_IPV6 + case PF_INET6: + { + struct v6_pair *p = storage; + struct sockaddr_in6 *sinp = &p->sockaddr_in6; + tmp = &p->addrinfo; + + if (port) + sinp->sin6_port = port; + + if (he->h_length != sizeof (sinp->sin6_addr)) + { + free (storage); + return EAI_SYSTEM; /* FIXME: Better return code? Set errno? */ + } + + memcpy (&sinp->sin6_addr, he->h_addr_list[0], sizeof sinp->sin6_addr); + + tmp->ai_addr = (struct sockaddr *) sinp; + tmp->ai_addrlen = sizeof *sinp; + } + break; +#endif + +#if HAVE_IPV4 + case PF_INET: + { + struct v4_pair *p = storage; + struct sockaddr_in *sinp = &p->sockaddr_in; + tmp = &p->addrinfo; + + if (port) + sinp->sin_port = port; + + if (he->h_length != sizeof (sinp->sin_addr)) + { + free (storage); + return EAI_SYSTEM; /* FIXME: Better return code? Set errno? */ + } + + memcpy (&sinp->sin_addr, he->h_addr_list[0], sizeof sinp->sin_addr); + + tmp->ai_addr = (struct sockaddr *) sinp; + tmp->ai_addrlen = sizeof *sinp; + } + break; +#endif + + default: + free (storage); + return EAI_NODATA; + } + + if (hints && hints->ai_flags & AI_CANONNAME) + { + const char *cn; + if (he->h_name) + cn = he->h_name; + else + cn = nodename; + + tmp->ai_canonname = strdup (cn); + if (!tmp->ai_canonname) + { + free (storage); + return EAI_MEMORY; + } + } + + tmp->ai_protocol = (hints) ? hints->ai_protocol : 0; + tmp->ai_socktype = (hints) ? hints->ai_socktype : 0; + tmp->ai_addr->sa_family = he->h_addrtype; + tmp->ai_family = he->h_addrtype; + + /* FIXME: If more than one address, create linked list of addrinfo's. */ + + *res = tmp; + + return 0; +} + +/* Free `addrinfo' structure AI including associated storage. */ +void +freeaddrinfo (struct addrinfo *ai) +{ +#ifdef WIN32_NATIVE + if (use_win32_p ()) + { + freeaddrinfo_ptr (ai); + return; + } +#endif + + while (ai) + { + struct addrinfo *cur; + + cur = ai; + ai = ai->ai_next; + + if (cur->ai_canonname) free (cur->ai_canonname); + free (cur); + } +} + +int getnameinfo(const struct sockaddr *restrict sa, socklen_t salen, + char *restrict node, socklen_t nodelen, + char *restrict service, socklen_t servicelen, + int flags) +{ +#ifdef WIN32_NATIVE + if (use_win32_p ()) + return getnameinfo_ptr (sa, salen, node, nodelen, + service, servicelen, flags); +#endif + + /* FIXME: Support other flags. */ + if ((node && nodelen > 0 && !(flags & NI_NUMERICHOST)) || + (service && servicelen > 0 && !(flags & NI_NUMERICHOST)) || + (flags & ~(NI_NUMERICHOST|NI_NUMERICSERV))) + return EAI_BADFLAGS; + + if (sa == NULL || salen < sizeof (sa->sa_family)) + return EAI_FAMILY; + + switch (sa->sa_family) + { +#if HAVE_IPV4 + case AF_INET: + if (salen < sizeof (struct sockaddr_in)) + return EAI_FAMILY; + break; +#endif +#if HAVE_IPV6 + case AF_INET6: + if (salen < sizeof (struct sockaddr_in6)) + return EAI_FAMILY; + break; +#endif + default: + return EAI_FAMILY; + } + + if (node && nodelen > 0 && flags & NI_NUMERICHOST) + { + switch (sa->sa_family) + { +#if HAVE_IPV4 + case AF_INET: + if (!inet_ntop (AF_INET, + &(((const struct sockaddr_in *) sa)->sin_addr), + node, nodelen)) + return EAI_SYSTEM; + break; +#endif + +#if HAVE_IPV6 + case AF_INET6: + if (!inet_ntop (AF_INET6, + &(((const struct sockaddr_in6 *) sa)->sin6_addr), + node, nodelen)) + return EAI_SYSTEM; + break; +#endif + + default: + return EAI_FAMILY; + } + } + + if (service && servicelen > 0 && flags & NI_NUMERICSERV) + switch (sa->sa_family) + { +#if HAVE_IPV4 + case AF_INET: +#endif +#if HAVE_IPV6 + case AF_INET6: +#endif + { + unsigned short int port + = ntohs (((const struct sockaddr_in *) sa)->sin_port); + if (servicelen <= snprintf (service, servicelen, "%u", port)) + return EAI_OVERFLOW; + } + break; + } + + return 0; +} diff --git a/gl/getaddrinfo.h b/gl/getaddrinfo.h new file mode 100644 index 00000000..b4ef242c --- /dev/null +++ b/gl/getaddrinfo.h @@ -0,0 +1,155 @@ +/* Get address information. + Copyright (C) 1996-2002, 2003, 2004, 2005, 2006 + Free Software Foundation, Inc. + Contributed by Simon Josefsson <simon@josefsson.org>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef GETADDRINFO_H +#define GETADDRINFO_H + +/* sys/socket.h in i386-unknown-freebsd4.10 and + powerpc-apple-darwin5.5 require sys/types.h, so include it first. + Then we'll also get 'socklen_t' and 'struct sockaddr' which are + used below. */ +#include <sys/types.h> +/* Get all getaddrinfo related declarations, if available. */ +#include <sys/socket.h> +#ifdef HAVE_NETDB_H +# include <netdb.h> +#endif + +#ifndef HAVE_STRUCT_ADDRINFO + +/* Structure to contain information about address of a service provider. */ +struct addrinfo +{ + int ai_flags; /* Input flags. */ + int ai_family; /* Protocol family for socket. */ + int ai_socktype; /* Socket type. */ + int ai_protocol; /* Protocol for socket. */ + socklen_t ai_addrlen; /* Length of socket address. */ + struct sockaddr *ai_addr; /* Socket address for socket. */ + char *ai_canonname; /* Canonical name for service location. */ + struct addrinfo *ai_next; /* Pointer to next in list. */ +}; +#endif + +/* Possible values for `ai_flags' field in `addrinfo' structure. */ +#ifndef AI_PASSIVE +# define AI_PASSIVE 0x0001 /* Socket address is intended for `bind'. */ +#endif +#ifndef AI_CANONNAME +# define AI_CANONNAME 0x0002 /* Request for canonical name. */ +#endif +#ifndef AI_NUMERICSERV +# define AI_NUMERICSERV 0x0400 /* Don't use name resolution. */ +#endif + +#if 0 +/* The commented out definitions below are not yet implemented in the + GNULIB getaddrinfo() replacement, so are not yet needed and may, in fact, + cause conflicts on systems with a getaddrinfo() function which does not + define them. + + If they are restored, be sure to protect the definitions with #ifndef. */ +#define AI_NUMERICHOST 0x0004 /* Don't use name resolution. */ +#define AI_V4MAPPED 0x0008 /* IPv4 mapped addresses are acceptable. */ +#define AI_ALL 0x0010 /* Return IPv4 mapped and IPv6 addresses. */ +#define AI_ADDRCONFIG 0x0020 /* Use configuration of this host to choose + returned address type.. */ +#endif /* 0 */ + +/* Error values for `getaddrinfo' function. */ +#ifndef EAI_BADFLAGS +# define EAI_BADFLAGS -1 /* Invalid value for `ai_flags' field. */ +# define EAI_NONAME -2 /* NAME or SERVICE is unknown. */ +# define EAI_AGAIN -3 /* Temporary failure in name resolution. */ +# define EAI_FAIL -4 /* Non-recoverable failure in name res. */ +# define EAI_NODATA -5 /* No address associated with NAME. */ +# define EAI_FAMILY -6 /* `ai_family' not supported. */ +# define EAI_SOCKTYPE -7 /* `ai_socktype' not supported. */ +# define EAI_SERVICE -8 /* SERVICE not supported for `ai_socktype'. */ +# define EAI_MEMORY -10 /* Memory allocation failure. */ +#endif +#ifndef EAI_OVERFLOW +/* Not defined on mingw32. */ +# define EAI_OVERFLOW -12 /* Argument buffer overflow. */ +#endif +#ifndef EAI_ADDRFAMILY +/* Not defined on mingw32. */ +# define EAI_ADDRFAMILY -9 /* Address family for NAME not supported. */ +#endif +#ifndef EAI_SYSTEM +/* Not defined on mingw32. */ +# define EAI_SYSTEM -11 /* System error returned in `errno'. */ +#endif + +#ifdef __USE_GNU +# ifndef EAI_INPROGRESS +# define EAI_INPROGRESS -100 /* Processing request in progress. */ +# define EAI_CANCELED -101 /* Request canceled. */ +# define EAI_NOTCANCELED -102 /* Request not canceled. */ +# define EAI_ALLDONE -103 /* All requests done. */ +# define EAI_INTR -104 /* Interrupted by a signal. */ +# define EAI_IDN_ENCODE -105 /* IDN encoding failed. */ +# endif +#endif + +#if !HAVE_DECL_GETADDRINFO +/* Translate name of a service location and/or a service name to set of + socket addresses. + For more details, see the POSIX:2001 specification + <http://www.opengroup.org/susv3xsh/getaddrinfo.html>. */ +extern int getaddrinfo (const char *restrict nodename, + const char *restrict servname, + const struct addrinfo *restrict hints, + struct addrinfo **restrict res); +#endif + +#if !HAVE_DECL_FREEADDRINFO +/* Free `addrinfo' structure AI including associated storage. + For more details, see the POSIX:2001 specification + <http://www.opengroup.org/susv3xsh/getaddrinfo.html>. */ +extern void freeaddrinfo (struct addrinfo *ai); +#endif + +#if !HAVE_DECL_GAI_STRERROR +/* Convert error return from getaddrinfo() to a string. + For more details, see the POSIX:2001 specification + <http://www.opengroup.org/susv3xsh/gai_strerror.html>. */ +extern const char *gai_strerror (int ecode); +#endif + +#if !HAVE_DECL_GETNAMEINFO +/* Convert socket address to printable node and service names. + For more details, see the POSIX:2001 specification + <http://www.opengroup.org/susv3xsh/getnameinfo.html>. */ +extern int getnameinfo(const struct sockaddr *restrict sa, socklen_t salen, + char *restrict node, socklen_t nodelen, + char *restrict service, socklen_t servicelen, + int flags); + +#endif + +/* Possible flags for getnameinfo. */ +#ifndef NI_NUMERICHOST +# define NI_NUMERICHOST 1 +#endif +#ifndef NI_NUMERICSERV +# define NI_NUMERICSERV 2 +#endif + +#endif /* GETADDRINFO_H */ diff --git a/gl/gethostname.c b/gl/gethostname.c new file mode 100644 index 00000000..eedc40ec --- /dev/null +++ b/gl/gethostname.c @@ -0,0 +1,52 @@ +/* gethostname emulation for SysV and POSIX.1. + + Copyright (C) 1992, 2003, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* David MacKenzie <djm@gnu.ai.mit.edu> */ + +#include <config.h> + +#ifdef HAVE_UNAME +# include <sys/utsname.h> +#endif + +/* Put up to LEN chars of the host name into NAME. + Null terminate it if the name is shorter than LEN. + Return 0 if ok, -1 if error. */ + +#include <stddef.h> + +int +gethostname (char *name, size_t len) +{ +#ifdef HAVE_UNAME + struct utsname uts; + + if (uname (&uts) == -1) + return -1; + if (len > sizeof (uts.nodename)) + { + /* More space than we need is available. */ + name[sizeof (uts.nodename)] = '\0'; + len = sizeof (uts.nodename); + } + strncpy (name, uts.nodename, len); +#else + strcpy (name, ""); /* Hardcode your system name if you want. */ +#endif + return 0; +} diff --git a/gl/getloadavg.c b/gl/getloadavg.c new file mode 100644 index 00000000..cfa62735 --- /dev/null +++ b/gl/getloadavg.c @@ -0,0 +1,1020 @@ +/* Get the system load averages. + + Copyright (C) 1985, 1986, 1987, 1988, 1989, 1991, 1992, 1993, 1994, + 1995, 1997, 1999, 2000, 2003, 2004, 2005, 2006 Free Software + Foundation, Inc. + + NOTE: The canonical source of this file is maintained with gnulib. + Bugs can be reported to bug-gnulib@gnu.org. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, + USA. */ + +/* Compile-time symbols that this file uses: + + HAVE_PSTAT_GETDYNAMIC Define this if your system has the + pstat_getdynamic function. I think it + is unique to HPUX9. The best way to get the + definition is through the AC_FUNC_GETLOADAVG + macro that comes with autoconf 2.13 or newer. + If that isn't an option, then just put + AC_CHECK_FUNCS(pstat_getdynamic) in your + configure.in file. + FIXUP_KERNEL_SYMBOL_ADDR() Adjust address in returned struct nlist. + KERNEL_FILE Name of the kernel file to nlist. + LDAV_CVT() Scale the load average from the kernel. + Returns a double. + LDAV_SYMBOL Name of kernel symbol giving load average. + LOAD_AVE_TYPE Type of the load average array in the kernel. + Must be defined unless one of + apollo, DGUX, NeXT, or UMAX is defined; + or we have libkstat; + otherwise, no load average is available. + HAVE_NLIST_H nlist.h is available. NLIST_STRUCT defaults + to this. + NLIST_STRUCT Include nlist.h, not a.out.h, and + the nlist n_name element is a pointer, + not an array. + HAVE_STRUCT_NLIST_N_UN_N_NAME `n_un.n_name' is member of `struct nlist'. + LINUX_LDAV_FILE [__linux__, __CYGWIN__]: File containing + load averages. + + Specific system predefines this file uses, aside from setting + default values if not emacs: + + apollo + BSD Real BSD, not just BSD-like. + convex + DGUX + eunice UNIX emulator under VMS. + hpux + __MSDOS__ No-op for MSDOS. + NeXT + sgi + sequent Sequent Dynix 3.x.x (BSD) + _SEQUENT_ Sequent DYNIX/ptx 1.x.x (SYSV) + sony_news NEWS-OS (works at least for 4.1C) + UMAX + UMAX4_3 + VMS + WINDOWS32 No-op for Windows95/NT. + __linux__ Linux: assumes /proc file system mounted. + Support from Michael K. Johnson. + __CYGWIN__ Cygwin emulates linux /proc/loadavg. + __NetBSD__ NetBSD: assumes /kern file system mounted. + + In addition, to avoid nesting many #ifdefs, we internally set + LDAV_DONE to indicate that the load average has been computed. + + We also #define LDAV_PRIVILEGED if a program will require + special installation to be able to call getloadavg. */ + +/* "configure" defines CONFIGURING_GETLOADAVG to sidestep problems + with partially-configured source directories. */ + +#ifndef CONFIGURING_GETLOADAVG +# include <config.h> +# include <stdbool.h> +#endif + +#include <errno.h> +#include <stdio.h> +#include <stdlib.h> + +/* Exclude all the code except the test program at the end + if the system has its own `getloadavg' function. */ + +#ifndef HAVE_GETLOADAVG + +# include <sys/types.h> + +/* Both the Emacs and non-Emacs sections want this. Some + configuration files' definitions for the LOAD_AVE_CVT macro (like + sparc.h's) use macros like FSCALE, defined here. */ +# if defined (unix) || defined (__unix) +# include <sys/param.h> +# endif + +# include "c-strtod.h" +# include "cloexec.h" +# include "intprops.h" +# include "xalloc.h" + +/* The existing Emacs configuration files define a macro called + LOAD_AVE_CVT, which accepts a value of type LOAD_AVE_TYPE, and + returns the load average multiplied by 100. What we actually want + is a macro called LDAV_CVT, which returns the load average as an + unmultiplied double. + + For backwards compatibility, we'll define LDAV_CVT in terms of + LOAD_AVE_CVT, but future machine config files should just define + LDAV_CVT directly. */ + +# if !defined (LDAV_CVT) && defined (LOAD_AVE_CVT) +# define LDAV_CVT(n) (LOAD_AVE_CVT (n) / 100.0) +# endif + +# if !defined (BSD) && defined (ultrix) +/* Ultrix behaves like BSD on Vaxen. */ +# define BSD +# endif + +# ifdef NeXT +/* NeXT in the 2.{0,1,2} releases defines BSD in <sys/param.h>, which + conflicts with the definition understood in this file, that this + really is BSD. */ +# undef BSD + +/* NeXT defines FSCALE in <sys/param.h>. However, we take FSCALE being + defined to mean that the nlist method should be used, which is not true. */ +# undef FSCALE +# endif + +/* Same issues as for NeXT apply to the HURD-based GNU system. */ +# ifdef __GNU__ +# undef BSD +# undef FSCALE +# endif /* __GNU__ */ + +/* Set values that are different from the defaults, which are + set a little farther down with #ifndef. */ + + +/* Some shorthands. */ + +# if defined (HPUX) && !defined (hpux) +# define hpux +# endif + +# if defined (__hpux) && !defined (hpux) +# define hpux +# endif + +# if defined (__sun) && !defined (sun) +# define sun +# endif + +# if defined (hp300) && !defined (hpux) +# define MORE_BSD +# endif + +# if defined (ultrix) && defined (mips) +# define decstation +# endif + +# if defined (__SVR4) && !defined (SVR4) +# define SVR4 +# endif + +# if (defined (sun) && defined (SVR4)) || defined (SOLARIS2) +# define SUNOS_5 +# endif + +# if defined (__osf__) && (defined (__alpha) || defined (__alpha__)) +# define OSF_ALPHA +# include <sys/mbuf.h> +# include <sys/socket.h> +# include <net/route.h> +# include <sys/table.h> +# endif + +# if defined (__osf__) && (defined (mips) || defined (__mips__)) +# define OSF_MIPS +# include <sys/table.h> +# endif + +/* UTek's /bin/cc on the 4300 has no architecture specific cpp define by + default, but _MACH_IND_SYS_TYPES is defined in <sys/types.h>. Combine + that with a couple of other things and we'll have a unique match. */ +# if !defined (tek4300) && defined (unix) && defined (m68k) && defined (mc68000) && defined (mc68020) && defined (_MACH_IND_SYS_TYPES) +# define tek4300 /* Define by emacs, but not by other users. */ +# endif + + +/* VAX C can't handle multi-line #ifs, or lines longer than 256 chars. */ +# ifndef LOAD_AVE_TYPE + +# ifdef MORE_BSD +# define LOAD_AVE_TYPE long +# endif + +# ifdef sun +# define LOAD_AVE_TYPE long +# endif + +# ifdef decstation +# define LOAD_AVE_TYPE long +# endif + +# ifdef _SEQUENT_ +# define LOAD_AVE_TYPE long +# endif + +# ifdef sgi +# define LOAD_AVE_TYPE long +# endif + +# ifdef SVR4 +# define LOAD_AVE_TYPE long +# endif + +# ifdef sony_news +# define LOAD_AVE_TYPE long +# endif + +# ifdef sequent +# define LOAD_AVE_TYPE long +# endif + +# ifdef OSF_ALPHA +# define LOAD_AVE_TYPE long +# endif + +# if defined (ardent) && defined (titan) +# define LOAD_AVE_TYPE long +# endif + +# ifdef tek4300 +# define LOAD_AVE_TYPE long +# endif + +# if defined (alliant) && defined (i860) /* Alliant FX/2800 */ +# define LOAD_AVE_TYPE long +# endif + +# ifdef _AIX +# define LOAD_AVE_TYPE long +# endif + +# ifdef convex +# define LOAD_AVE_TYPE double +# ifndef LDAV_CVT +# define LDAV_CVT(n) (n) +# endif +# endif + +# endif /* No LOAD_AVE_TYPE. */ + +# ifdef OSF_ALPHA +/* <sys/param.h> defines an incorrect value for FSCALE on Alpha OSF/1, + according to ghazi@noc.rutgers.edu. */ +# undef FSCALE +# define FSCALE 1024.0 +# endif + +# if defined (alliant) && defined (i860) /* Alliant FX/2800 */ +/* <sys/param.h> defines an incorrect value for FSCALE on an + Alliant FX/2800 Concentrix 2.2, according to ghazi@noc.rutgers.edu. */ +# undef FSCALE +# define FSCALE 100.0 +# endif + + +# ifndef FSCALE + +/* SunOS and some others define FSCALE in sys/param.h. */ + +# ifdef MORE_BSD +# define FSCALE 2048.0 +# endif + +# if defined (MIPS) || defined (SVR4) || defined (decstation) +# define FSCALE 256 +# endif + +# if defined (sgi) || defined (sequent) +/* Sometimes both MIPS and sgi are defined, so FSCALE was just defined + above under #ifdef MIPS. But we want the sgi value. */ +# undef FSCALE +# define FSCALE 1000.0 +# endif + +# if defined (ardent) && defined (titan) +# define FSCALE 65536.0 +# endif + +# ifdef tek4300 +# define FSCALE 100.0 +# endif + +# ifdef _AIX +# define FSCALE 65536.0 +# endif + +# endif /* Not FSCALE. */ + +# if !defined (LDAV_CVT) && defined (FSCALE) +# define LDAV_CVT(n) (((double) (n)) / FSCALE) +# endif + +# ifndef NLIST_STRUCT +# if HAVE_NLIST_H +# define NLIST_STRUCT +# endif +# endif + +# if defined (sgi) || (defined (mips) && !defined (BSD)) +# define FIXUP_KERNEL_SYMBOL_ADDR(nl) ((nl)[0].n_value &= ~(1 << 31)) +# endif + + +# if !defined (KERNEL_FILE) && defined (sequent) +# define KERNEL_FILE "/dynix" +# endif + +# if !defined (KERNEL_FILE) && defined (hpux) +# define KERNEL_FILE "/hp-ux" +# endif + +# if !defined (KERNEL_FILE) && (defined (_SEQUENT_) || defined (MIPS) || defined (SVR4) || defined (ISC) || defined (sgi) || (defined (ardent) && defined (titan))) +# define KERNEL_FILE "/unix" +# endif + + +# if !defined (LDAV_SYMBOL) && defined (alliant) +# define LDAV_SYMBOL "_Loadavg" +# endif + +# if !defined (LDAV_SYMBOL) && ((defined (hpux) && !defined (hp9000s300)) || defined (_SEQUENT_) || defined (SVR4) || defined (ISC) || defined (sgi) || (defined (ardent) && defined (titan)) || defined (_AIX)) +# define LDAV_SYMBOL "avenrun" +# endif + +# include <unistd.h> + +/* LOAD_AVE_TYPE should only get defined if we're going to use the + nlist method. */ +# if !defined (LOAD_AVE_TYPE) && (defined (BSD) || defined (LDAV_CVT) || defined (KERNEL_FILE) || defined (LDAV_SYMBOL)) +# define LOAD_AVE_TYPE double +# endif + +# ifdef LOAD_AVE_TYPE + +# ifndef __VMS +# ifndef __linux__ +# ifndef NLIST_STRUCT +# include <a.out.h> +# else /* NLIST_STRUCT */ +# include <nlist.h> +# endif /* NLIST_STRUCT */ + +# ifdef SUNOS_5 +# include <fcntl.h> +# include <kvm.h> +# include <kstat.h> +# endif + +# if defined (hpux) && defined (HAVE_PSTAT_GETDYNAMIC) +# include <sys/pstat.h> +# endif + +# ifndef KERNEL_FILE +# define KERNEL_FILE "/vmunix" +# endif /* KERNEL_FILE */ + +# ifndef LDAV_SYMBOL +# define LDAV_SYMBOL "_avenrun" +# endif /* LDAV_SYMBOL */ +# endif /* __linux__ */ + +# else /* __VMS */ + +# ifndef eunice +# include <iodef.h> +# include <descrip.h> +# else /* eunice */ +# include <vms/iodef.h> +# endif /* eunice */ +# endif /* __VMS */ + +# ifndef LDAV_CVT +# define LDAV_CVT(n) ((double) (n)) +# endif /* !LDAV_CVT */ + +# endif /* LOAD_AVE_TYPE */ + +# if defined (__GNU__) && !defined (NeXT) +/* Note that NeXT Openstep defines __GNU__ even though it should not. */ +/* GNU system acts much like NeXT, for load average purposes, + but not exactly. */ +# define NeXT +# define host_self mach_host_self +# endif + +# ifdef NeXT +# ifdef HAVE_MACH_MACH_H +# include <mach/mach.h> +# else +# include <mach.h> +# endif +# endif /* NeXT */ + +# ifdef sgi +# include <sys/sysmp.h> +# endif /* sgi */ + +# ifdef UMAX +# include <signal.h> +# include <sys/time.h> +# include <sys/wait.h> +# include <sys/syscall.h> + +# ifdef UMAX_43 +# include <machine/cpu.h> +# include <inq_stats/statistics.h> +# include <inq_stats/sysstats.h> +# include <inq_stats/cpustats.h> +# include <inq_stats/procstats.h> +# else /* Not UMAX_43. */ +# include <sys/sysdefs.h> +# include <sys/statistics.h> +# include <sys/sysstats.h> +# include <sys/cpudefs.h> +# include <sys/cpustats.h> +# include <sys/procstats.h> +# endif /* Not UMAX_43. */ +# endif /* UMAX */ + +# ifdef DGUX +# include <sys/dg_sys_info.h> +# endif + +# include "fcntl--.h" + +/* Avoid static vars inside a function since in HPUX they dump as pure. */ + +# ifdef NeXT +static processor_set_t default_set; +static bool getloadavg_initialized; +# endif /* NeXT */ + +# ifdef UMAX +static unsigned int cpus = 0; +static unsigned int samples; +# endif /* UMAX */ + +# ifdef DGUX +static struct dg_sys_info_load_info load_info; /* what-a-mouthful! */ +# endif /* DGUX */ + +# if !defined (HAVE_LIBKSTAT) && defined (LOAD_AVE_TYPE) +/* File descriptor open to /dev/kmem or VMS load ave driver. */ +static int channel; +/* True iff channel is valid. */ +static bool getloadavg_initialized; +/* Offset in kmem to seek to read load average, or 0 means invalid. */ +static long offset; + +# if ! defined __VMS && ! defined sgi && ! defined __linux__ +static struct nlist nl[2]; +# endif + +# ifdef SUNOS_5 +static kvm_t *kd; +# endif /* SUNOS_5 */ + +# endif /* LOAD_AVE_TYPE && !HAVE_LIBKSTAT */ + +/* Put the 1 minute, 5 minute and 15 minute load averages + into the first NELEM elements of LOADAVG. + Return the number written (never more than 3, but may be less than NELEM), + or -1 if an error occurred. */ + +int +getloadavg (double loadavg[], int nelem) +{ + int elem = 0; /* Return value. */ + +# ifdef NO_GET_LOAD_AVG +# define LDAV_DONE + /* Set errno to zero to indicate that there was no particular error; + this function just can't work at all on this system. */ + errno = 0; + elem = -1; +# endif + +# if !defined (LDAV_DONE) && defined (HAVE_LIBKSTAT) +/* Use libkstat because we don't have to be root. */ +# define LDAV_DONE + kstat_ctl_t *kc; + kstat_t *ksp; + kstat_named_t *kn; + + kc = kstat_open (); + if (kc == 0) + return -1; + ksp = kstat_lookup (kc, "unix", 0, "system_misc"); + if (ksp == 0) + return -1; + if (kstat_read (kc, ksp, 0) == -1) + return -1; + + + kn = kstat_data_lookup (ksp, "avenrun_1min"); + if (kn == 0) + { + /* Return -1 if no load average information is available. */ + nelem = 0; + elem = -1; + } + + if (nelem >= 1) + loadavg[elem++] = (double) kn->value.ul / FSCALE; + + if (nelem >= 2) + { + kn = kstat_data_lookup (ksp, "avenrun_5min"); + if (kn != 0) + { + loadavg[elem++] = (double) kn->value.ul / FSCALE; + + if (nelem >= 3) + { + kn = kstat_data_lookup (ksp, "avenrun_15min"); + if (kn != 0) + loadavg[elem++] = (double) kn->value.ul / FSCALE; + } + } + } + + kstat_close (kc); +# endif /* HAVE_LIBKSTAT */ + +# if !defined (LDAV_DONE) && defined (hpux) && defined (HAVE_PSTAT_GETDYNAMIC) +/* Use pstat_getdynamic() because we don't have to be root. */ +# define LDAV_DONE +# undef LOAD_AVE_TYPE + + struct pst_dynamic dyn_info; + if (pstat_getdynamic (&dyn_info, sizeof (dyn_info), 0, 0) < 0) + return -1; + if (nelem > 0) + loadavg[elem++] = dyn_info.psd_avg_1_min; + if (nelem > 1) + loadavg[elem++] = dyn_info.psd_avg_5_min; + if (nelem > 2) + loadavg[elem++] = dyn_info.psd_avg_15_min; + +# endif /* hpux && HAVE_PSTAT_GETDYNAMIC */ + +# if !defined (LDAV_DONE) && (defined (__linux__) || defined (__CYGWIN__)) +# define LDAV_DONE +# undef LOAD_AVE_TYPE + +# ifndef LINUX_LDAV_FILE +# define LINUX_LDAV_FILE "/proc/loadavg" +# endif + + char ldavgbuf[3 * (INT_STRLEN_BOUND (int) + sizeof ".00 ")]; + char const *ptr = ldavgbuf; + int fd, count; + + fd = open (LINUX_LDAV_FILE, O_RDONLY); + if (fd == -1) + return -1; + count = read (fd, ldavgbuf, sizeof ldavgbuf - 1); + (void) close (fd); + if (count <= 0) + return -1; + ldavgbuf[count] = '\0'; + + for (elem = 0; elem < nelem; elem++) + { + char *endptr; + double d = c_strtod (ptr, &endptr); + if (ptr == endptr) + { + if (elem == 0) + return -1; + break; + } + loadavg[elem] = d; + ptr = endptr; + } + + return elem; + +# endif /* __linux__ || __CYGWIN__ */ + +# if !defined (LDAV_DONE) && defined (__NetBSD__) +# define LDAV_DONE +# undef LOAD_AVE_TYPE + +# ifndef NETBSD_LDAV_FILE +# define NETBSD_LDAV_FILE "/kern/loadavg" +# endif + + unsigned long int load_ave[3], scale; + int count; + FILE *fp; + + fp = fopen (NETBSD_LDAV_FILE, "r"); + if (fp == NULL) + return -1; + count = fscanf (fp, "%lu %lu %lu %lu\n", + &load_ave[0], &load_ave[1], &load_ave[2], + &scale); + (void) fclose (fp); + if (count != 4) + return -1; + + for (elem = 0; elem < nelem; elem++) + loadavg[elem] = (double) load_ave[elem] / (double) scale; + + return elem; + +# endif /* __NetBSD__ */ + +# if !defined (LDAV_DONE) && defined (NeXT) +# define LDAV_DONE + /* The NeXT code was adapted from iscreen 3.2. */ + + host_t host; + struct processor_set_basic_info info; + unsigned int info_count; + + /* We only know how to get the 1-minute average for this system, + so even if the caller asks for more than 1, we only return 1. */ + + if (!getloadavg_initialized) + { + if (processor_set_default (host_self (), &default_set) == KERN_SUCCESS) + getloadavg_initialized = true; + } + + if (getloadavg_initialized) + { + info_count = PROCESSOR_SET_BASIC_INFO_COUNT; + if (processor_set_info (default_set, PROCESSOR_SET_BASIC_INFO, &host, + (processor_set_info_t) &info, &info_count) + != KERN_SUCCESS) + getloadavg_initialized = false; + else + { + if (nelem > 0) + loadavg[elem++] = (double) info.load_average / LOAD_SCALE; + } + } + + if (!getloadavg_initialized) + return -1; +# endif /* NeXT */ + +# if !defined (LDAV_DONE) && defined (UMAX) +# define LDAV_DONE +/* UMAX 4.2, which runs on the Encore Multimax multiprocessor, does not + have a /dev/kmem. Information about the workings of the running kernel + can be gathered with inq_stats system calls. + We only know how to get the 1-minute average for this system. */ + + struct proc_summary proc_sum_data; + struct stat_descr proc_info; + double load; + register unsigned int i, j; + + if (cpus == 0) + { + register unsigned int c, i; + struct cpu_config conf; + struct stat_descr desc; + + desc.sd_next = 0; + desc.sd_subsys = SUBSYS_CPU; + desc.sd_type = CPUTYPE_CONFIG; + desc.sd_addr = (char *) &conf; + desc.sd_size = sizeof conf; + + if (inq_stats (1, &desc)) + return -1; + + c = 0; + for (i = 0; i < conf.config_maxclass; ++i) + { + struct class_stats stats; + bzero ((char *) &stats, sizeof stats); + + desc.sd_type = CPUTYPE_CLASS; + desc.sd_objid = i; + desc.sd_addr = (char *) &stats; + desc.sd_size = sizeof stats; + + if (inq_stats (1, &desc)) + return -1; + + c += stats.class_numcpus; + } + cpus = c; + samples = cpus < 2 ? 3 : (2 * cpus / 3); + } + + proc_info.sd_next = 0; + proc_info.sd_subsys = SUBSYS_PROC; + proc_info.sd_type = PROCTYPE_SUMMARY; + proc_info.sd_addr = (char *) &proc_sum_data; + proc_info.sd_size = sizeof (struct proc_summary); + proc_info.sd_sizeused = 0; + + if (inq_stats (1, &proc_info) != 0) + return -1; + + load = proc_sum_data.ps_nrunnable; + j = 0; + for (i = samples - 1; i > 0; --i) + { + load += proc_sum_data.ps_nrun[j]; + if (j++ == PS_NRUNSIZE) + j = 0; + } + + if (nelem > 0) + loadavg[elem++] = load / samples / cpus; +# endif /* UMAX */ + +# if !defined (LDAV_DONE) && defined (DGUX) +# define LDAV_DONE + /* This call can return -1 for an error, but with good args + it's not supposed to fail. The first argument is for no + apparent reason of type `long int *'. */ + dg_sys_info ((long int *) &load_info, + DG_SYS_INFO_LOAD_INFO_TYPE, + DG_SYS_INFO_LOAD_VERSION_0); + + if (nelem > 0) + loadavg[elem++] = load_info.one_minute; + if (nelem > 1) + loadavg[elem++] = load_info.five_minute; + if (nelem > 2) + loadavg[elem++] = load_info.fifteen_minute; +# endif /* DGUX */ + +# if !defined (LDAV_DONE) && defined (apollo) +# define LDAV_DONE +/* Apollo code from lisch@mentorg.com (Ray Lischner). + + This system call is not documented. The load average is obtained as + three long integers, for the load average over the past minute, + five minutes, and fifteen minutes. Each value is a scaled integer, + with 16 bits of integer part and 16 bits of fraction part. + + I'm not sure which operating system first supported this system call, + but I know that SR10.2 supports it. */ + + extern void proc1_$get_loadav (); + unsigned long load_ave[3]; + + proc1_$get_loadav (load_ave); + + if (nelem > 0) + loadavg[elem++] = load_ave[0] / 65536.0; + if (nelem > 1) + loadavg[elem++] = load_ave[1] / 65536.0; + if (nelem > 2) + loadavg[elem++] = load_ave[2] / 65536.0; +# endif /* apollo */ + +# if !defined (LDAV_DONE) && defined (OSF_MIPS) +# define LDAV_DONE + + struct tbl_loadavg load_ave; + table (TBL_LOADAVG, 0, &load_ave, 1, sizeof (load_ave)); + loadavg[elem++] + = (load_ave.tl_lscale == 0 + ? load_ave.tl_avenrun.d[0] + : (load_ave.tl_avenrun.l[0] / (double) load_ave.tl_lscale)); +# endif /* OSF_MIPS */ + +# if !defined (LDAV_DONE) && (defined (__MSDOS__) || defined (WINDOWS32)) +# define LDAV_DONE + + /* A faithful emulation is going to have to be saved for a rainy day. */ + for ( ; elem < nelem; elem++) + { + loadavg[elem] = 0.0; + } +# endif /* __MSDOS__ || WINDOWS32 */ + +# if !defined (LDAV_DONE) && defined (OSF_ALPHA) +# define LDAV_DONE + + struct tbl_loadavg load_ave; + table (TBL_LOADAVG, 0, &load_ave, 1, sizeof (load_ave)); + for (elem = 0; elem < nelem; elem++) + loadavg[elem] + = (load_ave.tl_lscale == 0 + ? load_ave.tl_avenrun.d[elem] + : (load_ave.tl_avenrun.l[elem] / (double) load_ave.tl_lscale)); +# endif /* OSF_ALPHA */ + +# if ! defined LDAV_DONE && defined __VMS + /* VMS specific code -- read from the Load Ave driver. */ + + LOAD_AVE_TYPE load_ave[3]; + static bool getloadavg_initialized; +# ifdef eunice + struct + { + int dsc$w_length; + char *dsc$a_pointer; + } descriptor; +# endif + + /* Ensure that there is a channel open to the load ave device. */ + if (!getloadavg_initialized) + { + /* Attempt to open the channel. */ +# ifdef eunice + descriptor.dsc$w_length = 18; + descriptor.dsc$a_pointer = "$$VMS_LOAD_AVERAGE"; +# else + $DESCRIPTOR (descriptor, "LAV0:"); +# endif + if (sys$assign (&descriptor, &channel, 0, 0) & 1) + getloadavg_initialized = true; + } + + /* Read the load average vector. */ + if (getloadavg_initialized + && !(sys$qiow (0, channel, IO$_READVBLK, 0, 0, 0, + load_ave, 12, 0, 0, 0, 0) & 1)) + { + sys$dassgn (channel); + getloadavg_initialized = false; + } + + if (!getloadavg_initialized) + return -1; +# endif /* ! defined LDAV_DONE && defined __VMS */ + +# if ! defined LDAV_DONE && defined LOAD_AVE_TYPE && ! defined __VMS + + /* UNIX-specific code -- read the average from /dev/kmem. */ + +# define LDAV_PRIVILEGED /* This code requires special installation. */ + + LOAD_AVE_TYPE load_ave[3]; + + /* Get the address of LDAV_SYMBOL. */ + if (offset == 0) + { +# ifndef sgi +# ifndef NLIST_STRUCT + strcpy (nl[0].n_name, LDAV_SYMBOL); + strcpy (nl[1].n_name, ""); +# else /* NLIST_STRUCT */ +# ifdef HAVE_STRUCT_NLIST_N_UN_N_NAME + nl[0].n_un.n_name = LDAV_SYMBOL; + nl[1].n_un.n_name = 0; +# else /* not HAVE_STRUCT_NLIST_N_UN_N_NAME */ + nl[0].n_name = LDAV_SYMBOL; + nl[1].n_name = 0; +# endif /* not HAVE_STRUCT_NLIST_N_UN_N_NAME */ +# endif /* NLIST_STRUCT */ + +# ifndef SUNOS_5 + if ( +# if !(defined (_AIX) && !defined (ps2)) + nlist (KERNEL_FILE, nl) +# else /* _AIX */ + knlist (nl, 1, sizeof (nl[0])) +# endif + >= 0) + /* Omit "&& nl[0].n_type != 0 " -- it breaks on Sun386i. */ + { +# ifdef FIXUP_KERNEL_SYMBOL_ADDR + FIXUP_KERNEL_SYMBOL_ADDR (nl); +# endif + offset = nl[0].n_value; + } +# endif /* !SUNOS_5 */ +# else /* sgi */ + int ldav_off; + + ldav_off = sysmp (MP_KERNADDR, MPKA_AVENRUN); + if (ldav_off != -1) + offset = (long int) ldav_off & 0x7fffffff; +# endif /* sgi */ + } + + /* Make sure we have /dev/kmem open. */ + if (!getloadavg_initialized) + { +# ifndef SUNOS_5 + channel = open ("/dev/kmem", O_RDONLY); + if (channel >= 0) + { + /* Set the channel to close on exec, so it does not + litter any child's descriptor table. */ + set_cloexec_flag (channel, true); + getloadavg_initialized = true; + } +# else /* SUNOS_5 */ + /* We pass 0 for the kernel, corefile, and swapfile names + to use the currently running kernel. */ + kd = kvm_open (0, 0, 0, O_RDONLY, 0); + if (kd != 0) + { + /* nlist the currently running kernel. */ + kvm_nlist (kd, nl); + offset = nl[0].n_value; + getloadavg_initialized = true; + } +# endif /* SUNOS_5 */ + } + + /* If we can, get the load average values. */ + if (offset && getloadavg_initialized) + { + /* Try to read the load. */ +# ifndef SUNOS_5 + if (lseek (channel, offset, 0) == -1L + || read (channel, (char *) load_ave, sizeof (load_ave)) + != sizeof (load_ave)) + { + close (channel); + getloadavg_initialized = false; + } +# else /* SUNOS_5 */ + if (kvm_read (kd, offset, (char *) load_ave, sizeof (load_ave)) + != sizeof (load_ave)) + { + kvm_close (kd); + getloadavg_initialized = false; + } +# endif /* SUNOS_5 */ + } + + if (offset == 0 || !getloadavg_initialized) + return -1; +# endif /* ! defined LDAV_DONE && defined LOAD_AVE_TYPE && ! defined __VMS */ + +# if !defined (LDAV_DONE) && defined (LOAD_AVE_TYPE) /* Including VMS. */ + if (nelem > 0) + loadavg[elem++] = LDAV_CVT (load_ave[0]); + if (nelem > 1) + loadavg[elem++] = LDAV_CVT (load_ave[1]); + if (nelem > 2) + loadavg[elem++] = LDAV_CVT (load_ave[2]); + +# define LDAV_DONE +# endif /* !LDAV_DONE && LOAD_AVE_TYPE */ + +# if !defined LDAV_DONE + /* Set errno to zero to indicate that there was no particular error; + this function just can't work at all on this system. */ + errno = 0; + elem = -1; +# endif + return elem; +} + +#endif /* ! HAVE_GETLOADAVG */ + +#ifdef TEST +int +main (int argc, char **argv) +{ + int naptime = 0; + + if (argc > 1) + naptime = atoi (argv[1]); + + while (1) + { + double avg[3]; + int loads; + + errno = 0; /* Don't be misled if it doesn't set errno. */ + loads = getloadavg (avg, 3); + if (loads == -1) + { + perror ("Error getting load average"); + return EXIT_FAILURE; + } + if (loads > 0) + printf ("1-minute: %f ", avg[0]); + if (loads > 1) + printf ("5-minute: %f ", avg[1]); + if (loads > 2) + printf ("15-minute: %f ", avg[2]); + if (loads > 0) + putchar ('\n'); + + if (naptime == 0) + break; + sleep (naptime); + } + + return EXIT_SUCCESS; +} +#endif /* TEST */ diff --git a/gl/getopt.c b/gl/getopt.c new file mode 100644 index 00000000..3580ad82 --- /dev/null +++ b/gl/getopt.c @@ -0,0 +1,1191 @@ +/* Getopt for GNU. + NOTE: getopt is now part of the C library, so if you don't know what + "Keep this file name-space clean" means, talk to drepper@gnu.org + before changing it! + Copyright (C) 1987,88,89,90,91,92,93,94,95,96,98,99,2000,2001,2002,2003,2004,2006 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _LIBC +# include <config.h> +#endif + +#include "getopt.h" + +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <unistd.h> + +#ifdef __VMS +# include <unixlib.h> +#endif + +#ifdef _LIBC +# include <libintl.h> +#else +# include "gettext.h" +# define _(msgid) gettext (msgid) +#endif + +#if defined _LIBC && defined USE_IN_LIBIO +# include <wchar.h> +#endif + +#ifndef attribute_hidden +# define attribute_hidden +#endif + +/* Unlike standard Unix `getopt', functions like `getopt_long' + let the user intersperse the options with the other arguments. + + As `getopt_long' works, it permutes the elements of ARGV so that, + when it is done, all the options precede everything else. Thus + all application programs are extended to handle flexible argument order. + + Using `getopt' or setting the environment variable POSIXLY_CORRECT + disables permutation. + Then the application's behavior is completely standard. + + GNU application programs can use a third alternative mode in which + they can distinguish the relative order of options and other arguments. */ + +#include "getopt_int.h" + +/* For communication from `getopt' to the caller. + When `getopt' finds an option that takes an argument, + the argument value is returned here. + Also, when `ordering' is RETURN_IN_ORDER, + each non-option ARGV-element is returned here. */ + +char *optarg; + +/* Index in ARGV of the next element to be scanned. + This is used for communication to and from the caller + and for communication between successive calls to `getopt'. + + On entry to `getopt', zero means this is the first call; initialize. + + When `getopt' returns -1, this is the index of the first of the + non-option elements that the caller should itself scan. + + Otherwise, `optind' communicates from one call to the next + how much of ARGV has been scanned so far. */ + +/* 1003.2 says this must be 1 before any call. */ +int optind = 1; + +/* Callers store zero here to inhibit the error message + for unrecognized options. */ + +int opterr = 1; + +/* Set to an option character which was unrecognized. + This must be initialized on some systems to avoid linking in the + system's own getopt implementation. */ + +int optopt = '?'; + +/* Keep a global copy of all internal members of getopt_data. */ + +static struct _getopt_data getopt_data; + + +#if defined HAVE_DECL_GETENV && !HAVE_DECL_GETENV +extern char *getenv (); +#endif + +#ifdef _LIBC +/* Stored original parameters. + XXX This is no good solution. We should rather copy the args so + that we can compare them later. But we must not use malloc(3). */ +extern int __libc_argc; +extern char **__libc_argv; + +/* Bash 2.0 gives us an environment variable containing flags + indicating ARGV elements that should not be considered arguments. */ + +# ifdef USE_NONOPTION_FLAGS +/* Defined in getopt_init.c */ +extern char *__getopt_nonoption_flags; +# endif + +# ifdef USE_NONOPTION_FLAGS +# define SWAP_FLAGS(ch1, ch2) \ + if (d->__nonoption_flags_len > 0) \ + { \ + char __tmp = __getopt_nonoption_flags[ch1]; \ + __getopt_nonoption_flags[ch1] = __getopt_nonoption_flags[ch2]; \ + __getopt_nonoption_flags[ch2] = __tmp; \ + } +# else +# define SWAP_FLAGS(ch1, ch2) +# endif +#else /* !_LIBC */ +# define SWAP_FLAGS(ch1, ch2) +#endif /* _LIBC */ + +/* Exchange two adjacent subsequences of ARGV. + One subsequence is elements [first_nonopt,last_nonopt) + which contains all the non-options that have been skipped so far. + The other is elements [last_nonopt,optind), which contains all + the options processed since those non-options were skipped. + + `first_nonopt' and `last_nonopt' are relocated so that they describe + the new indices of the non-options in ARGV after they are moved. */ + +static void +exchange (char **argv, struct _getopt_data *d) +{ + int bottom = d->__first_nonopt; + int middle = d->__last_nonopt; + int top = d->optind; + char *tem; + + /* Exchange the shorter segment with the far end of the longer segment. + That puts the shorter segment into the right place. + It leaves the longer segment in the right place overall, + but it consists of two parts that need to be swapped next. */ + +#if defined _LIBC && defined USE_NONOPTION_FLAGS + /* First make sure the handling of the `__getopt_nonoption_flags' + string can work normally. Our top argument must be in the range + of the string. */ + if (d->__nonoption_flags_len > 0 && top >= d->__nonoption_flags_max_len) + { + /* We must extend the array. The user plays games with us and + presents new arguments. */ + char *new_str = malloc (top + 1); + if (new_str == NULL) + d->__nonoption_flags_len = d->__nonoption_flags_max_len = 0; + else + { + memset (__mempcpy (new_str, __getopt_nonoption_flags, + d->__nonoption_flags_max_len), + '\0', top + 1 - d->__nonoption_flags_max_len); + d->__nonoption_flags_max_len = top + 1; + __getopt_nonoption_flags = new_str; + } + } +#endif + + while (top > middle && middle > bottom) + { + if (top - middle > middle - bottom) + { + /* Bottom segment is the short one. */ + int len = middle - bottom; + register int i; + + /* Swap it with the top part of the top segment. */ + for (i = 0; i < len; i++) + { + tem = argv[bottom + i]; + argv[bottom + i] = argv[top - (middle - bottom) + i]; + argv[top - (middle - bottom) + i] = tem; + SWAP_FLAGS (bottom + i, top - (middle - bottom) + i); + } + /* Exclude the moved bottom segment from further swapping. */ + top -= len; + } + else + { + /* Top segment is the short one. */ + int len = top - middle; + register int i; + + /* Swap it with the bottom part of the bottom segment. */ + for (i = 0; i < len; i++) + { + tem = argv[bottom + i]; + argv[bottom + i] = argv[middle + i]; + argv[middle + i] = tem; + SWAP_FLAGS (bottom + i, middle + i); + } + /* Exclude the moved top segment from further swapping. */ + bottom += len; + } + } + + /* Update records for the slots the non-options now occupy. */ + + d->__first_nonopt += (d->optind - d->__last_nonopt); + d->__last_nonopt = d->optind; +} + +/* Initialize the internal data when the first call is made. */ + +static const char * +_getopt_initialize (int argc, char **argv, const char *optstring, + int posixly_correct, struct _getopt_data *d) +{ + /* Start processing options with ARGV-element 1 (since ARGV-element 0 + is the program name); the sequence of previously skipped + non-option ARGV-elements is empty. */ + + d->__first_nonopt = d->__last_nonopt = d->optind; + + d->__nextchar = NULL; + + d->__posixly_correct = posixly_correct || !!getenv ("POSIXLY_CORRECT"); + + /* Determine how to handle the ordering of options and nonoptions. */ + + if (optstring[0] == '-') + { + d->__ordering = RETURN_IN_ORDER; + ++optstring; + } + else if (optstring[0] == '+') + { + d->__ordering = REQUIRE_ORDER; + ++optstring; + } + else if (d->__posixly_correct) + d->__ordering = REQUIRE_ORDER; + else + d->__ordering = PERMUTE; + +#if defined _LIBC && defined USE_NONOPTION_FLAGS + if (!d->__posixly_correct + && argc == __libc_argc && argv == __libc_argv) + { + if (d->__nonoption_flags_max_len == 0) + { + if (__getopt_nonoption_flags == NULL + || __getopt_nonoption_flags[0] == '\0') + d->__nonoption_flags_max_len = -1; + else + { + const char *orig_str = __getopt_nonoption_flags; + int len = d->__nonoption_flags_max_len = strlen (orig_str); + if (d->__nonoption_flags_max_len < argc) + d->__nonoption_flags_max_len = argc; + __getopt_nonoption_flags = + (char *) malloc (d->__nonoption_flags_max_len); + if (__getopt_nonoption_flags == NULL) + d->__nonoption_flags_max_len = -1; + else + memset (__mempcpy (__getopt_nonoption_flags, orig_str, len), + '\0', d->__nonoption_flags_max_len - len); + } + } + d->__nonoption_flags_len = d->__nonoption_flags_max_len; + } + else + d->__nonoption_flags_len = 0; +#endif + + return optstring; +} + +/* Scan elements of ARGV (whose length is ARGC) for option characters + given in OPTSTRING. + + If an element of ARGV starts with '-', and is not exactly "-" or "--", + then it is an option element. The characters of this element + (aside from the initial '-') are option characters. If `getopt' + is called repeatedly, it returns successively each of the option characters + from each of the option elements. + + If `getopt' finds another option character, it returns that character, + updating `optind' and `nextchar' so that the next call to `getopt' can + resume the scan with the following option character or ARGV-element. + + If there are no more option characters, `getopt' returns -1. + Then `optind' is the index in ARGV of the first ARGV-element + that is not an option. (The ARGV-elements have been permuted + so that those that are not options now come last.) + + OPTSTRING is a string containing the legitimate option characters. + If an option character is seen that is not listed in OPTSTRING, + return '?' after printing an error message. If you set `opterr' to + zero, the error message is suppressed but we still return '?'. + + If a char in OPTSTRING is followed by a colon, that means it wants an arg, + so the following text in the same ARGV-element, or the text of the following + ARGV-element, is returned in `optarg'. Two colons mean an option that + wants an optional arg; if there is text in the current ARGV-element, + it is returned in `optarg', otherwise `optarg' is set to zero. + + If OPTSTRING starts with `-' or `+', it requests different methods of + handling the non-option ARGV-elements. + See the comments about RETURN_IN_ORDER and REQUIRE_ORDER, above. + + Long-named options begin with `--' instead of `-'. + Their names may be abbreviated as long as the abbreviation is unique + or is an exact match for some defined option. If they have an + argument, it follows the option name in the same ARGV-element, separated + from the option name by a `=', or else the in next ARGV-element. + When `getopt' finds a long-named option, it returns 0 if that option's + `flag' field is nonzero, the value of the option's `val' field + if the `flag' field is zero. + + LONGOPTS is a vector of `struct option' terminated by an + element containing a name which is zero. + + LONGIND returns the index in LONGOPT of the long-named option found. + It is only valid when a long-named option has been found by the most + recent call. + + If LONG_ONLY is nonzero, '-' as well as '--' can introduce + long-named options. + + If POSIXLY_CORRECT is nonzero, behave as if the POSIXLY_CORRECT + environment variable were set. */ + +int +_getopt_internal_r (int argc, char **argv, const char *optstring, + const struct option *longopts, int *longind, + int long_only, int posixly_correct, struct _getopt_data *d) +{ + int print_errors = d->opterr; + if (optstring[0] == ':') + print_errors = 0; + + if (argc < 1) + return -1; + + d->optarg = NULL; + + if (d->optind == 0 || !d->__initialized) + { + if (d->optind == 0) + d->optind = 1; /* Don't scan ARGV[0], the program name. */ + optstring = _getopt_initialize (argc, argv, optstring, + posixly_correct, d); + d->__initialized = 1; + } + + /* Test whether ARGV[optind] points to a non-option argument. + Either it does not have option syntax, or there is an environment flag + from the shell indicating it is not an option. The later information + is only used when the used in the GNU libc. */ +#if defined _LIBC && defined USE_NONOPTION_FLAGS +# define NONOPTION_P (argv[d->optind][0] != '-' || argv[d->optind][1] == '\0' \ + || (d->optind < d->__nonoption_flags_len \ + && __getopt_nonoption_flags[d->optind] == '1')) +#else +# define NONOPTION_P (argv[d->optind][0] != '-' || argv[d->optind][1] == '\0') +#endif + + if (d->__nextchar == NULL || *d->__nextchar == '\0') + { + /* Advance to the next ARGV-element. */ + + /* Give FIRST_NONOPT & LAST_NONOPT rational values if OPTIND has been + moved back by the user (who may also have changed the arguments). */ + if (d->__last_nonopt > d->optind) + d->__last_nonopt = d->optind; + if (d->__first_nonopt > d->optind) + d->__first_nonopt = d->optind; + + if (d->__ordering == PERMUTE) + { + /* If we have just processed some options following some non-options, + exchange them so that the options come first. */ + + if (d->__first_nonopt != d->__last_nonopt + && d->__last_nonopt != d->optind) + exchange ((char **) argv, d); + else if (d->__last_nonopt != d->optind) + d->__first_nonopt = d->optind; + + /* Skip any additional non-options + and extend the range of non-options previously skipped. */ + + while (d->optind < argc && NONOPTION_P) + d->optind++; + d->__last_nonopt = d->optind; + } + + /* The special ARGV-element `--' means premature end of options. + Skip it like a null option, + then exchange with previous non-options as if it were an option, + then skip everything else like a non-option. */ + + if (d->optind != argc && !strcmp (argv[d->optind], "--")) + { + d->optind++; + + if (d->__first_nonopt != d->__last_nonopt + && d->__last_nonopt != d->optind) + exchange ((char **) argv, d); + else if (d->__first_nonopt == d->__last_nonopt) + d->__first_nonopt = d->optind; + d->__last_nonopt = argc; + + d->optind = argc; + } + + /* If we have done all the ARGV-elements, stop the scan + and back over any non-options that we skipped and permuted. */ + + if (d->optind == argc) + { + /* Set the next-arg-index to point at the non-options + that we previously skipped, so the caller will digest them. */ + if (d->__first_nonopt != d->__last_nonopt) + d->optind = d->__first_nonopt; + return -1; + } + + /* If we have come to a non-option and did not permute it, + either stop the scan or describe it to the caller and pass it by. */ + + if (NONOPTION_P) + { + if (d->__ordering == REQUIRE_ORDER) + return -1; + d->optarg = argv[d->optind++]; + return 1; + } + + /* We have found another option-ARGV-element. + Skip the initial punctuation. */ + + d->__nextchar = (argv[d->optind] + 1 + + (longopts != NULL && argv[d->optind][1] == '-')); + } + + /* Decode the current option-ARGV-element. */ + + /* Check whether the ARGV-element is a long option. + + If long_only and the ARGV-element has the form "-f", where f is + a valid short option, don't consider it an abbreviated form of + a long option that starts with f. Otherwise there would be no + way to give the -f short option. + + On the other hand, if there's a long option "fubar" and + the ARGV-element is "-fu", do consider that an abbreviation of + the long option, just like "--fu", and not "-f" with arg "u". + + This distinction seems to be the most useful approach. */ + + if (longopts != NULL + && (argv[d->optind][1] == '-' + || (long_only && (argv[d->optind][2] + || !strchr (optstring, argv[d->optind][1]))))) + { + char *nameend; + const struct option *p; + const struct option *pfound = NULL; + int exact = 0; + int ambig = 0; + int indfound = -1; + int option_index; + + for (nameend = d->__nextchar; *nameend && *nameend != '='; nameend++) + /* Do nothing. */ ; + + /* Test all long options for either exact match + or abbreviated matches. */ + for (p = longopts, option_index = 0; p->name; p++, option_index++) + if (!strncmp (p->name, d->__nextchar, nameend - d->__nextchar)) + { + if ((unsigned int) (nameend - d->__nextchar) + == (unsigned int) strlen (p->name)) + { + /* Exact match found. */ + pfound = p; + indfound = option_index; + exact = 1; + break; + } + else if (pfound == NULL) + { + /* First nonexact match found. */ + pfound = p; + indfound = option_index; + } + else if (long_only + || pfound->has_arg != p->has_arg + || pfound->flag != p->flag + || pfound->val != p->val) + /* Second or later nonexact match found. */ + ambig = 1; + } + + if (ambig && !exact) + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("%s: option `%s' is ambiguous\n"), + argv[0], argv[d->optind]) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, _("%s: option `%s' is ambiguous\n"), + argv[0], argv[d->optind]); +#endif + } + d->__nextchar += strlen (d->__nextchar); + d->optind++; + d->optopt = 0; + return '?'; + } + + if (pfound != NULL) + { + option_index = indfound; + d->optind++; + if (*nameend) + { + /* Don't test has_arg with >, because some C compilers don't + allow it to be used on enums. */ + if (pfound->has_arg) + d->optarg = nameend + 1; + else + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + int n; +#endif + + if (argv[d->optind - 1][1] == '-') + { + /* --option */ +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("\ +%s: option `--%s' doesn't allow an argument\n"), + argv[0], pfound->name); +#else + fprintf (stderr, _("\ +%s: option `--%s' doesn't allow an argument\n"), + argv[0], pfound->name); +#endif + } + else + { + /* +option or -option */ +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("\ +%s: option `%c%s' doesn't allow an argument\n"), + argv[0], argv[d->optind - 1][0], + pfound->name); +#else + fprintf (stderr, _("\ +%s: option `%c%s' doesn't allow an argument\n"), + argv[0], argv[d->optind - 1][0], + pfound->name); +#endif + } + +#if defined _LIBC && defined USE_IN_LIBIO + if (n >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 + |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#endif + } + + d->__nextchar += strlen (d->__nextchar); + + d->optopt = pfound->val; + return '?'; + } + } + else if (pfound->has_arg == 1) + { + if (d->optind < argc) + d->optarg = argv[d->optind++]; + else + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("\ +%s: option `%s' requires an argument\n"), + argv[0], argv[d->optind - 1]) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 + |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, + _("%s: option `%s' requires an argument\n"), + argv[0], argv[d->optind - 1]); +#endif + } + d->__nextchar += strlen (d->__nextchar); + d->optopt = pfound->val; + return optstring[0] == ':' ? ':' : '?'; + } + } + d->__nextchar += strlen (d->__nextchar); + if (longind != NULL) + *longind = option_index; + if (pfound->flag) + { + *(pfound->flag) = pfound->val; + return 0; + } + return pfound->val; + } + + /* Can't find it as a long option. If this is not getopt_long_only, + or the option starts with '--' or is not a valid short + option, then it's an error. + Otherwise interpret it as a short option. */ + if (!long_only || argv[d->optind][1] == '-' + || strchr (optstring, *d->__nextchar) == NULL) + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + int n; +#endif + + if (argv[d->optind][1] == '-') + { + /* --option */ +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("%s: unrecognized option `--%s'\n"), + argv[0], d->__nextchar); +#else + fprintf (stderr, _("%s: unrecognized option `--%s'\n"), + argv[0], d->__nextchar); +#endif + } + else + { + /* +option or -option */ +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("%s: unrecognized option `%c%s'\n"), + argv[0], argv[d->optind][0], d->__nextchar); +#else + fprintf (stderr, _("%s: unrecognized option `%c%s'\n"), + argv[0], argv[d->optind][0], d->__nextchar); +#endif + } + +#if defined _LIBC && defined USE_IN_LIBIO + if (n >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#endif + } + d->__nextchar = (char *) ""; + d->optind++; + d->optopt = 0; + return '?'; + } + } + + /* Look at and handle the next short option-character. */ + + { + char c = *d->__nextchar++; + char *temp = strchr (optstring, c); + + /* Increment `optind' when we start to process its last character. */ + if (*d->__nextchar == '\0') + ++d->optind; + + if (temp == NULL || c == ':') + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + int n; +#endif + + if (d->__posixly_correct) + { + /* 1003.2 specifies the format of this message. */ +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("%s: illegal option -- %c\n"), + argv[0], c); +#else + fprintf (stderr, _("%s: illegal option -- %c\n"), argv[0], c); +#endif + } + else + { +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("%s: invalid option -- %c\n"), + argv[0], c); +#else + fprintf (stderr, _("%s: invalid option -- %c\n"), argv[0], c); +#endif + } + +#if defined _LIBC && defined USE_IN_LIBIO + if (n >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#endif + } + d->optopt = c; + return '?'; + } + /* Convenience. Treat POSIX -W foo same as long option --foo */ + if (temp[0] == 'W' && temp[1] == ';') + { + char *nameend; + const struct option *p; + const struct option *pfound = NULL; + int exact = 0; + int ambig = 0; + int indfound = 0; + int option_index; + + /* This is an option that requires an argument. */ + if (*d->__nextchar != '\0') + { + d->optarg = d->__nextchar; + /* If we end this ARGV-element by taking the rest as an arg, + we must advance to the next element now. */ + d->optind++; + } + else if (d->optind == argc) + { + if (print_errors) + { + /* 1003.2 specifies the format of this message. */ +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, + _("%s: option requires an argument -- %c\n"), + argv[0], c) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, _("%s: option requires an argument -- %c\n"), + argv[0], c); +#endif + } + d->optopt = c; + if (optstring[0] == ':') + c = ':'; + else + c = '?'; + return c; + } + else + /* We already incremented `d->optind' once; + increment it again when taking next ARGV-elt as argument. */ + d->optarg = argv[d->optind++]; + + /* optarg is now the argument, see if it's in the + table of longopts. */ + + for (d->__nextchar = nameend = d->optarg; *nameend && *nameend != '='; + nameend++) + /* Do nothing. */ ; + + /* Test all long options for either exact match + or abbreviated matches. */ + for (p = longopts, option_index = 0; p->name; p++, option_index++) + if (!strncmp (p->name, d->__nextchar, nameend - d->__nextchar)) + { + if ((unsigned int) (nameend - d->__nextchar) == strlen (p->name)) + { + /* Exact match found. */ + pfound = p; + indfound = option_index; + exact = 1; + break; + } + else if (pfound == NULL) + { + /* First nonexact match found. */ + pfound = p; + indfound = option_index; + } + else + /* Second or later nonexact match found. */ + ambig = 1; + } + if (ambig && !exact) + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("%s: option `-W %s' is ambiguous\n"), + argv[0], argv[d->optind]) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, _("%s: option `-W %s' is ambiguous\n"), + argv[0], argv[d->optind]); +#endif + } + d->__nextchar += strlen (d->__nextchar); + d->optind++; + return '?'; + } + if (pfound != NULL) + { + option_index = indfound; + if (*nameend) + { + /* Don't test has_arg with >, because some C compilers don't + allow it to be used on enums. */ + if (pfound->has_arg) + d->optarg = nameend + 1; + else + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("\ +%s: option `-W %s' doesn't allow an argument\n"), + argv[0], pfound->name) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 + |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, _("\ +%s: option `-W %s' doesn't allow an argument\n"), + argv[0], pfound->name); +#endif + } + + d->__nextchar += strlen (d->__nextchar); + return '?'; + } + } + else if (pfound->has_arg == 1) + { + if (d->optind < argc) + d->optarg = argv[d->optind++]; + else + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("\ +%s: option `%s' requires an argument\n"), + argv[0], argv[d->optind - 1]) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 + |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, + _("%s: option `%s' requires an argument\n"), + argv[0], argv[d->optind - 1]); +#endif + } + d->__nextchar += strlen (d->__nextchar); + return optstring[0] == ':' ? ':' : '?'; + } + } + d->__nextchar += strlen (d->__nextchar); + if (longind != NULL) + *longind = option_index; + if (pfound->flag) + { + *(pfound->flag) = pfound->val; + return 0; + } + return pfound->val; + } + d->__nextchar = NULL; + return 'W'; /* Let the application handle it. */ + } + if (temp[1] == ':') + { + if (temp[2] == ':') + { + /* This is an option that accepts an argument optionally. */ + if (*d->__nextchar != '\0') + { + d->optarg = d->__nextchar; + d->optind++; + } + else + d->optarg = NULL; + d->__nextchar = NULL; + } + else + { + /* This is an option that requires an argument. */ + if (*d->__nextchar != '\0') + { + d->optarg = d->__nextchar; + /* If we end this ARGV-element by taking the rest as an arg, + we must advance to the next element now. */ + d->optind++; + } + else if (d->optind == argc) + { + if (print_errors) + { + /* 1003.2 specifies the format of this message. */ +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("\ +%s: option requires an argument -- %c\n"), + argv[0], c) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, + _("%s: option requires an argument -- %c\n"), + argv[0], c); +#endif + } + d->optopt = c; + if (optstring[0] == ':') + c = ':'; + else + c = '?'; + } + else + /* We already incremented `optind' once; + increment it again when taking next ARGV-elt as argument. */ + d->optarg = argv[d->optind++]; + d->__nextchar = NULL; + } + } + return c; + } +} + +int +_getopt_internal (int argc, char **argv, const char *optstring, + const struct option *longopts, int *longind, + int long_only, int posixly_correct) +{ + int result; + + getopt_data.optind = optind; + getopt_data.opterr = opterr; + + result = _getopt_internal_r (argc, argv, optstring, longopts, longind, + long_only, posixly_correct, &getopt_data); + + optind = getopt_data.optind; + optarg = getopt_data.optarg; + optopt = getopt_data.optopt; + + return result; +} + +/* glibc gets a LSB-compliant getopt. + Standalone applications get a POSIX-compliant getopt. */ +#if _LIBC +enum { POSIXLY_CORRECT = 0 }; +#else +enum { POSIXLY_CORRECT = 1 }; +#endif + +int +getopt (int argc, char *const *argv, const char *optstring) +{ + return _getopt_internal (argc, (char **) argv, optstring, NULL, NULL, 0, + POSIXLY_CORRECT); +} + + +#ifdef TEST + +/* Compile with -DTEST to make an executable for use in testing + the above definition of `getopt'. */ + +int +main (int argc, char **argv) +{ + int c; + int digit_optind = 0; + + while (1) + { + int this_option_optind = optind ? optind : 1; + + c = getopt (argc, argv, "abc:d:0123456789"); + if (c == -1) + break; + + switch (c) + { + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + if (digit_optind != 0 && digit_optind != this_option_optind) + printf ("digits occur in two different argv-elements.\n"); + digit_optind = this_option_optind; + printf ("option %c\n", c); + break; + + case 'a': + printf ("option a\n"); + break; + + case 'b': + printf ("option b\n"); + break; + + case 'c': + printf ("option c with value `%s'\n", optarg); + break; + + case '?': + break; + + default: + printf ("?? getopt returned character code 0%o ??\n", c); + } + } + + if (optind < argc) + { + printf ("non-option ARGV-elements: "); + while (optind < argc) + printf ("%s ", argv[optind++]); + printf ("\n"); + } + + exit (0); +} + +#endif /* TEST */ diff --git a/gl/getopt1.c b/gl/getopt1.c new file mode 100644 index 00000000..cc0746ea --- /dev/null +++ b/gl/getopt1.c @@ -0,0 +1,171 @@ +/* getopt_long and getopt_long_only entry points for GNU getopt. + Copyright (C) 1987,88,89,90,91,92,93,94,96,97,98,2004,2006 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifdef _LIBC +# include <getopt.h> +#else +# include <config.h> +# include "getopt.h" +#endif +#include "getopt_int.h" + +#include <stdio.h> + +/* This needs to come after some library #include + to get __GNU_LIBRARY__ defined. */ +#ifdef __GNU_LIBRARY__ +#include <stdlib.h> +#endif + +#ifndef NULL +#define NULL 0 +#endif + +int +getopt_long (int argc, char *__getopt_argv_const *argv, const char *options, + const struct option *long_options, int *opt_index) +{ + return _getopt_internal (argc, (char **) argv, options, long_options, + opt_index, 0, 0); +} + +int +_getopt_long_r (int argc, char **argv, const char *options, + const struct option *long_options, int *opt_index, + struct _getopt_data *d) +{ + return _getopt_internal_r (argc, argv, options, long_options, opt_index, + 0, 0, d); +} + +/* Like getopt_long, but '-' as well as '--' can indicate a long option. + If an option that starts with '-' (not '--') doesn't match a long option, + but does match a short option, it is parsed as a short option + instead. */ + +int +getopt_long_only (int argc, char *__getopt_argv_const *argv, + const char *options, + const struct option *long_options, int *opt_index) +{ + return _getopt_internal (argc, (char **) argv, options, long_options, + opt_index, 1, 0); +} + +int +_getopt_long_only_r (int argc, char **argv, const char *options, + const struct option *long_options, int *opt_index, + struct _getopt_data *d) +{ + return _getopt_internal_r (argc, argv, options, long_options, opt_index, + 1, 0, d); +} + + +#ifdef TEST + +#include <stdio.h> + +int +main (int argc, char **argv) +{ + int c; + int digit_optind = 0; + + while (1) + { + int this_option_optind = optind ? optind : 1; + int option_index = 0; + static struct option long_options[] = + { + {"add", 1, 0, 0}, + {"append", 0, 0, 0}, + {"delete", 1, 0, 0}, + {"verbose", 0, 0, 0}, + {"create", 0, 0, 0}, + {"file", 1, 0, 0}, + {0, 0, 0, 0} + }; + + c = getopt_long (argc, argv, "abc:d:0123456789", + long_options, &option_index); + if (c == -1) + break; + + switch (c) + { + case 0: + printf ("option %s", long_options[option_index].name); + if (optarg) + printf (" with arg %s", optarg); + printf ("\n"); + break; + + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + if (digit_optind != 0 && digit_optind != this_option_optind) + printf ("digits occur in two different argv-elements.\n"); + digit_optind = this_option_optind; + printf ("option %c\n", c); + break; + + case 'a': + printf ("option a\n"); + break; + + case 'b': + printf ("option b\n"); + break; + + case 'c': + printf ("option c with value `%s'\n", optarg); + break; + + case 'd': + printf ("option d with value `%s'\n", optarg); + break; + + case '?': + break; + + default: + printf ("?? getopt returned character code 0%o ??\n", c); + } + } + + if (optind < argc) + { + printf ("non-option ARGV-elements: "); + while (optind < argc) + printf ("%s ", argv[optind++]); + printf ("\n"); + } + + exit (0); +} + +#endif /* TEST */ diff --git a/gl/getopt_.h b/gl/getopt_.h new file mode 100644 index 00000000..27fce3dc --- /dev/null +++ b/gl/getopt_.h @@ -0,0 +1,226 @@ +/* Declarations for getopt. + Copyright (C) 1989-1994,1996-1999,2001,2003,2004,2005,2006 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _GETOPT_H + +#ifndef __need_getopt +# define _GETOPT_H 1 +#endif + +/* Standalone applications should #define __GETOPT_PREFIX to an + identifier that prefixes the external functions and variables + defined in this header. When this happens, include the + headers that might declare getopt so that they will not cause + confusion if included after this file. Then systematically rename + identifiers so that they do not collide with the system functions + and variables. Renaming avoids problems with some compilers and + linkers. */ +#if defined __GETOPT_PREFIX && !defined __need_getopt +# include <stdlib.h> +# include <stdio.h> +# include <unistd.h> +# undef __need_getopt +# undef getopt +# undef getopt_long +# undef getopt_long_only +# undef optarg +# undef opterr +# undef optind +# undef optopt +# define __GETOPT_CONCAT(x, y) x ## y +# define __GETOPT_XCONCAT(x, y) __GETOPT_CONCAT (x, y) +# define __GETOPT_ID(y) __GETOPT_XCONCAT (__GETOPT_PREFIX, y) +# define getopt __GETOPT_ID (getopt) +# define getopt_long __GETOPT_ID (getopt_long) +# define getopt_long_only __GETOPT_ID (getopt_long_only) +# define optarg __GETOPT_ID (optarg) +# define opterr __GETOPT_ID (opterr) +# define optind __GETOPT_ID (optind) +# define optopt __GETOPT_ID (optopt) +#endif + +/* Standalone applications get correct prototypes for getopt_long and + getopt_long_only; they declare "char **argv". libc uses prototypes + with "char *const *argv" that are incorrect because getopt_long and + getopt_long_only can permute argv; this is required for backward + compatibility (e.g., for LSB 2.0.1). + + This used to be `#if defined __GETOPT_PREFIX && !defined __need_getopt', + but it caused redefinition warnings if both unistd.h and getopt.h were + included, since unistd.h includes getopt.h having previously defined + __need_getopt. + + The only place where __getopt_argv_const is used is in definitions + of getopt_long and getopt_long_only below, but these are visible + only if __need_getopt is not defined, so it is quite safe to rewrite + the conditional as follows: +*/ +#if !defined __need_getopt +# if defined __GETOPT_PREFIX +# define __getopt_argv_const /* empty */ +# else +# define __getopt_argv_const const +# endif +#endif + +/* If __GNU_LIBRARY__ is not already defined, either we are being used + standalone, or this is the first header included in the source file. + If we are being used with glibc, we need to include <features.h>, but + that does not exist if we are standalone. So: if __GNU_LIBRARY__ is + not defined, include <ctype.h>, which will pull in <features.h> for us + if it's from glibc. (Why ctype.h? It's guaranteed to exist and it + doesn't flood the namespace with stuff the way some other headers do.) */ +#if !defined __GNU_LIBRARY__ +# include <ctype.h> +#endif + +#ifndef __THROW +# ifndef __GNUC_PREREQ +# define __GNUC_PREREQ(maj, min) (0) +# endif +# if defined __cplusplus && __GNUC_PREREQ (2,8) +# define __THROW throw () +# else +# define __THROW +# endif +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +/* For communication from `getopt' to the caller. + When `getopt' finds an option that takes an argument, + the argument value is returned here. + Also, when `ordering' is RETURN_IN_ORDER, + each non-option ARGV-element is returned here. */ + +extern char *optarg; + +/* Index in ARGV of the next element to be scanned. + This is used for communication to and from the caller + and for communication between successive calls to `getopt'. + + On entry to `getopt', zero means this is the first call; initialize. + + When `getopt' returns -1, this is the index of the first of the + non-option elements that the caller should itself scan. + + Otherwise, `optind' communicates from one call to the next + how much of ARGV has been scanned so far. */ + +extern int optind; + +/* Callers store zero here to inhibit the error message `getopt' prints + for unrecognized options. */ + +extern int opterr; + +/* Set to an option character which was unrecognized. */ + +extern int optopt; + +#ifndef __need_getopt +/* Describe the long-named options requested by the application. + The LONG_OPTIONS argument to getopt_long or getopt_long_only is a vector + of `struct option' terminated by an element containing a name which is + zero. + + The field `has_arg' is: + no_argument (or 0) if the option does not take an argument, + required_argument (or 1) if the option requires an argument, + optional_argument (or 2) if the option takes an optional argument. + + If the field `flag' is not NULL, it points to a variable that is set + to the value given in the field `val' when the option is found, but + left unchanged if the option is not found. + + To have a long-named option do something other than set an `int' to + a compiled-in constant, such as set a value from `optarg', set the + option's `flag' field to zero and its `val' field to a nonzero + value (the equivalent single-letter option character, if there is + one). For long options that have a zero `flag' field, `getopt' + returns the contents of the `val' field. */ + +struct option +{ + const char *name; + /* has_arg can't be an enum because some compilers complain about + type mismatches in all the code that assumes it is an int. */ + int has_arg; + int *flag; + int val; +}; + +/* Names for the values of the `has_arg' field of `struct option'. */ + +# define no_argument 0 +# define required_argument 1 +# define optional_argument 2 +#endif /* need getopt */ + + +/* Get definitions and prototypes for functions to process the + arguments in ARGV (ARGC of them, minus the program name) for + options given in OPTS. + + Return the option character from OPTS just read. Return -1 when + there are no more options. For unrecognized options, or options + missing arguments, `optopt' is set to the option letter, and '?' is + returned. + + The OPTS string is a list of characters which are recognized option + letters, optionally followed by colons, specifying that that letter + takes an argument, to be placed in `optarg'. + + If a letter in OPTS is followed by two colons, its argument is + optional. This behavior is specific to the GNU `getopt'. + + The argument `--' causes premature termination of argument + scanning, explicitly telling `getopt' that there are no more + options. + + If OPTS begins with `-', then non-option arguments are treated as + arguments to the option '\1'. This behavior is specific to the GNU + `getopt'. If OPTS begins with `+', or POSIXLY_CORRECT is set in + the environment, then do not permute arguments. */ + +extern int getopt (int ___argc, char *const *___argv, const char *__shortopts) + __THROW; + +#ifndef __need_getopt +extern int getopt_long (int ___argc, char *__getopt_argv_const *___argv, + const char *__shortopts, + const struct option *__longopts, int *__longind) + __THROW; +extern int getopt_long_only (int ___argc, char *__getopt_argv_const *___argv, + const char *__shortopts, + const struct option *__longopts, int *__longind) + __THROW; + +#endif + +#ifdef __cplusplus +} +#endif + +/* Make sure we later can get all the definitions and declarations. */ +#undef __need_getopt + +#endif /* getopt.h */ diff --git a/gl/getopt_int.h b/gl/getopt_int.h new file mode 100644 index 00000000..401579fd --- /dev/null +++ b/gl/getopt_int.h @@ -0,0 +1,131 @@ +/* Internal declarations for getopt. + Copyright (C) 1989-1994,1996-1999,2001,2003,2004 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _GETOPT_INT_H +#define _GETOPT_INT_H 1 + +extern int _getopt_internal (int ___argc, char **___argv, + const char *__shortopts, + const struct option *__longopts, int *__longind, + int __long_only, int __posixly_correct); + + +/* Reentrant versions which can handle parsing multiple argument + vectors at the same time. */ + +/* Data type for reentrant functions. */ +struct _getopt_data +{ + /* These have exactly the same meaning as the corresponding global + variables, except that they are used for the reentrant + versions of getopt. */ + int optind; + int opterr; + int optopt; + char *optarg; + + /* Internal members. */ + + /* True if the internal members have been initialized. */ + int __initialized; + + /* The next char to be scanned in the option-element + in which the last option character we returned was found. + This allows us to pick up the scan where we left off. + + If this is zero, or a null string, it means resume the scan + by advancing to the next ARGV-element. */ + char *__nextchar; + + /* Describe how to deal with options that follow non-option ARGV-elements. + + If the caller did not specify anything, + the default is REQUIRE_ORDER if the environment variable + POSIXLY_CORRECT is defined, PERMUTE otherwise. + + REQUIRE_ORDER means don't recognize them as options; + stop option processing when the first non-option is seen. + This is what Unix does. + This mode of operation is selected by either setting the environment + variable POSIXLY_CORRECT, or using `+' as the first character + of the list of option characters, or by calling getopt. + + PERMUTE is the default. We permute the contents of ARGV as we + scan, so that eventually all the non-options are at the end. + This allows options to be given in any order, even with programs + that were not written to expect this. + + RETURN_IN_ORDER is an option available to programs that were + written to expect options and other ARGV-elements in any order + and that care about the ordering of the two. We describe each + non-option ARGV-element as if it were the argument of an option + with character code 1. Using `-' as the first character of the + list of option characters selects this mode of operation. + + The special argument `--' forces an end of option-scanning regardless + of the value of `ordering'. In the case of RETURN_IN_ORDER, only + `--' can cause `getopt' to return -1 with `optind' != ARGC. */ + + enum + { + REQUIRE_ORDER, PERMUTE, RETURN_IN_ORDER + } __ordering; + + /* If the POSIXLY_CORRECT environment variable is set + or getopt was called. */ + int __posixly_correct; + + + /* Handle permutation of arguments. */ + + /* Describe the part of ARGV that contains non-options that have + been skipped. `first_nonopt' is the index in ARGV of the first + of them; `last_nonopt' is the index after the last of them. */ + + int __first_nonopt; + int __last_nonopt; + +#if defined _LIBC && defined USE_NONOPTION_FLAGS + int __nonoption_flags_max_len; + int __nonoption_flags_len; +# endif +}; + +/* The initializer is necessary to set OPTIND and OPTERR to their + default values and to clear the initialization flag. */ +#define _GETOPT_DATA_INITIALIZER { 1, 1 } + +extern int _getopt_internal_r (int ___argc, char **___argv, + const char *__shortopts, + const struct option *__longopts, int *__longind, + int __long_only, int __posixly_correct, + struct _getopt_data *__data); + +extern int _getopt_long_r (int ___argc, char **___argv, + const char *__shortopts, + const struct option *__longopts, int *__longind, + struct _getopt_data *__data); + +extern int _getopt_long_only_r (int ___argc, char **___argv, + const char *__shortopts, + const struct option *__longopts, + int *__longind, + struct _getopt_data *__data); + +#endif /* getopt_int.h */ diff --git a/gl/gettext.h b/gl/gettext.h new file mode 100644 index 00000000..9d76ec9a --- /dev/null +++ b/gl/gettext.h @@ -0,0 +1,270 @@ +/* Convenience header for conditional use of GNU <libintl.h>. + Copyright (C) 1995-1998, 2000-2002, 2004-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _LIBGETTEXT_H +#define _LIBGETTEXT_H 1 + +/* NLS can be disabled through the configure --disable-nls option. */ +#if ENABLE_NLS + +/* Get declarations of GNU message catalog functions. */ +# include <libintl.h> + +/* You can set the DEFAULT_TEXT_DOMAIN macro to specify the domain used by + the gettext() and ngettext() macros. This is an alternative to calling + textdomain(), and is useful for libraries. */ +# ifdef DEFAULT_TEXT_DOMAIN +# undef gettext +# define gettext(Msgid) \ + dgettext (DEFAULT_TEXT_DOMAIN, Msgid) +# undef ngettext +# define ngettext(Msgid1, Msgid2, N) \ + dngettext (DEFAULT_TEXT_DOMAIN, Msgid1, Msgid2, N) +# endif + +#else + +/* Solaris /usr/include/locale.h includes /usr/include/libintl.h, which + chokes if dcgettext is defined as a macro. So include it now, to make + later inclusions of <locale.h> a NOP. We don't include <libintl.h> + as well because people using "gettext.h" will not include <libintl.h>, + and also including <libintl.h> would fail on SunOS 4, whereas <locale.h> + is OK. */ +#if defined(__sun) +# include <locale.h> +#endif + +/* Many header files from the libstdc++ coming with g++ 3.3 or newer include + <libintl.h>, which chokes if dcgettext is defined as a macro. So include + it now, to make later inclusions of <libintl.h> a NOP. */ +#if defined(__cplusplus) && defined(__GNUG__) && (__GNUC__ >= 3) +# include <cstdlib> +# if (__GLIBC__ >= 2) || _GLIBCXX_HAVE_LIBINTL_H +# include <libintl.h> +# endif +#endif + +/* Disabled NLS. + The casts to 'const char *' serve the purpose of producing warnings + for invalid uses of the value returned from these functions. + On pre-ANSI systems without 'const', the config.h file is supposed to + contain "#define const". */ +# define gettext(Msgid) ((const char *) (Msgid)) +# define dgettext(Domainname, Msgid) ((void) (Domainname), gettext (Msgid)) +# define dcgettext(Domainname, Msgid, Category) \ + ((void) (Category), dgettext (Domainname, Msgid)) +# define ngettext(Msgid1, Msgid2, N) \ + ((N) == 1 \ + ? ((void) (Msgid2), (const char *) (Msgid1)) \ + : ((void) (Msgid1), (const char *) (Msgid2))) +# define dngettext(Domainname, Msgid1, Msgid2, N) \ + ((void) (Domainname), ngettext (Msgid1, Msgid2, N)) +# define dcngettext(Domainname, Msgid1, Msgid2, N, Category) \ + ((void) (Category), dngettext(Domainname, Msgid1, Msgid2, N)) +# define textdomain(Domainname) ((const char *) (Domainname)) +# define bindtextdomain(Domainname, Dirname) \ + ((void) (Domainname), (const char *) (Dirname)) +# define bind_textdomain_codeset(Domainname, Codeset) \ + ((void) (Domainname), (const char *) (Codeset)) + +#endif + +/* A pseudo function call that serves as a marker for the automated + extraction of messages, but does not call gettext(). The run-time + translation is done at a different place in the code. + The argument, String, should be a literal string. Concatenated strings + and other string expressions won't work. + The macro's expansion is not parenthesized, so that it is suitable as + initializer for static 'char[]' or 'const char[]' variables. */ +#define gettext_noop(String) String + +/* The separator between msgctxt and msgid in a .mo file. */ +#define GETTEXT_CONTEXT_GLUE "\004" + +/* Pseudo function calls, taking a MSGCTXT and a MSGID instead of just a + MSGID. MSGCTXT and MSGID must be string literals. MSGCTXT should be + short and rarely need to change. + The letter 'p' stands for 'particular' or 'special'. */ +#ifdef DEFAULT_TEXT_DOMAIN +# define pgettext(Msgctxt, Msgid) \ + pgettext_aux (DEFAULT_TEXT_DOMAIN, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, LC_MESSAGES) +#else +# define pgettext(Msgctxt, Msgid) \ + pgettext_aux (NULL, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, LC_MESSAGES) +#endif +#define dpgettext(Domainname, Msgctxt, Msgid) \ + pgettext_aux (Domainname, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, LC_MESSAGES) +#define dcpgettext(Domainname, Msgctxt, Msgid, Category) \ + pgettext_aux (Domainname, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, Category) +#ifdef DEFAULT_TEXT_DOMAIN +# define npgettext(Msgctxt, Msgid, MsgidPlural, N) \ + npgettext_aux (DEFAULT_TEXT_DOMAIN, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, MsgidPlural, N, LC_MESSAGES) +#else +# define npgettext(Msgctxt, Msgid, MsgidPlural, N) \ + npgettext_aux (NULL, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, MsgidPlural, N, LC_MESSAGES) +#endif +#define dnpgettext(Domainname, Msgctxt, Msgid, MsgidPlural, N) \ + npgettext_aux (Domainname, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, MsgidPlural, N, LC_MESSAGES) +#define dcnpgettext(Domainname, Msgctxt, Msgid, MsgidPlural, N, Category) \ + npgettext_aux (Domainname, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, MsgidPlural, N, Category) + +#ifdef __GNUC__ +__inline +#else +#ifdef __cplusplus +inline +#endif +#endif +static const char * +pgettext_aux (const char *domain, + const char *msg_ctxt_id, const char *msgid, + int category) +{ + const char *translation = dcgettext (domain, msg_ctxt_id, category); + if (translation == msg_ctxt_id) + return msgid; + else + return translation; +} + +#ifdef __GNUC__ +__inline +#else +#ifdef __cplusplus +inline +#endif +#endif +static const char * +npgettext_aux (const char *domain, + const char *msg_ctxt_id, const char *msgid, + const char *msgid_plural, unsigned long int n, + int category) +{ + const char *translation = + dcngettext (domain, msg_ctxt_id, msgid_plural, n, category); + if (translation == msg_ctxt_id || translation == msgid_plural) + return (n == 1 ? msgid : msgid_plural); + else + return translation; +} + +/* The same thing extended for non-constant arguments. Here MSGCTXT and MSGID + can be arbitrary expressions. But for string literals these macros are + less efficient than those above. */ + +#include <string.h> + +#define _LIBGETTEXT_HAVE_VARIABLE_SIZE_ARRAYS \ + (((__GNUC__ >= 3 || __GNUG__ >= 2) && !__STRICT_ANSI__) \ + /* || __STDC_VERSION__ >= 199901L */ ) + +#if !_LIBGETTEXT_HAVE_VARIABLE_SIZE_ARRAYS +#include <stdlib.h> +#endif + +#define pgettext_expr(Msgctxt, Msgid) \ + dcpgettext_expr (NULL, Msgctxt, Msgid, LC_MESSAGES) +#define dpgettext_expr(Domainname, Msgctxt, Msgid) \ + dcpgettext_expr (Domainname, Msgctxt, Msgid, LC_MESSAGES) + +#ifdef __GNUC__ +__inline +#else +#ifdef __cplusplus +inline +#endif +#endif +static const char * +dcpgettext_expr (const char *domain, + const char *msgctxt, const char *msgid, + int category) +{ + size_t msgctxt_len = strlen (msgctxt) + 1; + size_t msgid_len = strlen (msgid) + 1; + const char *translation; +#if _LIBGETTEXT_HAVE_VARIABLE_SIZE_ARRAYS + char msg_ctxt_id[msgctxt_len + msgid_len]; +#else + char buf[1024]; + char *msg_ctxt_id = + (msgctxt_len + msgid_len <= sizeof (buf) + ? buf + : (char *) malloc (msgctxt_len + msgid_len)); + if (msg_ctxt_id != NULL) +#endif + { + memcpy (msg_ctxt_id, msgctxt, msgctxt_len - 1); + msg_ctxt_id[msgctxt_len - 1] = '\004'; + memcpy (msg_ctxt_id + msgctxt_len, msgid, msgid_len); + translation = dcgettext (domain, msg_ctxt_id, category); +#if !_LIBGETTEXT_HAVE_VARIABLE_SIZE_ARRAYS + if (msg_ctxt_id != buf) + free (msg_ctxt_id); +#endif + if (translation != msg_ctxt_id) + return translation; + } + return msgid; +} + +#define npgettext_expr(Msgctxt, Msgid, MsgidPlural, N) \ + dcnpgettext_expr (NULL, Msgctxt, Msgid, MsgidPlural, N, LC_MESSAGES) +#define dnpgettext_expr(Domainname, Msgctxt, Msgid, MsgidPlural, N) \ + dcnpgettext_expr (Domainname, Msgctxt, Msgid, MsgidPlural, N, LC_MESSAGES) + +#ifdef __GNUC__ +__inline +#else +#ifdef __cplusplus +inline +#endif +#endif +static const char * +dcnpgettext_expr (const char *domain, + const char *msgctxt, const char *msgid, + const char *msgid_plural, unsigned long int n, + int category) +{ + size_t msgctxt_len = strlen (msgctxt) + 1; + size_t msgid_len = strlen (msgid) + 1; + const char *translation; +#if _LIBGETTEXT_HAVE_VARIABLE_SIZE_ARRAYS + char msg_ctxt_id[msgctxt_len + msgid_len]; +#else + char buf[1024]; + char *msg_ctxt_id = + (msgctxt_len + msgid_len <= sizeof (buf) + ? buf + : (char *) malloc (msgctxt_len + msgid_len)); + if (msg_ctxt_id != NULL) +#endif + { + memcpy (msg_ctxt_id, msgctxt, msgctxt_len - 1); + msg_ctxt_id[msgctxt_len - 1] = '\004'; + memcpy (msg_ctxt_id + msgctxt_len, msgid, msgid_len); + translation = dcngettext (domain, msg_ctxt_id, msgid_plural, n, category); +#if !_LIBGETTEXT_HAVE_VARIABLE_SIZE_ARRAYS + if (msg_ctxt_id != buf) + free (msg_ctxt_id); +#endif + if (!(translation == msg_ctxt_id || translation == msgid_plural)) + return translation; + } + return (n == 1 ? msgid : msgid_plural); +} + +#endif /* _LIBGETTEXT_H */ diff --git a/gl/inet_ntop.c b/gl/inet_ntop.c new file mode 100644 index 00000000..537b59fa --- /dev/null +++ b/gl/inet_ntop.c @@ -0,0 +1,238 @@ +/* inet_ntop.c -- convert IPv4 and IPv6 addresses from binary to text form + + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* + * Copyright (c) 1996-1999 by Internet Software Consortium. + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND INTERNET SOFTWARE CONSORTIUM DISCLAIMS + * ALL WARRANTIES WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES + * OF MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL INTERNET SOFTWARE + * CONSORTIUM BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL + * DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR + * PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS + * ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS + * SOFTWARE. + */ + +#include <config.h> + +/* Specification. */ +#include "inet_ntop.h" + +#include <stdio.h> +#include <string.h> +#include <errno.h> + +#ifndef EAFNOSUPPORT +# define EAFNOSUPPORT EINVAL +#endif + +#define NS_IN6ADDRSZ 16 +#define NS_INT16SZ 2 + +/* + * WARNING: Don't even consider trying to compile this on a system where + * sizeof(int) < 4. sizeof(int) > 4 is fine; all the world's not a VAX. + */ +typedef int verify_int_size[2 * sizeof (int) - 7]; + +static const char *inet_ntop4 (const unsigned char *src, char *dst, socklen_t size); +#if HAVE_IPV6 +static const char *inet_ntop6 (const unsigned char *src, char *dst, socklen_t size); +#endif + + +/* char * + * inet_ntop(af, src, dst, size) + * convert a network format address to presentation format. + * return: + * pointer to presentation format address (`dst'), or NULL (see errno). + * author: + * Paul Vixie, 1996. + */ +const char * +inet_ntop (int af, const void *restrict src, + char *restrict dst, socklen_t cnt) +{ + switch (af) + { +#if HAVE_IPV4 + case AF_INET: + return (inet_ntop4 (src, dst, cnt)); +#endif + +#if HAVE_IPV6 + case AF_INET6: + return (inet_ntop6 (src, dst, cnt)); +#endif + + default: + errno = EAFNOSUPPORT; + return (NULL); + } + /* NOTREACHED */ +} + +/* const char * + * inet_ntop4(src, dst, size) + * format an IPv4 address + * return: + * `dst' (as a const) + * notes: + * (1) uses no statics + * (2) takes a u_char* not an in_addr as input + * author: + * Paul Vixie, 1996. + */ +static const char * +inet_ntop4 (const unsigned char *src, char *dst, socklen_t size) +{ + char tmp[sizeof "255.255.255.255"]; + int len; + + len = sprintf (tmp, "%u.%u.%u.%u", src[0], src[1], src[2], src[3]); + if (len < 0) + return NULL; + + if (len > size) + { + errno = ENOSPC; + return NULL; + } + + return strcpy (dst, tmp); +} + +#if HAVE_IPV6 + +/* const char * + * inet_ntop6(src, dst, size) + * convert IPv6 binary address into presentation (printable) format + * author: + * Paul Vixie, 1996. + */ +static const char * +inet_ntop6 (const unsigned char *src, char *dst, socklen_t size) +{ + /* + * Note that int32_t and int16_t need only be "at least" large enough + * to contain a value of the specified size. On some systems, like + * Crays, there is no such thing as an integer variable with 16 bits. + * Keep this in mind if you think this function should have been coded + * to use pointer overlays. All the world's not a VAX. + */ + char tmp[sizeof "ffff:ffff:ffff:ffff:ffff:ffff:255.255.255.255"], *tp; + struct + { + int base, len; + } best, cur; + unsigned int words[NS_IN6ADDRSZ / NS_INT16SZ]; + int i; + + /* + * Preprocess: + * Copy the input (bytewise) array into a wordwise array. + * Find the longest run of 0x00's in src[] for :: shorthanding. + */ + memset (words, '\0', sizeof words); + for (i = 0; i < NS_IN6ADDRSZ; i += 2) + words[i / 2] = (src[i] << 8) | src[i + 1]; + best.base = -1; + cur.base = -1; + for (i = 0; i < (NS_IN6ADDRSZ / NS_INT16SZ); i++) + { + if (words[i] == 0) + { + if (cur.base == -1) + cur.base = i, cur.len = 1; + else + cur.len++; + } + else + { + if (cur.base != -1) + { + if (best.base == -1 || cur.len > best.len) + best = cur; + cur.base = -1; + } + } + } + if (cur.base != -1) + { + if (best.base == -1 || cur.len > best.len) + best = cur; + } + if (best.base != -1 && best.len < 2) + best.base = -1; + + /* + * Format the result. + */ + tp = tmp; + for (i = 0; i < (NS_IN6ADDRSZ / NS_INT16SZ); i++) + { + /* Are we inside the best run of 0x00's? */ + if (best.base != -1 && i >= best.base && i < (best.base + best.len)) + { + if (i == best.base) + *tp++ = ':'; + continue; + } + /* Are we following an initial run of 0x00s or any real hex? */ + if (i != 0) + *tp++ = ':'; + /* Is this address an encapsulated IPv4? */ + if (i == 6 && best.base == 0 && + (best.len == 6 || (best.len == 5 && words[5] == 0xffff))) + { + if (!inet_ntop4 (src + 12, tp, sizeof tmp - (tp - tmp))) + return (NULL); + tp += strlen (tp); + break; + } + { + int len = sprintf (tp, "%x", words[i]); + if (len < 0) + return NULL; + tp += len; + } + } + /* Was it a trailing run of 0x00's? */ + if (best.base != -1 && (best.base + best.len) == + (NS_IN6ADDRSZ / NS_INT16SZ)) + *tp++ = ':'; + *tp++ = '\0'; + + /* + * Check for overflow, copy, and we're done. + */ + if ((socklen_t) (tp - tmp) > size) + { + errno = ENOSPC; + return NULL; + } + + return strcpy (dst, tmp); +} + +#endif diff --git a/gl/inet_ntop.h b/gl/inet_ntop.h new file mode 100644 index 00000000..bd1e085a --- /dev/null +++ b/gl/inet_ntop.h @@ -0,0 +1,42 @@ +/* Convert internet address from internal to printable, presentable format. + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <sys/types.h> +#include <sys/socket.h> +#include <netinet/in.h> +#include <arpa/inet.h> + +/* Converts an internet address from internal format to a printable, + presentable format. + AF is an internet address family, such as AF_INET or AF_INET6. + SRC points to a 'struct in_addr' (for AF_INET) or 'struct in6_addr' + (for AF_INET6). + DST points to a buffer having room for CNT bytes. + The printable representation of the address (in numeric form, not + surrounded by [...], no reverse DNS is done) is placed in DST, and + DST is returned. If an error occurs, the return value is NULL and + errno is set. If CNT bytes are not sufficient to hold the result, + the return value is NULL and errno is set to ENOSPC. A good value + for CNT is 46. + + For more details, see the POSIX:2001 specification + <http://www.opengroup.org/susv3xsh/inet_ntop.html>. */ + +#if !HAVE_DECL_INET_NTOP +extern const char *inet_ntop (int af, const void *restrict src, + char *restrict dst, socklen_t cnt); +#endif diff --git a/gl/intprops.h b/gl/intprops.h new file mode 100644 index 00000000..34f971cb --- /dev/null +++ b/gl/intprops.h @@ -0,0 +1,78 @@ +/* intprops.h -- properties of integer types + + Copyright (C) 2001, 2002, 2003, 2004, 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <limits.h> + +/* The extra casts in the following macros work around compiler bugs, + e.g., in Cray C 5.0.3.0. */ + +/* True if the arithmetic type T is an integer type. bool counts as + an integer. */ +#define TYPE_IS_INTEGER(t) ((t) 1.5 == 1) + +/* True if negative values of the signed integer type T use two's + complement, ones' complement, or signed magnitude representation, + respectively. Much GNU code assumes two's complement, but some + people like to be portable to all possible C hosts. */ +#define TYPE_TWOS_COMPLEMENT(t) ((t) ~ (t) 0 == (t) -1) +#define TYPE_ONES_COMPLEMENT(t) ((t) ~ (t) 0 == 0) +#define TYPE_SIGNED_MAGNITUDE(t) ((t) ~ (t) 0 < (t) -1) + +/* True if the arithmetic type T is signed. */ +#define TYPE_SIGNED(t) (! ((t) 0 < (t) -1)) + +/* The maximum and minimum values for the integer type T. These + macros have undefined behavior if T is signed and has padding bits. + If this is a problem for you, please let us know how to fix it for + your host. */ +#define TYPE_MINIMUM(t) \ + ((t) (! TYPE_SIGNED (t) \ + ? (t) 0 \ + : TYPE_SIGNED_MAGNITUDE (t) \ + ? ~ (t) 0 \ + : ~ (t) 0 << (sizeof (t) * CHAR_BIT - 1))) +#define TYPE_MAXIMUM(t) \ + ((t) (! TYPE_SIGNED (t) \ + ? (t) -1 \ + : ~ (~ (t) 0 << (sizeof (t) * CHAR_BIT - 1)))) + +/* Return zero if T can be determined to be an unsigned type. + Otherwise, return 1. + When compiling with GCC, INT_STRLEN_BOUND uses this macro to obtain a + tighter bound. Otherwise, it overestimates the true bound by one byte + when applied to unsigned types of size 2, 4, 16, ... bytes. + The symbol signed_type_or_expr__ is private to this header file. */ +#if __GNUC__ >= 2 +# define signed_type_or_expr__(t) TYPE_SIGNED (__typeof__ (t)) +#else +# define signed_type_or_expr__(t) 1 +#endif + +/* Bound on length of the string representing an integer type or expression T. + Subtract 1 for the sign bit if T is signed; log10 (2.0) < 146/485; + add 1 for integer division truncation; add 1 more for a minus sign + if needed. */ +#define INT_STRLEN_BOUND(t) \ + ((sizeof (t) * CHAR_BIT - signed_type_or_expr__ (t)) * 146 / 485 \ + + signed_type_or_expr__ (t) + 1) + +/* Bound on buffer size needed to represent an integer type or expression T, + including the terminating null. */ +#define INT_BUFSIZE_BOUND(t) (INT_STRLEN_BOUND (t) + 1) diff --git a/gl/m4/absolute-header.m4 b/gl/m4/absolute-header.m4 new file mode 100644 index 00000000..c649df08 --- /dev/null +++ b/gl/m4/absolute-header.m4 @@ -0,0 +1,44 @@ +# absolute-header.m4 serial 6 +dnl Copyright (C) 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Derek Price. + +# gl_ABSOLUTE_HEADER(HEADER1 HEADER2 ...) +# --------------------------------------- +# Find the absolute name of a header file, assuming the header exists. +# If the header were sys/inttypes.h, this macro would define +# ABSOLUTE_SYS_INTTYPES_H to the `""' quoted absolute name of sys/inttypes.h +# in config.h +# (e.g. `#define ABSOLUTE_SYS_INTTYPES_H "///usr/include/sys/inttypes.h"'). +# The three "///" are to pacify Sun C 5.8, which otherwise would say +# "warning: #include of /usr/include/... may be non-portable". +# Use `""', not `<>', so that the /// cannot be confused with a C99 comment. +AC_DEFUN([gl_ABSOLUTE_HEADER], +[AC_LANG_PREPROC_REQUIRE()dnl +AC_FOREACH([gl_HEADER_NAME], [$1], + [AS_VAR_PUSHDEF([gl_absolute_header], + [gl_cv_absolute_]m4_quote(m4_defn([gl_HEADER_NAME])))dnl + AC_CACHE_CHECK([absolute name of <]m4_quote(m4_defn([gl_HEADER_NAME]))[>], + m4_quote(m4_defn([gl_absolute_header])), + [AS_VAR_PUSHDEF([ac_header_exists], + [ac_cv_header_]m4_quote(m4_defn([gl_HEADER_NAME])))dnl + AC_CHECK_HEADERS_ONCE(m4_quote(m4_defn([gl_HEADER_NAME])))dnl + if test AS_VAR_GET(ac_header_exists) = yes; then + AC_LANG_CONFTEST([AC_LANG_SOURCE([[#include <]]m4_dquote(m4_defn([gl_HEADER_NAME]))[[>]])]) +dnl eval is necessary to expand ac_cpp. +dnl Ultrix and Pyramid sh refuse to redirect output of eval, so use subshell. + AS_VAR_SET(gl_absolute_header, +[`(eval "$ac_cpp conftest.$ac_ext") 2>&AS_MESSAGE_LOG_FD | +sed -n '\#/]m4_quote(m4_defn([gl_HEADER_NAME]))[#{s#.*"\(.*/]m4_quote(m4_defn([gl_HEADER_NAME]))[\)".*#\1#;s#^/[^/]#//&#;p;q;}'`]) + fi + AS_VAR_POPDEF([ac_header_exists])dnl + ])dnl + AC_DEFINE_UNQUOTED(AS_TR_CPP([ABSOLUTE_]m4_quote(m4_defn([gl_HEADER_NAME]))), + ["AS_VAR_GET(gl_absolute_header)"], + [Define this to an absolute name of <]m4_quote(m4_defn([gl_HEADER_NAME]))[>.]) + AS_VAR_POPDEF([gl_absolute_header])dnl +])dnl +])# gl_ABSOLUTE_HEADER diff --git a/gl/m4/alloca.m4 b/gl/m4/alloca.m4 new file mode 100644 index 00000000..eb62e0e7 --- /dev/null +++ b/gl/m4/alloca.m4 @@ -0,0 +1,50 @@ +# alloca.m4 serial 7 +dnl Copyright (C) 2002-2004, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_FUNC_ALLOCA], +[ + dnl Work around a bug of AC_EGREP_CPP in autoconf-2.57. + AC_REQUIRE([AC_PROG_CPP]) + AC_REQUIRE([AC_PROG_EGREP]) + + AC_REQUIRE([AC_FUNC_ALLOCA]) + if test $ac_cv_func_alloca_works = no; then + gl_PREREQ_ALLOCA + fi + + # Define an additional variable used in the Makefile substitution. + if test $ac_cv_working_alloca_h = yes; then + AC_CACHE_CHECK([for alloca as a compiler built-in], [gl_cv_rpl_alloca], [ + AC_EGREP_CPP([Need own alloca], [ +#if defined __GNUC__ || defined _AIX || defined _MSC_VER + Need own alloca +#endif + ], [gl_cv_rpl_alloca=yes], [gl_cv_rpl_alloca=no]) + ]) + if test $gl_cv_rpl_alloca = yes; then + dnl OK, alloca can be implemented through a compiler built-in. + AC_DEFINE([HAVE_ALLOCA], 1, + [Define to 1 if you have 'alloca' after including <alloca.h>, + a header that may be supplied by this distribution.]) + ALLOCA_H=alloca.h + else + dnl alloca exists as a library function, i.e. it is slow and probably + dnl a memory leak. Don't define HAVE_ALLOCA in this case. + ALLOCA_H= + fi + else + ALLOCA_H=alloca.h + fi + AC_SUBST([ALLOCA_H]) + + AC_DEFINE(HAVE_ALLOCA_H, 1, + [Define HAVE_ALLOCA_H for backward compatibility with older code + that includes <alloca.h> only if HAVE_ALLOCA_H is defined.]) +]) + +# Prerequisites of lib/alloca.c. +# STACK_DIRECTION is already handled by AC_FUNC_ALLOCA. +AC_DEFUN([gl_PREREQ_ALLOCA], [:]) diff --git a/gl/m4/arpa_inet_h.m4 b/gl/m4/arpa_inet_h.m4 new file mode 100644 index 00000000..d01d0984 --- /dev/null +++ b/gl/m4/arpa_inet_h.m4 @@ -0,0 +1,18 @@ +# arpa_inet_h.m4 serial 1 +dnl Copyright (C) 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl Written by Simon Josefsson + +AC_DEFUN([gl_HEADER_ARPA_INET], +[ + AC_CHECK_HEADERS_ONCE([arpa/inet.h]) + if test $ac_cv_header_arpa_inet_h = yes; then + ARPA_INET_H='' + else + ARPA_INET_H='arpa/inet.h' + fi + AC_SUBST(ARPA_INET_H) +]) diff --git a/gl/m4/c-strtod.m4 b/gl/m4/c-strtod.m4 new file mode 100644 index 00000000..7f206c71 --- /dev/null +++ b/gl/m4/c-strtod.m4 @@ -0,0 +1,53 @@ +# c-strtod.m4 serial 9 + +# Copyright (C) 2004, 2005, 2006 Free Software Foundation, Inc. +# This file is free software; the Free Software Foundation +# gives unlimited permission to copy and/or distribute it, +# with or without modifications, as long as this notice is preserved. + +# Written by Paul Eggert. + +AC_DEFUN([gl_C99_STRTOLD], +[ + AC_CACHE_CHECK([whether strtold conforms to C99], + [gl_cv_func_c99_strtold], + [AC_LINK_IFELSE( + [AC_LANG_PROGRAM( + [[/* On HP-UX before 11.23, strtold returns a struct instead of + long double. Reject implementations like that, by requiring + compatibility with the C99 prototype. */ + #include <stdlib.h> + static long double (*p) (char const *, char **) = strtold; + static long double + test (char const *nptr, char **endptr) + { + long double r; + r = strtold (nptr, endptr); + return r; + }]], + [[return test ("1.0", NULL) != 1 || p ("1.0", NULL) != 1;]])], + [gl_cv_func_c99_strtold=yes], + [gl_cv_func_c99_strtold=no])]) + if test $gl_cv_func_c99_strtold = yes; then + AC_DEFINE([HAVE_C99_STRTOLD], 1, [Define to 1 if strtold conforms to C99.]) + fi +]) + +AC_DEFUN([gl_C_STRTOD], +[ + AC_LIBOBJ([c-strtod]) + + dnl Prerequisites of lib/c-strtod.c. + AC_REQUIRE([gl_USE_SYSTEM_EXTENSIONS]) + : +]) + +AC_DEFUN([gl_C_STRTOLD], +[ + AC_LIBOBJ([c-strtold]) + + dnl Prerequisites of lib/c-strtold.c. + AC_REQUIRE([gl_USE_SYSTEM_EXTENSIONS]) + AC_REQUIRE([gl_C99_STRTOLD]) + : +]) diff --git a/gl/m4/cloexec.m4 b/gl/m4/cloexec.m4 new file mode 100644 index 00000000..4c4e26a1 --- /dev/null +++ b/gl/m4/cloexec.m4 @@ -0,0 +1,10 @@ +#serial 6 +dnl Copyright (C) 2004, 2005, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_CLOEXEC], +[ + AC_LIBOBJ([cloexec]) +]) diff --git a/gl/m4/codeset.m4 b/gl/m4/codeset.m4 new file mode 100644 index 00000000..223955b4 --- /dev/null +++ b/gl/m4/codeset.m4 @@ -0,0 +1,21 @@ +# codeset.m4 serial 2 (gettext-0.16) +dnl Copyright (C) 2000-2002, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Bruno Haible. + +AC_DEFUN([AM_LANGINFO_CODESET], +[ + AC_CACHE_CHECK([for nl_langinfo and CODESET], am_cv_langinfo_codeset, + [AC_TRY_LINK([#include <langinfo.h>], + [char* cs = nl_langinfo(CODESET); return !cs;], + am_cv_langinfo_codeset=yes, + am_cv_langinfo_codeset=no) + ]) + if test $am_cv_langinfo_codeset = yes; then + AC_DEFINE(HAVE_LANGINFO_CODESET, 1, + [Define if you have <langinfo.h> and nl_langinfo(CODESET).]) + fi +]) diff --git a/gl/m4/dirname.m4 b/gl/m4/dirname.m4 new file mode 100644 index 00000000..e35da965 --- /dev/null +++ b/gl/m4/dirname.m4 @@ -0,0 +1,18 @@ +#serial 7 -*- autoconf -*- +dnl Copyright (C) 2002-2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_DIRNAME], +[ + AC_LIBOBJ([basename]) + AC_LIBOBJ([dirname]) + AC_LIBOBJ([stripslash]) + + dnl Prerequisites of lib/dirname.h. + AC_REQUIRE([gl_AC_DOS]) + AC_REQUIRE([gl_DOUBLE_SLASH_ROOT]) + + dnl No prerequisites of lib/basename.c, lib/dirname.c, lib/stripslash.c. +]) diff --git a/gl/m4/dos.m4 b/gl/m4/dos.m4 new file mode 100644 index 00000000..dd59571c --- /dev/null +++ b/gl/m4/dos.m4 @@ -0,0 +1,71 @@ +#serial 10 -*- autoconf -*- + +# Define some macros required for proper operation of code in lib/*.c +# on MSDOS/Windows systems. + +# Copyright (C) 2000, 2001, 2004, 2005, 2006 Free Software Foundation, Inc. +# This file is free software; the Free Software Foundation +# gives unlimited permission to copy and/or distribute it, +# with or without modifications, as long as this notice is preserved. + +# From Jim Meyering. + +AC_DEFUN([gl_AC_DOS], + [ + AC_CACHE_CHECK([whether system is Windows or MSDOS], [ac_cv_win_or_dos], + [ + AC_TRY_COMPILE([], + [#if !defined _WIN32 && !defined __WIN32__ && !defined __MSDOS__ && !defined __CYGWIN__ +neither MSDOS nor Windows +#endif], + [ac_cv_win_or_dos=yes], + [ac_cv_win_or_dos=no]) + ]) + + if test x"$ac_cv_win_or_dos" = xyes; then + ac_fs_accepts_drive_letter_prefix=1 + ac_fs_backslash_is_file_name_separator=1 + AC_CACHE_CHECK([whether drive letter can start relative path], + [ac_cv_drive_letter_can_be_relative], + [ + AC_TRY_COMPILE([], + [#if defined __CYGWIN__ +drive letters are always absolute +#endif], + [ac_cv_drive_letter_can_be_relative=yes], + [ac_cv_drive_letter_can_be_relative=no]) + ]) + if test x"$ac_cv_drive_letter_can_be_relative" = xyes; then + ac_fs_drive_letter_can_be_relative=1 + else + ac_fs_drive_letter_can_be_relative=0 + fi + else + ac_fs_accepts_drive_letter_prefix=0 + ac_fs_backslash_is_file_name_separator=0 + ac_fs_drive_letter_can_be_relative=0 + fi + + AC_DEFINE_UNQUOTED([FILE_SYSTEM_ACCEPTS_DRIVE_LETTER_PREFIX], + $ac_fs_accepts_drive_letter_prefix, + [Define on systems for which file names may have a so-called + `drive letter' prefix, define this to compute the length of that + prefix, including the colon.]) + + AH_VERBATIM(ISSLASH, + [#if FILE_SYSTEM_BACKSLASH_IS_FILE_NAME_SEPARATOR +# define ISSLASH(C) ((C) == '/' || (C) == '\\') +#else +# define ISSLASH(C) ((C) == '/') +#endif]) + + AC_DEFINE_UNQUOTED([FILE_SYSTEM_BACKSLASH_IS_FILE_NAME_SEPARATOR], + $ac_fs_backslash_is_file_name_separator, + [Define if the backslash character may also serve as a file name + component separator.]) + + AC_DEFINE_UNQUOTED([FILE_SYSTEM_DRIVE_PREFIX_CAN_BE_RELATIVE], + $ac_fs_drive_letter_can_be_relative, + [Define if a drive letter prefix denotes a relative path if it is + not followed by a file name component separator.]) + ]) diff --git a/gl/m4/double-slash-root.m4 b/gl/m4/double-slash-root.m4 new file mode 100644 index 00000000..69d60d01 --- /dev/null +++ b/gl/m4/double-slash-root.m4 @@ -0,0 +1,38 @@ +# double-slash-root.m4 serial 2 -*- Autoconf -*- +dnl Copyright (C) 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_DOUBLE_SLASH_ROOT], +[ + AC_REQUIRE([AC_CANONICAL_HOST]) + AC_CACHE_CHECK([whether // is distinct from /], [gl_cv_double_slash_root], + [ if test x"$cross_compiling" = xyes ; then + # When cross-compiling, there is no way to tell whether // is special + # short of a list of hosts. However, the only known hosts to date + # that have a distinct // are Apollo DomainOS (too old to port to), + # Cygwin, and z/OS. If anyone knows of another system for which // has + # special semantics and is distinct from /, please report it to + # <bug-gnulib@gnu.org>. + case $host in + *-cygwin | i370-ibm-openedition) + gl_cv_double_slash_root=yes ;; + *) + # Be optimistic and assume that / and // are the same when we + # don't know. + gl_cv_double_slash_root='unknown, assuming no' ;; + esac + else + set x `ls -di / //` + if test $[2] = $[4] && wc //dev/null >/dev/null 2>&1; then + gl_cv_double_slash_root=no + else + gl_cv_double_slash_root=yes + fi + fi]) + if test "$gl_cv_double_slash_root" = yes; then + AC_DEFINE([DOUBLE_SLASH_IS_DISTINCT_ROOT], 1, + [Define to 1 if // is a file system root distinct from /.]) + fi +]) diff --git a/gl/m4/eoverflow.m4 b/gl/m4/eoverflow.m4 new file mode 100644 index 00000000..3bffd10e --- /dev/null +++ b/gl/m4/eoverflow.m4 @@ -0,0 +1,70 @@ +# eoverflow.m4 serial 2 +dnl Copyright (C) 2004, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Bruno Haible. + +# The EOVERFLOW errno value ought to be defined in <errno.h>, according to +# POSIX. But some systems (like AIX 3) don't define it, and some systems +# (like OSF/1) define it when _XOPEN_SOURCE_EXTENDED is defined. + +# Define EOVERFLOW as a C macro and as a substituted macro in such a way that +# 1. on all systems, after inclusion of <errno.h>, EOVERFLOW is usable, +# 2. on systems where EOVERFLOW is defined elsewhere, we use the same numeric +# value. + +AC_DEFUN([gl_EOVERFLOW], +[ + AC_REQUIRE([AC_PROG_CC])dnl + + AC_CACHE_CHECK([for EOVERFLOW], ac_cv_decl_EOVERFLOW, [ + AC_EGREP_CPP(yes,[ +#include <errno.h> +#ifdef EOVERFLOW +yes +#endif + ], have_eoverflow=1) + if test -n "$have_eoverflow"; then + dnl EOVERFLOW exists in <errno.h>. Don't need to define EOVERFLOW ourselves. + ac_cv_decl_EOVERFLOW=yes + else + AC_EGREP_CPP(yes,[ +#define _XOPEN_SOURCE_EXTENDED 1 +#include <errno.h> +#ifdef EOVERFLOW +yes +#endif + ], have_eoverflow=1) + if test -n "$have_eoverflow"; then + dnl EOVERFLOW exists but is hidden. + dnl Define it to the same value. + AC_COMPUTE_INT([ac_cv_decl_EOVERFLOW], [EOVERFLOW], [ +#define _XOPEN_SOURCE_EXTENDED 1 +#include <errno.h> +/* The following two lines are a workaround against an autoconf-2.52 bug. */ +#include <stdio.h> +#include <stdlib.h> +]) + else + dnl EOVERFLOW isn't defined by the system. Define EOVERFLOW ourselves, but + dnl don't define it as EINVAL, because snprintf() callers want to + dnl distinguish EINVAL and EOVERFLOW. + ac_cv_decl_EOVERFLOW=E2BIG + fi + fi + ]) + if test "$ac_cv_decl_EOVERFLOW" != yes; then + AC_DEFINE_UNQUOTED([EOVERFLOW], [$ac_cv_decl_EOVERFLOW], + [Define as good substitute value for EOVERFLOW.]) + EOVERFLOW="$ac_cv_decl_EOVERFLOW" + AC_SUBST(EOVERFLOW) + fi +]) + +dnl Autoconf >= 2.61 has AC_COMPUTE_INT built-in. +dnl Remove this when we can assume autoconf >= 2.61. +m4_ifdef([AC_COMPUTE_INT], [], [ + AC_DEFUN([AC_COMPUTE_INT], [_AC_COMPUTE_INT([$2],[$1],[$3],[$4])]) +]) diff --git a/gl/m4/error.m4 b/gl/m4/error.m4 new file mode 100644 index 00000000..7c7746e2 --- /dev/null +++ b/gl/m4/error.m4 @@ -0,0 +1,22 @@ +#serial 11 + +# Copyright (C) 1996, 1997, 1998, 2001, 2002, 2003, 2004 Free Software +# Foundation, Inc. +# +# This file is free software; the Free Software Foundation +# gives unlimited permission to copy and/or distribute it, +# with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_ERROR], +[ + AC_FUNC_ERROR_AT_LINE + dnl Note: AC_FUNC_ERROR_AT_LINE does AC_LIBSOURCES([error.h, error.c]). + gl_PREREQ_ERROR +]) + +# Prerequisites of lib/error.c. +AC_DEFUN([gl_PREREQ_ERROR], +[ + AC_REQUIRE([AC_FUNC_STRERROR_R]) + : +]) diff --git a/gl/m4/exitfail.m4 b/gl/m4/exitfail.m4 new file mode 100644 index 00000000..b7a691e5 --- /dev/null +++ b/gl/m4/exitfail.m4 @@ -0,0 +1,13 @@ +# exitfail.m4 serial 6 +dnl Copyright (C) 2002, 2003, 2005, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_EXITFAIL], +[ + AC_LIBOBJ([exitfail]) + + dnl No prerequisites of lib/exitfail.c. + : +]) diff --git a/gl/m4/extensions.m4 b/gl/m4/extensions.m4 new file mode 100644 index 00000000..143a9e54 --- /dev/null +++ b/gl/m4/extensions.m4 @@ -0,0 +1,58 @@ +# serial 4 -*- Autoconf -*- +# Enable extensions on systems that normally disable them. + +# Copyright (C) 2003, 2006 Free Software Foundation, Inc. +# This file is free software; the Free Software Foundation +# gives unlimited permission to copy and/or distribute it, +# with or without modifications, as long as this notice is preserved. + +# This definition of AC_USE_SYSTEM_EXTENSIONS is stolen from CVS +# Autoconf. Perhaps we can remove this once we can assume Autoconf +# 2.61 or later everywhere, but since CVS Autoconf mutates rapidly +# enough in this area it's likely we'll need to redefine +# AC_USE_SYSTEM_EXTENSIONS for quite some time. + +# AC_USE_SYSTEM_EXTENSIONS +# ------------------------ +# Enable extensions on systems that normally disable them, +# typically due to standards-conformance issues. +AC_DEFUN([AC_USE_SYSTEM_EXTENSIONS], +[ + AC_BEFORE([$0], [AC_COMPILE_IFELSE]) + AC_BEFORE([$0], [AC_RUN_IFELSE]) + + AC_REQUIRE([AC_GNU_SOURCE]) + AC_REQUIRE([AC_AIX]) + AC_REQUIRE([AC_MINIX]) + + AH_VERBATIM([__EXTENSIONS__], +[/* Enable extensions on Solaris. */ +#ifndef __EXTENSIONS__ +# undef __EXTENSIONS__ +#endif +#ifndef _POSIX_PTHREAD_SEMANTICS +# undef _POSIX_PTHREAD_SEMANTICS +#endif +#ifndef _TANDEM_SOURCE +# undef _TANDEM_SOURCE +#endif]) + AC_CACHE_CHECK([whether it is safe to define __EXTENSIONS__], + [ac_cv_safe_to_define___extensions__], + [AC_COMPILE_IFELSE( + [AC_LANG_PROGRAM([ +# define __EXTENSIONS__ 1 + AC_INCLUDES_DEFAULT])], + [ac_cv_safe_to_define___extensions__=yes], + [ac_cv_safe_to_define___extensions__=no])]) + test $ac_cv_safe_to_define___extensions__ = yes && + AC_DEFINE([__EXTENSIONS__]) + AC_DEFINE([_POSIX_PTHREAD_SEMANTICS]) + AC_DEFINE([_TANDEM_SOURCE]) +]) + +# gl_USE_SYSTEM_EXTENSIONS +# ------------------------ +# Enable extensions on systems that normally disable them, +# typically due to standards-conformance issues. +AC_DEFUN([gl_USE_SYSTEM_EXTENSIONS], + [AC_REQUIRE([AC_USE_SYSTEM_EXTENSIONS])]) diff --git a/gl/m4/fcntl-safer.m4 b/gl/m4/fcntl-safer.m4 new file mode 100644 index 00000000..3475b0a7 --- /dev/null +++ b/gl/m4/fcntl-safer.m4 @@ -0,0 +1,11 @@ +#serial 5 +dnl Copyright (C) 2005-2007 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_FCNTL_SAFER], +[ + AC_LIBOBJ([open-safer]) + AC_LIBOBJ([creat-safer]) +]) diff --git a/gl/m4/fstypename.m4 b/gl/m4/fstypename.m4 new file mode 100644 index 00000000..aa676f3a --- /dev/null +++ b/gl/m4/fstypename.m4 @@ -0,0 +1,22 @@ +#serial 6 + +dnl From Jim Meyering. +dnl +dnl See if struct statfs has the f_fstypename member. +dnl If so, define HAVE_STRUCT_STATFS_F_FSTYPENAME. +dnl + +# Copyright (C) 1998, 1999, 2001, 2004, 2006 Free Software Foundation, Inc. +# This file is free software; the Free Software Foundation +# gives unlimited permission to copy and/or distribute it, +# with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_FSTYPENAME], +[ + AC_CHECK_MEMBERS([struct statfs.f_fstypename],,, + [ + #include <sys/types.h> + #include <sys/param.h> + #include <sys/mount.h> + ]) +]) diff --git a/gl/m4/fsusage.m4 b/gl/m4/fsusage.m4 new file mode 100644 index 00000000..08bf06c9 --- /dev/null +++ b/gl/m4/fsusage.m4 @@ -0,0 +1,265 @@ +#serial 22 +# Obtaining file system usage information. + +# Copyright (C) 1997, 1998, 2000, 2001, 2003, 2004, 2005, 2006 Free Software +# Foundation, Inc. +# +# This file is free software; the Free Software Foundation +# gives unlimited permission to copy and/or distribute it, +# with or without modifications, as long as this notice is preserved. + +# Written by Jim Meyering. + +AC_DEFUN([gl_FSUSAGE], +[ + AC_CHECK_HEADERS_ONCE(sys/param.h) + AC_CHECK_HEADERS_ONCE(sys/vfs.h sys/fs_types.h) + AC_CHECK_HEADERS(sys/mount.h, [], [], + [AC_INCLUDES_DEFAULT + [#if HAVE_SYS_PARAM_H + #include <sys/param.h> + #endif]]) + gl_FILE_SYSTEM_USAGE([gl_cv_fs_space=yes], [gl_cv_fs_space=no]) + if test $gl_cv_fs_space = yes; then + AC_LIBOBJ(fsusage) + gl_PREREQ_FSUSAGE_EXTRA + fi +]) + +# Try to determine how a program can obtain file system usage information. +# If successful, define the appropriate symbol (see fsusage.c) and +# execute ACTION-IF-FOUND. Otherwise, execute ACTION-IF-NOT-FOUND. +# +# gl_FILE_SYSTEM_USAGE([ACTION-IF-FOUND[, ACTION-IF-NOT-FOUND]]) + +AC_DEFUN([gl_FILE_SYSTEM_USAGE], +[ + +AC_MSG_NOTICE([checking how to get file system space usage]) +ac_fsusage_space=no + +# Perform only the link test since it seems there are no variants of the +# statvfs function. This check is more than just AC_CHECK_FUNCS(statvfs) +# because that got a false positive on SCO OSR5. Adding the declaration +# of a `struct statvfs' causes this test to fail (as it should) on such +# systems. That system is reported to work fine with STAT_STATFS4 which +# is what it gets when this test fails. +if test $ac_fsusage_space = no; then + # SVR4 + AC_CACHE_CHECK([for statvfs function (SVR4)], fu_cv_sys_stat_statvfs, + [AC_TRY_LINK([#include <sys/types.h> +#if defined __GLIBC__ && !defined __BEOS__ +Do not use statvfs on systems with GNU libc, because that function stats +all preceding entries in /proc/mounts, and that makes df hang if even +one of the corresponding file systems is hard-mounted, but not available. +statvfs in GNU libc on BeOS operates differently: it only makes a system +call. +#endif +#include <sys/statvfs.h>], + [struct statvfs fsd; statvfs (0, &fsd);], + fu_cv_sys_stat_statvfs=yes, + fu_cv_sys_stat_statvfs=no)]) + if test $fu_cv_sys_stat_statvfs = yes; then + ac_fsusage_space=yes + AC_DEFINE(STAT_STATVFS, 1, + [ Define if there is a function named statvfs. (SVR4)]) + fi +fi + +if test $ac_fsusage_space = no; then + # DEC Alpha running OSF/1 + AC_MSG_CHECKING([for 3-argument statfs function (DEC OSF/1)]) + AC_CACHE_VAL(fu_cv_sys_stat_statfs3_osf1, + [AC_TRY_RUN([ +#include <sys/param.h> +#include <sys/types.h> +#include <sys/mount.h> + int + main () + { + struct statfs fsd; + fsd.f_fsize = 0; + return statfs (".", &fsd, sizeof (struct statfs)) != 0; + }], + fu_cv_sys_stat_statfs3_osf1=yes, + fu_cv_sys_stat_statfs3_osf1=no, + fu_cv_sys_stat_statfs3_osf1=no)]) + AC_MSG_RESULT($fu_cv_sys_stat_statfs3_osf1) + if test $fu_cv_sys_stat_statfs3_osf1 = yes; then + ac_fsusage_space=yes + AC_DEFINE(STAT_STATFS3_OSF1, 1, + [ Define if statfs takes 3 args. (DEC Alpha running OSF/1)]) + fi +fi + +if test $ac_fsusage_space = no; then +# AIX + AC_MSG_CHECKING([for two-argument statfs with statfs.bsize dnl +member (AIX, 4.3BSD)]) + AC_CACHE_VAL(fu_cv_sys_stat_statfs2_bsize, + [AC_TRY_RUN([ +#ifdef HAVE_SYS_PARAM_H +#include <sys/param.h> +#endif +#ifdef HAVE_SYS_MOUNT_H +#include <sys/mount.h> +#endif +#ifdef HAVE_SYS_VFS_H +#include <sys/vfs.h> +#endif + int + main () + { + struct statfs fsd; + fsd.f_bsize = 0; + return statfs (".", &fsd) != 0; + }], + fu_cv_sys_stat_statfs2_bsize=yes, + fu_cv_sys_stat_statfs2_bsize=no, + fu_cv_sys_stat_statfs2_bsize=no)]) + AC_MSG_RESULT($fu_cv_sys_stat_statfs2_bsize) + if test $fu_cv_sys_stat_statfs2_bsize = yes; then + ac_fsusage_space=yes + AC_DEFINE(STAT_STATFS2_BSIZE, 1, +[ Define if statfs takes 2 args and struct statfs has a field named f_bsize. + (4.3BSD, SunOS 4, HP-UX, AIX PS/2)]) + fi +fi + +if test $ac_fsusage_space = no; then +# SVR3 + AC_MSG_CHECKING([for four-argument statfs (AIX-3.2.5, SVR3)]) + AC_CACHE_VAL(fu_cv_sys_stat_statfs4, + [AC_TRY_RUN([#include <sys/types.h> +#include <sys/statfs.h> + int + main () + { + struct statfs fsd; + return statfs (".", &fsd, sizeof fsd, 0) != 0; + }], + fu_cv_sys_stat_statfs4=yes, + fu_cv_sys_stat_statfs4=no, + fu_cv_sys_stat_statfs4=no)]) + AC_MSG_RESULT($fu_cv_sys_stat_statfs4) + if test $fu_cv_sys_stat_statfs4 = yes; then + ac_fsusage_space=yes + AC_DEFINE(STAT_STATFS4, 1, + [ Define if statfs takes 4 args. (SVR3, Dynix, Irix, Dolphin)]) + fi +fi + +if test $ac_fsusage_space = no; then +# 4.4BSD and NetBSD + AC_MSG_CHECKING([for two-argument statfs with statfs.fsize dnl +member (4.4BSD and NetBSD)]) + AC_CACHE_VAL(fu_cv_sys_stat_statfs2_fsize, + [AC_TRY_RUN([#include <sys/types.h> +#ifdef HAVE_SYS_PARAM_H +#include <sys/param.h> +#endif +#ifdef HAVE_SYS_MOUNT_H +#include <sys/mount.h> +#endif + int + main () + { + struct statfs fsd; + fsd.f_fsize = 0; + return statfs (".", &fsd) != 0; + }], + fu_cv_sys_stat_statfs2_fsize=yes, + fu_cv_sys_stat_statfs2_fsize=no, + fu_cv_sys_stat_statfs2_fsize=no)]) + AC_MSG_RESULT($fu_cv_sys_stat_statfs2_fsize) + if test $fu_cv_sys_stat_statfs2_fsize = yes; then + ac_fsusage_space=yes + AC_DEFINE(STAT_STATFS2_FSIZE, 1, +[ Define if statfs takes 2 args and struct statfs has a field named f_fsize. + (4.4BSD, NetBSD)]) + fi +fi + +if test $ac_fsusage_space = no; then + # Ultrix + AC_MSG_CHECKING([for two-argument statfs with struct fs_data (Ultrix)]) + AC_CACHE_VAL(fu_cv_sys_stat_fs_data, + [AC_TRY_RUN([#include <sys/types.h> +#ifdef HAVE_SYS_PARAM_H +#include <sys/param.h> +#endif +#ifdef HAVE_SYS_MOUNT_H +#include <sys/mount.h> +#endif +#ifdef HAVE_SYS_FS_TYPES_H +#include <sys/fs_types.h> +#endif + int + main () + { + struct fs_data fsd; + /* Ultrix's statfs returns 1 for success, + 0 for not mounted, -1 for failure. */ + return statfs (".", &fsd) != 1; + }], + fu_cv_sys_stat_fs_data=yes, + fu_cv_sys_stat_fs_data=no, + fu_cv_sys_stat_fs_data=no)]) + AC_MSG_RESULT($fu_cv_sys_stat_fs_data) + if test $fu_cv_sys_stat_fs_data = yes; then + ac_fsusage_space=yes + AC_DEFINE(STAT_STATFS2_FS_DATA, 1, +[ Define if statfs takes 2 args and the second argument has + type struct fs_data. (Ultrix)]) + fi +fi + +if test $ac_fsusage_space = no; then + # SVR2 + AC_TRY_CPP([#include <sys/filsys.h> + ], + AC_DEFINE(STAT_READ_FILSYS, 1, + [Define if there is no specific function for reading file systems usage + information and you have the <sys/filsys.h> header file. (SVR2)]) + ac_fsusage_space=yes) +fi + +AS_IF([test $ac_fsusage_space = yes], [$1], [$2]) + +]) + + +# Check for SunOS statfs brokenness wrt partitions 2GB and larger. +# If <sys/vfs.h> exists and struct statfs has a member named f_spare, +# enable the work-around code in fsusage.c. +AC_DEFUN([gl_STATFS_TRUNCATES], +[ + AC_MSG_CHECKING([for statfs that truncates block counts]) + AC_CACHE_VAL(fu_cv_sys_truncating_statfs, + [AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[ +#if !defined(sun) && !defined(__sun) +choke -- this is a workaround for a Sun-specific problem +#endif +#include <sys/types.h> +#include <sys/vfs.h>]], + [[struct statfs t; long c = *(t.f_spare); + if (c) return 0;]])], + [fu_cv_sys_truncating_statfs=yes], + [fu_cv_sys_truncating_statfs=no])]) + if test $fu_cv_sys_truncating_statfs = yes; then + AC_DEFINE(STATFS_TRUNCATES_BLOCK_COUNTS, 1, + [Define if the block counts reported by statfs may be truncated to 2GB + and the correct values may be stored in the f_spare array. + (SunOS 4.1.2, 4.1.3, and 4.1.3_U1 are reported to have this problem. + SunOS 4.1.1 seems not to be affected.)]) + fi + AC_MSG_RESULT($fu_cv_sys_truncating_statfs) +]) + + +# Prerequisites of lib/fsusage.c not done by gl_FILE_SYSTEM_USAGE. +AC_DEFUN([gl_PREREQ_FSUSAGE_EXTRA], +[ + AC_CHECK_HEADERS(dustat.h sys/fs/s5param.h sys/filsys.h sys/statfs.h) + gl_STATFS_TRUNCATES +]) diff --git a/gl/m4/getaddrinfo.m4 b/gl/m4/getaddrinfo.m4 new file mode 100644 index 00000000..db285d94 --- /dev/null +++ b/gl/m4/getaddrinfo.m4 @@ -0,0 +1,88 @@ +# getaddrinfo.m4 serial 11 +dnl Copyright (C) 2004, 2005, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_GETADDRINFO], +[ + AC_MSG_NOTICE([checking how to do getaddrinfo, freeaddrinfo and getnameinfo]) + + AC_SEARCH_LIBS(getaddrinfo, [nsl socket]) + AC_CHECK_FUNCS(getaddrinfo,, [ + AC_CACHE_CHECK(for getaddrinfo in ws2tcpip.h and -lws2_32, + gl_cv_w32_getaddrinfo, [ + gl_cv_w32_getaddrinfo=no + am_save_LIBS="$LIBS" + LIBS="$LIBS -lws2_32" + AC_TRY_LINK([ +#ifdef HAVE_WS2TCPIP_H +#include <ws2tcpip.h> +#endif +], [getaddrinfo(0, 0, 0, 0);], gl_cv_w32_getaddrinfo=yes) + LIBS="$am_save_LIBS"]) + if test "$gl_cv_w32_getaddrinfo" = "yes"; then + LIBS="$LIBS -lws2_32" + else + AC_LIBOBJ(getaddrinfo) + fi + ]) + + AC_REPLACE_FUNCS(gai_strerror) + gl_PREREQ_GETADDRINFO +]) + +# Prerequisites of lib/getaddrinfo.h and lib/getaddrinfo.c. +AC_DEFUN([gl_PREREQ_GETADDRINFO], [ + AC_SEARCH_LIBS(gethostbyname, [inet nsl]) + AC_SEARCH_LIBS(getservbyname, [inet nsl socket xnet]) + AC_CHECK_FUNCS(gethostbyname,, [ + AC_CACHE_CHECK(for gethostbyname in winsock2.h and -lws2_32, + gl_cv_w32_gethostbyname, [ + gl_cv_w32_gethostbyname=no + am_save_LIBS="$LIBS" + LIBS="$LIBS -lws2_32" + AC_TRY_LINK([ +#ifdef HAVE_WINSOCK2_H +#include <winsock2.h> +#endif +], [gethostbyname(0);], gl_cv_w32_gethostbyname=yes) + LIBS="$am_save_LIBS"]) + if test "$gl_cv_w32_gethostbyname" = "yes"; then + LIBS="$LIBS -lws2_32" + fi + ]) + AC_REQUIRE([AC_C_RESTRICT]) + AC_REQUIRE([gl_SOCKET_FAMILIES]) + AC_REQUIRE([gl_HEADER_SYS_SOCKET]) + AC_REQUIRE([AC_C_INLINE]) + AC_REQUIRE([AC_GNU_SOURCE]) + AC_CHECK_HEADERS_ONCE(netinet/in.h netdb.h) + AC_CHECK_DECLS([getaddrinfo, freeaddrinfo, gai_strerror, getnameinfo],,,[ + /* sys/types.h is not needed according to POSIX, but the + sys/socket.h in i386-unknown-freebsd4.10 and + powerpc-apple-darwin5.5 required it. */ +#include <sys/types.h> +#ifdef HAVE_SYS_SOCKET_H +#include <sys/socket.h> +#endif +#ifdef HAVE_NETDB_H +#include <netdb.h> +#endif +#ifdef HAVE_WS2TCPIP_H +#include <ws2tcpip.h> +#endif +]) + AC_CHECK_TYPES([struct addrinfo],,,[ +#include <sys/types.h> +#ifdef HAVE_SYS_SOCKET_H +#include <sys/socket.h> +#endif +#ifdef HAVE_NETDB_H +#include <netdb.h> +#endif +#ifdef HAVE_WS2TCPIP_H +#include <ws2tcpip.h> +#endif +]) +]) diff --git a/gl/m4/gethostname.m4 b/gl/m4/gethostname.m4 new file mode 100644 index 00000000..1e9749d3 --- /dev/null +++ b/gl/m4/gethostname.m4 @@ -0,0 +1,18 @@ +# gethostname.m4 serial 2 +dnl Copyright (C) 2002 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_FUNC_GETHOSTNAME], +[ + AC_REPLACE_FUNCS(gethostname) + if test $ac_cv_func_gethostname = no; then + gl_PREREQ_GETHOSTNAME + fi +]) + +# Prerequisites of lib/gethostname.c. +AC_DEFUN([gl_PREREQ_GETHOSTNAME], [ + AC_CHECK_FUNCS(uname) +]) diff --git a/gl/m4/getloadavg.m4 b/gl/m4/getloadavg.m4 new file mode 100644 index 00000000..82437291 --- /dev/null +++ b/gl/m4/getloadavg.m4 @@ -0,0 +1,155 @@ +# Check for getloadavg. + +# Copyright (C) 1992, 1993, 1994, 1995, 1996, 1999, 2000, 2002, 2003, +# 2006 Free Software Foundation, Inc. + +# This file is free software; the Free Software Foundation +# gives unlimited permission to copy and/or distribute it, +# with or without modifications, as long as this notice is preserved. + +# Autoconf defines AC_FUNC_GETLOADAVG, but that is obsolescent. +# New applications should use gl_GETLOADAVG instead. + +# gl_GETLOADAVG(LIBOBJDIR) +# ------------------------ +AC_DEFUN([gl_GETLOADAVG], +[gl_have_func=no # yes means we've found a way to get the load average. + +# Make sure getloadavg.c is where it belongs, at configure-time. +test -f "$srcdir/$1/getloadavg.c" || + AC_MSG_ERROR([$srcdir/$1/getloadavg.c is missing]) + +gl_save_LIBS=$LIBS + +# Check for getloadavg, but be sure not to touch the cache variable. +(AC_CHECK_FUNC(getloadavg, exit 0, exit 1)) && gl_have_func=yes + +# On HPUX9, an unprivileged user can get load averages through this function. +AC_CHECK_FUNCS(pstat_getdynamic) + +# Solaris has libkstat which does not require root. +AC_CHECK_LIB(kstat, kstat_open) +test $ac_cv_lib_kstat_kstat_open = yes && gl_have_func=yes + +# Some systems with -lutil have (and need) -lkvm as well, some do not. +# On Solaris, -lkvm requires nlist from -lelf, so check that first +# to get the right answer into the cache. +# For kstat on solaris, we need libelf to force the definition of SVR4 below. +if test $gl_have_func = no; then + AC_CHECK_LIB(elf, elf_begin, LIBS="-lelf $LIBS") +fi +if test $gl_have_func = no; then + AC_CHECK_LIB(kvm, kvm_open, LIBS="-lkvm $LIBS") + # Check for the 4.4BSD definition of getloadavg. + AC_CHECK_LIB(util, getloadavg, + [LIBS="-lutil $LIBS" gl_have_func=yes gl_cv_func_getloadavg_setgid=yes]) +fi + +if test $gl_have_func = no; then + # There is a commonly available library for RS/6000 AIX. + # Since it is not a standard part of AIX, it might be installed locally. + gl_getloadavg_LIBS=$LIBS + LIBS="-L/usr/local/lib $LIBS" + AC_CHECK_LIB(getloadavg, getloadavg, + [LIBS="-lgetloadavg $LIBS"], [LIBS=$gl_getloadavg_LIBS]) +fi + +# Make sure it is really in the library, if we think we found it, +# otherwise set up the replacement function. +AC_CHECK_FUNCS(getloadavg, [], + [gl_PREREQ_GETLOADAVG]) + +# Some definitions of getloadavg require that the program be installed setgid. +AC_CACHE_CHECK(whether getloadavg requires setgid, + gl_cv_func_getloadavg_setgid, +[AC_EGREP_CPP([Yowza Am I SETGID yet], +[#define CONFIGURING_GETLOADAVG +#include "$srcdir/$1/getloadavg.c" +#ifdef LDAV_PRIVILEGED +Yowza Am I SETGID yet +#endif +], + gl_cv_func_getloadavg_setgid=yes, + gl_cv_func_getloadavg_setgid=no)]) +if test $gl_cv_func_getloadavg_setgid = yes; then + NEED_SETGID=true + AC_DEFINE(GETLOADAVG_PRIVILEGED, 1, + [Define to 1 if the `getloadavg' function needs to be run setuid + or setgid.]) +else + NEED_SETGID=false +fi +AC_SUBST(NEED_SETGID)dnl + +if test $gl_cv_func_getloadavg_setgid = yes; then + AC_CACHE_CHECK(group of /dev/kmem, gl_cv_group_kmem, +[ # On Solaris, /dev/kmem is a symlink. Get info on the real file. + ac_ls_output=`ls -lgL /dev/kmem 2>/dev/null` + # If we got an error (system does not support symlinks), try without -L. + test -z "$ac_ls_output" && ac_ls_output=`ls -lg /dev/kmem` + gl_cv_group_kmem=`echo $ac_ls_output \ + | sed -ne ['s/[ ][ ]*/ /g + s/^.[sSrwx-]* *[0-9]* *\([^0-9]*\) *.*/\1/ + / /s/.* //;p']` +]) + AC_SUBST(KMEM_GROUP, $gl_cv_group_kmem)dnl +fi +if test "x$gl_save_LIBS" = x; then + GETLOADAVG_LIBS=$LIBS +else + GETLOADAVG_LIBS=`echo "$LIBS" | sed "s!$gl_save_LIBS!!"` +fi +LIBS=$gl_save_LIBS + +AC_SUBST(GETLOADAVG_LIBS)dnl +])# gl_GETLOADAVG + + +# gl_PREREQ_GETLOADAVG +# -------------------- +# Set up the AC_LIBOBJ replacement of `getloadavg'. +AC_DEFUN([gl_PREREQ_GETLOADAVG], +[AC_LIBOBJ(getloadavg) +AC_DEFINE(C_GETLOADAVG, 1, [Define to 1 if using `getloadavg.c'.]) +# Figure out what our getloadavg.c needs. +gl_have_func=no +AC_CHECK_HEADER(sys/dg_sys_info.h, +[gl_have_func=yes + AC_DEFINE(DGUX, 1, [Define to 1 for DGUX with <sys/dg_sys_info.h>.]) + AC_CHECK_LIB(dgc, dg_sys_info)]) + +# We cannot check for <dwarf.h>, because Solaris 2 does not use dwarf (it +# uses stabs), but it is still SVR4. We cannot check for <elf.h> because +# Irix 4.0.5F has the header but not the library. +if test $gl_have_func = no && test "$ac_cv_lib_elf_elf_begin" = yes; then + gl_have_func=yes + AC_DEFINE(SVR4, 1, [Define to 1 on System V Release 4.]) +fi + +if test $gl_have_func = no; then + AC_CHECK_HEADER(inq_stats/cpustats.h, + [gl_have_func=yes + AC_DEFINE(UMAX, 1, [Define to 1 for Encore UMAX.]) + AC_DEFINE(UMAX4_3, 1, + [Define to 1 for Encore UMAX 4.3 that has <inq_status/cpustats.h> + instead of <sys/cpustats.h>.])]) +fi + +if test $gl_have_func = no; then + AC_CHECK_HEADER(sys/cpustats.h, + [gl_have_func=yes; AC_DEFINE(UMAX)]) +fi + +if test $gl_have_func = no; then + AC_CHECK_HEADERS(mach/mach.h) +fi + +AC_CHECK_HEADERS(nlist.h, +[AC_CHECK_MEMBERS([struct nlist.n_un.n_name], + [AC_DEFINE(NLIST_NAME_UNION, 1, + [Define to 1 if your `struct nlist' has an + `n_un' member. Obsolete, depend on + `HAVE_STRUCT_NLIST_N_UN_N_NAME])], [], + [@%:@include <nlist.h>]) +])dnl +])# gl_PREREQ_GETLOADAVG diff --git a/gl/m4/getopt.m4 b/gl/m4/getopt.m4 new file mode 100644 index 00000000..c0a73b2c --- /dev/null +++ b/gl/m4/getopt.m4 @@ -0,0 +1,83 @@ +# getopt.m4 serial 13 +dnl Copyright (C) 2002, 2003, 2004, 2005, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +# The getopt module assume you want GNU getopt, with getopt_long etc, +# rather than vanilla POSIX getopt. This means your code should +# always include <getopt.h> for the getopt prototypes. + +AC_DEFUN([gl_GETOPT_SUBSTITUTE], +[ + AC_LIBOBJ([getopt]) + AC_LIBOBJ([getopt1]) + gl_GETOPT_SUBSTITUTE_HEADER + gl_PREREQ_GETOPT +]) + +AC_DEFUN([gl_GETOPT_SUBSTITUTE_HEADER], +[ + GETOPT_H=getopt.h + AC_DEFINE([__GETOPT_PREFIX], [[rpl_]], + [Define to rpl_ if the getopt replacement functions and variables + should be used.]) + AC_SUBST([GETOPT_H]) +]) + +AC_DEFUN([gl_GETOPT_CHECK_HEADERS], +[ + if test -z "$GETOPT_H"; then + AC_CHECK_HEADERS([getopt.h], [], [GETOPT_H=getopt.h]) + fi + + if test -z "$GETOPT_H"; then + AC_CHECK_FUNCS([getopt_long_only], [], [GETOPT_H=getopt.h]) + fi + + dnl BSD getopt_long uses an incompatible method to reset option processing, + dnl and (as of 2004-10-15) mishandles optional option-arguments. + if test -z "$GETOPT_H"; then + AC_CHECK_DECL([optreset], [GETOPT_H=getopt.h], [], [#include <getopt.h>]) + fi + + dnl Solaris 10 getopt doesn't handle `+' as a leading character in an + dnl option string (as of 2005-05-05). + if test -z "$GETOPT_H"; then + AC_CACHE_CHECK([for working GNU getopt function], [gl_cv_func_gnu_getopt], + [AC_RUN_IFELSE( + [AC_LANG_PROGRAM([#include <getopt.h>], + [[ + char *myargv[3]; + myargv[0] = "conftest"; + myargv[1] = "-+"; + myargv[2] = 0; + return getopt (2, myargv, "+a") != '?'; + ]])], + [gl_cv_func_gnu_getopt=yes], + [gl_cv_func_gnu_getopt=no], + [dnl cross compiling - pessimistically guess based on decls + dnl Solaris 10 getopt doesn't handle `+' as a leading character in an + dnl option string (as of 2005-05-05). + AC_CHECK_DECL([getopt_clip], + [gl_cv_func_gnu_getopt=no], [gl_cv_func_gnu_getopt=yes], + [#include <getopt.h>])])]) + if test "$gl_cv_func_gnu_getopt" = "no"; then + GETOPT_H=getopt.h + fi + fi +]) + +AC_DEFUN([gl_GETOPT_IFELSE], +[ + AC_REQUIRE([gl_GETOPT_CHECK_HEADERS]) + AS_IF([test -n "$GETOPT_H"], [$1], [$2]) +]) + +AC_DEFUN([gl_GETOPT], [gl_GETOPT_IFELSE([gl_GETOPT_SUBSTITUTE])]) + +# Prerequisites of lib/getopt*. +AC_DEFUN([gl_PREREQ_GETOPT], +[ + AC_CHECK_DECLS_ONCE([getenv]) +]) diff --git a/gl/m4/gettext.m4 b/gl/m4/gettext.m4 new file mode 100644 index 00000000..91c345e9 --- /dev/null +++ b/gl/m4/gettext.m4 @@ -0,0 +1,419 @@ +# gettext.m4 serial 59 (gettext-0.16.1) +dnl Copyright (C) 1995-2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. +dnl +dnl This file can can be used in projects which are not available under +dnl the GNU General Public License or the GNU Library General Public +dnl License but which still want to provide support for the GNU gettext +dnl functionality. +dnl Please note that the actual code of the GNU gettext library is covered +dnl by the GNU Library General Public License, and the rest of the GNU +dnl gettext package package is covered by the GNU General Public License. +dnl They are *not* in the public domain. + +dnl Authors: +dnl Ulrich Drepper <drepper@cygnus.com>, 1995-2000. +dnl Bruno Haible <haible@clisp.cons.org>, 2000-2006. + +dnl Macro to add for using GNU gettext. + +dnl Usage: AM_GNU_GETTEXT([INTLSYMBOL], [NEEDSYMBOL], [INTLDIR]). +dnl INTLSYMBOL can be one of 'external', 'no-libtool', 'use-libtool'. The +dnl default (if it is not specified or empty) is 'no-libtool'. +dnl INTLSYMBOL should be 'external' for packages with no intl directory, +dnl and 'no-libtool' or 'use-libtool' for packages with an intl directory. +dnl If INTLSYMBOL is 'use-libtool', then a libtool library +dnl $(top_builddir)/intl/libintl.la will be created (shared and/or static, +dnl depending on --{enable,disable}-{shared,static} and on the presence of +dnl AM-DISABLE-SHARED). If INTLSYMBOL is 'no-libtool', a static library +dnl $(top_builddir)/intl/libintl.a will be created. +dnl If NEEDSYMBOL is specified and is 'need-ngettext', then GNU gettext +dnl implementations (in libc or libintl) without the ngettext() function +dnl will be ignored. If NEEDSYMBOL is specified and is +dnl 'need-formatstring-macros', then GNU gettext implementations that don't +dnl support the ISO C 99 <inttypes.h> formatstring macros will be ignored. +dnl INTLDIR is used to find the intl libraries. If empty, +dnl the value `$(top_builddir)/intl/' is used. +dnl +dnl The result of the configuration is one of three cases: +dnl 1) GNU gettext, as included in the intl subdirectory, will be compiled +dnl and used. +dnl Catalog format: GNU --> install in $(datadir) +dnl Catalog extension: .mo after installation, .gmo in source tree +dnl 2) GNU gettext has been found in the system's C library. +dnl Catalog format: GNU --> install in $(datadir) +dnl Catalog extension: .mo after installation, .gmo in source tree +dnl 3) No internationalization, always use English msgid. +dnl Catalog format: none +dnl Catalog extension: none +dnl If INTLSYMBOL is 'external', only cases 2 and 3 can occur. +dnl The use of .gmo is historical (it was needed to avoid overwriting the +dnl GNU format catalogs when building on a platform with an X/Open gettext), +dnl but we keep it in order not to force irrelevant filename changes on the +dnl maintainers. +dnl +AC_DEFUN([AM_GNU_GETTEXT], +[ + dnl Argument checking. + ifelse([$1], [], , [ifelse([$1], [external], , [ifelse([$1], [no-libtool], , [ifelse([$1], [use-libtool], , + [errprint([ERROR: invalid first argument to AM_GNU_GETTEXT +])])])])]) + ifelse([$2], [], , [ifelse([$2], [need-ngettext], , [ifelse([$2], [need-formatstring-macros], , + [errprint([ERROR: invalid second argument to AM_GNU_GETTEXT +])])])]) + define([gt_included_intl], + ifelse([$1], [external], + ifdef([AM_GNU_GETTEXT_][INTL_SUBDIR], [yes], [no]), + [yes])) + define([gt_libtool_suffix_prefix], ifelse([$1], [use-libtool], [l], [])) + gt_NEEDS_INIT + AM_GNU_GETTEXT_NEED([$2]) + + AC_REQUIRE([AM_PO_SUBDIRS])dnl + ifelse(gt_included_intl, yes, [ + AC_REQUIRE([AM_INTL_SUBDIR])dnl + ]) + + dnl Prerequisites of AC_LIB_LINKFLAGS_BODY. + AC_REQUIRE([AC_LIB_PREPARE_PREFIX]) + AC_REQUIRE([AC_LIB_RPATH]) + + dnl Sometimes libintl requires libiconv, so first search for libiconv. + dnl Ideally we would do this search only after the + dnl if test "$USE_NLS" = "yes"; then + dnl if { eval "gt_val=\$$gt_func_gnugettext_libc"; test "$gt_val" != "yes"; }; then + dnl tests. But if configure.in invokes AM_ICONV after AM_GNU_GETTEXT + dnl the configure script would need to contain the same shell code + dnl again, outside any 'if'. There are two solutions: + dnl - Invoke AM_ICONV_LINKFLAGS_BODY here, outside any 'if'. + dnl - Control the expansions in more detail using AC_PROVIDE_IFELSE. + dnl Since AC_PROVIDE_IFELSE is only in autoconf >= 2.52 and not + dnl documented, we avoid it. + ifelse(gt_included_intl, yes, , [ + AC_REQUIRE([AM_ICONV_LINKFLAGS_BODY]) + ]) + + dnl Sometimes, on MacOS X, libintl requires linking with CoreFoundation. + gt_INTL_MACOSX + + dnl Set USE_NLS. + AC_REQUIRE([AM_NLS]) + + ifelse(gt_included_intl, yes, [ + BUILD_INCLUDED_LIBINTL=no + USE_INCLUDED_LIBINTL=no + ]) + LIBINTL= + LTLIBINTL= + POSUB= + + dnl Add a version number to the cache macros. + case " $gt_needs " in + *" need-formatstring-macros "*) gt_api_version=3 ;; + *" need-ngettext "*) gt_api_version=2 ;; + *) gt_api_version=1 ;; + esac + gt_func_gnugettext_libc="gt_cv_func_gnugettext${gt_api_version}_libc" + gt_func_gnugettext_libintl="gt_cv_func_gnugettext${gt_api_version}_libintl" + + dnl If we use NLS figure out what method + if test "$USE_NLS" = "yes"; then + gt_use_preinstalled_gnugettext=no + ifelse(gt_included_intl, yes, [ + AC_MSG_CHECKING([whether included gettext is requested]) + AC_ARG_WITH(included-gettext, + [ --with-included-gettext use the GNU gettext library included here], + nls_cv_force_use_gnu_gettext=$withval, + nls_cv_force_use_gnu_gettext=no) + AC_MSG_RESULT($nls_cv_force_use_gnu_gettext) + + nls_cv_use_gnu_gettext="$nls_cv_force_use_gnu_gettext" + if test "$nls_cv_force_use_gnu_gettext" != "yes"; then + ]) + dnl User does not insist on using GNU NLS library. Figure out what + dnl to use. If GNU gettext is available we use this. Else we have + dnl to fall back to GNU NLS library. + + if test $gt_api_version -ge 3; then + gt_revision_test_code=' +#ifndef __GNU_GETTEXT_SUPPORTED_REVISION +#define __GNU_GETTEXT_SUPPORTED_REVISION(major) ((major) == 0 ? 0 : -1) +#endif +changequote(,)dnl +typedef int array [2 * (__GNU_GETTEXT_SUPPORTED_REVISION(0) >= 1) - 1]; +changequote([,])dnl +' + else + gt_revision_test_code= + fi + if test $gt_api_version -ge 2; then + gt_expression_test_code=' + * ngettext ("", "", 0)' + else + gt_expression_test_code= + fi + + AC_CACHE_CHECK([for GNU gettext in libc], [$gt_func_gnugettext_libc], + [AC_TRY_LINK([#include <libintl.h> +$gt_revision_test_code +extern int _nl_msg_cat_cntr; +extern int *_nl_domain_bindings;], + [bindtextdomain ("", ""); +return * gettext ("")$gt_expression_test_code + _nl_msg_cat_cntr + *_nl_domain_bindings], + [eval "$gt_func_gnugettext_libc=yes"], + [eval "$gt_func_gnugettext_libc=no"])]) + + if { eval "gt_val=\$$gt_func_gnugettext_libc"; test "$gt_val" != "yes"; }; then + dnl Sometimes libintl requires libiconv, so first search for libiconv. + ifelse(gt_included_intl, yes, , [ + AM_ICONV_LINK + ]) + dnl Search for libintl and define LIBINTL, LTLIBINTL and INCINTL + dnl accordingly. Don't use AC_LIB_LINKFLAGS_BODY([intl],[iconv]) + dnl because that would add "-liconv" to LIBINTL and LTLIBINTL + dnl even if libiconv doesn't exist. + AC_LIB_LINKFLAGS_BODY([intl]) + AC_CACHE_CHECK([for GNU gettext in libintl], + [$gt_func_gnugettext_libintl], + [gt_save_CPPFLAGS="$CPPFLAGS" + CPPFLAGS="$CPPFLAGS $INCINTL" + gt_save_LIBS="$LIBS" + LIBS="$LIBS $LIBINTL" + dnl Now see whether libintl exists and does not depend on libiconv. + AC_TRY_LINK([#include <libintl.h> +$gt_revision_test_code +extern int _nl_msg_cat_cntr; +extern +#ifdef __cplusplus +"C" +#endif +const char *_nl_expand_alias (const char *);], + [bindtextdomain ("", ""); +return * gettext ("")$gt_expression_test_code + _nl_msg_cat_cntr + *_nl_expand_alias ("")], + [eval "$gt_func_gnugettext_libintl=yes"], + [eval "$gt_func_gnugettext_libintl=no"]) + dnl Now see whether libintl exists and depends on libiconv. + if { eval "gt_val=\$$gt_func_gnugettext_libintl"; test "$gt_val" != yes; } && test -n "$LIBICONV"; then + LIBS="$LIBS $LIBICONV" + AC_TRY_LINK([#include <libintl.h> +$gt_revision_test_code +extern int _nl_msg_cat_cntr; +extern +#ifdef __cplusplus +"C" +#endif +const char *_nl_expand_alias (const char *);], + [bindtextdomain ("", ""); +return * gettext ("")$gt_expression_test_code + _nl_msg_cat_cntr + *_nl_expand_alias ("")], + [LIBINTL="$LIBINTL $LIBICONV" + LTLIBINTL="$LTLIBINTL $LTLIBICONV" + eval "$gt_func_gnugettext_libintl=yes" + ]) + fi + CPPFLAGS="$gt_save_CPPFLAGS" + LIBS="$gt_save_LIBS"]) + fi + + dnl If an already present or preinstalled GNU gettext() is found, + dnl use it. But if this macro is used in GNU gettext, and GNU + dnl gettext is already preinstalled in libintl, we update this + dnl libintl. (Cf. the install rule in intl/Makefile.in.) + if { eval "gt_val=\$$gt_func_gnugettext_libc"; test "$gt_val" = "yes"; } \ + || { { eval "gt_val=\$$gt_func_gnugettext_libintl"; test "$gt_val" = "yes"; } \ + && test "$PACKAGE" != gettext-runtime \ + && test "$PACKAGE" != gettext-tools; }; then + gt_use_preinstalled_gnugettext=yes + else + dnl Reset the values set by searching for libintl. + LIBINTL= + LTLIBINTL= + INCINTL= + fi + + ifelse(gt_included_intl, yes, [ + if test "$gt_use_preinstalled_gnugettext" != "yes"; then + dnl GNU gettext is not found in the C library. + dnl Fall back on included GNU gettext library. + nls_cv_use_gnu_gettext=yes + fi + fi + + if test "$nls_cv_use_gnu_gettext" = "yes"; then + dnl Mark actions used to generate GNU NLS library. + BUILD_INCLUDED_LIBINTL=yes + USE_INCLUDED_LIBINTL=yes + LIBINTL="ifelse([$3],[],\${top_builddir}/intl,[$3])/libintl.[]gt_libtool_suffix_prefix[]a $LIBICONV $LIBTHREAD" + LTLIBINTL="ifelse([$3],[],\${top_builddir}/intl,[$3])/libintl.[]gt_libtool_suffix_prefix[]a $LTLIBICONV $LTLIBTHREAD" + LIBS=`echo " $LIBS " | sed -e 's/ -lintl / /' -e 's/^ //' -e 's/ $//'` + fi + + CATOBJEXT= + if test "$gt_use_preinstalled_gnugettext" = "yes" \ + || test "$nls_cv_use_gnu_gettext" = "yes"; then + dnl Mark actions to use GNU gettext tools. + CATOBJEXT=.gmo + fi + ]) + + if test -n "$INTL_MACOSX_LIBS"; then + if test "$gt_use_preinstalled_gnugettext" = "yes" \ + || test "$nls_cv_use_gnu_gettext" = "yes"; then + dnl Some extra flags are needed during linking. + LIBINTL="$LIBINTL $INTL_MACOSX_LIBS" + LTLIBINTL="$LTLIBINTL $INTL_MACOSX_LIBS" + fi + fi + + if test "$gt_use_preinstalled_gnugettext" = "yes" \ + || test "$nls_cv_use_gnu_gettext" = "yes"; then + AC_DEFINE(ENABLE_NLS, 1, + [Define to 1 if translation of program messages to the user's native language + is requested.]) + else + USE_NLS=no + fi + fi + + AC_MSG_CHECKING([whether to use NLS]) + AC_MSG_RESULT([$USE_NLS]) + if test "$USE_NLS" = "yes"; then + AC_MSG_CHECKING([where the gettext function comes from]) + if test "$gt_use_preinstalled_gnugettext" = "yes"; then + if { eval "gt_val=\$$gt_func_gnugettext_libintl"; test "$gt_val" = "yes"; }; then + gt_source="external libintl" + else + gt_source="libc" + fi + else + gt_source="included intl directory" + fi + AC_MSG_RESULT([$gt_source]) + fi + + if test "$USE_NLS" = "yes"; then + + if test "$gt_use_preinstalled_gnugettext" = "yes"; then + if { eval "gt_val=\$$gt_func_gnugettext_libintl"; test "$gt_val" = "yes"; }; then + AC_MSG_CHECKING([how to link with libintl]) + AC_MSG_RESULT([$LIBINTL]) + AC_LIB_APPENDTOVAR([CPPFLAGS], [$INCINTL]) + fi + + dnl For backward compatibility. Some packages may be using this. + AC_DEFINE(HAVE_GETTEXT, 1, + [Define if the GNU gettext() function is already present or preinstalled.]) + AC_DEFINE(HAVE_DCGETTEXT, 1, + [Define if the GNU dcgettext() function is already present or preinstalled.]) + fi + + dnl We need to process the po/ directory. + POSUB=po + fi + + ifelse(gt_included_intl, yes, [ + dnl If this is used in GNU gettext we have to set BUILD_INCLUDED_LIBINTL + dnl to 'yes' because some of the testsuite requires it. + if test "$PACKAGE" = gettext-runtime || test "$PACKAGE" = gettext-tools; then + BUILD_INCLUDED_LIBINTL=yes + fi + + dnl Make all variables we use known to autoconf. + AC_SUBST(BUILD_INCLUDED_LIBINTL) + AC_SUBST(USE_INCLUDED_LIBINTL) + AC_SUBST(CATOBJEXT) + + dnl For backward compatibility. Some configure.ins may be using this. + nls_cv_header_intl= + nls_cv_header_libgt= + + dnl For backward compatibility. Some Makefiles may be using this. + DATADIRNAME=share + AC_SUBST(DATADIRNAME) + + dnl For backward compatibility. Some Makefiles may be using this. + INSTOBJEXT=.mo + AC_SUBST(INSTOBJEXT) + + dnl For backward compatibility. Some Makefiles may be using this. + GENCAT=gencat + AC_SUBST(GENCAT) + + dnl For backward compatibility. Some Makefiles may be using this. + INTLOBJS= + if test "$USE_INCLUDED_LIBINTL" = yes; then + INTLOBJS="\$(GETTOBJS)" + fi + AC_SUBST(INTLOBJS) + + dnl Enable libtool support if the surrounding package wishes it. + INTL_LIBTOOL_SUFFIX_PREFIX=gt_libtool_suffix_prefix + AC_SUBST(INTL_LIBTOOL_SUFFIX_PREFIX) + ]) + + dnl For backward compatibility. Some Makefiles may be using this. + INTLLIBS="$LIBINTL" + AC_SUBST(INTLLIBS) + + dnl Make all documented variables known to autoconf. + AC_SUBST(LIBINTL) + AC_SUBST(LTLIBINTL) + AC_SUBST(POSUB) +]) + + +dnl Checks for special options needed on MacOS X. +dnl Defines INTL_MACOSX_LIBS. +AC_DEFUN([gt_INTL_MACOSX], +[ + dnl Check for API introduced in MacOS X 10.2. + AC_CACHE_CHECK([for CFPreferencesCopyAppValue], + gt_cv_func_CFPreferencesCopyAppValue, + [gt_save_LIBS="$LIBS" + LIBS="$LIBS -Wl,-framework -Wl,CoreFoundation" + AC_TRY_LINK([#include <CoreFoundation/CFPreferences.h>], + [CFPreferencesCopyAppValue(NULL, NULL)], + [gt_cv_func_CFPreferencesCopyAppValue=yes], + [gt_cv_func_CFPreferencesCopyAppValue=no]) + LIBS="$gt_save_LIBS"]) + if test $gt_cv_func_CFPreferencesCopyAppValue = yes; then + AC_DEFINE([HAVE_CFPREFERENCESCOPYAPPVALUE], 1, + [Define to 1 if you have the MacOS X function CFPreferencesCopyAppValue in the CoreFoundation framework.]) + fi + dnl Check for API introduced in MacOS X 10.3. + AC_CACHE_CHECK([for CFLocaleCopyCurrent], gt_cv_func_CFLocaleCopyCurrent, + [gt_save_LIBS="$LIBS" + LIBS="$LIBS -Wl,-framework -Wl,CoreFoundation" + AC_TRY_LINK([#include <CoreFoundation/CFLocale.h>], [CFLocaleCopyCurrent();], + [gt_cv_func_CFLocaleCopyCurrent=yes], + [gt_cv_func_CFLocaleCopyCurrent=no]) + LIBS="$gt_save_LIBS"]) + if test $gt_cv_func_CFLocaleCopyCurrent = yes; then + AC_DEFINE([HAVE_CFLOCALECOPYCURRENT], 1, + [Define to 1 if you have the MacOS X function CFLocaleCopyCurrent in the CoreFoundation framework.]) + fi + INTL_MACOSX_LIBS= + if test $gt_cv_func_CFPreferencesCopyAppValue = yes || test $gt_cv_func_CFLocaleCopyCurrent = yes; then + INTL_MACOSX_LIBS="-Wl,-framework -Wl,CoreFoundation" + fi + AC_SUBST([INTL_MACOSX_LIBS]) +]) + + +dnl gt_NEEDS_INIT ensures that the gt_needs variable is initialized. +m4_define([gt_NEEDS_INIT], +[ + m4_divert_text([DEFAULTS], [gt_needs=]) + m4_define([gt_NEEDS_INIT], []) +]) + + +dnl Usage: AM_GNU_GETTEXT_NEED([NEEDSYMBOL]) +AC_DEFUN([AM_GNU_GETTEXT_NEED], +[ + m4_divert_text([INIT_PREPARE], [gt_needs="$gt_needs $1"]) +]) + + +dnl Usage: AM_GNU_GETTEXT_VERSION([gettext-version]) +AC_DEFUN([AM_GNU_GETTEXT_VERSION], []) diff --git a/gl/m4/glibc2.m4 b/gl/m4/glibc2.m4 new file mode 100644 index 00000000..e8f5bfe6 --- /dev/null +++ b/gl/m4/glibc2.m4 @@ -0,0 +1,30 @@ +# glibc2.m4 serial 1 +dnl Copyright (C) 2000-2002, 2004 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +# Test for the GNU C Library, version 2.0 or newer. +# From Bruno Haible. + +AC_DEFUN([gt_GLIBC2], + [ + AC_CACHE_CHECK(whether we are using the GNU C Library 2 or newer, + ac_cv_gnu_library_2, + [AC_EGREP_CPP([Lucky GNU user], + [ +#include <features.h> +#ifdef __GNU_LIBRARY__ + #if (__GLIBC__ >= 2) + Lucky GNU user + #endif +#endif + ], + ac_cv_gnu_library_2=yes, + ac_cv_gnu_library_2=no) + ] + ) + AC_SUBST(GLIBC2) + GLIBC2="$ac_cv_gnu_library_2" + ] +) diff --git a/gl/m4/glibc21.m4 b/gl/m4/glibc21.m4 new file mode 100644 index 00000000..d95fd986 --- /dev/null +++ b/gl/m4/glibc21.m4 @@ -0,0 +1,30 @@ +# glibc21.m4 serial 3 +dnl Copyright (C) 2000-2002, 2004 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +# Test for the GNU C Library, version 2.1 or newer. +# From Bruno Haible. + +AC_DEFUN([gl_GLIBC21], + [ + AC_CACHE_CHECK(whether we are using the GNU C Library 2.1 or newer, + ac_cv_gnu_library_2_1, + [AC_EGREP_CPP([Lucky GNU user], + [ +#include <features.h> +#ifdef __GNU_LIBRARY__ + #if (__GLIBC__ == 2 && __GLIBC_MINOR__ >= 1) || (__GLIBC__ > 2) + Lucky GNU user + #endif +#endif + ], + ac_cv_gnu_library_2_1=yes, + ac_cv_gnu_library_2_1=no) + ] + ) + AC_SUBST(GLIBC21) + GLIBC21="$ac_cv_gnu_library_2_1" + ] +) diff --git a/gl/m4/gnulib-cache.m4 b/gl/m4/gnulib-cache.m4 new file mode 100644 index 00000000..e78c8604 --- /dev/null +++ b/gl/m4/gnulib-cache.m4 @@ -0,0 +1,30 @@ +# Copyright (C) 2004-2007 Free Software Foundation, Inc. +# +# This file is free software, distributed under the terms of the GNU +# General Public License. As a special exception to the GNU General +# Public License, this file may be distributed as part of a program +# that contains a configuration script generated by Autoconf, under +# the same distribution terms as the rest of that program. +# +# Generated by gnulib-tool. +# +# This file represents the specification of how gnulib-tool is used. +# It acts as a cache: It is written and read by gnulib-tool. +# In projects using CVS, this file is meant to be stored in CVS, +# like the configure.ac and various Makefile.am files. + + +# Specification in the form of a command-line invocation: +# gnulib-tool --import --dir=. --lib=libgnu --source-base=gl --m4-base=gl/m4 --doc-base=doc --aux-dir=. --no-libtool --macro-prefix=gl dirname fsusage getaddrinfo gethostname getloadavg getopt gettext mountlist regex vasprintf vsnprintf + +# Specification in the form of a few gnulib-tool.m4 macro invocations: +gl_LOCAL_DIR([]) +gl_MODULES([dirname fsusage getaddrinfo gethostname getloadavg getopt gettext mountlist regex vasprintf vsnprintf]) +gl_AVOID([]) +gl_SOURCE_BASE([gl]) +gl_M4_BASE([gl/m4]) +gl_DOC_BASE([doc]) +gl_TESTS_BASE([tests]) +gl_LIB([libgnu]) +gl_MAKEFILE_NAME([]) +gl_MACRO_PREFIX([gl]) diff --git a/gl/m4/gnulib-common.m4 b/gl/m4/gnulib-common.m4 new file mode 100644 index 00000000..53980108 --- /dev/null +++ b/gl/m4/gnulib-common.m4 @@ -0,0 +1,22 @@ +# gnulib-common.m4 serial 2 +dnl Copyright (C) 2007 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +# gl_MODULE_INDICATOR([modulename]) +# defines a C macro indicating the presence of the given module. +AC_DEFUN([gl_MODULE_INDICATOR], +[ + AC_DEFINE([GNULIB_]translit([$1],[abcdefghijklmnopqrstuvwxyz./-],[ABCDEFGHIJKLMNOPQRSTUVWXYZ___]), [1], + [Define to 1 when using the gnulib module ]$1[.]) +]) + +# AC_PROG_MKDIR_P +# is a backport of autoconf-2.60's AC_PROG_MKDIR_P. +# Remove this macro when we can assume autoconf >= 2.60. +m4_ifdef([AC_PROG_MKDIR_P], [], [ + AC_DEFUN([AC_PROG_MKDIR_P], + [AC_REQUIRE([AM_PROG_MKDIR_P])dnl defined by automake + MKDIR_P='$(mkdir_p)' + AC_SUBST([MKDIR_P])])]) diff --git a/gl/m4/gnulib-comp.m4 b/gl/m4/gnulib-comp.m4 new file mode 100644 index 00000000..ed4da7d9 --- /dev/null +++ b/gl/m4/gnulib-comp.m4 @@ -0,0 +1,326 @@ +# DO NOT EDIT! GENERATED AUTOMATICALLY! +# Copyright (C) 2004-2007 Free Software Foundation, Inc. +# +# This file is free software, distributed under the terms of the GNU +# General Public License. As a special exception to the GNU General +# Public License, this file may be distributed as part of a program +# that contains a configuration script generated by Autoconf, under +# the same distribution terms as the rest of that program. +# +# Generated by gnulib-tool. +# +# This file represents the compiled summary of the specification in +# gnulib-cache.m4. It lists the computed macro invocations that need +# to be invoked from configure.ac. +# In projects using CVS, this file can be treated like other built files. + + +# This macro should be invoked from ./configure.in, in the section +# "Checks for programs", right after AC_PROG_CC, and certainly before +# any checks for libraries, header files, types and library functions. +AC_DEFUN([gl_EARLY], +[ + m4_pattern_forbid([^gl_[A-Z]])dnl the gnulib macro namespace + m4_pattern_allow([^gl_ES$])dnl a valid locale name + m4_pattern_allow([^gl_LIBOBJS$])dnl a variable + m4_pattern_allow([^gl_LTLIBOBJS$])dnl a variable + AC_REQUIRE([AC_PROG_RANLIB]) + AC_REQUIRE([AC_GNU_SOURCE]) + AC_REQUIRE([gl_USE_SYSTEM_EXTENSIONS]) + AC_REQUIRE([gl_LOCK_EARLY]) +]) + +# This macro should be invoked from ./configure.in, in the section +# "Check for header files, types and library functions". +AC_DEFUN([gl_INIT], +[ + m4_pushdef([AC_LIBOBJ], m4_defn([gl_LIBOBJ])) + m4_pushdef([AC_REPLACE_FUNCS], m4_defn([gl_REPLACE_FUNCS])) + m4_pushdef([AC_LIBSOURCES], m4_defn([gl_LIBSOURCES])) + AM_CONDITIONAL([GL_COND_LIBTOOL], [false]) + gl_cond_libtool=false + gl_libdeps= + gl_ltlibdeps= + gl_source_base='gl' + gl_FUNC_ALLOCA + gl_HEADER_ARPA_INET + gl_C_STRTOD + gl_CLOEXEC + gl_DIRNAME + gl_DOUBLE_SLASH_ROOT + gl_ERROR + gl_EXITFAIL + dnl gl_USE_SYSTEM_EXTENSIONS must be added quite early to configure.ac. + gl_FCNTL_SAFER + gl_MODULE_INDICATOR([fcntl-safer]) + gl_FSUSAGE + gl_GETADDRINFO + gl_FUNC_GETHOSTNAME + gl_GETLOADAVG([$gl_source_base]) + gl_GETOPT + dnl you must add AM_GNU_GETTEXT([external]) or similar to configure.ac. + AM_GNU_GETTEXT_VERSION([0.15]) + gl_INET_NTOP + gl_INLINE + AC_FUNC_MALLOC + gl_MBCHAR + gl_MBITER + gl_FUNC_MEMCHR + gl_MINMAX + gl_MOUNTLIST + gl_HEADER_NETINET_IN + gl_REGEX + gl_SAFE_READ + gl_SAFE_WRITE + gl_SIZE_MAX + gl_FUNC_SNPRINTF + gl_TYPE_SOCKLEN_T + gt_TYPE_SSIZE_T + AM_STDBOOL_H + gl_STDINT_H + gl_STRCASE + gl_FUNC_STRDUP + gl_FUNC_STRNDUP + gl_FUNC_STRNLEN + gl_HEADER_SYS_SOCKET + AC_PROG_MKDIR_P + gl_HEADER_UNISTD + gl_UNISTD_SAFER + gl_FUNC_VASNPRINTF + gl_FUNC_VASPRINTF + gl_FUNC_VSNPRINTF + gl_WCHAR_H + gl_WCTYPE_H + gl_FUNC_WCWIDTH + gl_XALLOC + gl_XSIZE + gl_XSTRNDUP + LIBGNU_LIBDEPS="$gl_libdeps" + AC_SUBST([LIBGNU_LIBDEPS]) + LIBGNU_LTLIBDEPS="$gl_ltlibdeps" + AC_SUBST([LIBGNU_LTLIBDEPS]) + m4_popdef([AC_LIBSOURCES]) + m4_popdef([AC_REPLACE_FUNCS]) + m4_popdef([AC_LIBOBJ]) + AC_CONFIG_COMMANDS_PRE([ + gl_libobjs= + gl_ltlibobjs= + if test -n "$gl_LIBOBJS"; then + # Remove the extension. + sed_drop_objext='s/\.o$//;s/\.obj$//' + for i in `for i in $gl_LIBOBJS; do echo "$i"; done | sed "$sed_drop_objext" | sort | uniq`; do + gl_libobjs="$gl_libobjs $i.$ac_objext" + gl_ltlibobjs="$gl_ltlibobjs $i.lo" + done + fi + AC_SUBST([gl_LIBOBJS], [$gl_libobjs]) + AC_SUBST([gl_LTLIBOBJS], [$gl_ltlibobjs]) + ]) +]) + +# Like AC_LIBOBJ, except that the module name goes +# into gl_LIBOBJS instead of into LIBOBJS. +AC_DEFUN([gl_LIBOBJ], + [gl_LIBOBJS="$gl_LIBOBJS $1.$ac_objext"]) + +# Like AC_REPLACE_FUNCS, except that the module name goes +# into gl_LIBOBJS instead of into LIBOBJS. +AC_DEFUN([gl_REPLACE_FUNCS], + [AC_CHECK_FUNCS([$1], , [gl_LIBOBJ($ac_func)])]) + +# Like AC_LIBSOURCES, except that it does nothing. +# We rely on EXTRA_lib..._SOURCES instead. +AC_DEFUN([gl_LIBSOURCES], + []) + +# This macro records the list of files which have been installed by +# gnulib-tool and may be removed by future gnulib-tool invocations. +AC_DEFUN([gl_FILE_LIST], [ + build-aux/config.rpath + lib/alloca.c + lib/alloca_.h + lib/asnprintf.c + lib/asprintf.c + lib/basename.c + lib/c-strtod.c + lib/c-strtod.h + lib/cloexec.c + lib/cloexec.h + lib/creat-safer.c + lib/dirname.c + lib/dirname.h + lib/dup-safer.c + lib/error.c + lib/error.h + lib/exit.h + lib/exitfail.c + lib/exitfail.h + lib/fcntl--.h + lib/fcntl-safer.h + lib/fd-safer.c + lib/fsusage.c + lib/fsusage.h + lib/full-read.c + lib/full-read.h + lib/full-write.c + lib/full-write.h + lib/gai_strerror.c + lib/getaddrinfo.c + lib/getaddrinfo.h + lib/gethostname.c + lib/getloadavg.c + lib/getopt.c + lib/getopt1.c + lib/getopt_.h + lib/getopt_int.h + lib/gettext.h + lib/inet_ntop.c + lib/inet_ntop.h + lib/intprops.h + lib/malloc.c + lib/mbchar.c + lib/mbchar.h + lib/mbuiter.h + lib/memchr.c + lib/minmax.h + lib/mountlist.c + lib/mountlist.h + lib/open-safer.c + lib/pipe-safer.c + lib/printf-args.c + lib/printf-args.h + lib/printf-parse.c + lib/printf-parse.h + lib/regcomp.c + lib/regex.c + lib/regex.h + lib/regex_internal.c + lib/regex_internal.h + lib/regexec.c + lib/safe-read.c + lib/safe-read.h + lib/safe-write.c + lib/safe-write.h + lib/size_max.h + lib/snprintf.c + lib/snprintf.h + lib/socket_.h + lib/stdbool_.h + lib/stdint_.h + lib/strcase.h + lib/strcasecmp.c + lib/strdup.c + lib/strdup.h + lib/stripslash.c + lib/strncasecmp.c + lib/strndup.c + lib/strndup.h + lib/strnlen.c + lib/strnlen.h + lib/strnlen1.c + lib/strnlen1.h + lib/unistd--.h + lib/unistd-safer.h + lib/unistd_.h + lib/vasnprintf.c + lib/vasnprintf.h + lib/vasprintf.c + lib/vasprintf.h + lib/vsnprintf.c + lib/vsnprintf.h + lib/wchar_.h + lib/wctype_.h + lib/wcwidth.h + lib/xalloc-die.c + lib/xalloc.h + lib/xmalloc.c + lib/xsize.h + lib/xstrndup.c + lib/xstrndup.h + m4/absolute-header.m4 + m4/alloca.m4 + m4/arpa_inet_h.m4 + m4/c-strtod.m4 + m4/cloexec.m4 + m4/codeset.m4 + m4/dirname.m4 + m4/dos.m4 + m4/double-slash-root.m4 + m4/eoverflow.m4 + m4/error.m4 + m4/exitfail.m4 + m4/extensions.m4 + m4/fcntl-safer.m4 + m4/fstypename.m4 + m4/fsusage.m4 + m4/getaddrinfo.m4 + m4/gethostname.m4 + m4/getloadavg.m4 + m4/getopt.m4 + m4/gettext.m4 + m4/glibc2.m4 + m4/glibc21.m4 + m4/gnulib-common.m4 + m4/iconv.m4 + m4/inet_ntop.m4 + m4/inline.m4 + m4/intdiv0.m4 + m4/intl.m4 + m4/intldir.m4 + m4/intmax.m4 + m4/intmax_t.m4 + m4/inttypes-pri.m4 + m4/inttypes_h.m4 + m4/lcmessage.m4 + m4/lib-ld.m4 + m4/lib-link.m4 + m4/lib-prefix.m4 + m4/lock.m4 + m4/longdouble.m4 + m4/longlong.m4 + m4/ls-mntd-fs.m4 + m4/mbchar.m4 + m4/mbiter.m4 + m4/mbrtowc.m4 + m4/memchr.m4 + m4/minmax.m4 + m4/mountlist.m4 + m4/netinet_in_h.m4 + m4/nls.m4 + m4/onceonly_2_57.m4 + m4/po.m4 + m4/printf-posix.m4 + m4/progtest.m4 + m4/regex.m4 + m4/safe-read.m4 + m4/safe-write.m4 + m4/size_max.m4 + m4/snprintf.m4 + m4/socklen.m4 + m4/sockpfaf.m4 + m4/ssize_t.m4 + m4/stdbool.m4 + m4/stdint.m4 + m4/stdint_h.m4 + m4/strcase.m4 + m4/strdup.m4 + m4/strndup.m4 + m4/strnlen.m4 + m4/sys_socket_h.m4 + m4/uintmax_t.m4 + m4/ulonglong.m4 + m4/unistd-safer.m4 + m4/unistd_h.m4 + m4/vasnprintf.m4 + m4/vasprintf.m4 + m4/visibility.m4 + m4/vsnprintf.m4 + m4/wchar.m4 + m4/wchar_t.m4 + m4/wctype.m4 + m4/wcwidth.m4 + m4/wint_t.m4 + m4/xalloc.m4 + m4/xsize.m4 + m4/xstrndup.m4 +]) diff --git a/gl/m4/gnulib-tool.m4 b/gl/m4/gnulib-tool.m4 new file mode 100644 index 00000000..ef593203 --- /dev/null +++ b/gl/m4/gnulib-tool.m4 @@ -0,0 +1,33 @@ +# gnulib-tool.m4 serial 1 +dnl Copyright (C) 2004-2005 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl The following macros need not be invoked explicitly. +dnl Invoking them does nothing except to declare default arguments +dnl for "gnulib-tool --import". + +dnl Usage: gl_MODULES([module1 module2 ...]) +AC_DEFUN([gl_MODULES], []) + +dnl Usage: gl_AVOID([module1 module2 ...]) +AC_DEFUN([gl_AVOID], []) + +dnl Usage: gl_SOURCE_BASE([DIR]) +AC_DEFUN([gl_SOURCE_BASE], []) + +dnl Usage: gl_M4_BASE([DIR]) +AC_DEFUN([gl_M4_BASE], []) + +dnl Usage: gl_LIB([LIBNAME]) +AC_DEFUN([gl_LIB], []) + +dnl Usage: gl_LGPL +AC_DEFUN([gl_LGPL], []) + +dnl Usage: gl_LIBTOOL +AC_DEFUN([gl_LIBTOOL], []) + +dnl Usage: gl_MACRO_PREFIX([PREFIX]) +AC_DEFUN([gl_MACRO_PREFIX], []) diff --git a/gl/m4/iconv.m4 b/gl/m4/iconv.m4 new file mode 100644 index 00000000..654c4158 --- /dev/null +++ b/gl/m4/iconv.m4 @@ -0,0 +1,101 @@ +# iconv.m4 serial AM4 (gettext-0.11.3) +dnl Copyright (C) 2000-2002 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Bruno Haible. + +AC_DEFUN([AM_ICONV_LINKFLAGS_BODY], +[ + dnl Prerequisites of AC_LIB_LINKFLAGS_BODY. + AC_REQUIRE([AC_LIB_PREPARE_PREFIX]) + AC_REQUIRE([AC_LIB_RPATH]) + + dnl Search for libiconv and define LIBICONV, LTLIBICONV and INCICONV + dnl accordingly. + AC_LIB_LINKFLAGS_BODY([iconv]) +]) + +AC_DEFUN([AM_ICONV_LINK], +[ + dnl Some systems have iconv in libc, some have it in libiconv (OSF/1 and + dnl those with the standalone portable GNU libiconv installed). + + dnl Search for libiconv and define LIBICONV, LTLIBICONV and INCICONV + dnl accordingly. + AC_REQUIRE([AM_ICONV_LINKFLAGS_BODY]) + + dnl Add $INCICONV to CPPFLAGS before performing the following checks, + dnl because if the user has installed libiconv and not disabled its use + dnl via --without-libiconv-prefix, he wants to use it. The first + dnl AC_TRY_LINK will then fail, the second AC_TRY_LINK will succeed. + am_save_CPPFLAGS="$CPPFLAGS" + AC_LIB_APPENDTOVAR([CPPFLAGS], [$INCICONV]) + + AC_CACHE_CHECK(for iconv, am_cv_func_iconv, [ + am_cv_func_iconv="no, consider installing GNU libiconv" + am_cv_lib_iconv=no + AC_TRY_LINK([#include <stdlib.h> +#include <iconv.h>], + [iconv_t cd = iconv_open("",""); + iconv(cd,NULL,NULL,NULL,NULL); + iconv_close(cd);], + am_cv_func_iconv=yes) + if test "$am_cv_func_iconv" != yes; then + am_save_LIBS="$LIBS" + LIBS="$LIBS $LIBICONV" + AC_TRY_LINK([#include <stdlib.h> +#include <iconv.h>], + [iconv_t cd = iconv_open("",""); + iconv(cd,NULL,NULL,NULL,NULL); + iconv_close(cd);], + am_cv_lib_iconv=yes + am_cv_func_iconv=yes) + LIBS="$am_save_LIBS" + fi + ]) + if test "$am_cv_func_iconv" = yes; then + AC_DEFINE(HAVE_ICONV, 1, [Define if you have the iconv() function.]) + fi + if test "$am_cv_lib_iconv" = yes; then + AC_MSG_CHECKING([how to link with libiconv]) + AC_MSG_RESULT([$LIBICONV]) + else + dnl If $LIBICONV didn't lead to a usable library, we don't need $INCICONV + dnl either. + CPPFLAGS="$am_save_CPPFLAGS" + LIBICONV= + LTLIBICONV= + fi + AC_SUBST(LIBICONV) + AC_SUBST(LTLIBICONV) +]) + +AC_DEFUN([AM_ICONV], +[ + AM_ICONV_LINK + if test "$am_cv_func_iconv" = yes; then + AC_MSG_CHECKING([for iconv declaration]) + AC_CACHE_VAL(am_cv_proto_iconv, [ + AC_TRY_COMPILE([ +#include <stdlib.h> +#include <iconv.h> +extern +#ifdef __cplusplus +"C" +#endif +#if defined(__STDC__) || defined(__cplusplus) +size_t iconv (iconv_t cd, char * *inbuf, size_t *inbytesleft, char * *outbuf, size_t *outbytesleft); +#else +size_t iconv(); +#endif +], [], am_cv_proto_iconv_arg1="", am_cv_proto_iconv_arg1="const") + am_cv_proto_iconv="extern size_t iconv (iconv_t cd, $am_cv_proto_iconv_arg1 char * *inbuf, size_t *inbytesleft, char * *outbuf, size_t *outbytesleft);"]) + am_cv_proto_iconv=`echo "[$]am_cv_proto_iconv" | tr -s ' ' | sed -e 's/( /(/'` + AC_MSG_RESULT([$]{ac_t:- + }[$]am_cv_proto_iconv) + AC_DEFINE_UNQUOTED(ICONV_CONST, $am_cv_proto_iconv_arg1, + [Define as const if the declaration of iconv() needs const.]) + fi +]) diff --git a/gl/m4/inet_ntop.m4 b/gl/m4/inet_ntop.m4 new file mode 100644 index 00000000..bb02d229 --- /dev/null +++ b/gl/m4/inet_ntop.m4 @@ -0,0 +1,19 @@ +# inet_ntop.m4 serial 3 +dnl Copyright (C) 2005, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_INET_NTOP], +[ + AC_REPLACE_FUNCS(inet_ntop) + gl_PREREQ_INET_NTOP +]) + +# Prerequisites of lib/inet_ntop.h and lib/inet_ntop.c. +AC_DEFUN([gl_PREREQ_INET_NTOP], [ + AC_CHECK_HEADERS_ONCE([netinet/in.h arpa/inet.h]) + AC_CHECK_DECLS([inet_ntop],,,[#include <arpa/inet.h>]) + AC_REQUIRE([gl_SOCKET_FAMILIES]) + AC_REQUIRE([AC_C_RESTRICT]) +]) diff --git a/gl/m4/inline.m4 b/gl/m4/inline.m4 new file mode 100644 index 00000000..a07076cd --- /dev/null +++ b/gl/m4/inline.m4 @@ -0,0 +1,40 @@ +# inline.m4 serial 3 +dnl Copyright (C) 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl Test for the 'inline' keyword or equivalent. +dnl Define 'inline' to a supported equivalent, or to nothing if not supported, +dnl like AC_C_INLINE does. Also, define HAVE_INLINE if 'inline' or an +dnl equivalent is effectively supported, i.e. if the compiler is likely to +dnl drop unused 'static inline' functions. +AC_DEFUN([gl_INLINE], +[ + AC_REQUIRE([AC_C_INLINE]) + AC_CACHE_CHECK([whether the compiler generally respects inline], + [gl_cv_c_inline_effective], + [if test $ac_cv_c_inline = no; then + gl_cv_c_inline_effective=no + else + dnl GCC defines __NO_INLINE__ if not optimizing or if -fno-inline is + dnl specified. + dnl Use AC_COMPILE_IFELSE here, not AC_EGREP_CPP, because the result + dnl depends on optimization flags, which can be in CFLAGS. + dnl (AC_EGREP_CPP looks only at the CPPFLAGS.) + AC_COMPILE_IFELSE( + [AC_LANG_PROGRAM([[]], + [[#ifdef __NO_INLINE__ + #error "inline is not effective" + #endif]])], + [gl_cv_c_inline_effective=yes], + [gl_cv_c_inline_effective=no]) + fi + ]) + if test $gl_cv_c_inline_effective = yes; then + AC_DEFINE([HAVE_INLINE], 1, + [Define to 1 if the compiler supports one of the keywords + 'inline', '__inline__', '__inline' and effectively inlines + functions marked as such.]) + fi +]) diff --git a/gl/m4/intdiv0.m4 b/gl/m4/intdiv0.m4 new file mode 100644 index 00000000..b8d78176 --- /dev/null +++ b/gl/m4/intdiv0.m4 @@ -0,0 +1,70 @@ +# intdiv0.m4 serial 1 (gettext-0.11.3) +dnl Copyright (C) 2002 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Bruno Haible. + +AC_DEFUN([gt_INTDIV0], +[ + AC_REQUIRE([AC_PROG_CC])dnl + AC_REQUIRE([AC_CANONICAL_HOST])dnl + + AC_CACHE_CHECK([whether integer division by zero raises SIGFPE], + gt_cv_int_divbyzero_sigfpe, + [ + AC_TRY_RUN([ +#include <stdlib.h> +#include <signal.h> + +static void +#ifdef __cplusplus +sigfpe_handler (int sig) +#else +sigfpe_handler (sig) int sig; +#endif +{ + /* Exit with code 0 if SIGFPE, with code 1 if any other signal. */ + exit (sig != SIGFPE); +} + +int x = 1; +int y = 0; +int z; +int nan; + +int main () +{ + signal (SIGFPE, sigfpe_handler); +/* IRIX and AIX (when "xlc -qcheck" is used) yield signal SIGTRAP. */ +#if (defined (__sgi) || defined (_AIX)) && defined (SIGTRAP) + signal (SIGTRAP, sigfpe_handler); +#endif +/* Linux/SPARC yields signal SIGILL. */ +#if defined (__sparc__) && defined (__linux__) + signal (SIGILL, sigfpe_handler); +#endif + + z = x / y; + nan = y / y; + exit (1); +} +], gt_cv_int_divbyzero_sigfpe=yes, gt_cv_int_divbyzero_sigfpe=no, + [ + # Guess based on the CPU. + case "$host_cpu" in + alpha* | i[34567]86 | m68k | s390*) + gt_cv_int_divbyzero_sigfpe="guessing yes";; + *) + gt_cv_int_divbyzero_sigfpe="guessing no";; + esac + ]) + ]) + case "$gt_cv_int_divbyzero_sigfpe" in + *yes) value=1;; + *) value=0;; + esac + AC_DEFINE_UNQUOTED(INTDIV0_RAISES_SIGFPE, $value, + [Define if integer division by zero raises signal SIGFPE.]) +]) diff --git a/gl/m4/intl.m4 b/gl/m4/intl.m4 new file mode 100644 index 00000000..dcefb118 --- /dev/null +++ b/gl/m4/intl.m4 @@ -0,0 +1,259 @@ +# intl.m4 serial 3 (gettext-0.16) +dnl Copyright (C) 1995-2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. +dnl +dnl This file can can be used in projects which are not available under +dnl the GNU General Public License or the GNU Library General Public +dnl License but which still want to provide support for the GNU gettext +dnl functionality. +dnl Please note that the actual code of the GNU gettext library is covered +dnl by the GNU Library General Public License, and the rest of the GNU +dnl gettext package package is covered by the GNU General Public License. +dnl They are *not* in the public domain. + +dnl Authors: +dnl Ulrich Drepper <drepper@cygnus.com>, 1995-2000. +dnl Bruno Haible <haible@clisp.cons.org>, 2000-2006. + +AC_PREREQ(2.52) + +dnl Checks for all prerequisites of the intl subdirectory, +dnl except for INTL_LIBTOOL_SUFFIX_PREFIX (and possibly LIBTOOL), INTLOBJS, +dnl USE_INCLUDED_LIBINTL, BUILD_INCLUDED_LIBINTL. +AC_DEFUN([AM_INTL_SUBDIR], +[ + AC_REQUIRE([AC_PROG_INSTALL])dnl + AC_REQUIRE([AM_PROG_MKDIR_P])dnl defined by automake + AC_REQUIRE([AC_PROG_CC])dnl + AC_REQUIRE([AC_CANONICAL_HOST])dnl + AC_REQUIRE([gt_GLIBC2])dnl + AC_REQUIRE([AC_PROG_RANLIB])dnl + AC_REQUIRE([gl_VISIBILITY])dnl + AC_REQUIRE([gt_INTL_SUBDIR_CORE])dnl + AC_REQUIRE([AC_TYPE_LONG_LONG_INT])dnl + AC_REQUIRE([gt_TYPE_LONGDOUBLE])dnl + AC_REQUIRE([gt_TYPE_WCHAR_T])dnl + AC_REQUIRE([gt_TYPE_WINT_T])dnl + AC_REQUIRE([gl_AC_HEADER_INTTYPES_H]) + AC_REQUIRE([gt_TYPE_INTMAX_T]) + AC_REQUIRE([gt_PRINTF_POSIX]) + AC_REQUIRE([gl_GLIBC21])dnl + AC_REQUIRE([gl_XSIZE])dnl + AC_REQUIRE([gt_INTL_MACOSX])dnl + + AC_CHECK_TYPE([ptrdiff_t], , + [AC_DEFINE([ptrdiff_t], [long], + [Define as the type of the result of subtracting two pointers, if the system doesn't define it.]) + ]) + AC_CHECK_HEADERS([stddef.h stdlib.h string.h]) + AC_CHECK_FUNCS([asprintf fwprintf putenv setenv setlocale snprintf wcslen]) + + dnl Use the _snprintf function only if it is declared (because on NetBSD it + dnl is defined as a weak alias of snprintf; we prefer to use the latter). + gt_CHECK_DECL(_snprintf, [#include <stdio.h>]) + gt_CHECK_DECL(_snwprintf, [#include <stdio.h>]) + + dnl Use the *_unlocked functions only if they are declared. + dnl (because some of them were defined without being declared in Solaris + dnl 2.5.1 but were removed in Solaris 2.6, whereas we want binaries built + dnl on Solaris 2.5.1 to run on Solaris 2.6). + dnl Don't use AC_CHECK_DECLS because it isn't supported in autoconf-2.13. + gt_CHECK_DECL(getc_unlocked, [#include <stdio.h>]) + + case $gt_cv_func_printf_posix in + *yes) HAVE_POSIX_PRINTF=1 ;; + *) HAVE_POSIX_PRINTF=0 ;; + esac + AC_SUBST([HAVE_POSIX_PRINTF]) + if test "$ac_cv_func_asprintf" = yes; then + HAVE_ASPRINTF=1 + else + HAVE_ASPRINTF=0 + fi + AC_SUBST([HAVE_ASPRINTF]) + if test "$ac_cv_func_snprintf" = yes; then + HAVE_SNPRINTF=1 + else + HAVE_SNPRINTF=0 + fi + AC_SUBST([HAVE_SNPRINTF]) + if test "$ac_cv_func_wprintf" = yes; then + HAVE_WPRINTF=1 + else + HAVE_WPRINTF=0 + fi + AC_SUBST([HAVE_WPRINTF]) + + AM_LANGINFO_CODESET + gt_LC_MESSAGES + + dnl Compilation on mingw and Cygwin needs special Makefile rules, because + dnl 1. when we install a shared library, we must arrange to export + dnl auxiliary pointer variables for every exported variable, + dnl 2. when we install a shared library and a static library simultaneously, + dnl the include file specifies __declspec(dllimport) and therefore we + dnl must arrange to define the auxiliary pointer variables for the + dnl exported variables _also_ in the static library. + if test "$enable_shared" = yes; then + case "$host_os" in + cygwin*) is_woe32dll=yes ;; + *) is_woe32dll=no ;; + esac + else + is_woe32dll=no + fi + WOE32DLL=$is_woe32dll + AC_SUBST([WOE32DLL]) + + dnl Rename some macros and functions used for locking. + AH_BOTTOM([ +#define __libc_lock_t gl_lock_t +#define __libc_lock_define gl_lock_define +#define __libc_lock_define_initialized gl_lock_define_initialized +#define __libc_lock_init gl_lock_init +#define __libc_lock_lock gl_lock_lock +#define __libc_lock_unlock gl_lock_unlock +#define __libc_lock_recursive_t gl_recursive_lock_t +#define __libc_lock_define_recursive gl_recursive_lock_define +#define __libc_lock_define_initialized_recursive gl_recursive_lock_define_initialized +#define __libc_lock_init_recursive gl_recursive_lock_init +#define __libc_lock_lock_recursive gl_recursive_lock_lock +#define __libc_lock_unlock_recursive gl_recursive_lock_unlock +#define glthread_in_use libintl_thread_in_use +#define glthread_lock_init libintl_lock_init +#define glthread_lock_lock libintl_lock_lock +#define glthread_lock_unlock libintl_lock_unlock +#define glthread_lock_destroy libintl_lock_destroy +#define glthread_rwlock_init libintl_rwlock_init +#define glthread_rwlock_rdlock libintl_rwlock_rdlock +#define glthread_rwlock_wrlock libintl_rwlock_wrlock +#define glthread_rwlock_unlock libintl_rwlock_unlock +#define glthread_rwlock_destroy libintl_rwlock_destroy +#define glthread_recursive_lock_init libintl_recursive_lock_init +#define glthread_recursive_lock_lock libintl_recursive_lock_lock +#define glthread_recursive_lock_unlock libintl_recursive_lock_unlock +#define glthread_recursive_lock_destroy libintl_recursive_lock_destroy +#define glthread_once libintl_once +#define glthread_once_call libintl_once_call +#define glthread_once_singlethreaded libintl_once_singlethreaded +]) +]) + + +dnl Checks for the core files of the intl subdirectory: +dnl dcigettext.c +dnl eval-plural.h +dnl explodename.c +dnl finddomain.c +dnl gettextP.h +dnl gmo.h +dnl hash-string.h hash-string.c +dnl l10nflist.c +dnl libgnuintl.h.in (except the *printf stuff) +dnl loadinfo.h +dnl loadmsgcat.c +dnl localealias.c +dnl log.c +dnl plural-exp.h plural-exp.c +dnl plural.y +dnl Used by libglocale. +AC_DEFUN([gt_INTL_SUBDIR_CORE], +[ + AC_REQUIRE([AC_C_INLINE])dnl + AC_REQUIRE([AC_TYPE_SIZE_T])dnl + AC_REQUIRE([gl_AC_HEADER_STDINT_H]) + AC_REQUIRE([AC_FUNC_ALLOCA])dnl + AC_REQUIRE([AC_FUNC_MMAP])dnl + AC_REQUIRE([gt_INTDIV0])dnl + AC_REQUIRE([gl_AC_TYPE_UINTMAX_T])dnl + AC_REQUIRE([gt_INTTYPES_PRI])dnl + AC_REQUIRE([gl_LOCK])dnl + + AC_TRY_LINK( + [int foo (int a) { a = __builtin_expect (a, 10); return a == 10 ? 0 : 1; }], + [], + [AC_DEFINE([HAVE_BUILTIN_EXPECT], 1, + [Define to 1 if the compiler understands __builtin_expect.])]) + + AC_CHECK_HEADERS([argz.h inttypes.h limits.h unistd.h sys/param.h]) + AC_CHECK_FUNCS([getcwd getegid geteuid getgid getuid mempcpy munmap \ + stpcpy strcasecmp strdup strtoul tsearch argz_count argz_stringify \ + argz_next __fsetlocking]) + + dnl Use the *_unlocked functions only if they are declared. + dnl (because some of them were defined without being declared in Solaris + dnl 2.5.1 but were removed in Solaris 2.6, whereas we want binaries built + dnl on Solaris 2.5.1 to run on Solaris 2.6). + dnl Don't use AC_CHECK_DECLS because it isn't supported in autoconf-2.13. + gt_CHECK_DECL(feof_unlocked, [#include <stdio.h>]) + gt_CHECK_DECL(fgets_unlocked, [#include <stdio.h>]) + + AM_ICONV + + dnl glibc >= 2.4 has a NL_LOCALE_NAME macro when _GNU_SOURCE is defined, + dnl and a _NL_LOCALE_NAME macro always. + AC_CACHE_CHECK([for NL_LOCALE_NAME macro], gt_cv_nl_locale_name, + [AC_TRY_LINK([#include <langinfo.h> +#include <locale.h>], + [char* cs = nl_langinfo(_NL_LOCALE_NAME(LC_MESSAGES));], + gt_cv_nl_locale_name=yes, + gt_cv_nl_locale_name=no) + ]) + if test $gt_cv_nl_locale_name = yes; then + AC_DEFINE(HAVE_NL_LOCALE_NAME, 1, + [Define if you have <langinfo.h> and it defines the NL_LOCALE_NAME macro if _GNU_SOURCE is defined.]) + fi + + dnl intl/plural.c is generated from intl/plural.y. It requires bison, + dnl because plural.y uses bison specific features. It requires at least + dnl bison-1.26 because earlier versions generate a plural.c that doesn't + dnl compile. + dnl bison is only needed for the maintainer (who touches plural.y). But in + dnl order to avoid separate Makefiles or --enable-maintainer-mode, we put + dnl the rule in general Makefile. Now, some people carelessly touch the + dnl files or have a broken "make" program, hence the plural.c rule will + dnl sometimes fire. To avoid an error, defines BISON to ":" if it is not + dnl present or too old. + AC_CHECK_PROGS([INTLBISON], [bison]) + if test -z "$INTLBISON"; then + ac_verc_fail=yes + else + dnl Found it, now check the version. + AC_MSG_CHECKING([version of bison]) +changequote(<<,>>)dnl + ac_prog_version=`$INTLBISON --version 2>&1 | sed -n 's/^.*GNU Bison.* \([0-9]*\.[0-9.]*\).*$/\1/p'` + case $ac_prog_version in + '') ac_prog_version="v. ?.??, bad"; ac_verc_fail=yes;; + 1.2[6-9]* | 1.[3-9][0-9]* | [2-9].*) +changequote([,])dnl + ac_prog_version="$ac_prog_version, ok"; ac_verc_fail=no;; + *) ac_prog_version="$ac_prog_version, bad"; ac_verc_fail=yes;; + esac + AC_MSG_RESULT([$ac_prog_version]) + fi + if test $ac_verc_fail = yes; then + INTLBISON=: + fi +]) + + +dnl gt_CHECK_DECL(FUNC, INCLUDES) +dnl Check whether a function is declared. +AC_DEFUN([gt_CHECK_DECL], +[ + AC_CACHE_CHECK([whether $1 is declared], ac_cv_have_decl_$1, + [AC_TRY_COMPILE([$2], [ +#ifndef $1 + char *p = (char *) $1; +#endif +], ac_cv_have_decl_$1=yes, ac_cv_have_decl_$1=no)]) + if test $ac_cv_have_decl_$1 = yes; then + gt_value=1 + else + gt_value=0 + fi + AC_DEFINE_UNQUOTED([HAVE_DECL_]translit($1, [a-z], [A-Z]), [$gt_value], + [Define to 1 if you have the declaration of `$1', and to 0 if you don't.]) +]) diff --git a/gl/m4/intldir.m4 b/gl/m4/intldir.m4 new file mode 100644 index 00000000..7a28843f --- /dev/null +++ b/gl/m4/intldir.m4 @@ -0,0 +1,19 @@ +# intldir.m4 serial 1 (gettext-0.16) +dnl Copyright (C) 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. +dnl +dnl This file can can be used in projects which are not available under +dnl the GNU General Public License or the GNU Library General Public +dnl License but which still want to provide support for the GNU gettext +dnl functionality. +dnl Please note that the actual code of the GNU gettext library is covered +dnl by the GNU Library General Public License, and the rest of the GNU +dnl gettext package package is covered by the GNU General Public License. +dnl They are *not* in the public domain. + +AC_PREREQ(2.52) + +dnl Tells the AM_GNU_GETTEXT macro to consider an intl/ directory. +AC_DEFUN([AM_GNU_GETTEXT_INTL_SUBDIR], []) diff --git a/gl/m4/intmax.m4 b/gl/m4/intmax.m4 new file mode 100644 index 00000000..ce7a8a49 --- /dev/null +++ b/gl/m4/intmax.m4 @@ -0,0 +1,33 @@ +# intmax.m4 serial 3 (gettext-0.16) +dnl Copyright (C) 2002-2005 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Bruno Haible. +dnl Test whether the system has the 'intmax_t' type, but don't attempt to +dnl find a replacement if it is lacking. + +AC_DEFUN([gt_TYPE_INTMAX_T], +[ + AC_REQUIRE([gl_AC_HEADER_INTTYPES_H]) + AC_REQUIRE([gl_AC_HEADER_STDINT_H]) + AC_CACHE_CHECK(for intmax_t, gt_cv_c_intmax_t, + [AC_TRY_COMPILE([ +#include <stddef.h> +#include <stdlib.h> +#if HAVE_STDINT_H_WITH_UINTMAX +#include <stdint.h> +#endif +#if HAVE_INTTYPES_H_WITH_UINTMAX +#include <inttypes.h> +#endif +], [intmax_t x = -1; + return !x;], + gt_cv_c_intmax_t=yes, + gt_cv_c_intmax_t=no)]) + if test $gt_cv_c_intmax_t = yes; then + AC_DEFINE(HAVE_INTMAX_T, 1, + [Define if you have the 'intmax_t' type in <stdint.h> or <inttypes.h>.]) + fi +]) diff --git a/gl/m4/intmax_t.m4 b/gl/m4/intmax_t.m4 new file mode 100644 index 00000000..17c7b0ae --- /dev/null +++ b/gl/m4/intmax_t.m4 @@ -0,0 +1,61 @@ +# intmax_t.m4 serial 5 +dnl Copyright (C) 1997-2004, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Paul Eggert. + +AC_PREREQ(2.13) + +# Define intmax_t to 'long' or 'long long' +# if it is not already defined in <stdint.h> or <inttypes.h>. + +AC_DEFUN([gl_AC_TYPE_INTMAX_T], +[ + dnl For simplicity, we assume that a header file defines 'intmax_t' if and + dnl only if it defines 'uintmax_t'. + AC_REQUIRE([gl_AC_HEADER_INTTYPES_H]) + AC_REQUIRE([gl_AC_HEADER_STDINT_H]) + if test $gl_cv_header_inttypes_h = no && test $gl_cv_header_stdint_h = no; then + AC_REQUIRE([gl_AC_TYPE_LONG_LONG]) + test $ac_cv_type_long_long = yes \ + && ac_type='long long' \ + || ac_type='long' + AC_DEFINE_UNQUOTED(intmax_t, $ac_type, + [Define to long or long long if <inttypes.h> and <stdint.h> don't define.]) + else + AC_DEFINE(HAVE_INTMAX_T, 1, + [Define if you have the 'intmax_t' type in <stdint.h> or <inttypes.h>.]) + fi +]) + +dnl An alternative would be to explicitly test for 'intmax_t'. + +AC_DEFUN([gt_AC_TYPE_INTMAX_T], +[ + AC_REQUIRE([gl_AC_HEADER_INTTYPES_H]) + AC_REQUIRE([gl_AC_HEADER_STDINT_H]) + AC_CACHE_CHECK(for intmax_t, gt_cv_c_intmax_t, + [AC_TRY_COMPILE([ +#include <stddef.h> +#include <stdlib.h> +#if HAVE_STDINT_H_WITH_UINTMAX +#include <stdint.h> +#endif +#if HAVE_INTTYPES_H_WITH_UINTMAX +#include <inttypes.h> +#endif +], [intmax_t x = -1; return !x;], gt_cv_c_intmax_t=yes, gt_cv_c_intmax_t=no)]) + if test $gt_cv_c_intmax_t = yes; then + AC_DEFINE(HAVE_INTMAX_T, 1, + [Define if you have the 'intmax_t' type in <stdint.h> or <inttypes.h>.]) + else + AC_REQUIRE([gl_AC_TYPE_LONG_LONG]) + test $ac_cv_type_long_long = yes \ + && ac_type='long long' \ + || ac_type='long' + AC_DEFINE_UNQUOTED(intmax_t, $ac_type, + [Define to long or long long if <stdint.h> and <inttypes.h> don't define.]) + fi +]) diff --git a/gl/m4/inttypes-pri.m4 b/gl/m4/inttypes-pri.m4 new file mode 100644 index 00000000..7c7f8940 --- /dev/null +++ b/gl/m4/inttypes-pri.m4 @@ -0,0 +1,36 @@ +# inttypes-pri.m4 serial 4 (gettext-0.16) +dnl Copyright (C) 1997-2002, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Bruno Haible. + +AC_PREREQ(2.52) + +# Define PRI_MACROS_BROKEN if <inttypes.h> exists and defines the PRI* +# macros to non-string values. This is the case on AIX 4.3.3. + +AC_DEFUN([gt_INTTYPES_PRI], +[ + AC_CHECK_HEADERS([inttypes.h]) + if test $ac_cv_header_inttypes_h = yes; then + AC_CACHE_CHECK([whether the inttypes.h PRIxNN macros are broken], + gt_cv_inttypes_pri_broken, + [ + AC_TRY_COMPILE([#include <inttypes.h> +#ifdef PRId32 +char *p = PRId32; +#endif +], [], gt_cv_inttypes_pri_broken=no, gt_cv_inttypes_pri_broken=yes) + ]) + fi + if test "$gt_cv_inttypes_pri_broken" = yes; then + AC_DEFINE_UNQUOTED(PRI_MACROS_BROKEN, 1, + [Define if <inttypes.h> exists and defines unusable PRI* macros.]) + PRI_MACROS_BROKEN=1 + else + PRI_MACROS_BROKEN=0 + fi + AC_SUBST([PRI_MACROS_BROKEN]) +]) diff --git a/gl/m4/inttypes_h.m4 b/gl/m4/inttypes_h.m4 new file mode 100644 index 00000000..edc8ecb2 --- /dev/null +++ b/gl/m4/inttypes_h.m4 @@ -0,0 +1,26 @@ +# inttypes_h.m4 serial 7 +dnl Copyright (C) 1997-2004, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Paul Eggert. + +# Define HAVE_INTTYPES_H_WITH_UINTMAX if <inttypes.h> exists, +# doesn't clash with <sys/types.h>, and declares uintmax_t. + +AC_DEFUN([gl_AC_HEADER_INTTYPES_H], +[ + AC_CACHE_CHECK([for inttypes.h], gl_cv_header_inttypes_h, + [AC_TRY_COMPILE( + [#include <sys/types.h> +#include <inttypes.h>], + [uintmax_t i = (uintmax_t) -1; return !i;], + gl_cv_header_inttypes_h=yes, + gl_cv_header_inttypes_h=no)]) + if test $gl_cv_header_inttypes_h = yes; then + AC_DEFINE_UNQUOTED(HAVE_INTTYPES_H_WITH_UINTMAX, 1, + [Define if <inttypes.h> exists, doesn't clash with <sys/types.h>, + and declares uintmax_t. ]) + fi +]) diff --git a/gl/m4/lcmessage.m4 b/gl/m4/lcmessage.m4 new file mode 100644 index 00000000..19aa77e4 --- /dev/null +++ b/gl/m4/lcmessage.m4 @@ -0,0 +1,30 @@ +# lcmessage.m4 serial 4 (gettext-0.14.2) +dnl Copyright (C) 1995-2002, 2004-2005 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. +dnl +dnl This file can can be used in projects which are not available under +dnl the GNU General Public License or the GNU Library General Public +dnl License but which still want to provide support for the GNU gettext +dnl functionality. +dnl Please note that the actual code of the GNU gettext library is covered +dnl by the GNU Library General Public License, and the rest of the GNU +dnl gettext package package is covered by the GNU General Public License. +dnl They are *not* in the public domain. + +dnl Authors: +dnl Ulrich Drepper <drepper@cygnus.com>, 1995. + +# Check whether LC_MESSAGES is available in <locale.h>. + +AC_DEFUN([gt_LC_MESSAGES], +[ + AC_CACHE_CHECK([for LC_MESSAGES], gt_cv_val_LC_MESSAGES, + [AC_TRY_LINK([#include <locale.h>], [return LC_MESSAGES], + gt_cv_val_LC_MESSAGES=yes, gt_cv_val_LC_MESSAGES=no)]) + if test $gt_cv_val_LC_MESSAGES = yes; then + AC_DEFINE(HAVE_LC_MESSAGES, 1, + [Define if your <locale.h> file defines LC_MESSAGES.]) + fi +]) diff --git a/gl/m4/lib-ld.m4 b/gl/m4/lib-ld.m4 new file mode 100644 index 00000000..96c4e2c3 --- /dev/null +++ b/gl/m4/lib-ld.m4 @@ -0,0 +1,110 @@ +# lib-ld.m4 serial 3 (gettext-0.13) +dnl Copyright (C) 1996-2003 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl Subroutines of libtool.m4, +dnl with replacements s/AC_/AC_LIB/ and s/lt_cv/acl_cv/ to avoid collision +dnl with libtool.m4. + +dnl From libtool-1.4. Sets the variable with_gnu_ld to yes or no. +AC_DEFUN([AC_LIB_PROG_LD_GNU], +[AC_CACHE_CHECK([if the linker ($LD) is GNU ld], acl_cv_prog_gnu_ld, +[# I'd rather use --version here, but apparently some GNU ld's only accept -v. +case `$LD -v 2>&1 </dev/null` in +*GNU* | *'with BFD'*) + acl_cv_prog_gnu_ld=yes ;; +*) + acl_cv_prog_gnu_ld=no ;; +esac]) +with_gnu_ld=$acl_cv_prog_gnu_ld +]) + +dnl From libtool-1.4. Sets the variable LD. +AC_DEFUN([AC_LIB_PROG_LD], +[AC_ARG_WITH(gnu-ld, +[ --with-gnu-ld assume the C compiler uses GNU ld [default=no]], +test "$withval" = no || with_gnu_ld=yes, with_gnu_ld=no) +AC_REQUIRE([AC_PROG_CC])dnl +AC_REQUIRE([AC_CANONICAL_HOST])dnl +# Prepare PATH_SEPARATOR. +# The user is always right. +if test "${PATH_SEPARATOR+set}" != set; then + echo "#! /bin/sh" >conf$$.sh + echo "exit 0" >>conf$$.sh + chmod +x conf$$.sh + if (PATH="/nonexistent;."; conf$$.sh) >/dev/null 2>&1; then + PATH_SEPARATOR=';' + else + PATH_SEPARATOR=: + fi + rm -f conf$$.sh +fi +ac_prog=ld +if test "$GCC" = yes; then + # Check if gcc -print-prog-name=ld gives a path. + AC_MSG_CHECKING([for ld used by GCC]) + case $host in + *-*-mingw*) + # gcc leaves a trailing carriage return which upsets mingw + ac_prog=`($CC -print-prog-name=ld) 2>&5 | tr -d '\015'` ;; + *) + ac_prog=`($CC -print-prog-name=ld) 2>&5` ;; + esac + case $ac_prog in + # Accept absolute paths. + [[\\/]* | [A-Za-z]:[\\/]*)] + [re_direlt='/[^/][^/]*/\.\./'] + # Canonicalize the path of ld + ac_prog=`echo $ac_prog| sed 's%\\\\%/%g'` + while echo $ac_prog | grep "$re_direlt" > /dev/null 2>&1; do + ac_prog=`echo $ac_prog| sed "s%$re_direlt%/%"` + done + test -z "$LD" && LD="$ac_prog" + ;; + "") + # If it fails, then pretend we aren't using GCC. + ac_prog=ld + ;; + *) + # If it is relative, then search for the first ld in PATH. + with_gnu_ld=unknown + ;; + esac +elif test "$with_gnu_ld" = yes; then + AC_MSG_CHECKING([for GNU ld]) +else + AC_MSG_CHECKING([for non-GNU ld]) +fi +AC_CACHE_VAL(acl_cv_path_LD, +[if test -z "$LD"; then + IFS="${IFS= }"; ac_save_ifs="$IFS"; IFS="${IFS}${PATH_SEPARATOR-:}" + for ac_dir in $PATH; do + test -z "$ac_dir" && ac_dir=. + if test -f "$ac_dir/$ac_prog" || test -f "$ac_dir/$ac_prog$ac_exeext"; then + acl_cv_path_LD="$ac_dir/$ac_prog" + # Check to see if the program is GNU ld. I'd rather use --version, + # but apparently some GNU ld's only accept -v. + # Break only if it was the GNU/non-GNU ld that we prefer. + case `"$acl_cv_path_LD" -v 2>&1 < /dev/null` in + *GNU* | *'with BFD'*) + test "$with_gnu_ld" != no && break ;; + *) + test "$with_gnu_ld" != yes && break ;; + esac + fi + done + IFS="$ac_save_ifs" +else + acl_cv_path_LD="$LD" # Let the user override the test with a path. +fi]) +LD="$acl_cv_path_LD" +if test -n "$LD"; then + AC_MSG_RESULT($LD) +else + AC_MSG_RESULT(no) +fi +test -z "$LD" && AC_MSG_ERROR([no acceptable ld found in \$PATH]) +AC_LIB_PROG_LD_GNU +]) diff --git a/gl/m4/lib-link.m4 b/gl/m4/lib-link.m4 new file mode 100644 index 00000000..f157d983 --- /dev/null +++ b/gl/m4/lib-link.m4 @@ -0,0 +1,709 @@ +# lib-link.m4 serial 13 (gettext-0.16.2) +dnl Copyright (C) 2001-2007 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Bruno Haible. + +AC_PREREQ(2.54) + +dnl AC_LIB_LINKFLAGS(name [, dependencies]) searches for libname and +dnl the libraries corresponding to explicit and implicit dependencies. +dnl Sets and AC_SUBSTs the LIB${NAME} and LTLIB${NAME} variables and +dnl augments the CPPFLAGS variable. +dnl Sets and AC_SUBSTs the LIB${NAME}_PREFIX variable to nonempty if libname +dnl was found in ${LIB${NAME}_PREFIX}/$acl_libdirstem. +AC_DEFUN([AC_LIB_LINKFLAGS], +[ + AC_REQUIRE([AC_LIB_PREPARE_PREFIX]) + AC_REQUIRE([AC_LIB_RPATH]) + define([Name],[translit([$1],[./-], [___])]) + define([NAME],[translit([$1],[abcdefghijklmnopqrstuvwxyz./-], + [ABCDEFGHIJKLMNOPQRSTUVWXYZ___])]) + AC_CACHE_CHECK([how to link with lib[]$1], [ac_cv_lib[]Name[]_libs], [ + AC_LIB_LINKFLAGS_BODY([$1], [$2]) + ac_cv_lib[]Name[]_libs="$LIB[]NAME" + ac_cv_lib[]Name[]_ltlibs="$LTLIB[]NAME" + ac_cv_lib[]Name[]_cppflags="$INC[]NAME" + ac_cv_lib[]Name[]_prefix="$LIB[]NAME[]_PREFIX" + ]) + LIB[]NAME="$ac_cv_lib[]Name[]_libs" + LTLIB[]NAME="$ac_cv_lib[]Name[]_ltlibs" + INC[]NAME="$ac_cv_lib[]Name[]_cppflags" + LIB[]NAME[]_PREFIX="$ac_cv_lib[]Name[]_prefix" + AC_LIB_APPENDTOVAR([CPPFLAGS], [$INC]NAME) + AC_SUBST([LIB]NAME) + AC_SUBST([LTLIB]NAME) + AC_SUBST([LIB]NAME[_PREFIX]) + dnl Also set HAVE_LIB[]NAME so that AC_LIB_HAVE_LINKFLAGS can reuse the + dnl results of this search when this library appears as a dependency. + HAVE_LIB[]NAME=yes + undefine([Name]) + undefine([NAME]) +]) + +dnl AC_LIB_HAVE_LINKFLAGS(name, dependencies, includes, testcode) +dnl searches for libname and the libraries corresponding to explicit and +dnl implicit dependencies, together with the specified include files and +dnl the ability to compile and link the specified testcode. If found, it +dnl sets and AC_SUBSTs HAVE_LIB${NAME}=yes and the LIB${NAME} and +dnl LTLIB${NAME} variables and augments the CPPFLAGS variable, and +dnl #defines HAVE_LIB${NAME} to 1. Otherwise, it sets and AC_SUBSTs +dnl HAVE_LIB${NAME}=no and LIB${NAME} and LTLIB${NAME} to empty. +dnl Sets and AC_SUBSTs the LIB${NAME}_PREFIX variable to nonempty if libname +dnl was found in ${LIB${NAME}_PREFIX}/$acl_libdirstem. +AC_DEFUN([AC_LIB_HAVE_LINKFLAGS], +[ + AC_REQUIRE([AC_LIB_PREPARE_PREFIX]) + AC_REQUIRE([AC_LIB_RPATH]) + define([Name],[translit([$1],[./-], [___])]) + define([NAME],[translit([$1],[abcdefghijklmnopqrstuvwxyz./-], + [ABCDEFGHIJKLMNOPQRSTUVWXYZ___])]) + + dnl Search for lib[]Name and define LIB[]NAME, LTLIB[]NAME and INC[]NAME + dnl accordingly. + AC_LIB_LINKFLAGS_BODY([$1], [$2]) + + dnl Add $INC[]NAME to CPPFLAGS before performing the following checks, + dnl because if the user has installed lib[]Name and not disabled its use + dnl via --without-lib[]Name-prefix, he wants to use it. + ac_save_CPPFLAGS="$CPPFLAGS" + AC_LIB_APPENDTOVAR([CPPFLAGS], [$INC]NAME) + + AC_CACHE_CHECK([for lib[]$1], [ac_cv_lib[]Name], [ + ac_save_LIBS="$LIBS" + LIBS="$LIBS $LIB[]NAME" + AC_TRY_LINK([$3], [$4], [ac_cv_lib[]Name=yes], [ac_cv_lib[]Name=no]) + LIBS="$ac_save_LIBS" + ]) + if test "$ac_cv_lib[]Name" = yes; then + HAVE_LIB[]NAME=yes + AC_DEFINE([HAVE_LIB]NAME, 1, [Define if you have the $1 library.]) + AC_MSG_CHECKING([how to link with lib[]$1]) + AC_MSG_RESULT([$LIB[]NAME]) + else + HAVE_LIB[]NAME=no + dnl If $LIB[]NAME didn't lead to a usable library, we don't need + dnl $INC[]NAME either. + CPPFLAGS="$ac_save_CPPFLAGS" + LIB[]NAME= + LTLIB[]NAME= + LIB[]NAME[]_PREFIX= + fi + AC_SUBST([HAVE_LIB]NAME) + AC_SUBST([LIB]NAME) + AC_SUBST([LTLIB]NAME) + AC_SUBST([LIB]NAME[_PREFIX]) + undefine([Name]) + undefine([NAME]) +]) + +dnl Determine the platform dependent parameters needed to use rpath: +dnl acl_libext, +dnl acl_shlibext, +dnl acl_hardcode_libdir_flag_spec, +dnl acl_hardcode_libdir_separator, +dnl acl_hardcode_direct, +dnl acl_hardcode_minus_L. +AC_DEFUN([AC_LIB_RPATH], +[ + dnl Tell automake >= 1.10 to complain if config.rpath is missing. + m4_ifdef([AC_REQUIRE_AUX_FILE], [AC_REQUIRE_AUX_FILE([config.rpath])]) + AC_REQUIRE([AC_PROG_CC]) dnl we use $CC, $GCC, $LDFLAGS + AC_REQUIRE([AC_LIB_PROG_LD]) dnl we use $LD, $with_gnu_ld + AC_REQUIRE([AC_CANONICAL_HOST]) dnl we use $host + AC_REQUIRE([AC_CONFIG_AUX_DIR_DEFAULT]) dnl we use $ac_aux_dir + AC_CACHE_CHECK([for shared library run path origin], acl_cv_rpath, [ + CC="$CC" GCC="$GCC" LDFLAGS="$LDFLAGS" LD="$LD" with_gnu_ld="$with_gnu_ld" \ + ${CONFIG_SHELL-/bin/sh} "$ac_aux_dir/config.rpath" "$host" > conftest.sh + . ./conftest.sh + rm -f ./conftest.sh + acl_cv_rpath=done + ]) + wl="$acl_cv_wl" + acl_libext="$acl_cv_libext" + acl_shlibext="$acl_cv_shlibext" + acl_libname_spec="$acl_cv_libname_spec" + acl_library_names_spec="$acl_cv_library_names_spec" + acl_hardcode_libdir_flag_spec="$acl_cv_hardcode_libdir_flag_spec" + acl_hardcode_libdir_separator="$acl_cv_hardcode_libdir_separator" + acl_hardcode_direct="$acl_cv_hardcode_direct" + acl_hardcode_minus_L="$acl_cv_hardcode_minus_L" + dnl Determine whether the user wants rpath handling at all. + AC_ARG_ENABLE(rpath, + [ --disable-rpath do not hardcode runtime library paths], + :, enable_rpath=yes) +]) + +dnl AC_LIB_LINKFLAGS_BODY(name [, dependencies]) searches for libname and +dnl the libraries corresponding to explicit and implicit dependencies. +dnl Sets the LIB${NAME}, LTLIB${NAME} and INC${NAME} variables. +dnl Also, sets the LIB${NAME}_PREFIX variable to nonempty if libname was found +dnl in ${LIB${NAME}_PREFIX}/$acl_libdirstem. +AC_DEFUN([AC_LIB_LINKFLAGS_BODY], +[ + AC_REQUIRE([AC_LIB_PREPARE_MULTILIB]) + define([NAME],[translit([$1],[abcdefghijklmnopqrstuvwxyz./-], + [ABCDEFGHIJKLMNOPQRSTUVWXYZ___])]) + dnl Autoconf >= 2.61 supports dots in --with options. + define([N_A_M_E],[m4_if(m4_version_compare(m4_defn([m4_PACKAGE_VERSION]),[2.61]),[-1],[translit([$1],[.],[_])],[$1])]) + dnl By default, look in $includedir and $libdir. + use_additional=yes + AC_LIB_WITH_FINAL_PREFIX([ + eval additional_includedir=\"$includedir\" + eval additional_libdir=\"$libdir\" + ]) + AC_LIB_ARG_WITH([lib]N_A_M_E[-prefix], +[ --with-lib]N_A_M_E[-prefix[=DIR] search for lib$1 in DIR/include and DIR/lib + --without-lib]N_A_M_E[-prefix don't search for lib$1 in includedir and libdir], +[ + if test "X$withval" = "Xno"; then + use_additional=no + else + if test "X$withval" = "X"; then + AC_LIB_WITH_FINAL_PREFIX([ + eval additional_includedir=\"$includedir\" + eval additional_libdir=\"$libdir\" + ]) + else + additional_includedir="$withval/include" + additional_libdir="$withval/$acl_libdirstem" + fi + fi +]) + dnl Search the library and its dependencies in $additional_libdir and + dnl $LDFLAGS. Using breadth-first-seach. + LIB[]NAME= + LTLIB[]NAME= + INC[]NAME= + LIB[]NAME[]_PREFIX= + rpathdirs= + ltrpathdirs= + names_already_handled= + names_next_round='$1 $2' + while test -n "$names_next_round"; do + names_this_round="$names_next_round" + names_next_round= + for name in $names_this_round; do + already_handled= + for n in $names_already_handled; do + if test "$n" = "$name"; then + already_handled=yes + break + fi + done + if test -z "$already_handled"; then + names_already_handled="$names_already_handled $name" + dnl See if it was already located by an earlier AC_LIB_LINKFLAGS + dnl or AC_LIB_HAVE_LINKFLAGS call. + uppername=`echo "$name" | sed -e 'y|abcdefghijklmnopqrstuvwxyz./-|ABCDEFGHIJKLMNOPQRSTUVWXYZ___|'` + eval value=\"\$HAVE_LIB$uppername\" + if test -n "$value"; then + if test "$value" = yes; then + eval value=\"\$LIB$uppername\" + test -z "$value" || LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$value" + eval value=\"\$LTLIB$uppername\" + test -z "$value" || LTLIB[]NAME="${LTLIB[]NAME}${LTLIB[]NAME:+ }$value" + else + dnl An earlier call to AC_LIB_HAVE_LINKFLAGS has determined + dnl that this library doesn't exist. So just drop it. + : + fi + else + dnl Search the library lib$name in $additional_libdir and $LDFLAGS + dnl and the already constructed $LIBNAME/$LTLIBNAME. + found_dir= + found_la= + found_so= + found_a= + eval libname=\"$acl_libname_spec\" # typically: libname=lib$name + if test -n "$acl_shlibext"; then + shrext=".$acl_shlibext" # typically: shrext=.so + else + shrext= + fi + if test $use_additional = yes; then + dir="$additional_libdir" + dnl The same code as in the loop below: + dnl First look for a shared library. + if test -n "$acl_shlibext"; then + if test -f "$dir/$libname$shrext"; then + found_dir="$dir" + found_so="$dir/$libname$shrext" + else + if test "$acl_library_names_spec" = '$libname$shrext$versuffix'; then + ver=`(cd "$dir" && \ + for f in "$libname$shrext".*; do echo "$f"; done \ + | sed -e "s,^$libname$shrext\\\\.,," \ + | sort -t '.' -n -r -k1,1 -k2,2 -k3,3 -k4,4 -k5,5 \ + | sed 1q ) 2>/dev/null` + if test -n "$ver" && test -f "$dir/$libname$shrext.$ver"; then + found_dir="$dir" + found_so="$dir/$libname$shrext.$ver" + fi + else + eval library_names=\"$acl_library_names_spec\" + for f in $library_names; do + if test -f "$dir/$f"; then + found_dir="$dir" + found_so="$dir/$f" + break + fi + done + fi + fi + fi + dnl Then look for a static library. + if test "X$found_dir" = "X"; then + if test -f "$dir/$libname.$acl_libext"; then + found_dir="$dir" + found_a="$dir/$libname.$acl_libext" + fi + fi + if test "X$found_dir" != "X"; then + if test -f "$dir/$libname.la"; then + found_la="$dir/$libname.la" + fi + fi + fi + if test "X$found_dir" = "X"; then + for x in $LDFLAGS $LTLIB[]NAME; do + AC_LIB_WITH_FINAL_PREFIX([eval x=\"$x\"]) + case "$x" in + -L*) + dir=`echo "X$x" | sed -e 's/^X-L//'` + dnl First look for a shared library. + if test -n "$acl_shlibext"; then + if test -f "$dir/$libname$shrext"; then + found_dir="$dir" + found_so="$dir/$libname$shrext" + else + if test "$acl_library_names_spec" = '$libname$shrext$versuffix'; then + ver=`(cd "$dir" && \ + for f in "$libname$shrext".*; do echo "$f"; done \ + | sed -e "s,^$libname$shrext\\\\.,," \ + | sort -t '.' -n -r -k1,1 -k2,2 -k3,3 -k4,4 -k5,5 \ + | sed 1q ) 2>/dev/null` + if test -n "$ver" && test -f "$dir/$libname$shrext.$ver"; then + found_dir="$dir" + found_so="$dir/$libname$shrext.$ver" + fi + else + eval library_names=\"$acl_library_names_spec\" + for f in $library_names; do + if test -f "$dir/$f"; then + found_dir="$dir" + found_so="$dir/$f" + break + fi + done + fi + fi + fi + dnl Then look for a static library. + if test "X$found_dir" = "X"; then + if test -f "$dir/$libname.$acl_libext"; then + found_dir="$dir" + found_a="$dir/$libname.$acl_libext" + fi + fi + if test "X$found_dir" != "X"; then + if test -f "$dir/$libname.la"; then + found_la="$dir/$libname.la" + fi + fi + ;; + esac + if test "X$found_dir" != "X"; then + break + fi + done + fi + if test "X$found_dir" != "X"; then + dnl Found the library. + LTLIB[]NAME="${LTLIB[]NAME}${LTLIB[]NAME:+ }-L$found_dir -l$name" + if test "X$found_so" != "X"; then + dnl Linking with a shared library. We attempt to hardcode its + dnl directory into the executable's runpath, unless it's the + dnl standard /usr/lib. + if test "$enable_rpath" = no || test "X$found_dir" = "X/usr/$acl_libdirstem"; then + dnl No hardcoding is needed. + LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$found_so" + else + dnl Use an explicit option to hardcode DIR into the resulting + dnl binary. + dnl Potentially add DIR to ltrpathdirs. + dnl The ltrpathdirs will be appended to $LTLIBNAME at the end. + haveit= + for x in $ltrpathdirs; do + if test "X$x" = "X$found_dir"; then + haveit=yes + break + fi + done + if test -z "$haveit"; then + ltrpathdirs="$ltrpathdirs $found_dir" + fi + dnl The hardcoding into $LIBNAME is system dependent. + if test "$acl_hardcode_direct" = yes; then + dnl Using DIR/libNAME.so during linking hardcodes DIR into the + dnl resulting binary. + LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$found_so" + else + if test -n "$acl_hardcode_libdir_flag_spec" && test "$acl_hardcode_minus_L" = no; then + dnl Use an explicit option to hardcode DIR into the resulting + dnl binary. + LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$found_so" + dnl Potentially add DIR to rpathdirs. + dnl The rpathdirs will be appended to $LIBNAME at the end. + haveit= + for x in $rpathdirs; do + if test "X$x" = "X$found_dir"; then + haveit=yes + break + fi + done + if test -z "$haveit"; then + rpathdirs="$rpathdirs $found_dir" + fi + else + dnl Rely on "-L$found_dir". + dnl But don't add it if it's already contained in the LDFLAGS + dnl or the already constructed $LIBNAME + haveit= + for x in $LDFLAGS $LIB[]NAME; do + AC_LIB_WITH_FINAL_PREFIX([eval x=\"$x\"]) + if test "X$x" = "X-L$found_dir"; then + haveit=yes + break + fi + done + if test -z "$haveit"; then + LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }-L$found_dir" + fi + if test "$acl_hardcode_minus_L" != no; then + dnl FIXME: Not sure whether we should use + dnl "-L$found_dir -l$name" or "-L$found_dir $found_so" + dnl here. + LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$found_so" + else + dnl We cannot use $acl_hardcode_runpath_var and LD_RUN_PATH + dnl here, because this doesn't fit in flags passed to the + dnl compiler. So give up. No hardcoding. This affects only + dnl very old systems. + dnl FIXME: Not sure whether we should use + dnl "-L$found_dir -l$name" or "-L$found_dir $found_so" + dnl here. + LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }-l$name" + fi + fi + fi + fi + else + if test "X$found_a" != "X"; then + dnl Linking with a static library. + LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$found_a" + else + dnl We shouldn't come here, but anyway it's good to have a + dnl fallback. + LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }-L$found_dir -l$name" + fi + fi + dnl Assume the include files are nearby. + additional_includedir= + case "$found_dir" in + */$acl_libdirstem | */$acl_libdirstem/) + basedir=`echo "X$found_dir" | sed -e 's,^X,,' -e "s,/$acl_libdirstem/"'*$,,'` + LIB[]NAME[]_PREFIX="$basedir" + additional_includedir="$basedir/include" + ;; + esac + if test "X$additional_includedir" != "X"; then + dnl Potentially add $additional_includedir to $INCNAME. + dnl But don't add it + dnl 1. if it's the standard /usr/include, + dnl 2. if it's /usr/local/include and we are using GCC on Linux, + dnl 3. if it's already present in $CPPFLAGS or the already + dnl constructed $INCNAME, + dnl 4. if it doesn't exist as a directory. + if test "X$additional_includedir" != "X/usr/include"; then + haveit= + if test "X$additional_includedir" = "X/usr/local/include"; then + if test -n "$GCC"; then + case $host_os in + linux* | gnu* | k*bsd*-gnu) haveit=yes;; + esac + fi + fi + if test -z "$haveit"; then + for x in $CPPFLAGS $INC[]NAME; do + AC_LIB_WITH_FINAL_PREFIX([eval x=\"$x\"]) + if test "X$x" = "X-I$additional_includedir"; then + haveit=yes + break + fi + done + if test -z "$haveit"; then + if test -d "$additional_includedir"; then + dnl Really add $additional_includedir to $INCNAME. + INC[]NAME="${INC[]NAME}${INC[]NAME:+ }-I$additional_includedir" + fi + fi + fi + fi + fi + dnl Look for dependencies. + if test -n "$found_la"; then + dnl Read the .la file. It defines the variables + dnl dlname, library_names, old_library, dependency_libs, current, + dnl age, revision, installed, dlopen, dlpreopen, libdir. + save_libdir="$libdir" + case "$found_la" in + */* | *\\*) . "$found_la" ;; + *) . "./$found_la" ;; + esac + libdir="$save_libdir" + dnl We use only dependency_libs. + for dep in $dependency_libs; do + case "$dep" in + -L*) + additional_libdir=`echo "X$dep" | sed -e 's/^X-L//'` + dnl Potentially add $additional_libdir to $LIBNAME and $LTLIBNAME. + dnl But don't add it + dnl 1. if it's the standard /usr/lib, + dnl 2. if it's /usr/local/lib and we are using GCC on Linux, + dnl 3. if it's already present in $LDFLAGS or the already + dnl constructed $LIBNAME, + dnl 4. if it doesn't exist as a directory. + if test "X$additional_libdir" != "X/usr/$acl_libdirstem"; then + haveit= + if test "X$additional_libdir" = "X/usr/local/$acl_libdirstem"; then + if test -n "$GCC"; then + case $host_os in + linux* | gnu* | k*bsd*-gnu) haveit=yes;; + esac + fi + fi + if test -z "$haveit"; then + haveit= + for x in $LDFLAGS $LIB[]NAME; do + AC_LIB_WITH_FINAL_PREFIX([eval x=\"$x\"]) + if test "X$x" = "X-L$additional_libdir"; then + haveit=yes + break + fi + done + if test -z "$haveit"; then + if test -d "$additional_libdir"; then + dnl Really add $additional_libdir to $LIBNAME. + LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }-L$additional_libdir" + fi + fi + haveit= + for x in $LDFLAGS $LTLIB[]NAME; do + AC_LIB_WITH_FINAL_PREFIX([eval x=\"$x\"]) + if test "X$x" = "X-L$additional_libdir"; then + haveit=yes + break + fi + done + if test -z "$haveit"; then + if test -d "$additional_libdir"; then + dnl Really add $additional_libdir to $LTLIBNAME. + LTLIB[]NAME="${LTLIB[]NAME}${LTLIB[]NAME:+ }-L$additional_libdir" + fi + fi + fi + fi + ;; + -R*) + dir=`echo "X$dep" | sed -e 's/^X-R//'` + if test "$enable_rpath" != no; then + dnl Potentially add DIR to rpathdirs. + dnl The rpathdirs will be appended to $LIBNAME at the end. + haveit= + for x in $rpathdirs; do + if test "X$x" = "X$dir"; then + haveit=yes + break + fi + done + if test -z "$haveit"; then + rpathdirs="$rpathdirs $dir" + fi + dnl Potentially add DIR to ltrpathdirs. + dnl The ltrpathdirs will be appended to $LTLIBNAME at the end. + haveit= + for x in $ltrpathdirs; do + if test "X$x" = "X$dir"; then + haveit=yes + break + fi + done + if test -z "$haveit"; then + ltrpathdirs="$ltrpathdirs $dir" + fi + fi + ;; + -l*) + dnl Handle this in the next round. + names_next_round="$names_next_round "`echo "X$dep" | sed -e 's/^X-l//'` + ;; + *.la) + dnl Handle this in the next round. Throw away the .la's + dnl directory; it is already contained in a preceding -L + dnl option. + names_next_round="$names_next_round "`echo "X$dep" | sed -e 's,^X.*/,,' -e 's,^lib,,' -e 's,\.la$,,'` + ;; + *) + dnl Most likely an immediate library name. + LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$dep" + LTLIB[]NAME="${LTLIB[]NAME}${LTLIB[]NAME:+ }$dep" + ;; + esac + done + fi + else + dnl Didn't find the library; assume it is in the system directories + dnl known to the linker and runtime loader. (All the system + dnl directories known to the linker should also be known to the + dnl runtime loader, otherwise the system is severely misconfigured.) + LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }-l$name" + LTLIB[]NAME="${LTLIB[]NAME}${LTLIB[]NAME:+ }-l$name" + fi + fi + fi + done + done + if test "X$rpathdirs" != "X"; then + if test -n "$acl_hardcode_libdir_separator"; then + dnl Weird platform: only the last -rpath option counts, the user must + dnl pass all path elements in one option. We can arrange that for a + dnl single library, but not when more than one $LIBNAMEs are used. + alldirs= + for found_dir in $rpathdirs; do + alldirs="${alldirs}${alldirs:+$acl_hardcode_libdir_separator}$found_dir" + done + dnl Note: acl_hardcode_libdir_flag_spec uses $libdir and $wl. + acl_save_libdir="$libdir" + libdir="$alldirs" + eval flag=\"$acl_hardcode_libdir_flag_spec\" + libdir="$acl_save_libdir" + LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$flag" + else + dnl The -rpath options are cumulative. + for found_dir in $rpathdirs; do + acl_save_libdir="$libdir" + libdir="$found_dir" + eval flag=\"$acl_hardcode_libdir_flag_spec\" + libdir="$acl_save_libdir" + LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$flag" + done + fi + fi + if test "X$ltrpathdirs" != "X"; then + dnl When using libtool, the option that works for both libraries and + dnl executables is -R. The -R options are cumulative. + for found_dir in $ltrpathdirs; do + LTLIB[]NAME="${LTLIB[]NAME}${LTLIB[]NAME:+ }-R$found_dir" + done + fi +]) + +dnl AC_LIB_APPENDTOVAR(VAR, CONTENTS) appends the elements of CONTENTS to VAR, +dnl unless already present in VAR. +dnl Works only for CPPFLAGS, not for LIB* variables because that sometimes +dnl contains two or three consecutive elements that belong together. +AC_DEFUN([AC_LIB_APPENDTOVAR], +[ + for element in [$2]; do + haveit= + for x in $[$1]; do + AC_LIB_WITH_FINAL_PREFIX([eval x=\"$x\"]) + if test "X$x" = "X$element"; then + haveit=yes + break + fi + done + if test -z "$haveit"; then + [$1]="${[$1]}${[$1]:+ }$element" + fi + done +]) + +dnl For those cases where a variable contains several -L and -l options +dnl referring to unknown libraries and directories, this macro determines the +dnl necessary additional linker options for the runtime path. +dnl AC_LIB_LINKFLAGS_FROM_LIBS([LDADDVAR], [LIBSVALUE], [USE-LIBTOOL]) +dnl sets LDADDVAR to linker options needed together with LIBSVALUE. +dnl If USE-LIBTOOL evaluates to non-empty, linking with libtool is assumed, +dnl otherwise linking without libtool is assumed. +AC_DEFUN([AC_LIB_LINKFLAGS_FROM_LIBS], +[ + AC_REQUIRE([AC_LIB_RPATH]) + AC_REQUIRE([AC_LIB_PREPARE_MULTILIB]) + $1= + if test "$enable_rpath" != no; then + if test -n "$acl_hardcode_libdir_flag_spec" && test "$acl_hardcode_minus_L" = no; then + dnl Use an explicit option to hardcode directories into the resulting + dnl binary. + rpathdirs= + next= + for opt in $2; do + if test -n "$next"; then + dir="$next" + dnl No need to hardcode the standard /usr/lib. + if test "X$dir" != "X/usr/$acl_libdirstem"; then + rpathdirs="$rpathdirs $dir" + fi + next= + else + case $opt in + -L) next=yes ;; + -L*) dir=`echo "X$opt" | sed -e 's,^X-L,,'` + dnl No need to hardcode the standard /usr/lib. + if test "X$dir" != "X/usr/$acl_libdirstem"; then + rpathdirs="$rpathdirs $dir" + fi + next= ;; + *) next= ;; + esac + fi + done + if test "X$rpathdirs" != "X"; then + if test -n ""$3""; then + dnl libtool is used for linking. Use -R options. + for dir in $rpathdirs; do + $1="${$1}${$1:+ }-R$dir" + done + else + dnl The linker is used for linking directly. + if test -n "$acl_hardcode_libdir_separator"; then + dnl Weird platform: only the last -rpath option counts, the user + dnl must pass all path elements in one option. + alldirs= + for dir in $rpathdirs; do + alldirs="${alldirs}${alldirs:+$acl_hardcode_libdir_separator}$dir" + done + acl_save_libdir="$libdir" + libdir="$alldirs" + eval flag=\"$acl_hardcode_libdir_flag_spec\" + libdir="$acl_save_libdir" + $1="$flag" + else + dnl The -rpath options are cumulative. + for dir in $rpathdirs; do + acl_save_libdir="$libdir" + libdir="$dir" + eval flag=\"$acl_hardcode_libdir_flag_spec\" + libdir="$acl_save_libdir" + $1="${$1}${$1:+ }$flag" + done + fi + fi + fi + fi + fi + AC_SUBST([$1]) +]) diff --git a/gl/m4/lib-prefix.m4 b/gl/m4/lib-prefix.m4 new file mode 100644 index 00000000..a8684e17 --- /dev/null +++ b/gl/m4/lib-prefix.m4 @@ -0,0 +1,185 @@ +# lib-prefix.m4 serial 5 (gettext-0.15) +dnl Copyright (C) 2001-2005 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Bruno Haible. + +dnl AC_LIB_ARG_WITH is synonymous to AC_ARG_WITH in autoconf-2.13, and +dnl similar to AC_ARG_WITH in autoconf 2.52...2.57 except that is doesn't +dnl require excessive bracketing. +ifdef([AC_HELP_STRING], +[AC_DEFUN([AC_LIB_ARG_WITH], [AC_ARG_WITH([$1],[[$2]],[$3],[$4])])], +[AC_DEFUN([AC_][LIB_ARG_WITH], [AC_ARG_WITH([$1],[$2],[$3],[$4])])]) + +dnl AC_LIB_PREFIX adds to the CPPFLAGS and LDFLAGS the flags that are needed +dnl to access previously installed libraries. The basic assumption is that +dnl a user will want packages to use other packages he previously installed +dnl with the same --prefix option. +dnl This macro is not needed if only AC_LIB_LINKFLAGS is used to locate +dnl libraries, but is otherwise very convenient. +AC_DEFUN([AC_LIB_PREFIX], +[ + AC_BEFORE([$0], [AC_LIB_LINKFLAGS]) + AC_REQUIRE([AC_PROG_CC]) + AC_REQUIRE([AC_CANONICAL_HOST]) + AC_REQUIRE([AC_LIB_PREPARE_MULTILIB]) + AC_REQUIRE([AC_LIB_PREPARE_PREFIX]) + dnl By default, look in $includedir and $libdir. + use_additional=yes + AC_LIB_WITH_FINAL_PREFIX([ + eval additional_includedir=\"$includedir\" + eval additional_libdir=\"$libdir\" + ]) + AC_LIB_ARG_WITH([lib-prefix], +[ --with-lib-prefix[=DIR] search for libraries in DIR/include and DIR/lib + --without-lib-prefix don't search for libraries in includedir and libdir], +[ + if test "X$withval" = "Xno"; then + use_additional=no + else + if test "X$withval" = "X"; then + AC_LIB_WITH_FINAL_PREFIX([ + eval additional_includedir=\"$includedir\" + eval additional_libdir=\"$libdir\" + ]) + else + additional_includedir="$withval/include" + additional_libdir="$withval/$acl_libdirstem" + fi + fi +]) + if test $use_additional = yes; then + dnl Potentially add $additional_includedir to $CPPFLAGS. + dnl But don't add it + dnl 1. if it's the standard /usr/include, + dnl 2. if it's already present in $CPPFLAGS, + dnl 3. if it's /usr/local/include and we are using GCC on Linux, + dnl 4. if it doesn't exist as a directory. + if test "X$additional_includedir" != "X/usr/include"; then + haveit= + for x in $CPPFLAGS; do + AC_LIB_WITH_FINAL_PREFIX([eval x=\"$x\"]) + if test "X$x" = "X-I$additional_includedir"; then + haveit=yes + break + fi + done + if test -z "$haveit"; then + if test "X$additional_includedir" = "X/usr/local/include"; then + if test -n "$GCC"; then + case $host_os in + linux* | gnu* | k*bsd*-gnu) haveit=yes;; + esac + fi + fi + if test -z "$haveit"; then + if test -d "$additional_includedir"; then + dnl Really add $additional_includedir to $CPPFLAGS. + CPPFLAGS="${CPPFLAGS}${CPPFLAGS:+ }-I$additional_includedir" + fi + fi + fi + fi + dnl Potentially add $additional_libdir to $LDFLAGS. + dnl But don't add it + dnl 1. if it's the standard /usr/lib, + dnl 2. if it's already present in $LDFLAGS, + dnl 3. if it's /usr/local/lib and we are using GCC on Linux, + dnl 4. if it doesn't exist as a directory. + if test "X$additional_libdir" != "X/usr/$acl_libdirstem"; then + haveit= + for x in $LDFLAGS; do + AC_LIB_WITH_FINAL_PREFIX([eval x=\"$x\"]) + if test "X$x" = "X-L$additional_libdir"; then + haveit=yes + break + fi + done + if test -z "$haveit"; then + if test "X$additional_libdir" = "X/usr/local/$acl_libdirstem"; then + if test -n "$GCC"; then + case $host_os in + linux*) haveit=yes;; + esac + fi + fi + if test -z "$haveit"; then + if test -d "$additional_libdir"; then + dnl Really add $additional_libdir to $LDFLAGS. + LDFLAGS="${LDFLAGS}${LDFLAGS:+ }-L$additional_libdir" + fi + fi + fi + fi + fi +]) + +dnl AC_LIB_PREPARE_PREFIX creates variables acl_final_prefix, +dnl acl_final_exec_prefix, containing the values to which $prefix and +dnl $exec_prefix will expand at the end of the configure script. +AC_DEFUN([AC_LIB_PREPARE_PREFIX], +[ + dnl Unfortunately, prefix and exec_prefix get only finally determined + dnl at the end of configure. + if test "X$prefix" = "XNONE"; then + acl_final_prefix="$ac_default_prefix" + else + acl_final_prefix="$prefix" + fi + if test "X$exec_prefix" = "XNONE"; then + acl_final_exec_prefix='${prefix}' + else + acl_final_exec_prefix="$exec_prefix" + fi + acl_save_prefix="$prefix" + prefix="$acl_final_prefix" + eval acl_final_exec_prefix=\"$acl_final_exec_prefix\" + prefix="$acl_save_prefix" +]) + +dnl AC_LIB_WITH_FINAL_PREFIX([statement]) evaluates statement, with the +dnl variables prefix and exec_prefix bound to the values they will have +dnl at the end of the configure script. +AC_DEFUN([AC_LIB_WITH_FINAL_PREFIX], +[ + acl_save_prefix="$prefix" + prefix="$acl_final_prefix" + acl_save_exec_prefix="$exec_prefix" + exec_prefix="$acl_final_exec_prefix" + $1 + exec_prefix="$acl_save_exec_prefix" + prefix="$acl_save_prefix" +]) + +dnl AC_LIB_PREPARE_MULTILIB creates a variable acl_libdirstem, containing +dnl the basename of the libdir, either "lib" or "lib64". +AC_DEFUN([AC_LIB_PREPARE_MULTILIB], +[ + dnl There is no formal standard regarding lib and lib64. The current + dnl practice is that on a system supporting 32-bit and 64-bit instruction + dnl sets or ABIs, 64-bit libraries go under $prefix/lib64 and 32-bit + dnl libraries go under $prefix/lib. We determine the compiler's default + dnl mode by looking at the compiler's library search path. If at least + dnl of its elements ends in /lib64 or points to a directory whose absolute + dnl pathname ends in /lib64, we assume a 64-bit ABI. Otherwise we use the + dnl default, namely "lib". + acl_libdirstem=lib + searchpath=`(LC_ALL=C $CC -print-search-dirs) 2>/dev/null | sed -n -e 's,^libraries: ,,p' | sed -e 's,^=,,'` + if test -n "$searchpath"; then + acl_save_IFS="${IFS= }"; IFS=":" + for searchdir in $searchpath; do + if test -d "$searchdir"; then + case "$searchdir" in + */lib64/ | */lib64 ) acl_libdirstem=lib64 ;; + *) searchdir=`cd "$searchdir" && pwd` + case "$searchdir" in + */lib64 ) acl_libdirstem=lib64 ;; + esac ;; + esac + fi + done + IFS="$acl_save_IFS" + fi +]) diff --git a/gl/m4/lock.m4 b/gl/m4/lock.m4 new file mode 100644 index 00000000..0224f2ff --- /dev/null +++ b/gl/m4/lock.m4 @@ -0,0 +1,311 @@ +# lock.m4 serial 6 (gettext-0.16) +dnl Copyright (C) 2005-2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Bruno Haible. + +dnl Tests for a multithreading library to be used. +dnl Defines at most one of the macros USE_POSIX_THREADS, USE_SOLARIS_THREADS, +dnl USE_PTH_THREADS, USE_WIN32_THREADS +dnl Sets the variables LIBTHREAD and LTLIBTHREAD to the linker options for use +dnl in a Makefile (LIBTHREAD for use without libtool, LTLIBTHREAD for use with +dnl libtool). +dnl Sets the variables LIBMULTITHREAD and LTLIBMULTITHREAD similarly, for +dnl programs that really need multithread functionality. The difference +dnl between LIBTHREAD and LIBMULTITHREAD is that on platforms supporting weak +dnl symbols, typically LIBTHREAD="" whereas LIBMULTITHREAD="-lpthread". +dnl Adds to CPPFLAGS the flag -D_REENTRANT or -D_THREAD_SAFE if needed for +dnl multithread-safe programs. + +AC_DEFUN([gl_LOCK_EARLY], +[ + AC_REQUIRE([gl_LOCK_EARLY_BODY]) +]) + +dnl The guts of gl_LOCK_EARLY. Needs to be expanded only once. + +AC_DEFUN([gl_LOCK_EARLY_BODY], +[ + dnl Ordering constraints: This macro modifies CPPFLAGS in a way that + dnl influences the result of the autoconf tests that test for *_unlocked + dnl declarations, on AIX 5 at least. Therefore it must come early. + AC_BEFORE([$0], [gl_FUNC_GLIBC_UNLOCKED_IO])dnl + AC_BEFORE([$0], [gl_ARGP])dnl + + AC_REQUIRE([AC_CANONICAL_HOST]) + AC_REQUIRE([AC_GNU_SOURCE]) dnl needed for pthread_rwlock_t on glibc systems + dnl Check for multithreading. + AC_ARG_ENABLE(threads, +AC_HELP_STRING([--enable-threads={posix|solaris|pth|win32}], [specify multithreading API]) +AC_HELP_STRING([--disable-threads], [build without multithread safety]), + [gl_use_threads=$enableval], + [case "$host_os" in + dnl Disable multithreading by default on OSF/1, because it interferes + dnl with fork()/exec(): When msgexec is linked with -lpthread, its child + dnl process gets an endless segmentation fault inside execvp(). + osf*) gl_use_threads=no ;; + *) gl_use_threads=yes ;; + esac + ]) + if test "$gl_use_threads" = yes || test "$gl_use_threads" = posix; then + # For using <pthread.h>: + case "$host_os" in + osf*) + # On OSF/1, the compiler needs the flag -D_REENTRANT so that it + # groks <pthread.h>. cc also understands the flag -pthread, but + # we don't use it because 1. gcc-2.95 doesn't understand -pthread, + # 2. putting a flag into CPPFLAGS that has an effect on the linker + # causes the AC_TRY_LINK test below to succeed unexpectedly, + # leading to wrong values of LIBTHREAD and LTLIBTHREAD. + CPPFLAGS="$CPPFLAGS -D_REENTRANT" + ;; + esac + # Some systems optimize for single-threaded programs by default, and + # need special flags to disable these optimizations. For example, the + # definition of 'errno' in <errno.h>. + case "$host_os" in + aix* | freebsd*) CPPFLAGS="$CPPFLAGS -D_THREAD_SAFE" ;; + solaris*) CPPFLAGS="$CPPFLAGS -D_REENTRANT" ;; + esac + fi +]) + +dnl The guts of gl_LOCK. Needs to be expanded only once. + +AC_DEFUN([gl_LOCK_BODY], +[ + AC_REQUIRE([gl_LOCK_EARLY_BODY]) + gl_threads_api=none + LIBTHREAD= + LTLIBTHREAD= + LIBMULTITHREAD= + LTLIBMULTITHREAD= + if test "$gl_use_threads" != no; then + dnl Check whether the compiler and linker support weak declarations. + AC_MSG_CHECKING([whether imported symbols can be declared weak]) + gl_have_weak=no + AC_TRY_LINK([extern void xyzzy (); +#pragma weak xyzzy], [xyzzy();], [gl_have_weak=yes]) + AC_MSG_RESULT([$gl_have_weak]) + if test "$gl_use_threads" = yes || test "$gl_use_threads" = posix; then + # On OSF/1, the compiler needs the flag -pthread or -D_REENTRANT so that + # it groks <pthread.h>. It's added above, in gl_LOCK_EARLY_BODY. + AC_CHECK_HEADER(pthread.h, gl_have_pthread_h=yes, gl_have_pthread_h=no) + if test "$gl_have_pthread_h" = yes; then + # Other possible tests: + # -lpthreads (FSU threads, PCthreads) + # -lgthreads + gl_have_pthread= + # Test whether both pthread_mutex_lock and pthread_mutexattr_init exist + # in libc. IRIX 6.5 has the first one in both libc and libpthread, but + # the second one only in libpthread, and lock.c needs it. + AC_TRY_LINK([#include <pthread.h>], + [pthread_mutex_lock((pthread_mutex_t*)0); + pthread_mutexattr_init((pthread_mutexattr_t*)0);], + [gl_have_pthread=yes]) + # Test for libpthread by looking for pthread_kill. (Not pthread_self, + # since it is defined as a macro on OSF/1.) + if test -n "$gl_have_pthread"; then + # The program links fine without libpthread. But it may actually + # need to link with libpthread in order to create multiple threads. + AC_CHECK_LIB(pthread, pthread_kill, + [LIBMULTITHREAD=-lpthread LTLIBMULTITHREAD=-lpthread + # On Solaris and HP-UX, most pthread functions exist also in libc. + # Therefore pthread_in_use() needs to actually try to create a + # thread: pthread_create from libc will fail, whereas + # pthread_create will actually create a thread. + case "$host_os" in + solaris* | hpux*) + AC_DEFINE([PTHREAD_IN_USE_DETECTION_HARD], 1, + [Define if the pthread_in_use() detection is hard.]) + esac + ]) + else + # Some library is needed. Try libpthread and libc_r. + AC_CHECK_LIB(pthread, pthread_kill, + [gl_have_pthread=yes + LIBTHREAD=-lpthread LTLIBTHREAD=-lpthread + LIBMULTITHREAD=-lpthread LTLIBMULTITHREAD=-lpthread]) + if test -z "$gl_have_pthread"; then + # For FreeBSD 4. + AC_CHECK_LIB(c_r, pthread_kill, + [gl_have_pthread=yes + LIBTHREAD=-lc_r LTLIBTHREAD=-lc_r + LIBMULTITHREAD=-lc_r LTLIBMULTITHREAD=-lc_r]) + fi + fi + if test -n "$gl_have_pthread"; then + gl_threads_api=posix + AC_DEFINE([USE_POSIX_THREADS], 1, + [Define if the POSIX multithreading library can be used.]) + if test -n "$LIBMULTITHREAD" || test -n "$LTLIBMULTITHREAD"; then + if test $gl_have_weak = yes; then + AC_DEFINE([USE_POSIX_THREADS_WEAK], 1, + [Define if references to the POSIX multithreading library should be made weak.]) + LIBTHREAD= + LTLIBTHREAD= + fi + fi + # OSF/1 4.0 and MacOS X 10.1 lack the pthread_rwlock_t type and the + # pthread_rwlock_* functions. + AC_CHECK_TYPE([pthread_rwlock_t], + [AC_DEFINE([HAVE_PTHREAD_RWLOCK], 1, + [Define if the POSIX multithreading library has read/write locks.])], + [], + [#include <pthread.h>]) + # glibc defines PTHREAD_MUTEX_RECURSIVE as enum, not as a macro. + AC_TRY_COMPILE([#include <pthread.h>], + [#if __FreeBSD__ == 4 +error "No, in FreeBSD 4.0 recursive mutexes actually don't work." +#else +int x = (int)PTHREAD_MUTEX_RECURSIVE; +return !x; +#endif], + [AC_DEFINE([HAVE_PTHREAD_MUTEX_RECURSIVE], 1, + [Define if the <pthread.h> defines PTHREAD_MUTEX_RECURSIVE.])]) + fi + fi + fi + if test -z "$gl_have_pthread"; then + if test "$gl_use_threads" = yes || test "$gl_use_threads" = solaris; then + gl_have_solaristhread= + gl_save_LIBS="$LIBS" + LIBS="$LIBS -lthread" + AC_TRY_LINK([#include <thread.h> +#include <synch.h>], + [thr_self();], + [gl_have_solaristhread=yes]) + LIBS="$gl_save_LIBS" + if test -n "$gl_have_solaristhread"; then + gl_threads_api=solaris + LIBTHREAD=-lthread + LTLIBTHREAD=-lthread + LIBMULTITHREAD="$LIBTHREAD" + LTLIBMULTITHREAD="$LTLIBTHREAD" + AC_DEFINE([USE_SOLARIS_THREADS], 1, + [Define if the old Solaris multithreading library can be used.]) + if test $gl_have_weak = yes; then + AC_DEFINE([USE_SOLARIS_THREADS_WEAK], 1, + [Define if references to the old Solaris multithreading library should be made weak.]) + LIBTHREAD= + LTLIBTHREAD= + fi + fi + fi + fi + if test "$gl_use_threads" = pth; then + gl_save_CPPFLAGS="$CPPFLAGS" + AC_LIB_LINKFLAGS(pth) + gl_have_pth= + gl_save_LIBS="$LIBS" + LIBS="$LIBS -lpth" + AC_TRY_LINK([#include <pth.h>], [pth_self();], gl_have_pth=yes) + LIBS="$gl_save_LIBS" + if test -n "$gl_have_pth"; then + gl_threads_api=pth + LIBTHREAD="$LIBPTH" + LTLIBTHREAD="$LTLIBPTH" + LIBMULTITHREAD="$LIBTHREAD" + LTLIBMULTITHREAD="$LTLIBTHREAD" + AC_DEFINE([USE_PTH_THREADS], 1, + [Define if the GNU Pth multithreading library can be used.]) + if test -n "$LIBMULTITHREAD" || test -n "$LTLIBMULTITHREAD"; then + if test $gl_have_weak = yes; then + AC_DEFINE([USE_PTH_THREADS_WEAK], 1, + [Define if references to the GNU Pth multithreading library should be made weak.]) + LIBTHREAD= + LTLIBTHREAD= + fi + fi + else + CPPFLAGS="$gl_save_CPPFLAGS" + fi + fi + if test -z "$gl_have_pthread"; then + if test "$gl_use_threads" = yes || test "$gl_use_threads" = win32; then + if { case "$host_os" in + mingw*) true;; + *) false;; + esac + }; then + gl_threads_api=win32 + AC_DEFINE([USE_WIN32_THREADS], 1, + [Define if the Win32 multithreading API can be used.]) + fi + fi + fi + fi + AC_MSG_CHECKING([for multithread API to use]) + AC_MSG_RESULT([$gl_threads_api]) + AC_SUBST(LIBTHREAD) + AC_SUBST(LTLIBTHREAD) + AC_SUBST(LIBMULTITHREAD) + AC_SUBST(LTLIBMULTITHREAD) +]) + +AC_DEFUN([gl_LOCK], +[ + AC_REQUIRE([gl_LOCK_EARLY]) + AC_REQUIRE([gl_LOCK_BODY]) + gl_PREREQ_LOCK +]) + +# Prerequisites of lib/lock.c. +AC_DEFUN([gl_PREREQ_LOCK], [ + AC_REQUIRE([AC_C_INLINE]) +]) + +dnl Survey of platforms: +dnl +dnl Platform Available Compiler Supports test-lock +dnl flavours option weak result +dnl --------------- --------- --------- -------- --------- +dnl Linux 2.4/glibc posix -lpthread Y OK +dnl +dnl GNU Hurd/glibc posix +dnl +dnl FreeBSD 5.3 posix -lc_r Y +dnl posix -lkse ? Y +dnl posix -lpthread ? Y +dnl posix -lthr Y +dnl +dnl FreeBSD 5.2 posix -lc_r Y +dnl posix -lkse Y +dnl posix -lthr Y +dnl +dnl FreeBSD 4.0,4.10 posix -lc_r Y OK +dnl +dnl NetBSD 1.6 -- +dnl +dnl OpenBSD 3.4 posix -lpthread Y OK +dnl +dnl MacOS X 10.[123] posix -lpthread Y OK +dnl +dnl Solaris 7,8,9 posix -lpthread Y Sol 7,8: 0.0; Sol 9: OK +dnl solaris -lthread Y Sol 7,8: 0.0; Sol 9: OK +dnl +dnl HP-UX 11 posix -lpthread N (cc) OK +dnl Y (gcc) +dnl +dnl IRIX 6.5 posix -lpthread Y 0.5 +dnl +dnl AIX 4.3,5.1 posix -lpthread N AIX 4: 0.5; AIX 5: OK +dnl +dnl OSF/1 4.0,5.1 posix -pthread (cc) N OK +dnl -lpthread (gcc) Y +dnl +dnl Cygwin posix -lpthread Y OK +dnl +dnl Any of the above pth -lpth 0.0 +dnl +dnl Mingw win32 N OK +dnl +dnl BeOS 5 -- +dnl +dnl The test-lock result shows what happens if in test-lock.c EXPLICIT_YIELD is +dnl turned off: +dnl OK if all three tests terminate OK, +dnl 0.5 if the first test terminates OK but the second one loops endlessly, +dnl 0.0 if the first test already loops endlessly. diff --git a/gl/m4/longdouble.m4 b/gl/m4/longdouble.m4 new file mode 100644 index 00000000..25590f47 --- /dev/null +++ b/gl/m4/longdouble.m4 @@ -0,0 +1,31 @@ +# longdouble.m4 serial 2 (gettext-0.15) +dnl Copyright (C) 2002-2003, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Bruno Haible. +dnl Test whether the compiler supports the 'long double' type. +dnl Prerequisite: AC_PROG_CC + +dnl This file is only needed in autoconf <= 2.59. Newer versions of autoconf +dnl have a macro AC_TYPE_LONG_DOUBLE with identical semantics. + +AC_DEFUN([gt_TYPE_LONGDOUBLE], +[ + AC_CACHE_CHECK([for long double], gt_cv_c_long_double, + [if test "$GCC" = yes; then + gt_cv_c_long_double=yes + else + AC_TRY_COMPILE([ + /* The Stardent Vistra knows sizeof(long double), but does not support it. */ + long double foo = 0.0; + /* On Ultrix 4.3 cc, long double is 4 and double is 8. */ + int array [2*(sizeof(long double) >= sizeof(double)) - 1]; + ], , + gt_cv_c_long_double=yes, gt_cv_c_long_double=no) + fi]) + if test $gt_cv_c_long_double = yes; then + AC_DEFINE(HAVE_LONG_DOUBLE, 1, [Define if you have the 'long double' type.]) + fi +]) diff --git a/gl/m4/longlong.m4 b/gl/m4/longlong.m4 new file mode 100644 index 00000000..1f9e862e --- /dev/null +++ b/gl/m4/longlong.m4 @@ -0,0 +1,72 @@ +# longlong.m4 serial 10 +dnl Copyright (C) 1999-2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Paul Eggert. + +# Define HAVE_LONG_LONG_INT if 'long long int' works. +# This fixes a bug in Autoconf 2.60, but can be removed once we +# assume 2.61 everywhere. + +# Note: If the type 'long long int' exists but is only 32 bits large +# (as on some very old compilers), HAVE_LONG_LONG_INT will not be +# defined. In this case you can treat 'long long int' like 'long int'. + +AC_DEFUN([AC_TYPE_LONG_LONG_INT], +[ + AC_CACHE_CHECK([for long long int], [ac_cv_type_long_long_int], + [AC_LINK_IFELSE( + [AC_LANG_PROGRAM( + [[long long int ll = 9223372036854775807ll; + long long int nll = -9223372036854775807LL; + typedef int a[((-9223372036854775807LL < 0 + && 0 < 9223372036854775807ll) + ? 1 : -1)]; + int i = 63;]], + [[long long int llmax = 9223372036854775807ll; + return ((ll << 63) | (ll >> 63) | (ll < i) | (ll > i) + | (llmax / ll) | (llmax % ll));]])], + [dnl This catches a bug in Tandem NonStop Kernel (OSS) cc -O circa 2004. + dnl If cross compiling, assume the bug isn't important, since + dnl nobody cross compiles for this platform as far as we know. + AC_RUN_IFELSE( + [AC_LANG_PROGRAM( + [[@%:@include <limits.h> + @%:@ifndef LLONG_MAX + @%:@ define HALF \ + (1LL << (sizeof (long long int) * CHAR_BIT - 2)) + @%:@ define LLONG_MAX (HALF - 1 + HALF) + @%:@endif]], + [[long long int n = 1; + int i; + for (i = 0; ; i++) + { + long long int m = n << i; + if (m >> i != n) + return 1; + if (LLONG_MAX / 2 < m) + break; + } + return 0;]])], + [ac_cv_type_long_long_int=yes], + [ac_cv_type_long_long_int=no], + [ac_cv_type_long_long_int=yes])], + [ac_cv_type_long_long_int=no])]) + if test $ac_cv_type_long_long_int = yes; then + AC_DEFINE([HAVE_LONG_LONG_INT], 1, + [Define to 1 if the system has the type `long long int'.]) + fi +]) + +# This macro is obsolescent and should go away soon. +AC_DEFUN([gl_AC_TYPE_LONG_LONG], +[ + AC_REQUIRE([AC_TYPE_LONG_LONG_INT]) + ac_cv_type_long_long=$ac_cv_type_long_long_int + if test $ac_cv_type_long_long = yes; then + AC_DEFINE(HAVE_LONG_LONG, 1, + [Define if you have the 'long long' type.]) + fi +]) diff --git a/gl/m4/ls-mntd-fs.m4 b/gl/m4/ls-mntd-fs.m4 new file mode 100644 index 00000000..21ac4e74 --- /dev/null +++ b/gl/m4/ls-mntd-fs.m4 @@ -0,0 +1,337 @@ +#serial 26 +# How to list mounted file systems. + +# Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2006 Free Software +# Foundation, Inc. +# +# This file is free software; the Free Software Foundation +# gives unlimited permission to copy and/or distribute it, +# with or without modifications, as long as this notice is preserved. + +dnl From Jim Meyering. +dnl +dnl This is not pretty. I've just taken the autoconf code and wrapped +dnl it in an AC_DEFUN and made some other fixes. +dnl + +# Replace Autoconf's AC_FUNC_GETMNTENT to work around a bug in Autoconf +# through Autoconf 2.59. We can remove this once we assume Autoconf 2.60 +# or later. +AC_DEFUN([AC_FUNC_GETMNTENT], +[# getmntent is in the standard C library on UNICOS, in -lsun on Irix 4, +# -lseq on Dynix/PTX, -lgen on Unixware. +AC_SEARCH_LIBS(getmntent, [sun seq gen]) +AC_CHECK_FUNCS(getmntent) +]) + +# gl_LIST_MOUNTED_FILE_SYSTEMS([ACTION-IF-FOUND[, ACTION-IF-NOT-FOUND]]) +AC_DEFUN([gl_LIST_MOUNTED_FILE_SYSTEMS], + [ +AC_CHECK_FUNCS(listmntent getmntinfo) +AC_CHECK_HEADERS_ONCE(sys/param.h sys/statvfs.h) + +# We must include grp.h before ucred.h on OSF V4.0, since ucred.h uses +# NGROUPS (as the array dimension for a struct member) without a definition. +AC_CHECK_HEADERS(sys/ucred.h, [], [], [#include <grp.h>]) + +AC_CHECK_HEADERS(sys/mount.h, [], [], + [AC_INCLUDES_DEFAULT + [#if HAVE_SYS_PARAM_H + #include <sys/param.h> + #endif]]) + +AC_CHECK_HEADERS(mntent.h sys/fs_types.h) + getfsstat_includes="\ +$ac_includes_default +#if HAVE_SYS_PARAM_H +# include <sys/param.h> /* needed by powerpc-apple-darwin1.3.7 */ +#endif +#if HAVE_SYS_UCRED_H +# include <grp.h> /* needed for definition of NGROUPS */ +# include <sys/ucred.h> /* needed by powerpc-apple-darwin1.3.7 */ +#endif +#if HAVE_SYS_MOUNT_H +# include <sys/mount.h> +#endif +#if HAVE_SYS_FS_TYPES_H +# include <sys/fs_types.h> /* needed by powerpc-apple-darwin1.3.7 */ +#endif +" +AC_CHECK_MEMBERS([struct fsstat.f_fstypename],,,[$getfsstat_includes]) + +# Determine how to get the list of mounted file systems. +ac_list_mounted_fs= + +# If the getmntent function is available but not in the standard library, +# make sure LIBS contains the appropriate -l option. +AC_FUNC_GETMNTENT + +# This test must precede the ones for getmntent because Unicos-9 is +# reported to have the getmntent function, but its support is incompatible +# with other getmntent implementations. + +# NOTE: Normally, I wouldn't use a check for system type as I've done for +# `CRAY' below since that goes against the whole autoconf philosophy. But +# I think there is too great a chance that some non-Cray system has a +# function named listmntent to risk the false positive. + +if test -z "$ac_list_mounted_fs"; then + # Cray UNICOS 9 + AC_MSG_CHECKING([for listmntent of Cray/Unicos-9]) + AC_CACHE_VAL(fu_cv_sys_mounted_cray_listmntent, + [fu_cv_sys_mounted_cray_listmntent=no + AC_EGREP_CPP(yes, + [#ifdef _CRAY +yes +#endif + ], [test $ac_cv_func_listmntent = yes \ + && fu_cv_sys_mounted_cray_listmntent=yes] + ) + ] + ) + AC_MSG_RESULT($fu_cv_sys_mounted_cray_listmntent) + if test $fu_cv_sys_mounted_cray_listmntent = yes; then + ac_list_mounted_fs=found + AC_DEFINE(MOUNTED_LISTMNTENT, 1, + [Define if there is a function named listmntent that can be used to + list all mounted file systems. (UNICOS)]) + fi +fi + +if test -z "$ac_list_mounted_fs"; then + # AIX. + AC_MSG_CHECKING([for mntctl function and struct vmount]) + AC_CACHE_VAL(fu_cv_sys_mounted_vmount, + [AC_TRY_CPP([#include <fshelp.h>], + fu_cv_sys_mounted_vmount=yes, + fu_cv_sys_mounted_vmount=no)]) + AC_MSG_RESULT($fu_cv_sys_mounted_vmount) + if test $fu_cv_sys_mounted_vmount = yes; then + ac_list_mounted_fs=found + AC_DEFINE(MOUNTED_VMOUNT, 1, + [Define if there is a function named mntctl that can be used to read + the list of mounted file systems, and there is a system header file + that declares `struct vmount.' (AIX)]) + fi +fi + +if test $ac_cv_func_getmntent = yes; then + + # This system has the getmntent function. + # Determine whether it's the one-argument variant or the two-argument one. + + if test -z "$ac_list_mounted_fs"; then + # 4.3BSD, SunOS, HP-UX, Dynix, Irix + AC_MSG_CHECKING([for one-argument getmntent function]) + AC_CACHE_VAL(fu_cv_sys_mounted_getmntent1, + [AC_TRY_COMPILE([ +/* SunOS 4.1.x /usr/include/mntent.h needs this for FILE */ +#include <stdio.h> + +#include <mntent.h> +#if !defined MOUNTED +# if defined _PATH_MOUNTED /* GNU libc */ +# define MOUNTED _PATH_MOUNTED +# endif +# if defined MNT_MNTTAB /* HP-UX. */ +# define MOUNTED MNT_MNTTAB +# endif +# if defined MNTTABNAME /* Dynix. */ +# define MOUNTED MNTTABNAME +# endif +#endif +], + [ struct mntent *mnt = 0; char *table = MOUNTED; + if (sizeof mnt && sizeof table) return 0;], + fu_cv_sys_mounted_getmntent1=yes, + fu_cv_sys_mounted_getmntent1=no)]) + AC_MSG_RESULT($fu_cv_sys_mounted_getmntent1) + if test $fu_cv_sys_mounted_getmntent1 = yes; then + ac_list_mounted_fs=found + AC_DEFINE(MOUNTED_GETMNTENT1, 1, + [Define if there is a function named getmntent for reading the list + of mounted file systems, and that function takes a single argument. + (4.3BSD, SunOS, HP-UX, Dynix, Irix)]) + fi + fi + + if test -z "$ac_list_mounted_fs"; then + # SVR4 + AC_MSG_CHECKING([for two-argument getmntent function]) + AC_CACHE_VAL(fu_cv_sys_mounted_getmntent2, + [AC_EGREP_HEADER(getmntent, sys/mnttab.h, + fu_cv_sys_mounted_getmntent2=yes, + fu_cv_sys_mounted_getmntent2=no)]) + AC_MSG_RESULT($fu_cv_sys_mounted_getmntent2) + if test $fu_cv_sys_mounted_getmntent2 = yes; then + ac_list_mounted_fs=found + AC_DEFINE(MOUNTED_GETMNTENT2, 1, + [Define if there is a function named getmntent for reading the list of + mounted file systems, and that function takes two arguments. (SVR4)]) + AC_CHECK_FUNCS(hasmntopt) + fi + fi + +fi + +if test -z "$ac_list_mounted_fs"; then + # DEC Alpha running OSF/1, and Apple Darwin 1.3. + # powerpc-apple-darwin1.3.7 needs sys/param.h sys/ucred.h sys/fs_types.h + + AC_MSG_CHECKING([for getfsstat function]) + AC_CACHE_VAL(fu_cv_sys_mounted_getfsstat, + [AC_TRY_LINK([ +#include <sys/types.h> +#if HAVE_STRUCT_FSSTAT_F_FSTYPENAME +# define FS_TYPE(Ent) ((Ent).f_fstypename) +#else +# define FS_TYPE(Ent) mnt_names[(Ent).f_type] +#endif +]$getfsstat_includes +, + [struct statfs *stats; + int numsys = getfsstat ((struct statfs *)0, 0L, MNT_WAIT); + char *t = FS_TYPE (*stats); ], + fu_cv_sys_mounted_getfsstat=yes, + fu_cv_sys_mounted_getfsstat=no)]) + AC_MSG_RESULT($fu_cv_sys_mounted_getfsstat) + if test $fu_cv_sys_mounted_getfsstat = yes; then + ac_list_mounted_fs=found + AC_DEFINE(MOUNTED_GETFSSTAT, 1, + [Define if there is a function named getfsstat for reading the + list of mounted file systems. (DEC Alpha running OSF/1)]) + fi +fi + +if test -z "$ac_list_mounted_fs"; then + # SVR3 + AC_MSG_CHECKING([for FIXME existence of three headers]) + AC_CACHE_VAL(fu_cv_sys_mounted_fread_fstyp, + [AC_TRY_CPP([ +#include <sys/statfs.h> +#include <sys/fstyp.h> +#include <mnttab.h>], + fu_cv_sys_mounted_fread_fstyp=yes, + fu_cv_sys_mounted_fread_fstyp=no)]) + AC_MSG_RESULT($fu_cv_sys_mounted_fread_fstyp) + if test $fu_cv_sys_mounted_fread_fstyp = yes; then + ac_list_mounted_fs=found + AC_DEFINE(MOUNTED_FREAD_FSTYP, 1, + [Define if (like SVR2) there is no specific function for reading the + list of mounted file systems, and your system has these header files: + <sys/fstyp.h> and <sys/statfs.h>. (SVR3)]) + fi +fi + +if test -z "$ac_list_mounted_fs"; then + # 4.4BSD and DEC OSF/1. + AC_MSG_CHECKING([for getmntinfo function]) + AC_CACHE_VAL(fu_cv_sys_mounted_getmntinfo, + [ + test "$ac_cv_func_getmntinfo" = yes \ + && fu_cv_sys_mounted_getmntinfo=yes \ + || fu_cv_sys_mounted_getmntinfo=no + ]) + AC_MSG_RESULT($fu_cv_sys_mounted_getmntinfo) + if test $fu_cv_sys_mounted_getmntinfo = yes; then + AC_MSG_CHECKING([whether getmntinfo returns statvfs structures]) + AC_CACHE_VAL(fu_cv_sys_mounted_getmntinfo2, + [ + AC_TRY_COMPILE([ +#if HAVE_SYS_PARAM_H +# include <sys/param.h> +#endif +#include <sys/types.h> +#if HAVE_SYS_MOUNT_H +# include <sys/mount.h> +#endif +#if HAVE_SYS_STATVFS_H +# include <sys/statvfs.h> +#endif +extern int getmntinfo (struct statfs **, int); + ], [], + [fu_cv_sys_mounted_getmntinfo2=no], + [fu_cv_sys_mounted_getmntinfo2=yes]) + ]) + AC_MSG_RESULT([$fu_cv_sys_mounted_getmntinfo2]) + if test $fu_cv_sys_mounted_getmntinfo2 = no; then + ac_list_mounted_fs=found + AC_DEFINE(MOUNTED_GETMNTINFO, 1, + [Define if there is a function named getmntinfo for reading the + list of mounted file systems and it returns an array of + 'struct statfs'. (4.4BSD, Darwin)]) + else + ac_list_mounted_fs=found + AC_DEFINE(MOUNTED_GETMNTINFO2, 1, + [Define if there is a function named getmntinfo for reading the + list of mounted file systems and it returns an array of + 'struct statvfs'. (NetBSD 3.0)]) + fi + fi +fi + +if test -z "$ac_list_mounted_fs"; then + # Ultrix + AC_MSG_CHECKING([for getmnt function]) + AC_CACHE_VAL(fu_cv_sys_mounted_getmnt, + [AC_TRY_CPP([ +#include <sys/fs_types.h> +#include <sys/mount.h>], + fu_cv_sys_mounted_getmnt=yes, + fu_cv_sys_mounted_getmnt=no)]) + AC_MSG_RESULT($fu_cv_sys_mounted_getmnt) + if test $fu_cv_sys_mounted_getmnt = yes; then + ac_list_mounted_fs=found + AC_DEFINE(MOUNTED_GETMNT, 1, + [Define if there is a function named getmnt for reading the list of + mounted file systems. (Ultrix)]) + fi +fi + +if test -z "$ac_list_mounted_fs"; then + # BeOS + AC_CHECK_FUNCS(next_dev fs_stat_dev) + AC_CHECK_HEADERS(fs_info.h) + AC_MSG_CHECKING([for BEOS mounted file system support functions]) + if test $ac_cv_header_fs_info_h = yes \ + && test $ac_cv_func_next_dev = yes \ + && test $ac_cv_func_fs_stat_dev = yes; then + fu_result=yes + else + fu_result=no + fi + AC_MSG_RESULT($fu_result) + if test $fu_result = yes; then + ac_list_mounted_fs=found + AC_DEFINE(MOUNTED_FS_STAT_DEV, 1, + [Define if there are functions named next_dev and fs_stat_dev for + reading the list of mounted file systems. (BeOS)]) + fi +fi + +if test -z "$ac_list_mounted_fs"; then + # SVR2 + AC_MSG_CHECKING([whether it is possible to resort to fread on /etc/mnttab]) + AC_CACHE_VAL(fu_cv_sys_mounted_fread, + [AC_TRY_CPP([#include <mnttab.h>], + fu_cv_sys_mounted_fread=yes, + fu_cv_sys_mounted_fread=no)]) + AC_MSG_RESULT($fu_cv_sys_mounted_fread) + if test $fu_cv_sys_mounted_fread = yes; then + ac_list_mounted_fs=found + AC_DEFINE(MOUNTED_FREAD, 1, + [Define if there is no specific function for reading the list of + mounted file systems. fread will be used to read /etc/mnttab. + (SVR2) ]) + fi +fi + +if test -z "$ac_list_mounted_fs"; then + AC_MSG_ERROR([could not determine how to read list of mounted file systems]) + # FIXME -- no need to abort building the whole package + # Can't build mountlist.c or anything that needs its functions +fi + +AS_IF([test $ac_list_mounted_fs = found], [$1], [$2]) + + ]) diff --git a/gl/m4/mbchar.m4 b/gl/m4/mbchar.m4 new file mode 100644 index 00000000..5380941b --- /dev/null +++ b/gl/m4/mbchar.m4 @@ -0,0 +1,14 @@ +# mbchar.m4 serial 5 +dnl Copyright (C) 2005-2007 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl autoconf tests required for use of mbchar.m4 +dnl From Bruno Haible. + +AC_DEFUN([gl_MBCHAR], +[ + AC_REQUIRE([AC_GNU_SOURCE]) + AC_LIBOBJ([mbchar]) +]) diff --git a/gl/m4/mbiter.m4 b/gl/m4/mbiter.m4 new file mode 100644 index 00000000..7d51af10 --- /dev/null +++ b/gl/m4/mbiter.m4 @@ -0,0 +1,17 @@ +# mbiter.m4 serial 2 +dnl Copyright (C) 2005 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl autoconf tests required for use of mbiter.h +dnl From Bruno Haible. + +AC_DEFUN([gl_MBITER], +[ + AC_REQUIRE([AC_TYPE_MBSTATE_T]) + dnl The following line is that so the user can test HAVE_MBRTOWC before + dnl #include "mbiter.h" or "mbuiter.h". + AC_REQUIRE([gl_FUNC_MBRTOWC]) + : +]) diff --git a/gl/m4/mbrtowc.m4 b/gl/m4/mbrtowc.m4 new file mode 100644 index 00000000..a3bd9114 --- /dev/null +++ b/gl/m4/mbrtowc.m4 @@ -0,0 +1,31 @@ +# mbrtowc.m4 serial 8 +dnl Copyright (C) 2001-2002, 2004-2005 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Paul Eggert + +dnl This file can be removed, and gl_FUNC_MBRTOWC replaced with +dnl AC_FUNC_MBRTOWC, when autoconf 2.60 can be assumed everywhere. + +AC_DEFUN([gl_FUNC_MBRTOWC], +[ + dnl Same as AC_FUNC_MBRTOWC in autoconf-2.60. + AC_CACHE_CHECK([whether mbrtowc and mbstate_t are properly declared], + gl_cv_func_mbrtowc, + [AC_LINK_IFELSE( + [AC_LANG_PROGRAM( + [[#include <wchar.h>]], + [[wchar_t wc; + char const s[] = ""; + size_t n = 1; + mbstate_t state; + return ! (sizeof state && (mbrtowc) (&wc, s, n, &state));]])], + gl_cv_func_mbrtowc=yes, + gl_cv_func_mbrtowc=no)]) + if test $gl_cv_func_mbrtowc = yes; then + AC_DEFINE([HAVE_MBRTOWC], 1, + [Define to 1 if mbrtowc and mbstate_t are properly declared.]) + fi +]) diff --git a/gl/m4/memchr.m4 b/gl/m4/memchr.m4 new file mode 100644 index 00000000..91b8636e --- /dev/null +++ b/gl/m4/memchr.m4 @@ -0,0 +1,18 @@ +# memchr.m4 serial 4 +dnl Copyright (C) 2002, 2003, 2004 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_FUNC_MEMCHR], +[ + AC_REPLACE_FUNCS(memchr) + if test $ac_cv_func_memchr = no; then + gl_PREREQ_MEMCHR + fi +]) + +# Prerequisites of lib/memchr.c. +AC_DEFUN([gl_PREREQ_MEMCHR], [ + AC_CHECK_HEADERS(bp-sym.h) +]) diff --git a/gl/m4/minmax.m4 b/gl/m4/minmax.m4 new file mode 100644 index 00000000..bbd1ba0b --- /dev/null +++ b/gl/m4/minmax.m4 @@ -0,0 +1,41 @@ +# minmax.m4 serial 2 +dnl Copyright (C) 2005 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_PREREQ(2.52) + +AC_DEFUN([gl_MINMAX], +[ + AC_REQUIRE([gl_PREREQ_MINMAX]) +]) + +# Prerequisites of lib/minmax.h. +AC_DEFUN([gl_PREREQ_MINMAX], +[ + gl_MINMAX_IN_HEADER([limits.h]) + gl_MINMAX_IN_HEADER([sys/param.h]) +]) + +dnl gl_MINMAX_IN_HEADER(HEADER) +dnl The parameter has to be a literal header name; it cannot be macro, +dnl nor a shell variable. (Because autoheader collects only AC_DEFINE +dnl invocations with a literal macro name.) +AC_DEFUN([gl_MINMAX_IN_HEADER], +[ + m4_pushdef([header], AS_TR_SH([$1])) + m4_pushdef([HEADER], AS_TR_CPP([$1])) + AC_CACHE_CHECK([whether <$1> defines MIN and MAX], + [gl_cv_minmax_in_]header, + [AC_TRY_COMPILE([#include <$1> +int x = MIN (42, 17);], [], + [gl_cv_minmax_in_]header[=yes], + [gl_cv_minmax_in_]header[=no])]) + if test $gl_cv_minmax_in_[]header = yes; then + AC_DEFINE([HAVE_MINMAX_IN_]HEADER, 1, + [Define to 1 if <$1> defines the MIN and MAX macros.]) + fi + m4_popdef([HEADER]) + m4_popdef([header]) +]) diff --git a/gl/m4/mountlist.m4 b/gl/m4/mountlist.m4 new file mode 100644 index 00000000..c25f44e7 --- /dev/null +++ b/gl/m4/mountlist.m4 @@ -0,0 +1,23 @@ +#serial 9 +dnl Copyright (C) 2002, 2003, 2004, 2005, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_MOUNTLIST], +[ + gl_LIST_MOUNTED_FILE_SYSTEMS([gl_cv_list_mounted_fs=yes], + [gl_cv_list_mounted_fs=no]) + if test $gl_cv_list_mounted_fs = yes; then + AC_LIBOBJ(mountlist) + gl_PREREQ_MOUNTLIST_EXTRA + fi +]) + +# Prerequisites of lib/mountlist.c not done by gl_LIST_MOUNTED_FILE_SYSTEMS. +AC_DEFUN([gl_PREREQ_MOUNTLIST_EXTRA], +[ + dnl Note gl_LIST_MOUNTED_FILE_SYSTEMS checks for mntent.h, not sys/mntent.h. + AC_CHECK_HEADERS(sys/mntent.h) + gl_FSTYPENAME +]) diff --git a/gl/m4/netinet_in_h.m4 b/gl/m4/netinet_in_h.m4 new file mode 100644 index 00000000..d73531a0 --- /dev/null +++ b/gl/m4/netinet_in_h.m4 @@ -0,0 +1,18 @@ +# netinet_in_h.m4 serial 1 +dnl Copyright (C) 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl Written by Simon Josefsson + +AC_DEFUN([gl_HEADER_NETINET_IN], +[ + AC_CHECK_HEADERS_ONCE([netinet/in.h]) + if test $ac_cv_header_netinet_in_h = yes; then + NETINET_IN_H='' + else + NETINET_IN_H='netinet/in.h' + fi + AC_SUBST(NETINET_IN_H) +]) diff --git a/gl/m4/nls.m4 b/gl/m4/nls.m4 new file mode 100644 index 00000000..7967cc2f --- /dev/null +++ b/gl/m4/nls.m4 @@ -0,0 +1,31 @@ +# nls.m4 serial 3 (gettext-0.15) +dnl Copyright (C) 1995-2003, 2005-2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. +dnl +dnl This file can can be used in projects which are not available under +dnl the GNU General Public License or the GNU Library General Public +dnl License but which still want to provide support for the GNU gettext +dnl functionality. +dnl Please note that the actual code of the GNU gettext library is covered +dnl by the GNU Library General Public License, and the rest of the GNU +dnl gettext package package is covered by the GNU General Public License. +dnl They are *not* in the public domain. + +dnl Authors: +dnl Ulrich Drepper <drepper@cygnus.com>, 1995-2000. +dnl Bruno Haible <haible@clisp.cons.org>, 2000-2003. + +AC_PREREQ(2.50) + +AC_DEFUN([AM_NLS], +[ + AC_MSG_CHECKING([whether NLS is requested]) + dnl Default is enabled NLS + AC_ARG_ENABLE(nls, + [ --disable-nls do not use Native Language Support], + USE_NLS=$enableval, USE_NLS=yes) + AC_MSG_RESULT($USE_NLS) + AC_SUBST(USE_NLS) +]) diff --git a/gl/m4/onceonly_2_57.m4 b/gl/m4/onceonly_2_57.m4 new file mode 100644 index 00000000..15884b3e --- /dev/null +++ b/gl/m4/onceonly_2_57.m4 @@ -0,0 +1,86 @@ +# onceonly_2_57.m4 serial 4 +dnl Copyright (C) 2002-2003, 2005-2006 Free Software Foundation, Inc. +dnl This file is free software, distributed under the terms of the GNU +dnl General Public License. As a special exception to the GNU General +dnl Public License, this file may be distributed as part of a program +dnl that contains a configuration script generated by Autoconf, under +dnl the same distribution terms as the rest of that program. + +dnl This file defines some "once only" variants of standard autoconf macros. +dnl AC_CHECK_HEADERS_ONCE like AC_CHECK_HEADERS +dnl AC_CHECK_FUNCS_ONCE like AC_CHECK_FUNCS +dnl AC_CHECK_DECLS_ONCE like AC_CHECK_DECLS +dnl AC_REQUIRE([AC_FUNC_STRCOLL]) like AC_FUNC_STRCOLL +dnl The advantage is that the check for each of the headers/functions/decls +dnl will be put only once into the 'configure' file. It keeps the size of +dnl the 'configure' file down, and avoids redundant output when 'configure' +dnl is run. +dnl The drawback is that the checks cannot be conditionalized. If you write +dnl if some_condition; then gl_CHECK_HEADERS(stdlib.h); fi +dnl inside an AC_DEFUNed function, the gl_CHECK_HEADERS macro call expands to +dnl empty, and the check will be inserted before the body of the AC_DEFUNed +dnl function. + +dnl This is like onceonly.m4, except that it uses diversions to named sections +dnl DEFAULTS and INIT_PREPARE in order to check all requested headers at once, +dnl thus reducing the size of 'configure'. Works with autoconf-2.57. The +dnl size reduction is ca. 9%. + +dnl Autoconf version 2.57 or newer is recommended. +AC_PREREQ(2.57) + +# AC_CHECK_HEADERS_ONCE(HEADER1 HEADER2 ...) is a once-only variant of +# AC_CHECK_HEADERS(HEADER1 HEADER2 ...). +AC_DEFUN([AC_CHECK_HEADERS_ONCE], [ + : + AC_FOREACH([gl_HEADER_NAME], [$1], [ + AC_DEFUN([gl_CHECK_HEADER_]m4_quote(translit(gl_HEADER_NAME, + [./-], [___])), [ + m4_divert_text([INIT_PREPARE], + [gl_header_list="$gl_header_list gl_HEADER_NAME"]) + gl_HEADERS_EXPANSION + AH_TEMPLATE(AS_TR_CPP([HAVE_]m4_defn([gl_HEADER_NAME])), + [Define to 1 if you have the <]m4_defn([gl_HEADER_NAME])[> header file.]) + ]) + AC_REQUIRE([gl_CHECK_HEADER_]m4_quote(translit(gl_HEADER_NAME, + [./-], [___]))) + ]) +]) +m4_define([gl_HEADERS_EXPANSION], [ + m4_divert_text([DEFAULTS], [gl_header_list=]) + AC_CHECK_HEADERS([$gl_header_list]) + m4_define([gl_HEADERS_EXPANSION], []) +]) + +# AC_CHECK_FUNCS_ONCE(FUNC1 FUNC2 ...) is a once-only variant of +# AC_CHECK_FUNCS(FUNC1 FUNC2 ...). +AC_DEFUN([AC_CHECK_FUNCS_ONCE], [ + : + AC_FOREACH([gl_FUNC_NAME], [$1], [ + AC_DEFUN([gl_CHECK_FUNC_]m4_defn([gl_FUNC_NAME]), [ + m4_divert_text([INIT_PREPARE], + [gl_func_list="$gl_func_list gl_FUNC_NAME"]) + gl_FUNCS_EXPANSION + AH_TEMPLATE(AS_TR_CPP([HAVE_]m4_defn([gl_FUNC_NAME])), + [Define to 1 if you have the `]m4_defn([gl_FUNC_NAME])[' function.]) + ]) + AC_REQUIRE([gl_CHECK_FUNC_]m4_defn([gl_FUNC_NAME])) + ]) +]) +m4_define([gl_FUNCS_EXPANSION], [ + m4_divert_text([DEFAULTS], [gl_func_list=]) + AC_CHECK_FUNCS([$gl_func_list]) + m4_define([gl_FUNCS_EXPANSION], []) +]) + +# AC_CHECK_DECLS_ONCE(DECL1 DECL2 ...) is a once-only variant of +# AC_CHECK_DECLS(DECL1, DECL2, ...). +AC_DEFUN([AC_CHECK_DECLS_ONCE], [ + : + AC_FOREACH([gl_DECL_NAME], [$1], [ + AC_DEFUN([gl_CHECK_DECL_]m4_defn([gl_DECL_NAME]), [ + AC_CHECK_DECLS(m4_defn([gl_DECL_NAME])) + ]) + AC_REQUIRE([gl_CHECK_DECL_]m4_defn([gl_DECL_NAME])) + ]) +]) diff --git a/gl/m4/po.m4 b/gl/m4/po.m4 new file mode 100644 index 00000000..00133ef3 --- /dev/null +++ b/gl/m4/po.m4 @@ -0,0 +1,428 @@ +# po.m4 serial 13 (gettext-0.15) +dnl Copyright (C) 1995-2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. +dnl +dnl This file can can be used in projects which are not available under +dnl the GNU General Public License or the GNU Library General Public +dnl License but which still want to provide support for the GNU gettext +dnl functionality. +dnl Please note that the actual code of the GNU gettext library is covered +dnl by the GNU Library General Public License, and the rest of the GNU +dnl gettext package package is covered by the GNU General Public License. +dnl They are *not* in the public domain. + +dnl Authors: +dnl Ulrich Drepper <drepper@cygnus.com>, 1995-2000. +dnl Bruno Haible <haible@clisp.cons.org>, 2000-2003. + +AC_PREREQ(2.50) + +dnl Checks for all prerequisites of the po subdirectory. +AC_DEFUN([AM_PO_SUBDIRS], +[ + AC_REQUIRE([AC_PROG_MAKE_SET])dnl + AC_REQUIRE([AC_PROG_INSTALL])dnl + AC_REQUIRE([AM_PROG_MKDIR_P])dnl defined by automake + AC_REQUIRE([AM_NLS])dnl + + dnl Perform the following tests also if --disable-nls has been given, + dnl because they are needed for "make dist" to work. + + dnl Search for GNU msgfmt in the PATH. + dnl The first test excludes Solaris msgfmt and early GNU msgfmt versions. + dnl The second test excludes FreeBSD msgfmt. + AM_PATH_PROG_WITH_TEST(MSGFMT, msgfmt, + [$ac_dir/$ac_word --statistics /dev/null >&]AS_MESSAGE_LOG_FD[ 2>&1 && + (if $ac_dir/$ac_word --statistics /dev/null 2>&1 >/dev/null | grep usage >/dev/null; then exit 1; else exit 0; fi)], + :) + AC_PATH_PROG(GMSGFMT, gmsgfmt, $MSGFMT) + + dnl Test whether it is GNU msgfmt >= 0.15. +changequote(,)dnl + case `$MSGFMT --version | sed 1q | sed -e 's,^[^0-9]*,,'` in + '' | 0.[0-9] | 0.[0-9].* | 0.1[0-4] | 0.1[0-4].*) MSGFMT_015=: ;; + *) MSGFMT_015=$MSGFMT ;; + esac +changequote([,])dnl + AC_SUBST([MSGFMT_015]) +changequote(,)dnl + case `$GMSGFMT --version | sed 1q | sed -e 's,^[^0-9]*,,'` in + '' | 0.[0-9] | 0.[0-9].* | 0.1[0-4] | 0.1[0-4].*) GMSGFMT_015=: ;; + *) GMSGFMT_015=$GMSGFMT ;; + esac +changequote([,])dnl + AC_SUBST([GMSGFMT_015]) + + dnl Search for GNU xgettext 0.12 or newer in the PATH. + dnl The first test excludes Solaris xgettext and early GNU xgettext versions. + dnl The second test excludes FreeBSD xgettext. + AM_PATH_PROG_WITH_TEST(XGETTEXT, xgettext, + [$ac_dir/$ac_word --omit-header --copyright-holder= --msgid-bugs-address= /dev/null >&]AS_MESSAGE_LOG_FD[ 2>&1 && + (if $ac_dir/$ac_word --omit-header --copyright-holder= --msgid-bugs-address= /dev/null 2>&1 >/dev/null | grep usage >/dev/null; then exit 1; else exit 0; fi)], + :) + dnl Remove leftover from FreeBSD xgettext call. + rm -f messages.po + + dnl Test whether it is GNU xgettext >= 0.15. +changequote(,)dnl + case `$XGETTEXT --version | sed 1q | sed -e 's,^[^0-9]*,,'` in + '' | 0.[0-9] | 0.[0-9].* | 0.1[0-4] | 0.1[0-4].*) XGETTEXT_015=: ;; + *) XGETTEXT_015=$XGETTEXT ;; + esac +changequote([,])dnl + AC_SUBST([XGETTEXT_015]) + + dnl Search for GNU msgmerge 0.11 or newer in the PATH. + AM_PATH_PROG_WITH_TEST(MSGMERGE, msgmerge, + [$ac_dir/$ac_word --update -q /dev/null /dev/null >&]AS_MESSAGE_LOG_FD[ 2>&1], :) + + dnl Installation directories. + dnl Autoconf >= 2.60 defines localedir. For older versions of autoconf, we + dnl have to define it here, so that it can be used in po/Makefile. + test -n "$localedir" || localedir='${datadir}/locale' + AC_SUBST([localedir]) + + AC_CONFIG_COMMANDS([po-directories], [[ + for ac_file in $CONFIG_FILES; do + # Support "outfile[:infile[:infile...]]" + case "$ac_file" in + *:*) ac_file=`echo "$ac_file"|sed 's%:.*%%'` ;; + esac + # PO directories have a Makefile.in generated from Makefile.in.in. + case "$ac_file" in */Makefile.in) + # Adjust a relative srcdir. + ac_dir=`echo "$ac_file"|sed 's%/[^/][^/]*$%%'` + ac_dir_suffix="/`echo "$ac_dir"|sed 's%^\./%%'`" + ac_dots=`echo "$ac_dir_suffix"|sed 's%/[^/]*%../%g'` + # In autoconf-2.13 it is called $ac_given_srcdir. + # In autoconf-2.50 it is called $srcdir. + test -n "$ac_given_srcdir" || ac_given_srcdir="$srcdir" + case "$ac_given_srcdir" in + .) top_srcdir=`echo $ac_dots|sed 's%/$%%'` ;; + /*) top_srcdir="$ac_given_srcdir" ;; + *) top_srcdir="$ac_dots$ac_given_srcdir" ;; + esac + # Treat a directory as a PO directory if and only if it has a + # POTFILES.in file. This allows packages to have multiple PO + # directories under different names or in different locations. + if test -f "$ac_given_srcdir/$ac_dir/POTFILES.in"; then + rm -f "$ac_dir/POTFILES" + test -n "$as_me" && echo "$as_me: creating $ac_dir/POTFILES" || echo "creating $ac_dir/POTFILES" + cat "$ac_given_srcdir/$ac_dir/POTFILES.in" | sed -e "/^#/d" -e "/^[ ]*\$/d" -e "s,.*, $top_srcdir/& \\\\," | sed -e "\$s/\(.*\) \\\\/\1/" > "$ac_dir/POTFILES" + POMAKEFILEDEPS="POTFILES.in" + # ALL_LINGUAS, POFILES, UPDATEPOFILES, DUMMYPOFILES, GMOFILES depend + # on $ac_dir but don't depend on user-specified configuration + # parameters. + if test -f "$ac_given_srcdir/$ac_dir/LINGUAS"; then + # The LINGUAS file contains the set of available languages. + if test -n "$OBSOLETE_ALL_LINGUAS"; then + test -n "$as_me" && echo "$as_me: setting ALL_LINGUAS in configure.in is obsolete" || echo "setting ALL_LINGUAS in configure.in is obsolete" + fi + ALL_LINGUAS_=`sed -e "/^#/d" -e "s/#.*//" "$ac_given_srcdir/$ac_dir/LINGUAS"` + # Hide the ALL_LINGUAS assigment from automake < 1.5. + eval 'ALL_LINGUAS''=$ALL_LINGUAS_' + POMAKEFILEDEPS="$POMAKEFILEDEPS LINGUAS" + else + # The set of available languages was given in configure.in. + # Hide the ALL_LINGUAS assigment from automake < 1.5. + eval 'ALL_LINGUAS''=$OBSOLETE_ALL_LINGUAS' + fi + # Compute POFILES + # as $(foreach lang, $(ALL_LINGUAS), $(srcdir)/$(lang).po) + # Compute UPDATEPOFILES + # as $(foreach lang, $(ALL_LINGUAS), $(lang).po-update) + # Compute DUMMYPOFILES + # as $(foreach lang, $(ALL_LINGUAS), $(lang).nop) + # Compute GMOFILES + # as $(foreach lang, $(ALL_LINGUAS), $(srcdir)/$(lang).gmo) + case "$ac_given_srcdir" in + .) srcdirpre= ;; + *) srcdirpre='$(srcdir)/' ;; + esac + POFILES= + UPDATEPOFILES= + DUMMYPOFILES= + GMOFILES= + for lang in $ALL_LINGUAS; do + POFILES="$POFILES $srcdirpre$lang.po" + UPDATEPOFILES="$UPDATEPOFILES $lang.po-update" + DUMMYPOFILES="$DUMMYPOFILES $lang.nop" + GMOFILES="$GMOFILES $srcdirpre$lang.gmo" + done + # CATALOGS depends on both $ac_dir and the user's LINGUAS + # environment variable. + INST_LINGUAS= + if test -n "$ALL_LINGUAS"; then + for presentlang in $ALL_LINGUAS; do + useit=no + if test "%UNSET%" != "$LINGUAS"; then + desiredlanguages="$LINGUAS" + else + desiredlanguages="$ALL_LINGUAS" + fi + for desiredlang in $desiredlanguages; do + # Use the presentlang catalog if desiredlang is + # a. equal to presentlang, or + # b. a variant of presentlang (because in this case, + # presentlang can be used as a fallback for messages + # which are not translated in the desiredlang catalog). + case "$desiredlang" in + "$presentlang"*) useit=yes;; + esac + done + if test $useit = yes; then + INST_LINGUAS="$INST_LINGUAS $presentlang" + fi + done + fi + CATALOGS= + if test -n "$INST_LINGUAS"; then + for lang in $INST_LINGUAS; do + CATALOGS="$CATALOGS $lang.gmo" + done + fi + test -n "$as_me" && echo "$as_me: creating $ac_dir/Makefile" || echo "creating $ac_dir/Makefile" + sed -e "/^POTFILES =/r $ac_dir/POTFILES" -e "/^# Makevars/r $ac_given_srcdir/$ac_dir/Makevars" -e "s|@POFILES@|$POFILES|g" -e "s|@UPDATEPOFILES@|$UPDATEPOFILES|g" -e "s|@DUMMYPOFILES@|$DUMMYPOFILES|g" -e "s|@GMOFILES@|$GMOFILES|g" -e "s|@CATALOGS@|$CATALOGS|g" -e "s|@POMAKEFILEDEPS@|$POMAKEFILEDEPS|g" "$ac_dir/Makefile.in" > "$ac_dir/Makefile" + for f in "$ac_given_srcdir/$ac_dir"/Rules-*; do + if test -f "$f"; then + case "$f" in + *.orig | *.bak | *~) ;; + *) cat "$f" >> "$ac_dir/Makefile" ;; + esac + fi + done + fi + ;; + esac + done]], + [# Capture the value of obsolete ALL_LINGUAS because we need it to compute + # POFILES, UPDATEPOFILES, DUMMYPOFILES, GMOFILES, CATALOGS. But hide it + # from automake < 1.5. + eval 'OBSOLETE_ALL_LINGUAS''="$ALL_LINGUAS"' + # Capture the value of LINGUAS because we need it to compute CATALOGS. + LINGUAS="${LINGUAS-%UNSET%}" + ]) +]) + +dnl Postprocesses a Makefile in a directory containing PO files. +AC_DEFUN([AM_POSTPROCESS_PO_MAKEFILE], +[ + # When this code is run, in config.status, two variables have already been + # set: + # - OBSOLETE_ALL_LINGUAS is the value of LINGUAS set in configure.in, + # - LINGUAS is the value of the environment variable LINGUAS at configure + # time. + +changequote(,)dnl + # Adjust a relative srcdir. + ac_dir=`echo "$ac_file"|sed 's%/[^/][^/]*$%%'` + ac_dir_suffix="/`echo "$ac_dir"|sed 's%^\./%%'`" + ac_dots=`echo "$ac_dir_suffix"|sed 's%/[^/]*%../%g'` + # In autoconf-2.13 it is called $ac_given_srcdir. + # In autoconf-2.50 it is called $srcdir. + test -n "$ac_given_srcdir" || ac_given_srcdir="$srcdir" + case "$ac_given_srcdir" in + .) top_srcdir=`echo $ac_dots|sed 's%/$%%'` ;; + /*) top_srcdir="$ac_given_srcdir" ;; + *) top_srcdir="$ac_dots$ac_given_srcdir" ;; + esac + + # Find a way to echo strings without interpreting backslash. + if test "X`(echo '\t') 2>/dev/null`" = 'X\t'; then + gt_echo='echo' + else + if test "X`(printf '%s\n' '\t') 2>/dev/null`" = 'X\t'; then + gt_echo='printf %s\n' + else + echo_func () { + cat <<EOT +$* +EOT + } + gt_echo='echo_func' + fi + fi + + # A sed script that extracts the value of VARIABLE from a Makefile. + sed_x_variable=' +# Test if the hold space is empty. +x +s/P/P/ +x +ta +# Yes it was empty. Look if we have the expected variable definition. +/^[ ]*VARIABLE[ ]*=/{ + # Seen the first line of the variable definition. + s/^[ ]*VARIABLE[ ]*=// + ba +} +bd +:a +# Here we are processing a line from the variable definition. +# Remove comment, more precisely replace it with a space. +s/#.*$/ / +# See if the line ends in a backslash. +tb +:b +s/\\$// +# Print the line, without the trailing backslash. +p +tc +# There was no trailing backslash. The end of the variable definition is +# reached. Clear the hold space. +s/^.*$// +x +bd +:c +# A trailing backslash means that the variable definition continues in the +# next line. Put a nonempty string into the hold space to indicate this. +s/^.*$/P/ +x +:d +' +changequote([,])dnl + + # Set POTFILES to the value of the Makefile variable POTFILES. + sed_x_POTFILES=`$gt_echo "$sed_x_variable" | sed -e '/^ *#/d' -e 's/VARIABLE/POTFILES/g'` + POTFILES=`sed -n -e "$sed_x_POTFILES" < "$ac_file"` + # Compute POTFILES_DEPS as + # $(foreach file, $(POTFILES), $(top_srcdir)/$(file)) + POTFILES_DEPS= + for file in $POTFILES; do + POTFILES_DEPS="$POTFILES_DEPS "'$(top_srcdir)/'"$file" + done + POMAKEFILEDEPS="" + + if test -n "$OBSOLETE_ALL_LINGUAS"; then + test -n "$as_me" && echo "$as_me: setting ALL_LINGUAS in configure.in is obsolete" || echo "setting ALL_LINGUAS in configure.in is obsolete" + fi + if test -f "$ac_given_srcdir/$ac_dir/LINGUAS"; then + # The LINGUAS file contains the set of available languages. + ALL_LINGUAS_=`sed -e "/^#/d" -e "s/#.*//" "$ac_given_srcdir/$ac_dir/LINGUAS"` + POMAKEFILEDEPS="$POMAKEFILEDEPS LINGUAS" + else + # Set ALL_LINGUAS to the value of the Makefile variable LINGUAS. + sed_x_LINGUAS=`$gt_echo "$sed_x_variable" | sed -e '/^ *#/d' -e 's/VARIABLE/LINGUAS/g'` + ALL_LINGUAS_=`sed -n -e "$sed_x_LINGUAS" < "$ac_file"` + fi + # Hide the ALL_LINGUAS assigment from automake < 1.5. + eval 'ALL_LINGUAS''=$ALL_LINGUAS_' + # Compute POFILES + # as $(foreach lang, $(ALL_LINGUAS), $(srcdir)/$(lang).po) + # Compute UPDATEPOFILES + # as $(foreach lang, $(ALL_LINGUAS), $(lang).po-update) + # Compute DUMMYPOFILES + # as $(foreach lang, $(ALL_LINGUAS), $(lang).nop) + # Compute GMOFILES + # as $(foreach lang, $(ALL_LINGUAS), $(srcdir)/$(lang).gmo) + # Compute PROPERTIESFILES + # as $(foreach lang, $(ALL_LINGUAS), $(top_srcdir)/$(DOMAIN)_$(lang).properties) + # Compute CLASSFILES + # as $(foreach lang, $(ALL_LINGUAS), $(top_srcdir)/$(DOMAIN)_$(lang).class) + # Compute QMFILES + # as $(foreach lang, $(ALL_LINGUAS), $(srcdir)/$(lang).qm) + # Compute MSGFILES + # as $(foreach lang, $(ALL_LINGUAS), $(srcdir)/$(frob $(lang)).msg) + # Compute RESOURCESDLLFILES + # as $(foreach lang, $(ALL_LINGUAS), $(srcdir)/$(frob $(lang))/$(DOMAIN).resources.dll) + case "$ac_given_srcdir" in + .) srcdirpre= ;; + *) srcdirpre='$(srcdir)/' ;; + esac + POFILES= + UPDATEPOFILES= + DUMMYPOFILES= + GMOFILES= + PROPERTIESFILES= + CLASSFILES= + QMFILES= + MSGFILES= + RESOURCESDLLFILES= + for lang in $ALL_LINGUAS; do + POFILES="$POFILES $srcdirpre$lang.po" + UPDATEPOFILES="$UPDATEPOFILES $lang.po-update" + DUMMYPOFILES="$DUMMYPOFILES $lang.nop" + GMOFILES="$GMOFILES $srcdirpre$lang.gmo" + PROPERTIESFILES="$PROPERTIESFILES \$(top_srcdir)/\$(DOMAIN)_$lang.properties" + CLASSFILES="$CLASSFILES \$(top_srcdir)/\$(DOMAIN)_$lang.class" + QMFILES="$QMFILES $srcdirpre$lang.qm" + frobbedlang=`echo $lang | sed -e 's/\..*$//' -e 'y/ABCDEFGHIJKLMNOPQRSTUVWXYZ/abcdefghijklmnopqrstuvwxyz/'` + MSGFILES="$MSGFILES $srcdirpre$frobbedlang.msg" + frobbedlang=`echo $lang | sed -e 's/_/-/g' -e 's/^sr-CS/sr-SP/' -e 's/@latin$/-Latn/' -e 's/@cyrillic$/-Cyrl/' -e 's/^sr-SP$/sr-SP-Latn/' -e 's/^uz-UZ$/uz-UZ-Latn/'` + RESOURCESDLLFILES="$RESOURCESDLLFILES $srcdirpre$frobbedlang/\$(DOMAIN).resources.dll" + done + # CATALOGS depends on both $ac_dir and the user's LINGUAS + # environment variable. + INST_LINGUAS= + if test -n "$ALL_LINGUAS"; then + for presentlang in $ALL_LINGUAS; do + useit=no + if test "%UNSET%" != "$LINGUAS"; then + desiredlanguages="$LINGUAS" + else + desiredlanguages="$ALL_LINGUAS" + fi + for desiredlang in $desiredlanguages; do + # Use the presentlang catalog if desiredlang is + # a. equal to presentlang, or + # b. a variant of presentlang (because in this case, + # presentlang can be used as a fallback for messages + # which are not translated in the desiredlang catalog). + case "$desiredlang" in + "$presentlang"*) useit=yes;; + esac + done + if test $useit = yes; then + INST_LINGUAS="$INST_LINGUAS $presentlang" + fi + done + fi + CATALOGS= + JAVACATALOGS= + QTCATALOGS= + TCLCATALOGS= + CSHARPCATALOGS= + if test -n "$INST_LINGUAS"; then + for lang in $INST_LINGUAS; do + CATALOGS="$CATALOGS $lang.gmo" + JAVACATALOGS="$JAVACATALOGS \$(DOMAIN)_$lang.properties" + QTCATALOGS="$QTCATALOGS $lang.qm" + frobbedlang=`echo $lang | sed -e 's/\..*$//' -e 'y/ABCDEFGHIJKLMNOPQRSTUVWXYZ/abcdefghijklmnopqrstuvwxyz/'` + TCLCATALOGS="$TCLCATALOGS $frobbedlang.msg" + frobbedlang=`echo $lang | sed -e 's/_/-/g' -e 's/^sr-CS/sr-SP/' -e 's/@latin$/-Latn/' -e 's/@cyrillic$/-Cyrl/' -e 's/^sr-SP$/sr-SP-Latn/' -e 's/^uz-UZ$/uz-UZ-Latn/'` + CSHARPCATALOGS="$CSHARPCATALOGS $frobbedlang/\$(DOMAIN).resources.dll" + done + fi + + sed -e "s|@POTFILES_DEPS@|$POTFILES_DEPS|g" -e "s|@POFILES@|$POFILES|g" -e "s|@UPDATEPOFILES@|$UPDATEPOFILES|g" -e "s|@DUMMYPOFILES@|$DUMMYPOFILES|g" -e "s|@GMOFILES@|$GMOFILES|g" -e "s|@PROPERTIESFILES@|$PROPERTIESFILES|g" -e "s|@CLASSFILES@|$CLASSFILES|g" -e "s|@QMFILES@|$QMFILES|g" -e "s|@MSGFILES@|$MSGFILES|g" -e "s|@RESOURCESDLLFILES@|$RESOURCESDLLFILES|g" -e "s|@CATALOGS@|$CATALOGS|g" -e "s|@JAVACATALOGS@|$JAVACATALOGS|g" -e "s|@QTCATALOGS@|$QTCATALOGS|g" -e "s|@TCLCATALOGS@|$TCLCATALOGS|g" -e "s|@CSHARPCATALOGS@|$CSHARPCATALOGS|g" -e 's,^#distdir:,distdir:,' < "$ac_file" > "$ac_file.tmp" + if grep -l '@TCLCATALOGS@' "$ac_file" > /dev/null; then + # Add dependencies that cannot be formulated as a simple suffix rule. + for lang in $ALL_LINGUAS; do + frobbedlang=`echo $lang | sed -e 's/\..*$//' -e 'y/ABCDEFGHIJKLMNOPQRSTUVWXYZ/abcdefghijklmnopqrstuvwxyz/'` + cat >> "$ac_file.tmp" <<EOF +$frobbedlang.msg: $lang.po + @echo "\$(MSGFMT) -c --tcl -d \$(srcdir) -l $lang $srcdirpre$lang.po"; \ + \$(MSGFMT) -c --tcl -d "\$(srcdir)" -l $lang $srcdirpre$lang.po || { rm -f "\$(srcdir)/$frobbedlang.msg"; exit 1; } +EOF + done + fi + if grep -l '@CSHARPCATALOGS@' "$ac_file" > /dev/null; then + # Add dependencies that cannot be formulated as a simple suffix rule. + for lang in $ALL_LINGUAS; do + frobbedlang=`echo $lang | sed -e 's/_/-/g' -e 's/^sr-CS/sr-SP/' -e 's/@latin$/-Latn/' -e 's/@cyrillic$/-Cyrl/' -e 's/^sr-SP$/sr-SP-Latn/' -e 's/^uz-UZ$/uz-UZ-Latn/'` + cat >> "$ac_file.tmp" <<EOF +$frobbedlang/\$(DOMAIN).resources.dll: $lang.po + @echo "\$(MSGFMT) -c --csharp -d \$(srcdir) -l $lang $srcdirpre$lang.po -r \$(DOMAIN)"; \ + \$(MSGFMT) -c --csharp -d "\$(srcdir)" -l $lang $srcdirpre$lang.po -r "\$(DOMAIN)" || { rm -f "\$(srcdir)/$frobbedlang.msg"; exit 1; } +EOF + done + fi + if test -n "$POMAKEFILEDEPS"; then + cat >> "$ac_file.tmp" <<EOF +Makefile: $POMAKEFILEDEPS +EOF + fi + mv "$ac_file.tmp" "$ac_file" +]) diff --git a/gl/m4/printf-posix.m4 b/gl/m4/printf-posix.m4 new file mode 100644 index 00000000..af10170a --- /dev/null +++ b/gl/m4/printf-posix.m4 @@ -0,0 +1,44 @@ +# printf-posix.m4 serial 2 (gettext-0.13.1) +dnl Copyright (C) 2003 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Bruno Haible. +dnl Test whether the printf() function supports POSIX/XSI format strings with +dnl positions. + +AC_DEFUN([gt_PRINTF_POSIX], +[ + AC_REQUIRE([AC_PROG_CC]) + AC_CACHE_CHECK([whether printf() supports POSIX/XSI format strings], + gt_cv_func_printf_posix, + [ + AC_TRY_RUN([ +#include <stdio.h> +#include <string.h> +/* The string "%2$d %1$d", with dollar characters protected from the shell's + dollar expansion (possibly an autoconf bug). */ +static char format[] = { '%', '2', '$', 'd', ' ', '%', '1', '$', 'd', '\0' }; +static char buf[100]; +int main () +{ + sprintf (buf, format, 33, 55); + return (strcmp (buf, "55 33") != 0); +}], gt_cv_func_printf_posix=yes, gt_cv_func_printf_posix=no, + [ + AC_EGREP_CPP(notposix, [ +#if defined __NetBSD__ || defined _MSC_VER || defined __MINGW32__ || defined __CYGWIN__ + notposix +#endif + ], gt_cv_func_printf_posix="guessing no", + gt_cv_func_printf_posix="guessing yes") + ]) + ]) + case $gt_cv_func_printf_posix in + *yes) + AC_DEFINE(HAVE_POSIX_PRINTF, 1, + [Define if your printf() function supports format strings with positions.]) + ;; + esac +]) diff --git a/gl/m4/progtest.m4 b/gl/m4/progtest.m4 new file mode 100644 index 00000000..a56365cd --- /dev/null +++ b/gl/m4/progtest.m4 @@ -0,0 +1,92 @@ +# progtest.m4 serial 4 (gettext-0.14.2) +dnl Copyright (C) 1996-2003, 2005 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. +dnl +dnl This file can can be used in projects which are not available under +dnl the GNU General Public License or the GNU Library General Public +dnl License but which still want to provide support for the GNU gettext +dnl functionality. +dnl Please note that the actual code of the GNU gettext library is covered +dnl by the GNU Library General Public License, and the rest of the GNU +dnl gettext package package is covered by the GNU General Public License. +dnl They are *not* in the public domain. + +dnl Authors: +dnl Ulrich Drepper <drepper@cygnus.com>, 1996. + +AC_PREREQ(2.50) + +# Search path for a program which passes the given test. + +dnl AM_PATH_PROG_WITH_TEST(VARIABLE, PROG-TO-CHECK-FOR, +dnl TEST-PERFORMED-ON-FOUND_PROGRAM [, VALUE-IF-NOT-FOUND [, PATH]]) +AC_DEFUN([AM_PATH_PROG_WITH_TEST], +[ +# Prepare PATH_SEPARATOR. +# The user is always right. +if test "${PATH_SEPARATOR+set}" != set; then + echo "#! /bin/sh" >conf$$.sh + echo "exit 0" >>conf$$.sh + chmod +x conf$$.sh + if (PATH="/nonexistent;."; conf$$.sh) >/dev/null 2>&1; then + PATH_SEPARATOR=';' + else + PATH_SEPARATOR=: + fi + rm -f conf$$.sh +fi + +# Find out how to test for executable files. Don't use a zero-byte file, +# as systems may use methods other than mode bits to determine executability. +cat >conf$$.file <<_ASEOF +#! /bin/sh +exit 0 +_ASEOF +chmod +x conf$$.file +if test -x conf$$.file >/dev/null 2>&1; then + ac_executable_p="test -x" +else + ac_executable_p="test -f" +fi +rm -f conf$$.file + +# Extract the first word of "$2", so it can be a program name with args. +set dummy $2; ac_word=[$]2 +AC_MSG_CHECKING([for $ac_word]) +AC_CACHE_VAL(ac_cv_path_$1, +[case "[$]$1" in + [[\\/]]* | ?:[[\\/]]*) + ac_cv_path_$1="[$]$1" # Let the user override the test with a path. + ;; + *) + ac_save_IFS="$IFS"; IFS=$PATH_SEPARATOR + for ac_dir in ifelse([$5], , $PATH, [$5]); do + IFS="$ac_save_IFS" + test -z "$ac_dir" && ac_dir=. + for ac_exec_ext in '' $ac_executable_extensions; do + if $ac_executable_p "$ac_dir/$ac_word$ac_exec_ext"; then + echo "$as_me: trying $ac_dir/$ac_word..." >&AS_MESSAGE_LOG_FD + if [$3]; then + ac_cv_path_$1="$ac_dir/$ac_word$ac_exec_ext" + break 2 + fi + fi + done + done + IFS="$ac_save_IFS" +dnl If no 4th arg is given, leave the cache variable unset, +dnl so AC_PATH_PROGS will keep looking. +ifelse([$4], , , [ test -z "[$]ac_cv_path_$1" && ac_cv_path_$1="$4" +])dnl + ;; +esac])dnl +$1="$ac_cv_path_$1" +if test ifelse([$4], , [-n "[$]$1"], ["[$]$1" != "$4"]); then + AC_MSG_RESULT([$]$1) +else + AC_MSG_RESULT(no) +fi +AC_SUBST($1)dnl +]) diff --git a/gl/m4/regex.m4 b/gl/m4/regex.m4 new file mode 100644 index 00000000..25da645e --- /dev/null +++ b/gl/m4/regex.m4 @@ -0,0 +1,198 @@ +#serial 42 + +# Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2003, 2004, 2005, +# 2006, 2007 Free Software Foundation, Inc. +# +# This file is free software; the Free Software Foundation +# gives unlimited permission to copy and/or distribute it, +# with or without modifications, as long as this notice is preserved. + +dnl Initially derived from code in GNU grep. +dnl Mostly written by Jim Meyering. + +AC_PREREQ([2.50]) + +AC_DEFUN([gl_REGEX], +[ + AC_CHECK_HEADERS_ONCE([locale.h]) + + AC_ARG_WITH([included-regex], + [AC_HELP_STRING([--without-included-regex], + [don't compile regex; this is the default on + systems with recent-enough versions of the GNU C + Library (use with caution on other systems)])]) + + case $with_included_regex in #( + yes|no) ac_use_included_regex=$with_included_regex + ;; + '') + # If the system regex support is good enough that it passes the + # following run test, then default to *not* using the included regex.c. + # If cross compiling, assume the test would fail and use the included + # regex.c. The first failing regular expression is from `Spencer ere + # test #75' in grep-2.3. + AC_CACHE_CHECK([for working re_compile_pattern], + [gl_cv_func_re_compile_pattern_working], + [AC_RUN_IFELSE( + [AC_LANG_PROGRAM( + [AC_INCLUDES_DEFAULT + #if HAVE_LOCALE_H + #include <locale.h> + #endif + #include <limits.h> + #include <regex.h> + ], + [[static struct re_pattern_buffer regex; + unsigned char folded_chars[UCHAR_MAX + 1]; + int i; + const char *s; + struct re_registers regs; + + #if HAVE_LOCALE_H + /* http://sourceware.org/ml/libc-hacker/2006-09/msg00008.html + This test needs valgrind to catch the bug on Debian + GNU/Linux 3.1 x86, but it might catch the bug better + on other platforms and it shouldn't hurt to try the + test here. */ + if (setlocale (LC_ALL, "en_US.UTF-8")) + { + static char const pat[] = "insert into"; + static char const data[] = + "\xFF\0\x12\xA2\xAA\xC4\xB1,K\x12\xC4\xB1*\xACK"; + re_set_syntax (RE_SYNTAX_GREP | RE_HAT_LISTS_NOT_NEWLINE + | RE_ICASE); + memset (®ex, 0, sizeof regex); + s = re_compile_pattern (pat, sizeof pat - 1, ®ex); + if (s) + return 1; + if (re_search (®ex, data, sizeof data - 1, + 0, sizeof data - 1, ®s) + != -1) + return 1; + if (! setlocale (LC_ALL, "C")) + return 1; + } + #endif + + re_set_syntax (RE_SYNTAX_POSIX_EGREP); + memset (®ex, 0, sizeof (regex)); + for (i = 0; i <= UCHAR_MAX; i++) + folded_chars[i] = i; + regex.translate = folded_chars; + s = re_compile_pattern ("a[[:@:>@:]]b\n", 11, ®ex); + /* This should fail with _Invalid character class name_ error. */ + if (!s) + exit (1); + + /* This should succeed, but does not for e.g. glibc-2.1.3. */ + memset (®ex, 0, sizeof (regex)); + s = re_compile_pattern ("{1", 2, ®ex); + + if (s) + exit (1); + + /* The following example is derived from a problem report + against gawk from Jorge Stolfi <stolfi@ic.unicamp.br>. */ + memset (®ex, 0, sizeof (regex)); + s = re_compile_pattern ("[an\371]*n", 7, ®ex); + if (s) + exit (1); + + /* This should match, but does not for e.g. glibc-2.2.1. */ + if (re_match (®ex, "an", 2, 0, ®s) != 2) + exit (1); + + memset (®ex, 0, sizeof (regex)); + s = re_compile_pattern ("x", 1, ®ex); + if (s) + exit (1); + + /* The version of regex.c in e.g. GNU libc-2.2.93 did not + work with a negative RANGE argument. */ + if (re_search (®ex, "wxy", 3, 2, -2, ®s) != 1) + exit (1); + + /* The version of regex.c in older versions of gnulib + ignored RE_ICASE. Detect that problem too. */ + memset (®ex, 0, sizeof (regex)); + re_set_syntax (RE_SYNTAX_EMACS | RE_ICASE); + s = re_compile_pattern ("x", 1, ®ex); + if (s) + exit (1); + + if (re_search (®ex, "WXY", 3, 0, 3, ®s) < 0) + exit (1); + + /* REG_STARTEND was added to glibc on 2004-01-15. + Reject older versions. */ + if (! REG_STARTEND) + exit (1); + + /* Reject hosts whose regoff_t values are too narrow. + These include glibc 2.3.5 on hosts with 64-bit ptrdiff_t + and 32-bit int. */ + if (sizeof (regoff_t) < sizeof (ptrdiff_t) + || sizeof (regoff_t) < sizeof (ssize_t)) + exit (1); + + exit (0);]])], + [gl_cv_func_re_compile_pattern_working=yes], + [gl_cv_func_re_compile_pattern_working=no], + dnl When crosscompiling, assume it is not working. + [gl_cv_func_re_compile_pattern_working=no])]) + case $gl_cv_func_re_compile_pattern_working in #( + yes) ac_use_included_regex=no;; #( + no) ac_use_included_regex=yes;; + esac + ;; + *) AC_MSG_ERROR([Invalid value for --with-included-regex: $with_included_regex]) + ;; + esac + + if test $ac_use_included_regex = yes; then + AC_DEFINE([_REGEX_LARGE_OFFSETS], 1, + [Define if you want regoff_t to be at least as wide POSIX requires.]) + AC_DEFINE([re_syntax_options], [rpl_re_syntax_options], + [Define to rpl_re_syntax_options if the replacement should be used.]) + AC_DEFINE([re_set_syntax], [rpl_re_set_syntax], + [Define to rpl_re_set_syntax if the replacement should be used.]) + AC_DEFINE([re_compile_pattern], [rpl_re_compile_pattern], + [Define to rpl_re_compile_pattern if the replacement should be used.]) + AC_DEFINE([re_compile_fastmap], [rpl_re_compile_fastmap], + [Define to rpl_re_compile_fastmap if the replacement should be used.]) + AC_DEFINE([re_search], [rpl_re_search], + [Define to rpl_re_search if the replacement should be used.]) + AC_DEFINE([re_search_2], [rpl_re_search_2], + [Define to rpl_re_search_2 if the replacement should be used.]) + AC_DEFINE([re_match], [rpl_re_match], + [Define to rpl_re_match if the replacement should be used.]) + AC_DEFINE([re_match_2], [rpl_re_match_2], + [Define to rpl_re_match_2 if the replacement should be used.]) + AC_DEFINE([re_set_registers], [rpl_re_set_registers], + [Define to rpl_re_set_registers if the replacement should be used.]) + AC_DEFINE([re_comp], [rpl_re_comp], + [Define to rpl_re_comp if the replacement should be used.]) + AC_DEFINE([re_exec], [rpl_re_exec], + [Define to rpl_re_exec if the replacement should be used.]) + AC_DEFINE([regcomp], [rpl_regcomp], + [Define to rpl_regcomp if the replacement should be used.]) + AC_DEFINE([regexec], [rpl_regexec], + [Define to rpl_regexec if the replacement should be used.]) + AC_DEFINE([regerror], [rpl_regerror], + [Define to rpl_regerror if the replacement should be used.]) + AC_DEFINE([regfree], [rpl_regfree], + [Define to rpl_regfree if the replacement should be used.]) + AC_LIBOBJ([regex]) + gl_PREREQ_REGEX + fi +]) + +# Prerequisites of lib/regex.c and lib/regex_internal.c. +AC_DEFUN([gl_PREREQ_REGEX], +[ + AC_REQUIRE([AC_GNU_SOURCE]) + AC_REQUIRE([AC_C_RESTRICT]) + AC_REQUIRE([AM_LANGINFO_CODESET]) + AC_CHECK_FUNCS_ONCE([iswctype mbrtowc mempcpy wcrtomb wcscoll]) + AC_CHECK_DECLS([isblank], [], [], [#include <ctype.h>]) +]) diff --git a/gl/m4/safe-read.m4 b/gl/m4/safe-read.m4 new file mode 100644 index 00000000..7a89d0a6 --- /dev/null +++ b/gl/m4/safe-read.m4 @@ -0,0 +1,18 @@ +# safe-read.m4 serial 5 +dnl Copyright (C) 2002, 2003, 2005, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_SAFE_READ], +[ + AC_LIBOBJ([safe-read]) + + gl_PREREQ_SAFE_READ +]) + +# Prerequisites of lib/safe-read.c. +AC_DEFUN([gl_PREREQ_SAFE_READ], +[ + AC_REQUIRE([gt_TYPE_SSIZE_T]) +]) diff --git a/gl/m4/safe-write.m4 b/gl/m4/safe-write.m4 new file mode 100644 index 00000000..db119ffa --- /dev/null +++ b/gl/m4/safe-write.m4 @@ -0,0 +1,18 @@ +# safe-write.m4 serial 3 +dnl Copyright (C) 2002, 2005, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_SAFE_WRITE], +[ + AC_LIBOBJ([safe-write]) + + gl_PREREQ_SAFE_WRITE +]) + +# Prerequisites of lib/safe-write.c. +AC_DEFUN([gl_PREREQ_SAFE_WRITE], +[ + gl_PREREQ_SAFE_READ +]) diff --git a/gl/m4/size_max.m4 b/gl/m4/size_max.m4 new file mode 100644 index 00000000..6cb48689 --- /dev/null +++ b/gl/m4/size_max.m4 @@ -0,0 +1,68 @@ +# size_max.m4 serial 6 +dnl Copyright (C) 2003, 2005-2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Bruno Haible. + +AC_DEFUN([gl_SIZE_MAX], +[ + AC_CHECK_HEADERS(stdint.h) + dnl First test whether the system already has SIZE_MAX. + AC_MSG_CHECKING([for SIZE_MAX]) + AC_CACHE_VAL([gl_cv_size_max], [ + gl_cv_size_max= + AC_EGREP_CPP([Found it], [ +#include <limits.h> +#if HAVE_STDINT_H +#include <stdint.h> +#endif +#ifdef SIZE_MAX +Found it +#endif +], gl_cv_size_max=yes) + if test -z "$gl_cv_size_max"; then + dnl Define it ourselves. Here we assume that the type 'size_t' is not wider + dnl than the type 'unsigned long'. Try hard to find a definition that can + dnl be used in a preprocessor #if, i.e. doesn't contain a cast. + AC_COMPUTE_INT([size_t_bits_minus_1], [sizeof (size_t) * CHAR_BIT - 1], + [#include <stddef.h> +#include <limits.h>], size_t_bits_minus_1=) + AC_COMPUTE_INT([fits_in_uint], [sizeof (size_t) <= sizeof (unsigned int)], + [#include <stddef.h>], fits_in_uint=) + if test -n "$size_t_bits_minus_1" && test -n "$fits_in_uint"; then + if test $fits_in_uint = 1; then + dnl Even though SIZE_MAX fits in an unsigned int, it must be of type + dnl 'unsigned long' if the type 'size_t' is the same as 'unsigned long'. + AC_TRY_COMPILE([#include <stddef.h> + extern size_t foo; + extern unsigned long foo; + ], [], fits_in_uint=0) + fi + dnl We cannot use 'expr' to simplify this expression, because 'expr' + dnl works only with 'long' integers in the host environment, while we + dnl might be cross-compiling from a 32-bit platform to a 64-bit platform. + if test $fits_in_uint = 1; then + gl_cv_size_max="(((1U << $size_t_bits_minus_1) - 1) * 2 + 1)" + else + gl_cv_size_max="(((1UL << $size_t_bits_minus_1) - 1) * 2 + 1)" + fi + else + dnl Shouldn't happen, but who knows... + gl_cv_size_max='((size_t)~(size_t)0)' + fi + fi + ]) + AC_MSG_RESULT([$gl_cv_size_max]) + if test "$gl_cv_size_max" != yes; then + AC_DEFINE_UNQUOTED([SIZE_MAX], [$gl_cv_size_max], + [Define as the maximum value of type 'size_t', if the system doesn't define it.]) + fi +]) + +dnl Autoconf >= 2.61 has AC_COMPUTE_INT built-in. +dnl Remove this when we can assume autoconf >= 2.61. +m4_ifdef([AC_COMPUTE_INT], [], [ + AC_DEFUN([AC_COMPUTE_INT], [_AC_COMPUTE_INT([$2],[$1],[$3],[$4])]) +]) diff --git a/gl/m4/snprintf.m4 b/gl/m4/snprintf.m4 new file mode 100644 index 00000000..432e0375 --- /dev/null +++ b/gl/m4/snprintf.m4 @@ -0,0 +1,15 @@ +# snprintf.m4 serial 2 +dnl Copyright (C) 2002, 2003, 2004 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_FUNC_SNPRINTF], +[ + AC_REPLACE_FUNCS(snprintf) + AC_CHECK_DECLS_ONCE(snprintf) + gl_PREREQ_SNPRINTF +]) + +# Prerequisites of lib/snprintf.c. +AC_DEFUN([gl_PREREQ_SNPRINTF], [:]) diff --git a/gl/m4/socklen.m4 b/gl/m4/socklen.m4 new file mode 100644 index 00000000..5e3765a6 --- /dev/null +++ b/gl/m4/socklen.m4 @@ -0,0 +1,52 @@ +# socklen.m4 serial 4 +dnl Copyright (C) 2005, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Albert Chin, Windows fixes from Simon Josefsson. + +dnl Check for socklen_t: historically on BSD it is an int, and in +dnl POSIX 1g it is a type of its own, but some platforms use different +dnl types for the argument to getsockopt, getpeername, etc. So we +dnl have to test to find something that will work. + +dnl On mingw32, socklen_t is in ws2tcpip.h ('int'), so we try to find +dnl it there first. That file is included by gnulib's socket_.h, which +dnl all users of this module should include. Cygwin must not include +dnl ws2tcpip.h. +AC_DEFUN([gl_TYPE_SOCKLEN_T], + [AC_REQUIRE([gl_HEADER_SYS_SOCKET])dnl + AC_CHECK_TYPE([socklen_t], , + [AC_MSG_CHECKING([for socklen_t equivalent]) + AC_CACHE_VAL([gl_cv_gl_cv_socklen_t_equiv], + [# Systems have either "struct sockaddr *" or + # "void *" as the second argument to getpeername + gl_cv_socklen_t_equiv= + for arg2 in "struct sockaddr" void; do + for t in int size_t "unsigned int" "long int" "unsigned long int"; do + AC_TRY_COMPILE( + [#include <sys/types.h> + #include <sys/socket.h> + + int getpeername (int, $arg2 *, $t *);], + [$t len; + getpeername (0, 0, &len);], + [gl_cv_socklen_t_equiv="$t"]) + test "$gl_cv_socklen_t_equiv" != "" && break + done + test "$gl_cv_socklen_t_equiv" != "" && break + done + ]) + if test "$gl_cv_socklen_t_equiv" = ""; then + AC_MSG_ERROR([Cannot find a type to use in place of socklen_t]) + fi + AC_MSG_RESULT([$gl_cv_socklen_t_equiv]) + AC_DEFINE_UNQUOTED([socklen_t], [$gl_cv_socklen_t_equiv], + [type to use in place of socklen_t if not defined])], + [#include <sys/types.h> + #if HAVE_SYS_SOCKET_H + # include <sys/socket.h> + #elif HAVE_WS2TCPIP_H + # include <ws2tcpip.h> + #endif])]) diff --git a/gl/m4/sockpfaf.m4 b/gl/m4/sockpfaf.m4 new file mode 100644 index 00000000..25d9755c --- /dev/null +++ b/gl/m4/sockpfaf.m4 @@ -0,0 +1,58 @@ +# sockpfaf.m4 serial 5 +dnl Copyright (C) 2004, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl Test for some common socket protocol families (PF_INET, PF_INET6, ...) +dnl and some common address families (AF_INET, AF_INET6, ...). +dnl This test assumes that a system supports an address family if and only if +dnl it supports the corresponding protocol family. + +dnl From Bruno Haible. + +AC_DEFUN([gl_SOCKET_FAMILIES], +[ + AC_REQUIRE([gl_HEADER_SYS_SOCKET]) + AC_CHECK_HEADERS_ONCE([netinet/in.h]) + + AC_MSG_CHECKING(for IPv4 sockets) + AC_CACHE_VAL(gl_cv_socket_ipv4, + [AC_TRY_COMPILE([#include <sys/types.h> +#ifdef HAVE_SYS_SOCKET_H +#include <sys/socket.h> +#endif +#ifdef HAVE_NETINET_IN_H +#include <netinet/in.h> +#endif +#ifdef HAVE_WINSOCK2_H +#include <winsock2.h> +#endif], +[int x = AF_INET; struct in_addr y; struct sockaddr_in z; + if (&x && &y && &z) return 0;], + gl_cv_socket_ipv4=yes, gl_cv_socket_ipv4=no)]) + AC_MSG_RESULT($gl_cv_socket_ipv4) + if test $gl_cv_socket_ipv4 = yes; then + AC_DEFINE(HAVE_IPV4, 1, [Define to 1 if <sys/socket.h> defines AF_INET.]) + fi + + AC_MSG_CHECKING(for IPv6 sockets) + AC_CACHE_VAL(gl_cv_socket_ipv6, + [AC_TRY_COMPILE([#include <sys/types.h> +#ifdef HAVE_SYS_SOCKET_H +#include <sys/socket.h> +#endif +#ifdef HAVE_NETINET_IN_H +#include <netinet/in.h> +#endif +#ifdef HAVE_WINSOCK2_H +#include <winsock2.h> +#endif], +[int x = AF_INET6; struct in6_addr y; struct sockaddr_in6 z; + if (&x && &y && &z) return 0;], + gl_cv_socket_ipv6=yes, gl_cv_socket_ipv6=no)]) + AC_MSG_RESULT($gl_cv_socket_ipv6) + if test $gl_cv_socket_ipv6 = yes; then + AC_DEFINE(HAVE_IPV6, 1, [Define to 1 if <sys/socket.h> defines AF_INET6.]) + fi +]) diff --git a/gl/m4/ssize_t.m4 b/gl/m4/ssize_t.m4 new file mode 100644 index 00000000..4eaef93c --- /dev/null +++ b/gl/m4/ssize_t.m4 @@ -0,0 +1,21 @@ +# ssize_t.m4 serial 4 (gettext-0.15) +dnl Copyright (C) 2001-2003, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Bruno Haible. +dnl Test whether ssize_t is defined. + +AC_DEFUN([gt_TYPE_SSIZE_T], +[ + AC_CACHE_CHECK([for ssize_t], [gt_cv_ssize_t], + [AC_TRY_COMPILE([#include <sys/types.h>], + [int x = sizeof (ssize_t *) + sizeof (ssize_t); + return !x;], + [gt_cv_ssize_t=yes], [gt_cv_ssize_t=no])]) + if test $gt_cv_ssize_t = no; then + AC_DEFINE([ssize_t], [int], + [Define as a signed type of the same size as size_t.]) + fi +]) diff --git a/gl/m4/stdbool.m4 b/gl/m4/stdbool.m4 new file mode 100644 index 00000000..2204ecd9 --- /dev/null +++ b/gl/m4/stdbool.m4 @@ -0,0 +1,115 @@ +# Check for stdbool.h that conforms to C99. + +dnl Copyright (C) 2002-2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +# Prepare for substituting <stdbool.h> if it is not supported. + +AC_DEFUN([AM_STDBOOL_H], +[ + AC_REQUIRE([AC_HEADER_STDBOOL]) + + # Define two additional variables used in the Makefile substitution. + + if test "$ac_cv_header_stdbool_h" = yes; then + STDBOOL_H='' + else + STDBOOL_H='stdbool.h' + fi + AC_SUBST([STDBOOL_H]) + + if test "$ac_cv_type__Bool" = yes; then + HAVE__BOOL=1 + else + HAVE__BOOL=0 + fi + AC_SUBST([HAVE__BOOL]) +]) + +# AM_STDBOOL_H will be renamed to gl_STDBOOL_H in the future. +AC_DEFUN([gl_STDBOOL_H], [AM_STDBOOL_H]) + +# This macro is only needed in autoconf <= 2.59. Newer versions of autoconf +# have this macro built-in. + +AC_DEFUN([AC_HEADER_STDBOOL], + [AC_CACHE_CHECK([for stdbool.h that conforms to C99], + [ac_cv_header_stdbool_h], + [AC_TRY_COMPILE( + [ + #include <stdbool.h> + #ifndef bool + "error: bool is not defined" + #endif + #ifndef false + "error: false is not defined" + #endif + #if false + "error: false is not 0" + #endif + #ifndef true + "error: true is not defined" + #endif + #if true != 1 + "error: true is not 1" + #endif + #ifndef __bool_true_false_are_defined + "error: __bool_true_false_are_defined is not defined" + #endif + + struct s { _Bool s: 1; _Bool t; } s; + + char a[true == 1 ? 1 : -1]; + char b[false == 0 ? 1 : -1]; + char c[__bool_true_false_are_defined == 1 ? 1 : -1]; + char d[(bool) 0.5 == true ? 1 : -1]; + bool e = &s; + char f[(_Bool) 0.0 == false ? 1 : -1]; + char g[true]; + char h[sizeof (_Bool)]; + char i[sizeof s.t]; + enum { j = false, k = true, l = false * true, m = true * 256 }; + _Bool n[m]; + char o[sizeof n == m * sizeof n[0] ? 1 : -1]; + char p[-1 - (_Bool) 0 < 0 && -1 - (bool) 0 < 0 ? 1 : -1]; + #if defined __xlc__ || defined __GNUC__ + /* Catch a bug in IBM AIX xlc compiler version 6.0.0.0 + reported by James Lemley on 2005-10-05; see + http://lists.gnu.org/archive/html/bug-coreutils/2005-10/msg00086.html + This test is not quite right, since xlc is allowed to + reject this program, as the initializer for xlcbug is + not one of the forms that C requires support for. + However, doing the test right would require a run-time + test, and that would make cross-compilation harder. + Let us hope that IBM fixes the xlc bug, and also adds + support for this kind of constant expression. In the + meantime, this test will reject xlc, which is OK, since + our stdbool.h substitute should suffice. We also test + this with GCC, where it should work, to detect more + quickly whether someone messes up the test in the + future. */ + char digs[] = "0123456789"; + int xlcbug = 1 / (&(digs + 5)[-2 + (bool) 1] == &digs[4] ? 1 : -1); + #endif + /* Catch a bug in an HP-UX C compiler. See + http://gcc.gnu.org/ml/gcc-patches/2003-12/msg02303.html + http://lists.gnu.org/archive/html/bug-coreutils/2005-11/msg00161.html + */ + _Bool q = true; + _Bool *pq = &q; + ], + [ + *pq |= q; + *pq |= ! q; + /* Refer to every declared value, to avoid compiler optimizations. */ + return (!a + !b + !c + !d + !e + !f + !g + !h + !i + !!j + !k + !!l + + !m + !n + !o + !p + !q + !pq); + ], + [ac_cv_header_stdbool_h=yes], + [ac_cv_header_stdbool_h=no])]) + AC_CHECK_TYPES([_Bool]) + if test $ac_cv_header_stdbool_h = yes; then + AC_DEFINE(HAVE_STDBOOL_H, 1, [Define to 1 if stdbool.h conforms to C99.]) + fi]) diff --git a/gl/m4/stdint.m4 b/gl/m4/stdint.m4 new file mode 100644 index 00000000..1a4b4a6a --- /dev/null +++ b/gl/m4/stdint.m4 @@ -0,0 +1,369 @@ +# stdint.m4 serial 22 +dnl Copyright (C) 2001-2002, 2004-2007 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Paul Eggert and Bruno Haible. +dnl Test whether <stdint.h> is supported or must be substituted. + +AC_DEFUN([gl_STDINT_H], +[ + AC_PREREQ(2.59)dnl + + dnl Check for long long int and unsigned long long int. + AC_REQUIRE([AC_TYPE_LONG_LONG_INT]) + if test $ac_cv_type_long_long_int = yes; then + HAVE_LONG_LONG_INT=1 + else + HAVE_LONG_LONG_INT=0 + fi + AC_SUBST([HAVE_LONG_LONG_INT]) + AC_REQUIRE([AC_TYPE_UNSIGNED_LONG_LONG_INT]) + if test $ac_cv_type_unsigned_long_long_int = yes; then + HAVE_UNSIGNED_LONG_LONG_INT=1 + else + HAVE_UNSIGNED_LONG_LONG_INT=0 + fi + AC_SUBST([HAVE_UNSIGNED_LONG_LONG_INT]) + + dnl Check for <inttypes.h>. + dnl AC_INCLUDES_DEFAULT defines $ac_cv_header_inttypes_h. + if test $ac_cv_header_inttypes_h = yes; then + HAVE_INTTYPES_H=1 + else + HAVE_INTTYPES_H=0 + fi + AC_SUBST([HAVE_INTTYPES_H]) + + dnl Check for <sys/types.h>. + dnl AC_INCLUDES_DEFAULT defines $ac_cv_header_sys_types_h. + if test $ac_cv_header_sys_types_h = yes; then + HAVE_SYS_TYPES_H=1 + else + HAVE_SYS_TYPES_H=0 + fi + AC_SUBST([HAVE_SYS_TYPES_H]) + + dnl AC_INCLUDES_DEFAULT defines $ac_cv_header_stdint_h. + if test $ac_cv_header_stdint_h = yes; then + gl_ABSOLUTE_HEADER([stdint.h]) + ABSOLUTE_STDINT_H=\"$gl_cv_absolute_stdint_h\" + HAVE_STDINT_H=1 + else + ABSOLUTE_STDINT_H=\"no/such/file/stdint.h\" + HAVE_STDINT_H=0 + fi + AC_SUBST([ABSOLUTE_STDINT_H]) + AC_SUBST([HAVE_STDINT_H]) + + dnl Now see whether we need a substitute <stdint.h>. Use + dnl ABSOLUTE_STDINT_H, not <stdint.h>, so that it also works during + dnl a "config.status --recheck" if a stdint.h has been + dnl created in the build directory. + if test $ac_cv_header_stdint_h = yes; then + AC_CACHE_CHECK([whether stdint.h conforms to C99], + [gl_cv_header_working_stdint_h], + [gl_cv_header_working_stdint_h=no + AC_COMPILE_IFELSE([ + AC_LANG_PROGRAM([[ +#include <stddef.h> +#define __STDC_LIMIT_MACROS 1 /* to make it work also in C++ mode */ +#define __STDC_CONSTANT_MACROS 1 /* to make it work also in C++ mode */ +#include ABSOLUTE_STDINT_H +#ifdef INT8_MAX +int8_t a1 = INT8_MAX; +int8_t a1min = INT8_MIN; +#endif +#ifdef INT16_MAX +int16_t a2 = INT16_MAX; +int16_t a2min = INT16_MIN; +#endif +#ifdef INT32_MAX +int32_t a3 = INT32_MAX; +int32_t a3min = INT32_MIN; +#endif +#ifdef INT64_MAX +int64_t a4 = INT64_MAX; +int64_t a4min = INT64_MIN; +#endif +#ifdef UINT8_MAX +uint8_t b1 = UINT8_MAX; +#else +typedef int b1[(unsigned char) -1 != 255 ? 1 : -1]; +#endif +#ifdef UINT16_MAX +uint16_t b2 = UINT16_MAX; +#endif +#ifdef UINT32_MAX +uint32_t b3 = UINT32_MAX; +#endif +#ifdef UINT64_MAX +uint64_t b4 = UINT64_MAX; +#endif +int_least8_t c1 = INT8_C (0x7f); +int_least8_t c1max = INT_LEAST8_MAX; +int_least8_t c1min = INT_LEAST8_MIN; +int_least16_t c2 = INT16_C (0x7fff); +int_least16_t c2max = INT_LEAST16_MAX; +int_least16_t c2min = INT_LEAST16_MIN; +int_least32_t c3 = INT32_C (0x7fffffff); +int_least32_t c3max = INT_LEAST32_MAX; +int_least32_t c3min = INT_LEAST32_MIN; +int_least64_t c4 = INT64_C (0x7fffffffffffffff); +int_least64_t c4max = INT_LEAST64_MAX; +int_least64_t c4min = INT_LEAST64_MIN; +uint_least8_t d1 = UINT8_C (0xff); +uint_least8_t d1max = UINT_LEAST8_MAX; +uint_least16_t d2 = UINT16_C (0xffff); +uint_least16_t d2max = UINT_LEAST16_MAX; +uint_least32_t d3 = UINT32_C (0xffffffff); +uint_least32_t d3max = UINT_LEAST32_MAX; +uint_least64_t d4 = UINT64_C (0xffffffffffffffff); +uint_least64_t d4max = UINT_LEAST64_MAX; +int_fast8_t e1 = INT_FAST8_MAX; +int_fast8_t e1min = INT_FAST8_MIN; +int_fast16_t e2 = INT_FAST16_MAX; +int_fast16_t e2min = INT_FAST16_MIN; +int_fast32_t e3 = INT_FAST32_MAX; +int_fast32_t e3min = INT_FAST32_MIN; +int_fast64_t e4 = INT_FAST64_MAX; +int_fast64_t e4min = INT_FAST64_MIN; +uint_fast8_t f1 = UINT_FAST8_MAX; +uint_fast16_t f2 = UINT_FAST16_MAX; +uint_fast32_t f3 = UINT_FAST32_MAX; +uint_fast64_t f4 = UINT_FAST64_MAX; +#ifdef INTPTR_MAX +intptr_t g = INTPTR_MAX; +intptr_t gmin = INTPTR_MIN; +#endif +#ifdef UINTPTR_MAX +uintptr_t h = UINTPTR_MAX; +#endif +intmax_t i = INTMAX_MAX; +uintmax_t j = UINTMAX_MAX; +struct s { + int check_PTRDIFF: PTRDIFF_MIN < 0 && 0 < PTRDIFF_MAX ? 1 : -1; + int check_SIG_ATOMIC: SIG_ATOMIC_MIN <= 0 && 0 < SIG_ATOMIC_MAX ? 1 : -1; + int check_SIZE: 0 < SIZE_MAX ? 1 : -1; + int check_WCHAR: WCHAR_MIN <= 0 && 0 < WCHAR_MAX ? 1 : -1; + int check_WINT: WINT_MIN <= 0 && 0 < WINT_MAX ? 1 : -1; + + /* Detect bugs in glibc 2.4 and Solaris 10 stdint.h, among others. */ + int check_UINT8_C: + (-1 < UINT8_C (0)) == (-1 < (uint_least8_t) 0) ? 1 : -1; + int check_UINT16_C: + (-1 < UINT16_C (0)) == (-1 < (uint_least16_t) 0) ? 1 : -1; + + /* Detect bugs in OpenBSD 3.9 stdint.h. */ +#ifdef UINT8_MAX + int check_uint8: (uint8_t) -1 == UINT8_MAX ? 1 : -1; +#endif +#ifdef UINT16_MAX + int check_uint16: (uint16_t) -1 == UINT16_MAX ? 1 : -1; +#endif +#ifdef UINT32_MAX + int check_uint32: (uint32_t) -1 == UINT32_MAX ? 1 : -1; +#endif +#ifdef UINT64_MAX + int check_uint64: (uint64_t) -1 == UINT64_MAX ? 1 : -1; +#endif + int check_uint_least8: (uint_least8_t) -1 == UINT_LEAST8_MAX ? 1 : -1; + int check_uint_least16: (uint_least16_t) -1 == UINT_LEAST16_MAX ? 1 : -1; + int check_uint_least32: (uint_least32_t) -1 == UINT_LEAST32_MAX ? 1 : -1; + int check_uint_least64: (uint_least64_t) -1 == UINT_LEAST64_MAX ? 1 : -1; + int check_uint_fast8: (uint_fast8_t) -1 == UINT_FAST8_MAX ? 1 : -1; + int check_uint_fast16: (uint_fast16_t) -1 == UINT_FAST16_MAX ? 1 : -1; + int check_uint_fast32: (uint_fast32_t) -1 == UINT_FAST32_MAX ? 1 : -1; + int check_uint_fast64: (uint_fast64_t) -1 == UINT_FAST64_MAX ? 1 : -1; + int check_uintptr: (uintptr_t) -1 == UINTPTR_MAX ? 1 : -1; + int check_uintmax: (uintmax_t) -1 == UINTMAX_MAX ? 1 : -1; + int check_size: (size_t) -1 == SIZE_MAX ? 1 : -1; +}; + ]])], + [gl_cv_header_working_stdint_h=yes])]) + fi + if test "$gl_cv_header_working_stdint_h" = yes; then + STDINT_H= + else + dnl Check for <sys/inttypes.h>, and for + dnl <sys/bitypes.h> (used in Linux libc4 >= 4.6.7 and libc5). + AC_CHECK_HEADERS([sys/inttypes.h sys/bitypes.h]) + if test $ac_cv_header_sys_inttypes_h = yes; then + HAVE_SYS_INTTYPES_H=1 + else + HAVE_SYS_INTTYPES_H=0 + fi + AC_SUBST([HAVE_SYS_INTTYPES_H]) + if test $ac_cv_header_sys_bitypes_h = yes; then + HAVE_SYS_BITYPES_H=1 + else + HAVE_SYS_BITYPES_H=0 + fi + AC_SUBST([HAVE_SYS_BITYPES_H]) + + gl_STDINT_TYPE_PROPERTIES + STDINT_H=stdint.h + fi + AC_SUBST(STDINT_H) +]) + +dnl gl_STDINT_BITSIZEOF(TYPES, INCLUDES) +dnl Determine the size of each of the given types in bits. +AC_DEFUN([gl_STDINT_BITSIZEOF], +[ + dnl Use a shell loop, to avoid bloating configure, and + dnl - extra AH_TEMPLATE calls, so that autoheader knows what to put into + dnl config.h.in, + dnl - extra AC_SUBST calls, so that the right substitutions are made. + AC_FOREACH([gltype], [$1], + [AH_TEMPLATE([BITSIZEOF_]translit(gltype,[abcdefghijklmnopqrstuvwxyz ],[ABCDEFGHIJKLMNOPQRSTUVWXYZ_]), + [Define to the number of bits in type ']gltype['.])]) + for gltype in $1 ; do + AC_CACHE_CHECK([for bit size of $gltype], [gl_cv_bitsizeof_${gltype}], + [AC_COMPUTE_INT([result], [sizeof ($gltype) * CHAR_BIT], + [$2 +#include <limits.h>], [result=unknown]) + eval gl_cv_bitsizeof_${gltype}=\$result + ]) + eval result=\$gl_cv_bitsizeof_${gltype} + if test $result = unknown; then + dnl Use a nonempty default, because some compilers, such as IRIX 5 cc, + dnl do a syntax check even on unused #if conditions and give an error + dnl on valid C code like this: + dnl #if 0 + dnl # if > 32 + dnl # endif + dnl #endif + result=0 + fi + GLTYPE=`echo "$gltype" | tr 'abcdefghijklmnopqrstuvwxyz ' 'ABCDEFGHIJKLMNOPQRSTUVWXYZ_'` + AC_DEFINE_UNQUOTED([BITSIZEOF_${GLTYPE}], [$result]) + eval BITSIZEOF_${GLTYPE}=\$result + done + AC_FOREACH([gltype], [$1], + [AC_SUBST([BITSIZEOF_]translit(gltype,[abcdefghijklmnopqrstuvwxyz ],[ABCDEFGHIJKLMNOPQRSTUVWXYZ_]))]) +]) + +dnl gl_CHECK_TYPES_SIGNED(TYPES, INCLUDES) +dnl Determine the signedness of each of the given types. +dnl Define HAVE_SIGNED_TYPE if type is signed. +AC_DEFUN([gl_CHECK_TYPES_SIGNED], +[ + dnl Use a shell loop, to avoid bloating configure, and + dnl - extra AH_TEMPLATE calls, so that autoheader knows what to put into + dnl config.h.in, + dnl - extra AC_SUBST calls, so that the right substitutions are made. + AC_FOREACH([gltype], [$1], + [AH_TEMPLATE([HAVE_SIGNED_]translit(gltype,[abcdefghijklmnopqrstuvwxyz ],[ABCDEFGHIJKLMNOPQRSTUVWXYZ_]), + [Define to 1 if ']gltype[' is a signed integer type.])]) + for gltype in $1 ; do + AC_CACHE_CHECK([whether $gltype is signed], [gl_cv_type_${gltype}_signed], + [AC_COMPILE_IFELSE( + [AC_LANG_PROGRAM([$2[ + int verify[2 * (($gltype) -1 < ($gltype) 0) - 1];]])], + result=yes, result=no) + eval gl_cv_type_${gltype}_signed=\$result + ]) + eval result=\$gl_cv_type_${gltype}_signed + GLTYPE=`echo $gltype | tr 'abcdefghijklmnopqrstuvwxyz ' 'ABCDEFGHIJKLMNOPQRSTUVWXYZ_'` + if test "$result" = yes; then + AC_DEFINE_UNQUOTED([HAVE_SIGNED_${GLTYPE}], 1) + eval HAVE_SIGNED_${GLTYPE}=1 + else + eval HAVE_SIGNED_${GLTYPE}=0 + fi + done + AC_FOREACH([gltype], [$1], + [AC_SUBST([HAVE_SIGNED_]translit(gltype,[abcdefghijklmnopqrstuvwxyz ],[ABCDEFGHIJKLMNOPQRSTUVWXYZ_]))]) +]) + +dnl gl_INTEGER_TYPE_SUFFIX(TYPES, INCLUDES) +dnl Determine the suffix to use for integer constants of the given types. +dnl Define t_SUFFIX for each such type. +AC_DEFUN([gl_INTEGER_TYPE_SUFFIX], +[ + dnl Use a shell loop, to avoid bloating configure, and + dnl - extra AH_TEMPLATE calls, so that autoheader knows what to put into + dnl config.h.in, + dnl - extra AC_SUBST calls, so that the right substitutions are made. + AC_FOREACH([gltype], [$1], + [AH_TEMPLATE(translit(gltype,[abcdefghijklmnopqrstuvwxyz ],[ABCDEFGHIJKLMNOPQRSTUVWXYZ_])[_SUFFIX], + [Define to l, ll, u, ul, ull, etc., as suitable for + constants of type ']gltype['.])]) + for gltype in $1 ; do + AC_CACHE_CHECK([for $gltype integer literal suffix], + [gl_cv_type_${gltype}_suffix], + [eval gl_cv_type_${gltype}_suffix=no + eval result=\$gl_cv_type_${gltype}_signed + if test "$result" = yes; then + glsufu= + else + glsufu=u + fi + for glsuf in "$glsufu" ${glsufu}l ${glsufu}ll ${glsufu}i64; do + case $glsuf in + '') gltype1='int';; + l) gltype1='long int';; + ll) gltype1='long long int';; + i64) gltype1='__int64';; + u) gltype1='unsigned int';; + ul) gltype1='unsigned long int';; + ull) gltype1='unsigned long long int';; + ui64)gltype1='unsigned __int64';; + esac + AC_COMPILE_IFELSE( + [AC_LANG_PROGRAM([$2 + extern $gltype foo; + extern $gltype1 foo;])], + [eval gl_cv_type_${gltype}_suffix=\$glsuf]) + eval result=\$gl_cv_type_${gltype}_suffix + test "$result" != no && break + done]) + GLTYPE=`echo $gltype | tr 'abcdefghijklmnopqrstuvwxyz ' 'ABCDEFGHIJKLMNOPQRSTUVWXYZ_'` + eval result=\$gl_cv_type_${gltype}_suffix + test "$result" = no && result= + eval ${GLTYPE}_SUFFIX=\$result + AC_DEFINE_UNQUOTED([${GLTYPE}_SUFFIX], $result) + done + AC_FOREACH([gltype], [$1], + [AC_SUBST(translit(gltype,[abcdefghijklmnopqrstuvwxyz ],[ABCDEFGHIJKLMNOPQRSTUVWXYZ_])[_SUFFIX])]) +]) + +dnl gl_STDINT_INCLUDES +AC_DEFUN([gl_STDINT_INCLUDES], +[[ + /* BSD/OS 4.0.1 has a bug: <stddef.h>, <stdio.h> and <time.h> must be + included before <wchar.h>. */ + #include <stddef.h> + #include <signal.h> + #include <stdio.h> + #include <time.h> + #include <wchar.h> +]]) + +dnl gl_STDINT_TYPE_PROPERTIES +dnl Compute HAVE_SIGNED_t, BITSIZEOF_t and t_SUFFIX, for all the types t +dnl of interest to stdint_.h. +AC_DEFUN([gl_STDINT_TYPE_PROPERTIES], +[ + gl_STDINT_BITSIZEOF([ptrdiff_t sig_atomic_t size_t wchar_t wint_t], + [gl_STDINT_INCLUDES]) + gl_CHECK_TYPES_SIGNED([sig_atomic_t wchar_t wint_t], + [gl_STDINT_INCLUDES]) + gl_cv_type_ptrdiff_t_signed=yes + gl_cv_type_size_t_signed=no + gl_INTEGER_TYPE_SUFFIX([ptrdiff_t sig_atomic_t size_t wchar_t wint_t], + [gl_STDINT_INCLUDES]) +]) + +dnl Autoconf >= 2.61 has AC_COMPUTE_INT built-in. +dnl Remove this when we can assume autoconf >= 2.61. +m4_ifdef([AC_COMPUTE_INT], [], [ + AC_DEFUN([AC_COMPUTE_INT], [_AC_COMPUTE_INT([$2],[$1],[$3],[$4])]) +]) + +# Hey Emacs! +# Local Variables: +# indent-tabs-mode: nil +# End: diff --git a/gl/m4/stdint_h.m4 b/gl/m4/stdint_h.m4 new file mode 100644 index 00000000..db9a8ac4 --- /dev/null +++ b/gl/m4/stdint_h.m4 @@ -0,0 +1,26 @@ +# stdint_h.m4 serial 6 +dnl Copyright (C) 1997-2004, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Paul Eggert. + +# Define HAVE_STDINT_H_WITH_UINTMAX if <stdint.h> exists, +# doesn't clash with <sys/types.h>, and declares uintmax_t. + +AC_DEFUN([gl_AC_HEADER_STDINT_H], +[ + AC_CACHE_CHECK([for stdint.h], gl_cv_header_stdint_h, + [AC_TRY_COMPILE( + [#include <sys/types.h> +#include <stdint.h>], + [uintmax_t i = (uintmax_t) -1; return !i;], + gl_cv_header_stdint_h=yes, + gl_cv_header_stdint_h=no)]) + if test $gl_cv_header_stdint_h = yes; then + AC_DEFINE_UNQUOTED(HAVE_STDINT_H_WITH_UINTMAX, 1, + [Define if <stdint.h> exists, doesn't clash with <sys/types.h>, + and declares uintmax_t. ]) + fi +]) diff --git a/gl/m4/strcase.m4 b/gl/m4/strcase.m4 new file mode 100644 index 00000000..ae827907 --- /dev/null +++ b/gl/m4/strcase.m4 @@ -0,0 +1,39 @@ +# strcase.m4 serial 4 +dnl Copyright (C) 2002, 2005-2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_STRCASE], +[ + gl_FUNC_STRCASECMP + gl_FUNC_STRNCASECMP +]) + +AC_DEFUN([gl_FUNC_STRCASECMP], +[ + dnl No known system has a strcasecmp() function that works correctly in + dnl multibyte locales. Therefore we use our version always. + AC_LIBOBJ(strcasecmp) + gl_PREREQ_STRCASECMP +]) + +AC_DEFUN([gl_FUNC_STRNCASECMP], +[ + AC_REPLACE_FUNCS(strncasecmp) + AC_CHECK_DECLS(strncasecmp) + if test $ac_cv_func_strncasecmp = no; then + gl_PREREQ_STRNCASECMP + fi +]) + +# Prerequisites of lib/strcasecmp.c. +AC_DEFUN([gl_PREREQ_STRCASECMP], [ + AC_REQUIRE([gl_FUNC_MBRTOWC]) + : +]) + +# Prerequisites of lib/strncasecmp.c. +AC_DEFUN([gl_PREREQ_STRNCASECMP], [ + : +]) diff --git a/gl/m4/strdup.m4 b/gl/m4/strdup.m4 new file mode 100644 index 00000000..f7786e94 --- /dev/null +++ b/gl/m4/strdup.m4 @@ -0,0 +1,15 @@ +# strdup.m4 serial 7 +dnl Copyright (C) 2002, 2003, 2004, 2005, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_FUNC_STRDUP], +[ + AC_REPLACE_FUNCS(strdup) + AC_CHECK_DECLS_ONCE(strdup) + gl_PREREQ_STRDUP +]) + +# Prerequisites of lib/strdup.c. +AC_DEFUN([gl_PREREQ_STRDUP], [:]) diff --git a/gl/m4/strndup.m4 b/gl/m4/strndup.m4 new file mode 100644 index 00000000..dd5780b6 --- /dev/null +++ b/gl/m4/strndup.m4 @@ -0,0 +1,48 @@ +# strndup.m4 serial 11 +dnl Copyright (C) 2002-2003, 2005-2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_FUNC_STRNDUP], +[ + dnl Persuade glibc <string.h> to declare strndup(). + AC_REQUIRE([AC_GNU_SOURCE]) + + AC_CHECK_DECLS_ONCE([strndup]) + + # AIX 4.3.3, AIX 5.1 have a function that fails to add the terminating '\0'. + AC_CACHE_CHECK([for working strndup], gl_cv_func_strndup, + [AC_RUN_IFELSE([ + AC_LANG_PROGRAM([#include <string.h> + #include <stdlib.h>], [[ +#ifndef HAVE_DECL_STRNDUP + extern char *strndup (const char *, size_t); +#endif + char *s; + s = strndup ("some longer string", 15); + free (s); + s = strndup ("shorter string", 13); + return s[13] != '\0';]])], + [gl_cv_func_strndup=yes], + [gl_cv_func_strndup=no], + [AC_CHECK_FUNC([strndup], + [AC_EGREP_CPP([too risky], [ +#ifdef _AIX + too risky +#endif + ], + [gl_cv_func_strndup=no], + [gl_cv_func_strndup=yes])], + [gl_cv_func_strndup=no])])]) + if test $gl_cv_func_strndup = yes; then + AC_DEFINE([HAVE_STRNDUP], 1, + [Define if you have the strndup() function and it works.]) + else + AC_LIBOBJ([strndup]) + gl_PREREQ_STRNDUP + fi +]) + +# Prerequisites of lib/strndup.c. +AC_DEFUN([gl_PREREQ_STRNDUP], [:]) diff --git a/gl/m4/strnlen.m4 b/gl/m4/strnlen.m4 new file mode 100644 index 00000000..0213a8ae --- /dev/null +++ b/gl/m4/strnlen.m4 @@ -0,0 +1,27 @@ +# strnlen.m4 serial 6 +dnl Copyright (C) 2002-2003, 2005, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_FUNC_STRNLEN], +[ + dnl Persuade glibc <string.h> to declare strnlen(). + AC_REQUIRE([AC_GNU_SOURCE]) + + AC_FUNC_STRNLEN + if test $ac_cv_func_strnlen_working = no; then + # This is necessary because automake-1.6.1 doens't understand + # that the above use of AC_FUNC_STRNLEN means we may have to use + # lib/strnlen.c. + #AC_LIBOBJ(strnlen) + AC_DEFINE(strnlen, rpl_strnlen, + [Define to rpl_strnlen if the replacement function should be used.]) + gl_PREREQ_STRNLEN + fi +]) + +# Prerequisites of lib/strnlen.c. +AC_DEFUN([gl_PREREQ_STRNLEN], [ + AC_CHECK_DECLS_ONCE(strnlen) +]) diff --git a/gl/m4/sys_socket_h.m4 b/gl/m4/sys_socket_h.m4 new file mode 100644 index 00000000..d3e45b48 --- /dev/null +++ b/gl/m4/sys_socket_h.m4 @@ -0,0 +1,23 @@ +# sys_socket_h.m4 serial 2 +dnl Copyright (C) 2005, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Simon Josefsson. + +AC_DEFUN([gl_HEADER_SYS_SOCKET], +[ + AC_CHECK_HEADERS_ONCE([sys/socket.h]) + if test $ac_cv_header_sys_socket_h = yes; then + SYS_SOCKET_H='' + else + dnl We cannot use AC_CHECK_HEADERS_ONCE here, because that would make + dnl the check for those headers unconditional; yet cygwin reports + dnl that the headers are present but cannot be compiled (since on + dnl cygwin, all socket information should come from sys/socket.h). + AC_CHECK_HEADERS([winsock2.h ws2tcpip.h]) + SYS_SOCKET_H='sys/socket.h' + fi + AC_SUBST(SYS_SOCKET_H) +]) diff --git a/gl/m4/uintmax_t.m4 b/gl/m4/uintmax_t.m4 new file mode 100644 index 00000000..bf83ed74 --- /dev/null +++ b/gl/m4/uintmax_t.m4 @@ -0,0 +1,30 @@ +# uintmax_t.m4 serial 9 +dnl Copyright (C) 1997-2004 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Paul Eggert. + +AC_PREREQ(2.13) + +# Define uintmax_t to 'unsigned long' or 'unsigned long long' +# if it is not already defined in <stdint.h> or <inttypes.h>. + +AC_DEFUN([gl_AC_TYPE_UINTMAX_T], +[ + AC_REQUIRE([gl_AC_HEADER_INTTYPES_H]) + AC_REQUIRE([gl_AC_HEADER_STDINT_H]) + if test $gl_cv_header_inttypes_h = no && test $gl_cv_header_stdint_h = no; then + AC_REQUIRE([gl_AC_TYPE_UNSIGNED_LONG_LONG]) + test $ac_cv_type_unsigned_long_long = yes \ + && ac_type='unsigned long long' \ + || ac_type='unsigned long' + AC_DEFINE_UNQUOTED(uintmax_t, $ac_type, + [Define to unsigned long or unsigned long long + if <stdint.h> and <inttypes.h> don't define.]) + else + AC_DEFINE(HAVE_UINTMAX_T, 1, + [Define if you have the 'uintmax_t' type in <stdint.h> or <inttypes.h>.]) + fi +]) diff --git a/gl/m4/ulonglong.m4 b/gl/m4/ulonglong.m4 new file mode 100644 index 00000000..9fae98e3 --- /dev/null +++ b/gl/m4/ulonglong.m4 @@ -0,0 +1,48 @@ +# ulonglong.m4 serial 6 +dnl Copyright (C) 1999-2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Paul Eggert. + +# Define HAVE_UNSIGNED_LONG_LONG_INT if 'unsigned long long int' works. +# This fixes a bug in Autoconf 2.60, but can be removed once we +# assume 2.61 everywhere. + +# Note: If the type 'unsigned long long int' exists but is only 32 bits +# large (as on some very old compilers), AC_TYPE_UNSIGNED_LONG_LONG_INT +# will not be defined. In this case you can treat 'unsigned long long int' +# like 'unsigned long int'. + +AC_DEFUN([AC_TYPE_UNSIGNED_LONG_LONG_INT], +[ + AC_CACHE_CHECK([for unsigned long long int], + [ac_cv_type_unsigned_long_long_int], + [AC_LINK_IFELSE( + [AC_LANG_PROGRAM( + [[unsigned long long int ull = 18446744073709551615ULL; + typedef int a[(18446744073709551615ULL <= (unsigned long long int) -1 + ? 1 : -1)]; + int i = 63;]], + [[unsigned long long int ullmax = 18446744073709551615ull; + return (ull << 63 | ull >> 63 | ull << i | ull >> i + | ullmax / ull | ullmax % ull);]])], + [ac_cv_type_unsigned_long_long_int=yes], + [ac_cv_type_unsigned_long_long_int=no])]) + if test $ac_cv_type_unsigned_long_long_int = yes; then + AC_DEFINE([HAVE_UNSIGNED_LONG_LONG_INT], 1, + [Define to 1 if the system has the type `unsigned long long int'.]) + fi +]) + +# This macro is obsolescent and should go away soon. +AC_DEFUN([gl_AC_TYPE_UNSIGNED_LONG_LONG], +[ + AC_REQUIRE([AC_TYPE_UNSIGNED_LONG_LONG_INT]) + ac_cv_type_unsigned_long_long=$ac_cv_type_unsigned_long_long_int + if test $ac_cv_type_unsigned_long_long = yes; then + AC_DEFINE(HAVE_UNSIGNED_LONG_LONG, 1, + [Define if you have the 'unsigned long long' type.]) + fi +]) diff --git a/gl/m4/unistd-safer.m4 b/gl/m4/unistd-safer.m4 new file mode 100644 index 00000000..09adf931 --- /dev/null +++ b/gl/m4/unistd-safer.m4 @@ -0,0 +1,13 @@ +#serial 8 +dnl Copyright (C) 2002, 2005, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_UNISTD_SAFER], +[ + AC_CHECK_FUNCS_ONCE([pipe]) + AC_LIBOBJ([dup-safer]) + AC_LIBOBJ([fd-safer]) + AC_LIBOBJ([pipe-safer]) +]) diff --git a/gl/m4/unistd_h.m4 b/gl/m4/unistd_h.m4 new file mode 100644 index 00000000..9d499dfe --- /dev/null +++ b/gl/m4/unistd_h.m4 @@ -0,0 +1,32 @@ +# unistd_h.m4 serial 3 +dnl Copyright (C) 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl Written by Simon Josefsson + +AC_DEFUN([gl_HEADER_UNISTD], +[ + AC_CHECK_HEADERS([unistd.h], [ + UNISTD_H='' + ], [ + UNISTD_H='unistd.h' + ]) + AC_SUBST(UNISTD_H) + dnl This module decides to build unistd.h if it is missing. + dnl The fchdir module decides to build unistd.h if fchdir() is missing. + dnl Therefore check for the prerequisites of lib/unistd.h always. + gl_PREREQ_UNISTD +]) + +dnl Prerequisites of lib/unistd.h. +AC_DEFUN([gl_PREREQ_UNISTD], +[ + AC_CHECK_HEADERS_ONCE([unistd.h]) + if test $ac_cv_header_unistd_h = yes; then + gl_ABSOLUTE_HEADER([unistd.h]) + ABSOLUTE_UNISTD_H=\"$gl_cv_absolute_unistd_h\" + fi + AC_SUBST([ABSOLUTE_UNISTD_H]) +]) diff --git a/gl/m4/vasnprintf.m4 b/gl/m4/vasnprintf.m4 new file mode 100644 index 00000000..72c9a13e --- /dev/null +++ b/gl/m4/vasnprintf.m4 @@ -0,0 +1,57 @@ +# vasnprintf.m4 serial 7 +dnl Copyright (C) 2002-2004, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_FUNC_VASNPRINTF], +[ + AC_REQUIRE([gl_EOVERFLOW]) + AC_REPLACE_FUNCS(vasnprintf) + if test $ac_cv_func_vasnprintf = no; then + AC_LIBOBJ(printf-args) + AC_LIBOBJ(printf-parse) + AC_LIBOBJ(asnprintf) + gl_PREREQ_PRINTF_ARGS + gl_PREREQ_PRINTF_PARSE + gl_PREREQ_VASNPRINTF + gl_PREREQ_ASNPRINTF + fi +]) + +# Prequisites of lib/printf-args.h, lib/printf-args.c. +AC_DEFUN([gl_PREREQ_PRINTF_ARGS], +[ + AC_REQUIRE([AC_TYPE_LONG_LONG_INT]) + AC_REQUIRE([gt_TYPE_LONGDOUBLE]) + AC_REQUIRE([gt_TYPE_WCHAR_T]) + AC_REQUIRE([gt_TYPE_WINT_T]) +]) + +# Prequisites of lib/printf-parse.h, lib/printf-parse.c. +AC_DEFUN([gl_PREREQ_PRINTF_PARSE], +[ + AC_REQUIRE([AC_TYPE_LONG_LONG_INT]) + AC_REQUIRE([gt_TYPE_LONGDOUBLE]) + AC_REQUIRE([gt_TYPE_WCHAR_T]) + AC_REQUIRE([gt_TYPE_WINT_T]) + AC_REQUIRE([AC_TYPE_SIZE_T]) + AC_CHECK_TYPES(ptrdiff_t) + AC_REQUIRE([gt_AC_TYPE_INTMAX_T]) +]) + +# Prerequisites of lib/vasnprintf.c. +AC_DEFUN([gl_PREREQ_VASNPRINTF], +[ + AC_REQUIRE([AC_FUNC_ALLOCA]) + AC_REQUIRE([AC_TYPE_LONG_LONG_INT]) + AC_REQUIRE([gt_TYPE_LONGDOUBLE]) + AC_REQUIRE([gt_TYPE_WCHAR_T]) + AC_REQUIRE([gt_TYPE_WINT_T]) + AC_CHECK_FUNCS(snprintf wcslen) +]) + +# Prerequisites of lib/asnprintf.c. +AC_DEFUN([gl_PREREQ_ASNPRINTF], +[ +]) diff --git a/gl/m4/vasprintf.m4 b/gl/m4/vasprintf.m4 new file mode 100644 index 00000000..18ca6327 --- /dev/null +++ b/gl/m4/vasprintf.m4 @@ -0,0 +1,33 @@ +# vasprintf.m4 serial 2 +dnl Copyright (C) 2002-2003, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_FUNC_VASPRINTF], +[ + AC_REPLACE_FUNCS(vasprintf) + if test $ac_cv_func_vasprintf = no; then + AC_LIBOBJ(asprintf) + gl_PREREQ_VASPRINTF_H + gl_PREREQ_VASPRINTF + gl_PREREQ_ASPRINTF + fi +]) + +# Prerequisites of lib/vasprintf.h. +AC_DEFUN([gl_PREREQ_VASPRINTF_H], +[ + dnl Persuade glibc <stdio.h> to declare asprintf() and vasprintf(). + AC_REQUIRE([AC_GNU_SOURCE]) +]) + +# Prerequisites of lib/vasprintf.c. +AC_DEFUN([gl_PREREQ_VASPRINTF], +[ +]) + +# Prerequisites of lib/asprintf.c. +AC_DEFUN([gl_PREREQ_ASPRINTF], +[ +]) diff --git a/gl/m4/visibility.m4 b/gl/m4/visibility.m4 new file mode 100644 index 00000000..2ff6330a --- /dev/null +++ b/gl/m4/visibility.m4 @@ -0,0 +1,52 @@ +# visibility.m4 serial 1 (gettext-0.15) +dnl Copyright (C) 2005 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Bruno Haible. + +dnl Tests whether the compiler supports the command-line option +dnl -fvisibility=hidden and the function and variable attributes +dnl __attribute__((__visibility__("hidden"))) and +dnl __attribute__((__visibility__("default"))). +dnl Does *not* test for __visibility__("protected") - which has tricky +dnl semantics (see the 'vismain' test in glibc) and does not exist e.g. on +dnl MacOS X. +dnl Does *not* test for __visibility__("internal") - which has processor +dnl dependent semantics. +dnl Does *not* test for #pragma GCC visibility push(hidden) - which is +dnl "really only recommended for legacy code". +dnl Set the variable CFLAG_VISIBILITY. +dnl Defines and sets the variable HAVE_VISIBILITY. + +AC_DEFUN([gl_VISIBILITY], +[ + AC_REQUIRE([AC_PROG_CC]) + CFLAG_VISIBILITY= + HAVE_VISIBILITY=0 + if test -n "$GCC"; then + AC_MSG_CHECKING([for simple visibility declarations]) + AC_CACHE_VAL(gl_cv_cc_visibility, [ + gl_save_CFLAGS="$CFLAGS" + CFLAGS="$CFLAGS -fvisibility=hidden" + AC_TRY_COMPILE( + [extern __attribute__((__visibility__("hidden"))) int hiddenvar; + extern __attribute__((__visibility__("default"))) int exportedvar; + extern __attribute__((__visibility__("hidden"))) int hiddenfunc (void); + extern __attribute__((__visibility__("default"))) int exportedfunc (void);], + [], + gl_cv_cc_visibility=yes, + gl_cv_cc_visibility=no) + CFLAGS="$gl_save_CFLAGS"]) + AC_MSG_RESULT([$gl_cv_cc_visibility]) + if test $gl_cv_cc_visibility = yes; then + CFLAG_VISIBILITY="-fvisibility=hidden" + HAVE_VISIBILITY=1 + fi + fi + AC_SUBST([CFLAG_VISIBILITY]) + AC_SUBST([HAVE_VISIBILITY]) + AC_DEFINE_UNQUOTED([HAVE_VISIBILITY], [$HAVE_VISIBILITY], + [Define to 1 or 0, depending whether the compiler supports simple visibility declarations.]) +]) diff --git a/gl/m4/vsnprintf.m4 b/gl/m4/vsnprintf.m4 new file mode 100644 index 00000000..cb8a9b18 --- /dev/null +++ b/gl/m4/vsnprintf.m4 @@ -0,0 +1,15 @@ +# vsnprintf.m4 serial 2 +dnl Copyright (C) 2002, 2003, 2004 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_FUNC_VSNPRINTF], +[ + AC_REPLACE_FUNCS(vsnprintf) + AC_CHECK_DECLS_ONCE(vsnprintf) + gl_PREREQ_VSNPRINTF +]) + +# Prerequisites of lib/vsnprintf.c. +AC_DEFUN([gl_PREREQ_VSNPRINTF], [:]) diff --git a/gl/m4/wchar.m4 b/gl/m4/wchar.m4 new file mode 100644 index 00000000..068f22d3 --- /dev/null +++ b/gl/m4/wchar.m4 @@ -0,0 +1,29 @@ +dnl A placeholder for ISO C99 <wchar.h>, for platforms that have issues. + +dnl Copyright (C) 2007 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl Written by Eric Blake. + +# wchar.m4 serial 1 + +AC_DEFUN([gl_WCHAR_H], +[ + AC_CACHE_CHECK([whether <wchar.h> is standalone], + [gl_cv_header_wchar_h_standalone], + [AC_COMPILE_IFELSE([[#include <wchar.h> +wchar_t w;]], + [gl_cv_header_wchar_h_standalone=yes], + [gl_cv_header_wchar_h_standalone=no])]) + if test $gl_cv_header_wchar_h_standalone = yes; then + WCHAR_H= + else + gl_ABSOLUTE_HEADER([wchar.h]) + ABSOLUTE_WCHAR_H=\"$gl_cv_absolute_wchar_h\" + WCHAR_H=wchar.h + fi + AC_SUBST([ABSOLUTE_WCHAR_H]) + AC_SUBST([WCHAR_H]) +]) diff --git a/gl/m4/wchar_t.m4 b/gl/m4/wchar_t.m4 new file mode 100644 index 00000000..cde2129a --- /dev/null +++ b/gl/m4/wchar_t.m4 @@ -0,0 +1,20 @@ +# wchar_t.m4 serial 1 (gettext-0.12) +dnl Copyright (C) 2002-2003 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Bruno Haible. +dnl Test whether <stddef.h> has the 'wchar_t' type. +dnl Prerequisite: AC_PROG_CC + +AC_DEFUN([gt_TYPE_WCHAR_T], +[ + AC_CACHE_CHECK([for wchar_t], gt_cv_c_wchar_t, + [AC_TRY_COMPILE([#include <stddef.h> + wchar_t foo = (wchar_t)'\0';], , + gt_cv_c_wchar_t=yes, gt_cv_c_wchar_t=no)]) + if test $gt_cv_c_wchar_t = yes; then + AC_DEFINE(HAVE_WCHAR_T, 1, [Define if you have the 'wchar_t' type.]) + fi +]) diff --git a/gl/m4/wctype.m4 b/gl/m4/wctype.m4 new file mode 100644 index 00000000..62994c62 --- /dev/null +++ b/gl/m4/wctype.m4 @@ -0,0 +1,41 @@ +dnl A placeholder for ISO C99 <wctype.h>, for platforms that lack it. + +dnl Copyright (C) 2006, 2007 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl Written by Paul Eggert. + +AC_DEFUN([gl_WCTYPE_H], +[ + AC_CHECK_FUNCS_ONCE([iswcntrl]) + AC_CHECK_HEADERS_ONCE([wctype.h]) + AC_REQUIRE([AC_C_INLINE]) + + AC_REQUIRE([gt_TYPE_WINT_T]) + if test $gt_cv_c_wint_t = yes; then + HAVE_WINT_T=1 + else + HAVE_WINT_T=0 + fi + AC_SUBST([HAVE_WINT_T]) + + WCTYPE_H=wctype.h + if test $ac_cv_header_wctype_h = yes; then + if test "$ac_cv_func_iswcntrl" = yes; then + WCTYPE_H= + fi + dnl Compute ABSOLUTE_WCTYPE_H even if WCTYPE_H is empty, + dnl for the benefit of builds from non-distclean directories. + gl_ABSOLUTE_HEADER([wctype.h]) + ABSOLUTE_WCTYPE_H=\"$gl_cv_absolute_wctype_h\" + HAVE_WCTYPE_H=1 + else + ABSOLUTE_WCTYPE_H=\"no/such/file/wctype.h\" + HAVE_WCTYPE_H=0 + fi + AC_SUBST([ABSOLUTE_WCTYPE_H]) + AC_SUBST([HAVE_WCTYPE_H]) + AC_SUBST([WCTYPE_H]) +]) diff --git a/gl/m4/wcwidth.m4 b/gl/m4/wcwidth.m4 new file mode 100644 index 00000000..b4834991 --- /dev/null +++ b/gl/m4/wcwidth.m4 @@ -0,0 +1,30 @@ +# wcwidth.m4 serial 8 +dnl Copyright (C) 2006, 2007 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_FUNC_WCWIDTH], +[ + dnl Persuade glibc <wchar.h> to declare wcwidth(). + AC_REQUIRE([AC_GNU_SOURCE]) + + AC_REQUIRE([AC_C_INLINE]) + AC_REQUIRE([gt_TYPE_WCHAR_T]) + AC_REQUIRE([gt_TYPE_WINT_T]) + + AC_CHECK_HEADERS_ONCE([wchar.h]) + AC_CHECK_FUNCS_ONCE([wcwidth]) + + AC_CHECK_DECLS([wcwidth], [], [], [ +/* AIX 3.2.5 declares wcwidth in <string.h>. */ +#include <string.h> +/* Tru64 with Desktop Toolkit C has a bug: <stdio.h> must be included before + <wchar.h>. + BSD/OS 4.0.1 has a bug: <stddef.h>, <stdio.h> and <time.h> must be included + before <wchar.h>. */ +#include <stddef.h> +#include <stdio.h> +#include <time.h> +#include <wchar.h> +])]) diff --git a/gl/m4/wint_t.m4 b/gl/m4/wint_t.m4 new file mode 100644 index 00000000..3706c047 --- /dev/null +++ b/gl/m4/wint_t.m4 @@ -0,0 +1,28 @@ +# wint_t.m4 serial 2 (gettext-0.12) +dnl Copyright (C) 2003, 2007 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +dnl From Bruno Haible. +dnl Test whether <wchar.h> has the 'wint_t' type. +dnl Prerequisite: AC_PROG_CC + +AC_DEFUN([gt_TYPE_WINT_T], +[ + AC_CACHE_CHECK([for wint_t], gt_cv_c_wint_t, + [AC_TRY_COMPILE([ +/* Tru64 with Desktop Toolkit C has a bug: <stdio.h> must be included before + <wchar.h>. + BSD/OS 4.0.1 has a bug: <stddef.h>, <stdio.h> and <time.h> must be included + before <wchar.h>. */ +#include <stddef.h> +#include <stdio.h> +#include <time.h> +#include <wchar.h> + wint_t foo = (wchar_t)'\0';], , + gt_cv_c_wint_t=yes, gt_cv_c_wint_t=no)]) + if test $gt_cv_c_wint_t = yes; then + AC_DEFINE(HAVE_WINT_T, 1, [Define if you have the 'wint_t' type.]) + fi +]) diff --git a/gl/m4/xalloc.m4 b/gl/m4/xalloc.m4 new file mode 100644 index 00000000..837a948c --- /dev/null +++ b/gl/m4/xalloc.m4 @@ -0,0 +1,24 @@ +# xalloc.m4 serial 16 +dnl Copyright (C) 2002, 2003, 2004, 2005, 2006 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_XALLOC], +[ + AC_LIBOBJ([xmalloc]) + + gl_PREREQ_XALLOC + gl_PREREQ_XMALLOC +]) + +# Prerequisites of lib/xalloc.h. +AC_DEFUN([gl_PREREQ_XALLOC], [ + AC_REQUIRE([gl_INLINE]) + : +]) + +# Prerequisites of lib/xmalloc.c. +AC_DEFUN([gl_PREREQ_XMALLOC], [ + : +]) diff --git a/gl/m4/xsize.m4 b/gl/m4/xsize.m4 new file mode 100644 index 00000000..85bb721e --- /dev/null +++ b/gl/m4/xsize.m4 @@ -0,0 +1,13 @@ +# xsize.m4 serial 3 +dnl Copyright (C) 2003-2004 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_XSIZE], +[ + dnl Prerequisites of lib/xsize.h. + AC_REQUIRE([gl_SIZE_MAX]) + AC_REQUIRE([AC_C_INLINE]) + AC_CHECK_HEADERS(stdint.h) +]) diff --git a/gl/m4/xstrndup.m4 b/gl/m4/xstrndup.m4 new file mode 100644 index 00000000..8a30ab15 --- /dev/null +++ b/gl/m4/xstrndup.m4 @@ -0,0 +1,15 @@ +# xstrndup.m4 serial 2 +dnl Copyright (C) 2003 Free Software Foundation, Inc. +dnl This file is free software; the Free Software Foundation +dnl gives unlimited permission to copy and/or distribute it, +dnl with or without modifications, as long as this notice is preserved. + +AC_DEFUN([gl_XSTRNDUP], +[ + gl_PREREQ_XSTRNDUP +]) + +# Prerequisites of lib/xstrndup.c. +AC_DEFUN([gl_PREREQ_XSTRNDUP], [ + : +]) diff --git a/gl/malloc.c b/gl/malloc.c new file mode 100644 index 00000000..d4dae3e1 --- /dev/null +++ b/gl/malloc.c @@ -0,0 +1,35 @@ +/* malloc() function that is glibc compatible. + + Copyright (C) 1997, 1998, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* written by Jim Meyering */ + +#include <config.h> +#undef malloc + +#include <stdlib.h> + +/* Allocate an N-byte block of memory from the heap. + If N is zero, allocate a 1-byte block. */ + +void * +rpl_malloc (size_t n) +{ + if (n == 0) + n = 1; + return malloc (n); +} diff --git a/gl/mbchar.c b/gl/mbchar.c new file mode 100644 index 00000000..95373f58 --- /dev/null +++ b/gl/mbchar.c @@ -0,0 +1,36 @@ +/* Copyright (C) 2001, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + + +#include <config.h> + +#include <limits.h> + +#include "mbchar.h" + +#if IS_BASIC_ASCII + +/* Bit table of characters in the ISO C "basic character set". */ +const unsigned int is_basic_table [UCHAR_MAX / 32 + 1] = +{ + 0x00001a00, /* '\t' '\v' '\f' */ + 0xffffffef, /* ' '...'#' '%'...'?' */ + 0xfffffffe, /* 'A'...'Z' '[' '\\' ']' '^' '_' */ + 0x7ffffffe /* 'a'...'z' '{' '|' '}' '~' */ + /* The remaining bits are 0. */ +}; + +#endif /* IS_BASIC_ASCII */ diff --git a/gl/mbchar.h b/gl/mbchar.h new file mode 100644 index 00000000..f3e28ef5 --- /dev/null +++ b/gl/mbchar.h @@ -0,0 +1,353 @@ +/* Multibyte character data type. + Copyright (C) 2001, 2005-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Bruno Haible <bruno@clisp.org>. */ + +/* A multibyte character is a short subsequence of a char* string, + representing a single wide character. + + We use multibyte characters instead of wide characters because of + the following goals: + 1) correct multibyte handling, i.e. operate according to the LC_CTYPE + locale, + 2) ease of maintenance, i.e. the maintainer needs not know all details + of the ISO C 99 standard, + 3) don't fail grossly if the input is not in the encoding set by the + locale, because often different encodings are in use in the same + countries (ISO-8859-1/UTF-8, EUC-JP/Shift_JIS, ...), + 4) fast in the case of ASCII characters, + 5) portability, i.e. don't make unportable assumptions about wchar_t. + + Multibyte characters are only accessed through the mb* macros. + + mb_ptr (mbc) + return a pointer to the beginning of the multibyte sequence. + + mb_len (mbc) + returns the number of bytes occupied by the multibyte sequence. + Always > 0. + + mb_iseq (mbc, sc) + returns true if mbc is the standard ASCII character sc. + + mb_isnul (mbc) + returns true if mbc is the nul character. + + mb_cmp (mbc1, mbc2) + returns a positive, zero, or negative value depending on whether mbc1 + sorts after, same or before mbc2. + + mb_casecmp (mbc1, mbc2) + returns a positive, zero, or negative value depending on whether mbc1 + sorts after, same or before mbc2, modulo upper/lowercase conversion. + + mb_equal (mbc1, mbc2) + returns true if mbc1 and mbc2 are equal. + + mb_caseequal (mbc1, mbc2) + returns true if mbc1 and mbc2 are equal modulo upper/lowercase conversion. + + mb_isalnum (mbc) + returns true if mbc is alphanumeric. + + mb_isalpha (mbc) + returns true if mbc is alphabetic. + + mb_isascii(mbc) + returns true if mbc is plain ASCII. + + mb_isblank (mbc) + returns true if mbc is a blank. + + mb_iscntrl (mbc) + returns true if mbc is a control character. + + mb_isdigit (mbc) + returns true if mbc is a decimal digit. + + mb_isgraph (mbc) + returns true if mbc is a graphic character. + + mb_islower (mbc) + returns true if mbc is lowercase. + + mb_isprint (mbc) + returns true if mbc is a printable character. + + mb_ispunct (mbc) + returns true if mbc is a punctuation character. + + mb_isspace (mbc) + returns true if mbc is a space character. + + mb_isupper (mbc) + returns true if mbc is uppercase. + + mb_isxdigit (mbc) + returns true if mbc is a hexadecimal digit. + + mb_width (mbc) + returns the number of columns on the output device occupied by mbc. + Always >= 0. + + mb_putc (mbc, stream) + outputs mbc on stream, a byte oriented FILE stream opened for output. + + mb_setascii (&mbc, sc) + assigns the standard ASCII character sc to mbc. + + mb_copy (&destmbc, &srcmbc) + copies srcmbc to destmbc. + + Here are the function prototypes of the macros. + + extern const char * mb_ptr (const mbchar_t mbc); + extern size_t mb_len (const mbchar_t mbc); + extern bool mb_iseq (const mbchar_t mbc, char sc); + extern bool mb_isnul (const mbchar_t mbc); + extern int mb_cmp (const mbchar_t mbc1, const mbchar_t mbc2); + extern int mb_casecmp (const mbchar_t mbc1, const mbchar_t mbc2); + extern bool mb_equal (const mbchar_t mbc1, const mbchar_t mbc2); + extern bool mb_caseequal (const mbchar_t mbc1, const mbchar_t mbc2); + extern bool mb_isalnum (const mbchar_t mbc); + extern bool mb_isalpha (const mbchar_t mbc); + extern bool mb_isascii (const mbchar_t mbc); + extern bool mb_isblank (const mbchar_t mbc); + extern bool mb_iscntrl (const mbchar_t mbc); + extern bool mb_isdigit (const mbchar_t mbc); + extern bool mb_isgraph (const mbchar_t mbc); + extern bool mb_islower (const mbchar_t mbc); + extern bool mb_isprint (const mbchar_t mbc); + extern bool mb_ispunct (const mbchar_t mbc); + extern bool mb_isspace (const mbchar_t mbc); + extern bool mb_isupper (const mbchar_t mbc); + extern bool mb_isxdigit (const mbchar_t mbc); + extern int mb_width (const mbchar_t mbc); + extern void mb_putc (const mbchar_t mbc, FILE *stream); + extern void mb_setascii (mbchar_t *new, char sc); + extern void mb_copy (mbchar_t *new, const mbchar_t *old); + */ + +#ifndef _MBCHAR_H +#define _MBCHAR_H 1 + +#include <stdbool.h> +#include <string.h> + +/* Tru64 with Desktop Toolkit C has a bug: <stdio.h> must be included before + <wchar.h>. + BSD/OS 4.1 has a bug: <stdio.h> and <time.h> must be included before + <wchar.h>. */ +#include <stdio.h> +#include <time.h> +#include <wchar.h> +#include <wctype.h> + +#include "wcwidth.h" + +#define MBCHAR_BUF_SIZE 24 + +struct mbchar +{ + const char *ptr; /* pointer to current character */ + size_t bytes; /* number of bytes of current character, > 0 */ + bool wc_valid; /* true if wc is a valid wide character */ + wchar_t wc; /* if wc_valid: the current character */ + char buf[MBCHAR_BUF_SIZE]; /* room for the bytes, used for file input only */ +}; + +/* EOF (not a real character) is represented with bytes = 0 and + wc_valid = false. */ + +typedef struct mbchar mbchar_t; + +/* Access the current character. */ +#define mb_ptr(mbc) ((mbc).ptr) +#define mb_len(mbc) ((mbc).bytes) + +/* Comparison of characters. */ +#define mb_iseq(mbc, sc) ((mbc).wc_valid && (mbc).wc == (sc)) +#define mb_isnul(mbc) ((mbc).wc_valid && (mbc).wc == 0) +#define mb_cmp(mbc1, mbc2) \ + ((mbc1).wc_valid \ + ? ((mbc2).wc_valid \ + ? (int) (mbc1).wc - (int) (mbc2).wc \ + : -1) \ + : ((mbc2).wc_valid \ + ? 1 \ + : (mbc1).bytes == (mbc2).bytes \ + ? memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) \ + : (mbc1).bytes < (mbc2).bytes \ + ? (memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) > 0 ? 1 : -1) \ + : (memcmp ((mbc1).ptr, (mbc2).ptr, (mbc2).bytes) >= 0 ? 1 : -1))) +#define mb_casecmp(mbc1, mbc2) \ + ((mbc1).wc_valid \ + ? ((mbc2).wc_valid \ + ? (int) towlower ((mbc1).wc) - (int) towlower ((mbc2).wc) \ + : -1) \ + : ((mbc2).wc_valid \ + ? 1 \ + : (mbc1).bytes == (mbc2).bytes \ + ? memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) \ + : (mbc1).bytes < (mbc2).bytes \ + ? (memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) > 0 ? 1 : -1) \ + : (memcmp ((mbc1).ptr, (mbc2).ptr, (mbc2).bytes) >= 0 ? 1 : -1))) +#define mb_equal(mbc1, mbc2) \ + ((mbc1).wc_valid && (mbc2).wc_valid \ + ? (mbc1).wc == (mbc2).wc \ + : (mbc1).bytes == (mbc2).bytes \ + && memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) == 0) +#define mb_caseequal(mbc1, mbc2) \ + ((mbc1).wc_valid && (mbc2).wc_valid \ + ? towlower ((mbc1).wc) == towlower ((mbc2).wc) \ + : (mbc1).bytes == (mbc2).bytes \ + && memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) == 0) + +/* <ctype.h>, <wctype.h> classification. */ +#define mb_isascii(mbc) \ + ((mbc).wc_valid && (mbc).wc >= 0 && (mbc).wc <= 127) +#define mb_isalnum(mbc) ((mbc).wc_valid && iswalnum ((mbc).wc)) +#define mb_isalpha(mbc) ((mbc).wc_valid && iswalpha ((mbc).wc)) +#define mb_isblank(mbc) ((mbc).wc_valid && iswblank ((mbc).wc)) +#define mb_iscntrl(mbc) ((mbc).wc_valid && iswcntrl ((mbc).wc)) +#define mb_isdigit(mbc) ((mbc).wc_valid && iswdigit ((mbc).wc)) +#define mb_isgraph(mbc) ((mbc).wc_valid && iswgraph ((mbc).wc)) +#define mb_islower(mbc) ((mbc).wc_valid && iswlower ((mbc).wc)) +#define mb_isprint(mbc) ((mbc).wc_valid && iswprint ((mbc).wc)) +#define mb_ispunct(mbc) ((mbc).wc_valid && iswpunct ((mbc).wc)) +#define mb_isspace(mbc) ((mbc).wc_valid && iswspace ((mbc).wc)) +#define mb_isupper(mbc) ((mbc).wc_valid && iswupper ((mbc).wc)) +#define mb_isxdigit(mbc) ((mbc).wc_valid && iswxdigit ((mbc).wc)) + +/* Extra <wchar.h> function. */ + +/* Unprintable characters appear as a small box of width 1. */ +#define MB_UNPRINTABLE_WIDTH 1 + +static inline int +mb_width_aux (wint_t wc) +{ + int w = wcwidth (wc); + /* For unprintable characters, arbitrarily return 0 for control characters + and MB_UNPRINTABLE_WIDTH otherwise. */ + return (w >= 0 ? w : iswcntrl (wc) ? 0 : MB_UNPRINTABLE_WIDTH); +} + +#define mb_width(mbc) \ + ((mbc).wc_valid ? mb_width_aux ((mbc).wc) : MB_UNPRINTABLE_WIDTH) + +/* Output. */ +#define mb_putc(mbc, stream) fwrite ((mbc).ptr, 1, (mbc).bytes, (stream)) + +/* Assignment. */ +#define mb_setascii(mbc, sc) \ + ((mbc)->ptr = (mbc)->buf, (mbc)->bytes = 1, (mbc)->wc_valid = 1, \ + (mbc)->wc = (mbc)->buf[0] = (sc)) + +/* Copying a character. */ +static inline void +mb_copy (mbchar_t *new_mbc, const mbchar_t *old_mbc) +{ + if (old_mbc->ptr == &old_mbc->buf[0]) + { + memcpy (&new_mbc->buf[0], &old_mbc->buf[0], old_mbc->bytes); + new_mbc->ptr = &new_mbc->buf[0]; + } + else + new_mbc->ptr = old_mbc->ptr; + new_mbc->bytes = old_mbc->bytes; + if ((new_mbc->wc_valid = old_mbc->wc_valid)) + new_mbc->wc = old_mbc->wc; +} + + +/* is_basic(c) tests whether the single-byte character c is in the + ISO C "basic character set". + This is a convenience function, and is in this file only to share code + between mbiter_multi.h and mbfile_multi.h. */ +#if (' ' == 32) && ('!' == 33) && ('"' == 34) && ('#' == 35) \ + && ('%' == 37) && ('&' == 38) && ('\'' == 39) && ('(' == 40) \ + && (')' == 41) && ('*' == 42) && ('+' == 43) && (',' == 44) \ + && ('-' == 45) && ('.' == 46) && ('/' == 47) && ('0' == 48) \ + && ('1' == 49) && ('2' == 50) && ('3' == 51) && ('4' == 52) \ + && ('5' == 53) && ('6' == 54) && ('7' == 55) && ('8' == 56) \ + && ('9' == 57) && (':' == 58) && (';' == 59) && ('<' == 60) \ + && ('=' == 61) && ('>' == 62) && ('?' == 63) && ('A' == 65) \ + && ('B' == 66) && ('C' == 67) && ('D' == 68) && ('E' == 69) \ + && ('F' == 70) && ('G' == 71) && ('H' == 72) && ('I' == 73) \ + && ('J' == 74) && ('K' == 75) && ('L' == 76) && ('M' == 77) \ + && ('N' == 78) && ('O' == 79) && ('P' == 80) && ('Q' == 81) \ + && ('R' == 82) && ('S' == 83) && ('T' == 84) && ('U' == 85) \ + && ('V' == 86) && ('W' == 87) && ('X' == 88) && ('Y' == 89) \ + && ('Z' == 90) && ('[' == 91) && ('\\' == 92) && (']' == 93) \ + && ('^' == 94) && ('_' == 95) && ('a' == 97) && ('b' == 98) \ + && ('c' == 99) && ('d' == 100) && ('e' == 101) && ('f' == 102) \ + && ('g' == 103) && ('h' == 104) && ('i' == 105) && ('j' == 106) \ + && ('k' == 107) && ('l' == 108) && ('m' == 109) && ('n' == 110) \ + && ('o' == 111) && ('p' == 112) && ('q' == 113) && ('r' == 114) \ + && ('s' == 115) && ('t' == 116) && ('u' == 117) && ('v' == 118) \ + && ('w' == 119) && ('x' == 120) && ('y' == 121) && ('z' == 122) \ + && ('{' == 123) && ('|' == 124) && ('}' == 125) && ('~' == 126) +/* The character set is ISO-646, not EBCDIC. */ +# define IS_BASIC_ASCII 1 + +extern const unsigned int is_basic_table[]; + +static inline bool +is_basic (char c) +{ + return (is_basic_table [(unsigned char) c >> 5] >> ((unsigned char) c & 31)) + & 1; +} + +#else + +static inline bool +is_basic (char c) +{ + switch (c) + { + case '\t': case '\v': case '\f': + case ' ': case '!': case '"': case '#': case '%': + case '&': case '\'': case '(': case ')': case '*': + case '+': case ',': case '-': case '.': case '/': + case '0': case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': + case ':': case ';': case '<': case '=': case '>': + case '?': + case 'A': case 'B': case 'C': case 'D': case 'E': + case 'F': case 'G': case 'H': case 'I': case 'J': + case 'K': case 'L': case 'M': case 'N': case 'O': + case 'P': case 'Q': case 'R': case 'S': case 'T': + case 'U': case 'V': case 'W': case 'X': case 'Y': + case 'Z': + case '[': case '\\': case ']': case '^': case '_': + case 'a': case 'b': case 'c': case 'd': case 'e': + case 'f': case 'g': case 'h': case 'i': case 'j': + case 'k': case 'l': case 'm': case 'n': case 'o': + case 'p': case 'q': case 'r': case 's': case 't': + case 'u': case 'v': case 'w': case 'x': case 'y': + case 'z': case '{': case '|': case '}': case '~': + return 1; + default: + return 0; + } +} + +#endif + +#endif /* _MBCHAR_H */ diff --git a/gl/mbuiter.h b/gl/mbuiter.h new file mode 100644 index 00000000..9da3a6c7 --- /dev/null +++ b/gl/mbuiter.h @@ -0,0 +1,203 @@ +/* Iterating through multibyte strings: macros for multi-byte encodings. + Copyright (C) 2001, 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Bruno Haible <bruno@clisp.org>. */ + +/* The macros in this file implement forward iteration through a + multi-byte string, without knowing its length a-priori. + + With these macros, an iteration loop that looks like + + char *iter; + for (iter = buf; *iter != '\0'; iter++) + { + do_something (*iter); + } + + becomes + + mbui_iterator_t iter; + for (mbui_init (iter, buf); mbui_avail (iter); mbui_advance (iter)) + { + do_something (mbui_cur_ptr (iter), mb_len (mbui_cur (iter))); + } + + The benefit of these macros over plain use of mbrtowc is: + - Handling of invalid multibyte sequences is possible without + making the code more complicated, while still preserving the + invalid multibyte sequences. + + Compared to mbiter.h, the macros here don't need to know the string's + length a-priori. The downside is that at each step, the look-ahead + that guards against overrunning the terminating '\0' is more expensive. + The mbui_* macros are therefore suitable when there is a high probability + that only the first few multibyte characters need to be inspected. + Whereas the mbi_* macros are better if usually the iteration runs + through the entire string. + + mbui_iterator_t + is a type usable for variable declarations. + + mbui_init (iter, startptr) + initializes the iterator, starting at startptr. + + mbui_avail (iter) + returns true if there are more multibyte chracters available before + the end of string is reached. In this case, mbui_cur (iter) is + initialized to the next multibyte chracter. + + mbui_advance (iter) + advances the iterator by one multibyte character. + + mbui_cur (iter) + returns the current multibyte character, of type mbchar_t. All the + macros defined in mbchar.h can be used on it. + + mbui_cur_ptr (iter) + return a pointer to the beginning of the current multibyte character. + + mbui_reloc (iter, ptrdiff) + relocates iterator when the string is moved by ptrdiff bytes. + + Here are the function prototypes of the macros. + + extern void mbui_init (mbui_iterator_t iter, const char *startptr); + extern bool mbui_avail (mbui_iterator_t iter); + extern void mbui_advance (mbui_iterator_t iter); + extern mbchar_t mbui_cur (mbui_iterator_t iter); + extern const char * mbui_cur_ptr (mbui_iterator_t iter); + extern void mbui_reloc (mbui_iterator_t iter, ptrdiff_t ptrdiff); + */ + +#ifndef _MBUITER_H +#define _MBUITER_H 1 + +#include <assert.h> +#include <stdbool.h> +#include <stdlib.h> + +/* Tru64 with Desktop Toolkit C has a bug: <stdio.h> must be included before + <wchar.h>. + BSD/OS 4.1 has a bug: <stdio.h> and <time.h> must be included before + <wchar.h>. */ +#include <stdio.h> +#include <time.h> +#include <wchar.h> + +#include "mbchar.h" +#include "strnlen1.h" + +struct mbuiter_multi +{ + bool in_shift; /* true if next byte may not be interpreted as ASCII */ + mbstate_t state; /* if in_shift: current shift state */ + bool next_done; /* true if mbui_avail has already filled the following */ + struct mbchar cur; /* the current character: + const char *cur.ptr pointer to current character + The following are only valid after mbui_avail. + size_t cur.bytes number of bytes of current character + bool cur.wc_valid true if wc is a valid wide character + wchar_t cur.wc if wc_valid: the current character + */ +}; + +static inline void +mbuiter_multi_next (struct mbuiter_multi *iter) +{ + if (iter->next_done) + return; + if (iter->in_shift) + goto with_shift; + /* Handle most ASCII characters quickly, without calling mbrtowc(). */ + if (is_basic (*iter->cur.ptr)) + { + /* These characters are part of the basic character set. ISO C 99 + guarantees that their wide character code is identical to their + char code. */ + iter->cur.bytes = 1; + iter->cur.wc = *iter->cur.ptr; + iter->cur.wc_valid = true; + } + else + { + assert (mbsinit (&iter->state)); + iter->in_shift = true; + with_shift: + iter->cur.bytes = mbrtowc (&iter->cur.wc, iter->cur.ptr, + strnlen1 (iter->cur.ptr, MB_CUR_MAX), + &iter->state); + if (iter->cur.bytes == (size_t) -1) + { + /* An invalid multibyte sequence was encountered. */ + iter->cur.bytes = 1; + iter->cur.wc_valid = false; + /* Whether to set iter->in_shift = false and reset iter->state + or not is not very important; the string is bogus anyway. */ + } + else if (iter->cur.bytes == (size_t) -2) + { + /* An incomplete multibyte character at the end. */ + iter->cur.bytes = strlen (iter->cur.ptr); + iter->cur.wc_valid = false; + /* Whether to set iter->in_shift = false and reset iter->state + or not is not important; the string end is reached anyway. */ + } + else + { + if (iter->cur.bytes == 0) + { + /* A null wide character was encountered. */ + iter->cur.bytes = 1; + assert (*iter->cur.ptr == '\0'); + assert (iter->cur.wc == 0); + } + iter->cur.wc_valid = true; + + /* When in the initial state, we can go back treating ASCII + characters more quickly. */ + if (mbsinit (&iter->state)) + iter->in_shift = false; + } + } + iter->next_done = true; +} + +static inline void +mbuiter_multi_reloc (struct mbuiter_multi *iter, ptrdiff_t ptrdiff) +{ + iter->cur.ptr += ptrdiff; +} + +/* Iteration macros. */ +typedef struct mbuiter_multi mbui_iterator_t; +#define mbui_init(iter, startptr) \ + ((iter).cur.ptr = (startptr), \ + (iter).in_shift = false, memset (&(iter).state, '\0', sizeof (mbstate_t)), \ + (iter).next_done = false) +#define mbui_avail(iter) \ + (mbuiter_multi_next (&(iter)), !mb_isnul ((iter).cur)) +#define mbui_advance(iter) \ + ((iter).cur.ptr += (iter).cur.bytes, (iter).next_done = false) + +/* Access to the current character. */ +#define mbui_cur(iter) (iter).cur +#define mbui_cur_ptr(iter) (iter).cur.ptr + +/* Relocation. */ +#define mbui_reloc(iter, ptrdiff) mbuiter_multi_reloc (&iter, ptrdiff) + +#endif /* _MBUITER_H */ diff --git a/gl/memchr.c b/gl/memchr.c new file mode 100644 index 00000000..d44ad6de --- /dev/null +++ b/gl/memchr.c @@ -0,0 +1,201 @@ +/* Copyright (C) 1991, 1993, 1996, 1997, 1999, 2000, 2003, 2004, 2006 Free + Software Foundation, Inc. + + Based on strlen implementation by Torbjorn Granlund (tege@sics.se), + with help from Dan Sahlin (dan@sics.se) and + commentary by Jim Blandy (jimb@ai.mit.edu); + adaptation to memchr suggested by Dick Karpinski (dick@cca.ucsf.edu), + and implemented by Roland McGrath (roland@ai.mit.edu). + +NOTE: The canonical source of this file is maintained with the GNU C Library. +Bugs can be reported to bug-glibc@prep.ai.mit.edu. + +This program is free software; you can redistribute it and/or modify it +under the terms of the GNU General Public License as published by the +Free Software Foundation; either version 2, or (at your option) any +later version. + +This program is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +GNU General Public License for more details. + +You should have received a copy of the GNU General Public License +along with this program; if not, write to the Free Software Foundation, +Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _LIBC +# include <config.h> +#endif + +#include <string.h> + +#include <stddef.h> + +#if defined _LIBC +# include <memcopy.h> +#else +# define reg_char char +#endif + +#include <limits.h> + +#if HAVE_BP_SYM_H || defined _LIBC +# include <bp-sym.h> +#else +# define BP_SYM(sym) sym +#endif + +#undef memchr +#undef __memchr + +/* Search no more than N bytes of S for C. */ +void * +__memchr (void const *s, int c_in, size_t n) +{ + const unsigned char *char_ptr; + const unsigned long int *longword_ptr; + unsigned long int longword, magic_bits, charmask; + unsigned reg_char c; + int i; + + c = (unsigned char) c_in; + + /* Handle the first few characters by reading one character at a time. + Do this until CHAR_PTR is aligned on a longword boundary. */ + for (char_ptr = (const unsigned char *) s; + n > 0 && (size_t) char_ptr % sizeof longword != 0; + --n, ++char_ptr) + if (*char_ptr == c) + return (void *) char_ptr; + + /* All these elucidatory comments refer to 4-byte longwords, + but the theory applies equally well to any size longwords. */ + + longword_ptr = (const unsigned long int *) char_ptr; + + /* Bits 31, 24, 16, and 8 of this number are zero. Call these bits + the "holes." Note that there is a hole just to the left of + each byte, with an extra at the end: + + bits: 01111110 11111110 11111110 11111111 + bytes: AAAAAAAA BBBBBBBB CCCCCCCC DDDDDDDD + + The 1-bits make sure that carries propagate to the next 0-bit. + The 0-bits provide holes for carries to fall into. */ + + /* Set MAGIC_BITS to be this pattern of 1 and 0 bits. + Set CHARMASK to be a longword, each of whose bytes is C. */ + + magic_bits = 0xfefefefe; + charmask = c | (c << 8); + charmask |= charmask << 16; +#if 0xffffffffU < ULONG_MAX + magic_bits |= magic_bits << 32; + charmask |= charmask << 32; + if (8 < sizeof longword) + for (i = 64; i < sizeof longword * 8; i *= 2) + { + magic_bits |= magic_bits << i; + charmask |= charmask << i; + } +#endif + magic_bits = (ULONG_MAX >> 1) & (magic_bits | 1); + + /* Instead of the traditional loop which tests each character, + we will test a longword at a time. The tricky part is testing + if *any of the four* bytes in the longword in question are zero. */ + while (n >= sizeof longword) + { + /* We tentatively exit the loop if adding MAGIC_BITS to + LONGWORD fails to change any of the hole bits of LONGWORD. + + 1) Is this safe? Will it catch all the zero bytes? + Suppose there is a byte with all zeros. Any carry bits + propagating from its left will fall into the hole at its + least significant bit and stop. Since there will be no + carry from its most significant bit, the LSB of the + byte to the left will be unchanged, and the zero will be + detected. + + 2) Is this worthwhile? Will it ignore everything except + zero bytes? Suppose every byte of LONGWORD has a bit set + somewhere. There will be a carry into bit 8. If bit 8 + is set, this will carry into bit 16. If bit 8 is clear, + one of bits 9-15 must be set, so there will be a carry + into bit 16. Similarly, there will be a carry into bit + 24. If one of bits 24-30 is set, there will be a carry + into bit 31, so all of the hole bits will be changed. + + The one misfire occurs when bits 24-30 are clear and bit + 31 is set; in this case, the hole at bit 31 is not + changed. If we had access to the processor carry flag, + we could close this loophole by putting the fourth hole + at bit 32! + + So it ignores everything except 128's, when they're aligned + properly. + + 3) But wait! Aren't we looking for C, not zero? + Good point. So what we do is XOR LONGWORD with a longword, + each of whose bytes is C. This turns each byte that is C + into a zero. */ + + longword = *longword_ptr++ ^ charmask; + + /* Add MAGIC_BITS to LONGWORD. */ + if ((((longword + magic_bits) + + /* Set those bits that were unchanged by the addition. */ + ^ ~longword) + + /* Look at only the hole bits. If any of the hole bits + are unchanged, most likely one of the bytes was a + zero. */ + & ~magic_bits) != 0) + { + /* Which of the bytes was C? If none of them were, it was + a misfire; continue the search. */ + + const unsigned char *cp = (const unsigned char *) (longword_ptr - 1); + + if (cp[0] == c) + return (void *) cp; + if (cp[1] == c) + return (void *) &cp[1]; + if (cp[2] == c) + return (void *) &cp[2]; + if (cp[3] == c) + return (void *) &cp[3]; + if (4 < sizeof longword && cp[4] == c) + return (void *) &cp[4]; + if (5 < sizeof longword && cp[5] == c) + return (void *) &cp[5]; + if (6 < sizeof longword && cp[6] == c) + return (void *) &cp[6]; + if (7 < sizeof longword && cp[7] == c) + return (void *) &cp[7]; + if (8 < sizeof longword) + for (i = 8; i < sizeof longword; i++) + if (cp[i] == c) + return (void *) &cp[i]; + } + + n -= sizeof longword; + } + + char_ptr = (const unsigned char *) longword_ptr; + + while (n-- > 0) + { + if (*char_ptr == c) + return (void *) char_ptr; + else + ++char_ptr; + } + + return 0; +} +#ifdef weak_alias +weak_alias (__memchr, BP_SYM (memchr)) +#endif diff --git a/gl/minmax.h b/gl/minmax.h new file mode 100644 index 00000000..975ea76d --- /dev/null +++ b/gl/minmax.h @@ -0,0 +1,60 @@ +/* MIN, MAX macros. + Copyright (C) 1995, 1998, 2001, 2003, 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _MINMAX_H +#define _MINMAX_H + +/* Note: MIN, MAX are also defined in <sys/param.h> on some systems + (glibc, IRIX, HP-UX, OSF/1). Therefore you might get warnings about + MIN, MAX macro redefinitions on some systems; the workaround is to + #include this file as the last one among the #include list. */ + +/* Before we define the following symbols we get the <limits.h> file + since otherwise we get redefinitions on some systems if <limits.h> is + included after this file. Likewise for <sys/param.h>. + If more than one of these system headers define MIN and MAX, pick just + one of the headers (because the definitions most likely are the same). */ +#if HAVE_MINMAX_IN_LIMITS_H +# include <limits.h> +#elif HAVE_MINMAX_IN_SYS_PARAM_H +# include <sys/param.h> +#endif + +/* Note: MIN and MAX should be used with two arguments of the + same type. They might not return the minimum and maximum of their two + arguments, if the arguments have different types or have unusual + floating-point values. For example, on a typical host with 32-bit 'int', + 64-bit 'long long', and 64-bit IEEE 754 'double' types: + + MAX (-1, 2147483648) returns 4294967295. + MAX (9007199254740992.0, 9007199254740993) returns 9007199254740992.0. + MAX (NaN, 0.0) returns 0.0. + MAX (+0.0, -0.0) returns -0.0. + + and in each case the answer is in some sense bogus. */ + +/* MAX(a,b) returns the maximum of A and B. */ +#ifndef MAX +# define MAX(a,b) ((a) > (b) ? (a) : (b)) +#endif + +/* MIN(a,b) returns the minimum of A and B. */ +#ifndef MIN +# define MIN(a,b) ((a) < (b) ? (a) : (b)) +#endif + +#endif /* _MINMAX_H */ diff --git a/gl/mountlist.c b/gl/mountlist.c new file mode 100644 index 00000000..bb01f91e --- /dev/null +++ b/gl/mountlist.c @@ -0,0 +1,889 @@ +/* mountlist.c -- return a list of mounted file systems + + Copyright (C) 1991, 1992, 1997, 1998, 1999, 2000, 2001, 2002, 2003, + 2004, 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "mountlist.h" + +#include <limits.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +#include "xalloc.h" + +#ifndef strstr +char *strstr (); +#endif + +#include <errno.h> + +#include <fcntl.h> + +#include <unistd.h> + +#if HAVE_SYS_PARAM_H +# include <sys/param.h> +#endif + +#if defined MOUNTED_GETFSSTAT /* OSF_1 and Darwin1.3.x */ +# if HAVE_SYS_UCRED_H +# include <grp.h> /* needed on OSF V4.0 for definition of NGROUPS, + NGROUPS is used as an array dimension in ucred.h */ +# include <sys/ucred.h> /* needed by powerpc-apple-darwin1.3.7 */ +# endif +# if HAVE_SYS_MOUNT_H +# include <sys/mount.h> +# endif +# if HAVE_SYS_FS_TYPES_H +# include <sys/fs_types.h> /* needed by powerpc-apple-darwin1.3.7 */ +# endif +# if HAVE_STRUCT_FSSTAT_F_FSTYPENAME +# define FS_TYPE(Ent) ((Ent).f_fstypename) +# else +# define FS_TYPE(Ent) mnt_names[(Ent).f_type] +# endif +#endif /* MOUNTED_GETFSSTAT */ + +#ifdef MOUNTED_GETMNTENT1 /* 4.3BSD, SunOS, HP-UX, Dynix, Irix. */ +# include <mntent.h> +# if !defined MOUNTED +# if defined _PATH_MOUNTED /* GNU libc */ +# define MOUNTED _PATH_MOUNTED +# endif +# if defined MNT_MNTTAB /* HP-UX. */ +# define MOUNTED MNT_MNTTAB +# endif +# if defined MNTTABNAME /* Dynix. */ +# define MOUNTED MNTTABNAME +# endif +# endif +#endif + +#ifdef MOUNTED_GETMNTINFO /* 4.4BSD. */ +# include <sys/mount.h> +#endif + +#ifdef MOUNTED_GETMNTINFO2 /* NetBSD 3.0. */ +# include <sys/statvfs.h> +#endif + +#ifdef MOUNTED_GETMNT /* Ultrix. */ +# include <sys/mount.h> +# include <sys/fs_types.h> +#endif + +#ifdef MOUNTED_FS_STAT_DEV /* BeOS. */ +# include <fs_info.h> +# include <dirent.h> +#endif + +#ifdef MOUNTED_FREAD /* SVR2. */ +# include <mnttab.h> +#endif + +#ifdef MOUNTED_FREAD_FSTYP /* SVR3. */ +# include <mnttab.h> +# include <sys/fstyp.h> +# include <sys/statfs.h> +#endif + +#ifdef MOUNTED_LISTMNTENT +# include <mntent.h> +#endif + +#ifdef MOUNTED_GETMNTENT2 /* SVR4. */ +# include <sys/mnttab.h> +#endif + +#ifdef MOUNTED_VMOUNT /* AIX. */ +# include <fshelp.h> +# include <sys/vfs.h> +#endif + +#ifdef DOLPHIN +/* So special that it's not worth putting this in autoconf. */ +# undef MOUNTED_FREAD_FSTYP +# define MOUNTED_GETMNTTBL +#endif + +#if HAVE_SYS_MNTENT_H +/* This is to get MNTOPT_IGNORE on e.g. SVR4. */ +# include <sys/mntent.h> +#endif + +#undef MNT_IGNORE +#if defined MNTOPT_IGNORE && defined HAVE_HASMNTOPT +# define MNT_IGNORE(M) hasmntopt ((M), MNTOPT_IGNORE) +#else +# define MNT_IGNORE(M) 0 +#endif + +#if USE_UNLOCKED_IO +# include "unlocked-io.h" +#endif + +#ifndef SIZE_MAX +# define SIZE_MAX ((size_t) -1) +#endif + +/* The results of open() in this file are not used with fchdir, + therefore save some unnecessary work in fchdir.c. */ +#undef open +#undef close + +/* The results of opendir() in this file are not used with dirfd and fchdir, + therefore save some unnecessary work in fchdir.c. */ +#undef opendir +#undef closedir + +#ifndef ME_DUMMY +# define ME_DUMMY(Fs_name, Fs_type) \ + (strcmp (Fs_type, "autofs") == 0 \ + || strcmp (Fs_type, "none") == 0 \ + || strcmp (Fs_type, "proc") == 0 \ + || strcmp (Fs_type, "subfs") == 0 \ + /* for NetBSD 3.0 */ \ + || strcmp (Fs_type, "kernfs") == 0 \ + /* for Irix 6.5 */ \ + || strcmp (Fs_type, "ignore") == 0) +#endif + +#ifndef ME_REMOTE +/* A file system is `remote' if its Fs_name contains a `:' + or if (it is of type (smbfs or cifs) and its Fs_name starts with `//'). */ +# define ME_REMOTE(Fs_name, Fs_type) \ + (strchr (Fs_name, ':') != NULL \ + || ((Fs_name)[0] == '/' \ + && (Fs_name)[1] == '/' \ + && (strcmp (Fs_type, "smbfs") == 0 \ + || strcmp (Fs_type, "cifs") == 0))) +#endif + +#if MOUNTED_GETMNTINFO + +# if ! HAVE_STRUCT_STATFS_F_FSTYPENAME +static char * +fstype_to_string (short int t) +{ + switch (t) + { +# ifdef MOUNT_PC + case MOUNT_PC: + return "pc"; +# endif +# ifdef MOUNT_MFS + case MOUNT_MFS: + return "mfs"; +# endif +# ifdef MOUNT_LO + case MOUNT_LO: + return "lo"; +# endif +# ifdef MOUNT_TFS + case MOUNT_TFS: + return "tfs"; +# endif +# ifdef MOUNT_TMP + case MOUNT_TMP: + return "tmp"; +# endif +# ifdef MOUNT_UFS + case MOUNT_UFS: + return "ufs" ; +# endif +# ifdef MOUNT_NFS + case MOUNT_NFS: + return "nfs" ; +# endif +# ifdef MOUNT_MSDOS + case MOUNT_MSDOS: + return "msdos" ; +# endif +# ifdef MOUNT_LFS + case MOUNT_LFS: + return "lfs" ; +# endif +# ifdef MOUNT_LOFS + case MOUNT_LOFS: + return "lofs" ; +# endif +# ifdef MOUNT_FDESC + case MOUNT_FDESC: + return "fdesc" ; +# endif +# ifdef MOUNT_PORTAL + case MOUNT_PORTAL: + return "portal" ; +# endif +# ifdef MOUNT_NULL + case MOUNT_NULL: + return "null" ; +# endif +# ifdef MOUNT_UMAP + case MOUNT_UMAP: + return "umap" ; +# endif +# ifdef MOUNT_KERNFS + case MOUNT_KERNFS: + return "kernfs" ; +# endif +# ifdef MOUNT_PROCFS + case MOUNT_PROCFS: + return "procfs" ; +# endif +# ifdef MOUNT_AFS + case MOUNT_AFS: + return "afs" ; +# endif +# ifdef MOUNT_CD9660 + case MOUNT_CD9660: + return "cd9660" ; +# endif +# ifdef MOUNT_UNION + case MOUNT_UNION: + return "union" ; +# endif +# ifdef MOUNT_DEVFS + case MOUNT_DEVFS: + return "devfs" ; +# endif +# ifdef MOUNT_EXT2FS + case MOUNT_EXT2FS: + return "ext2fs" ; +# endif + default: + return "?"; + } +} +# endif + +static char * +fsp_to_string (const struct statfs *fsp) +{ +# if HAVE_STRUCT_STATFS_F_FSTYPENAME + return (char *) (fsp->f_fstypename); +# else + return fstype_to_string (fsp->f_type); +# endif +} + +#endif /* MOUNTED_GETMNTINFO */ + +#ifdef MOUNTED_VMOUNT /* AIX. */ +static char * +fstype_to_string (int t) +{ + struct vfs_ent *e; + + e = getvfsbytype (t); + if (!e || !e->vfsent_name) + return "none"; + else + return e->vfsent_name; +} +#endif /* MOUNTED_VMOUNT */ + + +#if defined MOUNTED_GETMNTENT1 || defined MOUNTED_GETMNTENT2 + +/* Return the device number from MOUNT_OPTIONS, if possible. + Otherwise return (dev_t) -1. */ + +static dev_t +dev_from_mount_options (char const *mount_options) +{ + /* GNU/Linux allows file system implementations to define their own + meaning for "dev=" mount options, so don't trust the meaning + here. */ +# ifndef __linux__ + + static char const dev_pattern[] = ",dev="; + char const *devopt = strstr (mount_options, dev_pattern); + + if (devopt) + { + char const *optval = devopt + sizeof dev_pattern - 1; + char *optvalend; + unsigned long int dev; + errno = 0; + dev = strtoul (optval, &optvalend, 16); + if (optval != optvalend + && (*optvalend == '\0' || *optvalend == ',') + && ! (dev == ULONG_MAX && errno == ERANGE) + && dev == (dev_t) dev) + return dev; + } + +# endif + + return -1; +} + +#endif + +/* Return a list of the currently mounted file systems, or NULL on error. + Add each entry to the tail of the list so that they stay in order. + If NEED_FS_TYPE is true, ensure that the file system type fields in + the returned list are valid. Otherwise, they might not be. */ + +struct mount_entry * +read_file_system_list (bool need_fs_type) +{ + struct mount_entry *mount_list; + struct mount_entry *me; + struct mount_entry **mtail = &mount_list; + +#ifdef MOUNTED_LISTMNTENT + { + struct tabmntent *mntlist, *p; + struct mntent *mnt; + struct mount_entry *me; + + /* the third and fourth arguments could be used to filter mounts, + but Crays doesn't seem to have any mounts that we want to + remove. Specifically, automount create normal NFS mounts. + */ + + if (listmntent (&mntlist, KMTAB, NULL, NULL) < 0) + return NULL; + for (p = mntlist; p; p = p->next) { + mnt = p->ment; + me = xmalloc (sizeof *me); + me->me_devname = xstrdup (mnt->mnt_fsname); + me->me_mountdir = xstrdup (mnt->mnt_dir); + me->me_type = xstrdup (mnt->mnt_type); + me->me_type_malloced = 1; + me->me_dummy = ME_DUMMY (me->me_devname, me->me_type); + me->me_remote = ME_REMOTE (me->me_devname, me->me_type); + me->me_dev = -1; + *mtail = me; + mtail = &me->me_next; + } + freemntlist (mntlist); + } +#endif + +#ifdef MOUNTED_GETMNTENT1 /* GNU/Linux, 4.3BSD, SunOS, HP-UX, Dynix, Irix. */ + { + struct mntent *mnt; + char *table = MOUNTED; + FILE *fp; + + fp = setmntent (table, "r"); + if (fp == NULL) + return NULL; + + while ((mnt = getmntent (fp))) + { + me = xmalloc (sizeof *me); + me->me_devname = xstrdup (mnt->mnt_fsname); + me->me_mountdir = xstrdup (mnt->mnt_dir); + me->me_type = xstrdup (mnt->mnt_type); + me->me_type_malloced = 1; + me->me_dummy = ME_DUMMY (me->me_devname, me->me_type); + me->me_remote = ME_REMOTE (me->me_devname, me->me_type); + me->me_dev = dev_from_mount_options (mnt->mnt_opts); + + /* Add to the linked list. */ + *mtail = me; + mtail = &me->me_next; + } + + if (endmntent (fp) == 0) + goto free_then_fail; + } +#endif /* MOUNTED_GETMNTENT1. */ + +#ifdef MOUNTED_GETMNTINFO /* 4.4BSD. */ + { + struct statfs *fsp; + int entries; + + entries = getmntinfo (&fsp, MNT_NOWAIT); + if (entries < 0) + return NULL; + for (; entries-- > 0; fsp++) + { + char *fs_type = fsp_to_string (fsp); + + me = xmalloc (sizeof *me); + me->me_devname = xstrdup (fsp->f_mntfromname); + me->me_mountdir = xstrdup (fsp->f_mntonname); + me->me_type = fs_type; + me->me_type_malloced = 0; + me->me_dummy = ME_DUMMY (me->me_devname, me->me_type); + me->me_remote = ME_REMOTE (me->me_devname, me->me_type); + me->me_dev = (dev_t) -1; /* Magic; means not known yet. */ + + /* Add to the linked list. */ + *mtail = me; + mtail = &me->me_next; + } + } +#endif /* MOUNTED_GETMNTINFO */ + +#ifdef MOUNTED_GETMNTINFO2 /* NetBSD 3.0. */ + { + struct statvfs *fsp; + int entries; + + entries = getmntinfo (&fsp, MNT_NOWAIT); + if (entries < 0) + return NULL; + for (; entries-- > 0; fsp++) + { + me = xmalloc (sizeof *me); + me->me_devname = xstrdup (fsp->f_mntfromname); + me->me_mountdir = xstrdup (fsp->f_mntonname); + me->me_type = xstrdup (fsp->f_fstypename); + me->me_type_malloced = 1; + me->me_dummy = ME_DUMMY (me->me_devname, me->me_type); + me->me_remote = ME_REMOTE (me->me_devname, me->me_type); + me->me_dev = (dev_t) -1; /* Magic; means not known yet. */ + + /* Add to the linked list. */ + *mtail = me; + mtail = &me->me_next; + } + } +#endif /* MOUNTED_GETMNTINFO2 */ + +#ifdef MOUNTED_GETMNT /* Ultrix. */ + { + int offset = 0; + int val; + struct fs_data fsd; + + while (errno = 0, + 0 < (val = getmnt (&offset, &fsd, sizeof (fsd), NOSTAT_MANY, + (char *) 0))) + { + me = xmalloc (sizeof *me); + me->me_devname = xstrdup (fsd.fd_req.devname); + me->me_mountdir = xstrdup (fsd.fd_req.path); + me->me_type = gt_names[fsd.fd_req.fstype]; + me->me_type_malloced = 0; + me->me_dummy = ME_DUMMY (me->me_devname, me->me_type); + me->me_remote = ME_REMOTE (me->me_devname, me->me_type); + me->me_dev = fsd.fd_req.dev; + + /* Add to the linked list. */ + *mtail = me; + mtail = &me->me_next; + } + if (val < 0) + goto free_then_fail; + } +#endif /* MOUNTED_GETMNT. */ + +#if defined MOUNTED_FS_STAT_DEV /* BeOS */ + { + /* The next_dev() and fs_stat_dev() system calls give the list of + all file systems, including the information returned by statvfs() + (fs type, total blocks, free blocks etc.), but without the mount + point. But on BeOS all file systems except / are mounted in the + rootfs, directly under /. + The directory name of the mount point is often, but not always, + identical to the volume name of the device. + We therefore get the list of subdirectories of /, and the list + of all file systems, and match the two lists. */ + + DIR *dirp; + struct rootdir_entry + { + char *name; + dev_t dev; + ino_t ino; + struct rootdir_entry *next; + }; + struct rootdir_entry *rootdir_list; + struct rootdir_entry **rootdir_tail; + int32 pos; + dev_t dev; + fs_info fi; + + /* All volumes are mounted in the rootfs, directly under /. */ + rootdir_list = NULL; + rootdir_tail = &rootdir_list; + dirp = opendir ("/"); + if (dirp) + { + struct dirent *d; + + while ((d = readdir (dirp)) != NULL) + { + char *name; + struct stat statbuf; + + if (strcmp (d->d_name, "..") == 0) + continue; + + if (strcmp (d->d_name, ".") == 0) + name = xstrdup ("/"); + else + { + name = xmalloc (1 + strlen (d->d_name) + 1); + name[0] = '/'; + strcpy (name + 1, d->d_name); + } + + if (lstat (name, &statbuf) >= 0 && S_ISDIR (statbuf.st_mode)) + { + struct rootdir_entry *re = xmalloc (sizeof *re); + re->name = name; + re->dev = statbuf.st_dev; + re->ino = statbuf.st_ino; + + /* Add to the linked list. */ + *rootdir_tail = re; + rootdir_tail = &re->next; + } + else + free (name); + } + closedir (dirp); + } + *rootdir_tail = NULL; + + for (pos = 0; (dev = next_dev (&pos)) >= 0; ) + if (fs_stat_dev (dev, &fi) >= 0) + { + /* Note: fi.dev == dev. */ + struct rootdir_entry *re; + + for (re = rootdir_list; re; re = re->next) + if (re->dev == fi.dev && re->ino == fi.root) + break; + + me = xmalloc (sizeof *me); + me->me_devname = xstrdup (fi.device_name[0] != '\0' ? fi.device_name : fi.fsh_name); + me->me_mountdir = xstrdup (re != NULL ? re->name : fi.fsh_name); + me->me_type = xstrdup (fi.fsh_name); + me->me_type_malloced = 1; + me->me_dev = fi.dev; + me->me_dummy = 0; + me->me_remote = (fi.flags & B_FS_IS_SHARED) != 0; + + /* Add to the linked list. */ + *mtail = me; + mtail = &me->me_next; + } + *mtail = NULL; + + while (rootdir_list != NULL) + { + struct rootdir_entry *re = rootdir_list; + rootdir_list = re->next; + free (re->name); + free (re); + } + } +#endif /* MOUNTED_FS_STAT_DEV */ + +#if defined MOUNTED_GETFSSTAT /* __alpha running OSF_1 */ + { + int numsys, counter; + size_t bufsize; + struct statfs *stats; + + numsys = getfsstat ((struct statfs *)0, 0L, MNT_NOWAIT); + if (numsys < 0) + return (NULL); + if (SIZE_MAX / sizeof *stats <= numsys) + xalloc_die (); + + bufsize = (1 + numsys) * sizeof *stats; + stats = xmalloc (bufsize); + numsys = getfsstat (stats, bufsize, MNT_NOWAIT); + + if (numsys < 0) + { + free (stats); + return (NULL); + } + + for (counter = 0; counter < numsys; counter++) + { + me = xmalloc (sizeof *me); + me->me_devname = xstrdup (stats[counter].f_mntfromname); + me->me_mountdir = xstrdup (stats[counter].f_mntonname); + me->me_type = xstrdup (FS_TYPE (stats[counter])); + me->me_type_malloced = 1; + me->me_dummy = ME_DUMMY (me->me_devname, me->me_type); + me->me_remote = ME_REMOTE (me->me_devname, me->me_type); + me->me_dev = (dev_t) -1; /* Magic; means not known yet. */ + + /* Add to the linked list. */ + *mtail = me; + mtail = &me->me_next; + } + + free (stats); + } +#endif /* MOUNTED_GETFSSTAT */ + +#if defined MOUNTED_FREAD || defined MOUNTED_FREAD_FSTYP /* SVR[23]. */ + { + struct mnttab mnt; + char *table = "/etc/mnttab"; + FILE *fp; + + fp = fopen (table, "r"); + if (fp == NULL) + return NULL; + + while (fread (&mnt, sizeof mnt, 1, fp) > 0) + { + me = xmalloc (sizeof *me); +# ifdef GETFSTYP /* SVR3. */ + me->me_devname = xstrdup (mnt.mt_dev); +# else + me->me_devname = xmalloc (strlen (mnt.mt_dev) + 6); + strcpy (me->me_devname, "/dev/"); + strcpy (me->me_devname + 5, mnt.mt_dev); +# endif + me->me_mountdir = xstrdup (mnt.mt_filsys); + me->me_dev = (dev_t) -1; /* Magic; means not known yet. */ + me->me_type = ""; + me->me_type_malloced = 0; +# ifdef GETFSTYP /* SVR3. */ + if (need_fs_type) + { + struct statfs fsd; + char typebuf[FSTYPSZ]; + + if (statfs (me->me_mountdir, &fsd, sizeof fsd, 0) != -1 + && sysfs (GETFSTYP, fsd.f_fstyp, typebuf) != -1) + { + me->me_type = xstrdup (typebuf); + me->me_type_malloced = 1; + } + } +# endif + me->me_dummy = ME_DUMMY (me->me_devname, me->me_type); + me->me_remote = ME_REMOTE (me->me_devname, me->me_type); + + /* Add to the linked list. */ + *mtail = me; + mtail = &me->me_next; + } + + if (ferror (fp)) + { + /* The last fread() call must have failed. */ + int saved_errno = errno; + fclose (fp); + errno = saved_errno; + goto free_then_fail; + } + + if (fclose (fp) == EOF) + goto free_then_fail; + } +#endif /* MOUNTED_FREAD || MOUNTED_FREAD_FSTYP. */ + +#ifdef MOUNTED_GETMNTTBL /* DolphinOS goes its own way. */ + { + struct mntent **mnttbl = getmnttbl (), **ent; + for (ent=mnttbl;*ent;ent++) + { + me = xmalloc (sizeof *me); + me->me_devname = xstrdup ( (*ent)->mt_resource); + me->me_mountdir = xstrdup ( (*ent)->mt_directory); + me->me_type = xstrdup ((*ent)->mt_fstype); + me->me_type_malloced = 1; + me->me_dummy = ME_DUMMY (me->me_devname, me->me_type); + me->me_remote = ME_REMOTE (me->me_devname, me->me_type); + me->me_dev = (dev_t) -1; /* Magic; means not known yet. */ + + /* Add to the linked list. */ + *mtail = me; + mtail = &me->me_next; + } + endmnttbl (); + } +#endif + +#ifdef MOUNTED_GETMNTENT2 /* SVR4. */ + { + struct mnttab mnt; + char *table = MNTTAB; + FILE *fp; + int ret; + int lockfd = -1; + +# if defined F_RDLCK && defined F_SETLKW + /* MNTTAB_LOCK is a macro name of our own invention; it's not present in + e.g. Solaris 2.6. If the SVR4 folks ever define a macro + for this file name, we should use their macro name instead. + (Why not just lock MNTTAB directly? We don't know.) */ +# ifndef MNTTAB_LOCK +# define MNTTAB_LOCK "/etc/.mnttab.lock" +# endif + lockfd = open (MNTTAB_LOCK, O_RDONLY); + if (0 <= lockfd) + { + struct flock flock; + flock.l_type = F_RDLCK; + flock.l_whence = SEEK_SET; + flock.l_start = 0; + flock.l_len = 0; + while (fcntl (lockfd, F_SETLKW, &flock) == -1) + if (errno != EINTR) + { + int saved_errno = errno; + close (lockfd); + errno = saved_errno; + return NULL; + } + } + else if (errno != ENOENT) + return NULL; +# endif + + errno = 0; + fp = fopen (table, "r"); + if (fp == NULL) + ret = errno; + else + { + while ((ret = getmntent (fp, &mnt)) == 0) + { + me = xmalloc (sizeof *me); + me->me_devname = xstrdup (mnt.mnt_special); + me->me_mountdir = xstrdup (mnt.mnt_mountp); + me->me_type = xstrdup (mnt.mnt_fstype); + me->me_type_malloced = 1; + me->me_dummy = MNT_IGNORE (&mnt) != 0; + me->me_remote = ME_REMOTE (me->me_devname, me->me_type); + me->me_dev = dev_from_mount_options (mnt.mnt_mntopts); + + /* Add to the linked list. */ + *mtail = me; + mtail = &me->me_next; + } + + ret = fclose (fp) == EOF ? errno : 0 < ret ? 0 : -1; + } + + if (0 <= lockfd && close (lockfd) != 0) + ret = errno; + + if (0 <= ret) + { + errno = ret; + goto free_then_fail; + } + } +#endif /* MOUNTED_GETMNTENT2. */ + +#ifdef MOUNTED_VMOUNT /* AIX. */ + { + int bufsize; + char *entries, *thisent; + struct vmount *vmp; + int n_entries; + int i; + + /* Ask how many bytes to allocate for the mounted file system info. */ + if (mntctl (MCTL_QUERY, sizeof bufsize, (struct vmount *) &bufsize) != 0) + return NULL; + entries = xmalloc (bufsize); + + /* Get the list of mounted file systems. */ + n_entries = mntctl (MCTL_QUERY, bufsize, (struct vmount *) entries); + if (n_entries < 0) + { + int saved_errno = errno; + free (entries); + errno = saved_errno; + return NULL; + } + + for (i = 0, thisent = entries; + i < n_entries; + i++, thisent += vmp->vmt_length) + { + char *options, *ignore; + + vmp = (struct vmount *) thisent; + me = xmalloc (sizeof *me); + if (vmp->vmt_flags & MNT_REMOTE) + { + char *host, *dir; + + me->me_remote = 1; + /* Prepend the remote dirname. */ + host = thisent + vmp->vmt_data[VMT_HOSTNAME].vmt_off; + dir = thisent + vmp->vmt_data[VMT_OBJECT].vmt_off; + me->me_devname = xmalloc (strlen (host) + strlen (dir) + 2); + strcpy (me->me_devname, host); + strcat (me->me_devname, ":"); + strcat (me->me_devname, dir); + } + else + { + me->me_remote = 0; + me->me_devname = xstrdup (thisent + + vmp->vmt_data[VMT_OBJECT].vmt_off); + } + me->me_mountdir = xstrdup (thisent + vmp->vmt_data[VMT_STUB].vmt_off); + me->me_type = xstrdup (fstype_to_string (vmp->vmt_gfstype)); + me->me_type_malloced = 1; + options = thisent + vmp->vmt_data[VMT_ARGS].vmt_off; + ignore = strstr (options, "ignore"); + me->me_dummy = (ignore + && (ignore == options || ignore[-1] == ',') + && (ignore[sizeof "ignore" - 1] == ',' + || ignore[sizeof "ignore" - 1] == '\0')); + me->me_dev = (dev_t) -1; /* vmt_fsid might be the info we want. */ + + /* Add to the linked list. */ + *mtail = me; + mtail = &me->me_next; + } + free (entries); + } +#endif /* MOUNTED_VMOUNT. */ + + *mtail = NULL; + return mount_list; + + + free_then_fail: + { + int saved_errno = errno; + *mtail = NULL; + + while (mount_list) + { + me = mount_list->me_next; + free (mount_list->me_devname); + free (mount_list->me_mountdir); + if (mount_list->me_type_malloced) + free (mount_list->me_type); + free (mount_list); + mount_list = me; + } + + errno = saved_errno; + return NULL; + } +} diff --git a/gl/mountlist.h b/gl/mountlist.h new file mode 100644 index 00000000..7f5a6f77 --- /dev/null +++ b/gl/mountlist.h @@ -0,0 +1,41 @@ +/* mountlist.h -- declarations for list of mounted file systems + + Copyright (C) 1991, 1992, 1998, 2000, 2001, 2002, 2003, 2004, 2005 + Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef MOUNTLIST_H_ +# define MOUNTLIST_H_ + +# include <stdbool.h> +# include <sys/types.h> + +/* A mount table entry. */ +struct mount_entry +{ + char *me_devname; /* Device node name, including "/dev/". */ + char *me_mountdir; /* Mount point directory name. */ + char *me_type; /* "nfs", "4.2", etc. */ + dev_t me_dev; /* Device number of me_mountdir. */ + unsigned int me_dummy : 1; /* Nonzero for dummy file systems. */ + unsigned int me_remote : 1; /* Nonzero for remote fileystems. */ + unsigned int me_type_malloced : 1; /* Nonzero if me_type was malloced. */ + struct mount_entry *me_next; +}; + +struct mount_entry *read_file_system_list (bool need_fs_type); + +#endif diff --git a/gl/open-safer.c b/gl/open-safer.c new file mode 100644 index 00000000..04a72eb7 --- /dev/null +++ b/gl/open-safer.c @@ -0,0 +1,50 @@ +/* Invoke open, but avoid some glitches. + + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <config.h> + +#include "fcntl-safer.h" + +#include <fcntl.h> +#include <stdarg.h> +#include "unistd-safer.h" + +int +open_safer (char const *file, int flags, ...) +{ + mode_t mode = 0; + + if (flags & O_CREAT) + { + va_list ap; + va_start (ap, flags); + + /* Assume mode_t promotes to int if and only if it is smaller. + This assumption isn't guaranteed by the C standard, but we + don't know of any real-world counterexamples. */ + mode = (sizeof (mode_t) < sizeof (int) + ? va_arg (ap, int) + : va_arg (ap, mode_t)); + + va_end (ap); + } + + return fd_safer (open (file, flags, mode)); +} diff --git a/gl/pipe-safer.c b/gl/pipe-safer.c new file mode 100644 index 00000000..e4431b33 --- /dev/null +++ b/gl/pipe-safer.c @@ -0,0 +1,57 @@ +/* Invoke pipe, but avoid some glitches. + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Jim Meyering. */ + +#include <config.h> + +#include "unistd-safer.h" + +#include <unistd.h> +#include <errno.h> + +/* Like pipe, but ensure that neither of the file descriptors is + STDIN_FILENO, STDOUT_FILENO, or STDERR_FILENO. Fail with ENOSYS on + platforms that lack pipe. */ + +int +pipe_safer (int fd[2]) +{ +#if HAVE_PIPE + if (pipe (fd) == 0) + { + int i; + for (i = 0; i < 2; i++) + { + fd[i] = fd_safer (fd[i]); + if (fd[i] < 0) + { + int e = errno; + close (fd[1 - i]); + errno = e; + return -1; + } + } + + return 0; + } +#else + errno = ENOSYS; +#endif + + return -1; +} diff --git a/gl/printf-args.c b/gl/printf-args.c new file mode 100644 index 00000000..358801c9 --- /dev/null +++ b/gl/printf-args.c @@ -0,0 +1,141 @@ +/* Decomposed printf argument list. + Copyright (C) 1999, 2002-2003, 2005-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include "printf-args.h" + +#ifdef STATIC +STATIC +#endif +int +printf_fetchargs (va_list args, arguments *a) +{ + size_t i; + argument *ap; + + for (i = 0, ap = &a->arg[0]; i < a->count; i++, ap++) + switch (ap->type) + { + case TYPE_SCHAR: + ap->a.a_schar = va_arg (args, /*signed char*/ int); + break; + case TYPE_UCHAR: + ap->a.a_uchar = va_arg (args, /*unsigned char*/ int); + break; + case TYPE_SHORT: + ap->a.a_short = va_arg (args, /*short*/ int); + break; + case TYPE_USHORT: + ap->a.a_ushort = va_arg (args, /*unsigned short*/ int); + break; + case TYPE_INT: + ap->a.a_int = va_arg (args, int); + break; + case TYPE_UINT: + ap->a.a_uint = va_arg (args, unsigned int); + break; + case TYPE_LONGINT: + ap->a.a_longint = va_arg (args, long int); + break; + case TYPE_ULONGINT: + ap->a.a_ulongint = va_arg (args, unsigned long int); + break; +#ifdef HAVE_LONG_LONG_INT + case TYPE_LONGLONGINT: + ap->a.a_longlongint = va_arg (args, long long int); + break; + case TYPE_ULONGLONGINT: + ap->a.a_ulonglongint = va_arg (args, unsigned long long int); + break; +#endif + case TYPE_DOUBLE: + ap->a.a_double = va_arg (args, double); + break; +#ifdef HAVE_LONG_DOUBLE + case TYPE_LONGDOUBLE: + ap->a.a_longdouble = va_arg (args, long double); + break; +#endif + case TYPE_CHAR: + ap->a.a_char = va_arg (args, int); + break; +#ifdef HAVE_WINT_T + case TYPE_WIDE_CHAR: + /* Although ISO C 99 7.24.1.(2) says that wint_t is "unchanged by + default argument promotions", this is not the case in mingw32, + where wint_t is 'unsigned short'. */ + ap->a.a_wide_char = + (sizeof (wint_t) < sizeof (int) + ? va_arg (args, int) + : va_arg (args, wint_t)); + break; +#endif + case TYPE_STRING: + ap->a.a_string = va_arg (args, const char *); + /* A null pointer is an invalid argument for "%s", but in practice + it occurs quite frequently in printf statements that produce + debug output. Use a fallback in this case. */ + if (ap->a.a_string == NULL) + ap->a.a_string = "(NULL)"; + break; +#ifdef HAVE_WCHAR_T + case TYPE_WIDE_STRING: + ap->a.a_wide_string = va_arg (args, const wchar_t *); + /* A null pointer is an invalid argument for "%ls", but in practice + it occurs quite frequently in printf statements that produce + debug output. Use a fallback in this case. */ + if (ap->a.a_wide_string == NULL) + { + static const wchar_t wide_null_string[] = + { + (wchar_t)'(', + (wchar_t)'N', (wchar_t)'U', (wchar_t)'L', (wchar_t)'L', + (wchar_t)')', + (wchar_t)0 + }; + ap->a.a_wide_string = wide_null_string; + } + break; +#endif + case TYPE_POINTER: + ap->a.a_pointer = va_arg (args, void *); + break; + case TYPE_COUNT_SCHAR_POINTER: + ap->a.a_count_schar_pointer = va_arg (args, signed char *); + break; + case TYPE_COUNT_SHORT_POINTER: + ap->a.a_count_short_pointer = va_arg (args, short *); + break; + case TYPE_COUNT_INT_POINTER: + ap->a.a_count_int_pointer = va_arg (args, int *); + break; + case TYPE_COUNT_LONGINT_POINTER: + ap->a.a_count_longint_pointer = va_arg (args, long int *); + break; +#ifdef HAVE_LONG_LONG_INT + case TYPE_COUNT_LONGLONGINT_POINTER: + ap->a.a_count_longlongint_pointer = va_arg (args, long long int *); + break; +#endif + default: + /* Unknown type. */ + return -1; + } + return 0; +} diff --git a/gl/printf-args.h b/gl/printf-args.h new file mode 100644 index 00000000..5759da0e --- /dev/null +++ b/gl/printf-args.h @@ -0,0 +1,136 @@ +/* Decomposed printf argument list. + Copyright (C) 1999, 2002-2003, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _PRINTF_ARGS_H +#define _PRINTF_ARGS_H + +/* Get size_t. */ +#include <stddef.h> + +/* Get wchar_t. */ +#ifdef HAVE_WCHAR_T +# include <stddef.h> +#endif + +/* Get wint_t. */ +#ifdef HAVE_WINT_T +# include <wchar.h> +#endif + +/* Get va_list. */ +#include <stdarg.h> + + +/* Argument types */ +typedef enum +{ + TYPE_NONE, + TYPE_SCHAR, + TYPE_UCHAR, + TYPE_SHORT, + TYPE_USHORT, + TYPE_INT, + TYPE_UINT, + TYPE_LONGINT, + TYPE_ULONGINT, +#ifdef HAVE_LONG_LONG_INT + TYPE_LONGLONGINT, + TYPE_ULONGLONGINT, +#endif + TYPE_DOUBLE, +#ifdef HAVE_LONG_DOUBLE + TYPE_LONGDOUBLE, +#endif + TYPE_CHAR, +#ifdef HAVE_WINT_T + TYPE_WIDE_CHAR, +#endif + TYPE_STRING, +#ifdef HAVE_WCHAR_T + TYPE_WIDE_STRING, +#endif + TYPE_POINTER, + TYPE_COUNT_SCHAR_POINTER, + TYPE_COUNT_SHORT_POINTER, + TYPE_COUNT_INT_POINTER, + TYPE_COUNT_LONGINT_POINTER +#ifdef HAVE_LONG_LONG_INT +, TYPE_COUNT_LONGLONGINT_POINTER +#endif +} arg_type; + +/* Polymorphic argument */ +typedef struct +{ + arg_type type; + union + { + signed char a_schar; + unsigned char a_uchar; + short a_short; + unsigned short a_ushort; + int a_int; + unsigned int a_uint; + long int a_longint; + unsigned long int a_ulongint; +#ifdef HAVE_LONG_LONG_INT + long long int a_longlongint; + unsigned long long int a_ulonglongint; +#endif + float a_float; + double a_double; +#ifdef HAVE_LONG_DOUBLE + long double a_longdouble; +#endif + int a_char; +#ifdef HAVE_WINT_T + wint_t a_wide_char; +#endif + const char* a_string; +#ifdef HAVE_WCHAR_T + const wchar_t* a_wide_string; +#endif + void* a_pointer; + signed char * a_count_schar_pointer; + short * a_count_short_pointer; + int * a_count_int_pointer; + long int * a_count_longint_pointer; +#ifdef HAVE_LONG_LONG_INT + long long int * a_count_longlongint_pointer; +#endif + } + a; +} +argument; + +typedef struct +{ + size_t count; + argument *arg; +} +arguments; + + +/* Fetch the arguments, putting them into a. */ +#ifdef STATIC +STATIC +#else +extern +#endif +int printf_fetchargs (va_list args, arguments *a); + +#endif /* _PRINTF_ARGS_H */ diff --git a/gl/printf-parse.c b/gl/printf-parse.c new file mode 100644 index 00000000..9a86f773 --- /dev/null +++ b/gl/printf-parse.c @@ -0,0 +1,543 @@ +/* Formatted output to strings. + Copyright (C) 1999-2000, 2002-2003, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#if WIDE_CHAR_VERSION +# include "wprintf-parse.h" +#else +# include "printf-parse.h" +#endif + +/* Get size_t, NULL. */ +#include <stddef.h> + +/* Get intmax_t. */ +#if HAVE_STDINT_H_WITH_UINTMAX +# include <stdint.h> +#endif +#if HAVE_INTTYPES_H_WITH_UINTMAX +# include <inttypes.h> +#endif + +/* malloc(), realloc(), free(). */ +#include <stdlib.h> + +/* Checked size_t computations. */ +#include "xsize.h" + +#if WIDE_CHAR_VERSION +# define PRINTF_PARSE wprintf_parse +# define CHAR_T wchar_t +# define DIRECTIVE wchar_t_directive +# define DIRECTIVES wchar_t_directives +#else +# define PRINTF_PARSE printf_parse +# define CHAR_T char +# define DIRECTIVE char_directive +# define DIRECTIVES char_directives +#endif + +#ifdef STATIC +STATIC +#endif +int +PRINTF_PARSE (const CHAR_T *format, DIRECTIVES *d, arguments *a) +{ + const CHAR_T *cp = format; /* pointer into format */ + size_t arg_posn = 0; /* number of regular arguments consumed */ + size_t d_allocated; /* allocated elements of d->dir */ + size_t a_allocated; /* allocated elements of a->arg */ + size_t max_width_length = 0; + size_t max_precision_length = 0; + + d->count = 0; + d_allocated = 1; + d->dir = (DIRECTIVE *) malloc (d_allocated * sizeof (DIRECTIVE)); + if (d->dir == NULL) + /* Out of memory. */ + return -1; + + a->count = 0; + a_allocated = 0; + a->arg = NULL; + +#define REGISTER_ARG(_index_,_type_) \ + { \ + size_t n = (_index_); \ + if (n >= a_allocated) \ + { \ + size_t memory_size; \ + argument *memory; \ + \ + a_allocated = xtimes (a_allocated, 2); \ + if (a_allocated <= n) \ + a_allocated = xsum (n, 1); \ + memory_size = xtimes (a_allocated, sizeof (argument)); \ + if (size_overflow_p (memory_size)) \ + /* Overflow, would lead to out of memory. */ \ + goto error; \ + memory = (argument *) (a->arg \ + ? realloc (a->arg, memory_size) \ + : malloc (memory_size)); \ + if (memory == NULL) \ + /* Out of memory. */ \ + goto error; \ + a->arg = memory; \ + } \ + while (a->count <= n) \ + a->arg[a->count++].type = TYPE_NONE; \ + if (a->arg[n].type == TYPE_NONE) \ + a->arg[n].type = (_type_); \ + else if (a->arg[n].type != (_type_)) \ + /* Ambiguous type for positional argument. */ \ + goto error; \ + } + + while (*cp != '\0') + { + CHAR_T c = *cp++; + if (c == '%') + { + size_t arg_index = ARG_NONE; + DIRECTIVE *dp = &d->dir[d->count];/* pointer to next directive */ + + /* Initialize the next directive. */ + dp->dir_start = cp - 1; + dp->flags = 0; + dp->width_start = NULL; + dp->width_end = NULL; + dp->width_arg_index = ARG_NONE; + dp->precision_start = NULL; + dp->precision_end = NULL; + dp->precision_arg_index = ARG_NONE; + dp->arg_index = ARG_NONE; + + /* Test for positional argument. */ + if (*cp >= '0' && *cp <= '9') + { + const CHAR_T *np; + + for (np = cp; *np >= '0' && *np <= '9'; np++) + ; + if (*np == '$') + { + size_t n = 0; + + for (np = cp; *np >= '0' && *np <= '9'; np++) + n = xsum (xtimes (n, 10), *np - '0'); + if (n == 0) + /* Positional argument 0. */ + goto error; + if (size_overflow_p (n)) + /* n too large, would lead to out of memory later. */ + goto error; + arg_index = n - 1; + cp = np + 1; + } + } + + /* Read the flags. */ + for (;;) + { + if (*cp == '\'') + { + dp->flags |= FLAG_GROUP; + cp++; + } + else if (*cp == '-') + { + dp->flags |= FLAG_LEFT; + cp++; + } + else if (*cp == '+') + { + dp->flags |= FLAG_SHOWSIGN; + cp++; + } + else if (*cp == ' ') + { + dp->flags |= FLAG_SPACE; + cp++; + } + else if (*cp == '#') + { + dp->flags |= FLAG_ALT; + cp++; + } + else if (*cp == '0') + { + dp->flags |= FLAG_ZERO; + cp++; + } + else + break; + } + + /* Parse the field width. */ + if (*cp == '*') + { + dp->width_start = cp; + cp++; + dp->width_end = cp; + if (max_width_length < 1) + max_width_length = 1; + + /* Test for positional argument. */ + if (*cp >= '0' && *cp <= '9') + { + const CHAR_T *np; + + for (np = cp; *np >= '0' && *np <= '9'; np++) + ; + if (*np == '$') + { + size_t n = 0; + + for (np = cp; *np >= '0' && *np <= '9'; np++) + n = xsum (xtimes (n, 10), *np - '0'); + if (n == 0) + /* Positional argument 0. */ + goto error; + if (size_overflow_p (n)) + /* n too large, would lead to out of memory later. */ + goto error; + dp->width_arg_index = n - 1; + cp = np + 1; + } + } + if (dp->width_arg_index == ARG_NONE) + { + dp->width_arg_index = arg_posn++; + if (dp->width_arg_index == ARG_NONE) + /* arg_posn wrapped around. */ + goto error; + } + REGISTER_ARG (dp->width_arg_index, TYPE_INT); + } + else if (*cp >= '0' && *cp <= '9') + { + size_t width_length; + + dp->width_start = cp; + for (; *cp >= '0' && *cp <= '9'; cp++) + ; + dp->width_end = cp; + width_length = dp->width_end - dp->width_start; + if (max_width_length < width_length) + max_width_length = width_length; + } + + /* Parse the precision. */ + if (*cp == '.') + { + cp++; + if (*cp == '*') + { + dp->precision_start = cp - 1; + cp++; + dp->precision_end = cp; + if (max_precision_length < 2) + max_precision_length = 2; + + /* Test for positional argument. */ + if (*cp >= '0' && *cp <= '9') + { + const CHAR_T *np; + + for (np = cp; *np >= '0' && *np <= '9'; np++) + ; + if (*np == '$') + { + size_t n = 0; + + for (np = cp; *np >= '0' && *np <= '9'; np++) + n = xsum (xtimes (n, 10), *np - '0'); + if (n == 0) + /* Positional argument 0. */ + goto error; + if (size_overflow_p (n)) + /* n too large, would lead to out of memory + later. */ + goto error; + dp->precision_arg_index = n - 1; + cp = np + 1; + } + } + if (dp->precision_arg_index == ARG_NONE) + { + dp->precision_arg_index = arg_posn++; + if (dp->precision_arg_index == ARG_NONE) + /* arg_posn wrapped around. */ + goto error; + } + REGISTER_ARG (dp->precision_arg_index, TYPE_INT); + } + else + { + size_t precision_length; + + dp->precision_start = cp - 1; + for (; *cp >= '0' && *cp <= '9'; cp++) + ; + dp->precision_end = cp; + precision_length = dp->precision_end - dp->precision_start; + if (max_precision_length < precision_length) + max_precision_length = precision_length; + } + } + + { + arg_type type; + + /* Parse argument type/size specifiers. */ + { + int flags = 0; + + for (;;) + { + if (*cp == 'h') + { + flags |= (1 << (flags & 1)); + cp++; + } + else if (*cp == 'L') + { + flags |= 4; + cp++; + } + else if (*cp == 'l') + { + flags += 8; + cp++; + } +#ifdef HAVE_INTMAX_T + else if (*cp == 'j') + { + if (sizeof (intmax_t) > sizeof (long)) + { + /* intmax_t = long long */ + flags += 16; + } + else if (sizeof (intmax_t) > sizeof (int)) + { + /* intmax_t = long */ + flags += 8; + } + cp++; + } +#endif + else if (*cp == 'z' || *cp == 'Z') + { + /* 'z' is standardized in ISO C 99, but glibc uses 'Z' + because the warning facility in gcc-2.95.2 understands + only 'Z' (see gcc-2.95.2/gcc/c-common.c:1784). */ + if (sizeof (size_t) > sizeof (long)) + { + /* size_t = long long */ + flags += 16; + } + else if (sizeof (size_t) > sizeof (int)) + { + /* size_t = long */ + flags += 8; + } + cp++; + } + else if (*cp == 't') + { + if (sizeof (ptrdiff_t) > sizeof (long)) + { + /* ptrdiff_t = long long */ + flags += 16; + } + else if (sizeof (ptrdiff_t) > sizeof (int)) + { + /* ptrdiff_t = long */ + flags += 8; + } + cp++; + } + else + break; + } + + /* Read the conversion character. */ + c = *cp++; + switch (c) + { + case 'd': case 'i': +#ifdef HAVE_LONG_LONG_INT + /* If 'long long' exists and is larger than 'long': */ + if (flags >= 16 || (flags & 4)) + type = TYPE_LONGLONGINT; + else +#endif + /* If 'long long' exists and is the same as 'long', we parse + "lld" into TYPE_LONGINT. */ + if (flags >= 8) + type = TYPE_LONGINT; + else if (flags & 2) + type = TYPE_SCHAR; + else if (flags & 1) + type = TYPE_SHORT; + else + type = TYPE_INT; + break; + case 'o': case 'u': case 'x': case 'X': +#ifdef HAVE_LONG_LONG_INT + /* If 'long long' exists and is larger than 'long': */ + if (flags >= 16 || (flags & 4)) + type = TYPE_ULONGLONGINT; + else +#endif + /* If 'unsigned long long' exists and is the same as + 'unsigned long', we parse "llu" into TYPE_ULONGINT. */ + if (flags >= 8) + type = TYPE_ULONGINT; + else if (flags & 2) + type = TYPE_UCHAR; + else if (flags & 1) + type = TYPE_USHORT; + else + type = TYPE_UINT; + break; + case 'f': case 'F': case 'e': case 'E': case 'g': case 'G': + case 'a': case 'A': +#ifdef HAVE_LONG_DOUBLE + if (flags >= 16 || (flags & 4)) + type = TYPE_LONGDOUBLE; + else +#endif + type = TYPE_DOUBLE; + break; + case 'c': + if (flags >= 8) +#ifdef HAVE_WINT_T + type = TYPE_WIDE_CHAR; +#else + goto error; +#endif + else + type = TYPE_CHAR; + break; +#ifdef HAVE_WINT_T + case 'C': + type = TYPE_WIDE_CHAR; + c = 'c'; + break; +#endif + case 's': + if (flags >= 8) +#ifdef HAVE_WCHAR_T + type = TYPE_WIDE_STRING; +#else + goto error; +#endif + else + type = TYPE_STRING; + break; +#ifdef HAVE_WCHAR_T + case 'S': + type = TYPE_WIDE_STRING; + c = 's'; + break; +#endif + case 'p': + type = TYPE_POINTER; + break; + case 'n': +#ifdef HAVE_LONG_LONG_INT + /* If 'long long' exists and is larger than 'long': */ + if (flags >= 16 || (flags & 4)) + type = TYPE_COUNT_LONGLONGINT_POINTER; + else +#endif + /* If 'long long' exists and is the same as 'long', we parse + "lln" into TYPE_COUNT_LONGINT_POINTER. */ + if (flags >= 8) + type = TYPE_COUNT_LONGINT_POINTER; + else if (flags & 2) + type = TYPE_COUNT_SCHAR_POINTER; + else if (flags & 1) + type = TYPE_COUNT_SHORT_POINTER; + else + type = TYPE_COUNT_INT_POINTER; + break; + case '%': + type = TYPE_NONE; + break; + default: + /* Unknown conversion character. */ + goto error; + } + } + + if (type != TYPE_NONE) + { + dp->arg_index = arg_index; + if (dp->arg_index == ARG_NONE) + { + dp->arg_index = arg_posn++; + if (dp->arg_index == ARG_NONE) + /* arg_posn wrapped around. */ + goto error; + } + REGISTER_ARG (dp->arg_index, type); + } + dp->conversion = c; + dp->dir_end = cp; + } + + d->count++; + if (d->count >= d_allocated) + { + size_t memory_size; + DIRECTIVE *memory; + + d_allocated = xtimes (d_allocated, 2); + memory_size = xtimes (d_allocated, sizeof (DIRECTIVE)); + if (size_overflow_p (memory_size)) + /* Overflow, would lead to out of memory. */ + goto error; + memory = (DIRECTIVE *) realloc (d->dir, memory_size); + if (memory == NULL) + /* Out of memory. */ + goto error; + d->dir = memory; + } + } + } + d->dir[d->count].dir_start = cp; + + d->max_width_length = max_width_length; + d->max_precision_length = max_precision_length; + return 0; + +error: + if (a->arg) + free (a->arg); + if (d->dir) + free (d->dir); + return -1; +} + +#undef DIRECTIVES +#undef DIRECTIVE +#undef CHAR_T +#undef PRINTF_PARSE diff --git a/gl/printf-parse.h b/gl/printf-parse.h new file mode 100644 index 00000000..82a0d37c --- /dev/null +++ b/gl/printf-parse.h @@ -0,0 +1,74 @@ +/* Parse printf format string. + Copyright (C) 1999, 2002-2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _PRINTF_PARSE_H +#define _PRINTF_PARSE_H + +#include "printf-args.h" + + +/* Flags */ +#define FLAG_GROUP 1 /* ' flag */ +#define FLAG_LEFT 2 /* - flag */ +#define FLAG_SHOWSIGN 4 /* + flag */ +#define FLAG_SPACE 8 /* space flag */ +#define FLAG_ALT 16 /* # flag */ +#define FLAG_ZERO 32 + +/* arg_index value indicating that no argument is consumed. */ +#define ARG_NONE (~(size_t)0) + +/* A parsed directive. */ +typedef struct +{ + const char* dir_start; + const char* dir_end; + int flags; + const char* width_start; + const char* width_end; + size_t width_arg_index; + const char* precision_start; + const char* precision_end; + size_t precision_arg_index; + char conversion; /* d i o u x X f e E g G c s p n U % but not C S */ + size_t arg_index; +} +char_directive; + +/* A parsed format string. */ +typedef struct +{ + size_t count; + char_directive *dir; + size_t max_width_length; + size_t max_precision_length; +} +char_directives; + + +/* Parses the format string. Fills in the number N of directives, and fills + in directives[0], ..., directives[N-1], and sets directives[N].dir_start + to the end of the format string. Also fills in the arg_type fields of the + arguments and the needed count of arguments. */ +#ifdef STATIC +STATIC +#else +extern +#endif +int printf_parse (const char *format, char_directives *d, arguments *a); + +#endif /* _PRINTF_PARSE_H */ diff --git a/gl/regcomp.c b/gl/regcomp.c new file mode 100644 index 00000000..8df6bb80 --- /dev/null +++ b/gl/regcomp.c @@ -0,0 +1,3858 @@ +/* Extended regular expression matching and search library. + Copyright (C) 2002,2003,2004,2005,2006 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +static reg_errcode_t re_compile_internal (regex_t *preg, const char * pattern, + size_t length, reg_syntax_t syntax); +static void re_compile_fastmap_iter (regex_t *bufp, + const re_dfastate_t *init_state, + char *fastmap); +static reg_errcode_t init_dfa (re_dfa_t *dfa, size_t pat_len); +#ifdef RE_ENABLE_I18N +static void free_charset (re_charset_t *cset); +#endif /* RE_ENABLE_I18N */ +static void free_workarea_compile (regex_t *preg); +static reg_errcode_t create_initial_state (re_dfa_t *dfa); +#ifdef RE_ENABLE_I18N +static void optimize_utf8 (re_dfa_t *dfa); +#endif +static reg_errcode_t analyze (regex_t *preg); +static reg_errcode_t preorder (bin_tree_t *root, + reg_errcode_t (fn (void *, bin_tree_t *)), + void *extra); +static reg_errcode_t postorder (bin_tree_t *root, + reg_errcode_t (fn (void *, bin_tree_t *)), + void *extra); +static reg_errcode_t optimize_subexps (void *extra, bin_tree_t *node); +static reg_errcode_t lower_subexps (void *extra, bin_tree_t *node); +static bin_tree_t *lower_subexp (reg_errcode_t *err, regex_t *preg, + bin_tree_t *node); +static reg_errcode_t calc_first (void *extra, bin_tree_t *node); +static reg_errcode_t calc_next (void *extra, bin_tree_t *node); +static reg_errcode_t link_nfa_nodes (void *extra, bin_tree_t *node); +static Idx duplicate_node (re_dfa_t *dfa, Idx org_idx, unsigned int constraint); +static Idx search_duplicated_node (const re_dfa_t *dfa, Idx org_node, + unsigned int constraint); +static reg_errcode_t calc_eclosure (re_dfa_t *dfa); +static reg_errcode_t calc_eclosure_iter (re_node_set *new_set, re_dfa_t *dfa, + Idx node, bool root); +static reg_errcode_t calc_inveclosure (re_dfa_t *dfa); +static Idx fetch_number (re_string_t *input, re_token_t *token, + reg_syntax_t syntax); +static int peek_token (re_token_t *token, re_string_t *input, + reg_syntax_t syntax) internal_function; +static bin_tree_t *parse (re_string_t *regexp, regex_t *preg, + reg_syntax_t syntax, reg_errcode_t *err); +static bin_tree_t *parse_reg_exp (re_string_t *regexp, regex_t *preg, + re_token_t *token, reg_syntax_t syntax, + Idx nest, reg_errcode_t *err); +static bin_tree_t *parse_branch (re_string_t *regexp, regex_t *preg, + re_token_t *token, reg_syntax_t syntax, + Idx nest, reg_errcode_t *err); +static bin_tree_t *parse_expression (re_string_t *regexp, regex_t *preg, + re_token_t *token, reg_syntax_t syntax, + Idx nest, reg_errcode_t *err); +static bin_tree_t *parse_sub_exp (re_string_t *regexp, regex_t *preg, + re_token_t *token, reg_syntax_t syntax, + Idx nest, reg_errcode_t *err); +static bin_tree_t *parse_dup_op (bin_tree_t *dup_elem, re_string_t *regexp, + re_dfa_t *dfa, re_token_t *token, + reg_syntax_t syntax, reg_errcode_t *err); +static bin_tree_t *parse_bracket_exp (re_string_t *regexp, re_dfa_t *dfa, + re_token_t *token, reg_syntax_t syntax, + reg_errcode_t *err); +static reg_errcode_t parse_bracket_element (bracket_elem_t *elem, + re_string_t *regexp, + re_token_t *token, int token_len, + re_dfa_t *dfa, + reg_syntax_t syntax, + bool accept_hyphen); +static reg_errcode_t parse_bracket_symbol (bracket_elem_t *elem, + re_string_t *regexp, + re_token_t *token); +#ifdef RE_ENABLE_I18N +static reg_errcode_t build_equiv_class (bitset_t sbcset, + re_charset_t *mbcset, + Idx *equiv_class_alloc, + const unsigned char *name); +static reg_errcode_t build_charclass (RE_TRANSLATE_TYPE trans, + bitset_t sbcset, + re_charset_t *mbcset, + Idx *char_class_alloc, + const unsigned char *class_name, + reg_syntax_t syntax); +#else /* not RE_ENABLE_I18N */ +static reg_errcode_t build_equiv_class (bitset_t sbcset, + const unsigned char *name); +static reg_errcode_t build_charclass (RE_TRANSLATE_TYPE trans, + bitset_t sbcset, + const unsigned char *class_name, + reg_syntax_t syntax); +#endif /* not RE_ENABLE_I18N */ +static bin_tree_t *build_charclass_op (re_dfa_t *dfa, + RE_TRANSLATE_TYPE trans, + const unsigned char *class_name, + const unsigned char *extra, + bool non_match, reg_errcode_t *err); +static bin_tree_t *create_tree (re_dfa_t *dfa, + bin_tree_t *left, bin_tree_t *right, + re_token_type_t type); +static bin_tree_t *create_token_tree (re_dfa_t *dfa, + bin_tree_t *left, bin_tree_t *right, + const re_token_t *token); +static bin_tree_t *duplicate_tree (const bin_tree_t *src, re_dfa_t *dfa); +static void free_token (re_token_t *node); +static reg_errcode_t free_tree (void *extra, bin_tree_t *node); +static reg_errcode_t mark_opt_subexp (void *extra, bin_tree_t *node); + +/* This table gives an error message for each of the error codes listed + in regex.h. Obviously the order here has to be same as there. + POSIX doesn't require that we do anything for REG_NOERROR, + but why not be nice? */ + +static const char __re_error_msgid[] = + { +#define REG_NOERROR_IDX 0 + gettext_noop ("Success") /* REG_NOERROR */ + "\0" +#define REG_NOMATCH_IDX (REG_NOERROR_IDX + sizeof "Success") + gettext_noop ("No match") /* REG_NOMATCH */ + "\0" +#define REG_BADPAT_IDX (REG_NOMATCH_IDX + sizeof "No match") + gettext_noop ("Invalid regular expression") /* REG_BADPAT */ + "\0" +#define REG_ECOLLATE_IDX (REG_BADPAT_IDX + sizeof "Invalid regular expression") + gettext_noop ("Invalid collation character") /* REG_ECOLLATE */ + "\0" +#define REG_ECTYPE_IDX (REG_ECOLLATE_IDX + sizeof "Invalid collation character") + gettext_noop ("Invalid character class name") /* REG_ECTYPE */ + "\0" +#define REG_EESCAPE_IDX (REG_ECTYPE_IDX + sizeof "Invalid character class name") + gettext_noop ("Trailing backslash") /* REG_EESCAPE */ + "\0" +#define REG_ESUBREG_IDX (REG_EESCAPE_IDX + sizeof "Trailing backslash") + gettext_noop ("Invalid back reference") /* REG_ESUBREG */ + "\0" +#define REG_EBRACK_IDX (REG_ESUBREG_IDX + sizeof "Invalid back reference") + gettext_noop ("Unmatched [ or [^") /* REG_EBRACK */ + "\0" +#define REG_EPAREN_IDX (REG_EBRACK_IDX + sizeof "Unmatched [ or [^") + gettext_noop ("Unmatched ( or \\(") /* REG_EPAREN */ + "\0" +#define REG_EBRACE_IDX (REG_EPAREN_IDX + sizeof "Unmatched ( or \\(") + gettext_noop ("Unmatched \\{") /* REG_EBRACE */ + "\0" +#define REG_BADBR_IDX (REG_EBRACE_IDX + sizeof "Unmatched \\{") + gettext_noop ("Invalid content of \\{\\}") /* REG_BADBR */ + "\0" +#define REG_ERANGE_IDX (REG_BADBR_IDX + sizeof "Invalid content of \\{\\}") + gettext_noop ("Invalid range end") /* REG_ERANGE */ + "\0" +#define REG_ESPACE_IDX (REG_ERANGE_IDX + sizeof "Invalid range end") + gettext_noop ("Memory exhausted") /* REG_ESPACE */ + "\0" +#define REG_BADRPT_IDX (REG_ESPACE_IDX + sizeof "Memory exhausted") + gettext_noop ("Invalid preceding regular expression") /* REG_BADRPT */ + "\0" +#define REG_EEND_IDX (REG_BADRPT_IDX + sizeof "Invalid preceding regular expression") + gettext_noop ("Premature end of regular expression") /* REG_EEND */ + "\0" +#define REG_ESIZE_IDX (REG_EEND_IDX + sizeof "Premature end of regular expression") + gettext_noop ("Regular expression too big") /* REG_ESIZE */ + "\0" +#define REG_ERPAREN_IDX (REG_ESIZE_IDX + sizeof "Regular expression too big") + gettext_noop ("Unmatched ) or \\)") /* REG_ERPAREN */ + }; + +static const size_t __re_error_msgid_idx[] = + { + REG_NOERROR_IDX, + REG_NOMATCH_IDX, + REG_BADPAT_IDX, + REG_ECOLLATE_IDX, + REG_ECTYPE_IDX, + REG_EESCAPE_IDX, + REG_ESUBREG_IDX, + REG_EBRACK_IDX, + REG_EPAREN_IDX, + REG_EBRACE_IDX, + REG_BADBR_IDX, + REG_ERANGE_IDX, + REG_ESPACE_IDX, + REG_BADRPT_IDX, + REG_EEND_IDX, + REG_ESIZE_IDX, + REG_ERPAREN_IDX + }; + +/* Entry points for GNU code. */ + +/* re_compile_pattern is the GNU regular expression compiler: it + compiles PATTERN (of length LENGTH) and puts the result in BUFP. + Returns 0 if the pattern was valid, otherwise an error string. + + Assumes the `allocated' (and perhaps `buffer') and `translate' fields + are set in BUFP on entry. */ + +#ifdef _LIBC +const char * +re_compile_pattern (pattern, length, bufp) + const char *pattern; + size_t length; + struct re_pattern_buffer *bufp; +#else /* size_t might promote */ +const char * +re_compile_pattern (const char *pattern, size_t length, + struct re_pattern_buffer *bufp) +#endif +{ + reg_errcode_t ret; + + /* And GNU code determines whether or not to get register information + by passing null for the REGS argument to re_match, etc., not by + setting no_sub, unless RE_NO_SUB is set. */ + bufp->no_sub = !!(re_syntax_options & RE_NO_SUB); + + /* Match anchors at newline. */ + bufp->newline_anchor = 1; + + ret = re_compile_internal (bufp, pattern, length, re_syntax_options); + + if (!ret) + return NULL; + return gettext (__re_error_msgid + __re_error_msgid_idx[(int) ret]); +} +#ifdef _LIBC +weak_alias (__re_compile_pattern, re_compile_pattern) +#endif + +/* Set by `re_set_syntax' to the current regexp syntax to recognize. Can + also be assigned to arbitrarily: each pattern buffer stores its own + syntax, so it can be changed between regex compilations. */ +/* This has no initializer because initialized variables in Emacs + become read-only after dumping. */ +reg_syntax_t re_syntax_options; + + +/* Specify the precise syntax of regexps for compilation. This provides + for compatibility for various utilities which historically have + different, incompatible syntaxes. + + The argument SYNTAX is a bit mask comprised of the various bits + defined in regex.h. We return the old syntax. */ + +reg_syntax_t +re_set_syntax (syntax) + reg_syntax_t syntax; +{ + reg_syntax_t ret = re_syntax_options; + + re_syntax_options = syntax; + return ret; +} +#ifdef _LIBC +weak_alias (__re_set_syntax, re_set_syntax) +#endif + +int +re_compile_fastmap (bufp) + struct re_pattern_buffer *bufp; +{ + re_dfa_t *dfa = (re_dfa_t *) bufp->buffer; + char *fastmap = bufp->fastmap; + + memset (fastmap, '\0', sizeof (char) * SBC_MAX); + re_compile_fastmap_iter (bufp, dfa->init_state, fastmap); + if (dfa->init_state != dfa->init_state_word) + re_compile_fastmap_iter (bufp, dfa->init_state_word, fastmap); + if (dfa->init_state != dfa->init_state_nl) + re_compile_fastmap_iter (bufp, dfa->init_state_nl, fastmap); + if (dfa->init_state != dfa->init_state_begbuf) + re_compile_fastmap_iter (bufp, dfa->init_state_begbuf, fastmap); + bufp->fastmap_accurate = 1; + return 0; +} +#ifdef _LIBC +weak_alias (__re_compile_fastmap, re_compile_fastmap) +#endif + +static inline void +__attribute ((always_inline)) +re_set_fastmap (char *fastmap, bool icase, int ch) +{ + fastmap[ch] = 1; + if (icase) + fastmap[tolower (ch)] = 1; +} + +/* Helper function for re_compile_fastmap. + Compile fastmap for the initial_state INIT_STATE. */ + +static void +re_compile_fastmap_iter (regex_t *bufp, const re_dfastate_t *init_state, + char *fastmap) +{ + re_dfa_t *dfa = (re_dfa_t *) bufp->buffer; + Idx node_cnt; + bool icase = (dfa->mb_cur_max == 1 && (bufp->syntax & RE_ICASE)); + for (node_cnt = 0; node_cnt < init_state->nodes.nelem; ++node_cnt) + { + Idx node = init_state->nodes.elems[node_cnt]; + re_token_type_t type = dfa->nodes[node].type; + + if (type == CHARACTER) + { + re_set_fastmap (fastmap, icase, dfa->nodes[node].opr.c); +#ifdef RE_ENABLE_I18N + if ((bufp->syntax & RE_ICASE) && dfa->mb_cur_max > 1) + { + unsigned char buf[MB_LEN_MAX]; + unsigned char *p; + wchar_t wc; + mbstate_t state; + + p = buf; + *p++ = dfa->nodes[node].opr.c; + while (++node < dfa->nodes_len + && dfa->nodes[node].type == CHARACTER + && dfa->nodes[node].mb_partial) + *p++ = dfa->nodes[node].opr.c; + memset (&state, '\0', sizeof (state)); + if (mbrtowc (&wc, (const char *) buf, p - buf, + &state) == p - buf + && (__wcrtomb ((char *) buf, towlower (wc), &state) + != (size_t) -1)) + re_set_fastmap (fastmap, false, buf[0]); + } +#endif + } + else if (type == SIMPLE_BRACKET) + { + int i, ch; + for (i = 0, ch = 0; i < BITSET_WORDS; ++i) + { + int j; + bitset_word_t w = dfa->nodes[node].opr.sbcset[i]; + for (j = 0; j < BITSET_WORD_BITS; ++j, ++ch) + if (w & ((bitset_word_t) 1 << j)) + re_set_fastmap (fastmap, icase, ch); + } + } +#ifdef RE_ENABLE_I18N + else if (type == COMPLEX_BRACKET) + { + Idx i; + re_charset_t *cset = dfa->nodes[node].opr.mbcset; + if (cset->non_match || cset->ncoll_syms || cset->nequiv_classes + || cset->nranges || cset->nchar_classes) + { +# ifdef _LIBC + if (_NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES) != 0) + { + /* In this case we want to catch the bytes which are + the first byte of any collation elements. + e.g. In da_DK, we want to catch 'a' since "aa" + is a valid collation element, and don't catch + 'b' since 'b' is the only collation element + which starts from 'b'. */ + const int32_t *table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB); + for (i = 0; i < SBC_MAX; ++i) + if (table[i] < 0) + re_set_fastmap (fastmap, icase, i); + } +# else + if (dfa->mb_cur_max > 1) + for (i = 0; i < SBC_MAX; ++i) + if (__btowc (i) == WEOF) + re_set_fastmap (fastmap, icase, i); +# endif /* not _LIBC */ + } + for (i = 0; i < cset->nmbchars; ++i) + { + char buf[256]; + mbstate_t state; + memset (&state, '\0', sizeof (state)); + if (__wcrtomb (buf, cset->mbchars[i], &state) != (size_t) -1) + re_set_fastmap (fastmap, icase, *(unsigned char *) buf); + if ((bufp->syntax & RE_ICASE) && dfa->mb_cur_max > 1) + { + if (__wcrtomb (buf, towlower (cset->mbchars[i]), &state) + != (size_t) -1) + re_set_fastmap (fastmap, false, *(unsigned char *) buf); + } + } + } +#endif /* RE_ENABLE_I18N */ + else if (type == OP_PERIOD +#ifdef RE_ENABLE_I18N + || type == OP_UTF8_PERIOD +#endif /* RE_ENABLE_I18N */ + || type == END_OF_RE) + { + memset (fastmap, '\1', sizeof (char) * SBC_MAX); + if (type == END_OF_RE) + bufp->can_be_null = 1; + return; + } + } +} + +/* Entry point for POSIX code. */ +/* regcomp takes a regular expression as a string and compiles it. + + PREG is a regex_t *. We do not expect any fields to be initialized, + since POSIX says we shouldn't. Thus, we set + + `buffer' to the compiled pattern; + `used' to the length of the compiled pattern; + `syntax' to RE_SYNTAX_POSIX_EXTENDED if the + REG_EXTENDED bit in CFLAGS is set; otherwise, to + RE_SYNTAX_POSIX_BASIC; + `newline_anchor' to REG_NEWLINE being set in CFLAGS; + `fastmap' to an allocated space for the fastmap; + `fastmap_accurate' to zero; + `re_nsub' to the number of subexpressions in PATTERN. + + PATTERN is the address of the pattern string. + + CFLAGS is a series of bits which affect compilation. + + If REG_EXTENDED is set, we use POSIX extended syntax; otherwise, we + use POSIX basic syntax. + + If REG_NEWLINE is set, then . and [^...] don't match newline. + Also, regexec will try a match beginning after every newline. + + If REG_ICASE is set, then we considers upper- and lowercase + versions of letters to be equivalent when matching. + + If REG_NOSUB is set, then when PREG is passed to regexec, that + routine will report only success or failure, and nothing about the + registers. + + It returns 0 if it succeeds, nonzero if it doesn't. (See regex.h for + the return codes and their meanings.) */ + +int +regcomp (preg, pattern, cflags) + regex_t *__restrict preg; + const char *__restrict pattern; + int cflags; +{ + reg_errcode_t ret; + reg_syntax_t syntax = ((cflags & REG_EXTENDED) ? RE_SYNTAX_POSIX_EXTENDED + : RE_SYNTAX_POSIX_BASIC); + + preg->buffer = NULL; + preg->allocated = 0; + preg->used = 0; + + /* Try to allocate space for the fastmap. */ + preg->fastmap = re_malloc (char, SBC_MAX); + if (BE (preg->fastmap == NULL, 0)) + return REG_ESPACE; + + syntax |= (cflags & REG_ICASE) ? RE_ICASE : 0; + + /* If REG_NEWLINE is set, newlines are treated differently. */ + if (cflags & REG_NEWLINE) + { /* REG_NEWLINE implies neither . nor [^...] match newline. */ + syntax &= ~RE_DOT_NEWLINE; + syntax |= RE_HAT_LISTS_NOT_NEWLINE; + /* It also changes the matching behavior. */ + preg->newline_anchor = 1; + } + else + preg->newline_anchor = 0; + preg->no_sub = !!(cflags & REG_NOSUB); + preg->translate = NULL; + + ret = re_compile_internal (preg, pattern, strlen (pattern), syntax); + + /* POSIX doesn't distinguish between an unmatched open-group and an + unmatched close-group: both are REG_EPAREN. */ + if (ret == REG_ERPAREN) + ret = REG_EPAREN; + + /* We have already checked preg->fastmap != NULL. */ + if (BE (ret == REG_NOERROR, 1)) + /* Compute the fastmap now, since regexec cannot modify the pattern + buffer. This function never fails in this implementation. */ + (void) re_compile_fastmap (preg); + else + { + /* Some error occurred while compiling the expression. */ + re_free (preg->fastmap); + preg->fastmap = NULL; + } + + return (int) ret; +} +#ifdef _LIBC +weak_alias (__regcomp, regcomp) +#endif + +/* Returns a message corresponding to an error code, ERRCODE, returned + from either regcomp or regexec. We don't use PREG here. */ + +#ifdef _LIBC +size_t +regerror (errcode, preg, errbuf, errbuf_size) + int errcode; + const regex_t *__restrict preg; + char *__restrict errbuf; + size_t errbuf_size; +#else /* size_t might promote */ +size_t +regerror (int errcode, const regex_t *__restrict preg, + char *__restrict errbuf, size_t errbuf_size) +#endif +{ + const char *msg; + size_t msg_size; + + if (BE (errcode < 0 + || errcode >= (int) (sizeof (__re_error_msgid_idx) + / sizeof (__re_error_msgid_idx[0])), 0)) + /* Only error codes returned by the rest of the code should be passed + to this routine. If we are given anything else, or if other regex + code generates an invalid error code, then the program has a bug. + Dump core so we can fix it. */ + abort (); + + msg = gettext (__re_error_msgid + __re_error_msgid_idx[errcode]); + + msg_size = strlen (msg) + 1; /* Includes the null. */ + + if (BE (errbuf_size != 0, 1)) + { + if (BE (msg_size > errbuf_size, 0)) + { +#if defined HAVE_MEMPCPY || defined _LIBC + *((char *) __mempcpy (errbuf, msg, errbuf_size - 1)) = '\0'; +#else + memcpy (errbuf, msg, errbuf_size - 1); + errbuf[errbuf_size - 1] = 0; +#endif + } + else + memcpy (errbuf, msg, msg_size); + } + + return msg_size; +} +#ifdef _LIBC +weak_alias (__regerror, regerror) +#endif + + +#ifdef RE_ENABLE_I18N +/* This static array is used for the map to single-byte characters when + UTF-8 is used. Otherwise we would allocate memory just to initialize + it the same all the time. UTF-8 is the preferred encoding so this is + a worthwhile optimization. */ +static const bitset_t utf8_sb_map = +{ + /* Set the first 128 bits. */ +# if 4 * BITSET_WORD_BITS < ASCII_CHARS +# error "bitset_word_t is narrower than 32 bits" +# elif 3 * BITSET_WORD_BITS < ASCII_CHARS + BITSET_WORD_MAX, BITSET_WORD_MAX, BITSET_WORD_MAX, +# elif 2 * BITSET_WORD_BITS < ASCII_CHARS + BITSET_WORD_MAX, BITSET_WORD_MAX, +# elif 1 * BITSET_WORD_BITS < ASCII_CHARS + BITSET_WORD_MAX, +# endif + (BITSET_WORD_MAX + >> (SBC_MAX % BITSET_WORD_BITS == 0 + ? 0 + : BITSET_WORD_BITS - SBC_MAX % BITSET_WORD_BITS)) +}; +#endif + + +static void +free_dfa_content (re_dfa_t *dfa) +{ + Idx i, j; + + if (dfa->nodes) + for (i = 0; i < dfa->nodes_len; ++i) + free_token (dfa->nodes + i); + re_free (dfa->nexts); + for (i = 0; i < dfa->nodes_len; ++i) + { + if (dfa->eclosures != NULL) + re_node_set_free (dfa->eclosures + i); + if (dfa->inveclosures != NULL) + re_node_set_free (dfa->inveclosures + i); + if (dfa->edests != NULL) + re_node_set_free (dfa->edests + i); + } + re_free (dfa->edests); + re_free (dfa->eclosures); + re_free (dfa->inveclosures); + re_free (dfa->nodes); + + if (dfa->state_table) + for (i = 0; i <= dfa->state_hash_mask; ++i) + { + struct re_state_table_entry *entry = dfa->state_table + i; + for (j = 0; j < entry->num; ++j) + { + re_dfastate_t *state = entry->array[j]; + free_state (state); + } + re_free (entry->array); + } + re_free (dfa->state_table); +#ifdef RE_ENABLE_I18N + if (dfa->sb_char != utf8_sb_map) + re_free (dfa->sb_char); +#endif + re_free (dfa->subexp_map); +#ifdef DEBUG + re_free (dfa->re_str); +#endif + + re_free (dfa); +} + + +/* Free dynamically allocated space used by PREG. */ + +void +regfree (preg) + regex_t *preg; +{ + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + if (BE (dfa != NULL, 1)) + free_dfa_content (dfa); + preg->buffer = NULL; + preg->allocated = 0; + + re_free (preg->fastmap); + preg->fastmap = NULL; + + re_free (preg->translate); + preg->translate = NULL; +} +#ifdef _LIBC +weak_alias (__regfree, regfree) +#endif + +/* Entry points compatible with 4.2 BSD regex library. We don't define + them unless specifically requested. */ + +#if defined _REGEX_RE_COMP || defined _LIBC + +/* BSD has one and only one pattern buffer. */ +static struct re_pattern_buffer re_comp_buf; + +char * +# ifdef _LIBC +/* Make these definitions weak in libc, so POSIX programs can redefine + these names if they don't use our functions, and still use + regcomp/regexec above without link errors. */ +weak_function +# endif +re_comp (s) + const char *s; +{ + reg_errcode_t ret; + char *fastmap; + + if (!s) + { + if (!re_comp_buf.buffer) + return gettext ("No previous regular expression"); + return 0; + } + + if (re_comp_buf.buffer) + { + fastmap = re_comp_buf.fastmap; + re_comp_buf.fastmap = NULL; + __regfree (&re_comp_buf); + memset (&re_comp_buf, '\0', sizeof (re_comp_buf)); + re_comp_buf.fastmap = fastmap; + } + + if (re_comp_buf.fastmap == NULL) + { + re_comp_buf.fastmap = (char *) malloc (SBC_MAX); + if (re_comp_buf.fastmap == NULL) + return (char *) gettext (__re_error_msgid + + __re_error_msgid_idx[(int) REG_ESPACE]); + } + + /* Since `re_exec' always passes NULL for the `regs' argument, we + don't need to initialize the pattern buffer fields which affect it. */ + + /* Match anchors at newlines. */ + re_comp_buf.newline_anchor = 1; + + ret = re_compile_internal (&re_comp_buf, s, strlen (s), re_syntax_options); + + if (!ret) + return NULL; + + /* Yes, we're discarding `const' here if !HAVE_LIBINTL. */ + return (char *) gettext (__re_error_msgid + __re_error_msgid_idx[(int) ret]); +} + +#ifdef _LIBC +libc_freeres_fn (free_mem) +{ + __regfree (&re_comp_buf); +} +#endif + +#endif /* _REGEX_RE_COMP */ + +/* Internal entry point. + Compile the regular expression PATTERN, whose length is LENGTH. + SYNTAX indicate regular expression's syntax. */ + +static reg_errcode_t +re_compile_internal (regex_t *preg, const char * pattern, size_t length, + reg_syntax_t syntax) +{ + reg_errcode_t err = REG_NOERROR; + re_dfa_t *dfa; + re_string_t regexp; + + /* Initialize the pattern buffer. */ + preg->fastmap_accurate = 0; + preg->syntax = syntax; + preg->not_bol = preg->not_eol = 0; + preg->used = 0; + preg->re_nsub = 0; + preg->can_be_null = 0; + preg->regs_allocated = REGS_UNALLOCATED; + + /* Initialize the dfa. */ + dfa = (re_dfa_t *) preg->buffer; + if (BE (preg->allocated < sizeof (re_dfa_t), 0)) + { + /* If zero allocated, but buffer is non-null, try to realloc + enough space. This loses if buffer's address is bogus, but + that is the user's responsibility. If ->buffer is NULL this + is a simple allocation. */ + dfa = re_realloc (preg->buffer, re_dfa_t, 1); + if (dfa == NULL) + return REG_ESPACE; + preg->allocated = sizeof (re_dfa_t); + preg->buffer = (unsigned char *) dfa; + } + preg->used = sizeof (re_dfa_t); + + err = init_dfa (dfa, length); + if (BE (err != REG_NOERROR, 0)) + { + free_dfa_content (dfa); + preg->buffer = NULL; + preg->allocated = 0; + return err; + } +#ifdef DEBUG + /* Note: length+1 will not overflow since it is checked in init_dfa. */ + dfa->re_str = re_malloc (char, length + 1); + strncpy (dfa->re_str, pattern, length + 1); +#endif + + __libc_lock_init (dfa->lock); + + err = re_string_construct (®exp, pattern, length, preg->translate, + syntax & RE_ICASE, dfa); + if (BE (err != REG_NOERROR, 0)) + { + re_compile_internal_free_return: + free_workarea_compile (preg); + re_string_destruct (®exp); + free_dfa_content (dfa); + preg->buffer = NULL; + preg->allocated = 0; + return err; + } + + /* Parse the regular expression, and build a structure tree. */ + preg->re_nsub = 0; + dfa->str_tree = parse (®exp, preg, syntax, &err); + if (BE (dfa->str_tree == NULL, 0)) + goto re_compile_internal_free_return; + + /* Analyze the tree and create the nfa. */ + err = analyze (preg); + if (BE (err != REG_NOERROR, 0)) + goto re_compile_internal_free_return; + +#ifdef RE_ENABLE_I18N + /* If possible, do searching in single byte encoding to speed things up. */ + if (dfa->is_utf8 && !(syntax & RE_ICASE) && preg->translate == NULL) + optimize_utf8 (dfa); +#endif + + /* Then create the initial state of the dfa. */ + err = create_initial_state (dfa); + + /* Release work areas. */ + free_workarea_compile (preg); + re_string_destruct (®exp); + + if (BE (err != REG_NOERROR, 0)) + { + free_dfa_content (dfa); + preg->buffer = NULL; + preg->allocated = 0; + } + + return err; +} + +/* Initialize DFA. We use the length of the regular expression PAT_LEN + as the initial length of some arrays. */ + +static reg_errcode_t +init_dfa (re_dfa_t *dfa, size_t pat_len) +{ + __re_size_t table_size; +#ifndef _LIBC + char *codeset_name; +#endif +#ifdef RE_ENABLE_I18N + size_t max_i18n_object_size = MAX (sizeof (wchar_t), sizeof (wctype_t)); +#else + size_t max_i18n_object_size = 0; +#endif + size_t max_object_size = + MAX (sizeof (struct re_state_table_entry), + MAX (sizeof (re_token_t), + MAX (sizeof (re_node_set), + MAX (sizeof (regmatch_t), + max_i18n_object_size)))); + + memset (dfa, '\0', sizeof (re_dfa_t)); + + /* Force allocation of str_tree_storage the first time. */ + dfa->str_tree_storage_idx = BIN_TREE_STORAGE_SIZE; + + /* Avoid overflows. The extra "/ 2" is for the table_size doubling + calculation below, and for similar doubling calculations + elsewhere. And it's <= rather than <, because some of the + doubling calculations add 1 afterwards. */ + if (BE (SIZE_MAX / max_object_size / 2 <= pat_len, 0)) + return REG_ESPACE; + + dfa->nodes_alloc = pat_len + 1; + dfa->nodes = re_malloc (re_token_t, dfa->nodes_alloc); + + /* table_size = 2 ^ ceil(log pat_len) */ + for (table_size = 1; ; table_size <<= 1) + if (table_size > pat_len) + break; + + dfa->state_table = calloc (sizeof (struct re_state_table_entry), table_size); + dfa->state_hash_mask = table_size - 1; + + dfa->mb_cur_max = MB_CUR_MAX; +#ifdef _LIBC + if (dfa->mb_cur_max == 6 + && strcmp (_NL_CURRENT (LC_CTYPE, _NL_CTYPE_CODESET_NAME), "UTF-8") == 0) + dfa->is_utf8 = 1; + dfa->map_notascii = (_NL_CURRENT_WORD (LC_CTYPE, _NL_CTYPE_MAP_TO_NONASCII) + != 0); +#else +# ifdef HAVE_LANGINFO_CODESET + codeset_name = nl_langinfo (CODESET); +# else + codeset_name = getenv ("LC_ALL"); + if (codeset_name == NULL || codeset_name[0] == '\0') + codeset_name = getenv ("LC_CTYPE"); + if (codeset_name == NULL || codeset_name[0] == '\0') + codeset_name = getenv ("LANG"); + if (codeset_name == NULL) + codeset_name = ""; + else if (strchr (codeset_name, '.') != NULL) + codeset_name = strchr (codeset_name, '.') + 1; +# endif + + if (strcasecmp (codeset_name, "UTF-8") == 0 + || strcasecmp (codeset_name, "UTF8") == 0) + dfa->is_utf8 = 1; + + /* We check exhaustively in the loop below if this charset is a + superset of ASCII. */ + dfa->map_notascii = 0; +#endif + +#ifdef RE_ENABLE_I18N + if (dfa->mb_cur_max > 1) + { + if (dfa->is_utf8) + dfa->sb_char = (re_bitset_ptr_t) utf8_sb_map; + else + { + int i, j, ch; + + dfa->sb_char = (re_bitset_ptr_t) calloc (sizeof (bitset_t), 1); + if (BE (dfa->sb_char == NULL, 0)) + return REG_ESPACE; + + /* Set the bits corresponding to single byte chars. */ + for (i = 0, ch = 0; i < BITSET_WORDS; ++i) + for (j = 0; j < BITSET_WORD_BITS; ++j, ++ch) + { + wint_t wch = __btowc (ch); + if (wch != WEOF) + dfa->sb_char[i] |= (bitset_word_t) 1 << j; +# ifndef _LIBC + if (isascii (ch) && wch != ch) + dfa->map_notascii = 1; +# endif + } + } + } +#endif + + if (BE (dfa->nodes == NULL || dfa->state_table == NULL, 0)) + return REG_ESPACE; + return REG_NOERROR; +} + +/* Initialize WORD_CHAR table, which indicate which character is + "word". In this case "word" means that it is the word construction + character used by some operators like "\<", "\>", etc. */ + +static void +internal_function +init_word_char (re_dfa_t *dfa) +{ + int i, j, ch; + dfa->word_ops_used = 1; + for (i = 0, ch = 0; i < BITSET_WORDS; ++i) + for (j = 0; j < BITSET_WORD_BITS; ++j, ++ch) + if (isalnum (ch) || ch == '_') + dfa->word_char[i] |= (bitset_word_t) 1 << j; +} + +/* Free the work area which are only used while compiling. */ + +static void +free_workarea_compile (regex_t *preg) +{ + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + bin_tree_storage_t *storage, *next; + for (storage = dfa->str_tree_storage; storage; storage = next) + { + next = storage->next; + re_free (storage); + } + dfa->str_tree_storage = NULL; + dfa->str_tree_storage_idx = BIN_TREE_STORAGE_SIZE; + dfa->str_tree = NULL; + re_free (dfa->org_indices); + dfa->org_indices = NULL; +} + +/* Create initial states for all contexts. */ + +static reg_errcode_t +create_initial_state (re_dfa_t *dfa) +{ + Idx first, i; + reg_errcode_t err; + re_node_set init_nodes; + + /* Initial states have the epsilon closure of the node which is + the first node of the regular expression. */ + first = dfa->str_tree->first->node_idx; + dfa->init_node = first; + err = re_node_set_init_copy (&init_nodes, dfa->eclosures + first); + if (BE (err != REG_NOERROR, 0)) + return err; + + /* The back-references which are in initial states can epsilon transit, + since in this case all of the subexpressions can be null. + Then we add epsilon closures of the nodes which are the next nodes of + the back-references. */ + if (dfa->nbackref > 0) + for (i = 0; i < init_nodes.nelem; ++i) + { + Idx node_idx = init_nodes.elems[i]; + re_token_type_t type = dfa->nodes[node_idx].type; + + Idx clexp_idx; + if (type != OP_BACK_REF) + continue; + for (clexp_idx = 0; clexp_idx < init_nodes.nelem; ++clexp_idx) + { + re_token_t *clexp_node; + clexp_node = dfa->nodes + init_nodes.elems[clexp_idx]; + if (clexp_node->type == OP_CLOSE_SUBEXP + && clexp_node->opr.idx == dfa->nodes[node_idx].opr.idx) + break; + } + if (clexp_idx == init_nodes.nelem) + continue; + + if (type == OP_BACK_REF) + { + Idx dest_idx = dfa->edests[node_idx].elems[0]; + if (!re_node_set_contains (&init_nodes, dest_idx)) + { + re_node_set_merge (&init_nodes, dfa->eclosures + dest_idx); + i = 0; + } + } + } + + /* It must be the first time to invoke acquire_state. */ + dfa->init_state = re_acquire_state_context (&err, dfa, &init_nodes, 0); + /* We don't check ERR here, since the initial state must not be NULL. */ + if (BE (dfa->init_state == NULL, 0)) + return err; + if (dfa->init_state->has_constraint) + { + dfa->init_state_word = re_acquire_state_context (&err, dfa, &init_nodes, + CONTEXT_WORD); + dfa->init_state_nl = re_acquire_state_context (&err, dfa, &init_nodes, + CONTEXT_NEWLINE); + dfa->init_state_begbuf = re_acquire_state_context (&err, dfa, + &init_nodes, + CONTEXT_NEWLINE + | CONTEXT_BEGBUF); + if (BE (dfa->init_state_word == NULL || dfa->init_state_nl == NULL + || dfa->init_state_begbuf == NULL, 0)) + return err; + } + else + dfa->init_state_word = dfa->init_state_nl + = dfa->init_state_begbuf = dfa->init_state; + + re_node_set_free (&init_nodes); + return REG_NOERROR; +} + +#ifdef RE_ENABLE_I18N +/* If it is possible to do searching in single byte encoding instead of UTF-8 + to speed things up, set dfa->mb_cur_max to 1, clear is_utf8 and change + DFA nodes where needed. */ + +static void +optimize_utf8 (re_dfa_t *dfa) +{ + Idx node; + int i; + bool mb_chars = false; + bool has_period = false; + + for (node = 0; node < dfa->nodes_len; ++node) + switch (dfa->nodes[node].type) + { + case CHARACTER: + if (dfa->nodes[node].opr.c >= ASCII_CHARS) + mb_chars = true; + break; + case ANCHOR: + switch (dfa->nodes[node].opr.idx) + { + case LINE_FIRST: + case LINE_LAST: + case BUF_FIRST: + case BUF_LAST: + break; + default: + /* Word anchors etc. cannot be handled. */ + return; + } + break; + case OP_PERIOD: + has_period = true; + break; + case OP_BACK_REF: + case OP_ALT: + case END_OF_RE: + case OP_DUP_ASTERISK: + case OP_OPEN_SUBEXP: + case OP_CLOSE_SUBEXP: + break; + case COMPLEX_BRACKET: + return; + case SIMPLE_BRACKET: + /* Just double check. */ + { + int rshift = (ASCII_CHARS % BITSET_WORD_BITS == 0 + ? 0 + : BITSET_WORD_BITS - ASCII_CHARS % BITSET_WORD_BITS); + for (i = ASCII_CHARS / BITSET_WORD_BITS; i < BITSET_WORDS; ++i) + { + if (dfa->nodes[node].opr.sbcset[i] >> rshift != 0) + return; + rshift = 0; + } + } + break; + default: + abort (); + } + + if (mb_chars || has_period) + for (node = 0; node < dfa->nodes_len; ++node) + { + if (dfa->nodes[node].type == CHARACTER + && dfa->nodes[node].opr.c >= ASCII_CHARS) + dfa->nodes[node].mb_partial = 0; + else if (dfa->nodes[node].type == OP_PERIOD) + dfa->nodes[node].type = OP_UTF8_PERIOD; + } + + /* The search can be in single byte locale. */ + dfa->mb_cur_max = 1; + dfa->is_utf8 = 0; + dfa->has_mb_node = dfa->nbackref > 0 || has_period; +} +#endif + +/* Analyze the structure tree, and calculate "first", "next", "edest", + "eclosure", and "inveclosure". */ + +static reg_errcode_t +analyze (regex_t *preg) +{ + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + reg_errcode_t ret; + + /* Allocate arrays. */ + dfa->nexts = re_malloc (Idx, dfa->nodes_alloc); + dfa->org_indices = re_malloc (Idx, dfa->nodes_alloc); + dfa->edests = re_malloc (re_node_set, dfa->nodes_alloc); + dfa->eclosures = re_malloc (re_node_set, dfa->nodes_alloc); + if (BE (dfa->nexts == NULL || dfa->org_indices == NULL || dfa->edests == NULL + || dfa->eclosures == NULL, 0)) + return REG_ESPACE; + + dfa->subexp_map = re_malloc (Idx, preg->re_nsub); + if (dfa->subexp_map != NULL) + { + Idx i; + for (i = 0; i < preg->re_nsub; i++) + dfa->subexp_map[i] = i; + preorder (dfa->str_tree, optimize_subexps, dfa); + for (i = 0; i < preg->re_nsub; i++) + if (dfa->subexp_map[i] != i) + break; + if (i == preg->re_nsub) + { + free (dfa->subexp_map); + dfa->subexp_map = NULL; + } + } + + ret = postorder (dfa->str_tree, lower_subexps, preg); + if (BE (ret != REG_NOERROR, 0)) + return ret; + ret = postorder (dfa->str_tree, calc_first, dfa); + if (BE (ret != REG_NOERROR, 0)) + return ret; + preorder (dfa->str_tree, calc_next, dfa); + ret = preorder (dfa->str_tree, link_nfa_nodes, dfa); + if (BE (ret != REG_NOERROR, 0)) + return ret; + ret = calc_eclosure (dfa); + if (BE (ret != REG_NOERROR, 0)) + return ret; + + /* We only need this during the prune_impossible_nodes pass in regexec.c; + skip it if p_i_n will not run, as calc_inveclosure can be quadratic. */ + if ((!preg->no_sub && preg->re_nsub > 0 && dfa->has_plural_match) + || dfa->nbackref) + { + dfa->inveclosures = re_malloc (re_node_set, dfa->nodes_len); + if (BE (dfa->inveclosures == NULL, 0)) + return REG_ESPACE; + ret = calc_inveclosure (dfa); + } + + return ret; +} + +/* Our parse trees are very unbalanced, so we cannot use a stack to + implement parse tree visits. Instead, we use parent pointers and + some hairy code in these two functions. */ +static reg_errcode_t +postorder (bin_tree_t *root, reg_errcode_t (fn (void *, bin_tree_t *)), + void *extra) +{ + bin_tree_t *node, *prev; + + for (node = root; ; ) + { + /* Descend down the tree, preferably to the left (or to the right + if that's the only child). */ + while (node->left || node->right) + if (node->left) + node = node->left; + else + node = node->right; + + do + { + reg_errcode_t err = fn (extra, node); + if (BE (err != REG_NOERROR, 0)) + return err; + if (node->parent == NULL) + return REG_NOERROR; + prev = node; + node = node->parent; + } + /* Go up while we have a node that is reached from the right. */ + while (node->right == prev || node->right == NULL); + node = node->right; + } +} + +static reg_errcode_t +preorder (bin_tree_t *root, reg_errcode_t (fn (void *, bin_tree_t *)), + void *extra) +{ + bin_tree_t *node; + + for (node = root; ; ) + { + reg_errcode_t err = fn (extra, node); + if (BE (err != REG_NOERROR, 0)) + return err; + + /* Go to the left node, or up and to the right. */ + if (node->left) + node = node->left; + else + { + bin_tree_t *prev = NULL; + while (node->right == prev || node->right == NULL) + { + prev = node; + node = node->parent; + if (!node) + return REG_NOERROR; + } + node = node->right; + } + } +} + +/* Optimization pass: if a SUBEXP is entirely contained, strip it and tell + re_search_internal to map the inner one's opr.idx to this one's. Adjust + backreferences as well. Requires a preorder visit. */ +static reg_errcode_t +optimize_subexps (void *extra, bin_tree_t *node) +{ + re_dfa_t *dfa = (re_dfa_t *) extra; + + if (node->token.type == OP_BACK_REF && dfa->subexp_map) + { + int idx = node->token.opr.idx; + node->token.opr.idx = dfa->subexp_map[idx]; + dfa->used_bkref_map |= 1 << node->token.opr.idx; + } + + else if (node->token.type == SUBEXP + && node->left && node->left->token.type == SUBEXP) + { + Idx other_idx = node->left->token.opr.idx; + + node->left = node->left->left; + if (node->left) + node->left->parent = node; + + dfa->subexp_map[other_idx] = dfa->subexp_map[node->token.opr.idx]; + if (other_idx < BITSET_WORD_BITS) + dfa->used_bkref_map &= ~((bitset_word_t) 1 << other_idx); + } + + return REG_NOERROR; +} + +/* Lowering pass: Turn each SUBEXP node into the appropriate concatenation + of OP_OPEN_SUBEXP, the body of the SUBEXP (if any) and OP_CLOSE_SUBEXP. */ +static reg_errcode_t +lower_subexps (void *extra, bin_tree_t *node) +{ + regex_t *preg = (regex_t *) extra; + reg_errcode_t err = REG_NOERROR; + + if (node->left && node->left->token.type == SUBEXP) + { + node->left = lower_subexp (&err, preg, node->left); + if (node->left) + node->left->parent = node; + } + if (node->right && node->right->token.type == SUBEXP) + { + node->right = lower_subexp (&err, preg, node->right); + if (node->right) + node->right->parent = node; + } + + return err; +} + +static bin_tree_t * +lower_subexp (reg_errcode_t *err, regex_t *preg, bin_tree_t *node) +{ + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + bin_tree_t *body = node->left; + bin_tree_t *op, *cls, *tree1, *tree; + + if (preg->no_sub + /* We do not optimize empty subexpressions, because otherwise we may + have bad CONCAT nodes with NULL children. This is obviously not + very common, so we do not lose much. An example that triggers + this case is the sed "script" /\(\)/x. */ + && node->left != NULL + && (node->token.opr.idx >= BITSET_WORD_BITS + || !(dfa->used_bkref_map + & ((bitset_word_t) 1 << node->token.opr.idx)))) + return node->left; + + /* Convert the SUBEXP node to the concatenation of an + OP_OPEN_SUBEXP, the contents, and an OP_CLOSE_SUBEXP. */ + op = create_tree (dfa, NULL, NULL, OP_OPEN_SUBEXP); + cls = create_tree (dfa, NULL, NULL, OP_CLOSE_SUBEXP); + tree1 = body ? create_tree (dfa, body, cls, CONCAT) : cls; + tree = create_tree (dfa, op, tree1, CONCAT); + if (BE (tree == NULL || tree1 == NULL || op == NULL || cls == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + + op->token.opr.idx = cls->token.opr.idx = node->token.opr.idx; + op->token.opt_subexp = cls->token.opt_subexp = node->token.opt_subexp; + return tree; +} + +/* Pass 1 in building the NFA: compute FIRST and create unlinked automaton + nodes. Requires a postorder visit. */ +static reg_errcode_t +calc_first (void *extra, bin_tree_t *node) +{ + re_dfa_t *dfa = (re_dfa_t *) extra; + if (node->token.type == CONCAT) + { + node->first = node->left->first; + node->node_idx = node->left->node_idx; + } + else + { + node->first = node; + node->node_idx = re_dfa_add_node (dfa, node->token); + if (BE (node->node_idx == REG_MISSING, 0)) + return REG_ESPACE; + } + return REG_NOERROR; +} + +/* Pass 2: compute NEXT on the tree. Preorder visit. */ +static reg_errcode_t +calc_next (void *extra, bin_tree_t *node) +{ + switch (node->token.type) + { + case OP_DUP_ASTERISK: + node->left->next = node; + break; + case CONCAT: + node->left->next = node->right->first; + node->right->next = node->next; + break; + default: + if (node->left) + node->left->next = node->next; + if (node->right) + node->right->next = node->next; + break; + } + return REG_NOERROR; +} + +/* Pass 3: link all DFA nodes to their NEXT node (any order will do). */ +static reg_errcode_t +link_nfa_nodes (void *extra, bin_tree_t *node) +{ + re_dfa_t *dfa = (re_dfa_t *) extra; + Idx idx = node->node_idx; + reg_errcode_t err = REG_NOERROR; + + switch (node->token.type) + { + case CONCAT: + break; + + case END_OF_RE: + assert (node->next == NULL); + break; + + case OP_DUP_ASTERISK: + case OP_ALT: + { + Idx left, right; + dfa->has_plural_match = 1; + if (node->left != NULL) + left = node->left->first->node_idx; + else + left = node->next->node_idx; + if (node->right != NULL) + right = node->right->first->node_idx; + else + right = node->next->node_idx; + assert (REG_VALID_INDEX (left)); + assert (REG_VALID_INDEX (right)); + err = re_node_set_init_2 (dfa->edests + idx, left, right); + } + break; + + case ANCHOR: + case OP_OPEN_SUBEXP: + case OP_CLOSE_SUBEXP: + err = re_node_set_init_1 (dfa->edests + idx, node->next->node_idx); + break; + + case OP_BACK_REF: + dfa->nexts[idx] = node->next->node_idx; + if (node->token.type == OP_BACK_REF) + re_node_set_init_1 (dfa->edests + idx, dfa->nexts[idx]); + break; + + default: + assert (!IS_EPSILON_NODE (node->token.type)); + dfa->nexts[idx] = node->next->node_idx; + break; + } + + return err; +} + +/* Duplicate the epsilon closure of the node ROOT_NODE. + Note that duplicated nodes have constraint INIT_CONSTRAINT in addition + to their own constraint. */ + +static reg_errcode_t +internal_function +duplicate_node_closure (re_dfa_t *dfa, Idx top_org_node, Idx top_clone_node, + Idx root_node, unsigned int init_constraint) +{ + Idx org_node, clone_node; + bool ok; + unsigned int constraint = init_constraint; + for (org_node = top_org_node, clone_node = top_clone_node;;) + { + Idx org_dest, clone_dest; + if (dfa->nodes[org_node].type == OP_BACK_REF) + { + /* If the back reference epsilon-transit, its destination must + also have the constraint. Then duplicate the epsilon closure + of the destination of the back reference, and store it in + edests of the back reference. */ + org_dest = dfa->nexts[org_node]; + re_node_set_empty (dfa->edests + clone_node); + clone_dest = duplicate_node (dfa, org_dest, constraint); + if (BE (clone_dest == REG_MISSING, 0)) + return REG_ESPACE; + dfa->nexts[clone_node] = dfa->nexts[org_node]; + ok = re_node_set_insert (dfa->edests + clone_node, clone_dest); + if (BE (! ok, 0)) + return REG_ESPACE; + } + else if (dfa->edests[org_node].nelem == 0) + { + /* In case of the node can't epsilon-transit, don't duplicate the + destination and store the original destination as the + destination of the node. */ + dfa->nexts[clone_node] = dfa->nexts[org_node]; + break; + } + else if (dfa->edests[org_node].nelem == 1) + { + /* In case of the node can epsilon-transit, and it has only one + destination. */ + org_dest = dfa->edests[org_node].elems[0]; + re_node_set_empty (dfa->edests + clone_node); + if (dfa->nodes[org_node].type == ANCHOR) + { + /* In case of the node has another constraint, append it. */ + if (org_node == root_node && clone_node != org_node) + { + /* ...but if the node is root_node itself, it means the + epsilon closure have a loop, then tie it to the + destination of the root_node. */ + ok = re_node_set_insert (dfa->edests + clone_node, org_dest); + if (BE (! ok, 0)) + return REG_ESPACE; + break; + } + constraint |= dfa->nodes[org_node].opr.ctx_type; + } + clone_dest = duplicate_node (dfa, org_dest, constraint); + if (BE (clone_dest == REG_MISSING, 0)) + return REG_ESPACE; + ok = re_node_set_insert (dfa->edests + clone_node, clone_dest); + if (BE (! ok, 0)) + return REG_ESPACE; + } + else /* dfa->edests[org_node].nelem == 2 */ + { + /* In case of the node can epsilon-transit, and it has two + destinations. In the bin_tree_t and DFA, that's '|' and '*'. */ + org_dest = dfa->edests[org_node].elems[0]; + re_node_set_empty (dfa->edests + clone_node); + /* Search for a duplicated node which satisfies the constraint. */ + clone_dest = search_duplicated_node (dfa, org_dest, constraint); + if (clone_dest == REG_MISSING) + { + /* There are no such a duplicated node, create a new one. */ + reg_errcode_t err; + clone_dest = duplicate_node (dfa, org_dest, constraint); + if (BE (clone_dest == REG_MISSING, 0)) + return REG_ESPACE; + ok = re_node_set_insert (dfa->edests + clone_node, clone_dest); + if (BE (! ok, 0)) + return REG_ESPACE; + err = duplicate_node_closure (dfa, org_dest, clone_dest, + root_node, constraint); + if (BE (err != REG_NOERROR, 0)) + return err; + } + else + { + /* There are a duplicated node which satisfy the constraint, + use it to avoid infinite loop. */ + ok = re_node_set_insert (dfa->edests + clone_node, clone_dest); + if (BE (! ok, 0)) + return REG_ESPACE; + } + + org_dest = dfa->edests[org_node].elems[1]; + clone_dest = duplicate_node (dfa, org_dest, constraint); + if (BE (clone_dest == REG_MISSING, 0)) + return REG_ESPACE; + ok = re_node_set_insert (dfa->edests + clone_node, clone_dest); + if (BE (! ok, 0)) + return REG_ESPACE; + } + org_node = org_dest; + clone_node = clone_dest; + } + return REG_NOERROR; +} + +/* Search for a node which is duplicated from the node ORG_NODE, and + satisfies the constraint CONSTRAINT. */ + +static Idx +search_duplicated_node (const re_dfa_t *dfa, Idx org_node, + unsigned int constraint) +{ + Idx idx; + for (idx = dfa->nodes_len - 1; dfa->nodes[idx].duplicated && idx > 0; --idx) + { + if (org_node == dfa->org_indices[idx] + && constraint == dfa->nodes[idx].constraint) + return idx; /* Found. */ + } + return REG_MISSING; /* Not found. */ +} + +/* Duplicate the node whose index is ORG_IDX and set the constraint CONSTRAINT. + Return the index of the new node, or REG_MISSING if insufficient storage is + available. */ + +static Idx +duplicate_node (re_dfa_t *dfa, Idx org_idx, unsigned int constraint) +{ + Idx dup_idx = re_dfa_add_node (dfa, dfa->nodes[org_idx]); + if (BE (dup_idx != REG_MISSING, 1)) + { + dfa->nodes[dup_idx].constraint = constraint; + if (dfa->nodes[org_idx].type == ANCHOR) + dfa->nodes[dup_idx].constraint |= dfa->nodes[org_idx].opr.ctx_type; + dfa->nodes[dup_idx].duplicated = 1; + + /* Store the index of the original node. */ + dfa->org_indices[dup_idx] = org_idx; + } + return dup_idx; +} + +static reg_errcode_t +calc_inveclosure (re_dfa_t *dfa) +{ + Idx src, idx; + bool ok; + for (idx = 0; idx < dfa->nodes_len; ++idx) + re_node_set_init_empty (dfa->inveclosures + idx); + + for (src = 0; src < dfa->nodes_len; ++src) + { + Idx *elems = dfa->eclosures[src].elems; + for (idx = 0; idx < dfa->eclosures[src].nelem; ++idx) + { + ok = re_node_set_insert_last (dfa->inveclosures + elems[idx], src); + if (BE (! ok, 0)) + return REG_ESPACE; + } + } + + return REG_NOERROR; +} + +/* Calculate "eclosure" for all the node in DFA. */ + +static reg_errcode_t +calc_eclosure (re_dfa_t *dfa) +{ + Idx node_idx; + bool incomplete; +#ifdef DEBUG + assert (dfa->nodes_len > 0); +#endif + incomplete = false; + /* For each nodes, calculate epsilon closure. */ + for (node_idx = 0; ; ++node_idx) + { + reg_errcode_t err; + re_node_set eclosure_elem; + if (node_idx == dfa->nodes_len) + { + if (!incomplete) + break; + incomplete = false; + node_idx = 0; + } + +#ifdef DEBUG + assert (dfa->eclosures[node_idx].nelem != REG_MISSING); +#endif + + /* If we have already calculated, skip it. */ + if (dfa->eclosures[node_idx].nelem != 0) + continue; + /* Calculate epsilon closure of `node_idx'. */ + err = calc_eclosure_iter (&eclosure_elem, dfa, node_idx, true); + if (BE (err != REG_NOERROR, 0)) + return err; + + if (dfa->eclosures[node_idx].nelem == 0) + { + incomplete = true; + re_node_set_free (&eclosure_elem); + } + } + return REG_NOERROR; +} + +/* Calculate epsilon closure of NODE. */ + +static reg_errcode_t +calc_eclosure_iter (re_node_set *new_set, re_dfa_t *dfa, Idx node, bool root) +{ + reg_errcode_t err; + unsigned int constraint; + Idx i; + bool incomplete; + bool ok; + re_node_set eclosure; + incomplete = false; + err = re_node_set_alloc (&eclosure, dfa->edests[node].nelem + 1); + if (BE (err != REG_NOERROR, 0)) + return err; + + /* This indicates that we are calculating this node now. + We reference this value to avoid infinite loop. */ + dfa->eclosures[node].nelem = REG_MISSING; + + constraint = ((dfa->nodes[node].type == ANCHOR) + ? dfa->nodes[node].opr.ctx_type : 0); + /* If the current node has constraints, duplicate all nodes. + Since they must inherit the constraints. */ + if (constraint + && dfa->edests[node].nelem + && !dfa->nodes[dfa->edests[node].elems[0]].duplicated) + { + err = duplicate_node_closure (dfa, node, node, node, constraint); + if (BE (err != REG_NOERROR, 0)) + return err; + } + + /* Expand each epsilon destination nodes. */ + if (IS_EPSILON_NODE(dfa->nodes[node].type)) + for (i = 0; i < dfa->edests[node].nelem; ++i) + { + re_node_set eclosure_elem; + Idx edest = dfa->edests[node].elems[i]; + /* If calculating the epsilon closure of `edest' is in progress, + return intermediate result. */ + if (dfa->eclosures[edest].nelem == REG_MISSING) + { + incomplete = true; + continue; + } + /* If we haven't calculated the epsilon closure of `edest' yet, + calculate now. Otherwise use calculated epsilon closure. */ + if (dfa->eclosures[edest].nelem == 0) + { + err = calc_eclosure_iter (&eclosure_elem, dfa, edest, false); + if (BE (err != REG_NOERROR, 0)) + return err; + } + else + eclosure_elem = dfa->eclosures[edest]; + /* Merge the epsilon closure of `edest'. */ + re_node_set_merge (&eclosure, &eclosure_elem); + /* If the epsilon closure of `edest' is incomplete, + the epsilon closure of this node is also incomplete. */ + if (dfa->eclosures[edest].nelem == 0) + { + incomplete = true; + re_node_set_free (&eclosure_elem); + } + } + + /* Epsilon closures include itself. */ + ok = re_node_set_insert (&eclosure, node); + if (BE (! ok, 0)) + return REG_ESPACE; + if (incomplete && !root) + dfa->eclosures[node].nelem = 0; + else + dfa->eclosures[node] = eclosure; + *new_set = eclosure; + return REG_NOERROR; +} + +/* Functions for token which are used in the parser. */ + +/* Fetch a token from INPUT. + We must not use this function inside bracket expressions. */ + +static void +internal_function +fetch_token (re_token_t *result, re_string_t *input, reg_syntax_t syntax) +{ + re_string_skip_bytes (input, peek_token (result, input, syntax)); +} + +/* Peek a token from INPUT, and return the length of the token. + We must not use this function inside bracket expressions. */ + +static int +internal_function +peek_token (re_token_t *token, re_string_t *input, reg_syntax_t syntax) +{ + unsigned char c; + + if (re_string_eoi (input)) + { + token->type = END_OF_RE; + return 0; + } + + c = re_string_peek_byte (input, 0); + token->opr.c = c; + + token->word_char = 0; +#ifdef RE_ENABLE_I18N + token->mb_partial = 0; + if (input->mb_cur_max > 1 && + !re_string_first_byte (input, re_string_cur_idx (input))) + { + token->type = CHARACTER; + token->mb_partial = 1; + return 1; + } +#endif + if (c == '\\') + { + unsigned char c2; + if (re_string_cur_idx (input) + 1 >= re_string_length (input)) + { + token->type = BACK_SLASH; + return 1; + } + + c2 = re_string_peek_byte_case (input, 1); + token->opr.c = c2; + token->type = CHARACTER; +#ifdef RE_ENABLE_I18N + if (input->mb_cur_max > 1) + { + wint_t wc = re_string_wchar_at (input, + re_string_cur_idx (input) + 1); + token->word_char = IS_WIDE_WORD_CHAR (wc) != 0; + } + else +#endif + token->word_char = IS_WORD_CHAR (c2) != 0; + + switch (c2) + { + case '|': + if (!(syntax & RE_LIMITED_OPS) && !(syntax & RE_NO_BK_VBAR)) + token->type = OP_ALT; + break; + case '1': case '2': case '3': case '4': case '5': + case '6': case '7': case '8': case '9': + if (!(syntax & RE_NO_BK_REFS)) + { + token->type = OP_BACK_REF; + token->opr.idx = c2 - '1'; + } + break; + case '<': + if (!(syntax & RE_NO_GNU_OPS)) + { + token->type = ANCHOR; + token->opr.ctx_type = WORD_FIRST; + } + break; + case '>': + if (!(syntax & RE_NO_GNU_OPS)) + { + token->type = ANCHOR; + token->opr.ctx_type = WORD_LAST; + } + break; + case 'b': + if (!(syntax & RE_NO_GNU_OPS)) + { + token->type = ANCHOR; + token->opr.ctx_type = WORD_DELIM; + } + break; + case 'B': + if (!(syntax & RE_NO_GNU_OPS)) + { + token->type = ANCHOR; + token->opr.ctx_type = NOT_WORD_DELIM; + } + break; + case 'w': + if (!(syntax & RE_NO_GNU_OPS)) + token->type = OP_WORD; + break; + case 'W': + if (!(syntax & RE_NO_GNU_OPS)) + token->type = OP_NOTWORD; + break; + case 's': + if (!(syntax & RE_NO_GNU_OPS)) + token->type = OP_SPACE; + break; + case 'S': + if (!(syntax & RE_NO_GNU_OPS)) + token->type = OP_NOTSPACE; + break; + case '`': + if (!(syntax & RE_NO_GNU_OPS)) + { + token->type = ANCHOR; + token->opr.ctx_type = BUF_FIRST; + } + break; + case '\'': + if (!(syntax & RE_NO_GNU_OPS)) + { + token->type = ANCHOR; + token->opr.ctx_type = BUF_LAST; + } + break; + case '(': + if (!(syntax & RE_NO_BK_PARENS)) + token->type = OP_OPEN_SUBEXP; + break; + case ')': + if (!(syntax & RE_NO_BK_PARENS)) + token->type = OP_CLOSE_SUBEXP; + break; + case '+': + if (!(syntax & RE_LIMITED_OPS) && (syntax & RE_BK_PLUS_QM)) + token->type = OP_DUP_PLUS; + break; + case '?': + if (!(syntax & RE_LIMITED_OPS) && (syntax & RE_BK_PLUS_QM)) + token->type = OP_DUP_QUESTION; + break; + case '{': + if ((syntax & RE_INTERVALS) && (!(syntax & RE_NO_BK_BRACES))) + token->type = OP_OPEN_DUP_NUM; + break; + case '}': + if ((syntax & RE_INTERVALS) && (!(syntax & RE_NO_BK_BRACES))) + token->type = OP_CLOSE_DUP_NUM; + break; + default: + break; + } + return 2; + } + + token->type = CHARACTER; +#ifdef RE_ENABLE_I18N + if (input->mb_cur_max > 1) + { + wint_t wc = re_string_wchar_at (input, re_string_cur_idx (input)); + token->word_char = IS_WIDE_WORD_CHAR (wc) != 0; + } + else +#endif + token->word_char = IS_WORD_CHAR (token->opr.c); + + switch (c) + { + case '\n': + if (syntax & RE_NEWLINE_ALT) + token->type = OP_ALT; + break; + case '|': + if (!(syntax & RE_LIMITED_OPS) && (syntax & RE_NO_BK_VBAR)) + token->type = OP_ALT; + break; + case '*': + token->type = OP_DUP_ASTERISK; + break; + case '+': + if (!(syntax & RE_LIMITED_OPS) && !(syntax & RE_BK_PLUS_QM)) + token->type = OP_DUP_PLUS; + break; + case '?': + if (!(syntax & RE_LIMITED_OPS) && !(syntax & RE_BK_PLUS_QM)) + token->type = OP_DUP_QUESTION; + break; + case '{': + if ((syntax & RE_INTERVALS) && (syntax & RE_NO_BK_BRACES)) + token->type = OP_OPEN_DUP_NUM; + break; + case '}': + if ((syntax & RE_INTERVALS) && (syntax & RE_NO_BK_BRACES)) + token->type = OP_CLOSE_DUP_NUM; + break; + case '(': + if (syntax & RE_NO_BK_PARENS) + token->type = OP_OPEN_SUBEXP; + break; + case ')': + if (syntax & RE_NO_BK_PARENS) + token->type = OP_CLOSE_SUBEXP; + break; + case '[': + token->type = OP_OPEN_BRACKET; + break; + case '.': + token->type = OP_PERIOD; + break; + case '^': + if (!(syntax & (RE_CONTEXT_INDEP_ANCHORS | RE_CARET_ANCHORS_HERE)) && + re_string_cur_idx (input) != 0) + { + char prev = re_string_peek_byte (input, -1); + if (!(syntax & RE_NEWLINE_ALT) || prev != '\n') + break; + } + token->type = ANCHOR; + token->opr.ctx_type = LINE_FIRST; + break; + case '$': + if (!(syntax & RE_CONTEXT_INDEP_ANCHORS) && + re_string_cur_idx (input) + 1 != re_string_length (input)) + { + re_token_t next; + re_string_skip_bytes (input, 1); + peek_token (&next, input, syntax); + re_string_skip_bytes (input, -1); + if (next.type != OP_ALT && next.type != OP_CLOSE_SUBEXP) + break; + } + token->type = ANCHOR; + token->opr.ctx_type = LINE_LAST; + break; + default: + break; + } + return 1; +} + +/* Peek a token from INPUT, and return the length of the token. + We must not use this function out of bracket expressions. */ + +static int +internal_function +peek_token_bracket (re_token_t *token, re_string_t *input, reg_syntax_t syntax) +{ + unsigned char c; + if (re_string_eoi (input)) + { + token->type = END_OF_RE; + return 0; + } + c = re_string_peek_byte (input, 0); + token->opr.c = c; + +#ifdef RE_ENABLE_I18N + if (input->mb_cur_max > 1 && + !re_string_first_byte (input, re_string_cur_idx (input))) + { + token->type = CHARACTER; + return 1; + } +#endif /* RE_ENABLE_I18N */ + + if (c == '\\' && (syntax & RE_BACKSLASH_ESCAPE_IN_LISTS) + && re_string_cur_idx (input) + 1 < re_string_length (input)) + { + /* In this case, '\' escape a character. */ + unsigned char c2; + re_string_skip_bytes (input, 1); + c2 = re_string_peek_byte (input, 0); + token->opr.c = c2; + token->type = CHARACTER; + return 1; + } + if (c == '[') /* '[' is a special char in a bracket exps. */ + { + unsigned char c2; + int token_len; + if (re_string_cur_idx (input) + 1 < re_string_length (input)) + c2 = re_string_peek_byte (input, 1); + else + c2 = 0; + token->opr.c = c2; + token_len = 2; + switch (c2) + { + case '.': + token->type = OP_OPEN_COLL_ELEM; + break; + case '=': + token->type = OP_OPEN_EQUIV_CLASS; + break; + case ':': + if (syntax & RE_CHAR_CLASSES) + { + token->type = OP_OPEN_CHAR_CLASS; + break; + } + /* else fall through. */ + default: + token->type = CHARACTER; + token->opr.c = c; + token_len = 1; + break; + } + return token_len; + } + switch (c) + { + case '-': + token->type = OP_CHARSET_RANGE; + break; + case ']': + token->type = OP_CLOSE_BRACKET; + break; + case '^': + token->type = OP_NON_MATCH_LIST; + break; + default: + token->type = CHARACTER; + } + return 1; +} + +/* Functions for parser. */ + +/* Entry point of the parser. + Parse the regular expression REGEXP and return the structure tree. + If an error is occured, ERR is set by error code, and return NULL. + This function build the following tree, from regular expression <reg_exp>: + CAT + / \ + / \ + <reg_exp> EOR + + CAT means concatenation. + EOR means end of regular expression. */ + +static bin_tree_t * +parse (re_string_t *regexp, regex_t *preg, reg_syntax_t syntax, + reg_errcode_t *err) +{ + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + bin_tree_t *tree, *eor, *root; + re_token_t current_token; + dfa->syntax = syntax; + fetch_token (¤t_token, regexp, syntax | RE_CARET_ANCHORS_HERE); + tree = parse_reg_exp (regexp, preg, ¤t_token, syntax, 0, err); + if (BE (*err != REG_NOERROR && tree == NULL, 0)) + return NULL; + eor = create_tree (dfa, NULL, NULL, END_OF_RE); + if (tree != NULL) + root = create_tree (dfa, tree, eor, CONCAT); + else + root = eor; + if (BE (eor == NULL || root == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + return root; +} + +/* This function build the following tree, from regular expression + <branch1>|<branch2>: + ALT + / \ + / \ + <branch1> <branch2> + + ALT means alternative, which represents the operator `|'. */ + +static bin_tree_t * +parse_reg_exp (re_string_t *regexp, regex_t *preg, re_token_t *token, + reg_syntax_t syntax, Idx nest, reg_errcode_t *err) +{ + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + bin_tree_t *tree, *branch = NULL; + tree = parse_branch (regexp, preg, token, syntax, nest, err); + if (BE (*err != REG_NOERROR && tree == NULL, 0)) + return NULL; + + while (token->type == OP_ALT) + { + fetch_token (token, regexp, syntax | RE_CARET_ANCHORS_HERE); + if (token->type != OP_ALT && token->type != END_OF_RE + && (nest == 0 || token->type != OP_CLOSE_SUBEXP)) + { + branch = parse_branch (regexp, preg, token, syntax, nest, err); + if (BE (*err != REG_NOERROR && branch == NULL, 0)) + return NULL; + } + else + branch = NULL; + tree = create_tree (dfa, tree, branch, OP_ALT); + if (BE (tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + } + return tree; +} + +/* This function build the following tree, from regular expression + <exp1><exp2>: + CAT + / \ + / \ + <exp1> <exp2> + + CAT means concatenation. */ + +static bin_tree_t * +parse_branch (re_string_t *regexp, regex_t *preg, re_token_t *token, + reg_syntax_t syntax, Idx nest, reg_errcode_t *err) +{ + bin_tree_t *tree, *expr; + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + tree = parse_expression (regexp, preg, token, syntax, nest, err); + if (BE (*err != REG_NOERROR && tree == NULL, 0)) + return NULL; + + while (token->type != OP_ALT && token->type != END_OF_RE + && (nest == 0 || token->type != OP_CLOSE_SUBEXP)) + { + expr = parse_expression (regexp, preg, token, syntax, nest, err); + if (BE (*err != REG_NOERROR && expr == NULL, 0)) + { + return NULL; + } + if (tree != NULL && expr != NULL) + { + tree = create_tree (dfa, tree, expr, CONCAT); + if (tree == NULL) + { + *err = REG_ESPACE; + return NULL; + } + } + else if (tree == NULL) + tree = expr; + /* Otherwise expr == NULL, we don't need to create new tree. */ + } + return tree; +} + +/* This function build the following tree, from regular expression a*: + * + | + a +*/ + +static bin_tree_t * +parse_expression (re_string_t *regexp, regex_t *preg, re_token_t *token, + reg_syntax_t syntax, Idx nest, reg_errcode_t *err) +{ + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + bin_tree_t *tree; + switch (token->type) + { + case CHARACTER: + tree = create_token_tree (dfa, NULL, NULL, token); + if (BE (tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } +#ifdef RE_ENABLE_I18N + if (dfa->mb_cur_max > 1) + { + while (!re_string_eoi (regexp) + && !re_string_first_byte (regexp, re_string_cur_idx (regexp))) + { + bin_tree_t *mbc_remain; + fetch_token (token, regexp, syntax); + mbc_remain = create_token_tree (dfa, NULL, NULL, token); + tree = create_tree (dfa, tree, mbc_remain, CONCAT); + if (BE (mbc_remain == NULL || tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + } + } +#endif + break; + case OP_OPEN_SUBEXP: + tree = parse_sub_exp (regexp, preg, token, syntax, nest + 1, err); + if (BE (*err != REG_NOERROR && tree == NULL, 0)) + return NULL; + break; + case OP_OPEN_BRACKET: + tree = parse_bracket_exp (regexp, dfa, token, syntax, err); + if (BE (*err != REG_NOERROR && tree == NULL, 0)) + return NULL; + break; + case OP_BACK_REF: + if (!BE (dfa->completed_bkref_map & (1 << token->opr.idx), 1)) + { + *err = REG_ESUBREG; + return NULL; + } + dfa->used_bkref_map |= 1 << token->opr.idx; + tree = create_token_tree (dfa, NULL, NULL, token); + if (BE (tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + ++dfa->nbackref; + dfa->has_mb_node = 1; + break; + case OP_OPEN_DUP_NUM: + if (syntax & RE_CONTEXT_INVALID_DUP) + { + *err = REG_BADRPT; + return NULL; + } + /* FALLTHROUGH */ + case OP_DUP_ASTERISK: + case OP_DUP_PLUS: + case OP_DUP_QUESTION: + if (syntax & RE_CONTEXT_INVALID_OPS) + { + *err = REG_BADRPT; + return NULL; + } + else if (syntax & RE_CONTEXT_INDEP_OPS) + { + fetch_token (token, regexp, syntax); + return parse_expression (regexp, preg, token, syntax, nest, err); + } + /* else fall through */ + case OP_CLOSE_SUBEXP: + if ((token->type == OP_CLOSE_SUBEXP) && + !(syntax & RE_UNMATCHED_RIGHT_PAREN_ORD)) + { + *err = REG_ERPAREN; + return NULL; + } + /* else fall through */ + case OP_CLOSE_DUP_NUM: + /* We treat it as a normal character. */ + + /* Then we can these characters as normal characters. */ + token->type = CHARACTER; + /* mb_partial and word_char bits should be initialized already + by peek_token. */ + tree = create_token_tree (dfa, NULL, NULL, token); + if (BE (tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + break; + case ANCHOR: + if ((token->opr.ctx_type + & (WORD_DELIM | NOT_WORD_DELIM | WORD_FIRST | WORD_LAST)) + && dfa->word_ops_used == 0) + init_word_char (dfa); + if (token->opr.ctx_type == WORD_DELIM + || token->opr.ctx_type == NOT_WORD_DELIM) + { + bin_tree_t *tree_first, *tree_last; + if (token->opr.ctx_type == WORD_DELIM) + { + token->opr.ctx_type = WORD_FIRST; + tree_first = create_token_tree (dfa, NULL, NULL, token); + token->opr.ctx_type = WORD_LAST; + } + else + { + token->opr.ctx_type = INSIDE_WORD; + tree_first = create_token_tree (dfa, NULL, NULL, token); + token->opr.ctx_type = INSIDE_NOTWORD; + } + tree_last = create_token_tree (dfa, NULL, NULL, token); + tree = create_tree (dfa, tree_first, tree_last, OP_ALT); + if (BE (tree_first == NULL || tree_last == NULL || tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + } + else + { + tree = create_token_tree (dfa, NULL, NULL, token); + if (BE (tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + } + /* We must return here, since ANCHORs can't be followed + by repetition operators. + eg. RE"^*" is invalid or "<ANCHOR(^)><CHAR(*)>", + it must not be "<ANCHOR(^)><REPEAT(*)>". */ + fetch_token (token, regexp, syntax); + return tree; + case OP_PERIOD: + tree = create_token_tree (dfa, NULL, NULL, token); + if (BE (tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + if (dfa->mb_cur_max > 1) + dfa->has_mb_node = 1; + break; + case OP_WORD: + case OP_NOTWORD: + tree = build_charclass_op (dfa, regexp->trans, + (const unsigned char *) "alnum", + (const unsigned char *) "_", + token->type == OP_NOTWORD, err); + if (BE (*err != REG_NOERROR && tree == NULL, 0)) + return NULL; + break; + case OP_SPACE: + case OP_NOTSPACE: + tree = build_charclass_op (dfa, regexp->trans, + (const unsigned char *) "space", + (const unsigned char *) "", + token->type == OP_NOTSPACE, err); + if (BE (*err != REG_NOERROR && tree == NULL, 0)) + return NULL; + break; + case OP_ALT: + case END_OF_RE: + return NULL; + case BACK_SLASH: + *err = REG_EESCAPE; + return NULL; + default: + /* Must not happen? */ +#ifdef DEBUG + assert (0); +#endif + return NULL; + } + fetch_token (token, regexp, syntax); + + while (token->type == OP_DUP_ASTERISK || token->type == OP_DUP_PLUS + || token->type == OP_DUP_QUESTION || token->type == OP_OPEN_DUP_NUM) + { + tree = parse_dup_op (tree, regexp, dfa, token, syntax, err); + if (BE (*err != REG_NOERROR && tree == NULL, 0)) + return NULL; + /* In BRE consecutive duplications are not allowed. */ + if ((syntax & RE_CONTEXT_INVALID_DUP) + && (token->type == OP_DUP_ASTERISK + || token->type == OP_OPEN_DUP_NUM)) + { + *err = REG_BADRPT; + return NULL; + } + } + + return tree; +} + +/* This function build the following tree, from regular expression + (<reg_exp>): + SUBEXP + | + <reg_exp> +*/ + +static bin_tree_t * +parse_sub_exp (re_string_t *regexp, regex_t *preg, re_token_t *token, + reg_syntax_t syntax, Idx nest, reg_errcode_t *err) +{ + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + bin_tree_t *tree; + size_t cur_nsub; + cur_nsub = preg->re_nsub++; + + fetch_token (token, regexp, syntax | RE_CARET_ANCHORS_HERE); + + /* The subexpression may be a null string. */ + if (token->type == OP_CLOSE_SUBEXP) + tree = NULL; + else + { + tree = parse_reg_exp (regexp, preg, token, syntax, nest, err); + if (BE (*err == REG_NOERROR && token->type != OP_CLOSE_SUBEXP, 0)) + *err = REG_EPAREN; + if (BE (*err != REG_NOERROR, 0)) + return NULL; + } + + if (cur_nsub <= '9' - '1') + dfa->completed_bkref_map |= 1 << cur_nsub; + + tree = create_tree (dfa, tree, NULL, SUBEXP); + if (BE (tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + tree->token.opr.idx = cur_nsub; + return tree; +} + +/* This function parse repetition operators like "*", "+", "{1,3}" etc. */ + +static bin_tree_t * +parse_dup_op (bin_tree_t *elem, re_string_t *regexp, re_dfa_t *dfa, + re_token_t *token, reg_syntax_t syntax, reg_errcode_t *err) +{ + bin_tree_t *tree = NULL, *old_tree = NULL; + Idx i, start, end, start_idx = re_string_cur_idx (regexp); + re_token_t start_token = *token; + + if (token->type == OP_OPEN_DUP_NUM) + { + end = 0; + start = fetch_number (regexp, token, syntax); + if (start == REG_MISSING) + { + if (token->type == CHARACTER && token->opr.c == ',') + start = 0; /* We treat "{,m}" as "{0,m}". */ + else + { + *err = REG_BADBR; /* <re>{} is invalid. */ + return NULL; + } + } + if (BE (start != REG_ERROR, 1)) + { + /* We treat "{n}" as "{n,n}". */ + end = ((token->type == OP_CLOSE_DUP_NUM) ? start + : ((token->type == CHARACTER && token->opr.c == ',') + ? fetch_number (regexp, token, syntax) : REG_ERROR)); + } + if (BE (start == REG_ERROR || end == REG_ERROR, 0)) + { + /* Invalid sequence. */ + if (BE (!(syntax & RE_INVALID_INTERVAL_ORD), 0)) + { + if (token->type == END_OF_RE) + *err = REG_EBRACE; + else + *err = REG_BADBR; + + return NULL; + } + + /* If the syntax bit is set, rollback. */ + re_string_set_index (regexp, start_idx); + *token = start_token; + token->type = CHARACTER; + /* mb_partial and word_char bits should be already initialized by + peek_token. */ + return elem; + } + + if (BE (end != REG_MISSING && start > end, 0)) + { + /* First number greater than second. */ + *err = REG_BADBR; + return NULL; + } + } + else + { + start = (token->type == OP_DUP_PLUS) ? 1 : 0; + end = (token->type == OP_DUP_QUESTION) ? 1 : REG_MISSING; + } + + fetch_token (token, regexp, syntax); + + if (BE (elem == NULL, 0)) + return NULL; + if (BE (start == 0 && end == 0, 0)) + { + postorder (elem, free_tree, NULL); + return NULL; + } + + /* Extract "<re>{n,m}" to "<re><re>...<re><re>{0,<m-n>}". */ + if (BE (start > 0, 0)) + { + tree = elem; + for (i = 2; i <= start; ++i) + { + elem = duplicate_tree (elem, dfa); + tree = create_tree (dfa, tree, elem, CONCAT); + if (BE (elem == NULL || tree == NULL, 0)) + goto parse_dup_op_espace; + } + + if (start == end) + return tree; + + /* Duplicate ELEM before it is marked optional. */ + elem = duplicate_tree (elem, dfa); + old_tree = tree; + } + else + old_tree = NULL; + + if (elem->token.type == SUBEXP) + postorder (elem, mark_opt_subexp, (void *) (long) elem->token.opr.idx); + + tree = create_tree (dfa, elem, NULL, + (end == REG_MISSING ? OP_DUP_ASTERISK : OP_ALT)); + if (BE (tree == NULL, 0)) + goto parse_dup_op_espace; + + /* This loop is actually executed only when end != REG_MISSING, + to rewrite <re>{0,n} as (<re>(<re>...<re>?)?)?... We have + already created the start+1-th copy. */ + if ((Idx) -1 < 0 || end != REG_MISSING) + for (i = start + 2; i <= end; ++i) + { + elem = duplicate_tree (elem, dfa); + tree = create_tree (dfa, tree, elem, CONCAT); + if (BE (elem == NULL || tree == NULL, 0)) + goto parse_dup_op_espace; + + tree = create_tree (dfa, tree, NULL, OP_ALT); + if (BE (tree == NULL, 0)) + goto parse_dup_op_espace; + } + + if (old_tree) + tree = create_tree (dfa, old_tree, tree, CONCAT); + + return tree; + + parse_dup_op_espace: + *err = REG_ESPACE; + return NULL; +} + +/* Size of the names for collating symbol/equivalence_class/character_class. + I'm not sure, but maybe enough. */ +#define BRACKET_NAME_BUF_SIZE 32 + +#ifndef _LIBC + /* Local function for parse_bracket_exp only used in case of NOT _LIBC. + Build the range expression which starts from START_ELEM, and ends + at END_ELEM. The result are written to MBCSET and SBCSET. + RANGE_ALLOC is the allocated size of mbcset->range_starts, and + mbcset->range_ends, is a pointer argument sinse we may + update it. */ + +static reg_errcode_t +internal_function +# ifdef RE_ENABLE_I18N +build_range_exp (bitset_t sbcset, re_charset_t *mbcset, Idx *range_alloc, + bracket_elem_t *start_elem, bracket_elem_t *end_elem) +# else /* not RE_ENABLE_I18N */ +build_range_exp (bitset_t sbcset, bracket_elem_t *start_elem, + bracket_elem_t *end_elem) +# endif /* not RE_ENABLE_I18N */ +{ + unsigned int start_ch, end_ch; + /* Equivalence Classes and Character Classes can't be a range start/end. */ + if (BE (start_elem->type == EQUIV_CLASS || start_elem->type == CHAR_CLASS + || end_elem->type == EQUIV_CLASS || end_elem->type == CHAR_CLASS, + 0)) + return REG_ERANGE; + + /* We can handle no multi character collating elements without libc + support. */ + if (BE ((start_elem->type == COLL_SYM + && strlen ((char *) start_elem->opr.name) > 1) + || (end_elem->type == COLL_SYM + && strlen ((char *) end_elem->opr.name) > 1), 0)) + return REG_ECOLLATE; + +# ifdef RE_ENABLE_I18N + { + wchar_t wc; + wint_t start_wc; + wint_t end_wc; + wchar_t cmp_buf[6] = {L'\0', L'\0', L'\0', L'\0', L'\0', L'\0'}; + + start_ch = ((start_elem->type == SB_CHAR) ? start_elem->opr.ch + : ((start_elem->type == COLL_SYM) ? start_elem->opr.name[0] + : 0)); + end_ch = ((end_elem->type == SB_CHAR) ? end_elem->opr.ch + : ((end_elem->type == COLL_SYM) ? end_elem->opr.name[0] + : 0)); + start_wc = ((start_elem->type == SB_CHAR || start_elem->type == COLL_SYM) + ? __btowc (start_ch) : start_elem->opr.wch); + end_wc = ((end_elem->type == SB_CHAR || end_elem->type == COLL_SYM) + ? __btowc (end_ch) : end_elem->opr.wch); + if (start_wc == WEOF || end_wc == WEOF) + return REG_ECOLLATE; + cmp_buf[0] = start_wc; + cmp_buf[4] = end_wc; + if (wcscoll (cmp_buf, cmp_buf + 4) > 0) + return REG_ERANGE; + + /* Got valid collation sequence values, add them as a new entry. + However, for !_LIBC we have no collation elements: if the + character set is single byte, the single byte character set + that we build below suffices. parse_bracket_exp passes + no MBCSET if dfa->mb_cur_max == 1. */ + if (mbcset) + { + /* Check the space of the arrays. */ + if (BE (*range_alloc == mbcset->nranges, 0)) + { + /* There is not enough space, need realloc. */ + wchar_t *new_array_start, *new_array_end; + Idx new_nranges; + + /* +1 in case of mbcset->nranges is 0. */ + new_nranges = 2 * mbcset->nranges + 1; + /* Use realloc since mbcset->range_starts and mbcset->range_ends + are NULL if *range_alloc == 0. */ + new_array_start = re_realloc (mbcset->range_starts, wchar_t, + new_nranges); + new_array_end = re_realloc (mbcset->range_ends, wchar_t, + new_nranges); + + if (BE (new_array_start == NULL || new_array_end == NULL, 0)) + return REG_ESPACE; + + mbcset->range_starts = new_array_start; + mbcset->range_ends = new_array_end; + *range_alloc = new_nranges; + } + + mbcset->range_starts[mbcset->nranges] = start_wc; + mbcset->range_ends[mbcset->nranges++] = end_wc; + } + + /* Build the table for single byte characters. */ + for (wc = 0; wc < SBC_MAX; ++wc) + { + cmp_buf[2] = wc; + if (wcscoll (cmp_buf, cmp_buf + 2) <= 0 + && wcscoll (cmp_buf + 2, cmp_buf + 4) <= 0) + bitset_set (sbcset, wc); + } + } +# else /* not RE_ENABLE_I18N */ + { + unsigned int ch; + start_ch = ((start_elem->type == SB_CHAR ) ? start_elem->opr.ch + : ((start_elem->type == COLL_SYM) ? start_elem->opr.name[0] + : 0)); + end_ch = ((end_elem->type == SB_CHAR ) ? end_elem->opr.ch + : ((end_elem->type == COLL_SYM) ? end_elem->opr.name[0] + : 0)); + if (start_ch > end_ch) + return REG_ERANGE; + /* Build the table for single byte characters. */ + for (ch = 0; ch < SBC_MAX; ++ch) + if (start_ch <= ch && ch <= end_ch) + bitset_set (sbcset, ch); + } +# endif /* not RE_ENABLE_I18N */ + return REG_NOERROR; +} +#endif /* not _LIBC */ + +#ifndef _LIBC +/* Helper function for parse_bracket_exp only used in case of NOT _LIBC.. + Build the collating element which is represented by NAME. + The result are written to MBCSET and SBCSET. + COLL_SYM_ALLOC is the allocated size of mbcset->coll_sym, is a + pointer argument since we may update it. */ + +static reg_errcode_t +internal_function +build_collating_symbol (bitset_t sbcset, +# ifdef RE_ENABLE_I18N + re_charset_t *mbcset, Idx *coll_sym_alloc, +# endif + const unsigned char *name) +{ + size_t name_len = strlen ((const char *) name); + if (BE (name_len != 1, 0)) + return REG_ECOLLATE; + else + { + bitset_set (sbcset, name[0]); + return REG_NOERROR; + } +} +#endif /* not _LIBC */ + +/* This function parse bracket expression like "[abc]", "[a-c]", + "[[.a-a.]]" etc. */ + +static bin_tree_t * +parse_bracket_exp (re_string_t *regexp, re_dfa_t *dfa, re_token_t *token, + reg_syntax_t syntax, reg_errcode_t *err) +{ +#ifdef _LIBC + const unsigned char *collseqmb; + const char *collseqwc; + uint32_t nrules; + int32_t table_size; + const int32_t *symb_table; + const unsigned char *extra; + + /* Local function for parse_bracket_exp used in _LIBC environement. + Seek the collating symbol entry correspondings to NAME. + Return the index of the symbol in the SYMB_TABLE. */ + + auto inline int32_t + __attribute ((always_inline)) + seek_collating_symbol_entry (name, name_len) + const unsigned char *name; + size_t name_len; + { + int32_t hash = elem_hash ((const char *) name, name_len); + int32_t elem = hash % table_size; + if (symb_table[2 * elem] != 0) + { + int32_t second = hash % (table_size - 2) + 1; + + do + { + /* First compare the hashing value. */ + if (symb_table[2 * elem] == hash + /* Compare the length of the name. */ + && name_len == extra[symb_table[2 * elem + 1]] + /* Compare the name. */ + && memcmp (name, &extra[symb_table[2 * elem + 1] + 1], + name_len) == 0) + { + /* Yep, this is the entry. */ + break; + } + + /* Next entry. */ + elem += second; + } + while (symb_table[2 * elem] != 0); + } + return elem; + } + + /* Local function for parse_bracket_exp used in _LIBC environement. + Look up the collation sequence value of BR_ELEM. + Return the value if succeeded, UINT_MAX otherwise. */ + + auto inline unsigned int + __attribute ((always_inline)) + lookup_collation_sequence_value (br_elem) + bracket_elem_t *br_elem; + { + if (br_elem->type == SB_CHAR) + { + /* + if (MB_CUR_MAX == 1) + */ + if (nrules == 0) + return collseqmb[br_elem->opr.ch]; + else + { + wint_t wc = __btowc (br_elem->opr.ch); + return __collseq_table_lookup (collseqwc, wc); + } + } + else if (br_elem->type == MB_CHAR) + { + return __collseq_table_lookup (collseqwc, br_elem->opr.wch); + } + else if (br_elem->type == COLL_SYM) + { + size_t sym_name_len = strlen ((char *) br_elem->opr.name); + if (nrules != 0) + { + int32_t elem, idx; + elem = seek_collating_symbol_entry (br_elem->opr.name, + sym_name_len); + if (symb_table[2 * elem] != 0) + { + /* We found the entry. */ + idx = symb_table[2 * elem + 1]; + /* Skip the name of collating element name. */ + idx += 1 + extra[idx]; + /* Skip the byte sequence of the collating element. */ + idx += 1 + extra[idx]; + /* Adjust for the alignment. */ + idx = (idx + 3) & ~3; + /* Skip the multibyte collation sequence value. */ + idx += sizeof (unsigned int); + /* Skip the wide char sequence of the collating element. */ + idx += sizeof (unsigned int) * + (1 + *(unsigned int *) (extra + idx)); + /* Return the collation sequence value. */ + return *(unsigned int *) (extra + idx); + } + else if (symb_table[2 * elem] == 0 && sym_name_len == 1) + { + /* No valid character. Match it as a single byte + character. */ + return collseqmb[br_elem->opr.name[0]]; + } + } + else if (sym_name_len == 1) + return collseqmb[br_elem->opr.name[0]]; + } + return UINT_MAX; + } + + /* Local function for parse_bracket_exp used in _LIBC environement. + Build the range expression which starts from START_ELEM, and ends + at END_ELEM. The result are written to MBCSET and SBCSET. + RANGE_ALLOC is the allocated size of mbcset->range_starts, and + mbcset->range_ends, is a pointer argument sinse we may + update it. */ + + auto inline reg_errcode_t + __attribute ((always_inline)) + build_range_exp (sbcset, mbcset, range_alloc, start_elem, end_elem) + re_charset_t *mbcset; + Idx *range_alloc; + bitset_t sbcset; + bracket_elem_t *start_elem, *end_elem; + { + unsigned int ch; + uint32_t start_collseq; + uint32_t end_collseq; + + /* Equivalence Classes and Character Classes can't be a range + start/end. */ + if (BE (start_elem->type == EQUIV_CLASS || start_elem->type == CHAR_CLASS + || end_elem->type == EQUIV_CLASS || end_elem->type == CHAR_CLASS, + 0)) + return REG_ERANGE; + + start_collseq = lookup_collation_sequence_value (start_elem); + end_collseq = lookup_collation_sequence_value (end_elem); + /* Check start/end collation sequence values. */ + if (BE (start_collseq == UINT_MAX || end_collseq == UINT_MAX, 0)) + return REG_ECOLLATE; + if (BE ((syntax & RE_NO_EMPTY_RANGES) && start_collseq > end_collseq, 0)) + return REG_ERANGE; + + /* Got valid collation sequence values, add them as a new entry. + However, if we have no collation elements, and the character set + is single byte, the single byte character set that we + build below suffices. */ + if (nrules > 0 || dfa->mb_cur_max > 1) + { + /* Check the space of the arrays. */ + if (BE (*range_alloc == mbcset->nranges, 0)) + { + /* There is not enough space, need realloc. */ + uint32_t *new_array_start; + uint32_t *new_array_end; + Idx new_nranges; + + /* +1 in case of mbcset->nranges is 0. */ + new_nranges = 2 * mbcset->nranges + 1; + new_array_start = re_realloc (mbcset->range_starts, uint32_t, + new_nranges); + new_array_end = re_realloc (mbcset->range_ends, uint32_t, + new_nranges); + + if (BE (new_array_start == NULL || new_array_end == NULL, 0)) + return REG_ESPACE; + + mbcset->range_starts = new_array_start; + mbcset->range_ends = new_array_end; + *range_alloc = new_nranges; + } + + mbcset->range_starts[mbcset->nranges] = start_collseq; + mbcset->range_ends[mbcset->nranges++] = end_collseq; + } + + /* Build the table for single byte characters. */ + for (ch = 0; ch < SBC_MAX; ch++) + { + uint32_t ch_collseq; + /* + if (MB_CUR_MAX == 1) + */ + if (nrules == 0) + ch_collseq = collseqmb[ch]; + else + ch_collseq = __collseq_table_lookup (collseqwc, __btowc (ch)); + if (start_collseq <= ch_collseq && ch_collseq <= end_collseq) + bitset_set (sbcset, ch); + } + return REG_NOERROR; + } + + /* Local function for parse_bracket_exp used in _LIBC environement. + Build the collating element which is represented by NAME. + The result are written to MBCSET and SBCSET. + COLL_SYM_ALLOC is the allocated size of mbcset->coll_sym, is a + pointer argument sinse we may update it. */ + + auto inline reg_errcode_t + __attribute ((always_inline)) + build_collating_symbol (sbcset, mbcset, coll_sym_alloc, name) + re_charset_t *mbcset; + Idx *coll_sym_alloc; + bitset_t sbcset; + const unsigned char *name; + { + int32_t elem, idx; + size_t name_len = strlen ((const char *) name); + if (nrules != 0) + { + elem = seek_collating_symbol_entry (name, name_len); + if (symb_table[2 * elem] != 0) + { + /* We found the entry. */ + idx = symb_table[2 * elem + 1]; + /* Skip the name of collating element name. */ + idx += 1 + extra[idx]; + } + else if (symb_table[2 * elem] == 0 && name_len == 1) + { + /* No valid character, treat it as a normal + character. */ + bitset_set (sbcset, name[0]); + return REG_NOERROR; + } + else + return REG_ECOLLATE; + + /* Got valid collation sequence, add it as a new entry. */ + /* Check the space of the arrays. */ + if (BE (*coll_sym_alloc == mbcset->ncoll_syms, 0)) + { + /* Not enough, realloc it. */ + /* +1 in case of mbcset->ncoll_syms is 0. */ + Idx new_coll_sym_alloc = 2 * mbcset->ncoll_syms + 1; + /* Use realloc since mbcset->coll_syms is NULL + if *alloc == 0. */ + int32_t *new_coll_syms = re_realloc (mbcset->coll_syms, int32_t, + new_coll_sym_alloc); + if (BE (new_coll_syms == NULL, 0)) + return REG_ESPACE; + mbcset->coll_syms = new_coll_syms; + *coll_sym_alloc = new_coll_sym_alloc; + } + mbcset->coll_syms[mbcset->ncoll_syms++] = idx; + return REG_NOERROR; + } + else + { + if (BE (name_len != 1, 0)) + return REG_ECOLLATE; + else + { + bitset_set (sbcset, name[0]); + return REG_NOERROR; + } + } + } +#endif + + re_token_t br_token; + re_bitset_ptr_t sbcset; +#ifdef RE_ENABLE_I18N + re_charset_t *mbcset; + Idx coll_sym_alloc = 0, range_alloc = 0, mbchar_alloc = 0; + Idx equiv_class_alloc = 0, char_class_alloc = 0; +#endif /* not RE_ENABLE_I18N */ + bool non_match = false; + bin_tree_t *work_tree; + int token_len; + bool first_round = true; +#ifdef _LIBC + collseqmb = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQMB); + nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + if (nrules) + { + /* + if (MB_CUR_MAX > 1) + */ + collseqwc = _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQWC); + table_size = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_SYMB_HASH_SIZEMB); + symb_table = (const int32_t *) _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_TABLEMB); + extra = (const unsigned char *) _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_EXTRAMB); + } +#endif + sbcset = (re_bitset_ptr_t) calloc (sizeof (bitset_t), 1); +#ifdef RE_ENABLE_I18N + mbcset = (re_charset_t *) calloc (sizeof (re_charset_t), 1); +#endif /* RE_ENABLE_I18N */ +#ifdef RE_ENABLE_I18N + if (BE (sbcset == NULL || mbcset == NULL, 0)) +#else + if (BE (sbcset == NULL, 0)) +#endif /* RE_ENABLE_I18N */ + { + *err = REG_ESPACE; + return NULL; + } + + token_len = peek_token_bracket (token, regexp, syntax); + if (BE (token->type == END_OF_RE, 0)) + { + *err = REG_BADPAT; + goto parse_bracket_exp_free_return; + } + if (token->type == OP_NON_MATCH_LIST) + { +#ifdef RE_ENABLE_I18N + mbcset->non_match = 1; +#endif /* not RE_ENABLE_I18N */ + non_match = true; + if (syntax & RE_HAT_LISTS_NOT_NEWLINE) + bitset_set (sbcset, '\0'); + re_string_skip_bytes (regexp, token_len); /* Skip a token. */ + token_len = peek_token_bracket (token, regexp, syntax); + if (BE (token->type == END_OF_RE, 0)) + { + *err = REG_BADPAT; + goto parse_bracket_exp_free_return; + } + } + + /* We treat the first ']' as a normal character. */ + if (token->type == OP_CLOSE_BRACKET) + token->type = CHARACTER; + + while (1) + { + bracket_elem_t start_elem, end_elem; + unsigned char start_name_buf[BRACKET_NAME_BUF_SIZE]; + unsigned char end_name_buf[BRACKET_NAME_BUF_SIZE]; + reg_errcode_t ret; + int token_len2 = 0; + bool is_range_exp = false; + re_token_t token2; + + start_elem.opr.name = start_name_buf; + ret = parse_bracket_element (&start_elem, regexp, token, token_len, dfa, + syntax, first_round); + if (BE (ret != REG_NOERROR, 0)) + { + *err = ret; + goto parse_bracket_exp_free_return; + } + first_round = false; + + /* Get information about the next token. We need it in any case. */ + token_len = peek_token_bracket (token, regexp, syntax); + + /* Do not check for ranges if we know they are not allowed. */ + if (start_elem.type != CHAR_CLASS && start_elem.type != EQUIV_CLASS) + { + if (BE (token->type == END_OF_RE, 0)) + { + *err = REG_EBRACK; + goto parse_bracket_exp_free_return; + } + if (token->type == OP_CHARSET_RANGE) + { + re_string_skip_bytes (regexp, token_len); /* Skip '-'. */ + token_len2 = peek_token_bracket (&token2, regexp, syntax); + if (BE (token2.type == END_OF_RE, 0)) + { + *err = REG_EBRACK; + goto parse_bracket_exp_free_return; + } + if (token2.type == OP_CLOSE_BRACKET) + { + /* We treat the last '-' as a normal character. */ + re_string_skip_bytes (regexp, -token_len); + token->type = CHARACTER; + } + else + is_range_exp = true; + } + } + + if (is_range_exp == true) + { + end_elem.opr.name = end_name_buf; + ret = parse_bracket_element (&end_elem, regexp, &token2, token_len2, + dfa, syntax, true); + if (BE (ret != REG_NOERROR, 0)) + { + *err = ret; + goto parse_bracket_exp_free_return; + } + + token_len = peek_token_bracket (token, regexp, syntax); + +#ifdef _LIBC + *err = build_range_exp (sbcset, mbcset, &range_alloc, + &start_elem, &end_elem); +#else +# ifdef RE_ENABLE_I18N + *err = build_range_exp (sbcset, + dfa->mb_cur_max > 1 ? mbcset : NULL, + &range_alloc, &start_elem, &end_elem); +# else + *err = build_range_exp (sbcset, &start_elem, &end_elem); +# endif +#endif /* RE_ENABLE_I18N */ + if (BE (*err != REG_NOERROR, 0)) + goto parse_bracket_exp_free_return; + } + else + { + switch (start_elem.type) + { + case SB_CHAR: + bitset_set (sbcset, start_elem.opr.ch); + break; +#ifdef RE_ENABLE_I18N + case MB_CHAR: + /* Check whether the array has enough space. */ + if (BE (mbchar_alloc == mbcset->nmbchars, 0)) + { + wchar_t *new_mbchars; + /* Not enough, realloc it. */ + /* +1 in case of mbcset->nmbchars is 0. */ + mbchar_alloc = 2 * mbcset->nmbchars + 1; + /* Use realloc since array is NULL if *alloc == 0. */ + new_mbchars = re_realloc (mbcset->mbchars, wchar_t, + mbchar_alloc); + if (BE (new_mbchars == NULL, 0)) + goto parse_bracket_exp_espace; + mbcset->mbchars = new_mbchars; + } + mbcset->mbchars[mbcset->nmbchars++] = start_elem.opr.wch; + break; +#endif /* RE_ENABLE_I18N */ + case EQUIV_CLASS: + *err = build_equiv_class (sbcset, +#ifdef RE_ENABLE_I18N + mbcset, &equiv_class_alloc, +#endif /* RE_ENABLE_I18N */ + start_elem.opr.name); + if (BE (*err != REG_NOERROR, 0)) + goto parse_bracket_exp_free_return; + break; + case COLL_SYM: + *err = build_collating_symbol (sbcset, +#ifdef RE_ENABLE_I18N + mbcset, &coll_sym_alloc, +#endif /* RE_ENABLE_I18N */ + start_elem.opr.name); + if (BE (*err != REG_NOERROR, 0)) + goto parse_bracket_exp_free_return; + break; + case CHAR_CLASS: + *err = build_charclass (regexp->trans, sbcset, +#ifdef RE_ENABLE_I18N + mbcset, &char_class_alloc, +#endif /* RE_ENABLE_I18N */ + start_elem.opr.name, syntax); + if (BE (*err != REG_NOERROR, 0)) + goto parse_bracket_exp_free_return; + break; + default: + assert (0); + break; + } + } + if (BE (token->type == END_OF_RE, 0)) + { + *err = REG_EBRACK; + goto parse_bracket_exp_free_return; + } + if (token->type == OP_CLOSE_BRACKET) + break; + } + + re_string_skip_bytes (regexp, token_len); /* Skip a token. */ + + /* If it is non-matching list. */ + if (non_match) + bitset_not (sbcset); + +#ifdef RE_ENABLE_I18N + /* Ensure only single byte characters are set. */ + if (dfa->mb_cur_max > 1) + bitset_mask (sbcset, dfa->sb_char); + + if (mbcset->nmbchars || mbcset->ncoll_syms || mbcset->nequiv_classes + || mbcset->nranges || (dfa->mb_cur_max > 1 && (mbcset->nchar_classes + || mbcset->non_match))) + { + bin_tree_t *mbc_tree; + int sbc_idx; + /* Build a tree for complex bracket. */ + dfa->has_mb_node = 1; + br_token.type = COMPLEX_BRACKET; + br_token.opr.mbcset = mbcset; + mbc_tree = create_token_tree (dfa, NULL, NULL, &br_token); + if (BE (mbc_tree == NULL, 0)) + goto parse_bracket_exp_espace; + for (sbc_idx = 0; sbc_idx < BITSET_WORDS; ++sbc_idx) + if (sbcset[sbc_idx]) + break; + /* If there are no bits set in sbcset, there is no point + of having both SIMPLE_BRACKET and COMPLEX_BRACKET. */ + if (sbc_idx < BITSET_WORDS) + { + /* Build a tree for simple bracket. */ + br_token.type = SIMPLE_BRACKET; + br_token.opr.sbcset = sbcset; + work_tree = create_token_tree (dfa, NULL, NULL, &br_token); + if (BE (work_tree == NULL, 0)) + goto parse_bracket_exp_espace; + + /* Then join them by ALT node. */ + work_tree = create_tree (dfa, work_tree, mbc_tree, OP_ALT); + if (BE (work_tree == NULL, 0)) + goto parse_bracket_exp_espace; + } + else + { + re_free (sbcset); + work_tree = mbc_tree; + } + } + else +#endif /* not RE_ENABLE_I18N */ + { +#ifdef RE_ENABLE_I18N + free_charset (mbcset); +#endif + /* Build a tree for simple bracket. */ + br_token.type = SIMPLE_BRACKET; + br_token.opr.sbcset = sbcset; + work_tree = create_token_tree (dfa, NULL, NULL, &br_token); + if (BE (work_tree == NULL, 0)) + goto parse_bracket_exp_espace; + } + return work_tree; + + parse_bracket_exp_espace: + *err = REG_ESPACE; + parse_bracket_exp_free_return: + re_free (sbcset); +#ifdef RE_ENABLE_I18N + free_charset (mbcset); +#endif /* RE_ENABLE_I18N */ + return NULL; +} + +/* Parse an element in the bracket expression. */ + +static reg_errcode_t +parse_bracket_element (bracket_elem_t *elem, re_string_t *regexp, + re_token_t *token, int token_len, re_dfa_t *dfa, + reg_syntax_t syntax, bool accept_hyphen) +{ +#ifdef RE_ENABLE_I18N + int cur_char_size; + cur_char_size = re_string_char_size_at (regexp, re_string_cur_idx (regexp)); + if (cur_char_size > 1) + { + elem->type = MB_CHAR; + elem->opr.wch = re_string_wchar_at (regexp, re_string_cur_idx (regexp)); + re_string_skip_bytes (regexp, cur_char_size); + return REG_NOERROR; + } +#endif /* RE_ENABLE_I18N */ + re_string_skip_bytes (regexp, token_len); /* Skip a token. */ + if (token->type == OP_OPEN_COLL_ELEM || token->type == OP_OPEN_CHAR_CLASS + || token->type == OP_OPEN_EQUIV_CLASS) + return parse_bracket_symbol (elem, regexp, token); + if (BE (token->type == OP_CHARSET_RANGE, 0) && !accept_hyphen) + { + /* A '-' must only appear as anything but a range indicator before + the closing bracket. Everything else is an error. */ + re_token_t token2; + (void) peek_token_bracket (&token2, regexp, syntax); + if (token2.type != OP_CLOSE_BRACKET) + /* The actual error value is not standardized since this whole + case is undefined. But ERANGE makes good sense. */ + return REG_ERANGE; + } + elem->type = SB_CHAR; + elem->opr.ch = token->opr.c; + return REG_NOERROR; +} + +/* Parse a bracket symbol in the bracket expression. Bracket symbols are + such as [:<character_class>:], [.<collating_element>.], and + [=<equivalent_class>=]. */ + +static reg_errcode_t +parse_bracket_symbol (bracket_elem_t *elem, re_string_t *regexp, + re_token_t *token) +{ + unsigned char ch, delim = token->opr.c; + int i = 0; + if (re_string_eoi(regexp)) + return REG_EBRACK; + for (;; ++i) + { + if (i >= BRACKET_NAME_BUF_SIZE) + return REG_EBRACK; + if (token->type == OP_OPEN_CHAR_CLASS) + ch = re_string_fetch_byte_case (regexp); + else + ch = re_string_fetch_byte (regexp); + if (re_string_eoi(regexp)) + return REG_EBRACK; + if (ch == delim && re_string_peek_byte (regexp, 0) == ']') + break; + elem->opr.name[i] = ch; + } + re_string_skip_bytes (regexp, 1); + elem->opr.name[i] = '\0'; + switch (token->type) + { + case OP_OPEN_COLL_ELEM: + elem->type = COLL_SYM; + break; + case OP_OPEN_EQUIV_CLASS: + elem->type = EQUIV_CLASS; + break; + case OP_OPEN_CHAR_CLASS: + elem->type = CHAR_CLASS; + break; + default: + break; + } + return REG_NOERROR; +} + + /* Helper function for parse_bracket_exp. + Build the equivalence class which is represented by NAME. + The result are written to MBCSET and SBCSET. + EQUIV_CLASS_ALLOC is the allocated size of mbcset->equiv_classes, + is a pointer argument sinse we may update it. */ + +static reg_errcode_t +#ifdef RE_ENABLE_I18N +build_equiv_class (bitset_t sbcset, re_charset_t *mbcset, + Idx *equiv_class_alloc, const unsigned char *name) +#else /* not RE_ENABLE_I18N */ +build_equiv_class (bitset_t sbcset, const unsigned char *name) +#endif /* not RE_ENABLE_I18N */ +{ +#ifdef _LIBC + uint32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + if (nrules != 0) + { + const int32_t *table, *indirect; + const unsigned char *weights, *extra, *cp; + unsigned char char_buf[2]; + int32_t idx1, idx2; + unsigned int ch; + size_t len; + /* This #include defines a local function! */ +# include <locale/weight.h> + /* Calculate the index for equivalence class. */ + cp = name; + table = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB); + weights = (const unsigned char *) _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_WEIGHTMB); + extra = (const unsigned char *) _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_EXTRAMB); + indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_INDIRECTMB); + idx1 = findidx (&cp); + if (BE (idx1 == 0 || cp < name + strlen ((const char *) name), 0)) + /* This isn't a valid character. */ + return REG_ECOLLATE; + + /* Build single byte matcing table for this equivalence class. */ + char_buf[1] = (unsigned char) '\0'; + len = weights[idx1]; + for (ch = 0; ch < SBC_MAX; ++ch) + { + char_buf[0] = ch; + cp = char_buf; + idx2 = findidx (&cp); +/* + idx2 = table[ch]; +*/ + if (idx2 == 0) + /* This isn't a valid character. */ + continue; + if (len == weights[idx2]) + { + int cnt = 0; + while (cnt <= len && + weights[idx1 + 1 + cnt] == weights[idx2 + 1 + cnt]) + ++cnt; + + if (cnt > len) + bitset_set (sbcset, ch); + } + } + /* Check whether the array has enough space. */ + if (BE (*equiv_class_alloc == mbcset->nequiv_classes, 0)) + { + /* Not enough, realloc it. */ + /* +1 in case of mbcset->nequiv_classes is 0. */ + Idx new_equiv_class_alloc = 2 * mbcset->nequiv_classes + 1; + /* Use realloc since the array is NULL if *alloc == 0. */ + int32_t *new_equiv_classes = re_realloc (mbcset->equiv_classes, + int32_t, + new_equiv_class_alloc); + if (BE (new_equiv_classes == NULL, 0)) + return REG_ESPACE; + mbcset->equiv_classes = new_equiv_classes; + *equiv_class_alloc = new_equiv_class_alloc; + } + mbcset->equiv_classes[mbcset->nequiv_classes++] = idx1; + } + else +#endif /* _LIBC */ + { + if (BE (strlen ((const char *) name) != 1, 0)) + return REG_ECOLLATE; + bitset_set (sbcset, *name); + } + return REG_NOERROR; +} + + /* Helper function for parse_bracket_exp. + Build the character class which is represented by NAME. + The result are written to MBCSET and SBCSET. + CHAR_CLASS_ALLOC is the allocated size of mbcset->char_classes, + is a pointer argument sinse we may update it. */ + +static reg_errcode_t +#ifdef RE_ENABLE_I18N +build_charclass (RE_TRANSLATE_TYPE trans, bitset_t sbcset, + re_charset_t *mbcset, Idx *char_class_alloc, + const unsigned char *class_name, reg_syntax_t syntax) +#else /* not RE_ENABLE_I18N */ +build_charclass (RE_TRANSLATE_TYPE trans, bitset_t sbcset, + const unsigned char *class_name, reg_syntax_t syntax) +#endif /* not RE_ENABLE_I18N */ +{ + int i; + const char *name = (const char *) class_name; + + /* In case of REG_ICASE "upper" and "lower" match the both of + upper and lower cases. */ + if ((syntax & RE_ICASE) + && (strcmp (name, "upper") == 0 || strcmp (name, "lower") == 0)) + name = "alpha"; + +#ifdef RE_ENABLE_I18N + /* Check the space of the arrays. */ + if (BE (*char_class_alloc == mbcset->nchar_classes, 0)) + { + /* Not enough, realloc it. */ + /* +1 in case of mbcset->nchar_classes is 0. */ + Idx new_char_class_alloc = 2 * mbcset->nchar_classes + 1; + /* Use realloc since array is NULL if *alloc == 0. */ + wctype_t *new_char_classes = re_realloc (mbcset->char_classes, wctype_t, + new_char_class_alloc); + if (BE (new_char_classes == NULL, 0)) + return REG_ESPACE; + mbcset->char_classes = new_char_classes; + *char_class_alloc = new_char_class_alloc; + } + mbcset->char_classes[mbcset->nchar_classes++] = __wctype (name); +#endif /* RE_ENABLE_I18N */ + +#define BUILD_CHARCLASS_LOOP(ctype_func) \ + do { \ + if (BE (trans != NULL, 0)) \ + { \ + for (i = 0; i < SBC_MAX; ++i) \ + if (ctype_func (i)) \ + bitset_set (sbcset, trans[i]); \ + } \ + else \ + { \ + for (i = 0; i < SBC_MAX; ++i) \ + if (ctype_func (i)) \ + bitset_set (sbcset, i); \ + } \ + } while (0) + + if (strcmp (name, "alnum") == 0) + BUILD_CHARCLASS_LOOP (isalnum); + else if (strcmp (name, "cntrl") == 0) + BUILD_CHARCLASS_LOOP (iscntrl); + else if (strcmp (name, "lower") == 0) + BUILD_CHARCLASS_LOOP (islower); + else if (strcmp (name, "space") == 0) + BUILD_CHARCLASS_LOOP (isspace); + else if (strcmp (name, "alpha") == 0) + BUILD_CHARCLASS_LOOP (isalpha); + else if (strcmp (name, "digit") == 0) + BUILD_CHARCLASS_LOOP (isdigit); + else if (strcmp (name, "print") == 0) + BUILD_CHARCLASS_LOOP (isprint); + else if (strcmp (name, "upper") == 0) + BUILD_CHARCLASS_LOOP (isupper); + else if (strcmp (name, "blank") == 0) + BUILD_CHARCLASS_LOOP (isblank); + else if (strcmp (name, "graph") == 0) + BUILD_CHARCLASS_LOOP (isgraph); + else if (strcmp (name, "punct") == 0) + BUILD_CHARCLASS_LOOP (ispunct); + else if (strcmp (name, "xdigit") == 0) + BUILD_CHARCLASS_LOOP (isxdigit); + else + return REG_ECTYPE; + + return REG_NOERROR; +} + +static bin_tree_t * +build_charclass_op (re_dfa_t *dfa, RE_TRANSLATE_TYPE trans, + const unsigned char *class_name, + const unsigned char *extra, bool non_match, + reg_errcode_t *err) +{ + re_bitset_ptr_t sbcset; +#ifdef RE_ENABLE_I18N + re_charset_t *mbcset; + Idx alloc = 0; +#endif /* not RE_ENABLE_I18N */ + reg_errcode_t ret; + re_token_t br_token; + bin_tree_t *tree; + + sbcset = (re_bitset_ptr_t) calloc (sizeof (bitset_t), 1); +#ifdef RE_ENABLE_I18N + mbcset = (re_charset_t *) calloc (sizeof (re_charset_t), 1); +#endif /* RE_ENABLE_I18N */ + +#ifdef RE_ENABLE_I18N + if (BE (sbcset == NULL || mbcset == NULL, 0)) +#else /* not RE_ENABLE_I18N */ + if (BE (sbcset == NULL, 0)) +#endif /* not RE_ENABLE_I18N */ + { + *err = REG_ESPACE; + return NULL; + } + + if (non_match) + { +#ifdef RE_ENABLE_I18N + /* + if (syntax & RE_HAT_LISTS_NOT_NEWLINE) + bitset_set(cset->sbcset, '\0'); + */ + mbcset->non_match = 1; +#endif /* not RE_ENABLE_I18N */ + } + + /* We don't care the syntax in this case. */ + ret = build_charclass (trans, sbcset, +#ifdef RE_ENABLE_I18N + mbcset, &alloc, +#endif /* RE_ENABLE_I18N */ + class_name, 0); + + if (BE (ret != REG_NOERROR, 0)) + { + re_free (sbcset); +#ifdef RE_ENABLE_I18N + free_charset (mbcset); +#endif /* RE_ENABLE_I18N */ + *err = ret; + return NULL; + } + /* \w match '_' also. */ + for (; *extra; extra++) + bitset_set (sbcset, *extra); + + /* If it is non-matching list. */ + if (non_match) + bitset_not (sbcset); + +#ifdef RE_ENABLE_I18N + /* Ensure only single byte characters are set. */ + if (dfa->mb_cur_max > 1) + bitset_mask (sbcset, dfa->sb_char); +#endif + + /* Build a tree for simple bracket. */ + br_token.type = SIMPLE_BRACKET; + br_token.opr.sbcset = sbcset; + tree = create_token_tree (dfa, NULL, NULL, &br_token); + if (BE (tree == NULL, 0)) + goto build_word_op_espace; + +#ifdef RE_ENABLE_I18N + if (dfa->mb_cur_max > 1) + { + bin_tree_t *mbc_tree; + /* Build a tree for complex bracket. */ + br_token.type = COMPLEX_BRACKET; + br_token.opr.mbcset = mbcset; + dfa->has_mb_node = 1; + mbc_tree = create_token_tree (dfa, NULL, NULL, &br_token); + if (BE (mbc_tree == NULL, 0)) + goto build_word_op_espace; + /* Then join them by ALT node. */ + tree = create_tree (dfa, tree, mbc_tree, OP_ALT); + if (BE (mbc_tree != NULL, 1)) + return tree; + } + else + { + free_charset (mbcset); + return tree; + } +#else /* not RE_ENABLE_I18N */ + return tree; +#endif /* not RE_ENABLE_I18N */ + + build_word_op_espace: + re_free (sbcset); +#ifdef RE_ENABLE_I18N + free_charset (mbcset); +#endif /* RE_ENABLE_I18N */ + *err = REG_ESPACE; + return NULL; +} + +/* This is intended for the expressions like "a{1,3}". + Fetch a number from `input', and return the number. + Return REG_MISSING if the number field is empty like "{,1}". + Return REG_ERROR if an error occurred. */ + +static Idx +fetch_number (re_string_t *input, re_token_t *token, reg_syntax_t syntax) +{ + Idx num = REG_MISSING; + unsigned char c; + while (1) + { + fetch_token (token, input, syntax); + c = token->opr.c; + if (BE (token->type == END_OF_RE, 0)) + return REG_ERROR; + if (token->type == OP_CLOSE_DUP_NUM || c == ',') + break; + num = ((token->type != CHARACTER || c < '0' || '9' < c + || num == REG_ERROR) + ? REG_ERROR + : ((num == REG_MISSING) ? c - '0' : num * 10 + c - '0')); + num = (num > RE_DUP_MAX) ? REG_ERROR : num; + } + return num; +} + +#ifdef RE_ENABLE_I18N +static void +free_charset (re_charset_t *cset) +{ + re_free (cset->mbchars); +# ifdef _LIBC + re_free (cset->coll_syms); + re_free (cset->equiv_classes); + re_free (cset->range_starts); + re_free (cset->range_ends); +# endif + re_free (cset->char_classes); + re_free (cset); +} +#endif /* RE_ENABLE_I18N */ + +/* Functions for binary tree operation. */ + +/* Create a tree node. */ + +static bin_tree_t * +create_tree (re_dfa_t *dfa, bin_tree_t *left, bin_tree_t *right, + re_token_type_t type) +{ + re_token_t t; + t.type = type; + return create_token_tree (dfa, left, right, &t); +} + +static bin_tree_t * +create_token_tree (re_dfa_t *dfa, bin_tree_t *left, bin_tree_t *right, + const re_token_t *token) +{ + bin_tree_t *tree; + if (BE (dfa->str_tree_storage_idx == BIN_TREE_STORAGE_SIZE, 0)) + { + bin_tree_storage_t *storage = re_malloc (bin_tree_storage_t, 1); + + if (storage == NULL) + return NULL; + storage->next = dfa->str_tree_storage; + dfa->str_tree_storage = storage; + dfa->str_tree_storage_idx = 0; + } + tree = &dfa->str_tree_storage->data[dfa->str_tree_storage_idx++]; + + tree->parent = NULL; + tree->left = left; + tree->right = right; + tree->token = *token; + tree->token.duplicated = 0; + tree->token.opt_subexp = 0; + tree->first = NULL; + tree->next = NULL; + tree->node_idx = REG_MISSING; + + if (left != NULL) + left->parent = tree; + if (right != NULL) + right->parent = tree; + return tree; +} + +/* Mark the tree SRC as an optional subexpression. + To be called from preorder or postorder. */ + +static reg_errcode_t +mark_opt_subexp (void *extra, bin_tree_t *node) +{ + Idx idx = (Idx) (long) extra; + if (node->token.type == SUBEXP && node->token.opr.idx == idx) + node->token.opt_subexp = 1; + + return REG_NOERROR; +} + +/* Free the allocated memory inside NODE. */ + +static void +free_token (re_token_t *node) +{ +#ifdef RE_ENABLE_I18N + if (node->type == COMPLEX_BRACKET && node->duplicated == 0) + free_charset (node->opr.mbcset); + else +#endif /* RE_ENABLE_I18N */ + if (node->type == SIMPLE_BRACKET && node->duplicated == 0) + re_free (node->opr.sbcset); +} + +/* Worker function for tree walking. Free the allocated memory inside NODE + and its children. */ + +static reg_errcode_t +free_tree (void *extra, bin_tree_t *node) +{ + free_token (&node->token); + return REG_NOERROR; +} + + +/* Duplicate the node SRC, and return new node. This is a preorder + visit similar to the one implemented by the generic visitor, but + we need more infrastructure to maintain two parallel trees --- so, + it's easier to duplicate. */ + +static bin_tree_t * +duplicate_tree (const bin_tree_t *root, re_dfa_t *dfa) +{ + const bin_tree_t *node; + bin_tree_t *dup_root; + bin_tree_t **p_new = &dup_root, *dup_node = root->parent; + + for (node = root; ; ) + { + /* Create a new tree and link it back to the current parent. */ + *p_new = create_token_tree (dfa, NULL, NULL, &node->token); + if (*p_new == NULL) + return NULL; + (*p_new)->parent = dup_node; + (*p_new)->token.duplicated = 1; + dup_node = *p_new; + + /* Go to the left node, or up and to the right. */ + if (node->left) + { + node = node->left; + p_new = &dup_node->left; + } + else + { + const bin_tree_t *prev = NULL; + while (node->right == prev || node->right == NULL) + { + prev = node; + node = node->parent; + dup_node = dup_node->parent; + if (!node) + return dup_root; + } + node = node->right; + p_new = &dup_node->right; + } + } +} diff --git a/gl/regex.c b/gl/regex.c new file mode 100644 index 00000000..d4eb726b --- /dev/null +++ b/gl/regex.c @@ -0,0 +1,71 @@ +/* Extended regular expression matching and search library. + Copyright (C) 2002, 2003, 2005, 2006 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Make sure noone compiles this code with a C++ compiler. */ +#if defined __cplusplus && defined _LIBC +# error "This is C code, use a C compiler" +#endif + +#ifdef _LIBC +/* We have to keep the namespace clean. */ +# define regfree(preg) __regfree (preg) +# define regexec(pr, st, nm, pm, ef) __regexec (pr, st, nm, pm, ef) +# define regcomp(preg, pattern, cflags) __regcomp (preg, pattern, cflags) +# define regerror(errcode, preg, errbuf, errbuf_size) \ + __regerror(errcode, preg, errbuf, errbuf_size) +# define re_set_registers(bu, re, nu, st, en) \ + __re_set_registers (bu, re, nu, st, en) +# define re_match_2(bufp, string1, size1, string2, size2, pos, regs, stop) \ + __re_match_2 (bufp, string1, size1, string2, size2, pos, regs, stop) +# define re_match(bufp, string, size, pos, regs) \ + __re_match (bufp, string, size, pos, regs) +# define re_search(bufp, string, size, startpos, range, regs) \ + __re_search (bufp, string, size, startpos, range, regs) +# define re_compile_pattern(pattern, length, bufp) \ + __re_compile_pattern (pattern, length, bufp) +# define re_set_syntax(syntax) __re_set_syntax (syntax) +# define re_search_2(bufp, st1, s1, st2, s2, startpos, range, regs, stop) \ + __re_search_2 (bufp, st1, s1, st2, s2, startpos, range, regs, stop) +# define re_compile_fastmap(bufp) __re_compile_fastmap (bufp) + +# include "../locale/localeinfo.h" +#endif + +/* On some systems, limits.h sets RE_DUP_MAX to a lower value than + GNU regex allows. Include it before <regex.h>, which correctly + #undefs RE_DUP_MAX and sets it to the right value. */ +#include <limits.h> + +#include <regex.h> +#include "regex_internal.h" + +#include "regex_internal.c" +#include "regcomp.c" +#include "regexec.c" + +/* Binary backward compatibility. */ +#if _LIBC +# include <shlib-compat.h> +# if SHLIB_COMPAT (libc, GLIBC_2_0, GLIBC_2_3) +link_warning (re_max_failures, "the 're_max_failures' variable is obsolete and will go away.") +int re_max_failures = 2000; +# endif +#endif diff --git a/gl/regex.h b/gl/regex.h new file mode 100644 index 00000000..6885ebdf --- /dev/null +++ b/gl/regex.h @@ -0,0 +1,671 @@ +/* Definitions for data structures and routines for the regular + expression library. + Copyright (C) 1985,1989-93,1995-98,2000,2001,2002,2003,2005,2006 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _REGEX_H +#define _REGEX_H 1 + +#include <sys/types.h> + +/* Allow the use in C++ code. */ +#ifdef __cplusplus +extern "C" { +#endif + +/* Define __USE_GNU_REGEX to declare GNU extensions that violate the + POSIX name space rules. */ +#undef __USE_GNU_REGEX +#if (defined _GNU_SOURCE \ + || (!defined _POSIX_C_SOURCE && !defined _POSIX_SOURCE \ + && !defined _XOPEN_SOURCE)) +# define __USE_GNU_REGEX 1 +#endif + +#ifdef _REGEX_LARGE_OFFSETS + +/* Use types and values that are wide enough to represent signed and + unsigned byte offsets in memory. This currently works only when + the regex code is used outside of the GNU C library; it is not yet + supported within glibc itself, and glibc users should not define + _REGEX_LARGE_OFFSETS. */ + +/* The type of the offset of a byte within a string. + For historical reasons POSIX 1003.1-2004 requires that regoff_t be + at least as wide as off_t. However, many common POSIX platforms set + regoff_t to the more-sensible ssize_t and the Open Group has + signalled its intention to change the requirement to be that + regoff_t be at least as wide as ptrdiff_t and ssize_t; see XBD ERN + 60 (2005-08-25). We don't know of any hosts where ssize_t or + ptrdiff_t is wider than ssize_t, so ssize_t is safe. */ +typedef ssize_t regoff_t; + +/* The type of nonnegative object indexes. Traditionally, GNU regex + uses 'int' for these. Code that uses __re_idx_t should work + regardless of whether the type is signed. */ +typedef size_t __re_idx_t; + +/* The type of object sizes. */ +typedef size_t __re_size_t; + +/* The type of object sizes, in places where the traditional code + uses unsigned long int. */ +typedef size_t __re_long_size_t; + +#else + +/* Use types that are binary-compatible with the traditional GNU regex + implementation, which mishandles strings longer than INT_MAX. */ + +typedef int regoff_t; +typedef int __re_idx_t; +typedef unsigned int __re_size_t; +typedef unsigned long int __re_long_size_t; + +#endif + +/* The following two types have to be signed and unsigned integer type + wide enough to hold a value of a pointer. For most ANSI compilers + ptrdiff_t and size_t should be likely OK. Still size of these two + types is 2 for Microsoft C. Ugh... */ +typedef long int s_reg_t; +typedef unsigned long int active_reg_t; + +/* The following bits are used to determine the regexp syntax we + recognize. The set/not-set meanings are chosen so that Emacs syntax + remains the value 0. The bits are given in alphabetical order, and + the definitions shifted by one from the previous bit; thus, when we + add or remove a bit, only one other definition need change. */ +typedef unsigned long int reg_syntax_t; + +#ifdef __USE_GNU_REGEX + +/* If this bit is not set, then \ inside a bracket expression is literal. + If set, then such a \ quotes the following character. */ +# define RE_BACKSLASH_ESCAPE_IN_LISTS ((unsigned long int) 1) + +/* If this bit is not set, then + and ? are operators, and \+ and \? are + literals. + If set, then \+ and \? are operators and + and ? are literals. */ +# define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1) + +/* If this bit is set, then character classes are supported. They are: + [:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:], + [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:]. + If not set, then character classes are not supported. */ +# define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1) + +/* If this bit is set, then ^ and $ are always anchors (outside bracket + expressions, of course). + If this bit is not set, then it depends: + ^ is an anchor if it is at the beginning of a regular + expression or after an open-group or an alternation operator; + $ is an anchor if it is at the end of a regular expression, or + before a close-group or an alternation operator. + + This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because + POSIX draft 11.2 says that * etc. in leading positions is undefined. + We already implemented a previous draft which made those constructs + invalid, though, so we haven't changed the code back. */ +# define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1) + +/* If this bit is set, then special characters are always special + regardless of where they are in the pattern. + If this bit is not set, then special characters are special only in + some contexts; otherwise they are ordinary. Specifically, + * + ? and intervals are only special when not after the beginning, + open-group, or alternation operator. */ +# define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1) + +/* If this bit is set, then *, +, ?, and { cannot be first in an re or + immediately after an alternation or begin-group operator. */ +# define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1) + +/* If this bit is set, then . matches newline. + If not set, then it doesn't. */ +# define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1) + +/* If this bit is set, then . doesn't match NUL. + If not set, then it does. */ +# define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1) + +/* If this bit is set, nonmatching lists [^...] do not match newline. + If not set, they do. */ +# define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1) + +/* If this bit is set, either \{...\} or {...} defines an + interval, depending on RE_NO_BK_BRACES. + If not set, \{, \}, {, and } are literals. */ +# define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1) + +/* If this bit is set, +, ? and | aren't recognized as operators. + If not set, they are. */ +# define RE_LIMITED_OPS (RE_INTERVALS << 1) + +/* If this bit is set, newline is an alternation operator. + If not set, newline is literal. */ +# define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1) + +/* If this bit is set, then `{...}' defines an interval, and \{ and \} + are literals. + If not set, then `\{...\}' defines an interval. */ +# define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1) + +/* If this bit is set, (...) defines a group, and \( and \) are literals. + If not set, \(...\) defines a group, and ( and ) are literals. */ +# define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1) + +/* If this bit is set, then \<digit> matches <digit>. + If not set, then \<digit> is a back-reference. */ +# define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1) + +/* If this bit is set, then | is an alternation operator, and \| is literal. + If not set, then \| is an alternation operator, and | is literal. */ +# define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1) + +/* If this bit is set, then an ending range point collating higher + than the starting range point, as in [z-a], is invalid. + If not set, then when ending range point collates higher than the + starting range point, the range is ignored. */ +# define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1) + +/* If this bit is set, then an unmatched ) is ordinary. + If not set, then an unmatched ) is invalid. */ +# define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1) + +/* If this bit is set, succeed as soon as we match the whole pattern, + without further backtracking. */ +# define RE_NO_POSIX_BACKTRACKING (RE_UNMATCHED_RIGHT_PAREN_ORD << 1) + +/* If this bit is set, do not process the GNU regex operators. + If not set, then the GNU regex operators are recognized. */ +# define RE_NO_GNU_OPS (RE_NO_POSIX_BACKTRACKING << 1) + +/* If this bit is set, turn on internal regex debugging. + If not set, and debugging was on, turn it off. + This only works if regex.c is compiled -DDEBUG. + We define this bit always, so that all that's needed to turn on + debugging is to recompile regex.c; the calling code can always have + this bit set, and it won't affect anything in the normal case. */ +# define RE_DEBUG (RE_NO_GNU_OPS << 1) + +/* If this bit is set, a syntactically invalid interval is treated as + a string of ordinary characters. For example, the ERE 'a{1' is + treated as 'a\{1'. */ +# define RE_INVALID_INTERVAL_ORD (RE_DEBUG << 1) + +/* If this bit is set, then ignore case when matching. + If not set, then case is significant. */ +# define RE_ICASE (RE_INVALID_INTERVAL_ORD << 1) + +/* This bit is used internally like RE_CONTEXT_INDEP_ANCHORS but only + for ^, because it is difficult to scan the regex backwards to find + whether ^ should be special. */ +# define RE_CARET_ANCHORS_HERE (RE_ICASE << 1) + +/* If this bit is set, then \{ cannot be first in an bre or + immediately after an alternation or begin-group operator. */ +# define RE_CONTEXT_INVALID_DUP (RE_CARET_ANCHORS_HERE << 1) + +/* If this bit is set, then no_sub will be set to 1 during + re_compile_pattern. */ +# define RE_NO_SUB (RE_CONTEXT_INVALID_DUP << 1) + +#endif /* defined __USE_GNU_REGEX */ + +/* This global variable defines the particular regexp syntax to use (for + some interfaces). When a regexp is compiled, the syntax used is + stored in the pattern buffer, so changing this does not affect + already-compiled regexps. */ +extern reg_syntax_t re_syntax_options; + +#ifdef __USE_GNU_REGEX +/* Define combinations of the above bits for the standard possibilities. + (The [[[ comments delimit what gets put into the Texinfo file, so + don't delete them!) */ +/* [[[begin syntaxes]]] */ +# define RE_SYNTAX_EMACS 0 + +# define RE_SYNTAX_AWK \ + (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \ + | RE_NO_BK_PARENS | RE_NO_BK_REFS \ + | RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \ + | RE_DOT_NEWLINE | RE_CONTEXT_INDEP_ANCHORS \ + | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_GNU_OPS) + +# define RE_SYNTAX_GNU_AWK \ + ((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DEBUG) \ + & ~(RE_DOT_NOT_NULL | RE_INTERVALS | RE_CONTEXT_INDEP_OPS \ + | RE_CONTEXT_INVALID_OPS )) + +# define RE_SYNTAX_POSIX_AWK \ + (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \ + | RE_INTERVALS | RE_NO_GNU_OPS) + +# define RE_SYNTAX_GREP \ + (RE_BK_PLUS_QM | RE_CHAR_CLASSES \ + | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS \ + | RE_NEWLINE_ALT) + +# define RE_SYNTAX_EGREP \ + (RE_CHAR_CLASSES | RE_CONTEXT_INDEP_ANCHORS \ + | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE \ + | RE_NEWLINE_ALT | RE_NO_BK_PARENS \ + | RE_NO_BK_VBAR) + +# define RE_SYNTAX_POSIX_EGREP \ + (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES \ + | RE_INVALID_INTERVAL_ORD) + +/* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */ +# define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC + +# define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC + +/* Syntax bits common to both basic and extended POSIX regex syntax. */ +# define _RE_SYNTAX_POSIX_COMMON \ + (RE_CHAR_CLASSES | RE_DOT_NEWLINE | RE_DOT_NOT_NULL \ + | RE_INTERVALS | RE_NO_EMPTY_RANGES) + +# define RE_SYNTAX_POSIX_BASIC \ + (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM | RE_CONTEXT_INVALID_DUP) + +/* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes + RE_LIMITED_OPS, i.e., \? \+ \| are not recognized. Actually, this + isn't minimal, since other operators, such as \`, aren't disabled. */ +# define RE_SYNTAX_POSIX_MINIMAL_BASIC \ + (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS) + +# define RE_SYNTAX_POSIX_EXTENDED \ + (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ + | RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \ + | RE_NO_BK_PARENS | RE_NO_BK_VBAR \ + | RE_CONTEXT_INVALID_OPS | RE_UNMATCHED_RIGHT_PAREN_ORD) + +/* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INDEP_OPS is + removed and RE_NO_BK_REFS is added. */ +# define RE_SYNTAX_POSIX_MINIMAL_EXTENDED \ + (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ + | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES \ + | RE_NO_BK_PARENS | RE_NO_BK_REFS \ + | RE_NO_BK_VBAR | RE_UNMATCHED_RIGHT_PAREN_ORD) +/* [[[end syntaxes]]] */ + +#endif /* defined __USE_GNU_REGEX */ + +#ifdef __USE_GNU_REGEX + +/* Maximum number of duplicates an interval can allow. POSIX-conforming + systems might define this in <limits.h>, but we want our + value, so remove any previous define. */ +# ifdef RE_DUP_MAX +# undef RE_DUP_MAX +# endif + +/* RE_DUP_MAX is 2**15 - 1 because an earlier implementation stored + the counter as a 2-byte signed integer. This is no longer true, so + RE_DUP_MAX could be increased to (INT_MAX / 10 - 1), or to + ((SIZE_MAX - 2) / 10 - 1) if _REGEX_LARGE_OFFSETS is defined. + However, there would be a huge performance problem if someone + actually used a pattern like a\{214748363\}, so RE_DUP_MAX retains + its historical value. */ +# define RE_DUP_MAX (0x7fff) + +#endif /* defined __USE_GNU_REGEX */ + + +/* POSIX `cflags' bits (i.e., information for `regcomp'). */ + +/* If this bit is set, then use extended regular expression syntax. + If not set, then use basic regular expression syntax. */ +#define REG_EXTENDED 1 + +/* If this bit is set, then ignore case when matching. + If not set, then case is significant. */ +#define REG_ICASE (1 << 1) + +/* If this bit is set, then anchors do not match at newline + characters in the string. + If not set, then anchors do match at newlines. */ +#define REG_NEWLINE (1 << 2) + +/* If this bit is set, then report only success or fail in regexec. + If not set, then returns differ between not matching and errors. */ +#define REG_NOSUB (1 << 3) + + +/* POSIX `eflags' bits (i.e., information for regexec). */ + +/* If this bit is set, then the beginning-of-line operator doesn't match + the beginning of the string (presumably because it's not the + beginning of a line). + If not set, then the beginning-of-line operator does match the + beginning of the string. */ +#define REG_NOTBOL 1 + +/* Like REG_NOTBOL, except for the end-of-line. */ +#define REG_NOTEOL (1 << 1) + +/* Use PMATCH[0] to delimit the start and end of the search in the + buffer. */ +#define REG_STARTEND (1 << 2) + + +/* If any error codes are removed, changed, or added, update the + `__re_error_msgid' table in regcomp.c. */ + +typedef enum +{ + _REG_ENOSYS = -1, /* This will never happen for this implementation. */ + _REG_NOERROR = 0, /* Success. */ + _REG_NOMATCH, /* Didn't find a match (for regexec). */ + + /* POSIX regcomp return error codes. (In the order listed in the + standard.) */ + _REG_BADPAT, /* Invalid pattern. */ + _REG_ECOLLATE, /* Invalid collating element. */ + _REG_ECTYPE, /* Invalid character class name. */ + _REG_EESCAPE, /* Trailing backslash. */ + _REG_ESUBREG, /* Invalid back reference. */ + _REG_EBRACK, /* Unmatched left bracket. */ + _REG_EPAREN, /* Parenthesis imbalance. */ + _REG_EBRACE, /* Unmatched \{. */ + _REG_BADBR, /* Invalid contents of \{\}. */ + _REG_ERANGE, /* Invalid range end. */ + _REG_ESPACE, /* Ran out of memory. */ + _REG_BADRPT, /* No preceding re for repetition op. */ + + /* Error codes we've added. */ + _REG_EEND, /* Premature end. */ + _REG_ESIZE, /* Compiled pattern bigger than 2^16 bytes. */ + _REG_ERPAREN /* Unmatched ) or \); not returned from regcomp. */ +} reg_errcode_t; + +#ifdef _XOPEN_SOURCE +# define REG_ENOSYS _REG_ENOSYS +#endif +#define REG_NOERROR _REG_NOERROR +#define REG_NOMATCH _REG_NOMATCH +#define REG_BADPAT _REG_BADPAT +#define REG_ECOLLATE _REG_ECOLLATE +#define REG_ECTYPE _REG_ECTYPE +#define REG_EESCAPE _REG_EESCAPE +#define REG_ESUBREG _REG_ESUBREG +#define REG_EBRACK _REG_EBRACK +#define REG_EPAREN _REG_EPAREN +#define REG_EBRACE _REG_EBRACE +#define REG_BADBR _REG_BADBR +#define REG_ERANGE _REG_ERANGE +#define REG_ESPACE _REG_ESPACE +#define REG_BADRPT _REG_BADRPT +#define REG_EEND _REG_EEND +#define REG_ESIZE _REG_ESIZE +#define REG_ERPAREN _REG_ERPAREN + +/* struct re_pattern_buffer normally uses member names like `buffer' + that POSIX does not allow. In POSIX mode these members have names + with leading `re_' (e.g., `re_buffer'). */ +#ifdef __USE_GNU_REGEX +# define _REG_RE_NAME(id) id +# define _REG_RM_NAME(id) id +#else +# define _REG_RE_NAME(id) re_##id +# define _REG_RM_NAME(id) rm_##id +#endif + +/* The user can specify the type of the re_translate member by + defining the macro RE_TRANSLATE_TYPE, which defaults to unsigned + char *. This pollutes the POSIX name space, so in POSIX mode just + use unsigned char *. */ +#ifdef __USE_GNU_REGEX +# ifndef RE_TRANSLATE_TYPE +# define RE_TRANSLATE_TYPE unsigned char * +# endif +# define REG_TRANSLATE_TYPE RE_TRANSLATE_TYPE +#else +# define REG_TRANSLATE_TYPE unsigned char * +#endif + +/* This data structure represents a compiled pattern. Before calling + the pattern compiler, the fields `buffer', `allocated', `fastmap', + `translate', and `no_sub' can be set. After the pattern has been + compiled, the `re_nsub' field is available. All other fields are + private to the regex routines. */ + +struct re_pattern_buffer +{ + /* Space that holds the compiled pattern. It is declared as + `unsigned char *' because its elements are sometimes used as + array indexes. */ + unsigned char *_REG_RE_NAME (buffer); + + /* Number of bytes to which `buffer' points. */ + __re_long_size_t _REG_RE_NAME (allocated); + + /* Number of bytes actually used in `buffer'. */ + __re_long_size_t _REG_RE_NAME (used); + + /* Syntax setting with which the pattern was compiled. */ + reg_syntax_t _REG_RE_NAME (syntax); + + /* Pointer to a fastmap, if any, otherwise zero. re_search uses the + fastmap, if there is one, to skip over impossible starting points + for matches. */ + char *_REG_RE_NAME (fastmap); + + /* Either a translate table to apply to all characters before + comparing them, or zero for no translation. The translation is + applied to a pattern when it is compiled and to a string when it + is matched. */ + REG_TRANSLATE_TYPE _REG_RE_NAME (translate); + + /* Number of subexpressions found by the compiler. */ + size_t re_nsub; + + /* Zero if this pattern cannot match the empty string, one else. + Well, in truth it's used only in `re_search_2', to see whether or + not we should use the fastmap, so we don't set this absolutely + perfectly; see `re_compile_fastmap' (the `duplicate' case). */ + unsigned int _REG_RE_NAME (can_be_null) : 1; + + /* If REGS_UNALLOCATED, allocate space in the `regs' structure + for `max (RE_NREGS, re_nsub + 1)' groups. + If REGS_REALLOCATE, reallocate space if necessary. + If REGS_FIXED, use what's there. */ +#ifdef __USE_GNU_REGEX +# define REGS_UNALLOCATED 0 +# define REGS_REALLOCATE 1 +# define REGS_FIXED 2 +#endif + unsigned int _REG_RE_NAME (regs_allocated) : 2; + + /* Set to zero when `regex_compile' compiles a pattern; set to one + by `re_compile_fastmap' if it updates the fastmap. */ + unsigned int _REG_RE_NAME (fastmap_accurate) : 1; + + /* If set, `re_match_2' does not return information about + subexpressions. */ + unsigned int _REG_RE_NAME (no_sub) : 1; + + /* If set, a beginning-of-line anchor doesn't match at the beginning + of the string. */ + unsigned int _REG_RE_NAME (not_bol) : 1; + + /* Similarly for an end-of-line anchor. */ + unsigned int _REG_RE_NAME (not_eol) : 1; + + /* If true, an anchor at a newline matches. */ + unsigned int _REG_RE_NAME (newline_anchor) : 1; + +/* [[[end pattern_buffer]]] */ +}; + +typedef struct re_pattern_buffer regex_t; + +/* This is the structure we store register match data in. See + regex.texinfo for a full description of what registers match. */ +struct re_registers +{ + __re_size_t _REG_RM_NAME (num_regs); + regoff_t *_REG_RM_NAME (start); + regoff_t *_REG_RM_NAME (end); +}; + + +/* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer, + `re_match_2' returns information about at least this many registers + the first time a `regs' structure is passed. */ +#if !defined RE_NREGS && defined __USE_GNU_REGEX +# define RE_NREGS 30 +#endif + + +/* POSIX specification for registers. Aside from the different names than + `re_registers', POSIX uses an array of structures, instead of a + structure of arrays. */ +typedef struct +{ + regoff_t rm_so; /* Byte offset from string's start to substring's start. */ + regoff_t rm_eo; /* Byte offset from string's start to substring's end. */ +} regmatch_t; + +/* Declarations for routines. */ + +/* Sets the current default syntax to SYNTAX, and return the old syntax. + You can also simply assign to the `re_syntax_options' variable. */ +extern reg_syntax_t re_set_syntax (reg_syntax_t __syntax); + +/* Compile the regular expression PATTERN, with length LENGTH + and syntax given by the global `re_syntax_options', into the buffer + BUFFER. Return NULL if successful, and an error string if not. */ +extern const char *re_compile_pattern (const char *__pattern, size_t __length, + struct re_pattern_buffer *__buffer); + + +/* Compile a fastmap for the compiled pattern in BUFFER; used to + accelerate searches. Return 0 if successful and -2 if was an + internal error. */ +extern int re_compile_fastmap (struct re_pattern_buffer *__buffer); + + +/* Search in the string STRING (with length LENGTH) for the pattern + compiled into BUFFER. Start searching at position START, for RANGE + characters. Return the starting position of the match, -1 for no + match, or -2 for an internal error. Also return register + information in REGS (if REGS and BUFFER->no_sub are nonzero). */ +extern regoff_t re_search (struct re_pattern_buffer *__buffer, + const char *__string, __re_idx_t __length, + __re_idx_t __start, regoff_t __range, + struct re_registers *__regs); + + +/* Like `re_search', but search in the concatenation of STRING1 and + STRING2. Also, stop searching at index START + STOP. */ +extern regoff_t re_search_2 (struct re_pattern_buffer *__buffer, + const char *__string1, __re_idx_t __length1, + const char *__string2, __re_idx_t __length2, + __re_idx_t __start, regoff_t __range, + struct re_registers *__regs, + __re_idx_t __stop); + + +/* Like `re_search', but return how many characters in STRING the regexp + in BUFFER matched, starting at position START. */ +extern regoff_t re_match (struct re_pattern_buffer *__buffer, + const char *__string, __re_idx_t __length, + __re_idx_t __start, struct re_registers *__regs); + + +/* Relates to `re_match' as `re_search_2' relates to `re_search'. */ +extern regoff_t re_match_2 (struct re_pattern_buffer *__buffer, + const char *__string1, __re_idx_t __length1, + const char *__string2, __re_idx_t __length2, + __re_idx_t __start, struct re_registers *__regs, + __re_idx_t __stop); + + +/* Set REGS to hold NUM_REGS registers, storing them in STARTS and + ENDS. Subsequent matches using BUFFER and REGS will use this memory + for recording register information. STARTS and ENDS must be + allocated with malloc, and must each be at least `NUM_REGS * sizeof + (regoff_t)' bytes long. + + If NUM_REGS == 0, then subsequent matches should allocate their own + register data. + + Unless this function is called, the first search or match using + PATTERN_BUFFER will allocate its own register data, without + freeing the old data. */ +extern void re_set_registers (struct re_pattern_buffer *__buffer, + struct re_registers *__regs, + __re_size_t __num_regs, + regoff_t *__starts, regoff_t *__ends); + +#if defined _REGEX_RE_COMP || defined _LIBC +# ifndef _CRAY +/* 4.2 bsd compatibility. */ +extern char *re_comp (const char *); +extern int re_exec (const char *); +# endif +#endif + +/* GCC 2.95 and later have "__restrict"; C99 compilers have + "restrict", and "configure" may have defined "restrict". */ +#ifndef __restrict +# if ! (2 < __GNUC__ || (2 == __GNUC__ && 95 <= __GNUC_MINOR__)) +# if defined restrict || 199901L <= __STDC_VERSION__ +# define __restrict restrict +# else +# define __restrict +# endif +# endif +#endif +/* gcc 3.1 and up support the [restrict] syntax. Don't trust + sys/cdefs.h's definition of __restrict_arr, though, as it + mishandles gcc -ansi -pedantic. */ +#undef __restrict_arr +#if ((199901L <= __STDC_VERSION__ \ + || ((3 < __GNUC__ || (3 == __GNUC__ && 1 <= __GNUC_MINOR__)) \ + && !__STRICT_ANSI__)) \ + && !defined __GNUG__) +# define __restrict_arr __restrict +#else +# define __restrict_arr +#endif + +/* POSIX compatibility. */ +extern int regcomp (regex_t *__restrict __preg, + const char *__restrict __pattern, + int __cflags); + +extern int regexec (const regex_t *__restrict __preg, + const char *__restrict __string, size_t __nmatch, + regmatch_t __pmatch[__restrict_arr], + int __eflags); + +extern size_t regerror (int __errcode, const regex_t *__restrict __preg, + char *__restrict __errbuf, size_t __errbuf_size); + +extern void regfree (regex_t *__preg); + + +#ifdef __cplusplus +} +#endif /* C++ */ + +#endif /* regex.h */ diff --git a/gl/regex_internal.c b/gl/regex_internal.c new file mode 100644 index 00000000..78e16f33 --- /dev/null +++ b/gl/regex_internal.c @@ -0,0 +1,1742 @@ +/* Extended regular expression matching and search library. + Copyright (C) 2002, 2003, 2004, 2005, 2006 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +static void re_string_construct_common (const char *str, Idx len, + re_string_t *pstr, + RE_TRANSLATE_TYPE trans, bool icase, + const re_dfa_t *dfa) internal_function; +static re_dfastate_t *create_ci_newstate (const re_dfa_t *dfa, + const re_node_set *nodes, + re_hashval_t hash) internal_function; +static re_dfastate_t *create_cd_newstate (const re_dfa_t *dfa, + const re_node_set *nodes, + unsigned int context, + re_hashval_t hash) internal_function; + +/* Functions for string operation. */ + +/* This function allocate the buffers. It is necessary to call + re_string_reconstruct before using the object. */ + +static reg_errcode_t +internal_function +re_string_allocate (re_string_t *pstr, const char *str, Idx len, Idx init_len, + RE_TRANSLATE_TYPE trans, bool icase, const re_dfa_t *dfa) +{ + reg_errcode_t ret; + Idx init_buf_len; + + /* Ensure at least one character fits into the buffers. */ + if (init_len < dfa->mb_cur_max) + init_len = dfa->mb_cur_max; + init_buf_len = (len + 1 < init_len) ? len + 1: init_len; + re_string_construct_common (str, len, pstr, trans, icase, dfa); + + ret = re_string_realloc_buffers (pstr, init_buf_len); + if (BE (ret != REG_NOERROR, 0)) + return ret; + + pstr->word_char = dfa->word_char; + pstr->word_ops_used = dfa->word_ops_used; + pstr->mbs = pstr->mbs_allocated ? pstr->mbs : (unsigned char *) str; + pstr->valid_len = (pstr->mbs_allocated || dfa->mb_cur_max > 1) ? 0 : len; + pstr->valid_raw_len = pstr->valid_len; + return REG_NOERROR; +} + +/* This function allocate the buffers, and initialize them. */ + +static reg_errcode_t +internal_function +re_string_construct (re_string_t *pstr, const char *str, Idx len, + RE_TRANSLATE_TYPE trans, bool icase, const re_dfa_t *dfa) +{ + reg_errcode_t ret; + memset (pstr, '\0', sizeof (re_string_t)); + re_string_construct_common (str, len, pstr, trans, icase, dfa); + + if (len > 0) + { + ret = re_string_realloc_buffers (pstr, len + 1); + if (BE (ret != REG_NOERROR, 0)) + return ret; + } + pstr->mbs = pstr->mbs_allocated ? pstr->mbs : (unsigned char *) str; + + if (icase) + { +#ifdef RE_ENABLE_I18N + if (dfa->mb_cur_max > 1) + { + while (1) + { + ret = build_wcs_upper_buffer (pstr); + if (BE (ret != REG_NOERROR, 0)) + return ret; + if (pstr->valid_raw_len >= len) + break; + if (pstr->bufs_len > pstr->valid_len + dfa->mb_cur_max) + break; + ret = re_string_realloc_buffers (pstr, pstr->bufs_len * 2); + if (BE (ret != REG_NOERROR, 0)) + return ret; + } + } + else +#endif /* RE_ENABLE_I18N */ + build_upper_buffer (pstr); + } + else + { +#ifdef RE_ENABLE_I18N + if (dfa->mb_cur_max > 1) + build_wcs_buffer (pstr); + else +#endif /* RE_ENABLE_I18N */ + { + if (trans != NULL) + re_string_translate_buffer (pstr); + else + { + pstr->valid_len = pstr->bufs_len; + pstr->valid_raw_len = pstr->bufs_len; + } + } + } + + return REG_NOERROR; +} + +/* Helper functions for re_string_allocate, and re_string_construct. */ + +static reg_errcode_t +internal_function +re_string_realloc_buffers (re_string_t *pstr, Idx new_buf_len) +{ +#ifdef RE_ENABLE_I18N + if (pstr->mb_cur_max > 1) + { + wint_t *new_wcs; + + /* Avoid overflow. */ + size_t max_object_size = MAX (sizeof (wint_t), sizeof (Idx)); + if (BE (SIZE_MAX / max_object_size < new_buf_len, 0)) + return REG_ESPACE; + + new_wcs = re_realloc (pstr->wcs, wint_t, new_buf_len); + if (BE (new_wcs == NULL, 0)) + return REG_ESPACE; + pstr->wcs = new_wcs; + if (pstr->offsets != NULL) + { + Idx *new_offsets = re_realloc (pstr->offsets, Idx, new_buf_len); + if (BE (new_offsets == NULL, 0)) + return REG_ESPACE; + pstr->offsets = new_offsets; + } + } +#endif /* RE_ENABLE_I18N */ + if (pstr->mbs_allocated) + { + unsigned char *new_mbs = re_realloc (pstr->mbs, unsigned char, + new_buf_len); + if (BE (new_mbs == NULL, 0)) + return REG_ESPACE; + pstr->mbs = new_mbs; + } + pstr->bufs_len = new_buf_len; + return REG_NOERROR; +} + + +static void +internal_function +re_string_construct_common (const char *str, Idx len, re_string_t *pstr, + RE_TRANSLATE_TYPE trans, bool icase, + const re_dfa_t *dfa) +{ + pstr->raw_mbs = (const unsigned char *) str; + pstr->len = len; + pstr->raw_len = len; + pstr->trans = trans; + pstr->icase = icase; + pstr->mbs_allocated = (trans != NULL || icase); + pstr->mb_cur_max = dfa->mb_cur_max; + pstr->is_utf8 = dfa->is_utf8; + pstr->map_notascii = dfa->map_notascii; + pstr->stop = pstr->len; + pstr->raw_stop = pstr->stop; +} + +#ifdef RE_ENABLE_I18N + +/* Build wide character buffer PSTR->WCS. + If the byte sequence of the string are: + <mb1>(0), <mb1>(1), <mb2>(0), <mb2>(1), <sb3> + Then wide character buffer will be: + <wc1> , WEOF , <wc2> , WEOF , <wc3> + We use WEOF for padding, they indicate that the position isn't + a first byte of a multibyte character. + + Note that this function assumes PSTR->VALID_LEN elements are already + built and starts from PSTR->VALID_LEN. */ + +static void +internal_function +build_wcs_buffer (re_string_t *pstr) +{ +#ifdef _LIBC + unsigned char buf[MB_LEN_MAX]; + assert (MB_LEN_MAX >= pstr->mb_cur_max); +#else + unsigned char buf[64]; +#endif + mbstate_t prev_st; + Idx byte_idx, end_idx, remain_len; + size_t mbclen; + + /* Build the buffers from pstr->valid_len to either pstr->len or + pstr->bufs_len. */ + end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len; + for (byte_idx = pstr->valid_len; byte_idx < end_idx;) + { + wchar_t wc; + const char *p; + + remain_len = end_idx - byte_idx; + prev_st = pstr->cur_state; + /* Apply the translation if we need. */ + if (BE (pstr->trans != NULL, 0)) + { + int i, ch; + + for (i = 0; i < pstr->mb_cur_max && i < remain_len; ++i) + { + ch = pstr->raw_mbs [pstr->raw_mbs_idx + byte_idx + i]; + buf[i] = pstr->mbs[byte_idx + i] = pstr->trans[ch]; + } + p = (const char *) buf; + } + else + p = (const char *) pstr->raw_mbs + pstr->raw_mbs_idx + byte_idx; + mbclen = mbrtowc (&wc, p, remain_len, &pstr->cur_state); + if (BE (mbclen == (size_t) -2, 0)) + { + /* The buffer doesn't have enough space, finish to build. */ + pstr->cur_state = prev_st; + break; + } + else if (BE (mbclen == (size_t) -1 || mbclen == 0, 0)) + { + /* We treat these cases as a singlebyte character. */ + mbclen = 1; + wc = (wchar_t) pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]; + if (BE (pstr->trans != NULL, 0)) + wc = pstr->trans[wc]; + pstr->cur_state = prev_st; + } + + /* Write wide character and padding. */ + pstr->wcs[byte_idx++] = wc; + /* Write paddings. */ + for (remain_len = byte_idx + mbclen - 1; byte_idx < remain_len ;) + pstr->wcs[byte_idx++] = WEOF; + } + pstr->valid_len = byte_idx; + pstr->valid_raw_len = byte_idx; +} + +/* Build wide character buffer PSTR->WCS like build_wcs_buffer, + but for REG_ICASE. */ + +static reg_errcode_t +internal_function +build_wcs_upper_buffer (re_string_t *pstr) +{ + mbstate_t prev_st; + Idx src_idx, byte_idx, end_idx, remain_len; + size_t mbclen; +#ifdef _LIBC + char buf[MB_LEN_MAX]; + assert (MB_LEN_MAX >= pstr->mb_cur_max); +#else + char buf[64]; +#endif + + byte_idx = pstr->valid_len; + end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len; + + /* The following optimization assumes that ASCII characters can be + mapped to wide characters with a simple cast. */ + if (! pstr->map_notascii && pstr->trans == NULL && !pstr->offsets_needed) + { + while (byte_idx < end_idx) + { + wchar_t wc; + + if (isascii (pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]) + && mbsinit (&pstr->cur_state)) + { + /* In case of a singlebyte character. */ + pstr->mbs[byte_idx] + = toupper (pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]); + /* The next step uses the assumption that wchar_t is encoded + ASCII-safe: all ASCII values can be converted like this. */ + pstr->wcs[byte_idx] = (wchar_t) pstr->mbs[byte_idx]; + ++byte_idx; + continue; + } + + remain_len = end_idx - byte_idx; + prev_st = pstr->cur_state; + mbclen = mbrtowc (&wc, + ((const char *) pstr->raw_mbs + pstr->raw_mbs_idx + + byte_idx), remain_len, &pstr->cur_state); + if (BE (mbclen < (size_t) -2, 1)) + { + wchar_t wcu = wc; + if (iswlower (wc)) + { + size_t mbcdlen; + + wcu = towupper (wc); + mbcdlen = wcrtomb (buf, wcu, &prev_st); + if (BE (mbclen == mbcdlen, 1)) + memcpy (pstr->mbs + byte_idx, buf, mbclen); + else + { + src_idx = byte_idx; + goto offsets_needed; + } + } + else + memcpy (pstr->mbs + byte_idx, + pstr->raw_mbs + pstr->raw_mbs_idx + byte_idx, mbclen); + pstr->wcs[byte_idx++] = wcu; + /* Write paddings. */ + for (remain_len = byte_idx + mbclen - 1; byte_idx < remain_len ;) + pstr->wcs[byte_idx++] = WEOF; + } + else if (mbclen == (size_t) -1 || mbclen == 0) + { + /* It is an invalid character or '\0'. Just use the byte. */ + int ch = pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]; + pstr->mbs[byte_idx] = ch; + /* And also cast it to wide char. */ + pstr->wcs[byte_idx++] = (wchar_t) ch; + if (BE (mbclen == (size_t) -1, 0)) + pstr->cur_state = prev_st; + } + else + { + /* The buffer doesn't have enough space, finish to build. */ + pstr->cur_state = prev_st; + break; + } + } + pstr->valid_len = byte_idx; + pstr->valid_raw_len = byte_idx; + return REG_NOERROR; + } + else + for (src_idx = pstr->valid_raw_len; byte_idx < end_idx;) + { + wchar_t wc; + const char *p; + offsets_needed: + remain_len = end_idx - byte_idx; + prev_st = pstr->cur_state; + if (BE (pstr->trans != NULL, 0)) + { + int i, ch; + + for (i = 0; i < pstr->mb_cur_max && i < remain_len; ++i) + { + ch = pstr->raw_mbs [pstr->raw_mbs_idx + src_idx + i]; + buf[i] = pstr->trans[ch]; + } + p = (const char *) buf; + } + else + p = (const char *) pstr->raw_mbs + pstr->raw_mbs_idx + src_idx; + mbclen = mbrtowc (&wc, p, remain_len, &pstr->cur_state); + if (BE (mbclen < (size_t) -2, 1)) + { + wchar_t wcu = wc; + if (iswlower (wc)) + { + size_t mbcdlen; + + wcu = towupper (wc); + mbcdlen = wcrtomb ((char *) buf, wcu, &prev_st); + if (BE (mbclen == mbcdlen, 1)) + memcpy (pstr->mbs + byte_idx, buf, mbclen); + else if (mbcdlen != (size_t) -1) + { + size_t i; + + if (byte_idx + mbcdlen > pstr->bufs_len) + { + pstr->cur_state = prev_st; + break; + } + + if (pstr->offsets == NULL) + { + pstr->offsets = re_malloc (Idx, pstr->bufs_len); + + if (pstr->offsets == NULL) + return REG_ESPACE; + } + if (!pstr->offsets_needed) + { + for (i = 0; i < (size_t) byte_idx; ++i) + pstr->offsets[i] = i; + pstr->offsets_needed = 1; + } + + memcpy (pstr->mbs + byte_idx, buf, mbcdlen); + pstr->wcs[byte_idx] = wcu; + pstr->offsets[byte_idx] = src_idx; + for (i = 1; i < mbcdlen; ++i) + { + pstr->offsets[byte_idx + i] + = src_idx + (i < mbclen ? i : mbclen - 1); + pstr->wcs[byte_idx + i] = WEOF; + } + pstr->len += mbcdlen - mbclen; + if (pstr->raw_stop > src_idx) + pstr->stop += mbcdlen - mbclen; + end_idx = (pstr->bufs_len > pstr->len) + ? pstr->len : pstr->bufs_len; + byte_idx += mbcdlen; + src_idx += mbclen; + continue; + } + else + memcpy (pstr->mbs + byte_idx, p, mbclen); + } + else + memcpy (pstr->mbs + byte_idx, p, mbclen); + + if (BE (pstr->offsets_needed != 0, 0)) + { + size_t i; + for (i = 0; i < mbclen; ++i) + pstr->offsets[byte_idx + i] = src_idx + i; + } + src_idx += mbclen; + + pstr->wcs[byte_idx++] = wcu; + /* Write paddings. */ + for (remain_len = byte_idx + mbclen - 1; byte_idx < remain_len ;) + pstr->wcs[byte_idx++] = WEOF; + } + else if (mbclen == (size_t) -1 || mbclen == 0) + { + /* It is an invalid character or '\0'. Just use the byte. */ + int ch = pstr->raw_mbs[pstr->raw_mbs_idx + src_idx]; + + if (BE (pstr->trans != NULL, 0)) + ch = pstr->trans [ch]; + pstr->mbs[byte_idx] = ch; + + if (BE (pstr->offsets_needed != 0, 0)) + pstr->offsets[byte_idx] = src_idx; + ++src_idx; + + /* And also cast it to wide char. */ + pstr->wcs[byte_idx++] = (wchar_t) ch; + if (BE (mbclen == (size_t) -1, 0)) + pstr->cur_state = prev_st; + } + else + { + /* The buffer doesn't have enough space, finish to build. */ + pstr->cur_state = prev_st; + break; + } + } + pstr->valid_len = byte_idx; + pstr->valid_raw_len = src_idx; + return REG_NOERROR; +} + +/* Skip characters until the index becomes greater than NEW_RAW_IDX. + Return the index. */ + +static Idx +internal_function +re_string_skip_chars (re_string_t *pstr, Idx new_raw_idx, wint_t *last_wc) +{ + mbstate_t prev_st; + Idx rawbuf_idx; + size_t mbclen; + wint_t wc = WEOF; + + /* Skip the characters which are not necessary to check. */ + for (rawbuf_idx = pstr->raw_mbs_idx + pstr->valid_raw_len; + rawbuf_idx < new_raw_idx;) + { + wchar_t wc2; + Idx remain_len; + remain_len = pstr->len - rawbuf_idx; + prev_st = pstr->cur_state; + mbclen = mbrtowc (&wc2, (const char *) pstr->raw_mbs + rawbuf_idx, + remain_len, &pstr->cur_state); + if (BE (mbclen == (size_t) -2 || mbclen == (size_t) -1 || mbclen == 0, 0)) + { + /* We treat these cases as a single byte character. */ + if (mbclen == 0 || remain_len == 0) + wc = L'\0'; + else + wc = *(unsigned char *) (pstr->raw_mbs + rawbuf_idx); + mbclen = 1; + pstr->cur_state = prev_st; + } + else + wc = wc2; + /* Then proceed the next character. */ + rawbuf_idx += mbclen; + } + *last_wc = wc; + return rawbuf_idx; +} +#endif /* RE_ENABLE_I18N */ + +/* Build the buffer PSTR->MBS, and apply the translation if we need. + This function is used in case of REG_ICASE. */ + +static void +internal_function +build_upper_buffer (re_string_t *pstr) +{ + Idx char_idx, end_idx; + end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len; + + for (char_idx = pstr->valid_len; char_idx < end_idx; ++char_idx) + { + int ch = pstr->raw_mbs[pstr->raw_mbs_idx + char_idx]; + if (BE (pstr->trans != NULL, 0)) + ch = pstr->trans[ch]; + if (islower (ch)) + pstr->mbs[char_idx] = toupper (ch); + else + pstr->mbs[char_idx] = ch; + } + pstr->valid_len = char_idx; + pstr->valid_raw_len = char_idx; +} + +/* Apply TRANS to the buffer in PSTR. */ + +static void +internal_function +re_string_translate_buffer (re_string_t *pstr) +{ + Idx buf_idx, end_idx; + end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len; + + for (buf_idx = pstr->valid_len; buf_idx < end_idx; ++buf_idx) + { + int ch = pstr->raw_mbs[pstr->raw_mbs_idx + buf_idx]; + pstr->mbs[buf_idx] = pstr->trans[ch]; + } + + pstr->valid_len = buf_idx; + pstr->valid_raw_len = buf_idx; +} + +/* This function re-construct the buffers. + Concretely, convert to wide character in case of pstr->mb_cur_max > 1, + convert to upper case in case of REG_ICASE, apply translation. */ + +static reg_errcode_t +internal_function +re_string_reconstruct (re_string_t *pstr, Idx idx, int eflags) +{ + Idx offset; + + if (BE (pstr->raw_mbs_idx <= idx, 0)) + offset = idx - pstr->raw_mbs_idx; + else + { + /* Reset buffer. */ +#ifdef RE_ENABLE_I18N + if (pstr->mb_cur_max > 1) + memset (&pstr->cur_state, '\0', sizeof (mbstate_t)); +#endif /* RE_ENABLE_I18N */ + pstr->len = pstr->raw_len; + pstr->stop = pstr->raw_stop; + pstr->valid_len = 0; + pstr->raw_mbs_idx = 0; + pstr->valid_raw_len = 0; + pstr->offsets_needed = 0; + pstr->tip_context = ((eflags & REG_NOTBOL) ? CONTEXT_BEGBUF + : CONTEXT_NEWLINE | CONTEXT_BEGBUF); + if (!pstr->mbs_allocated) + pstr->mbs = (unsigned char *) pstr->raw_mbs; + offset = idx; + } + + if (BE (offset != 0, 1)) + { + /* Should the already checked characters be kept? */ + if (BE (offset < pstr->valid_raw_len, 1)) + { + /* Yes, move them to the front of the buffer. */ +#ifdef RE_ENABLE_I18N + if (BE (pstr->offsets_needed, 0)) + { + Idx low = 0, high = pstr->valid_len, mid; + do + { + mid = (high + low) / 2; + if (pstr->offsets[mid] > offset) + high = mid; + else if (pstr->offsets[mid] < offset) + low = mid + 1; + else + break; + } + while (low < high); + if (pstr->offsets[mid] < offset) + ++mid; + pstr->tip_context = re_string_context_at (pstr, mid - 1, + eflags); + /* This can be quite complicated, so handle specially + only the common and easy case where the character with + different length representation of lower and upper + case is present at or after offset. */ + if (pstr->valid_len > offset + && mid == offset && pstr->offsets[mid] == offset) + { + memmove (pstr->wcs, pstr->wcs + offset, + (pstr->valid_len - offset) * sizeof (wint_t)); + memmove (pstr->mbs, pstr->mbs + offset, pstr->valid_len - offset); + pstr->valid_len -= offset; + pstr->valid_raw_len -= offset; + for (low = 0; low < pstr->valid_len; low++) + pstr->offsets[low] = pstr->offsets[low + offset] - offset; + } + else + { + /* Otherwise, just find out how long the partial multibyte + character at offset is and fill it with WEOF/255. */ + pstr->len = pstr->raw_len - idx + offset; + pstr->stop = pstr->raw_stop - idx + offset; + pstr->offsets_needed = 0; + while (mid > 0 && pstr->offsets[mid - 1] == offset) + --mid; + while (mid < pstr->valid_len) + if (pstr->wcs[mid] != WEOF) + break; + else + ++mid; + if (mid == pstr->valid_len) + pstr->valid_len = 0; + else + { + pstr->valid_len = pstr->offsets[mid] - offset; + if (pstr->valid_len) + { + for (low = 0; low < pstr->valid_len; ++low) + pstr->wcs[low] = WEOF; + memset (pstr->mbs, 255, pstr->valid_len); + } + } + pstr->valid_raw_len = pstr->valid_len; + } + } + else +#endif + { + pstr->tip_context = re_string_context_at (pstr, offset - 1, + eflags); +#ifdef RE_ENABLE_I18N + if (pstr->mb_cur_max > 1) + memmove (pstr->wcs, pstr->wcs + offset, + (pstr->valid_len - offset) * sizeof (wint_t)); +#endif /* RE_ENABLE_I18N */ + if (BE (pstr->mbs_allocated, 0)) + memmove (pstr->mbs, pstr->mbs + offset, + pstr->valid_len - offset); + pstr->valid_len -= offset; + pstr->valid_raw_len -= offset; +#if DEBUG + assert (pstr->valid_len > 0); +#endif + } + } + else + { + /* No, skip all characters until IDX. */ + Idx prev_valid_len = pstr->valid_len; + +#ifdef RE_ENABLE_I18N + if (BE (pstr->offsets_needed, 0)) + { + pstr->len = pstr->raw_len - idx + offset; + pstr->stop = pstr->raw_stop - idx + offset; + pstr->offsets_needed = 0; + } +#endif + pstr->valid_len = 0; +#ifdef RE_ENABLE_I18N + if (pstr->mb_cur_max > 1) + { + Idx wcs_idx; + wint_t wc = WEOF; + + if (pstr->is_utf8) + { + const unsigned char *raw, *p, *q, *end; + + /* Special case UTF-8. Multi-byte chars start with any + byte other than 0x80 - 0xbf. */ + raw = pstr->raw_mbs + pstr->raw_mbs_idx; + end = raw + (offset - pstr->mb_cur_max); + if (end < pstr->raw_mbs) + end = pstr->raw_mbs; + p = raw + offset - 1; +#ifdef _LIBC + /* We know the wchar_t encoding is UCS4, so for the simple + case, ASCII characters, skip the conversion step. */ + if (isascii (*p) && BE (pstr->trans == NULL, 1)) + { + memset (&pstr->cur_state, '\0', sizeof (mbstate_t)); + /* pstr->valid_len = 0; */ + wc = (wchar_t) *p; + } + else +#endif + for (; p >= end; --p) + if ((*p & 0xc0) != 0x80) + { + mbstate_t cur_state; + wchar_t wc2; + Idx mlen = raw + pstr->len - p; + unsigned char buf[6]; + size_t mbclen; + + q = p; + if (BE (pstr->trans != NULL, 0)) + { + int i = mlen < 6 ? mlen : 6; + while (--i >= 0) + buf[i] = pstr->trans[p[i]]; + q = buf; + } + /* XXX Don't use mbrtowc, we know which conversion + to use (UTF-8 -> UCS4). */ + memset (&cur_state, 0, sizeof (cur_state)); + mbclen = mbrtowc (&wc2, (const char *) p, mlen, + &cur_state); + if (raw + offset - p <= mbclen + && mbclen < (size_t) -2) + { + memset (&pstr->cur_state, '\0', + sizeof (mbstate_t)); + pstr->valid_len = mbclen - (raw + offset - p); + wc = wc2; + } + break; + } + } + + if (wc == WEOF) + pstr->valid_len = re_string_skip_chars (pstr, idx, &wc) - idx; + if (wc == WEOF) + pstr->tip_context + = re_string_context_at (pstr, prev_valid_len - 1, eflags); + else + pstr->tip_context = ((BE (pstr->word_ops_used != 0, 0) + && IS_WIDE_WORD_CHAR (wc)) + ? CONTEXT_WORD + : ((IS_WIDE_NEWLINE (wc) + && pstr->newline_anchor) + ? CONTEXT_NEWLINE : 0)); + if (BE (pstr->valid_len, 0)) + { + for (wcs_idx = 0; wcs_idx < pstr->valid_len; ++wcs_idx) + pstr->wcs[wcs_idx] = WEOF; + if (pstr->mbs_allocated) + memset (pstr->mbs, 255, pstr->valid_len); + } + pstr->valid_raw_len = pstr->valid_len; + } + else +#endif /* RE_ENABLE_I18N */ + { + int c = pstr->raw_mbs[pstr->raw_mbs_idx + offset - 1]; + pstr->valid_raw_len = 0; + if (pstr->trans) + c = pstr->trans[c]; + pstr->tip_context = (bitset_contain (pstr->word_char, c) + ? CONTEXT_WORD + : ((IS_NEWLINE (c) && pstr->newline_anchor) + ? CONTEXT_NEWLINE : 0)); + } + } + if (!BE (pstr->mbs_allocated, 0)) + pstr->mbs += offset; + } + pstr->raw_mbs_idx = idx; + pstr->len -= offset; + pstr->stop -= offset; + + /* Then build the buffers. */ +#ifdef RE_ENABLE_I18N + if (pstr->mb_cur_max > 1) + { + if (pstr->icase) + { + reg_errcode_t ret = build_wcs_upper_buffer (pstr); + if (BE (ret != REG_NOERROR, 0)) + return ret; + } + else + build_wcs_buffer (pstr); + } + else +#endif /* RE_ENABLE_I18N */ + if (BE (pstr->mbs_allocated, 0)) + { + if (pstr->icase) + build_upper_buffer (pstr); + else if (pstr->trans != NULL) + re_string_translate_buffer (pstr); + } + else + pstr->valid_len = pstr->len; + + pstr->cur_idx = 0; + return REG_NOERROR; +} + +static unsigned char +internal_function __attribute ((pure)) +re_string_peek_byte_case (const re_string_t *pstr, Idx idx) +{ + int ch; + Idx off; + + /* Handle the common (easiest) cases first. */ + if (BE (!pstr->mbs_allocated, 1)) + return re_string_peek_byte (pstr, idx); + +#ifdef RE_ENABLE_I18N + if (pstr->mb_cur_max > 1 + && ! re_string_is_single_byte_char (pstr, pstr->cur_idx + idx)) + return re_string_peek_byte (pstr, idx); +#endif + + off = pstr->cur_idx + idx; +#ifdef RE_ENABLE_I18N + if (pstr->offsets_needed) + off = pstr->offsets[off]; +#endif + + ch = pstr->raw_mbs[pstr->raw_mbs_idx + off]; + +#ifdef RE_ENABLE_I18N + /* Ensure that e.g. for tr_TR.UTF-8 BACKSLASH DOTLESS SMALL LETTER I + this function returns CAPITAL LETTER I instead of first byte of + DOTLESS SMALL LETTER I. The latter would confuse the parser, + since peek_byte_case doesn't advance cur_idx in any way. */ + if (pstr->offsets_needed && !isascii (ch)) + return re_string_peek_byte (pstr, idx); +#endif + + return ch; +} + +static unsigned char +internal_function __attribute ((pure)) +re_string_fetch_byte_case (re_string_t *pstr) +{ + if (BE (!pstr->mbs_allocated, 1)) + return re_string_fetch_byte (pstr); + +#ifdef RE_ENABLE_I18N + if (pstr->offsets_needed) + { + Idx off; + int ch; + + /* For tr_TR.UTF-8 [[:islower:]] there is + [[: CAPITAL LETTER I WITH DOT lower:]] in mbs. Skip + in that case the whole multi-byte character and return + the original letter. On the other side, with + [[: DOTLESS SMALL LETTER I return [[:I, as doing + anything else would complicate things too much. */ + + if (!re_string_first_byte (pstr, pstr->cur_idx)) + return re_string_fetch_byte (pstr); + + off = pstr->offsets[pstr->cur_idx]; + ch = pstr->raw_mbs[pstr->raw_mbs_idx + off]; + + if (! isascii (ch)) + return re_string_fetch_byte (pstr); + + re_string_skip_bytes (pstr, + re_string_char_size_at (pstr, pstr->cur_idx)); + return ch; + } +#endif + + return pstr->raw_mbs[pstr->raw_mbs_idx + pstr->cur_idx++]; +} + +static void +internal_function +re_string_destruct (re_string_t *pstr) +{ +#ifdef RE_ENABLE_I18N + re_free (pstr->wcs); + re_free (pstr->offsets); +#endif /* RE_ENABLE_I18N */ + if (pstr->mbs_allocated) + re_free (pstr->mbs); +} + +/* Return the context at IDX in INPUT. */ + +static unsigned int +internal_function +re_string_context_at (const re_string_t *input, Idx idx, int eflags) +{ + int c; + if (BE (! REG_VALID_INDEX (idx), 0)) + /* In this case, we use the value stored in input->tip_context, + since we can't know the character in input->mbs[-1] here. */ + return input->tip_context; + if (BE (idx == input->len, 0)) + return ((eflags & REG_NOTEOL) ? CONTEXT_ENDBUF + : CONTEXT_NEWLINE | CONTEXT_ENDBUF); +#ifdef RE_ENABLE_I18N + if (input->mb_cur_max > 1) + { + wint_t wc; + Idx wc_idx = idx; + while(input->wcs[wc_idx] == WEOF) + { +#ifdef DEBUG + /* It must not happen. */ + assert (REG_VALID_INDEX (wc_idx)); +#endif + --wc_idx; + if (! REG_VALID_INDEX (wc_idx)) + return input->tip_context; + } + wc = input->wcs[wc_idx]; + if (BE (input->word_ops_used != 0, 0) && IS_WIDE_WORD_CHAR (wc)) + return CONTEXT_WORD; + return (IS_WIDE_NEWLINE (wc) && input->newline_anchor + ? CONTEXT_NEWLINE : 0); + } + else +#endif + { + c = re_string_byte_at (input, idx); + if (bitset_contain (input->word_char, c)) + return CONTEXT_WORD; + return IS_NEWLINE (c) && input->newline_anchor ? CONTEXT_NEWLINE : 0; + } +} + +/* Functions for set operation. */ + +static reg_errcode_t +internal_function +re_node_set_alloc (re_node_set *set, Idx size) +{ + set->alloc = size; + set->nelem = 0; + set->elems = re_malloc (Idx, size); + if (BE (set->elems == NULL, 0)) + return REG_ESPACE; + return REG_NOERROR; +} + +static reg_errcode_t +internal_function +re_node_set_init_1 (re_node_set *set, Idx elem) +{ + set->alloc = 1; + set->nelem = 1; + set->elems = re_malloc (Idx, 1); + if (BE (set->elems == NULL, 0)) + { + set->alloc = set->nelem = 0; + return REG_ESPACE; + } + set->elems[0] = elem; + return REG_NOERROR; +} + +static reg_errcode_t +internal_function +re_node_set_init_2 (re_node_set *set, Idx elem1, Idx elem2) +{ + set->alloc = 2; + set->elems = re_malloc (Idx, 2); + if (BE (set->elems == NULL, 0)) + return REG_ESPACE; + if (elem1 == elem2) + { + set->nelem = 1; + set->elems[0] = elem1; + } + else + { + set->nelem = 2; + if (elem1 < elem2) + { + set->elems[0] = elem1; + set->elems[1] = elem2; + } + else + { + set->elems[0] = elem2; + set->elems[1] = elem1; + } + } + return REG_NOERROR; +} + +static reg_errcode_t +internal_function +re_node_set_init_copy (re_node_set *dest, const re_node_set *src) +{ + dest->nelem = src->nelem; + if (src->nelem > 0) + { + dest->alloc = dest->nelem; + dest->elems = re_malloc (Idx, dest->alloc); + if (BE (dest->elems == NULL, 0)) + { + dest->alloc = dest->nelem = 0; + return REG_ESPACE; + } + memcpy (dest->elems, src->elems, src->nelem * sizeof (Idx)); + } + else + re_node_set_init_empty (dest); + return REG_NOERROR; +} + +/* Calculate the intersection of the sets SRC1 and SRC2. And merge it to + DEST. Return value indicate the error code or REG_NOERROR if succeeded. + Note: We assume dest->elems is NULL, when dest->alloc is 0. */ + +static reg_errcode_t +internal_function +re_node_set_add_intersect (re_node_set *dest, const re_node_set *src1, + const re_node_set *src2) +{ + Idx i1, i2, is, id, delta, sbase; + if (src1->nelem == 0 || src2->nelem == 0) + return REG_NOERROR; + + /* We need dest->nelem + 2 * elems_in_intersection; this is a + conservative estimate. */ + if (src1->nelem + src2->nelem + dest->nelem > dest->alloc) + { + Idx new_alloc = src1->nelem + src2->nelem + dest->alloc; + Idx *new_elems = re_realloc (dest->elems, Idx, new_alloc); + if (BE (new_elems == NULL, 0)) + return REG_ESPACE; + dest->elems = new_elems; + dest->alloc = new_alloc; + } + + /* Find the items in the intersection of SRC1 and SRC2, and copy + into the top of DEST those that are not already in DEST itself. */ + sbase = dest->nelem + src1->nelem + src2->nelem; + i1 = src1->nelem - 1; + i2 = src2->nelem - 1; + id = dest->nelem - 1; + for (;;) + { + if (src1->elems[i1] == src2->elems[i2]) + { + /* Try to find the item in DEST. Maybe we could binary search? */ + while (REG_VALID_INDEX (id) && dest->elems[id] > src1->elems[i1]) + --id; + + if (! REG_VALID_INDEX (id) || dest->elems[id] != src1->elems[i1]) + dest->elems[--sbase] = src1->elems[i1]; + + if (! REG_VALID_INDEX (--i1) || ! REG_VALID_INDEX (--i2)) + break; + } + + /* Lower the highest of the two items. */ + else if (src1->elems[i1] < src2->elems[i2]) + { + if (! REG_VALID_INDEX (--i2)) + break; + } + else + { + if (! REG_VALID_INDEX (--i1)) + break; + } + } + + id = dest->nelem - 1; + is = dest->nelem + src1->nelem + src2->nelem - 1; + delta = is - sbase + 1; + + /* Now copy. When DELTA becomes zero, the remaining + DEST elements are already in place; this is more or + less the same loop that is in re_node_set_merge. */ + dest->nelem += delta; + if (delta > 0 && REG_VALID_INDEX (id)) + for (;;) + { + if (dest->elems[is] > dest->elems[id]) + { + /* Copy from the top. */ + dest->elems[id + delta--] = dest->elems[is--]; + if (delta == 0) + break; + } + else + { + /* Slide from the bottom. */ + dest->elems[id + delta] = dest->elems[id]; + if (! REG_VALID_INDEX (--id)) + break; + } + } + + /* Copy remaining SRC elements. */ + memcpy (dest->elems, dest->elems + sbase, delta * sizeof (Idx)); + + return REG_NOERROR; +} + +/* Calculate the union set of the sets SRC1 and SRC2. And store it to + DEST. Return value indicate the error code or REG_NOERROR if succeeded. */ + +static reg_errcode_t +internal_function +re_node_set_init_union (re_node_set *dest, const re_node_set *src1, + const re_node_set *src2) +{ + Idx i1, i2, id; + if (src1 != NULL && src1->nelem > 0 && src2 != NULL && src2->nelem > 0) + { + dest->alloc = src1->nelem + src2->nelem; + dest->elems = re_malloc (Idx, dest->alloc); + if (BE (dest->elems == NULL, 0)) + return REG_ESPACE; + } + else + { + if (src1 != NULL && src1->nelem > 0) + return re_node_set_init_copy (dest, src1); + else if (src2 != NULL && src2->nelem > 0) + return re_node_set_init_copy (dest, src2); + else + re_node_set_init_empty (dest); + return REG_NOERROR; + } + for (i1 = i2 = id = 0 ; i1 < src1->nelem && i2 < src2->nelem ;) + { + if (src1->elems[i1] > src2->elems[i2]) + { + dest->elems[id++] = src2->elems[i2++]; + continue; + } + if (src1->elems[i1] == src2->elems[i2]) + ++i2; + dest->elems[id++] = src1->elems[i1++]; + } + if (i1 < src1->nelem) + { + memcpy (dest->elems + id, src1->elems + i1, + (src1->nelem - i1) * sizeof (Idx)); + id += src1->nelem - i1; + } + else if (i2 < src2->nelem) + { + memcpy (dest->elems + id, src2->elems + i2, + (src2->nelem - i2) * sizeof (Idx)); + id += src2->nelem - i2; + } + dest->nelem = id; + return REG_NOERROR; +} + +/* Calculate the union set of the sets DEST and SRC. And store it to + DEST. Return value indicate the error code or REG_NOERROR if succeeded. */ + +static reg_errcode_t +internal_function +re_node_set_merge (re_node_set *dest, const re_node_set *src) +{ + Idx is, id, sbase, delta; + if (src == NULL || src->nelem == 0) + return REG_NOERROR; + if (dest->alloc < 2 * src->nelem + dest->nelem) + { + Idx new_alloc = 2 * (src->nelem + dest->alloc); + Idx *new_buffer = re_realloc (dest->elems, Idx, new_alloc); + if (BE (new_buffer == NULL, 0)) + return REG_ESPACE; + dest->elems = new_buffer; + dest->alloc = new_alloc; + } + + if (BE (dest->nelem == 0, 0)) + { + dest->nelem = src->nelem; + memcpy (dest->elems, src->elems, src->nelem * sizeof (Idx)); + return REG_NOERROR; + } + + /* Copy into the top of DEST the items of SRC that are not + found in DEST. Maybe we could binary search in DEST? */ + for (sbase = dest->nelem + 2 * src->nelem, + is = src->nelem - 1, id = dest->nelem - 1; + REG_VALID_INDEX (is) && REG_VALID_INDEX (id); ) + { + if (dest->elems[id] == src->elems[is]) + is--, id--; + else if (dest->elems[id] < src->elems[is]) + dest->elems[--sbase] = src->elems[is--]; + else /* if (dest->elems[id] > src->elems[is]) */ + --id; + } + + if (REG_VALID_INDEX (is)) + { + /* If DEST is exhausted, the remaining items of SRC must be unique. */ + sbase -= is + 1; + memcpy (dest->elems + sbase, src->elems, (is + 1) * sizeof (Idx)); + } + + id = dest->nelem - 1; + is = dest->nelem + 2 * src->nelem - 1; + delta = is - sbase + 1; + if (delta == 0) + return REG_NOERROR; + + /* Now copy. When DELTA becomes zero, the remaining + DEST elements are already in place. */ + dest->nelem += delta; + for (;;) + { + if (dest->elems[is] > dest->elems[id]) + { + /* Copy from the top. */ + dest->elems[id + delta--] = dest->elems[is--]; + if (delta == 0) + break; + } + else + { + /* Slide from the bottom. */ + dest->elems[id + delta] = dest->elems[id]; + if (! REG_VALID_INDEX (--id)) + { + /* Copy remaining SRC elements. */ + memcpy (dest->elems, dest->elems + sbase, + delta * sizeof (Idx)); + break; + } + } + } + + return REG_NOERROR; +} + +/* Insert the new element ELEM to the re_node_set* SET. + SET should not already have ELEM. + Return true if successful. */ + +static bool +internal_function +re_node_set_insert (re_node_set *set, Idx elem) +{ + Idx idx; + /* In case the set is empty. */ + if (set->alloc == 0) + return BE (re_node_set_init_1 (set, elem) == REG_NOERROR, 1); + + if (BE (set->nelem, 0) == 0) + { + /* We already guaranteed above that set->alloc != 0. */ + set->elems[0] = elem; + ++set->nelem; + return true; + } + + /* Realloc if we need. */ + if (set->alloc == set->nelem) + { + Idx *new_elems; + set->alloc = set->alloc * 2; + new_elems = re_realloc (set->elems, Idx, set->alloc); + if (BE (new_elems == NULL, 0)) + return false; + set->elems = new_elems; + } + + /* Move the elements which follows the new element. Test the + first element separately to skip a check in the inner loop. */ + if (elem < set->elems[0]) + { + idx = 0; + for (idx = set->nelem; idx > 0; idx--) + set->elems[idx] = set->elems[idx - 1]; + } + else + { + for (idx = set->nelem; set->elems[idx - 1] > elem; idx--) + set->elems[idx] = set->elems[idx - 1]; + } + + /* Insert the new element. */ + set->elems[idx] = elem; + ++set->nelem; + return true; +} + +/* Insert the new element ELEM to the re_node_set* SET. + SET should not already have any element greater than or equal to ELEM. + Return true if successful. */ + +static bool +internal_function +re_node_set_insert_last (re_node_set *set, Idx elem) +{ + /* Realloc if we need. */ + if (set->alloc == set->nelem) + { + Idx *new_elems; + set->alloc = (set->alloc + 1) * 2; + new_elems = re_realloc (set->elems, Idx, set->alloc); + if (BE (new_elems == NULL, 0)) + return false; + set->elems = new_elems; + } + + /* Insert the new element. */ + set->elems[set->nelem++] = elem; + return true; +} + +/* Compare two node sets SET1 and SET2. + Return true if SET1 and SET2 are equivalent. */ + +static bool +internal_function __attribute ((pure)) +re_node_set_compare (const re_node_set *set1, const re_node_set *set2) +{ + Idx i; + if (set1 == NULL || set2 == NULL || set1->nelem != set2->nelem) + return false; + for (i = set1->nelem ; REG_VALID_INDEX (--i) ; ) + if (set1->elems[i] != set2->elems[i]) + return false; + return true; +} + +/* Return (idx + 1) if SET contains the element ELEM, return 0 otherwise. */ + +static Idx +internal_function __attribute ((pure)) +re_node_set_contains (const re_node_set *set, Idx elem) +{ + __re_size_t idx, right, mid; + if (! REG_VALID_NONZERO_INDEX (set->nelem)) + return 0; + + /* Binary search the element. */ + idx = 0; + right = set->nelem - 1; + while (idx < right) + { + mid = (idx + right) / 2; + if (set->elems[mid] < elem) + idx = mid + 1; + else + right = mid; + } + return set->elems[idx] == elem ? idx + 1 : 0; +} + +static void +internal_function +re_node_set_remove_at (re_node_set *set, Idx idx) +{ + if (idx < 0 || idx >= set->nelem) + return; + --set->nelem; + for (; idx < set->nelem; idx++) + set->elems[idx] = set->elems[idx + 1]; +} + + +/* Add the token TOKEN to dfa->nodes, and return the index of the token. + Or return REG_MISSING if an error occurred. */ + +static Idx +internal_function +re_dfa_add_node (re_dfa_t *dfa, re_token_t token) +{ + if (BE (dfa->nodes_len >= dfa->nodes_alloc, 0)) + { + size_t new_nodes_alloc = dfa->nodes_alloc * 2; + Idx *new_nexts, *new_indices; + re_node_set *new_edests, *new_eclosures; + re_token_t *new_nodes; + size_t max_object_size = + MAX (sizeof (re_token_t), + MAX (sizeof (re_node_set), + sizeof (Idx))); + + /* Avoid overflows. */ + if (BE (SIZE_MAX / 2 / max_object_size < dfa->nodes_alloc, 0)) + return REG_MISSING; + + new_nodes = re_realloc (dfa->nodes, re_token_t, new_nodes_alloc); + if (BE (new_nodes == NULL, 0)) + return REG_MISSING; + dfa->nodes = new_nodes; + new_nexts = re_realloc (dfa->nexts, Idx, new_nodes_alloc); + new_indices = re_realloc (dfa->org_indices, Idx, new_nodes_alloc); + new_edests = re_realloc (dfa->edests, re_node_set, new_nodes_alloc); + new_eclosures = re_realloc (dfa->eclosures, re_node_set, new_nodes_alloc); + if (BE (new_nexts == NULL || new_indices == NULL + || new_edests == NULL || new_eclosures == NULL, 0)) + return REG_MISSING; + dfa->nexts = new_nexts; + dfa->org_indices = new_indices; + dfa->edests = new_edests; + dfa->eclosures = new_eclosures; + dfa->nodes_alloc = new_nodes_alloc; + } + dfa->nodes[dfa->nodes_len] = token; + dfa->nodes[dfa->nodes_len].constraint = 0; +#ifdef RE_ENABLE_I18N + { + int type = token.type; + dfa->nodes[dfa->nodes_len].accept_mb = + (type == OP_PERIOD && dfa->mb_cur_max > 1) || type == COMPLEX_BRACKET; + } +#endif + dfa->nexts[dfa->nodes_len] = REG_MISSING; + re_node_set_init_empty (dfa->edests + dfa->nodes_len); + re_node_set_init_empty (dfa->eclosures + dfa->nodes_len); + return dfa->nodes_len++; +} + +static inline re_hashval_t +internal_function +calc_state_hash (const re_node_set *nodes, unsigned int context) +{ + re_hashval_t hash = nodes->nelem + context; + Idx i; + for (i = 0 ; i < nodes->nelem ; i++) + hash += nodes->elems[i]; + return hash; +} + +/* Search for the state whose node_set is equivalent to NODES. + Return the pointer to the state, if we found it in the DFA. + Otherwise create the new one and return it. In case of an error + return NULL and set the error code in ERR. + Note: - We assume NULL as the invalid state, then it is possible that + return value is NULL and ERR is REG_NOERROR. + - We never return non-NULL value in case of any errors, it is for + optimization. */ + +static re_dfastate_t * +internal_function +re_acquire_state (reg_errcode_t *err, const re_dfa_t *dfa, + const re_node_set *nodes) +{ + re_hashval_t hash; + re_dfastate_t *new_state; + struct re_state_table_entry *spot; + Idx i; +#ifdef lint + /* Suppress bogus uninitialized-variable warnings. */ + *err = REG_NOERROR; +#endif + if (BE (nodes->nelem == 0, 0)) + { + *err = REG_NOERROR; + return NULL; + } + hash = calc_state_hash (nodes, 0); + spot = dfa->state_table + (hash & dfa->state_hash_mask); + + for (i = 0 ; i < spot->num ; i++) + { + re_dfastate_t *state = spot->array[i]; + if (hash != state->hash) + continue; + if (re_node_set_compare (&state->nodes, nodes)) + return state; + } + + /* There are no appropriate state in the dfa, create the new one. */ + new_state = create_ci_newstate (dfa, nodes, hash); + if (BE (new_state == NULL, 0)) + *err = REG_ESPACE; + + return new_state; +} + +/* Search for the state whose node_set is equivalent to NODES and + whose context is equivalent to CONTEXT. + Return the pointer to the state, if we found it in the DFA. + Otherwise create the new one and return it. In case of an error + return NULL and set the error code in ERR. + Note: - We assume NULL as the invalid state, then it is possible that + return value is NULL and ERR is REG_NOERROR. + - We never return non-NULL value in case of any errors, it is for + optimization. */ + +static re_dfastate_t * +internal_function +re_acquire_state_context (reg_errcode_t *err, const re_dfa_t *dfa, + const re_node_set *nodes, unsigned int context) +{ + re_hashval_t hash; + re_dfastate_t *new_state; + struct re_state_table_entry *spot; + Idx i; +#ifdef lint + /* Suppress bogus uninitialized-variable warnings. */ + *err = REG_NOERROR; +#endif + if (nodes->nelem == 0) + { + *err = REG_NOERROR; + return NULL; + } + hash = calc_state_hash (nodes, context); + spot = dfa->state_table + (hash & dfa->state_hash_mask); + + for (i = 0 ; i < spot->num ; i++) + { + re_dfastate_t *state = spot->array[i]; + if (state->hash == hash + && state->context == context + && re_node_set_compare (state->entrance_nodes, nodes)) + return state; + } + /* There are no appropriate state in `dfa', create the new one. */ + new_state = create_cd_newstate (dfa, nodes, context, hash); + if (BE (new_state == NULL, 0)) + *err = REG_ESPACE; + + return new_state; +} + +/* Finish initialization of the new state NEWSTATE, and using its hash value + HASH put in the appropriate bucket of DFA's state table. Return value + indicates the error code if failed. */ + +static reg_errcode_t +register_state (const re_dfa_t *dfa, re_dfastate_t *newstate, + re_hashval_t hash) +{ + struct re_state_table_entry *spot; + reg_errcode_t err; + Idx i; + + newstate->hash = hash; + err = re_node_set_alloc (&newstate->non_eps_nodes, newstate->nodes.nelem); + if (BE (err != REG_NOERROR, 0)) + return REG_ESPACE; + for (i = 0; i < newstate->nodes.nelem; i++) + { + Idx elem = newstate->nodes.elems[i]; + if (!IS_EPSILON_NODE (dfa->nodes[elem].type)) + if (BE (! re_node_set_insert_last (&newstate->non_eps_nodes, elem), 0)) + return REG_ESPACE; + } + + spot = dfa->state_table + (hash & dfa->state_hash_mask); + if (BE (spot->alloc <= spot->num, 0)) + { + Idx new_alloc = 2 * spot->num + 2; + re_dfastate_t **new_array = re_realloc (spot->array, re_dfastate_t *, + new_alloc); + if (BE (new_array == NULL, 0)) + return REG_ESPACE; + spot->array = new_array; + spot->alloc = new_alloc; + } + spot->array[spot->num++] = newstate; + return REG_NOERROR; +} + +static void +free_state (re_dfastate_t *state) +{ + re_node_set_free (&state->non_eps_nodes); + re_node_set_free (&state->inveclosure); + if (state->entrance_nodes != &state->nodes) + { + re_node_set_free (state->entrance_nodes); + re_free (state->entrance_nodes); + } + re_node_set_free (&state->nodes); + re_free (state->word_trtable); + re_free (state->trtable); + re_free (state); +} + +/* Create the new state which is independ of contexts. + Return the new state if succeeded, otherwise return NULL. */ + +static re_dfastate_t * +internal_function +create_ci_newstate (const re_dfa_t *dfa, const re_node_set *nodes, + re_hashval_t hash) +{ + Idx i; + reg_errcode_t err; + re_dfastate_t *newstate; + + newstate = (re_dfastate_t *) calloc (sizeof (re_dfastate_t), 1); + if (BE (newstate == NULL, 0)) + return NULL; + err = re_node_set_init_copy (&newstate->nodes, nodes); + if (BE (err != REG_NOERROR, 0)) + { + re_free (newstate); + return NULL; + } + + newstate->entrance_nodes = &newstate->nodes; + for (i = 0 ; i < nodes->nelem ; i++) + { + re_token_t *node = dfa->nodes + nodes->elems[i]; + re_token_type_t type = node->type; + if (type == CHARACTER && !node->constraint) + continue; +#ifdef RE_ENABLE_I18N + newstate->accept_mb |= node->accept_mb; +#endif /* RE_ENABLE_I18N */ + + /* If the state has the halt node, the state is a halt state. */ + if (type == END_OF_RE) + newstate->halt = 1; + else if (type == OP_BACK_REF) + newstate->has_backref = 1; + else if (type == ANCHOR || node->constraint) + newstate->has_constraint = 1; + } + err = register_state (dfa, newstate, hash); + if (BE (err != REG_NOERROR, 0)) + { + free_state (newstate); + newstate = NULL; + } + return newstate; +} + +/* Create the new state which is depend on the context CONTEXT. + Return the new state if succeeded, otherwise return NULL. */ + +static re_dfastate_t * +internal_function +create_cd_newstate (const re_dfa_t *dfa, const re_node_set *nodes, + unsigned int context, re_hashval_t hash) +{ + Idx i, nctx_nodes = 0; + reg_errcode_t err; + re_dfastate_t *newstate; + + newstate = (re_dfastate_t *) calloc (sizeof (re_dfastate_t), 1); + if (BE (newstate == NULL, 0)) + return NULL; + err = re_node_set_init_copy (&newstate->nodes, nodes); + if (BE (err != REG_NOERROR, 0)) + { + re_free (newstate); + return NULL; + } + + newstate->context = context; + newstate->entrance_nodes = &newstate->nodes; + + for (i = 0 ; i < nodes->nelem ; i++) + { + unsigned int constraint = 0; + re_token_t *node = dfa->nodes + nodes->elems[i]; + re_token_type_t type = node->type; + if (node->constraint) + constraint = node->constraint; + + if (type == CHARACTER && !constraint) + continue; +#ifdef RE_ENABLE_I18N + newstate->accept_mb |= node->accept_mb; +#endif /* RE_ENABLE_I18N */ + + /* If the state has the halt node, the state is a halt state. */ + if (type == END_OF_RE) + newstate->halt = 1; + else if (type == OP_BACK_REF) + newstate->has_backref = 1; + else if (type == ANCHOR) + constraint = node->opr.ctx_type; + + if (constraint) + { + if (newstate->entrance_nodes == &newstate->nodes) + { + newstate->entrance_nodes = re_malloc (re_node_set, 1); + if (BE (newstate->entrance_nodes == NULL, 0)) + { + free_state (newstate); + return NULL; + } + re_node_set_init_copy (newstate->entrance_nodes, nodes); + nctx_nodes = 0; + newstate->has_constraint = 1; + } + + if (NOT_SATISFY_PREV_CONSTRAINT (constraint,context)) + { + re_node_set_remove_at (&newstate->nodes, i - nctx_nodes); + ++nctx_nodes; + } + } + } + err = register_state (dfa, newstate, hash); + if (BE (err != REG_NOERROR, 0)) + { + free_state (newstate); + newstate = NULL; + } + return newstate; +} diff --git a/gl/regex_internal.h b/gl/regex_internal.h new file mode 100644 index 00000000..b0f7e657 --- /dev/null +++ b/gl/regex_internal.h @@ -0,0 +1,862 @@ +/* Extended regular expression matching and search library. + Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _REGEX_INTERNAL_H +#define _REGEX_INTERNAL_H 1 + +#include <assert.h> +#include <ctype.h> +#include <stdbool.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +#ifndef _LIBC +# include "strcase.h" +#endif + +#if defined HAVE_LANGINFO_H || defined HAVE_LANGINFO_CODESET || defined _LIBC +# include <langinfo.h> +#endif +#if defined HAVE_LOCALE_H || defined _LIBC +# include <locale.h> +#endif + +#include <wchar.h> +#include <wctype.h> +#include <stdint.h> +#if defined _LIBC +# include <bits/libc-lock.h> +#else +# define __libc_lock_init(NAME) do { } while (0) +# define __libc_lock_lock(NAME) do { } while (0) +# define __libc_lock_unlock(NAME) do { } while (0) +#endif + +/* In case that the system doesn't have isblank(). */ +#if !defined _LIBC && !HAVE_DECL_ISBLANK && !defined isblank +# define isblank(ch) ((ch) == ' ' || (ch) == '\t') +#endif + +#ifdef _LIBC +# ifndef _RE_DEFINE_LOCALE_FUNCTIONS +# define _RE_DEFINE_LOCALE_FUNCTIONS 1 +# include <locale/localeinfo.h> +# include <locale/elem-hash.h> +# include <locale/coll-lookup.h> +# endif +#endif + +/* This is for other GNU distributions with internationalized messages. */ +#if (HAVE_LIBINTL_H && ENABLE_NLS) || defined _LIBC +# include <libintl.h> +# ifdef _LIBC +# undef gettext +# define gettext(msgid) \ + INTUSE(__dcgettext) (_libc_intl_domainname, msgid, LC_MESSAGES) +# endif +#else +# define gettext(msgid) (msgid) +#endif + +#ifndef gettext_noop +/* This define is so xgettext can find the internationalizable + strings. */ +# define gettext_noop(String) String +#endif + +/* For loser systems without the definition. */ +#ifndef SIZE_MAX +# define SIZE_MAX ((size_t) -1) +#endif + +#if (defined MB_CUR_MAX && HAVE_LOCALE_H && HAVE_WCTYPE_H && HAVE_ISWCTYPE && HAVE_WCRTOMB && HAVE_MBRTOWC && HAVE_WCSCOLL) || _LIBC +# define RE_ENABLE_I18N +#endif + +#if __GNUC__ >= 3 +# define BE(expr, val) __builtin_expect (expr, val) +#else +# define BE(expr, val) (expr) +# ifdef _LIBC +# define inline +# endif +#endif + +/* Number of ASCII characters. */ +#define ASCII_CHARS 0x80 + +/* Number of single byte characters. */ +#define SBC_MAX (UCHAR_MAX + 1) + +#define COLL_ELEM_LEN_MAX 8 + +/* The character which represents newline. */ +#define NEWLINE_CHAR '\n' +#define WIDE_NEWLINE_CHAR L'\n' + +/* Rename to standard API for using out of glibc. */ +#ifndef _LIBC +# define __wctype wctype +# define __iswctype iswctype +# define __btowc btowc +# ifndef __mempcpy +# define __mempcpy mempcpy +# endif +# define __wcrtomb wcrtomb +# define __regfree regfree +# define attribute_hidden +#endif /* not _LIBC */ + +#if __GNUC__ >= 4 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 1) +# define __attribute(arg) __attribute__ (arg) +#else +# define __attribute(arg) +#endif + +typedef __re_idx_t Idx; + +/* Special return value for failure to match. */ +#define REG_MISSING ((Idx) -1) + +/* Special return value for internal error. */ +#define REG_ERROR ((Idx) -2) + +/* Test whether N is a valid index, and is not one of the above. */ +#ifdef _REGEX_LARGE_OFFSETS +# define REG_VALID_INDEX(n) ((Idx) (n) < REG_ERROR) +#else +# define REG_VALID_INDEX(n) (0 <= (n)) +#endif + +/* Test whether N is a valid nonzero index. */ +#ifdef _REGEX_LARGE_OFFSETS +# define REG_VALID_NONZERO_INDEX(n) ((Idx) ((n) - 1) < (Idx) (REG_ERROR - 1)) +#else +# define REG_VALID_NONZERO_INDEX(n) (0 < (n)) +#endif + +/* A hash value, suitable for computing hash tables. */ +typedef __re_size_t re_hashval_t; + +/* An integer used to represent a set of bits. It must be unsigned, + and must be at least as wide as unsigned int. */ +typedef unsigned long int bitset_word_t; +/* All bits set in a bitset_word_t. */ +#define BITSET_WORD_MAX ULONG_MAX + +/* Number of bits in a bitset_word_t. For portability to hosts with + padding bits, do not use '(sizeof (bitset_word_t) * CHAR_BIT)'; + instead, deduce it directly from BITSET_WORD_MAX. Avoid + greater-than-32-bit integers and unconditional shifts by more than + 31 bits, as they're not portable. */ +#if BITSET_WORD_MAX == 0xffffffff +# define BITSET_WORD_BITS 32 +#elif BITSET_WORD_MAX >> 31 >> 5 == 1 +# define BITSET_WORD_BITS 36 +#elif BITSET_WORD_MAX >> 31 >> 16 == 1 +# define BITSET_WORD_BITS 48 +#elif BITSET_WORD_MAX >> 31 >> 28 == 1 +# define BITSET_WORD_BITS 60 +#elif BITSET_WORD_MAX >> 31 >> 31 >> 1 == 1 +# define BITSET_WORD_BITS 64 +#elif BITSET_WORD_MAX >> 31 >> 31 >> 9 == 1 +# define BITSET_WORD_BITS 72 +#elif BITSET_WORD_MAX >> 31 >> 31 >> 31 >> 31 >> 3 == 1 +# define BITSET_WORD_BITS 128 +#elif BITSET_WORD_MAX >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 7 == 1 +# define BITSET_WORD_BITS 256 +#elif BITSET_WORD_MAX >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 7 > 1 +# define BITSET_WORD_BITS 257 /* any value > SBC_MAX will do here */ +# if BITSET_WORD_BITS <= SBC_MAX +# error "Invalid SBC_MAX" +# endif +#elif BITSET_WORD_MAX == (0xffffffff + 2) * 0xffffffff +/* Work around a bug in 64-bit PGC (before version 6.1-2), where the + preprocessor mishandles large unsigned values as if they were signed. */ +# define BITSET_WORD_BITS 64 +#else +# error "Add case for new bitset_word_t size" +#endif + +/* Number of bitset_word_t values in a bitset_t. */ +#define BITSET_WORDS ((SBC_MAX + BITSET_WORD_BITS - 1) / BITSET_WORD_BITS) + +typedef bitset_word_t bitset_t[BITSET_WORDS]; +typedef bitset_word_t *re_bitset_ptr_t; +typedef const bitset_word_t *re_const_bitset_ptr_t; + +#define PREV_WORD_CONSTRAINT 0x0001 +#define PREV_NOTWORD_CONSTRAINT 0x0002 +#define NEXT_WORD_CONSTRAINT 0x0004 +#define NEXT_NOTWORD_CONSTRAINT 0x0008 +#define PREV_NEWLINE_CONSTRAINT 0x0010 +#define NEXT_NEWLINE_CONSTRAINT 0x0020 +#define PREV_BEGBUF_CONSTRAINT 0x0040 +#define NEXT_ENDBUF_CONSTRAINT 0x0080 +#define WORD_DELIM_CONSTRAINT 0x0100 +#define NOT_WORD_DELIM_CONSTRAINT 0x0200 + +typedef enum +{ + INSIDE_WORD = PREV_WORD_CONSTRAINT | NEXT_WORD_CONSTRAINT, + WORD_FIRST = PREV_NOTWORD_CONSTRAINT | NEXT_WORD_CONSTRAINT, + WORD_LAST = PREV_WORD_CONSTRAINT | NEXT_NOTWORD_CONSTRAINT, + INSIDE_NOTWORD = PREV_NOTWORD_CONSTRAINT | NEXT_NOTWORD_CONSTRAINT, + LINE_FIRST = PREV_NEWLINE_CONSTRAINT, + LINE_LAST = NEXT_NEWLINE_CONSTRAINT, + BUF_FIRST = PREV_BEGBUF_CONSTRAINT, + BUF_LAST = NEXT_ENDBUF_CONSTRAINT, + WORD_DELIM = WORD_DELIM_CONSTRAINT, + NOT_WORD_DELIM = NOT_WORD_DELIM_CONSTRAINT +} re_context_type; + +typedef struct +{ + Idx alloc; + Idx nelem; + Idx *elems; +} re_node_set; + +typedef enum +{ + NON_TYPE = 0, + + /* Node type, These are used by token, node, tree. */ + CHARACTER = 1, + END_OF_RE = 2, + SIMPLE_BRACKET = 3, + OP_BACK_REF = 4, + OP_PERIOD = 5, +#ifdef RE_ENABLE_I18N + COMPLEX_BRACKET = 6, + OP_UTF8_PERIOD = 7, +#endif /* RE_ENABLE_I18N */ + + /* We define EPSILON_BIT as a macro so that OP_OPEN_SUBEXP is used + when the debugger shows values of this enum type. */ +#define EPSILON_BIT 8 + OP_OPEN_SUBEXP = EPSILON_BIT | 0, + OP_CLOSE_SUBEXP = EPSILON_BIT | 1, + OP_ALT = EPSILON_BIT | 2, + OP_DUP_ASTERISK = EPSILON_BIT | 3, + ANCHOR = EPSILON_BIT | 4, + + /* Tree type, these are used only by tree. */ + CONCAT = 16, + SUBEXP = 17, + + /* Token type, these are used only by token. */ + OP_DUP_PLUS = 18, + OP_DUP_QUESTION, + OP_OPEN_BRACKET, + OP_CLOSE_BRACKET, + OP_CHARSET_RANGE, + OP_OPEN_DUP_NUM, + OP_CLOSE_DUP_NUM, + OP_NON_MATCH_LIST, + OP_OPEN_COLL_ELEM, + OP_CLOSE_COLL_ELEM, + OP_OPEN_EQUIV_CLASS, + OP_CLOSE_EQUIV_CLASS, + OP_OPEN_CHAR_CLASS, + OP_CLOSE_CHAR_CLASS, + OP_WORD, + OP_NOTWORD, + OP_SPACE, + OP_NOTSPACE, + BACK_SLASH + +} re_token_type_t; + +#ifdef RE_ENABLE_I18N +typedef struct +{ + /* Multibyte characters. */ + wchar_t *mbchars; + + /* Collating symbols. */ +# ifdef _LIBC + int32_t *coll_syms; +# endif + + /* Equivalence classes. */ +# ifdef _LIBC + int32_t *equiv_classes; +# endif + + /* Range expressions. */ +# ifdef _LIBC + uint32_t *range_starts; + uint32_t *range_ends; +# else /* not _LIBC */ + wchar_t *range_starts; + wchar_t *range_ends; +# endif /* not _LIBC */ + + /* Character classes. */ + wctype_t *char_classes; + + /* If this character set is the non-matching list. */ + unsigned int non_match : 1; + + /* # of multibyte characters. */ + Idx nmbchars; + + /* # of collating symbols. */ + Idx ncoll_syms; + + /* # of equivalence classes. */ + Idx nequiv_classes; + + /* # of range expressions. */ + Idx nranges; + + /* # of character classes. */ + Idx nchar_classes; +} re_charset_t; +#endif /* RE_ENABLE_I18N */ + +typedef struct +{ + union + { + unsigned char c; /* for CHARACTER */ + re_bitset_ptr_t sbcset; /* for SIMPLE_BRACKET */ +#ifdef RE_ENABLE_I18N + re_charset_t *mbcset; /* for COMPLEX_BRACKET */ +#endif /* RE_ENABLE_I18N */ + Idx idx; /* for BACK_REF */ + re_context_type ctx_type; /* for ANCHOR */ + } opr; +#if __GNUC__ >= 2 && !__STRICT_ANSI__ + re_token_type_t type : 8; +#else + re_token_type_t type; +#endif + unsigned int constraint : 10; /* context constraint */ + unsigned int duplicated : 1; + unsigned int opt_subexp : 1; +#ifdef RE_ENABLE_I18N + unsigned int accept_mb : 1; + /* These 2 bits can be moved into the union if needed (e.g. if running out + of bits; move opr.c to opr.c.c and move the flags to opr.c.flags). */ + unsigned int mb_partial : 1; +#endif + unsigned int word_char : 1; +} re_token_t; + +#define IS_EPSILON_NODE(type) ((type) & EPSILON_BIT) + +struct re_string_t +{ + /* Indicate the raw buffer which is the original string passed as an + argument of regexec(), re_search(), etc.. */ + const unsigned char *raw_mbs; + /* Store the multibyte string. In case of "case insensitive mode" like + REG_ICASE, upper cases of the string are stored, otherwise MBS points + the same address that RAW_MBS points. */ + unsigned char *mbs; +#ifdef RE_ENABLE_I18N + /* Store the wide character string which is corresponding to MBS. */ + wint_t *wcs; + Idx *offsets; + mbstate_t cur_state; +#endif + /* Index in RAW_MBS. Each character mbs[i] corresponds to + raw_mbs[raw_mbs_idx + i]. */ + Idx raw_mbs_idx; + /* The length of the valid characters in the buffers. */ + Idx valid_len; + /* The corresponding number of bytes in raw_mbs array. */ + Idx valid_raw_len; + /* The length of the buffers MBS and WCS. */ + Idx bufs_len; + /* The index in MBS, which is updated by re_string_fetch_byte. */ + Idx cur_idx; + /* length of RAW_MBS array. */ + Idx raw_len; + /* This is RAW_LEN - RAW_MBS_IDX + VALID_LEN - VALID_RAW_LEN. */ + Idx len; + /* End of the buffer may be shorter than its length in the cases such + as re_match_2, re_search_2. Then, we use STOP for end of the buffer + instead of LEN. */ + Idx raw_stop; + /* This is RAW_STOP - RAW_MBS_IDX adjusted through OFFSETS. */ + Idx stop; + + /* The context of mbs[0]. We store the context independently, since + the context of mbs[0] may be different from raw_mbs[0], which is + the beginning of the input string. */ + unsigned int tip_context; + /* The translation passed as a part of an argument of re_compile_pattern. */ + RE_TRANSLATE_TYPE trans; + /* Copy of re_dfa_t's word_char. */ + re_const_bitset_ptr_t word_char; + /* true if REG_ICASE. */ + unsigned char icase; + unsigned char is_utf8; + unsigned char map_notascii; + unsigned char mbs_allocated; + unsigned char offsets_needed; + unsigned char newline_anchor; + unsigned char word_ops_used; + int mb_cur_max; +}; +typedef struct re_string_t re_string_t; + + +struct re_dfa_t; +typedef struct re_dfa_t re_dfa_t; + +#ifndef _LIBC +# ifdef __i386__ +# define internal_function __attribute ((regparm (3), stdcall)) +# else +# define internal_function +# endif +#endif + +static reg_errcode_t re_string_realloc_buffers (re_string_t *pstr, + Idx new_buf_len) + internal_function; +#ifdef RE_ENABLE_I18N +static void build_wcs_buffer (re_string_t *pstr) internal_function; +static reg_errcode_t build_wcs_upper_buffer (re_string_t *pstr) + internal_function; +#endif /* RE_ENABLE_I18N */ +static void build_upper_buffer (re_string_t *pstr) internal_function; +static void re_string_translate_buffer (re_string_t *pstr) internal_function; +static unsigned int re_string_context_at (const re_string_t *input, Idx idx, + int eflags) + internal_function __attribute ((pure)); +#define re_string_peek_byte(pstr, offset) \ + ((pstr)->mbs[(pstr)->cur_idx + offset]) +#define re_string_fetch_byte(pstr) \ + ((pstr)->mbs[(pstr)->cur_idx++]) +#define re_string_first_byte(pstr, idx) \ + ((idx) == (pstr)->valid_len || (pstr)->wcs[idx] != WEOF) +#define re_string_is_single_byte_char(pstr, idx) \ + ((pstr)->wcs[idx] != WEOF && ((pstr)->valid_len == (idx) + 1 \ + || (pstr)->wcs[(idx) + 1] != WEOF)) +#define re_string_eoi(pstr) ((pstr)->stop <= (pstr)->cur_idx) +#define re_string_cur_idx(pstr) ((pstr)->cur_idx) +#define re_string_get_buffer(pstr) ((pstr)->mbs) +#define re_string_length(pstr) ((pstr)->len) +#define re_string_byte_at(pstr,idx) ((pstr)->mbs[idx]) +#define re_string_skip_bytes(pstr,idx) ((pstr)->cur_idx += (idx)) +#define re_string_set_index(pstr,idx) ((pstr)->cur_idx = (idx)) + +#include <alloca.h> + +#ifndef _LIBC +# if HAVE_ALLOCA +/* The OS usually guarantees only one guard page at the bottom of the stack, + and a page size can be as small as 4096 bytes. So we cannot safely + allocate anything larger than 4096 bytes. Also care for the possibility + of a few compiler-allocated temporary stack slots. */ +# define __libc_use_alloca(n) ((n) < 4032) +# else +/* alloca is implemented with malloc, so just use malloc. */ +# define __libc_use_alloca(n) 0 +# endif +#endif + +#ifndef MAX +# define MAX(a,b) ((a) < (b) ? (b) : (a)) +#endif + +#define re_malloc(t,n) ((t *) malloc ((n) * sizeof (t))) +#define re_realloc(p,t,n) ((t *) realloc (p, (n) * sizeof (t))) +#define re_free(p) free (p) + +struct bin_tree_t +{ + struct bin_tree_t *parent; + struct bin_tree_t *left; + struct bin_tree_t *right; + struct bin_tree_t *first; + struct bin_tree_t *next; + + re_token_t token; + + /* `node_idx' is the index in dfa->nodes, if `type' == 0. + Otherwise `type' indicate the type of this node. */ + Idx node_idx; +}; +typedef struct bin_tree_t bin_tree_t; + +#define BIN_TREE_STORAGE_SIZE \ + ((1024 - sizeof (void *)) / sizeof (bin_tree_t)) + +struct bin_tree_storage_t +{ + struct bin_tree_storage_t *next; + bin_tree_t data[BIN_TREE_STORAGE_SIZE]; +}; +typedef struct bin_tree_storage_t bin_tree_storage_t; + +#define CONTEXT_WORD 1 +#define CONTEXT_NEWLINE (CONTEXT_WORD << 1) +#define CONTEXT_BEGBUF (CONTEXT_NEWLINE << 1) +#define CONTEXT_ENDBUF (CONTEXT_BEGBUF << 1) + +#define IS_WORD_CONTEXT(c) ((c) & CONTEXT_WORD) +#define IS_NEWLINE_CONTEXT(c) ((c) & CONTEXT_NEWLINE) +#define IS_BEGBUF_CONTEXT(c) ((c) & CONTEXT_BEGBUF) +#define IS_ENDBUF_CONTEXT(c) ((c) & CONTEXT_ENDBUF) +#define IS_ORDINARY_CONTEXT(c) ((c) == 0) + +#define IS_WORD_CHAR(ch) (isalnum (ch) || (ch) == '_') +#define IS_NEWLINE(ch) ((ch) == NEWLINE_CHAR) +#define IS_WIDE_WORD_CHAR(ch) (iswalnum (ch) || (ch) == L'_') +#define IS_WIDE_NEWLINE(ch) ((ch) == WIDE_NEWLINE_CHAR) + +#define NOT_SATISFY_PREV_CONSTRAINT(constraint,context) \ + ((((constraint) & PREV_WORD_CONSTRAINT) && !IS_WORD_CONTEXT (context)) \ + || ((constraint & PREV_NOTWORD_CONSTRAINT) && IS_WORD_CONTEXT (context)) \ + || ((constraint & PREV_NEWLINE_CONSTRAINT) && !IS_NEWLINE_CONTEXT (context))\ + || ((constraint & PREV_BEGBUF_CONSTRAINT) && !IS_BEGBUF_CONTEXT (context))) + +#define NOT_SATISFY_NEXT_CONSTRAINT(constraint,context) \ + ((((constraint) & NEXT_WORD_CONSTRAINT) && !IS_WORD_CONTEXT (context)) \ + || (((constraint) & NEXT_NOTWORD_CONSTRAINT) && IS_WORD_CONTEXT (context)) \ + || (((constraint) & NEXT_NEWLINE_CONSTRAINT) && !IS_NEWLINE_CONTEXT (context)) \ + || (((constraint) & NEXT_ENDBUF_CONSTRAINT) && !IS_ENDBUF_CONTEXT (context))) + +struct re_dfastate_t +{ + re_hashval_t hash; + re_node_set nodes; + re_node_set non_eps_nodes; + re_node_set inveclosure; + re_node_set *entrance_nodes; + struct re_dfastate_t **trtable, **word_trtable; + unsigned int context : 4; + unsigned int halt : 1; + /* If this state can accept `multi byte'. + Note that we refer to multibyte characters, and multi character + collating elements as `multi byte'. */ + unsigned int accept_mb : 1; + /* If this state has backreference node(s). */ + unsigned int has_backref : 1; + unsigned int has_constraint : 1; +}; +typedef struct re_dfastate_t re_dfastate_t; + +struct re_state_table_entry +{ + Idx num; + Idx alloc; + re_dfastate_t **array; +}; + +/* Array type used in re_sub_match_last_t and re_sub_match_top_t. */ + +typedef struct +{ + Idx next_idx; + Idx alloc; + re_dfastate_t **array; +} state_array_t; + +/* Store information about the node NODE whose type is OP_CLOSE_SUBEXP. */ + +typedef struct +{ + Idx node; + Idx str_idx; /* The position NODE match at. */ + state_array_t path; +} re_sub_match_last_t; + +/* Store information about the node NODE whose type is OP_OPEN_SUBEXP. + And information about the node, whose type is OP_CLOSE_SUBEXP, + corresponding to NODE is stored in LASTS. */ + +typedef struct +{ + Idx str_idx; + Idx node; + state_array_t *path; + Idx alasts; /* Allocation size of LASTS. */ + Idx nlasts; /* The number of LASTS. */ + re_sub_match_last_t **lasts; +} re_sub_match_top_t; + +struct re_backref_cache_entry +{ + Idx node; + Idx str_idx; + Idx subexp_from; + Idx subexp_to; + char more; + char unused; + unsigned short int eps_reachable_subexps_map; +}; + +typedef struct +{ + /* The string object corresponding to the input string. */ + re_string_t input; +#if defined _LIBC || (defined __STDC_VERSION__ && __STDC_VERSION__ >= 199901L) + const re_dfa_t *const dfa; +#else + const re_dfa_t *dfa; +#endif + /* EFLAGS of the argument of regexec. */ + int eflags; + /* Where the matching ends. */ + Idx match_last; + Idx last_node; + /* The state log used by the matcher. */ + re_dfastate_t **state_log; + Idx state_log_top; + /* Back reference cache. */ + Idx nbkref_ents; + Idx abkref_ents; + struct re_backref_cache_entry *bkref_ents; + int max_mb_elem_len; + Idx nsub_tops; + Idx asub_tops; + re_sub_match_top_t **sub_tops; +} re_match_context_t; + +typedef struct +{ + re_dfastate_t **sifted_states; + re_dfastate_t **limited_states; + Idx last_node; + Idx last_str_idx; + re_node_set limits; +} re_sift_context_t; + +struct re_fail_stack_ent_t +{ + Idx idx; + Idx node; + regmatch_t *regs; + re_node_set eps_via_nodes; +}; + +struct re_fail_stack_t +{ + Idx num; + Idx alloc; + struct re_fail_stack_ent_t *stack; +}; + +struct re_dfa_t +{ + re_token_t *nodes; + size_t nodes_alloc; + size_t nodes_len; + Idx *nexts; + Idx *org_indices; + re_node_set *edests; + re_node_set *eclosures; + re_node_set *inveclosures; + struct re_state_table_entry *state_table; + re_dfastate_t *init_state; + re_dfastate_t *init_state_word; + re_dfastate_t *init_state_nl; + re_dfastate_t *init_state_begbuf; + bin_tree_t *str_tree; + bin_tree_storage_t *str_tree_storage; + re_bitset_ptr_t sb_char; + int str_tree_storage_idx; + + /* number of subexpressions `re_nsub' is in regex_t. */ + re_hashval_t state_hash_mask; + Idx init_node; + Idx nbackref; /* The number of backreference in this dfa. */ + + /* Bitmap expressing which backreference is used. */ + bitset_word_t used_bkref_map; + bitset_word_t completed_bkref_map; + + unsigned int has_plural_match : 1; + /* If this dfa has "multibyte node", which is a backreference or + a node which can accept multibyte character or multi character + collating element. */ + unsigned int has_mb_node : 1; + unsigned int is_utf8 : 1; + unsigned int map_notascii : 1; + unsigned int word_ops_used : 1; + int mb_cur_max; + bitset_t word_char; + reg_syntax_t syntax; + Idx *subexp_map; +#ifdef DEBUG + char* re_str; +#endif +#ifdef _LIBC + __libc_lock_define (, lock) +#endif +}; + +#define re_node_set_init_empty(set) memset (set, '\0', sizeof (re_node_set)) +#define re_node_set_remove(set,id) \ + (re_node_set_remove_at (set, re_node_set_contains (set, id) - 1)) +#define re_node_set_empty(p) ((p)->nelem = 0) +#define re_node_set_free(set) re_free ((set)->elems) + + +typedef enum +{ + SB_CHAR, + MB_CHAR, + EQUIV_CLASS, + COLL_SYM, + CHAR_CLASS +} bracket_elem_type; + +typedef struct +{ + bracket_elem_type type; + union + { + unsigned char ch; + unsigned char *name; + wchar_t wch; + } opr; +} bracket_elem_t; + + +/* Inline functions for bitset_t operation. */ + +static inline void +bitset_set (bitset_t set, Idx i) +{ + set[i / BITSET_WORD_BITS] |= (bitset_word_t) 1 << i % BITSET_WORD_BITS; +} + +static inline void +bitset_clear (bitset_t set, Idx i) +{ + set[i / BITSET_WORD_BITS] &= ~ ((bitset_word_t) 1 << i % BITSET_WORD_BITS); +} + +static inline bool +bitset_contain (const bitset_t set, Idx i) +{ + return (set[i / BITSET_WORD_BITS] >> i % BITSET_WORD_BITS) & 1; +} + +static inline void +bitset_empty (bitset_t set) +{ + memset (set, '\0', sizeof (bitset_t)); +} + +static inline void +bitset_set_all (bitset_t set) +{ + memset (set, -1, sizeof (bitset_word_t) * (SBC_MAX / BITSET_WORD_BITS)); + if (SBC_MAX % BITSET_WORD_BITS != 0) + set[BITSET_WORDS - 1] = + ((bitset_word_t) 1 << SBC_MAX % BITSET_WORD_BITS) - 1; +} + +static inline void +bitset_copy (bitset_t dest, const bitset_t src) +{ + memcpy (dest, src, sizeof (bitset_t)); +} + +static inline void +bitset_not (bitset_t set) +{ + int bitset_i; + for (bitset_i = 0; bitset_i < SBC_MAX / BITSET_WORD_BITS; ++bitset_i) + set[bitset_i] = ~set[bitset_i]; + if (SBC_MAX % BITSET_WORD_BITS != 0) + set[BITSET_WORDS - 1] = + ((((bitset_word_t) 1 << SBC_MAX % BITSET_WORD_BITS) - 1) + & ~set[BITSET_WORDS - 1]); +} + +static inline void +bitset_merge (bitset_t dest, const bitset_t src) +{ + int bitset_i; + for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i) + dest[bitset_i] |= src[bitset_i]; +} + +static inline void +bitset_mask (bitset_t dest, const bitset_t src) +{ + int bitset_i; + for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i) + dest[bitset_i] &= src[bitset_i]; +} + +#ifdef RE_ENABLE_I18N +/* Inline functions for re_string. */ +static inline int +internal_function __attribute ((pure)) +re_string_char_size_at (const re_string_t *pstr, Idx idx) +{ + int byte_idx; + if (pstr->mb_cur_max == 1) + return 1; + for (byte_idx = 1; idx + byte_idx < pstr->valid_len; ++byte_idx) + if (pstr->wcs[idx + byte_idx] != WEOF) + break; + return byte_idx; +} + +static inline wint_t +internal_function __attribute ((pure)) +re_string_wchar_at (const re_string_t *pstr, Idx idx) +{ + if (pstr->mb_cur_max == 1) + return (wint_t) pstr->mbs[idx]; + return (wint_t) pstr->wcs[idx]; +} + +static int +internal_function __attribute ((pure)) +re_string_elem_size_at (const re_string_t *pstr, Idx idx) +{ +# ifdef _LIBC + const unsigned char *p, *extra; + const int32_t *table, *indirect; + int32_t tmp; +# include <locale/weight.h> + uint_fast32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + + if (nrules != 0) + { + table = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB); + indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_INDIRECTMB); + p = pstr->mbs + idx; + tmp = findidx (&p); + return p - pstr->mbs - idx; + } + else +# endif /* _LIBC */ + return 1; +} +#endif /* RE_ENABLE_I18N */ + +#endif /* _REGEX_INTERNAL_H */ diff --git a/gl/regexec.c b/gl/regexec.c new file mode 100644 index 00000000..7c186aa2 --- /dev/null +++ b/gl/regexec.c @@ -0,0 +1,4398 @@ +/* Extended regular expression matching and search library. + Copyright (C) 2002, 2003, 2004, 2005, 2006 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +static reg_errcode_t match_ctx_init (re_match_context_t *cache, int eflags, + Idx n) internal_function; +static void match_ctx_clean (re_match_context_t *mctx) internal_function; +static void match_ctx_free (re_match_context_t *cache) internal_function; +static reg_errcode_t match_ctx_add_entry (re_match_context_t *cache, Idx node, + Idx str_idx, Idx from, Idx to) + internal_function; +static Idx search_cur_bkref_entry (const re_match_context_t *mctx, Idx str_idx) + internal_function; +static reg_errcode_t match_ctx_add_subtop (re_match_context_t *mctx, Idx node, + Idx str_idx) internal_function; +static re_sub_match_last_t * match_ctx_add_sublast (re_sub_match_top_t *subtop, + Idx node, Idx str_idx) + internal_function; +static void sift_ctx_init (re_sift_context_t *sctx, re_dfastate_t **sifted_sts, + re_dfastate_t **limited_sts, Idx last_node, + Idx last_str_idx) + internal_function; +static reg_errcode_t re_search_internal (const regex_t *preg, + const char *string, Idx length, + Idx start, Idx last_start, Idx stop, + size_t nmatch, regmatch_t pmatch[], + int eflags) internal_function; +static regoff_t re_search_2_stub (struct re_pattern_buffer *bufp, + const char *string1, Idx length1, + const char *string2, Idx length2, + Idx start, regoff_t range, + struct re_registers *regs, + Idx stop, bool ret_len) internal_function; +static regoff_t re_search_stub (struct re_pattern_buffer *bufp, + const char *string, Idx length, Idx start, + regoff_t range, Idx stop, + struct re_registers *regs, + bool ret_len) internal_function; +static unsigned int re_copy_regs (struct re_registers *regs, regmatch_t *pmatch, + Idx nregs, int regs_allocated) + internal_function; +static reg_errcode_t prune_impossible_nodes (re_match_context_t *mctx) + internal_function; +static Idx check_matching (re_match_context_t *mctx, bool fl_longest_match, + Idx *p_match_first) internal_function; +static Idx check_halt_state_context (const re_match_context_t *mctx, + const re_dfastate_t *state, Idx idx) + internal_function; +static void update_regs (const re_dfa_t *dfa, regmatch_t *pmatch, + regmatch_t *prev_idx_match, Idx cur_node, + Idx cur_idx, Idx nmatch) internal_function; +static reg_errcode_t push_fail_stack (struct re_fail_stack_t *fs, + Idx str_idx, Idx dest_node, Idx nregs, + regmatch_t *regs, + re_node_set *eps_via_nodes) + internal_function; +static reg_errcode_t set_regs (const regex_t *preg, + const re_match_context_t *mctx, + size_t nmatch, regmatch_t *pmatch, + bool fl_backtrack) internal_function; +static reg_errcode_t free_fail_stack_return (struct re_fail_stack_t *fs) + internal_function; + +#ifdef RE_ENABLE_I18N +static int sift_states_iter_mb (const re_match_context_t *mctx, + re_sift_context_t *sctx, + Idx node_idx, Idx str_idx, Idx max_str_idx) + internal_function; +#endif /* RE_ENABLE_I18N */ +static reg_errcode_t sift_states_backward (const re_match_context_t *mctx, + re_sift_context_t *sctx) + internal_function; +static reg_errcode_t build_sifted_states (const re_match_context_t *mctx, + re_sift_context_t *sctx, Idx str_idx, + re_node_set *cur_dest) + internal_function; +static reg_errcode_t update_cur_sifted_state (const re_match_context_t *mctx, + re_sift_context_t *sctx, + Idx str_idx, + re_node_set *dest_nodes) + internal_function; +static reg_errcode_t add_epsilon_src_nodes (const re_dfa_t *dfa, + re_node_set *dest_nodes, + const re_node_set *candidates) + internal_function; +static bool check_dst_limits (const re_match_context_t *mctx, + const re_node_set *limits, + Idx dst_node, Idx dst_idx, Idx src_node, + Idx src_idx) internal_function; +static int check_dst_limits_calc_pos_1 (const re_match_context_t *mctx, + int boundaries, Idx subexp_idx, + Idx from_node, Idx bkref_idx) + internal_function; +static int check_dst_limits_calc_pos (const re_match_context_t *mctx, + Idx limit, Idx subexp_idx, + Idx node, Idx str_idx, + Idx bkref_idx) internal_function; +static reg_errcode_t check_subexp_limits (const re_dfa_t *dfa, + re_node_set *dest_nodes, + const re_node_set *candidates, + re_node_set *limits, + struct re_backref_cache_entry *bkref_ents, + Idx str_idx) internal_function; +static reg_errcode_t sift_states_bkref (const re_match_context_t *mctx, + re_sift_context_t *sctx, + Idx str_idx, const re_node_set *candidates) + internal_function; +static reg_errcode_t merge_state_array (const re_dfa_t *dfa, + re_dfastate_t **dst, + re_dfastate_t **src, Idx num) + internal_function; +static re_dfastate_t *find_recover_state (reg_errcode_t *err, + re_match_context_t *mctx) internal_function; +static re_dfastate_t *transit_state (reg_errcode_t *err, + re_match_context_t *mctx, + re_dfastate_t *state) internal_function; +static re_dfastate_t *merge_state_with_log (reg_errcode_t *err, + re_match_context_t *mctx, + re_dfastate_t *next_state) + internal_function; +static reg_errcode_t check_subexp_matching_top (re_match_context_t *mctx, + re_node_set *cur_nodes, + Idx str_idx) internal_function; +#if 0 +static re_dfastate_t *transit_state_sb (reg_errcode_t *err, + re_match_context_t *mctx, + re_dfastate_t *pstate) + internal_function; +#endif +#ifdef RE_ENABLE_I18N +static reg_errcode_t transit_state_mb (re_match_context_t *mctx, + re_dfastate_t *pstate) + internal_function; +#endif /* RE_ENABLE_I18N */ +static reg_errcode_t transit_state_bkref (re_match_context_t *mctx, + const re_node_set *nodes) + internal_function; +static reg_errcode_t get_subexp (re_match_context_t *mctx, + Idx bkref_node, Idx bkref_str_idx) + internal_function; +static reg_errcode_t get_subexp_sub (re_match_context_t *mctx, + const re_sub_match_top_t *sub_top, + re_sub_match_last_t *sub_last, + Idx bkref_node, Idx bkref_str) + internal_function; +static Idx find_subexp_node (const re_dfa_t *dfa, const re_node_set *nodes, + Idx subexp_idx, int type) internal_function; +static reg_errcode_t check_arrival (re_match_context_t *mctx, + state_array_t *path, Idx top_node, + Idx top_str, Idx last_node, Idx last_str, + int type) internal_function; +static reg_errcode_t check_arrival_add_next_nodes (re_match_context_t *mctx, + Idx str_idx, + re_node_set *cur_nodes, + re_node_set *next_nodes) + internal_function; +static reg_errcode_t check_arrival_expand_ecl (const re_dfa_t *dfa, + re_node_set *cur_nodes, + Idx ex_subexp, int type) + internal_function; +static reg_errcode_t check_arrival_expand_ecl_sub (const re_dfa_t *dfa, + re_node_set *dst_nodes, + Idx target, Idx ex_subexp, + int type) internal_function; +static reg_errcode_t expand_bkref_cache (re_match_context_t *mctx, + re_node_set *cur_nodes, Idx cur_str, + Idx subexp_num, int type) + internal_function; +static bool build_trtable (const re_dfa_t *dfa, + re_dfastate_t *state) internal_function; +#ifdef RE_ENABLE_I18N +static int check_node_accept_bytes (const re_dfa_t *dfa, Idx node_idx, + const re_string_t *input, Idx idx) + internal_function; +# ifdef _LIBC +static unsigned int find_collation_sequence_value (const unsigned char *mbs, + size_t name_len) + internal_function; +# endif /* _LIBC */ +#endif /* RE_ENABLE_I18N */ +static Idx group_nodes_into_DFAstates (const re_dfa_t *dfa, + const re_dfastate_t *state, + re_node_set *states_node, + bitset_t *states_ch) internal_function; +static bool check_node_accept (const re_match_context_t *mctx, + const re_token_t *node, Idx idx) + internal_function; +static reg_errcode_t extend_buffers (re_match_context_t *mctx) + internal_function; + +/* Entry point for POSIX code. */ + +/* regexec searches for a given pattern, specified by PREG, in the + string STRING. + + If NMATCH is zero or REG_NOSUB was set in the cflags argument to + `regcomp', we ignore PMATCH. Otherwise, we assume PMATCH has at + least NMATCH elements, and we set them to the offsets of the + corresponding matched substrings. + + EFLAGS specifies `execution flags' which affect matching: if + REG_NOTBOL is set, then ^ does not match at the beginning of the + string; if REG_NOTEOL is set, then $ does not match at the end. + + We return 0 if we find a match and REG_NOMATCH if not. */ + +int +regexec (preg, string, nmatch, pmatch, eflags) + const regex_t *__restrict preg; + const char *__restrict string; + size_t nmatch; + regmatch_t pmatch[]; + int eflags; +{ + reg_errcode_t err; + Idx start, length; +#ifdef _LIBC + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; +#endif + + if (eflags & ~(REG_NOTBOL | REG_NOTEOL | REG_STARTEND)) + return REG_BADPAT; + + if (eflags & REG_STARTEND) + { + start = pmatch[0].rm_so; + length = pmatch[0].rm_eo; + } + else + { + start = 0; + length = strlen (string); + } + + __libc_lock_lock (dfa->lock); + if (preg->no_sub) + err = re_search_internal (preg, string, length, start, length, + length, 0, NULL, eflags); + else + err = re_search_internal (preg, string, length, start, length, + length, nmatch, pmatch, eflags); + __libc_lock_unlock (dfa->lock); + return err != REG_NOERROR; +} + +#ifdef _LIBC +# include <shlib-compat.h> +versioned_symbol (libc, __regexec, regexec, GLIBC_2_3_4); + +# if SHLIB_COMPAT (libc, GLIBC_2_0, GLIBC_2_3_4) +__typeof__ (__regexec) __compat_regexec; + +int +attribute_compat_text_section +__compat_regexec (const regex_t *__restrict preg, + const char *__restrict string, size_t nmatch, + regmatch_t pmatch[], int eflags) +{ + return regexec (preg, string, nmatch, pmatch, + eflags & (REG_NOTBOL | REG_NOTEOL)); +} +compat_symbol (libc, __compat_regexec, regexec, GLIBC_2_0); +# endif +#endif + +/* Entry points for GNU code. */ + +/* re_match, re_search, re_match_2, re_search_2 + + The former two functions operate on STRING with length LENGTH, + while the later two operate on concatenation of STRING1 and STRING2 + with lengths LENGTH1 and LENGTH2, respectively. + + re_match() matches the compiled pattern in BUFP against the string, + starting at index START. + + re_search() first tries matching at index START, then it tries to match + starting from index START + 1, and so on. The last start position tried + is START + RANGE. (Thus RANGE = 0 forces re_search to operate the same + way as re_match().) + + The parameter STOP of re_{match,search}_2 specifies that no match exceeding + the first STOP characters of the concatenation of the strings should be + concerned. + + If REGS is not NULL, and BUFP->no_sub is not set, the offsets of the match + and all groups is stored in REGS. (For the "_2" variants, the offsets are + computed relative to the concatenation, not relative to the individual + strings.) + + On success, re_match* functions return the length of the match, re_search* + return the position of the start of the match. Return value -1 means no + match was found and -2 indicates an internal error. */ + +regoff_t +re_match (bufp, string, length, start, regs) + struct re_pattern_buffer *bufp; + const char *string; + Idx length, start; + struct re_registers *regs; +{ + return re_search_stub (bufp, string, length, start, 0, length, regs, true); +} +#ifdef _LIBC +weak_alias (__re_match, re_match) +#endif + +regoff_t +re_search (bufp, string, length, start, range, regs) + struct re_pattern_buffer *bufp; + const char *string; + Idx length, start; + regoff_t range; + struct re_registers *regs; +{ + return re_search_stub (bufp, string, length, start, range, length, regs, + false); +} +#ifdef _LIBC +weak_alias (__re_search, re_search) +#endif + +regoff_t +re_match_2 (bufp, string1, length1, string2, length2, start, regs, stop) + struct re_pattern_buffer *bufp; + const char *string1, *string2; + Idx length1, length2, start, stop; + struct re_registers *regs; +{ + return re_search_2_stub (bufp, string1, length1, string2, length2, + start, 0, regs, stop, true); +} +#ifdef _LIBC +weak_alias (__re_match_2, re_match_2) +#endif + +regoff_t +re_search_2 (bufp, string1, length1, string2, length2, start, range, regs, stop) + struct re_pattern_buffer *bufp; + const char *string1, *string2; + Idx length1, length2, start, stop; + regoff_t range; + struct re_registers *regs; +{ + return re_search_2_stub (bufp, string1, length1, string2, length2, + start, range, regs, stop, false); +} +#ifdef _LIBC +weak_alias (__re_search_2, re_search_2) +#endif + +static regoff_t +internal_function +re_search_2_stub (struct re_pattern_buffer *bufp, + const char *string1, Idx length1, + const char *string2, Idx length2, + Idx start, regoff_t range, struct re_registers *regs, + Idx stop, bool ret_len) +{ + const char *str; + regoff_t rval; + Idx len = length1 + length2; + char *s = NULL; + + if (BE (length1 < 0 || length2 < 0 || stop < 0 || len < length1, 0)) + return -2; + + /* Concatenate the strings. */ + if (length2 > 0) + if (length1 > 0) + { + s = re_malloc (char, len); + + if (BE (s == NULL, 0)) + return -2; +#ifdef _LIBC + memcpy (__mempcpy (s, string1, length1), string2, length2); +#else + memcpy (s, string1, length1); + memcpy (s + length1, string2, length2); +#endif + str = s; + } + else + str = string2; + else + str = string1; + + rval = re_search_stub (bufp, str, len, start, range, stop, regs, + ret_len); + re_free (s); + return rval; +} + +/* The parameters have the same meaning as those of re_search. + Additional parameters: + If RET_LEN is true the length of the match is returned (re_match style); + otherwise the position of the match is returned. */ + +static regoff_t +internal_function +re_search_stub (struct re_pattern_buffer *bufp, + const char *string, Idx length, + Idx start, regoff_t range, Idx stop, struct re_registers *regs, + bool ret_len) +{ + reg_errcode_t result; + regmatch_t *pmatch; + Idx nregs; + regoff_t rval; + int eflags = 0; +#ifdef _LIBC + re_dfa_t *dfa = (re_dfa_t *) bufp->buffer; +#endif + Idx last_start = start + range; + + /* Check for out-of-range. */ + if (BE (start < 0 || start > length, 0)) + return -1; + if (BE (length < last_start || (0 <= range && last_start < start), 0)) + last_start = length; + else if (BE (last_start < 0 || (range < 0 && start <= last_start), 0)) + last_start = 0; + + __libc_lock_lock (dfa->lock); + + eflags |= (bufp->not_bol) ? REG_NOTBOL : 0; + eflags |= (bufp->not_eol) ? REG_NOTEOL : 0; + + /* Compile fastmap if we haven't yet. */ + if (start < last_start && bufp->fastmap != NULL && !bufp->fastmap_accurate) + re_compile_fastmap (bufp); + + if (BE (bufp->no_sub, 0)) + regs = NULL; + + /* We need at least 1 register. */ + if (regs == NULL) + nregs = 1; + else if (BE (bufp->regs_allocated == REGS_FIXED + && regs->num_regs <= bufp->re_nsub, 0)) + { + nregs = regs->num_regs; + if (BE (nregs < 1, 0)) + { + /* Nothing can be copied to regs. */ + regs = NULL; + nregs = 1; + } + } + else + nregs = bufp->re_nsub + 1; + pmatch = re_malloc (regmatch_t, nregs); + if (BE (pmatch == NULL, 0)) + { + rval = -2; + goto out; + } + + result = re_search_internal (bufp, string, length, start, last_start, stop, + nregs, pmatch, eflags); + + rval = 0; + + /* I hope we needn't fill ther regs with -1's when no match was found. */ + if (result != REG_NOERROR) + rval = -1; + else if (regs != NULL) + { + /* If caller wants register contents data back, copy them. */ + bufp->regs_allocated = re_copy_regs (regs, pmatch, nregs, + bufp->regs_allocated); + if (BE (bufp->regs_allocated == REGS_UNALLOCATED, 0)) + rval = -2; + } + + if (BE (rval == 0, 1)) + { + if (ret_len) + { + assert (pmatch[0].rm_so == start); + rval = pmatch[0].rm_eo - start; + } + else + rval = pmatch[0].rm_so; + } + re_free (pmatch); + out: + __libc_lock_unlock (dfa->lock); + return rval; +} + +static unsigned int +internal_function +re_copy_regs (struct re_registers *regs, regmatch_t *pmatch, Idx nregs, + int regs_allocated) +{ + int rval = REGS_REALLOCATE; + Idx i; + Idx need_regs = nregs + 1; + /* We need one extra element beyond `num_regs' for the `-1' marker GNU code + uses. */ + + /* Have the register data arrays been allocated? */ + if (regs_allocated == REGS_UNALLOCATED) + { /* No. So allocate them with malloc. */ + regs->start = re_malloc (regoff_t, need_regs); + if (BE (regs->start == NULL, 0)) + return REGS_UNALLOCATED; + regs->end = re_malloc (regoff_t, need_regs); + if (BE (regs->end == NULL, 0)) + { + re_free (regs->start); + return REGS_UNALLOCATED; + } + regs->num_regs = need_regs; + } + else if (regs_allocated == REGS_REALLOCATE) + { /* Yes. If we need more elements than were already + allocated, reallocate them. If we need fewer, just + leave it alone. */ + if (BE (need_regs > regs->num_regs, 0)) + { + regoff_t *new_start = re_realloc (regs->start, regoff_t, need_regs); + regoff_t *new_end; + if (BE (new_start == NULL, 0)) + return REGS_UNALLOCATED; + new_end = re_realloc (regs->end, regoff_t, need_regs); + if (BE (new_end == NULL, 0)) + { + re_free (new_start); + return REGS_UNALLOCATED; + } + regs->start = new_start; + regs->end = new_end; + regs->num_regs = need_regs; + } + } + else + { + assert (regs_allocated == REGS_FIXED); + /* This function may not be called with REGS_FIXED and nregs too big. */ + assert (regs->num_regs >= nregs); + rval = REGS_FIXED; + } + + /* Copy the regs. */ + for (i = 0; i < nregs; ++i) + { + regs->start[i] = pmatch[i].rm_so; + regs->end[i] = pmatch[i].rm_eo; + } + for ( ; i < regs->num_regs; ++i) + regs->start[i] = regs->end[i] = -1; + + return rval; +} + +/* Set REGS to hold NUM_REGS registers, storing them in STARTS and + ENDS. Subsequent matches using PATTERN_BUFFER and REGS will use + this memory for recording register information. STARTS and ENDS + must be allocated using the malloc library routine, and must each + be at least NUM_REGS * sizeof (regoff_t) bytes long. + + If NUM_REGS == 0, then subsequent matches should allocate their own + register data. + + Unless this function is called, the first search or match using + PATTERN_BUFFER will allocate its own register data, without + freeing the old data. */ + +void +re_set_registers (bufp, regs, num_regs, starts, ends) + struct re_pattern_buffer *bufp; + struct re_registers *regs; + __re_size_t num_regs; + regoff_t *starts, *ends; +{ + if (num_regs) + { + bufp->regs_allocated = REGS_REALLOCATE; + regs->num_regs = num_regs; + regs->start = starts; + regs->end = ends; + } + else + { + bufp->regs_allocated = REGS_UNALLOCATED; + regs->num_regs = 0; + regs->start = regs->end = NULL; + } +} +#ifdef _LIBC +weak_alias (__re_set_registers, re_set_registers) +#endif + +/* Entry points compatible with 4.2 BSD regex library. We don't define + them unless specifically requested. */ + +#if defined _REGEX_RE_COMP || defined _LIBC +int +# ifdef _LIBC +weak_function +# endif +re_exec (s) + const char *s; +{ + return 0 == regexec (&re_comp_buf, s, 0, NULL, 0); +} +#endif /* _REGEX_RE_COMP */ + +/* Internal entry point. */ + +/* Searches for a compiled pattern PREG in the string STRING, whose + length is LENGTH. NMATCH, PMATCH, and EFLAGS have the same + meaning as with regexec. LAST_START is START + RANGE, where + START and RANGE have the same meaning as with re_search. + Return REG_NOERROR if we find a match, and REG_NOMATCH if not, + otherwise return the error code. + Note: We assume front end functions already check ranges. + (0 <= LAST_START && LAST_START <= LENGTH) */ + +static reg_errcode_t +internal_function +re_search_internal (const regex_t *preg, + const char *string, Idx length, + Idx start, Idx last_start, Idx stop, + size_t nmatch, regmatch_t pmatch[], + int eflags) +{ + reg_errcode_t err; + const re_dfa_t *dfa = (const re_dfa_t *) preg->buffer; + Idx left_lim, right_lim; + int incr; + bool fl_longest_match; + int match_kind; + Idx match_first; + Idx match_last = REG_MISSING; + Idx extra_nmatch; + bool sb; + int ch; +#if defined _LIBC || (defined __STDC_VERSION__ && __STDC_VERSION__ >= 199901L) + re_match_context_t mctx = { .dfa = dfa }; +#else + re_match_context_t mctx; +#endif + char *fastmap = ((preg->fastmap != NULL && preg->fastmap_accurate + && start != last_start && !preg->can_be_null) + ? preg->fastmap : NULL); + RE_TRANSLATE_TYPE t = preg->translate; + +#if !(defined _LIBC || (defined __STDC_VERSION__ && __STDC_VERSION__ >= 199901L)) + memset (&mctx, '\0', sizeof (re_match_context_t)); + mctx.dfa = dfa; +#endif + + extra_nmatch = (nmatch > preg->re_nsub) ? nmatch - (preg->re_nsub + 1) : 0; + nmatch -= extra_nmatch; + + /* Check if the DFA haven't been compiled. */ + if (BE (preg->used == 0 || dfa->init_state == NULL + || dfa->init_state_word == NULL || dfa->init_state_nl == NULL + || dfa->init_state_begbuf == NULL, 0)) + return REG_NOMATCH; + +#ifdef DEBUG + /* We assume front-end functions already check them. */ + assert (0 <= last_start && last_start <= length); +#endif + + /* If initial states with non-begbuf contexts have no elements, + the regex must be anchored. If preg->newline_anchor is set, + we'll never use init_state_nl, so do not check it. */ + if (dfa->init_state->nodes.nelem == 0 + && dfa->init_state_word->nodes.nelem == 0 + && (dfa->init_state_nl->nodes.nelem == 0 + || !preg->newline_anchor)) + { + if (start != 0 && last_start != 0) + return REG_NOMATCH; + start = last_start = 0; + } + + /* We must check the longest matching, if nmatch > 0. */ + fl_longest_match = (nmatch != 0 || dfa->nbackref); + + err = re_string_allocate (&mctx.input, string, length, dfa->nodes_len + 1, + preg->translate, preg->syntax & RE_ICASE, dfa); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + mctx.input.stop = stop; + mctx.input.raw_stop = stop; + mctx.input.newline_anchor = preg->newline_anchor; + + err = match_ctx_init (&mctx, eflags, dfa->nbackref * 2); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + + /* We will log all the DFA states through which the dfa pass, + if nmatch > 1, or this dfa has "multibyte node", which is a + back-reference or a node which can accept multibyte character or + multi character collating element. */ + if (nmatch > 1 || dfa->has_mb_node) + { + /* Avoid overflow. */ + if (BE (SIZE_MAX / sizeof (re_dfastate_t *) <= mctx.input.bufs_len, 0)) + { + err = REG_ESPACE; + goto free_return; + } + + mctx.state_log = re_malloc (re_dfastate_t *, mctx.input.bufs_len + 1); + if (BE (mctx.state_log == NULL, 0)) + { + err = REG_ESPACE; + goto free_return; + } + } + else + mctx.state_log = NULL; + + match_first = start; + mctx.input.tip_context = (eflags & REG_NOTBOL) ? CONTEXT_BEGBUF + : CONTEXT_NEWLINE | CONTEXT_BEGBUF; + + /* Check incrementally whether of not the input string match. */ + incr = (last_start < start) ? -1 : 1; + left_lim = (last_start < start) ? last_start : start; + right_lim = (last_start < start) ? start : last_start; + sb = dfa->mb_cur_max == 1; + match_kind = + (fastmap + ? ((sb || !(preg->syntax & RE_ICASE || t) ? 4 : 0) + | (start <= last_start ? 2 : 0) + | (t != NULL ? 1 : 0)) + : 8); + + for (;; match_first += incr) + { + err = REG_NOMATCH; + if (match_first < left_lim || right_lim < match_first) + goto free_return; + + /* Advance as rapidly as possible through the string, until we + find a plausible place to start matching. This may be done + with varying efficiency, so there are various possibilities: + only the most common of them are specialized, in order to + save on code size. We use a switch statement for speed. */ + switch (match_kind) + { + case 8: + /* No fastmap. */ + break; + + case 7: + /* Fastmap with single-byte translation, match forward. */ + while (BE (match_first < right_lim, 1) + && !fastmap[t[(unsigned char) string[match_first]]]) + ++match_first; + goto forward_match_found_start_or_reached_end; + + case 6: + /* Fastmap without translation, match forward. */ + while (BE (match_first < right_lim, 1) + && !fastmap[(unsigned char) string[match_first]]) + ++match_first; + + forward_match_found_start_or_reached_end: + if (BE (match_first == right_lim, 0)) + { + ch = match_first >= length + ? 0 : (unsigned char) string[match_first]; + if (!fastmap[t ? t[ch] : ch]) + goto free_return; + } + break; + + case 4: + case 5: + /* Fastmap without multi-byte translation, match backwards. */ + while (match_first >= left_lim) + { + ch = match_first >= length + ? 0 : (unsigned char) string[match_first]; + if (fastmap[t ? t[ch] : ch]) + break; + --match_first; + } + if (match_first < left_lim) + goto free_return; + break; + + default: + /* In this case, we can't determine easily the current byte, + since it might be a component byte of a multibyte + character. Then we use the constructed buffer instead. */ + for (;;) + { + /* If MATCH_FIRST is out of the valid range, reconstruct the + buffers. */ + __re_size_t offset = match_first - mctx.input.raw_mbs_idx; + if (BE (offset >= (__re_size_t) mctx.input.valid_raw_len, 0)) + { + err = re_string_reconstruct (&mctx.input, match_first, + eflags); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + + offset = match_first - mctx.input.raw_mbs_idx; + } + /* If MATCH_FIRST is out of the buffer, leave it as '\0'. + Note that MATCH_FIRST must not be smaller than 0. */ + ch = (match_first >= length + ? 0 : re_string_byte_at (&mctx.input, offset)); + if (fastmap[ch]) + break; + match_first += incr; + if (match_first < left_lim || match_first > right_lim) + { + err = REG_NOMATCH; + goto free_return; + } + } + break; + } + + /* Reconstruct the buffers so that the matcher can assume that + the matching starts from the beginning of the buffer. */ + err = re_string_reconstruct (&mctx.input, match_first, eflags); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + +#ifdef RE_ENABLE_I18N + /* Don't consider this char as a possible match start if it part, + yet isn't the head, of a multibyte character. */ + if (!sb && !re_string_first_byte (&mctx.input, 0)) + continue; +#endif + + /* It seems to be appropriate one, then use the matcher. */ + /* We assume that the matching starts from 0. */ + mctx.state_log_top = mctx.nbkref_ents = mctx.max_mb_elem_len = 0; + match_last = check_matching (&mctx, fl_longest_match, + start <= last_start ? &match_first : NULL); + if (match_last != REG_MISSING) + { + if (BE (match_last == REG_ERROR, 0)) + { + err = REG_ESPACE; + goto free_return; + } + else + { + mctx.match_last = match_last; + if ((!preg->no_sub && nmatch > 1) || dfa->nbackref) + { + re_dfastate_t *pstate = mctx.state_log[match_last]; + mctx.last_node = check_halt_state_context (&mctx, pstate, + match_last); + } + if ((!preg->no_sub && nmatch > 1 && dfa->has_plural_match) + || dfa->nbackref) + { + err = prune_impossible_nodes (&mctx); + if (err == REG_NOERROR) + break; + if (BE (err != REG_NOMATCH, 0)) + goto free_return; + match_last = REG_MISSING; + } + else + break; /* We found a match. */ + } + } + + match_ctx_clean (&mctx); + } + +#ifdef DEBUG + assert (match_last != REG_MISSING); + assert (err == REG_NOERROR); +#endif + + /* Set pmatch[] if we need. */ + if (nmatch > 0) + { + Idx reg_idx; + + /* Initialize registers. */ + for (reg_idx = 1; reg_idx < nmatch; ++reg_idx) + pmatch[reg_idx].rm_so = pmatch[reg_idx].rm_eo = -1; + + /* Set the points where matching start/end. */ + pmatch[0].rm_so = 0; + pmatch[0].rm_eo = mctx.match_last; + /* FIXME: This function should fail if mctx.match_last exceeds + the maximum possible regoff_t value. We need a new error + code REG_OVERFLOW. */ + + if (!preg->no_sub && nmatch > 1) + { + err = set_regs (preg, &mctx, nmatch, pmatch, + dfa->has_plural_match && dfa->nbackref > 0); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + } + + /* At last, add the offset to the each registers, since we slided + the buffers so that we could assume that the matching starts + from 0. */ + for (reg_idx = 0; reg_idx < nmatch; ++reg_idx) + if (pmatch[reg_idx].rm_so != -1) + { +#ifdef RE_ENABLE_I18N + if (BE (mctx.input.offsets_needed != 0, 0)) + { + pmatch[reg_idx].rm_so = + (pmatch[reg_idx].rm_so == mctx.input.valid_len + ? mctx.input.valid_raw_len + : mctx.input.offsets[pmatch[reg_idx].rm_so]); + pmatch[reg_idx].rm_eo = + (pmatch[reg_idx].rm_eo == mctx.input.valid_len + ? mctx.input.valid_raw_len + : mctx.input.offsets[pmatch[reg_idx].rm_eo]); + } +#else + assert (mctx.input.offsets_needed == 0); +#endif + pmatch[reg_idx].rm_so += match_first; + pmatch[reg_idx].rm_eo += match_first; + } + for (reg_idx = 0; reg_idx < extra_nmatch; ++reg_idx) + { + pmatch[nmatch + reg_idx].rm_so = -1; + pmatch[nmatch + reg_idx].rm_eo = -1; + } + + if (dfa->subexp_map) + for (reg_idx = 0; reg_idx + 1 < nmatch; reg_idx++) + if (dfa->subexp_map[reg_idx] != reg_idx) + { + pmatch[reg_idx + 1].rm_so + = pmatch[dfa->subexp_map[reg_idx] + 1].rm_so; + pmatch[reg_idx + 1].rm_eo + = pmatch[dfa->subexp_map[reg_idx] + 1].rm_eo; + } + } + + free_return: + re_free (mctx.state_log); + if (dfa->nbackref) + match_ctx_free (&mctx); + re_string_destruct (&mctx.input); + return err; +} + +static reg_errcode_t +internal_function +prune_impossible_nodes (re_match_context_t *mctx) +{ + const re_dfa_t *const dfa = mctx->dfa; + Idx halt_node, match_last; + reg_errcode_t ret; + re_dfastate_t **sifted_states; + re_dfastate_t **lim_states = NULL; + re_sift_context_t sctx; +#ifdef DEBUG + assert (mctx->state_log != NULL); +#endif + match_last = mctx->match_last; + halt_node = mctx->last_node; + + /* Avoid overflow. */ + if (BE (SIZE_MAX / sizeof (re_dfastate_t *) <= match_last, 0)) + return REG_ESPACE; + + sifted_states = re_malloc (re_dfastate_t *, match_last + 1); + if (BE (sifted_states == NULL, 0)) + { + ret = REG_ESPACE; + goto free_return; + } + if (dfa->nbackref) + { + lim_states = re_malloc (re_dfastate_t *, match_last + 1); + if (BE (lim_states == NULL, 0)) + { + ret = REG_ESPACE; + goto free_return; + } + while (1) + { + memset (lim_states, '\0', + sizeof (re_dfastate_t *) * (match_last + 1)); + sift_ctx_init (&sctx, sifted_states, lim_states, halt_node, + match_last); + ret = sift_states_backward (mctx, &sctx); + re_node_set_free (&sctx.limits); + if (BE (ret != REG_NOERROR, 0)) + goto free_return; + if (sifted_states[0] != NULL || lim_states[0] != NULL) + break; + do + { + --match_last; + if (! REG_VALID_INDEX (match_last)) + { + ret = REG_NOMATCH; + goto free_return; + } + } while (mctx->state_log[match_last] == NULL + || !mctx->state_log[match_last]->halt); + halt_node = check_halt_state_context (mctx, + mctx->state_log[match_last], + match_last); + } + ret = merge_state_array (dfa, sifted_states, lim_states, + match_last + 1); + re_free (lim_states); + lim_states = NULL; + if (BE (ret != REG_NOERROR, 0)) + goto free_return; + } + else + { + sift_ctx_init (&sctx, sifted_states, lim_states, halt_node, match_last); + ret = sift_states_backward (mctx, &sctx); + re_node_set_free (&sctx.limits); + if (BE (ret != REG_NOERROR, 0)) + goto free_return; + } + re_free (mctx->state_log); + mctx->state_log = sifted_states; + sifted_states = NULL; + mctx->last_node = halt_node; + mctx->match_last = match_last; + ret = REG_NOERROR; + free_return: + re_free (sifted_states); + re_free (lim_states); + return ret; +} + +/* Acquire an initial state and return it. + We must select appropriate initial state depending on the context, + since initial states may have constraints like "\<", "^", etc.. */ + +static inline re_dfastate_t * +__attribute ((always_inline)) internal_function +acquire_init_state_context (reg_errcode_t *err, const re_match_context_t *mctx, + Idx idx) +{ + const re_dfa_t *const dfa = mctx->dfa; + if (dfa->init_state->has_constraint) + { + unsigned int context; + context = re_string_context_at (&mctx->input, idx - 1, mctx->eflags); + if (IS_WORD_CONTEXT (context)) + return dfa->init_state_word; + else if (IS_ORDINARY_CONTEXT (context)) + return dfa->init_state; + else if (IS_BEGBUF_CONTEXT (context) && IS_NEWLINE_CONTEXT (context)) + return dfa->init_state_begbuf; + else if (IS_NEWLINE_CONTEXT (context)) + return dfa->init_state_nl; + else if (IS_BEGBUF_CONTEXT (context)) + { + /* It is relatively rare case, then calculate on demand. */ + return re_acquire_state_context (err, dfa, + dfa->init_state->entrance_nodes, + context); + } + else + /* Must not happen? */ + return dfa->init_state; + } + else + return dfa->init_state; +} + +/* Check whether the regular expression match input string INPUT or not, + and return the index where the matching end. Return REG_MISSING if + there is no match, and return REG_ERROR in case of an error. + FL_LONGEST_MATCH means we want the POSIX longest matching. + If P_MATCH_FIRST is not NULL, and the match fails, it is set to the + next place where we may want to try matching. + Note that the matcher assume that the maching starts from the current + index of the buffer. */ + +static Idx +internal_function +check_matching (re_match_context_t *mctx, bool fl_longest_match, + Idx *p_match_first) +{ + const re_dfa_t *const dfa = mctx->dfa; + reg_errcode_t err; + Idx match = 0; + Idx match_last = REG_MISSING; + Idx cur_str_idx = re_string_cur_idx (&mctx->input); + re_dfastate_t *cur_state; + bool at_init_state = p_match_first != NULL; + Idx next_start_idx = cur_str_idx; + + err = REG_NOERROR; + cur_state = acquire_init_state_context (&err, mctx, cur_str_idx); + /* An initial state must not be NULL (invalid). */ + if (BE (cur_state == NULL, 0)) + { + assert (err == REG_ESPACE); + return REG_ERROR; + } + + if (mctx->state_log != NULL) + { + mctx->state_log[cur_str_idx] = cur_state; + + /* Check OP_OPEN_SUBEXP in the initial state in case that we use them + later. E.g. Processing back references. */ + if (BE (dfa->nbackref, 0)) + { + at_init_state = false; + err = check_subexp_matching_top (mctx, &cur_state->nodes, 0); + if (BE (err != REG_NOERROR, 0)) + return err; + + if (cur_state->has_backref) + { + err = transit_state_bkref (mctx, &cur_state->nodes); + if (BE (err != REG_NOERROR, 0)) + return err; + } + } + } + + /* If the RE accepts NULL string. */ + if (BE (cur_state->halt, 0)) + { + if (!cur_state->has_constraint + || check_halt_state_context (mctx, cur_state, cur_str_idx)) + { + if (!fl_longest_match) + return cur_str_idx; + else + { + match_last = cur_str_idx; + match = 1; + } + } + } + + while (!re_string_eoi (&mctx->input)) + { + re_dfastate_t *old_state = cur_state; + Idx next_char_idx = re_string_cur_idx (&mctx->input) + 1; + + if (BE (next_char_idx >= mctx->input.bufs_len, 0) + || (BE (next_char_idx >= mctx->input.valid_len, 0) + && mctx->input.valid_len < mctx->input.len)) + { + err = extend_buffers (mctx); + if (BE (err != REG_NOERROR, 0)) + { + assert (err == REG_ESPACE); + return REG_ERROR; + } + } + + cur_state = transit_state (&err, mctx, cur_state); + if (mctx->state_log != NULL) + cur_state = merge_state_with_log (&err, mctx, cur_state); + + if (cur_state == NULL) + { + /* Reached the invalid state or an error. Try to recover a valid + state using the state log, if available and if we have not + already found a valid (even if not the longest) match. */ + if (BE (err != REG_NOERROR, 0)) + return REG_ERROR; + + if (mctx->state_log == NULL + || (match && !fl_longest_match) + || (cur_state = find_recover_state (&err, mctx)) == NULL) + break; + } + + if (BE (at_init_state, 0)) + { + if (old_state == cur_state) + next_start_idx = next_char_idx; + else + at_init_state = false; + } + + if (cur_state->halt) + { + /* Reached a halt state. + Check the halt state can satisfy the current context. */ + if (!cur_state->has_constraint + || check_halt_state_context (mctx, cur_state, + re_string_cur_idx (&mctx->input))) + { + /* We found an appropriate halt state. */ + match_last = re_string_cur_idx (&mctx->input); + match = 1; + + /* We found a match, do not modify match_first below. */ + p_match_first = NULL; + if (!fl_longest_match) + break; + } + } + } + + if (p_match_first) + *p_match_first += next_start_idx; + + return match_last; +} + +/* Check NODE match the current context. */ + +static bool +internal_function +check_halt_node_context (const re_dfa_t *dfa, Idx node, unsigned int context) +{ + re_token_type_t type = dfa->nodes[node].type; + unsigned int constraint = dfa->nodes[node].constraint; + if (type != END_OF_RE) + return false; + if (!constraint) + return true; + if (NOT_SATISFY_NEXT_CONSTRAINT (constraint, context)) + return false; + return true; +} + +/* Check the halt state STATE match the current context. + Return 0 if not match, if the node, STATE has, is a halt node and + match the context, return the node. */ + +static Idx +internal_function +check_halt_state_context (const re_match_context_t *mctx, + const re_dfastate_t *state, Idx idx) +{ + Idx i; + unsigned int context; +#ifdef DEBUG + assert (state->halt); +#endif + context = re_string_context_at (&mctx->input, idx, mctx->eflags); + for (i = 0; i < state->nodes.nelem; ++i) + if (check_halt_node_context (mctx->dfa, state->nodes.elems[i], context)) + return state->nodes.elems[i]; + return 0; +} + +/* Compute the next node to which "NFA" transit from NODE("NFA" is a NFA + corresponding to the DFA). + Return the destination node, and update EPS_VIA_NODES; + return REG_MISSING in case of errors. */ + +static Idx +internal_function +proceed_next_node (const re_match_context_t *mctx, Idx nregs, regmatch_t *regs, + Idx *pidx, Idx node, re_node_set *eps_via_nodes, + struct re_fail_stack_t *fs) +{ + const re_dfa_t *const dfa = mctx->dfa; + Idx i; + bool ok; + if (IS_EPSILON_NODE (dfa->nodes[node].type)) + { + re_node_set *cur_nodes = &mctx->state_log[*pidx]->nodes; + re_node_set *edests = &dfa->edests[node]; + Idx dest_node; + ok = re_node_set_insert (eps_via_nodes, node); + if (BE (! ok, 0)) + return REG_ERROR; + /* Pick up a valid destination, or return REG_MISSING if none + is found. */ + for (dest_node = REG_MISSING, i = 0; i < edests->nelem; ++i) + { + Idx candidate = edests->elems[i]; + if (!re_node_set_contains (cur_nodes, candidate)) + continue; + if (dest_node == REG_MISSING) + dest_node = candidate; + + else + { + /* In order to avoid infinite loop like "(a*)*", return the second + epsilon-transition if the first was already considered. */ + if (re_node_set_contains (eps_via_nodes, dest_node)) + return candidate; + + /* Otherwise, push the second epsilon-transition on the fail stack. */ + else if (fs != NULL + && push_fail_stack (fs, *pidx, candidate, nregs, regs, + eps_via_nodes)) + return REG_ERROR; + + /* We know we are going to exit. */ + break; + } + } + return dest_node; + } + else + { + Idx naccepted = 0; + re_token_type_t type = dfa->nodes[node].type; + +#ifdef RE_ENABLE_I18N + if (dfa->nodes[node].accept_mb) + naccepted = check_node_accept_bytes (dfa, node, &mctx->input, *pidx); + else +#endif /* RE_ENABLE_I18N */ + if (type == OP_BACK_REF) + { + Idx subexp_idx = dfa->nodes[node].opr.idx + 1; + naccepted = regs[subexp_idx].rm_eo - regs[subexp_idx].rm_so; + if (fs != NULL) + { + if (regs[subexp_idx].rm_so == -1 || regs[subexp_idx].rm_eo == -1) + return REG_MISSING; + else if (naccepted) + { + char *buf = (char *) re_string_get_buffer (&mctx->input); + if (memcmp (buf + regs[subexp_idx].rm_so, buf + *pidx, + naccepted) != 0) + return REG_MISSING; + } + } + + if (naccepted == 0) + { + Idx dest_node; + ok = re_node_set_insert (eps_via_nodes, node); + if (BE (! ok, 0)) + return REG_ERROR; + dest_node = dfa->edests[node].elems[0]; + if (re_node_set_contains (&mctx->state_log[*pidx]->nodes, + dest_node)) + return dest_node; + } + } + + if (naccepted != 0 + || check_node_accept (mctx, dfa->nodes + node, *pidx)) + { + Idx dest_node = dfa->nexts[node]; + *pidx = (naccepted == 0) ? *pidx + 1 : *pidx + naccepted; + if (fs && (*pidx > mctx->match_last || mctx->state_log[*pidx] == NULL + || !re_node_set_contains (&mctx->state_log[*pidx]->nodes, + dest_node))) + return REG_MISSING; + re_node_set_empty (eps_via_nodes); + return dest_node; + } + } + return REG_MISSING; +} + +static reg_errcode_t +internal_function +push_fail_stack (struct re_fail_stack_t *fs, Idx str_idx, Idx dest_node, + Idx nregs, regmatch_t *regs, re_node_set *eps_via_nodes) +{ + reg_errcode_t err; + Idx num = fs->num++; + if (fs->num == fs->alloc) + { + struct re_fail_stack_ent_t *new_array; + new_array = realloc (fs->stack, (sizeof (struct re_fail_stack_ent_t) + * fs->alloc * 2)); + if (new_array == NULL) + return REG_ESPACE; + fs->alloc *= 2; + fs->stack = new_array; + } + fs->stack[num].idx = str_idx; + fs->stack[num].node = dest_node; + fs->stack[num].regs = re_malloc (regmatch_t, nregs); + if (fs->stack[num].regs == NULL) + return REG_ESPACE; + memcpy (fs->stack[num].regs, regs, sizeof (regmatch_t) * nregs); + err = re_node_set_init_copy (&fs->stack[num].eps_via_nodes, eps_via_nodes); + return err; +} + +static Idx +internal_function +pop_fail_stack (struct re_fail_stack_t *fs, Idx *pidx, Idx nregs, + regmatch_t *regs, re_node_set *eps_via_nodes) +{ + Idx num = --fs->num; + assert (REG_VALID_INDEX (num)); + *pidx = fs->stack[num].idx; + memcpy (regs, fs->stack[num].regs, sizeof (regmatch_t) * nregs); + re_node_set_free (eps_via_nodes); + re_free (fs->stack[num].regs); + *eps_via_nodes = fs->stack[num].eps_via_nodes; + return fs->stack[num].node; +} + +/* Set the positions where the subexpressions are starts/ends to registers + PMATCH. + Note: We assume that pmatch[0] is already set, and + pmatch[i].rm_so == pmatch[i].rm_eo == -1 for 0 < i < nmatch. */ + +static reg_errcode_t +internal_function +set_regs (const regex_t *preg, const re_match_context_t *mctx, size_t nmatch, + regmatch_t *pmatch, bool fl_backtrack) +{ + const re_dfa_t *dfa = (const re_dfa_t *) preg->buffer; + Idx idx, cur_node; + re_node_set eps_via_nodes; + struct re_fail_stack_t *fs; + struct re_fail_stack_t fs_body = { 0, 2, NULL }; + regmatch_t *prev_idx_match; + bool prev_idx_match_malloced = false; + +#ifdef DEBUG + assert (nmatch > 1); + assert (mctx->state_log != NULL); +#endif + if (fl_backtrack) + { + fs = &fs_body; + fs->stack = re_malloc (struct re_fail_stack_ent_t, fs->alloc); + if (fs->stack == NULL) + return REG_ESPACE; + } + else + fs = NULL; + + cur_node = dfa->init_node; + re_node_set_init_empty (&eps_via_nodes); + + if (__libc_use_alloca (nmatch * sizeof (regmatch_t))) + prev_idx_match = (regmatch_t *) alloca (nmatch * sizeof (regmatch_t)); + else + { + prev_idx_match = re_malloc (regmatch_t, nmatch); + if (prev_idx_match == NULL) + { + free_fail_stack_return (fs); + return REG_ESPACE; + } + prev_idx_match_malloced = true; + } + memcpy (prev_idx_match, pmatch, sizeof (regmatch_t) * nmatch); + + for (idx = pmatch[0].rm_so; idx <= pmatch[0].rm_eo ;) + { + update_regs (dfa, pmatch, prev_idx_match, cur_node, idx, nmatch); + + if (idx == pmatch[0].rm_eo && cur_node == mctx->last_node) + { + Idx reg_idx; + if (fs) + { + for (reg_idx = 0; reg_idx < nmatch; ++reg_idx) + if (pmatch[reg_idx].rm_so > -1 && pmatch[reg_idx].rm_eo == -1) + break; + if (reg_idx == nmatch) + { + re_node_set_free (&eps_via_nodes); + if (prev_idx_match_malloced) + re_free (prev_idx_match); + return free_fail_stack_return (fs); + } + cur_node = pop_fail_stack (fs, &idx, nmatch, pmatch, + &eps_via_nodes); + } + else + { + re_node_set_free (&eps_via_nodes); + if (prev_idx_match_malloced) + re_free (prev_idx_match); + return REG_NOERROR; + } + } + + /* Proceed to next node. */ + cur_node = proceed_next_node (mctx, nmatch, pmatch, &idx, cur_node, + &eps_via_nodes, fs); + + if (BE (! REG_VALID_INDEX (cur_node), 0)) + { + if (BE (cur_node == REG_ERROR, 0)) + { + re_node_set_free (&eps_via_nodes); + if (prev_idx_match_malloced) + re_free (prev_idx_match); + free_fail_stack_return (fs); + return REG_ESPACE; + } + if (fs) + cur_node = pop_fail_stack (fs, &idx, nmatch, pmatch, + &eps_via_nodes); + else + { + re_node_set_free (&eps_via_nodes); + if (prev_idx_match_malloced) + re_free (prev_idx_match); + return REG_NOMATCH; + } + } + } + re_node_set_free (&eps_via_nodes); + if (prev_idx_match_malloced) + re_free (prev_idx_match); + return free_fail_stack_return (fs); +} + +static reg_errcode_t +internal_function +free_fail_stack_return (struct re_fail_stack_t *fs) +{ + if (fs) + { + Idx fs_idx; + for (fs_idx = 0; fs_idx < fs->num; ++fs_idx) + { + re_node_set_free (&fs->stack[fs_idx].eps_via_nodes); + re_free (fs->stack[fs_idx].regs); + } + re_free (fs->stack); + } + return REG_NOERROR; +} + +static void +internal_function +update_regs (const re_dfa_t *dfa, regmatch_t *pmatch, + regmatch_t *prev_idx_match, Idx cur_node, Idx cur_idx, Idx nmatch) +{ + int type = dfa->nodes[cur_node].type; + if (type == OP_OPEN_SUBEXP) + { + Idx reg_num = dfa->nodes[cur_node].opr.idx + 1; + + /* We are at the first node of this sub expression. */ + if (reg_num < nmatch) + { + pmatch[reg_num].rm_so = cur_idx; + pmatch[reg_num].rm_eo = -1; + } + } + else if (type == OP_CLOSE_SUBEXP) + { + Idx reg_num = dfa->nodes[cur_node].opr.idx + 1; + if (reg_num < nmatch) + { + /* We are at the last node of this sub expression. */ + if (pmatch[reg_num].rm_so < cur_idx) + { + pmatch[reg_num].rm_eo = cur_idx; + /* This is a non-empty match or we are not inside an optional + subexpression. Accept this right away. */ + memcpy (prev_idx_match, pmatch, sizeof (regmatch_t) * nmatch); + } + else + { + if (dfa->nodes[cur_node].opt_subexp + && prev_idx_match[reg_num].rm_so != -1) + /* We transited through an empty match for an optional + subexpression, like (a?)*, and this is not the subexp's + first match. Copy back the old content of the registers + so that matches of an inner subexpression are undone as + well, like in ((a?))*. */ + memcpy (pmatch, prev_idx_match, sizeof (regmatch_t) * nmatch); + else + /* We completed a subexpression, but it may be part of + an optional one, so do not update PREV_IDX_MATCH. */ + pmatch[reg_num].rm_eo = cur_idx; + } + } + } +} + +/* This function checks the STATE_LOG from the SCTX->last_str_idx to 0 + and sift the nodes in each states according to the following rules. + Updated state_log will be wrote to STATE_LOG. + + Rules: We throw away the Node `a' in the STATE_LOG[STR_IDX] if... + 1. When STR_IDX == MATCH_LAST(the last index in the state_log): + If `a' isn't the LAST_NODE and `a' can't epsilon transit to + the LAST_NODE, we throw away the node `a'. + 2. When 0 <= STR_IDX < MATCH_LAST and `a' accepts + string `s' and transit to `b': + i. If 'b' isn't in the STATE_LOG[STR_IDX+strlen('s')], we throw + away the node `a'. + ii. If 'b' is in the STATE_LOG[STR_IDX+strlen('s')] but 'b' is + thrown away, we throw away the node `a'. + 3. When 0 <= STR_IDX < MATCH_LAST and 'a' epsilon transit to 'b': + i. If 'b' isn't in the STATE_LOG[STR_IDX], we throw away the + node `a'. + ii. If 'b' is in the STATE_LOG[STR_IDX] but 'b' is thrown away, + we throw away the node `a'. */ + +#define STATE_NODE_CONTAINS(state,node) \ + ((state) != NULL && re_node_set_contains (&(state)->nodes, node)) + +static reg_errcode_t +internal_function +sift_states_backward (const re_match_context_t *mctx, re_sift_context_t *sctx) +{ + reg_errcode_t err; + int null_cnt = 0; + Idx str_idx = sctx->last_str_idx; + re_node_set cur_dest; + +#ifdef DEBUG + assert (mctx->state_log != NULL && mctx->state_log[str_idx] != NULL); +#endif + + /* Build sifted state_log[str_idx]. It has the nodes which can epsilon + transit to the last_node and the last_node itself. */ + err = re_node_set_init_1 (&cur_dest, sctx->last_node); + if (BE (err != REG_NOERROR, 0)) + return err; + err = update_cur_sifted_state (mctx, sctx, str_idx, &cur_dest); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + + /* Then check each states in the state_log. */ + while (str_idx > 0) + { + /* Update counters. */ + null_cnt = (sctx->sifted_states[str_idx] == NULL) ? null_cnt + 1 : 0; + if (null_cnt > mctx->max_mb_elem_len) + { + memset (sctx->sifted_states, '\0', + sizeof (re_dfastate_t *) * str_idx); + re_node_set_free (&cur_dest); + return REG_NOERROR; + } + re_node_set_empty (&cur_dest); + --str_idx; + + if (mctx->state_log[str_idx]) + { + err = build_sifted_states (mctx, sctx, str_idx, &cur_dest); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + } + + /* Add all the nodes which satisfy the following conditions: + - It can epsilon transit to a node in CUR_DEST. + - It is in CUR_SRC. + And update state_log. */ + err = update_cur_sifted_state (mctx, sctx, str_idx, &cur_dest); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + } + err = REG_NOERROR; + free_return: + re_node_set_free (&cur_dest); + return err; +} + +static reg_errcode_t +internal_function +build_sifted_states (const re_match_context_t *mctx, re_sift_context_t *sctx, + Idx str_idx, re_node_set *cur_dest) +{ + const re_dfa_t *const dfa = mctx->dfa; + const re_node_set *cur_src = &mctx->state_log[str_idx]->non_eps_nodes; + Idx i; + + /* Then build the next sifted state. + We build the next sifted state on `cur_dest', and update + `sifted_states[str_idx]' with `cur_dest'. + Note: + `cur_dest' is the sifted state from `state_log[str_idx + 1]'. + `cur_src' points the node_set of the old `state_log[str_idx]' + (with the epsilon nodes pre-filtered out). */ + for (i = 0; i < cur_src->nelem; i++) + { + Idx prev_node = cur_src->elems[i]; + int naccepted = 0; + bool ok; + +#ifdef DEBUG + re_token_type_t type = dfa->nodes[prev_node].type; + assert (!IS_EPSILON_NODE (type)); +#endif +#ifdef RE_ENABLE_I18N + /* If the node may accept `multi byte'. */ + if (dfa->nodes[prev_node].accept_mb) + naccepted = sift_states_iter_mb (mctx, sctx, prev_node, + str_idx, sctx->last_str_idx); +#endif /* RE_ENABLE_I18N */ + + /* We don't check backreferences here. + See update_cur_sifted_state(). */ + if (!naccepted + && check_node_accept (mctx, dfa->nodes + prev_node, str_idx) + && STATE_NODE_CONTAINS (sctx->sifted_states[str_idx + 1], + dfa->nexts[prev_node])) + naccepted = 1; + + if (naccepted == 0) + continue; + + if (sctx->limits.nelem) + { + Idx to_idx = str_idx + naccepted; + if (check_dst_limits (mctx, &sctx->limits, + dfa->nexts[prev_node], to_idx, + prev_node, str_idx)) + continue; + } + ok = re_node_set_insert (cur_dest, prev_node); + if (BE (! ok, 0)) + return REG_ESPACE; + } + + return REG_NOERROR; +} + +/* Helper functions. */ + +static reg_errcode_t +internal_function +clean_state_log_if_needed (re_match_context_t *mctx, Idx next_state_log_idx) +{ + Idx top = mctx->state_log_top; + + if (next_state_log_idx >= mctx->input.bufs_len + || (next_state_log_idx >= mctx->input.valid_len + && mctx->input.valid_len < mctx->input.len)) + { + reg_errcode_t err; + err = extend_buffers (mctx); + if (BE (err != REG_NOERROR, 0)) + return err; + } + + if (top < next_state_log_idx) + { + memset (mctx->state_log + top + 1, '\0', + sizeof (re_dfastate_t *) * (next_state_log_idx - top)); + mctx->state_log_top = next_state_log_idx; + } + return REG_NOERROR; +} + +static reg_errcode_t +internal_function +merge_state_array (const re_dfa_t *dfa, re_dfastate_t **dst, + re_dfastate_t **src, Idx num) +{ + Idx st_idx; + reg_errcode_t err; + for (st_idx = 0; st_idx < num; ++st_idx) + { + if (dst[st_idx] == NULL) + dst[st_idx] = src[st_idx]; + else if (src[st_idx] != NULL) + { + re_node_set merged_set; + err = re_node_set_init_union (&merged_set, &dst[st_idx]->nodes, + &src[st_idx]->nodes); + if (BE (err != REG_NOERROR, 0)) + return err; + dst[st_idx] = re_acquire_state (&err, dfa, &merged_set); + re_node_set_free (&merged_set); + if (BE (err != REG_NOERROR, 0)) + return err; + } + } + return REG_NOERROR; +} + +static reg_errcode_t +internal_function +update_cur_sifted_state (const re_match_context_t *mctx, + re_sift_context_t *sctx, Idx str_idx, + re_node_set *dest_nodes) +{ + const re_dfa_t *const dfa = mctx->dfa; + reg_errcode_t err = REG_NOERROR; + const re_node_set *candidates; + candidates = ((mctx->state_log[str_idx] == NULL) ? NULL + : &mctx->state_log[str_idx]->nodes); + + if (dest_nodes->nelem == 0) + sctx->sifted_states[str_idx] = NULL; + else + { + if (candidates) + { + /* At first, add the nodes which can epsilon transit to a node in + DEST_NODE. */ + err = add_epsilon_src_nodes (dfa, dest_nodes, candidates); + if (BE (err != REG_NOERROR, 0)) + return err; + + /* Then, check the limitations in the current sift_context. */ + if (sctx->limits.nelem) + { + err = check_subexp_limits (dfa, dest_nodes, candidates, &sctx->limits, + mctx->bkref_ents, str_idx); + if (BE (err != REG_NOERROR, 0)) + return err; + } + } + + sctx->sifted_states[str_idx] = re_acquire_state (&err, dfa, dest_nodes); + if (BE (err != REG_NOERROR, 0)) + return err; + } + + if (candidates && mctx->state_log[str_idx]->has_backref) + { + err = sift_states_bkref (mctx, sctx, str_idx, candidates); + if (BE (err != REG_NOERROR, 0)) + return err; + } + return REG_NOERROR; +} + +static reg_errcode_t +internal_function +add_epsilon_src_nodes (const re_dfa_t *dfa, re_node_set *dest_nodes, + const re_node_set *candidates) +{ + reg_errcode_t err = REG_NOERROR; + Idx i; + + re_dfastate_t *state = re_acquire_state (&err, dfa, dest_nodes); + if (BE (err != REG_NOERROR, 0)) + return err; + + if (!state->inveclosure.alloc) + { + err = re_node_set_alloc (&state->inveclosure, dest_nodes->nelem); + if (BE (err != REG_NOERROR, 0)) + return REG_ESPACE; + for (i = 0; i < dest_nodes->nelem; i++) + re_node_set_merge (&state->inveclosure, + dfa->inveclosures + dest_nodes->elems[i]); + } + return re_node_set_add_intersect (dest_nodes, candidates, + &state->inveclosure); +} + +static reg_errcode_t +internal_function +sub_epsilon_src_nodes (const re_dfa_t *dfa, Idx node, re_node_set *dest_nodes, + const re_node_set *candidates) +{ + Idx ecl_idx; + reg_errcode_t err; + re_node_set *inv_eclosure = dfa->inveclosures + node; + re_node_set except_nodes; + re_node_set_init_empty (&except_nodes); + for (ecl_idx = 0; ecl_idx < inv_eclosure->nelem; ++ecl_idx) + { + Idx cur_node = inv_eclosure->elems[ecl_idx]; + if (cur_node == node) + continue; + if (IS_EPSILON_NODE (dfa->nodes[cur_node].type)) + { + Idx edst1 = dfa->edests[cur_node].elems[0]; + Idx edst2 = ((dfa->edests[cur_node].nelem > 1) + ? dfa->edests[cur_node].elems[1] : REG_MISSING); + if ((!re_node_set_contains (inv_eclosure, edst1) + && re_node_set_contains (dest_nodes, edst1)) + || (REG_VALID_NONZERO_INDEX (edst2) + && !re_node_set_contains (inv_eclosure, edst2) + && re_node_set_contains (dest_nodes, edst2))) + { + err = re_node_set_add_intersect (&except_nodes, candidates, + dfa->inveclosures + cur_node); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&except_nodes); + return err; + } + } + } + } + for (ecl_idx = 0; ecl_idx < inv_eclosure->nelem; ++ecl_idx) + { + Idx cur_node = inv_eclosure->elems[ecl_idx]; + if (!re_node_set_contains (&except_nodes, cur_node)) + { + Idx idx = re_node_set_contains (dest_nodes, cur_node) - 1; + re_node_set_remove_at (dest_nodes, idx); + } + } + re_node_set_free (&except_nodes); + return REG_NOERROR; +} + +static bool +internal_function +check_dst_limits (const re_match_context_t *mctx, const re_node_set *limits, + Idx dst_node, Idx dst_idx, Idx src_node, Idx src_idx) +{ + const re_dfa_t *const dfa = mctx->dfa; + Idx lim_idx, src_pos, dst_pos; + + Idx dst_bkref_idx = search_cur_bkref_entry (mctx, dst_idx); + Idx src_bkref_idx = search_cur_bkref_entry (mctx, src_idx); + for (lim_idx = 0; lim_idx < limits->nelem; ++lim_idx) + { + Idx subexp_idx; + struct re_backref_cache_entry *ent; + ent = mctx->bkref_ents + limits->elems[lim_idx]; + subexp_idx = dfa->nodes[ent->node].opr.idx; + + dst_pos = check_dst_limits_calc_pos (mctx, limits->elems[lim_idx], + subexp_idx, dst_node, dst_idx, + dst_bkref_idx); + src_pos = check_dst_limits_calc_pos (mctx, limits->elems[lim_idx], + subexp_idx, src_node, src_idx, + src_bkref_idx); + + /* In case of: + <src> <dst> ( <subexp> ) + ( <subexp> ) <src> <dst> + ( <subexp1> <src> <subexp2> <dst> <subexp3> ) */ + if (src_pos == dst_pos) + continue; /* This is unrelated limitation. */ + else + return true; + } + return false; +} + +static int +internal_function +check_dst_limits_calc_pos_1 (const re_match_context_t *mctx, int boundaries, + Idx subexp_idx, Idx from_node, Idx bkref_idx) +{ + const re_dfa_t *const dfa = mctx->dfa; + const re_node_set *eclosures = dfa->eclosures + from_node; + Idx node_idx; + + /* Else, we are on the boundary: examine the nodes on the epsilon + closure. */ + for (node_idx = 0; node_idx < eclosures->nelem; ++node_idx) + { + Idx node = eclosures->elems[node_idx]; + switch (dfa->nodes[node].type) + { + case OP_BACK_REF: + if (bkref_idx != REG_MISSING) + { + struct re_backref_cache_entry *ent = mctx->bkref_ents + bkref_idx; + do + { + Idx dst; + int cpos; + + if (ent->node != node) + continue; + + if (subexp_idx < BITSET_WORD_BITS + && !(ent->eps_reachable_subexps_map + & ((bitset_word_t) 1 << subexp_idx))) + continue; + + /* Recurse trying to reach the OP_OPEN_SUBEXP and + OP_CLOSE_SUBEXP cases below. But, if the + destination node is the same node as the source + node, don't recurse because it would cause an + infinite loop: a regex that exhibits this behavior + is ()\1*\1* */ + dst = dfa->edests[node].elems[0]; + if (dst == from_node) + { + if (boundaries & 1) + return -1; + else /* if (boundaries & 2) */ + return 0; + } + + cpos = + check_dst_limits_calc_pos_1 (mctx, boundaries, subexp_idx, + dst, bkref_idx); + if (cpos == -1 /* && (boundaries & 1) */) + return -1; + if (cpos == 0 && (boundaries & 2)) + return 0; + + if (subexp_idx < BITSET_WORD_BITS) + ent->eps_reachable_subexps_map + &= ~((bitset_word_t) 1 << subexp_idx); + } + while (ent++->more); + } + break; + + case OP_OPEN_SUBEXP: + if ((boundaries & 1) && subexp_idx == dfa->nodes[node].opr.idx) + return -1; + break; + + case OP_CLOSE_SUBEXP: + if ((boundaries & 2) && subexp_idx == dfa->nodes[node].opr.idx) + return 0; + break; + + default: + break; + } + } + + return (boundaries & 2) ? 1 : 0; +} + +static int +internal_function +check_dst_limits_calc_pos (const re_match_context_t *mctx, Idx limit, + Idx subexp_idx, Idx from_node, Idx str_idx, + Idx bkref_idx) +{ + struct re_backref_cache_entry *lim = mctx->bkref_ents + limit; + int boundaries; + + /* If we are outside the range of the subexpression, return -1 or 1. */ + if (str_idx < lim->subexp_from) + return -1; + + if (lim->subexp_to < str_idx) + return 1; + + /* If we are within the subexpression, return 0. */ + boundaries = (str_idx == lim->subexp_from); + boundaries |= (str_idx == lim->subexp_to) << 1; + if (boundaries == 0) + return 0; + + /* Else, examine epsilon closure. */ + return check_dst_limits_calc_pos_1 (mctx, boundaries, subexp_idx, + from_node, bkref_idx); +} + +/* Check the limitations of sub expressions LIMITS, and remove the nodes + which are against limitations from DEST_NODES. */ + +static reg_errcode_t +internal_function +check_subexp_limits (const re_dfa_t *dfa, re_node_set *dest_nodes, + const re_node_set *candidates, re_node_set *limits, + struct re_backref_cache_entry *bkref_ents, Idx str_idx) +{ + reg_errcode_t err; + Idx node_idx, lim_idx; + + for (lim_idx = 0; lim_idx < limits->nelem; ++lim_idx) + { + Idx subexp_idx; + struct re_backref_cache_entry *ent; + ent = bkref_ents + limits->elems[lim_idx]; + + if (str_idx <= ent->subexp_from || ent->str_idx < str_idx) + continue; /* This is unrelated limitation. */ + + subexp_idx = dfa->nodes[ent->node].opr.idx; + if (ent->subexp_to == str_idx) + { + Idx ops_node = REG_MISSING; + Idx cls_node = REG_MISSING; + for (node_idx = 0; node_idx < dest_nodes->nelem; ++node_idx) + { + Idx node = dest_nodes->elems[node_idx]; + re_token_type_t type = dfa->nodes[node].type; + if (type == OP_OPEN_SUBEXP + && subexp_idx == dfa->nodes[node].opr.idx) + ops_node = node; + else if (type == OP_CLOSE_SUBEXP + && subexp_idx == dfa->nodes[node].opr.idx) + cls_node = node; + } + + /* Check the limitation of the open subexpression. */ + /* Note that (ent->subexp_to = str_idx != ent->subexp_from). */ + if (REG_VALID_INDEX (ops_node)) + { + err = sub_epsilon_src_nodes (dfa, ops_node, dest_nodes, + candidates); + if (BE (err != REG_NOERROR, 0)) + return err; + } + + /* Check the limitation of the close subexpression. */ + if (REG_VALID_INDEX (cls_node)) + for (node_idx = 0; node_idx < dest_nodes->nelem; ++node_idx) + { + Idx node = dest_nodes->elems[node_idx]; + if (!re_node_set_contains (dfa->inveclosures + node, + cls_node) + && !re_node_set_contains (dfa->eclosures + node, + cls_node)) + { + /* It is against this limitation. + Remove it form the current sifted state. */ + err = sub_epsilon_src_nodes (dfa, node, dest_nodes, + candidates); + if (BE (err != REG_NOERROR, 0)) + return err; + --node_idx; + } + } + } + else /* (ent->subexp_to != str_idx) */ + { + for (node_idx = 0; node_idx < dest_nodes->nelem; ++node_idx) + { + Idx node = dest_nodes->elems[node_idx]; + re_token_type_t type = dfa->nodes[node].type; + if (type == OP_CLOSE_SUBEXP || type == OP_OPEN_SUBEXP) + { + if (subexp_idx != dfa->nodes[node].opr.idx) + continue; + /* It is against this limitation. + Remove it form the current sifted state. */ + err = sub_epsilon_src_nodes (dfa, node, dest_nodes, + candidates); + if (BE (err != REG_NOERROR, 0)) + return err; + } + } + } + } + return REG_NOERROR; +} + +static reg_errcode_t +internal_function +sift_states_bkref (const re_match_context_t *mctx, re_sift_context_t *sctx, + Idx str_idx, const re_node_set *candidates) +{ + const re_dfa_t *const dfa = mctx->dfa; + reg_errcode_t err; + Idx node_idx, node; + re_sift_context_t local_sctx; + Idx first_idx = search_cur_bkref_entry (mctx, str_idx); + + if (first_idx == REG_MISSING) + return REG_NOERROR; + + local_sctx.sifted_states = NULL; /* Mark that it hasn't been initialized. */ + + for (node_idx = 0; node_idx < candidates->nelem; ++node_idx) + { + Idx enabled_idx; + re_token_type_t type; + struct re_backref_cache_entry *entry; + node = candidates->elems[node_idx]; + type = dfa->nodes[node].type; + /* Avoid infinite loop for the REs like "()\1+". */ + if (node == sctx->last_node && str_idx == sctx->last_str_idx) + continue; + if (type != OP_BACK_REF) + continue; + + entry = mctx->bkref_ents + first_idx; + enabled_idx = first_idx; + do + { + Idx subexp_len; + Idx to_idx; + Idx dst_node; + bool ok; + re_dfastate_t *cur_state; + + if (entry->node != node) + continue; + subexp_len = entry->subexp_to - entry->subexp_from; + to_idx = str_idx + subexp_len; + dst_node = (subexp_len ? dfa->nexts[node] + : dfa->edests[node].elems[0]); + + if (to_idx > sctx->last_str_idx + || sctx->sifted_states[to_idx] == NULL + || !STATE_NODE_CONTAINS (sctx->sifted_states[to_idx], dst_node) + || check_dst_limits (mctx, &sctx->limits, node, + str_idx, dst_node, to_idx)) + continue; + + if (local_sctx.sifted_states == NULL) + { + local_sctx = *sctx; + err = re_node_set_init_copy (&local_sctx.limits, &sctx->limits); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + } + local_sctx.last_node = node; + local_sctx.last_str_idx = str_idx; + ok = re_node_set_insert (&local_sctx.limits, enabled_idx); + if (BE (! ok, 0)) + { + err = REG_ESPACE; + goto free_return; + } + cur_state = local_sctx.sifted_states[str_idx]; + err = sift_states_backward (mctx, &local_sctx); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + if (sctx->limited_states != NULL) + { + err = merge_state_array (dfa, sctx->limited_states, + local_sctx.sifted_states, + str_idx + 1); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + } + local_sctx.sifted_states[str_idx] = cur_state; + re_node_set_remove (&local_sctx.limits, enabled_idx); + + /* mctx->bkref_ents may have changed, reload the pointer. */ + entry = mctx->bkref_ents + enabled_idx; + } + while (enabled_idx++, entry++->more); + } + err = REG_NOERROR; + free_return: + if (local_sctx.sifted_states != NULL) + { + re_node_set_free (&local_sctx.limits); + } + + return err; +} + + +#ifdef RE_ENABLE_I18N +static int +internal_function +sift_states_iter_mb (const re_match_context_t *mctx, re_sift_context_t *sctx, + Idx node_idx, Idx str_idx, Idx max_str_idx) +{ + const re_dfa_t *const dfa = mctx->dfa; + int naccepted; + /* Check the node can accept `multi byte'. */ + naccepted = check_node_accept_bytes (dfa, node_idx, &mctx->input, str_idx); + if (naccepted > 0 && str_idx + naccepted <= max_str_idx && + !STATE_NODE_CONTAINS (sctx->sifted_states[str_idx + naccepted], + dfa->nexts[node_idx])) + /* The node can't accept the `multi byte', or the + destination was already thrown away, then the node + could't accept the current input `multi byte'. */ + naccepted = 0; + /* Otherwise, it is sure that the node could accept + `naccepted' bytes input. */ + return naccepted; +} +#endif /* RE_ENABLE_I18N */ + + +/* Functions for state transition. */ + +/* Return the next state to which the current state STATE will transit by + accepting the current input byte, and update STATE_LOG if necessary. + If STATE can accept a multibyte char/collating element/back reference + update the destination of STATE_LOG. */ + +static re_dfastate_t * +internal_function +transit_state (reg_errcode_t *err, re_match_context_t *mctx, + re_dfastate_t *state) +{ + re_dfastate_t **trtable; + unsigned char ch; + +#ifdef RE_ENABLE_I18N + /* If the current state can accept multibyte. */ + if (BE (state->accept_mb, 0)) + { + *err = transit_state_mb (mctx, state); + if (BE (*err != REG_NOERROR, 0)) + return NULL; + } +#endif /* RE_ENABLE_I18N */ + + /* Then decide the next state with the single byte. */ +#if 0 + if (0) + /* don't use transition table */ + return transit_state_sb (err, mctx, state); +#endif + + /* Use transition table */ + ch = re_string_fetch_byte (&mctx->input); + for (;;) + { + trtable = state->trtable; + if (BE (trtable != NULL, 1)) + return trtable[ch]; + + trtable = state->word_trtable; + if (BE (trtable != NULL, 1)) + { + unsigned int context; + context + = re_string_context_at (&mctx->input, + re_string_cur_idx (&mctx->input) - 1, + mctx->eflags); + if (IS_WORD_CONTEXT (context)) + return trtable[ch + SBC_MAX]; + else + return trtable[ch]; + } + + if (!build_trtable (mctx->dfa, state)) + { + *err = REG_ESPACE; + return NULL; + } + + /* Retry, we now have a transition table. */ + } +} + +/* Update the state_log if we need */ +re_dfastate_t * +internal_function +merge_state_with_log (reg_errcode_t *err, re_match_context_t *mctx, + re_dfastate_t *next_state) +{ + const re_dfa_t *const dfa = mctx->dfa; + Idx cur_idx = re_string_cur_idx (&mctx->input); + + if (cur_idx > mctx->state_log_top) + { + mctx->state_log[cur_idx] = next_state; + mctx->state_log_top = cur_idx; + } + else if (mctx->state_log[cur_idx] == 0) + { + mctx->state_log[cur_idx] = next_state; + } + else + { + re_dfastate_t *pstate; + unsigned int context; + re_node_set next_nodes, *log_nodes, *table_nodes = NULL; + /* If (state_log[cur_idx] != 0), it implies that cur_idx is + the destination of a multibyte char/collating element/ + back reference. Then the next state is the union set of + these destinations and the results of the transition table. */ + pstate = mctx->state_log[cur_idx]; + log_nodes = pstate->entrance_nodes; + if (next_state != NULL) + { + table_nodes = next_state->entrance_nodes; + *err = re_node_set_init_union (&next_nodes, table_nodes, + log_nodes); + if (BE (*err != REG_NOERROR, 0)) + return NULL; + } + else + next_nodes = *log_nodes; + /* Note: We already add the nodes of the initial state, + then we don't need to add them here. */ + + context = re_string_context_at (&mctx->input, + re_string_cur_idx (&mctx->input) - 1, + mctx->eflags); + next_state = mctx->state_log[cur_idx] + = re_acquire_state_context (err, dfa, &next_nodes, context); + /* We don't need to check errors here, since the return value of + this function is next_state and ERR is already set. */ + + if (table_nodes != NULL) + re_node_set_free (&next_nodes); + } + + if (BE (dfa->nbackref, 0) && next_state != NULL) + { + /* Check OP_OPEN_SUBEXP in the current state in case that we use them + later. We must check them here, since the back references in the + next state might use them. */ + *err = check_subexp_matching_top (mctx, &next_state->nodes, + cur_idx); + if (BE (*err != REG_NOERROR, 0)) + return NULL; + + /* If the next state has back references. */ + if (next_state->has_backref) + { + *err = transit_state_bkref (mctx, &next_state->nodes); + if (BE (*err != REG_NOERROR, 0)) + return NULL; + next_state = mctx->state_log[cur_idx]; + } + } + + return next_state; +} + +/* Skip bytes in the input that correspond to part of a + multi-byte match, then look in the log for a state + from which to restart matching. */ +static re_dfastate_t * +internal_function +find_recover_state (reg_errcode_t *err, re_match_context_t *mctx) +{ + re_dfastate_t *cur_state; + do + { + Idx max = mctx->state_log_top; + Idx cur_str_idx = re_string_cur_idx (&mctx->input); + + do + { + if (++cur_str_idx > max) + return NULL; + re_string_skip_bytes (&mctx->input, 1); + } + while (mctx->state_log[cur_str_idx] == NULL); + + cur_state = merge_state_with_log (err, mctx, NULL); + } + while (*err == REG_NOERROR && cur_state == NULL); + return cur_state; +} + +/* Helper functions for transit_state. */ + +/* From the node set CUR_NODES, pick up the nodes whose types are + OP_OPEN_SUBEXP and which have corresponding back references in the regular + expression. And register them to use them later for evaluating the + correspoding back references. */ + +static reg_errcode_t +internal_function +check_subexp_matching_top (re_match_context_t *mctx, re_node_set *cur_nodes, + Idx str_idx) +{ + const re_dfa_t *const dfa = mctx->dfa; + Idx node_idx; + reg_errcode_t err; + + /* TODO: This isn't efficient. + Because there might be more than one nodes whose types are + OP_OPEN_SUBEXP and whose index is SUBEXP_IDX, we must check all + nodes. + E.g. RE: (a){2} */ + for (node_idx = 0; node_idx < cur_nodes->nelem; ++node_idx) + { + Idx node = cur_nodes->elems[node_idx]; + if (dfa->nodes[node].type == OP_OPEN_SUBEXP + && dfa->nodes[node].opr.idx < BITSET_WORD_BITS + && (dfa->used_bkref_map + & ((bitset_word_t) 1 << dfa->nodes[node].opr.idx))) + { + err = match_ctx_add_subtop (mctx, node, str_idx); + if (BE (err != REG_NOERROR, 0)) + return err; + } + } + return REG_NOERROR; +} + +#if 0 +/* Return the next state to which the current state STATE will transit by + accepting the current input byte. */ + +static re_dfastate_t * +transit_state_sb (reg_errcode_t *err, re_match_context_t *mctx, + re_dfastate_t *state) +{ + const re_dfa_t *const dfa = mctx->dfa; + re_node_set next_nodes; + re_dfastate_t *next_state; + Idx node_cnt, cur_str_idx = re_string_cur_idx (&mctx->input); + unsigned int context; + + *err = re_node_set_alloc (&next_nodes, state->nodes.nelem + 1); + if (BE (*err != REG_NOERROR, 0)) + return NULL; + for (node_cnt = 0; node_cnt < state->nodes.nelem; ++node_cnt) + { + Idx cur_node = state->nodes.elems[node_cnt]; + if (check_node_accept (mctx, dfa->nodes + cur_node, cur_str_idx)) + { + *err = re_node_set_merge (&next_nodes, + dfa->eclosures + dfa->nexts[cur_node]); + if (BE (*err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return NULL; + } + } + } + context = re_string_context_at (&mctx->input, cur_str_idx, mctx->eflags); + next_state = re_acquire_state_context (err, dfa, &next_nodes, context); + /* We don't need to check errors here, since the return value of + this function is next_state and ERR is already set. */ + + re_node_set_free (&next_nodes); + re_string_skip_bytes (&mctx->input, 1); + return next_state; +} +#endif + +#ifdef RE_ENABLE_I18N +static reg_errcode_t +internal_function +transit_state_mb (re_match_context_t *mctx, re_dfastate_t *pstate) +{ + const re_dfa_t *const dfa = mctx->dfa; + reg_errcode_t err; + Idx i; + + for (i = 0; i < pstate->nodes.nelem; ++i) + { + re_node_set dest_nodes, *new_nodes; + Idx cur_node_idx = pstate->nodes.elems[i]; + int naccepted; + Idx dest_idx; + unsigned int context; + re_dfastate_t *dest_state; + + if (!dfa->nodes[cur_node_idx].accept_mb) + continue; + + if (dfa->nodes[cur_node_idx].constraint) + { + context = re_string_context_at (&mctx->input, + re_string_cur_idx (&mctx->input), + mctx->eflags); + if (NOT_SATISFY_NEXT_CONSTRAINT (dfa->nodes[cur_node_idx].constraint, + context)) + continue; + } + + /* How many bytes the node can accept? */ + naccepted = check_node_accept_bytes (dfa, cur_node_idx, &mctx->input, + re_string_cur_idx (&mctx->input)); + if (naccepted == 0) + continue; + + /* The node can accepts `naccepted' bytes. */ + dest_idx = re_string_cur_idx (&mctx->input) + naccepted; + mctx->max_mb_elem_len = ((mctx->max_mb_elem_len < naccepted) ? naccepted + : mctx->max_mb_elem_len); + err = clean_state_log_if_needed (mctx, dest_idx); + if (BE (err != REG_NOERROR, 0)) + return err; +#ifdef DEBUG + assert (dfa->nexts[cur_node_idx] != REG_MISSING); +#endif + new_nodes = dfa->eclosures + dfa->nexts[cur_node_idx]; + + dest_state = mctx->state_log[dest_idx]; + if (dest_state == NULL) + dest_nodes = *new_nodes; + else + { + err = re_node_set_init_union (&dest_nodes, + dest_state->entrance_nodes, new_nodes); + if (BE (err != REG_NOERROR, 0)) + return err; + } + context = re_string_context_at (&mctx->input, dest_idx - 1, + mctx->eflags); + mctx->state_log[dest_idx] + = re_acquire_state_context (&err, dfa, &dest_nodes, context); + if (dest_state != NULL) + re_node_set_free (&dest_nodes); + if (BE (mctx->state_log[dest_idx] == NULL && err != REG_NOERROR, 0)) + return err; + } + return REG_NOERROR; +} +#endif /* RE_ENABLE_I18N */ + +static reg_errcode_t +internal_function +transit_state_bkref (re_match_context_t *mctx, const re_node_set *nodes) +{ + const re_dfa_t *const dfa = mctx->dfa; + reg_errcode_t err; + Idx i; + Idx cur_str_idx = re_string_cur_idx (&mctx->input); + + for (i = 0; i < nodes->nelem; ++i) + { + Idx dest_str_idx, prev_nelem, bkc_idx; + Idx node_idx = nodes->elems[i]; + unsigned int context; + const re_token_t *node = dfa->nodes + node_idx; + re_node_set *new_dest_nodes; + + /* Check whether `node' is a backreference or not. */ + if (node->type != OP_BACK_REF) + continue; + + if (node->constraint) + { + context = re_string_context_at (&mctx->input, cur_str_idx, + mctx->eflags); + if (NOT_SATISFY_NEXT_CONSTRAINT (node->constraint, context)) + continue; + } + + /* `node' is a backreference. + Check the substring which the substring matched. */ + bkc_idx = mctx->nbkref_ents; + err = get_subexp (mctx, node_idx, cur_str_idx); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + + /* And add the epsilon closures (which is `new_dest_nodes') of + the backreference to appropriate state_log. */ +#ifdef DEBUG + assert (dfa->nexts[node_idx] != REG_MISSING); +#endif + for (; bkc_idx < mctx->nbkref_ents; ++bkc_idx) + { + Idx subexp_len; + re_dfastate_t *dest_state; + struct re_backref_cache_entry *bkref_ent; + bkref_ent = mctx->bkref_ents + bkc_idx; + if (bkref_ent->node != node_idx || bkref_ent->str_idx != cur_str_idx) + continue; + subexp_len = bkref_ent->subexp_to - bkref_ent->subexp_from; + new_dest_nodes = (subexp_len == 0 + ? dfa->eclosures + dfa->edests[node_idx].elems[0] + : dfa->eclosures + dfa->nexts[node_idx]); + dest_str_idx = (cur_str_idx + bkref_ent->subexp_to + - bkref_ent->subexp_from); + context = re_string_context_at (&mctx->input, dest_str_idx - 1, + mctx->eflags); + dest_state = mctx->state_log[dest_str_idx]; + prev_nelem = ((mctx->state_log[cur_str_idx] == NULL) ? 0 + : mctx->state_log[cur_str_idx]->nodes.nelem); + /* Add `new_dest_node' to state_log. */ + if (dest_state == NULL) + { + mctx->state_log[dest_str_idx] + = re_acquire_state_context (&err, dfa, new_dest_nodes, + context); + if (BE (mctx->state_log[dest_str_idx] == NULL + && err != REG_NOERROR, 0)) + goto free_return; + } + else + { + re_node_set dest_nodes; + err = re_node_set_init_union (&dest_nodes, + dest_state->entrance_nodes, + new_dest_nodes); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&dest_nodes); + goto free_return; + } + mctx->state_log[dest_str_idx] + = re_acquire_state_context (&err, dfa, &dest_nodes, context); + re_node_set_free (&dest_nodes); + if (BE (mctx->state_log[dest_str_idx] == NULL + && err != REG_NOERROR, 0)) + goto free_return; + } + /* We need to check recursively if the backreference can epsilon + transit. */ + if (subexp_len == 0 + && mctx->state_log[cur_str_idx]->nodes.nelem > prev_nelem) + { + err = check_subexp_matching_top (mctx, new_dest_nodes, + cur_str_idx); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + err = transit_state_bkref (mctx, new_dest_nodes); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + } + } + } + err = REG_NOERROR; + free_return: + return err; +} + +/* Enumerate all the candidates which the backreference BKREF_NODE can match + at BKREF_STR_IDX, and register them by match_ctx_add_entry(). + Note that we might collect inappropriate candidates here. + However, the cost of checking them strictly here is too high, then we + delay these checking for prune_impossible_nodes(). */ + +static reg_errcode_t +internal_function +get_subexp (re_match_context_t *mctx, Idx bkref_node, Idx bkref_str_idx) +{ + const re_dfa_t *const dfa = mctx->dfa; + Idx subexp_num, sub_top_idx; + const char *buf = (const char *) re_string_get_buffer (&mctx->input); + /* Return if we have already checked BKREF_NODE at BKREF_STR_IDX. */ + Idx cache_idx = search_cur_bkref_entry (mctx, bkref_str_idx); + if (cache_idx != REG_MISSING) + { + const struct re_backref_cache_entry *entry + = mctx->bkref_ents + cache_idx; + do + if (entry->node == bkref_node) + return REG_NOERROR; /* We already checked it. */ + while (entry++->more); + } + + subexp_num = dfa->nodes[bkref_node].opr.idx; + + /* For each sub expression */ + for (sub_top_idx = 0; sub_top_idx < mctx->nsub_tops; ++sub_top_idx) + { + reg_errcode_t err; + re_sub_match_top_t *sub_top = mctx->sub_tops[sub_top_idx]; + re_sub_match_last_t *sub_last; + Idx sub_last_idx, sl_str, bkref_str_off; + + if (dfa->nodes[sub_top->node].opr.idx != subexp_num) + continue; /* It isn't related. */ + + sl_str = sub_top->str_idx; + bkref_str_off = bkref_str_idx; + /* At first, check the last node of sub expressions we already + evaluated. */ + for (sub_last_idx = 0; sub_last_idx < sub_top->nlasts; ++sub_last_idx) + { + regoff_t sl_str_diff; + sub_last = sub_top->lasts[sub_last_idx]; + sl_str_diff = sub_last->str_idx - sl_str; + /* The matched string by the sub expression match with the substring + at the back reference? */ + if (sl_str_diff > 0) + { + if (BE (bkref_str_off + sl_str_diff > mctx->input.valid_len, 0)) + { + /* Not enough chars for a successful match. */ + if (bkref_str_off + sl_str_diff > mctx->input.len) + break; + + err = clean_state_log_if_needed (mctx, + bkref_str_off + + sl_str_diff); + if (BE (err != REG_NOERROR, 0)) + return err; + buf = (const char *) re_string_get_buffer (&mctx->input); + } + if (memcmp (buf + bkref_str_off, buf + sl_str, sl_str_diff) != 0) + /* We don't need to search this sub expression any more. */ + break; + } + bkref_str_off += sl_str_diff; + sl_str += sl_str_diff; + err = get_subexp_sub (mctx, sub_top, sub_last, bkref_node, + bkref_str_idx); + + /* Reload buf, since the preceding call might have reallocated + the buffer. */ + buf = (const char *) re_string_get_buffer (&mctx->input); + + if (err == REG_NOMATCH) + continue; + if (BE (err != REG_NOERROR, 0)) + return err; + } + + if (sub_last_idx < sub_top->nlasts) + continue; + if (sub_last_idx > 0) + ++sl_str; + /* Then, search for the other last nodes of the sub expression. */ + for (; sl_str <= bkref_str_idx; ++sl_str) + { + Idx cls_node; + regoff_t sl_str_off; + const re_node_set *nodes; + sl_str_off = sl_str - sub_top->str_idx; + /* The matched string by the sub expression match with the substring + at the back reference? */ + if (sl_str_off > 0) + { + if (BE (bkref_str_off >= mctx->input.valid_len, 0)) + { + /* If we are at the end of the input, we cannot match. */ + if (bkref_str_off >= mctx->input.len) + break; + + err = extend_buffers (mctx); + if (BE (err != REG_NOERROR, 0)) + return err; + + buf = (const char *) re_string_get_buffer (&mctx->input); + } + if (buf [bkref_str_off++] != buf[sl_str - 1]) + break; /* We don't need to search this sub expression + any more. */ + } + if (mctx->state_log[sl_str] == NULL) + continue; + /* Does this state have a ')' of the sub expression? */ + nodes = &mctx->state_log[sl_str]->nodes; + cls_node = find_subexp_node (dfa, nodes, subexp_num, + OP_CLOSE_SUBEXP); + if (cls_node == REG_MISSING) + continue; /* No. */ + if (sub_top->path == NULL) + { + sub_top->path = calloc (sizeof (state_array_t), + sl_str - sub_top->str_idx + 1); + if (sub_top->path == NULL) + return REG_ESPACE; + } + /* Can the OP_OPEN_SUBEXP node arrive the OP_CLOSE_SUBEXP node + in the current context? */ + err = check_arrival (mctx, sub_top->path, sub_top->node, + sub_top->str_idx, cls_node, sl_str, + OP_CLOSE_SUBEXP); + if (err == REG_NOMATCH) + continue; + if (BE (err != REG_NOERROR, 0)) + return err; + sub_last = match_ctx_add_sublast (sub_top, cls_node, sl_str); + if (BE (sub_last == NULL, 0)) + return REG_ESPACE; + err = get_subexp_sub (mctx, sub_top, sub_last, bkref_node, + bkref_str_idx); + if (err == REG_NOMATCH) + continue; + } + } + return REG_NOERROR; +} + +/* Helper functions for get_subexp(). */ + +/* Check SUB_LAST can arrive to the back reference BKREF_NODE at BKREF_STR. + If it can arrive, register the sub expression expressed with SUB_TOP + and SUB_LAST. */ + +static reg_errcode_t +internal_function +get_subexp_sub (re_match_context_t *mctx, const re_sub_match_top_t *sub_top, + re_sub_match_last_t *sub_last, Idx bkref_node, Idx bkref_str) +{ + reg_errcode_t err; + Idx to_idx; + /* Can the subexpression arrive the back reference? */ + err = check_arrival (mctx, &sub_last->path, sub_last->node, + sub_last->str_idx, bkref_node, bkref_str, + OP_OPEN_SUBEXP); + if (err != REG_NOERROR) + return err; + err = match_ctx_add_entry (mctx, bkref_node, bkref_str, sub_top->str_idx, + sub_last->str_idx); + if (BE (err != REG_NOERROR, 0)) + return err; + to_idx = bkref_str + sub_last->str_idx - sub_top->str_idx; + return clean_state_log_if_needed (mctx, to_idx); +} + +/* Find the first node which is '(' or ')' and whose index is SUBEXP_IDX. + Search '(' if FL_OPEN, or search ')' otherwise. + TODO: This function isn't efficient... + Because there might be more than one nodes whose types are + OP_OPEN_SUBEXP and whose index is SUBEXP_IDX, we must check all + nodes. + E.g. RE: (a){2} */ + +static Idx +internal_function +find_subexp_node (const re_dfa_t *dfa, const re_node_set *nodes, + Idx subexp_idx, int type) +{ + Idx cls_idx; + for (cls_idx = 0; cls_idx < nodes->nelem; ++cls_idx) + { + Idx cls_node = nodes->elems[cls_idx]; + const re_token_t *node = dfa->nodes + cls_node; + if (node->type == type + && node->opr.idx == subexp_idx) + return cls_node; + } + return REG_MISSING; +} + +/* Check whether the node TOP_NODE at TOP_STR can arrive to the node + LAST_NODE at LAST_STR. We record the path onto PATH since it will be + heavily reused. + Return REG_NOERROR if it can arrive, or REG_NOMATCH otherwise. */ + +static reg_errcode_t +internal_function +check_arrival (re_match_context_t *mctx, state_array_t *path, Idx top_node, + Idx top_str, Idx last_node, Idx last_str, int type) +{ + const re_dfa_t *const dfa = mctx->dfa; + reg_errcode_t err = REG_NOERROR; + Idx subexp_num, backup_cur_idx, str_idx, null_cnt; + re_dfastate_t *cur_state = NULL; + re_node_set *cur_nodes, next_nodes; + re_dfastate_t **backup_state_log; + unsigned int context; + + subexp_num = dfa->nodes[top_node].opr.idx; + /* Extend the buffer if we need. */ + if (BE (path->alloc < last_str + mctx->max_mb_elem_len + 1, 0)) + { + re_dfastate_t **new_array; + Idx old_alloc = path->alloc; + Idx new_alloc = old_alloc + last_str + mctx->max_mb_elem_len + 1; + if (BE (new_alloc < old_alloc, 0) + || BE (SIZE_MAX / sizeof (re_dfastate_t *) < new_alloc, 0)) + return REG_ESPACE; + new_array = re_realloc (path->array, re_dfastate_t *, new_alloc); + if (BE (new_array == NULL, 0)) + return REG_ESPACE; + path->array = new_array; + path->alloc = new_alloc; + memset (new_array + old_alloc, '\0', + sizeof (re_dfastate_t *) * (path->alloc - old_alloc)); + } + + str_idx = path->next_idx ? path->next_idx : top_str; + + /* Temporary modify MCTX. */ + backup_state_log = mctx->state_log; + backup_cur_idx = mctx->input.cur_idx; + mctx->state_log = path->array; + mctx->input.cur_idx = str_idx; + + /* Setup initial node set. */ + context = re_string_context_at (&mctx->input, str_idx - 1, mctx->eflags); + if (str_idx == top_str) + { + err = re_node_set_init_1 (&next_nodes, top_node); + if (BE (err != REG_NOERROR, 0)) + return err; + err = check_arrival_expand_ecl (dfa, &next_nodes, subexp_num, type); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return err; + } + } + else + { + cur_state = mctx->state_log[str_idx]; + if (cur_state && cur_state->has_backref) + { + err = re_node_set_init_copy (&next_nodes, &cur_state->nodes); + if (BE (err != REG_NOERROR, 0)) + return err; + } + else + re_node_set_init_empty (&next_nodes); + } + if (str_idx == top_str || (cur_state && cur_state->has_backref)) + { + if (next_nodes.nelem) + { + err = expand_bkref_cache (mctx, &next_nodes, str_idx, + subexp_num, type); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return err; + } + } + cur_state = re_acquire_state_context (&err, dfa, &next_nodes, context); + if (BE (cur_state == NULL && err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return err; + } + mctx->state_log[str_idx] = cur_state; + } + + for (null_cnt = 0; str_idx < last_str && null_cnt <= mctx->max_mb_elem_len;) + { + re_node_set_empty (&next_nodes); + if (mctx->state_log[str_idx + 1]) + { + err = re_node_set_merge (&next_nodes, + &mctx->state_log[str_idx + 1]->nodes); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return err; + } + } + if (cur_state) + { + err = check_arrival_add_next_nodes (mctx, str_idx, + &cur_state->non_eps_nodes, + &next_nodes); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return err; + } + } + ++str_idx; + if (next_nodes.nelem) + { + err = check_arrival_expand_ecl (dfa, &next_nodes, subexp_num, type); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return err; + } + err = expand_bkref_cache (mctx, &next_nodes, str_idx, + subexp_num, type); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return err; + } + } + context = re_string_context_at (&mctx->input, str_idx - 1, mctx->eflags); + cur_state = re_acquire_state_context (&err, dfa, &next_nodes, context); + if (BE (cur_state == NULL && err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return err; + } + mctx->state_log[str_idx] = cur_state; + null_cnt = cur_state == NULL ? null_cnt + 1 : 0; + } + re_node_set_free (&next_nodes); + cur_nodes = (mctx->state_log[last_str] == NULL ? NULL + : &mctx->state_log[last_str]->nodes); + path->next_idx = str_idx; + + /* Fix MCTX. */ + mctx->state_log = backup_state_log; + mctx->input.cur_idx = backup_cur_idx; + + /* Then check the current node set has the node LAST_NODE. */ + if (cur_nodes != NULL && re_node_set_contains (cur_nodes, last_node)) + return REG_NOERROR; + + return REG_NOMATCH; +} + +/* Helper functions for check_arrival. */ + +/* Calculate the destination nodes of CUR_NODES at STR_IDX, and append them + to NEXT_NODES. + TODO: This function is similar to the functions transit_state*(), + however this function has many additional works. + Can't we unify them? */ + +static reg_errcode_t +internal_function +check_arrival_add_next_nodes (re_match_context_t *mctx, Idx str_idx, + re_node_set *cur_nodes, re_node_set *next_nodes) +{ + const re_dfa_t *const dfa = mctx->dfa; + bool ok; + Idx cur_idx; + reg_errcode_t err = REG_NOERROR; + re_node_set union_set; + re_node_set_init_empty (&union_set); + for (cur_idx = 0; cur_idx < cur_nodes->nelem; ++cur_idx) + { + int naccepted = 0; + Idx cur_node = cur_nodes->elems[cur_idx]; +#ifdef DEBUG + re_token_type_t type = dfa->nodes[cur_node].type; + assert (!IS_EPSILON_NODE (type)); +#endif +#ifdef RE_ENABLE_I18N + /* If the node may accept `multi byte'. */ + if (dfa->nodes[cur_node].accept_mb) + { + naccepted = check_node_accept_bytes (dfa, cur_node, &mctx->input, + str_idx); + if (naccepted > 1) + { + re_dfastate_t *dest_state; + Idx next_node = dfa->nexts[cur_node]; + Idx next_idx = str_idx + naccepted; + dest_state = mctx->state_log[next_idx]; + re_node_set_empty (&union_set); + if (dest_state) + { + err = re_node_set_merge (&union_set, &dest_state->nodes); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&union_set); + return err; + } + } + ok = re_node_set_insert (&union_set, next_node); + if (BE (! ok, 0)) + { + re_node_set_free (&union_set); + return REG_ESPACE; + } + mctx->state_log[next_idx] = re_acquire_state (&err, dfa, + &union_set); + if (BE (mctx->state_log[next_idx] == NULL + && err != REG_NOERROR, 0)) + { + re_node_set_free (&union_set); + return err; + } + } + } +#endif /* RE_ENABLE_I18N */ + if (naccepted + || check_node_accept (mctx, dfa->nodes + cur_node, str_idx)) + { + ok = re_node_set_insert (next_nodes, dfa->nexts[cur_node]); + if (BE (! ok, 0)) + { + re_node_set_free (&union_set); + return REG_ESPACE; + } + } + } + re_node_set_free (&union_set); + return REG_NOERROR; +} + +/* For all the nodes in CUR_NODES, add the epsilon closures of them to + CUR_NODES, however exclude the nodes which are: + - inside the sub expression whose number is EX_SUBEXP, if FL_OPEN. + - out of the sub expression whose number is EX_SUBEXP, if !FL_OPEN. +*/ + +static reg_errcode_t +internal_function +check_arrival_expand_ecl (const re_dfa_t *dfa, re_node_set *cur_nodes, + Idx ex_subexp, int type) +{ + reg_errcode_t err; + Idx idx, outside_node; + re_node_set new_nodes; +#ifdef DEBUG + assert (cur_nodes->nelem); +#endif + err = re_node_set_alloc (&new_nodes, cur_nodes->nelem); + if (BE (err != REG_NOERROR, 0)) + return err; + /* Create a new node set NEW_NODES with the nodes which are epsilon + closures of the node in CUR_NODES. */ + + for (idx = 0; idx < cur_nodes->nelem; ++idx) + { + Idx cur_node = cur_nodes->elems[idx]; + const re_node_set *eclosure = dfa->eclosures + cur_node; + outside_node = find_subexp_node (dfa, eclosure, ex_subexp, type); + if (outside_node == REG_MISSING) + { + /* There are no problematic nodes, just merge them. */ + err = re_node_set_merge (&new_nodes, eclosure); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&new_nodes); + return err; + } + } + else + { + /* There are problematic nodes, re-calculate incrementally. */ + err = check_arrival_expand_ecl_sub (dfa, &new_nodes, cur_node, + ex_subexp, type); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&new_nodes); + return err; + } + } + } + re_node_set_free (cur_nodes); + *cur_nodes = new_nodes; + return REG_NOERROR; +} + +/* Helper function for check_arrival_expand_ecl. + Check incrementally the epsilon closure of TARGET, and if it isn't + problematic append it to DST_NODES. */ + +static reg_errcode_t +internal_function +check_arrival_expand_ecl_sub (const re_dfa_t *dfa, re_node_set *dst_nodes, + Idx target, Idx ex_subexp, int type) +{ + Idx cur_node; + for (cur_node = target; !re_node_set_contains (dst_nodes, cur_node);) + { + bool ok; + + if (dfa->nodes[cur_node].type == type + && dfa->nodes[cur_node].opr.idx == ex_subexp) + { + if (type == OP_CLOSE_SUBEXP) + { + ok = re_node_set_insert (dst_nodes, cur_node); + if (BE (! ok, 0)) + return REG_ESPACE; + } + break; + } + ok = re_node_set_insert (dst_nodes, cur_node); + if (BE (! ok, 0)) + return REG_ESPACE; + if (dfa->edests[cur_node].nelem == 0) + break; + if (dfa->edests[cur_node].nelem == 2) + { + reg_errcode_t err; + err = check_arrival_expand_ecl_sub (dfa, dst_nodes, + dfa->edests[cur_node].elems[1], + ex_subexp, type); + if (BE (err != REG_NOERROR, 0)) + return err; + } + cur_node = dfa->edests[cur_node].elems[0]; + } + return REG_NOERROR; +} + + +/* For all the back references in the current state, calculate the + destination of the back references by the appropriate entry + in MCTX->BKREF_ENTS. */ + +static reg_errcode_t +internal_function +expand_bkref_cache (re_match_context_t *mctx, re_node_set *cur_nodes, + Idx cur_str, Idx subexp_num, int type) +{ + const re_dfa_t *const dfa = mctx->dfa; + reg_errcode_t err; + Idx cache_idx_start = search_cur_bkref_entry (mctx, cur_str); + struct re_backref_cache_entry *ent; + + if (cache_idx_start == REG_MISSING) + return REG_NOERROR; + + restart: + ent = mctx->bkref_ents + cache_idx_start; + do + { + Idx to_idx, next_node; + + /* Is this entry ENT is appropriate? */ + if (!re_node_set_contains (cur_nodes, ent->node)) + continue; /* No. */ + + to_idx = cur_str + ent->subexp_to - ent->subexp_from; + /* Calculate the destination of the back reference, and append it + to MCTX->STATE_LOG. */ + if (to_idx == cur_str) + { + /* The backreference did epsilon transit, we must re-check all the + node in the current state. */ + re_node_set new_dests; + reg_errcode_t err2, err3; + next_node = dfa->edests[ent->node].elems[0]; + if (re_node_set_contains (cur_nodes, next_node)) + continue; + err = re_node_set_init_1 (&new_dests, next_node); + err2 = check_arrival_expand_ecl (dfa, &new_dests, subexp_num, type); + err3 = re_node_set_merge (cur_nodes, &new_dests); + re_node_set_free (&new_dests); + if (BE (err != REG_NOERROR || err2 != REG_NOERROR + || err3 != REG_NOERROR, 0)) + { + err = (err != REG_NOERROR ? err + : (err2 != REG_NOERROR ? err2 : err3)); + return err; + } + /* TODO: It is still inefficient... */ + goto restart; + } + else + { + re_node_set union_set; + next_node = dfa->nexts[ent->node]; + if (mctx->state_log[to_idx]) + { + bool ok; + if (re_node_set_contains (&mctx->state_log[to_idx]->nodes, + next_node)) + continue; + err = re_node_set_init_copy (&union_set, + &mctx->state_log[to_idx]->nodes); + ok = re_node_set_insert (&union_set, next_node); + if (BE (err != REG_NOERROR || ! ok, 0)) + { + re_node_set_free (&union_set); + err = err != REG_NOERROR ? err : REG_ESPACE; + return err; + } + } + else + { + err = re_node_set_init_1 (&union_set, next_node); + if (BE (err != REG_NOERROR, 0)) + return err; + } + mctx->state_log[to_idx] = re_acquire_state (&err, dfa, &union_set); + re_node_set_free (&union_set); + if (BE (mctx->state_log[to_idx] == NULL + && err != REG_NOERROR, 0)) + return err; + } + } + while (ent++->more); + return REG_NOERROR; +} + +/* Build transition table for the state. + Return true if successful. */ + +static bool +internal_function +build_trtable (const re_dfa_t *dfa, re_dfastate_t *state) +{ + reg_errcode_t err; + Idx i, j; + int ch; + bool need_word_trtable = false; + bitset_word_t elem, mask; + bool dests_node_malloced = false; + bool dest_states_malloced = false; + Idx ndests; /* Number of the destination states from `state'. */ + re_dfastate_t **trtable; + re_dfastate_t **dest_states = NULL, **dest_states_word, **dest_states_nl; + re_node_set follows, *dests_node; + bitset_t *dests_ch; + bitset_t acceptable; + + struct dests_alloc + { + re_node_set dests_node[SBC_MAX]; + bitset_t dests_ch[SBC_MAX]; + } *dests_alloc; + + /* We build DFA states which corresponds to the destination nodes + from `state'. `dests_node[i]' represents the nodes which i-th + destination state contains, and `dests_ch[i]' represents the + characters which i-th destination state accepts. */ + if (__libc_use_alloca (sizeof (struct dests_alloc))) + dests_alloc = (struct dests_alloc *) alloca (sizeof (struct dests_alloc)); + else + { + dests_alloc = re_malloc (struct dests_alloc, 1); + if (BE (dests_alloc == NULL, 0)) + return false; + dests_node_malloced = true; + } + dests_node = dests_alloc->dests_node; + dests_ch = dests_alloc->dests_ch; + + /* Initialize transiton table. */ + state->word_trtable = state->trtable = NULL; + + /* At first, group all nodes belonging to `state' into several + destinations. */ + ndests = group_nodes_into_DFAstates (dfa, state, dests_node, dests_ch); + if (BE (! REG_VALID_NONZERO_INDEX (ndests), 0)) + { + if (dests_node_malloced) + free (dests_alloc); + if (ndests == 0) + { + state->trtable = (re_dfastate_t **) + calloc (sizeof (re_dfastate_t *), SBC_MAX); + return true; + } + return false; + } + + err = re_node_set_alloc (&follows, ndests + 1); + if (BE (err != REG_NOERROR, 0)) + goto out_free; + + /* Avoid arithmetic overflow in size calculation. */ + if (BE ((((SIZE_MAX - (sizeof (re_node_set) + sizeof (bitset_t)) * SBC_MAX) + / (3 * sizeof (re_dfastate_t *))) + < ndests), + 0)) + goto out_free; + + if (__libc_use_alloca ((sizeof (re_node_set) + sizeof (bitset_t)) * SBC_MAX + + ndests * 3 * sizeof (re_dfastate_t *))) + dest_states = (re_dfastate_t **) + alloca (ndests * 3 * sizeof (re_dfastate_t *)); + else + { + dest_states = (re_dfastate_t **) + malloc (ndests * 3 * sizeof (re_dfastate_t *)); + if (BE (dest_states == NULL, 0)) + { +out_free: + if (dest_states_malloced) + free (dest_states); + re_node_set_free (&follows); + for (i = 0; i < ndests; ++i) + re_node_set_free (dests_node + i); + if (dests_node_malloced) + free (dests_alloc); + return false; + } + dest_states_malloced = true; + } + dest_states_word = dest_states + ndests; + dest_states_nl = dest_states_word + ndests; + bitset_empty (acceptable); + + /* Then build the states for all destinations. */ + for (i = 0; i < ndests; ++i) + { + Idx next_node; + re_node_set_empty (&follows); + /* Merge the follows of this destination states. */ + for (j = 0; j < dests_node[i].nelem; ++j) + { + next_node = dfa->nexts[dests_node[i].elems[j]]; + if (next_node != REG_MISSING) + { + err = re_node_set_merge (&follows, dfa->eclosures + next_node); + if (BE (err != REG_NOERROR, 0)) + goto out_free; + } + } + dest_states[i] = re_acquire_state_context (&err, dfa, &follows, 0); + if (BE (dest_states[i] == NULL && err != REG_NOERROR, 0)) + goto out_free; + /* If the new state has context constraint, + build appropriate states for these contexts. */ + if (dest_states[i]->has_constraint) + { + dest_states_word[i] = re_acquire_state_context (&err, dfa, &follows, + CONTEXT_WORD); + if (BE (dest_states_word[i] == NULL && err != REG_NOERROR, 0)) + goto out_free; + + if (dest_states[i] != dest_states_word[i] && dfa->mb_cur_max > 1) + need_word_trtable = true; + + dest_states_nl[i] = re_acquire_state_context (&err, dfa, &follows, + CONTEXT_NEWLINE); + if (BE (dest_states_nl[i] == NULL && err != REG_NOERROR, 0)) + goto out_free; + } + else + { + dest_states_word[i] = dest_states[i]; + dest_states_nl[i] = dest_states[i]; + } + bitset_merge (acceptable, dests_ch[i]); + } + + if (!BE (need_word_trtable, 0)) + { + /* We don't care about whether the following character is a word + character, or we are in a single-byte character set so we can + discern by looking at the character code: allocate a + 256-entry transition table. */ + trtable = state->trtable = + (re_dfastate_t **) calloc (sizeof (re_dfastate_t *), SBC_MAX); + if (BE (trtable == NULL, 0)) + goto out_free; + + /* For all characters ch...: */ + for (i = 0; i < BITSET_WORDS; ++i) + for (ch = i * BITSET_WORD_BITS, elem = acceptable[i], mask = 1; + elem; + mask <<= 1, elem >>= 1, ++ch) + if (BE (elem & 1, 0)) + { + /* There must be exactly one destination which accepts + character ch. See group_nodes_into_DFAstates. */ + for (j = 0; (dests_ch[j][i] & mask) == 0; ++j) + ; + + /* j-th destination accepts the word character ch. */ + if (dfa->word_char[i] & mask) + trtable[ch] = dest_states_word[j]; + else + trtable[ch] = dest_states[j]; + } + } + else + { + /* We care about whether the following character is a word + character, and we are in a multi-byte character set: discern + by looking at the character code: build two 256-entry + transition tables, one starting at trtable[0] and one + starting at trtable[SBC_MAX]. */ + trtable = state->word_trtable = + (re_dfastate_t **) calloc (sizeof (re_dfastate_t *), 2 * SBC_MAX); + if (BE (trtable == NULL, 0)) + goto out_free; + + /* For all characters ch...: */ + for (i = 0; i < BITSET_WORDS; ++i) + for (ch = i * BITSET_WORD_BITS, elem = acceptable[i], mask = 1; + elem; + mask <<= 1, elem >>= 1, ++ch) + if (BE (elem & 1, 0)) + { + /* There must be exactly one destination which accepts + character ch. See group_nodes_into_DFAstates. */ + for (j = 0; (dests_ch[j][i] & mask) == 0; ++j) + ; + + /* j-th destination accepts the word character ch. */ + trtable[ch] = dest_states[j]; + trtable[ch + SBC_MAX] = dest_states_word[j]; + } + } + + /* new line */ + if (bitset_contain (acceptable, NEWLINE_CHAR)) + { + /* The current state accepts newline character. */ + for (j = 0; j < ndests; ++j) + if (bitset_contain (dests_ch[j], NEWLINE_CHAR)) + { + /* k-th destination accepts newline character. */ + trtable[NEWLINE_CHAR] = dest_states_nl[j]; + if (need_word_trtable) + trtable[NEWLINE_CHAR + SBC_MAX] = dest_states_nl[j]; + /* There must be only one destination which accepts + newline. See group_nodes_into_DFAstates. */ + break; + } + } + + if (dest_states_malloced) + free (dest_states); + + re_node_set_free (&follows); + for (i = 0; i < ndests; ++i) + re_node_set_free (dests_node + i); + + if (dests_node_malloced) + free (dests_alloc); + + return true; +} + +/* Group all nodes belonging to STATE into several destinations. + Then for all destinations, set the nodes belonging to the destination + to DESTS_NODE[i] and set the characters accepted by the destination + to DEST_CH[i]. This function return the number of destinations. */ + +static Idx +internal_function +group_nodes_into_DFAstates (const re_dfa_t *dfa, const re_dfastate_t *state, + re_node_set *dests_node, bitset_t *dests_ch) +{ + reg_errcode_t err; + bool ok; + Idx i, j, k; + Idx ndests; /* Number of the destinations from `state'. */ + bitset_t accepts; /* Characters a node can accept. */ + const re_node_set *cur_nodes = &state->nodes; + bitset_empty (accepts); + ndests = 0; + + /* For all the nodes belonging to `state', */ + for (i = 0; i < cur_nodes->nelem; ++i) + { + re_token_t *node = &dfa->nodes[cur_nodes->elems[i]]; + re_token_type_t type = node->type; + unsigned int constraint = node->constraint; + + /* Enumerate all single byte character this node can accept. */ + if (type == CHARACTER) + bitset_set (accepts, node->opr.c); + else if (type == SIMPLE_BRACKET) + { + bitset_merge (accepts, node->opr.sbcset); + } + else if (type == OP_PERIOD) + { +#ifdef RE_ENABLE_I18N + if (dfa->mb_cur_max > 1) + bitset_merge (accepts, dfa->sb_char); + else +#endif + bitset_set_all (accepts); + if (!(dfa->syntax & RE_DOT_NEWLINE)) + bitset_clear (accepts, '\n'); + if (dfa->syntax & RE_DOT_NOT_NULL) + bitset_clear (accepts, '\0'); + } +#ifdef RE_ENABLE_I18N + else if (type == OP_UTF8_PERIOD) + { + if (ASCII_CHARS % BITSET_WORD_BITS == 0) + memset (accepts, -1, ASCII_CHARS / CHAR_BIT); + else + bitset_merge (accepts, utf8_sb_map); + if (!(dfa->syntax & RE_DOT_NEWLINE)) + bitset_clear (accepts, '\n'); + if (dfa->syntax & RE_DOT_NOT_NULL) + bitset_clear (accepts, '\0'); + } +#endif + else + continue; + + /* Check the `accepts' and sift the characters which are not + match it the context. */ + if (constraint) + { + if (constraint & NEXT_NEWLINE_CONSTRAINT) + { + bool accepts_newline = bitset_contain (accepts, NEWLINE_CHAR); + bitset_empty (accepts); + if (accepts_newline) + bitset_set (accepts, NEWLINE_CHAR); + else + continue; + } + if (constraint & NEXT_ENDBUF_CONSTRAINT) + { + bitset_empty (accepts); + continue; + } + + if (constraint & NEXT_WORD_CONSTRAINT) + { + bitset_word_t any_set = 0; + if (type == CHARACTER && !node->word_char) + { + bitset_empty (accepts); + continue; + } +#ifdef RE_ENABLE_I18N + if (dfa->mb_cur_max > 1) + for (j = 0; j < BITSET_WORDS; ++j) + any_set |= (accepts[j] &= (dfa->word_char[j] | ~dfa->sb_char[j])); + else +#endif + for (j = 0; j < BITSET_WORDS; ++j) + any_set |= (accepts[j] &= dfa->word_char[j]); + if (!any_set) + continue; + } + if (constraint & NEXT_NOTWORD_CONSTRAINT) + { + bitset_word_t any_set = 0; + if (type == CHARACTER && node->word_char) + { + bitset_empty (accepts); + continue; + } +#ifdef RE_ENABLE_I18N + if (dfa->mb_cur_max > 1) + for (j = 0; j < BITSET_WORDS; ++j) + any_set |= (accepts[j] &= ~(dfa->word_char[j] & dfa->sb_char[j])); + else +#endif + for (j = 0; j < BITSET_WORDS; ++j) + any_set |= (accepts[j] &= ~dfa->word_char[j]); + if (!any_set) + continue; + } + } + + /* Then divide `accepts' into DFA states, or create a new + state. Above, we make sure that accepts is not empty. */ + for (j = 0; j < ndests; ++j) + { + bitset_t intersec; /* Intersection sets, see below. */ + bitset_t remains; + /* Flags, see below. */ + bitset_word_t has_intersec, not_subset, not_consumed; + + /* Optimization, skip if this state doesn't accept the character. */ + if (type == CHARACTER && !bitset_contain (dests_ch[j], node->opr.c)) + continue; + + /* Enumerate the intersection set of this state and `accepts'. */ + has_intersec = 0; + for (k = 0; k < BITSET_WORDS; ++k) + has_intersec |= intersec[k] = accepts[k] & dests_ch[j][k]; + /* And skip if the intersection set is empty. */ + if (!has_intersec) + continue; + + /* Then check if this state is a subset of `accepts'. */ + not_subset = not_consumed = 0; + for (k = 0; k < BITSET_WORDS; ++k) + { + not_subset |= remains[k] = ~accepts[k] & dests_ch[j][k]; + not_consumed |= accepts[k] = accepts[k] & ~dests_ch[j][k]; + } + + /* If this state isn't a subset of `accepts', create a + new group state, which has the `remains'. */ + if (not_subset) + { + bitset_copy (dests_ch[ndests], remains); + bitset_copy (dests_ch[j], intersec); + err = re_node_set_init_copy (dests_node + ndests, &dests_node[j]); + if (BE (err != REG_NOERROR, 0)) + goto error_return; + ++ndests; + } + + /* Put the position in the current group. */ + ok = re_node_set_insert (&dests_node[j], cur_nodes->elems[i]); + if (BE (! ok, 0)) + goto error_return; + + /* If all characters are consumed, go to next node. */ + if (!not_consumed) + break; + } + /* Some characters remain, create a new group. */ + if (j == ndests) + { + bitset_copy (dests_ch[ndests], accepts); + err = re_node_set_init_1 (dests_node + ndests, cur_nodes->elems[i]); + if (BE (err != REG_NOERROR, 0)) + goto error_return; + ++ndests; + bitset_empty (accepts); + } + } + return ndests; + error_return: + for (j = 0; j < ndests; ++j) + re_node_set_free (dests_node + j); + return REG_MISSING; +} + +#ifdef RE_ENABLE_I18N +/* Check how many bytes the node `dfa->nodes[node_idx]' accepts. + Return the number of the bytes the node accepts. + STR_IDX is the current index of the input string. + + This function handles the nodes which can accept one character, or + one collating element like '.', '[a-z]', opposite to the other nodes + can only accept one byte. */ + +static int +internal_function +check_node_accept_bytes (const re_dfa_t *dfa, Idx node_idx, + const re_string_t *input, Idx str_idx) +{ + const re_token_t *node = dfa->nodes + node_idx; + int char_len, elem_len; + Idx i; + + if (BE (node->type == OP_UTF8_PERIOD, 0)) + { + unsigned char c = re_string_byte_at (input, str_idx), d; + if (BE (c < 0xc2, 1)) + return 0; + + if (str_idx + 2 > input->len) + return 0; + + d = re_string_byte_at (input, str_idx + 1); + if (c < 0xe0) + return (d < 0x80 || d > 0xbf) ? 0 : 2; + else if (c < 0xf0) + { + char_len = 3; + if (c == 0xe0 && d < 0xa0) + return 0; + } + else if (c < 0xf8) + { + char_len = 4; + if (c == 0xf0 && d < 0x90) + return 0; + } + else if (c < 0xfc) + { + char_len = 5; + if (c == 0xf8 && d < 0x88) + return 0; + } + else if (c < 0xfe) + { + char_len = 6; + if (c == 0xfc && d < 0x84) + return 0; + } + else + return 0; + + if (str_idx + char_len > input->len) + return 0; + + for (i = 1; i < char_len; ++i) + { + d = re_string_byte_at (input, str_idx + i); + if (d < 0x80 || d > 0xbf) + return 0; + } + return char_len; + } + + char_len = re_string_char_size_at (input, str_idx); + if (node->type == OP_PERIOD) + { + if (char_len <= 1) + return 0; + /* FIXME: I don't think this if is needed, as both '\n' + and '\0' are char_len == 1. */ + /* '.' accepts any one character except the following two cases. */ + if ((!(dfa->syntax & RE_DOT_NEWLINE) && + re_string_byte_at (input, str_idx) == '\n') || + ((dfa->syntax & RE_DOT_NOT_NULL) && + re_string_byte_at (input, str_idx) == '\0')) + return 0; + return char_len; + } + + elem_len = re_string_elem_size_at (input, str_idx); + if ((elem_len <= 1 && char_len <= 1) || char_len == 0) + return 0; + + if (node->type == COMPLEX_BRACKET) + { + const re_charset_t *cset = node->opr.mbcset; +# ifdef _LIBC + const unsigned char *pin + = ((const unsigned char *) re_string_get_buffer (input) + str_idx); + Idx j; + uint32_t nrules; +# endif /* _LIBC */ + int match_len = 0; + wchar_t wc = ((cset->nranges || cset->nchar_classes || cset->nmbchars) + ? re_string_wchar_at (input, str_idx) : 0); + + /* match with multibyte character? */ + for (i = 0; i < cset->nmbchars; ++i) + if (wc == cset->mbchars[i]) + { + match_len = char_len; + goto check_node_accept_bytes_match; + } + /* match with character_class? */ + for (i = 0; i < cset->nchar_classes; ++i) + { + wctype_t wt = cset->char_classes[i]; + if (__iswctype (wc, wt)) + { + match_len = char_len; + goto check_node_accept_bytes_match; + } + } + +# ifdef _LIBC + nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + if (nrules != 0) + { + unsigned int in_collseq = 0; + const int32_t *table, *indirect; + const unsigned char *weights, *extra; + const char *collseqwc; + int32_t idx; + /* This #include defines a local function! */ +# include <locale/weight.h> + + /* match with collating_symbol? */ + if (cset->ncoll_syms) + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_SYMB_EXTRAMB); + for (i = 0; i < cset->ncoll_syms; ++i) + { + const unsigned char *coll_sym = extra + cset->coll_syms[i]; + /* Compare the length of input collating element and + the length of current collating element. */ + if (*coll_sym != elem_len) + continue; + /* Compare each bytes. */ + for (j = 0; j < *coll_sym; j++) + if (pin[j] != coll_sym[1 + j]) + break; + if (j == *coll_sym) + { + /* Match if every bytes is equal. */ + match_len = j; + goto check_node_accept_bytes_match; + } + } + + if (cset->nranges) + { + if (elem_len <= char_len) + { + collseqwc = _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQWC); + in_collseq = __collseq_table_lookup (collseqwc, wc); + } + else + in_collseq = find_collation_sequence_value (pin, elem_len); + } + /* match with range expression? */ + for (i = 0; i < cset->nranges; ++i) + if (cset->range_starts[i] <= in_collseq + && in_collseq <= cset->range_ends[i]) + { + match_len = elem_len; + goto check_node_accept_bytes_match; + } + + /* match with equivalence_class? */ + if (cset->nequiv_classes) + { + const unsigned char *cp = pin; + table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB); + weights = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB); + indirect = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTMB); + idx = findidx (&cp); + if (idx > 0) + for (i = 0; i < cset->nequiv_classes; ++i) + { + int32_t equiv_class_idx = cset->equiv_classes[i]; + size_t weight_len = weights[idx]; + if (weight_len == weights[equiv_class_idx]) + { + Idx cnt = 0; + while (cnt <= weight_len + && (weights[equiv_class_idx + 1 + cnt] + == weights[idx + 1 + cnt])) + ++cnt; + if (cnt > weight_len) + { + match_len = elem_len; + goto check_node_accept_bytes_match; + } + } + } + } + } + else +# endif /* _LIBC */ + { + /* match with range expression? */ +#if __GNUC__ >= 2 && ! (__STDC_VERSION__ < 199901L && __STRICT_ANSI__) + wchar_t cmp_buf[] = {L'\0', L'\0', wc, L'\0', L'\0', L'\0'}; +#else + wchar_t cmp_buf[] = {L'\0', L'\0', L'\0', L'\0', L'\0', L'\0'}; + cmp_buf[2] = wc; +#endif + for (i = 0; i < cset->nranges; ++i) + { + cmp_buf[0] = cset->range_starts[i]; + cmp_buf[4] = cset->range_ends[i]; + if (wcscoll (cmp_buf, cmp_buf + 2) <= 0 + && wcscoll (cmp_buf + 2, cmp_buf + 4) <= 0) + { + match_len = char_len; + goto check_node_accept_bytes_match; + } + } + } + check_node_accept_bytes_match: + if (!cset->non_match) + return match_len; + else + { + if (match_len > 0) + return 0; + else + return (elem_len > char_len) ? elem_len : char_len; + } + } + return 0; +} + +# ifdef _LIBC +static unsigned int +internal_function +find_collation_sequence_value (const unsigned char *mbs, size_t mbs_len) +{ + uint32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + if (nrules == 0) + { + if (mbs_len == 1) + { + /* No valid character. Match it as a single byte character. */ + const unsigned char *collseq = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQMB); + return collseq[mbs[0]]; + } + return UINT_MAX; + } + else + { + int32_t idx; + const unsigned char *extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_SYMB_EXTRAMB); + int32_t extrasize = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_SYMB_EXTRAMB + 1) - extra; + + for (idx = 0; idx < extrasize;) + { + int mbs_cnt; + bool found = false; + int32_t elem_mbs_len; + /* Skip the name of collating element name. */ + idx = idx + extra[idx] + 1; + elem_mbs_len = extra[idx++]; + if (mbs_len == elem_mbs_len) + { + for (mbs_cnt = 0; mbs_cnt < elem_mbs_len; ++mbs_cnt) + if (extra[idx + mbs_cnt] != mbs[mbs_cnt]) + break; + if (mbs_cnt == elem_mbs_len) + /* Found the entry. */ + found = true; + } + /* Skip the byte sequence of the collating element. */ + idx += elem_mbs_len; + /* Adjust for the alignment. */ + idx = (idx + 3) & ~3; + /* Skip the collation sequence value. */ + idx += sizeof (uint32_t); + /* Skip the wide char sequence of the collating element. */ + idx = idx + sizeof (uint32_t) * (extra[idx] + 1); + /* If we found the entry, return the sequence value. */ + if (found) + return *(uint32_t *) (extra + idx); + /* Skip the collation sequence value. */ + idx += sizeof (uint32_t); + } + return UINT_MAX; + } +} +# endif /* _LIBC */ +#endif /* RE_ENABLE_I18N */ + +/* Check whether the node accepts the byte which is IDX-th + byte of the INPUT. */ + +static bool +internal_function +check_node_accept (const re_match_context_t *mctx, const re_token_t *node, + Idx idx) +{ + unsigned char ch; + ch = re_string_byte_at (&mctx->input, idx); + switch (node->type) + { + case CHARACTER: + if (node->opr.c != ch) + return false; + break; + + case SIMPLE_BRACKET: + if (!bitset_contain (node->opr.sbcset, ch)) + return false; + break; + +#ifdef RE_ENABLE_I18N + case OP_UTF8_PERIOD: + if (ch >= ASCII_CHARS) + return false; + /* FALLTHROUGH */ +#endif + case OP_PERIOD: + if ((ch == '\n' && !(mctx->dfa->syntax & RE_DOT_NEWLINE)) + || (ch == '\0' && (mctx->dfa->syntax & RE_DOT_NOT_NULL))) + return false; + break; + + default: + return false; + } + + if (node->constraint) + { + /* The node has constraints. Check whether the current context + satisfies the constraints. */ + unsigned int context = re_string_context_at (&mctx->input, idx, + mctx->eflags); + if (NOT_SATISFY_NEXT_CONSTRAINT (node->constraint, context)) + return false; + } + + return true; +} + +/* Extend the buffers, if the buffers have run out. */ + +static reg_errcode_t +internal_function +extend_buffers (re_match_context_t *mctx) +{ + reg_errcode_t ret; + re_string_t *pstr = &mctx->input; + + /* Avoid overflow. */ + if (BE (SIZE_MAX / 2 / sizeof (re_dfastate_t *) <= pstr->bufs_len, 0)) + return REG_ESPACE; + + /* Double the lengthes of the buffers. */ + ret = re_string_realloc_buffers (pstr, pstr->bufs_len * 2); + if (BE (ret != REG_NOERROR, 0)) + return ret; + + if (mctx->state_log != NULL) + { + /* And double the length of state_log. */ + /* XXX We have no indication of the size of this buffer. If this + allocation fail we have no indication that the state_log array + does not have the right size. */ + re_dfastate_t **new_array = re_realloc (mctx->state_log, re_dfastate_t *, + pstr->bufs_len + 1); + if (BE (new_array == NULL, 0)) + return REG_ESPACE; + mctx->state_log = new_array; + } + + /* Then reconstruct the buffers. */ + if (pstr->icase) + { +#ifdef RE_ENABLE_I18N + if (pstr->mb_cur_max > 1) + { + ret = build_wcs_upper_buffer (pstr); + if (BE (ret != REG_NOERROR, 0)) + return ret; + } + else +#endif /* RE_ENABLE_I18N */ + build_upper_buffer (pstr); + } + else + { +#ifdef RE_ENABLE_I18N + if (pstr->mb_cur_max > 1) + build_wcs_buffer (pstr); + else +#endif /* RE_ENABLE_I18N */ + { + if (pstr->trans != NULL) + re_string_translate_buffer (pstr); + } + } + return REG_NOERROR; +} + + +/* Functions for matching context. */ + +/* Initialize MCTX. */ + +static reg_errcode_t +internal_function +match_ctx_init (re_match_context_t *mctx, int eflags, Idx n) +{ + mctx->eflags = eflags; + mctx->match_last = REG_MISSING; + if (n > 0) + { + /* Avoid overflow. */ + size_t max_object_size = + MAX (sizeof (struct re_backref_cache_entry), + sizeof (re_sub_match_top_t *)); + if (BE (SIZE_MAX / max_object_size < n, 0)) + return REG_ESPACE; + + mctx->bkref_ents = re_malloc (struct re_backref_cache_entry, n); + mctx->sub_tops = re_malloc (re_sub_match_top_t *, n); + if (BE (mctx->bkref_ents == NULL || mctx->sub_tops == NULL, 0)) + return REG_ESPACE; + } + /* Already zero-ed by the caller. + else + mctx->bkref_ents = NULL; + mctx->nbkref_ents = 0; + mctx->nsub_tops = 0; */ + mctx->abkref_ents = n; + mctx->max_mb_elem_len = 1; + mctx->asub_tops = n; + return REG_NOERROR; +} + +/* Clean the entries which depend on the current input in MCTX. + This function must be invoked when the matcher changes the start index + of the input, or changes the input string. */ + +static void +internal_function +match_ctx_clean (re_match_context_t *mctx) +{ + Idx st_idx; + for (st_idx = 0; st_idx < mctx->nsub_tops; ++st_idx) + { + Idx sl_idx; + re_sub_match_top_t *top = mctx->sub_tops[st_idx]; + for (sl_idx = 0; sl_idx < top->nlasts; ++sl_idx) + { + re_sub_match_last_t *last = top->lasts[sl_idx]; + re_free (last->path.array); + re_free (last); + } + re_free (top->lasts); + if (top->path) + { + re_free (top->path->array); + re_free (top->path); + } + free (top); + } + + mctx->nsub_tops = 0; + mctx->nbkref_ents = 0; +} + +/* Free all the memory associated with MCTX. */ + +static void +internal_function +match_ctx_free (re_match_context_t *mctx) +{ + /* First, free all the memory associated with MCTX->SUB_TOPS. */ + match_ctx_clean (mctx); + re_free (mctx->sub_tops); + re_free (mctx->bkref_ents); +} + +/* Add a new backreference entry to MCTX. + Note that we assume that caller never call this function with duplicate + entry, and call with STR_IDX which isn't smaller than any existing entry. +*/ + +static reg_errcode_t +internal_function +match_ctx_add_entry (re_match_context_t *mctx, Idx node, Idx str_idx, Idx from, + Idx to) +{ + if (mctx->nbkref_ents >= mctx->abkref_ents) + { + struct re_backref_cache_entry* new_entry; + new_entry = re_realloc (mctx->bkref_ents, struct re_backref_cache_entry, + mctx->abkref_ents * 2); + if (BE (new_entry == NULL, 0)) + { + re_free (mctx->bkref_ents); + return REG_ESPACE; + } + mctx->bkref_ents = new_entry; + memset (mctx->bkref_ents + mctx->nbkref_ents, '\0', + sizeof (struct re_backref_cache_entry) * mctx->abkref_ents); + mctx->abkref_ents *= 2; + } + if (mctx->nbkref_ents > 0 + && mctx->bkref_ents[mctx->nbkref_ents - 1].str_idx == str_idx) + mctx->bkref_ents[mctx->nbkref_ents - 1].more = 1; + + mctx->bkref_ents[mctx->nbkref_ents].node = node; + mctx->bkref_ents[mctx->nbkref_ents].str_idx = str_idx; + mctx->bkref_ents[mctx->nbkref_ents].subexp_from = from; + mctx->bkref_ents[mctx->nbkref_ents].subexp_to = to; + + /* This is a cache that saves negative results of check_dst_limits_calc_pos. + If bit N is clear, means that this entry won't epsilon-transition to + an OP_OPEN_SUBEXP or OP_CLOSE_SUBEXP for the N+1-th subexpression. If + it is set, check_dst_limits_calc_pos_1 will recurse and try to find one + such node. + + A backreference does not epsilon-transition unless it is empty, so set + to all zeros if FROM != TO. */ + mctx->bkref_ents[mctx->nbkref_ents].eps_reachable_subexps_map + = (from == to ? -1 : 0); + + mctx->bkref_ents[mctx->nbkref_ents++].more = 0; + if (mctx->max_mb_elem_len < to - from) + mctx->max_mb_elem_len = to - from; + return REG_NOERROR; +} + +/* Return the first entry with the same str_idx, or REG_MISSING if none is + found. Note that MCTX->BKREF_ENTS is already sorted by MCTX->STR_IDX. */ + +static Idx +internal_function +search_cur_bkref_entry (const re_match_context_t *mctx, Idx str_idx) +{ + Idx left, right, mid, last; + last = right = mctx->nbkref_ents; + for (left = 0; left < right;) + { + mid = (left + right) / 2; + if (mctx->bkref_ents[mid].str_idx < str_idx) + left = mid + 1; + else + right = mid; + } + if (left < last && mctx->bkref_ents[left].str_idx == str_idx) + return left; + else + return REG_MISSING; +} + +/* Register the node NODE, whose type is OP_OPEN_SUBEXP, and which matches + at STR_IDX. */ + +static reg_errcode_t +internal_function +match_ctx_add_subtop (re_match_context_t *mctx, Idx node, Idx str_idx) +{ +#ifdef DEBUG + assert (mctx->sub_tops != NULL); + assert (mctx->asub_tops > 0); +#endif + if (BE (mctx->nsub_tops == mctx->asub_tops, 0)) + { + Idx new_asub_tops = mctx->asub_tops * 2; + re_sub_match_top_t **new_array = re_realloc (mctx->sub_tops, + re_sub_match_top_t *, + new_asub_tops); + if (BE (new_array == NULL, 0)) + return REG_ESPACE; + mctx->sub_tops = new_array; + mctx->asub_tops = new_asub_tops; + } + mctx->sub_tops[mctx->nsub_tops] = calloc (1, sizeof (re_sub_match_top_t)); + if (BE (mctx->sub_tops[mctx->nsub_tops] == NULL, 0)) + return REG_ESPACE; + mctx->sub_tops[mctx->nsub_tops]->node = node; + mctx->sub_tops[mctx->nsub_tops++]->str_idx = str_idx; + return REG_NOERROR; +} + +/* Register the node NODE, whose type is OP_CLOSE_SUBEXP, and which matches + at STR_IDX, whose corresponding OP_OPEN_SUBEXP is SUB_TOP. */ + +static re_sub_match_last_t * +internal_function +match_ctx_add_sublast (re_sub_match_top_t *subtop, Idx node, Idx str_idx) +{ + re_sub_match_last_t *new_entry; + if (BE (subtop->nlasts == subtop->alasts, 0)) + { + Idx new_alasts = 2 * subtop->alasts + 1; + re_sub_match_last_t **new_array = re_realloc (subtop->lasts, + re_sub_match_last_t *, + new_alasts); + if (BE (new_array == NULL, 0)) + return NULL; + subtop->lasts = new_array; + subtop->alasts = new_alasts; + } + new_entry = calloc (1, sizeof (re_sub_match_last_t)); + if (BE (new_entry != NULL, 1)) + { + subtop->lasts[subtop->nlasts] = new_entry; + new_entry->node = node; + new_entry->str_idx = str_idx; + ++subtop->nlasts; + } + return new_entry; +} + +static void +internal_function +sift_ctx_init (re_sift_context_t *sctx, re_dfastate_t **sifted_sts, + re_dfastate_t **limited_sts, Idx last_node, Idx last_str_idx) +{ + sctx->sifted_states = sifted_sts; + sctx->limited_states = limited_sts; + sctx->last_node = last_node; + sctx->last_str_idx = last_str_idx; + re_node_set_init_empty (&sctx->limits); +} diff --git a/gl/safe-read.c b/gl/safe-read.c new file mode 100644 index 00000000..b7bf1d5c --- /dev/null +++ b/gl/safe-read.c @@ -0,0 +1,78 @@ +/* An interface to read and write that retries after interrupts. + + Copyright (C) 1993, 1994, 1998, 2002, 2003, 2004, 2005, 2006 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#ifdef SAFE_WRITE +# include "safe-write.h" +#else +# include "safe-read.h" +#endif + +/* Get ssize_t. */ +#include <sys/types.h> +#include <unistd.h> + +#include <errno.h> + +#ifdef EINTR +# define IS_EINTR(x) ((x) == EINTR) +#else +# define IS_EINTR(x) 0 +#endif + +#include <limits.h> + +#ifdef SAFE_WRITE +# define safe_rw safe_write +# define rw write +#else +# define safe_rw safe_read +# define rw read +# undef const +# define const /* empty */ +#endif + +/* Read(write) up to COUNT bytes at BUF from(to) descriptor FD, retrying if + interrupted. Return the actual number of bytes read(written), zero for EOF, + or SAFE_READ_ERROR(SAFE_WRITE_ERROR) upon error. */ +size_t +safe_rw (int fd, void const *buf, size_t count) +{ + /* Work around a bug in Tru64 5.1. Attempting to read more than + INT_MAX bytes fails with errno == EINVAL. See + <http://lists.gnu.org/archive/html/bug-gnu-utils/2002-04/msg00010.html>. + When decreasing COUNT, keep it block-aligned. */ + enum { BUGGY_READ_MAXIMUM = INT_MAX & ~8191 }; + + for (;;) + { + ssize_t result = rw (fd, buf, count); + + if (0 <= result) + return result; + else if (IS_EINTR (errno)) + continue; + else if (errno == EINVAL && BUGGY_READ_MAXIMUM < count) + count = BUGGY_READ_MAXIMUM; + else + return result; + } +} diff --git a/gl/safe-read.h b/gl/safe-read.h new file mode 100644 index 00000000..3451955a --- /dev/null +++ b/gl/safe-read.h @@ -0,0 +1,35 @@ +/* An interface to read() that retries after interrupts. + Copyright (C) 2002, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <stddef.h> + +#ifdef __cplusplus +extern "C" { +#endif + + +#define SAFE_READ_ERROR ((size_t) -1) + +/* Read up to COUNT bytes at BUF from descriptor FD, retrying if interrupted. + Return the actual number of bytes read, zero for EOF, or SAFE_READ_ERROR + upon error. */ +extern size_t safe_read (int fd, void *buf, size_t count); + + +#ifdef __cplusplus +} +#endif diff --git a/gl/safe-write.c b/gl/safe-write.c new file mode 100644 index 00000000..4c375a6c --- /dev/null +++ b/gl/safe-write.c @@ -0,0 +1,19 @@ +/* An interface to write that retries after interrupts. + Copyright (C) 2002 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#define SAFE_WRITE +#include "safe-read.c" diff --git a/gl/safe-write.h b/gl/safe-write.h new file mode 100644 index 00000000..c1946362 --- /dev/null +++ b/gl/safe-write.h @@ -0,0 +1,25 @@ +/* An interface to write() that retries after interrupts. + Copyright (C) 2002 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <stddef.h> + +#define SAFE_WRITE_ERROR ((size_t) -1) + +/* Write up to COUNT bytes at BUF to descriptor FD, retrying if interrupted. + Return the actual number of bytes written, zero for EOF, or SAFE_WRITE_ERROR + upon error. */ +extern size_t safe_write (int fd, const void *buf, size_t count); diff --git a/gl/size_max.h b/gl/size_max.h new file mode 100644 index 00000000..ed0bc137 --- /dev/null +++ b/gl/size_max.h @@ -0,0 +1,31 @@ +/* size_max.h -- declare SIZE_MAX through system headers + Copyright (C) 2005-2006 Free Software Foundation, Inc. + Written by Simon Josefsson. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef GNULIB_SIZE_MAX_H +#define GNULIB_SIZE_MAX_H + +/* Get SIZE_MAX declaration on systems like Solaris 7/8/9. */ +# include <limits.h> +/* Get SIZE_MAX declaration on systems like glibc 2. */ +# if HAVE_STDINT_H +# include <stdint.h> +# endif +/* On systems where these include files don't define it, SIZE_MAX is defined + in config.h. */ + +#endif /* GNULIB_SIZE_MAX_H */ diff --git a/gl/snprintf.c b/gl/snprintf.c new file mode 100644 index 00000000..db1ca9af --- /dev/null +++ b/gl/snprintf.c @@ -0,0 +1,76 @@ +/* Formatted output to strings. + Copyright (C) 2004, 2006 Free Software Foundation, Inc. + Written by Simon Josefsson and Paul Eggert. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "snprintf.h" + +#include <errno.h> +#include <limits.h> +#include <stdarg.h> +#include <stdlib.h> +#include <string.h> + +#include "vasnprintf.h" + +/* Some systems, like OSF/1 4.0 and Woe32, don't have EOVERFLOW. */ +#ifndef EOVERFLOW +# define EOVERFLOW E2BIG +#endif + +/* Print formatted output to string STR. Similar to sprintf, but + additional length SIZE limit how much is written into STR. Returns + string length of formatted string (which may be larger than SIZE). + STR may be NULL, in which case nothing will be written. On error, + return a negative value. */ +int +snprintf (char *str, size_t size, const char *format, ...) +{ + char *output; + size_t len; + size_t lenbuf = size; + va_list args; + + va_start (args, format); + output = vasnprintf (str, &lenbuf, format, args); + len = lenbuf; + va_end (args); + + if (!output) + return -1; + + if (output != str) + { + if (size) + { + size_t pruned_len = (len < size ? len : size - 1); + memcpy (str, output, pruned_len); + str[pruned_len] = '\0'; + } + + free (output); + } + + if (INT_MAX < len) + { + errno = EOVERFLOW; + return -1; + } + + return len; +} diff --git a/gl/snprintf.h b/gl/snprintf.h new file mode 100644 index 00000000..5032b9e8 --- /dev/null +++ b/gl/snprintf.h @@ -0,0 +1,29 @@ +/* Formatted output to strings. + Copyright (C) 2004 Free Software Foundation, Inc. + Written by Simon Josefsson. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef SNPRINTF_H +#define SNPRINTF_H + +/* Get snprintf declaration, if available. */ +#include <stdio.h> + +#if defined HAVE_DECL_SNPRINTF && !HAVE_DECL_SNPRINTF +int snprintf (char *str, size_t size, const char *format, ...); +#endif + +#endif /* SNPRINTF_H */ diff --git a/gl/socket_.h b/gl/socket_.h new file mode 100644 index 00000000..8b28b5ed --- /dev/null +++ b/gl/socket_.h @@ -0,0 +1,70 @@ +/* Provide a sys/socket header file for systems lacking it (read: MinGW). + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + Written by Simon Josefsson. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _SYS_SOCKET_H +#define _SYS_SOCKET_H + +/* This file is supposed to be used on platforms that lack + sys/socket.h. It is intended to provide definitions and prototypes + needed by an application. + + Currently only MinGW is supported. See the gnulib manual regarding + Windows sockets. MinGW has the header files winsock2.h and + ws2tcpip.h that declare the sys/socket.h definitions we need. Note + that you can influence which definitions you get by setting the + WINVER symbol before including these two files. For example, + getaddrinfo is only available if _WIN32_WINNT >= 0x0501 (that + symbol is set indiriectly through WINVER). You can set this by + adding AC_DEFINE(WINVER, 0x0501) to configure.ac. Note that your + code may not run on older Windows releases then. My Windows 2000 + box was not able to run the code, for example. The situation is + slightly confusing because: + http://msdn.microsoft.com/library/default.asp?url=/library/en-us/winsock/winsock/getaddrinfo_2.asp + suggests that getaddrinfo should be available on all Windows + releases. */ + + +#if HAVE_WINSOCK2_H +# include <winsock2.h> +#endif +#if HAVE_WS2TCPIP_H +# include <ws2tcpip.h> +#endif + +/* For shutdown(). */ +#if !defined SHUT_RD && defined SD_RECEIVE +# define SHUT_RD SD_RECEIVE +#endif +#if !defined SHUT_WR && defined SD_SEND +# define SHUT_WR SD_SEND +#endif +#if !defined SHUT_RDWR && defined SD_BOTH +# define SHUT_RDWR SD_BOTH +#endif + +#if defined _WIN32 || defined __WIN32__ +# define ENOTSOCK WSAENOTSOCK +# define EADDRINUSE WSAEADDRINUSE +# define ENETRESET WSAENETRESET +# define ECONNABORTED WSAECONNABORTED +# define ECONNRESET WSAECONNRESET +# define ENOTCONN WSAENOTCONN +# define ESHUTDOWN WSAESHUTDOWN +#endif + +#endif /* _SYS_SOCKET_H */ diff --git a/gl/stdbool_.h b/gl/stdbool_.h new file mode 100644 index 00000000..efa80ba9 --- /dev/null +++ b/gl/stdbool_.h @@ -0,0 +1,115 @@ +/* Copyright (C) 2001, 2002, 2003, 2006 Free Software Foundation, Inc. + Written by Bruno Haible <haible@clisp.cons.org>, 2001. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _STDBOOL_H +#define _STDBOOL_H + +/* ISO C 99 <stdbool.h> for platforms that lack it. */ + +/* Usage suggestions: + + Programs that use <stdbool.h> should be aware of some limitations + and standards compliance issues. + + Standards compliance: + + - <stdbool.h> must be #included before 'bool', 'false', 'true' + can be used. + + - You cannot assume that sizeof (bool) == 1. + + - Programs should not undefine the macros bool, true, and false, + as C99 lists that as an "obsolescent feature". + + Limitations of this substitute, when used in a C89 environment: + + - <stdbool.h> must be #included before the '_Bool' type can be used. + + - You cannot assume that _Bool is a typedef; it might be a macro. + + - In C99, casts and automatic conversions to '_Bool' or 'bool' are + performed in such a way that every nonzero value gets converted + to 'true', and zero gets converted to 'false'. This doesn't work + with this substitute. With this substitute, only the values 0 and 1 + give the expected result when converted to _Bool' or 'bool'. + + Also, it is suggested that programs use 'bool' rather than '_Bool'; + this isn't required, but 'bool' is more common. */ + + +/* 7.16. Boolean type and values */ + +/* BeOS <sys/socket.h> already #defines false 0, true 1. We use the same + definitions below, but temporarily we have to #undef them. */ +#ifdef __BEOS__ +# include <OS.h> /* defines bool but not _Bool */ +# undef false +# undef true +#endif + +/* For the sake of symbolic names in gdb, we define true and false as + enum constants, not only as macros. + It is tempting to write + typedef enum { false = 0, true = 1 } _Bool; + so that gdb prints values of type 'bool' symbolically. But if we do + this, values of type '_Bool' may promote to 'int' or 'unsigned int' + (see ISO C 99 6.7.2.2.(4)); however, '_Bool' must promote to 'int' + (see ISO C 99 6.3.1.1.(2)). So we add a negative value to the + enum; this ensures that '_Bool' promotes to 'int'. */ +#if defined __cplusplus || defined __BEOS__ + /* A compiler known to have 'bool'. */ + /* If the compiler already has both 'bool' and '_Bool', we can assume they + are the same types. */ +# if !@HAVE__BOOL@ +typedef bool _Bool; +# endif +#else +# if !defined __GNUC__ + /* If @HAVE__BOOL@: + Some HP-UX cc and AIX IBM C compiler versions have compiler bugs when + the built-in _Bool type is used. See + http://gcc.gnu.org/ml/gcc-patches/2003-12/msg02303.html + http://lists.gnu.org/archive/html/bug-coreutils/2005-11/msg00161.html + http://lists.gnu.org/archive/html/bug-coreutils/2005-10/msg00086.html + Similar bugs are likely with other compilers as well; this file + wouldn't be used if <stdbool.h> was working. + So we override the _Bool type. + If !@HAVE__BOOL@: + Need to define _Bool ourselves. As 'signed char' or as an enum type? + Use of a typedef, with SunPRO C, leads to a stupid + "warning: _Bool is a keyword in ISO C99". + Use of an enum type, with IRIX cc, leads to a stupid + "warning(1185): enumerated type mixed with another type". + The only benefit of the enum type, debuggability, is not important + with these compilers. So use 'signed char' and no typedef. */ +# define _Bool signed char +enum { false = 0, true = 1 }; +# else + /* With this compiler, trust the _Bool type if the compiler has it. */ +# if !@HAVE__BOOL@ +typedef enum { _Bool_must_promote_to_int = -1, false = 0, true = 1 } _Bool; +# endif +# endif +#endif +#define bool _Bool + +/* The other macros must be usable in preprocessor directives. */ +#define false 0 +#define true 1 +#define __bool_true_false_are_defined 1 + +#endif /* _STDBOOL_H */ diff --git a/gl/stdint_.h b/gl/stdint_.h new file mode 100644 index 00000000..64ec8c5b --- /dev/null +++ b/gl/stdint_.h @@ -0,0 +1,489 @@ +/* Copyright (C) 2001-2002, 2004-2007 Free Software Foundation, Inc. + Written by Paul Eggert, Bruno Haible, Sam Steingold, Peter Burwood. + This file is part of gnulib. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _GL_STDINT_H +#define _GL_STDINT_H + +/* + * ISO C 99 <stdint.h> for platforms that lack it. + * <http://www.opengroup.org/susv3xbd/stdint.h.html> + */ + +/* Get those types that are already defined in other system include + files, so that we can "#define int8_t signed char" below without + worrying about a later system include file containing a "typedef + signed char int8_t;" that will get messed up by our macro. Our + macros should all be consistent with the system versions, except + for the "fast" types and macros, which we recommend against using + in public interfaces due to compiler differences. */ + +#if @HAVE_STDINT_H@ +# if defined __sgi && ! defined __c99 + /* Bypass IRIX's <stdint.h> if in C89 mode, since it merely annoys users + with "This header file is to be used only for c99 mode compilations" + diagnostics. */ +# define __STDINT_H__ +# endif + /* Other systems may have an incomplete or buggy <stdint.h>. + Include it before <inttypes.h>, since any "#include <stdint.h>" + in <inttypes.h> would reinclude us, skipping our contents because + _GL_STDINT_H is defined. */ +# include @ABSOLUTE_STDINT_H@ +#endif + +/* <sys/types.h> defines some of the stdint.h types as well, on glibc, + IRIX 6.5, and OpenBSD 3.8 (via <machine/types.h>). + AIX 5.2 <sys/types.h> isn't needed and causes troubles. + MacOS X 10.4.6 <sys/types.h> includes <stdint.h> (which is us), but + relies on the system <stdint.h> definitions, so include + <sys/types.h> after @ABSOLUTE_STDINT_H@. */ +#if @HAVE_SYS_TYPES_H@ && ! defined _AIX +# include <sys/types.h> +#endif + +/* Get LONG_MIN, LONG_MAX, ULONG_MAX. */ +#include <limits.h> + +#if @HAVE_INTTYPES_H@ + /* In OpenBSD 3.8, <inttypes.h> includes <machine/types.h>, which defines + int{8,16,32,64}_t, uint{8,16,32,64}_t and __BIT_TYPES_DEFINED__. + <inttypes.h> also defines intptr_t and uintptr_t. */ +# define _GL_JUST_INCLUDE_ABSOLUTE_INTTYPES_H +# include <inttypes.h> +# undef _GL_JUST_INCLUDE_ABSOLUTE_INTTYPES_H +#elif @HAVE_SYS_INTTYPES_H@ + /* Solaris 7 <sys/inttypes.h> has the types except the *_fast*_t types, and + the macros except for *_FAST*_*, INTPTR_MIN, PTRDIFF_MIN, PTRDIFF_MAX. */ +# include <sys/inttypes.h> +#endif + +#if @HAVE_SYS_BITYPES_H@ && ! defined __BIT_TYPES_DEFINED__ + /* Linux libc4 >= 4.6.7 and libc5 have a <sys/bitypes.h> that defines + int{8,16,32,64}_t and __BIT_TYPES_DEFINED__. In libc5 >= 5.2.2 it is + included by <sys/types.h>. */ +# include <sys/bitypes.h> +#endif + +#if ! defined __cplusplus || defined __STDC_CONSTANT_MACROS + +/* Get WCHAR_MIN, WCHAR_MAX. */ +# if ! (defined WCHAR_MIN && defined WCHAR_MAX) +# include <wchar.h> +# endif + +#endif + +/* Minimum and maximum values for a integer type under the usual assumption. + Return an unspecified value if BITS == 0, adding a check to pacify + picky compilers. */ + +#define _STDINT_MIN(signed, bits, zero) \ + ((signed) ? (- ((zero) + 1) << ((bits) ? (bits) - 1 : 0)) : (zero)) + +#define _STDINT_MAX(signed, bits, zero) \ + ((signed) \ + ? ~ _STDINT_MIN (signed, bits, zero) \ + : ((((zero) + 1) << ((bits) ? (bits) - 1 : 0)) - 1) * 2 + 1) + +/* 7.18.1.1. Exact-width integer types */ + +/* Here we assume a standard architecture where the hardware integer + types have 8, 16, 32, optionally 64 bits. */ + +#undef int8_t +#undef uint8_t +#define int8_t signed char +#define uint8_t unsigned char + +#undef int16_t +#undef uint16_t +#define int16_t short int +#define uint16_t unsigned short int + +#undef int32_t +#undef uint32_t +#define int32_t int +#define uint32_t unsigned int + +#undef int64_t +#if LONG_MAX >> 31 >> 31 == 1 +# define int64_t long int +#elif defined _MSC_VER +# define int64_t __int64 +#elif @HAVE_LONG_LONG_INT@ +# define int64_t long long int +#endif + +#undef uint64_t +#if ULONG_MAX >> 31 >> 31 >> 1 == 1 +# define uint64_t unsigned long int +#elif defined _MSC_VER +# define uint64_t unsigned __int64 +#elif @HAVE_UNSIGNED_LONG_LONG_INT@ +# define uint64_t unsigned long long int +#endif + +/* Avoid collision with Solaris 2.5.1 <pthread.h> etc. */ +#define _UINT8_T +#define _UINT32_T +#define _UINT64_T + + +/* 7.18.1.2. Minimum-width integer types */ + +/* Here we assume a standard architecture where the hardware integer + types have 8, 16, 32, optionally 64 bits. Therefore the leastN_t types + are the same as the corresponding N_t types. */ + +#undef int_least8_t +#undef uint_least8_t +#undef int_least16_t +#undef uint_least16_t +#undef int_least32_t +#undef uint_least32_t +#undef int_least64_t +#undef uint_least64_t +#define int_least8_t int8_t +#define uint_least8_t uint8_t +#define int_least16_t int16_t +#define uint_least16_t uint16_t +#define int_least32_t int32_t +#define uint_least32_t uint32_t +#ifdef int64_t +# define int_least64_t int64_t +#endif +#ifdef uint64_t +# define uint_least64_t uint64_t +#endif + +/* 7.18.1.3. Fastest minimum-width integer types */ + +/* Note: Other <stdint.h> substitutes may define these types differently. + It is not recommended to use these types in public header files. */ + +/* Here we assume a standard architecture where the hardware integer + types have 8, 16, 32, optionally 64 bits. Therefore the fastN_t types + are taken from the same list of types. Assume that 'long int' + is fast enough for all narrower integers. */ + +#undef int_fast8_t +#undef uint_fast8_t +#undef int_fast16_t +#undef uint_fast16_t +#undef int_fast32_t +#undef uint_fast32_t +#undef int_fast64_t +#undef uint_fast64_t +#define int_fast8_t long int +#define uint_fast8_t unsigned int_fast8_t +#define int_fast16_t long int +#define uint_fast16_t unsigned int_fast16_t +#define int_fast32_t long int +#define uint_fast32_t unsigned int_fast32_t +#ifdef int64_t +# define int_fast64_t int64_t +#endif +#ifdef uint64_t +# define uint_fast64_t uint64_t +#endif + +/* 7.18.1.4. Integer types capable of holding object pointers */ + +#undef intptr_t +#undef uintptr_t +#define intptr_t long int +#define uintptr_t unsigned long int + +/* 7.18.1.5. Greatest-width integer types */ + +/* Note: These types are compiler dependent. It may be unwise to use them in + public header files. */ + +#undef intmax_t +#if @HAVE_LONG_LONG_INT@ && LONG_MAX >> 30 == 1 +# define intmax_t long long int +#elif defined int64_t +# define intmax_t int64_t +#else +# define intmax_t long int +#endif + +#undef uintmax_t +#if @HAVE_UNSIGNED_LONG_LONG_INT@ && ULONG_MAX >> 31 == 1 +# define uintmax_t unsigned long long int +#elif defined uint64_t +# define uintmax_t uint64_t +#else +# define uintmax_t unsigned long int +#endif + +/* 7.18.2. Limits of specified-width integer types */ + +#if ! defined __cplusplus || defined __STDC_LIMIT_MACROS + +/* 7.18.2.1. Limits of exact-width integer types */ + +/* Here we assume a standard architecture where the hardware integer + types have 8, 16, 32, optionally 64 bits. */ + +#undef INT8_MIN +#undef INT8_MAX +#undef UINT8_MAX +#define INT8_MIN (~ INT8_MAX) +#define INT8_MAX 127 +#define UINT8_MAX 255 + +#undef INT16_MIN +#undef INT16_MAX +#undef UINT16_MAX +#define INT16_MIN (~ INT16_MAX) +#define INT16_MAX 32767 +#define UINT16_MAX 65535 + +#undef INT32_MIN +#undef INT32_MAX +#undef UINT32_MAX +#define INT32_MIN (~ INT32_MAX) +#define INT32_MAX 2147483647 +#define UINT32_MAX 4294967295U + +#undef INT64_MIN +#undef INT64_MAX +#ifdef int64_t +# define INT64_MIN (~ INT64_MAX) +# define INT64_MAX INTMAX_C (9223372036854775807) +#endif + +#undef UINT64_MAX +#ifdef uint64_t +# define UINT64_MAX UINTMAX_C (18446744073709551615) +#endif + +/* 7.18.2.2. Limits of minimum-width integer types */ + +/* Here we assume a standard architecture where the hardware integer + types have 8, 16, 32, optionally 64 bits. Therefore the leastN_t types + are the same as the corresponding N_t types. */ + +#undef INT_LEAST8_MIN +#undef INT_LEAST8_MAX +#undef UINT_LEAST8_MAX +#define INT_LEAST8_MIN INT8_MIN +#define INT_LEAST8_MAX INT8_MAX +#define UINT_LEAST8_MAX UINT8_MAX + +#undef INT_LEAST16_MIN +#undef INT_LEAST16_MAX +#undef UINT_LEAST16_MAX +#define INT_LEAST16_MIN INT16_MIN +#define INT_LEAST16_MAX INT16_MAX +#define UINT_LEAST16_MAX UINT16_MAX + +#undef INT_LEAST32_MIN +#undef INT_LEAST32_MAX +#undef UINT_LEAST32_MAX +#define INT_LEAST32_MIN INT32_MIN +#define INT_LEAST32_MAX INT32_MAX +#define UINT_LEAST32_MAX UINT32_MAX + +#undef INT_LEAST64_MIN +#undef INT_LEAST64_MAX +#ifdef int64_t +# define INT_LEAST64_MIN INT64_MIN +# define INT_LEAST64_MAX INT64_MAX +#endif + +#undef UINT_LEAST64_MAX +#ifdef uint64_t +# define UINT_LEAST64_MAX UINT64_MAX +#endif + +/* 7.18.2.3. Limits of fastest minimum-width integer types */ + +/* Here we assume a standard architecture where the hardware integer + types have 8, 16, 32, optionally 64 bits. Therefore the fastN_t types + are taken from the same list of types. */ + +#undef INT_FAST8_MIN +#undef INT_FAST8_MAX +#undef UINT_FAST8_MAX +#define INT_FAST8_MIN LONG_MIN +#define INT_FAST8_MAX LONG_MAX +#define UINT_FAST8_MAX ULONG_MAX + +#undef INT_FAST16_MIN +#undef INT_FAST16_MAX +#undef UINT_FAST16_MAX +#define INT_FAST16_MIN LONG_MIN +#define INT_FAST16_MAX LONG_MAX +#define UINT_FAST16_MAX ULONG_MAX + +#undef INT_FAST32_MIN +#undef INT_FAST32_MAX +#undef UINT_FAST32_MAX +#define INT_FAST32_MIN LONG_MIN +#define INT_FAST32_MAX LONG_MAX +#define UINT_FAST32_MAX ULONG_MAX + +#undef INT_FAST64_MIN +#undef INT_FAST64_MAX +#ifdef int64_t +# define INT_FAST64_MIN INT64_MIN +# define INT_FAST64_MAX INT64_MAX +#endif + +#undef UINT_FAST64_MAX +#ifdef uint64_t +# define UINT_FAST64_MAX UINT64_MAX +#endif + +/* 7.18.2.4. Limits of integer types capable of holding object pointers */ + +#undef INTPTR_MIN +#undef INTPTR_MAX +#undef UINTPTR_MAX +#define INTPTR_MIN LONG_MIN +#define INTPTR_MAX LONG_MAX +#define UINTPTR_MAX ULONG_MAX + +/* 7.18.2.5. Limits of greatest-width integer types */ + +#undef INTMAX_MIN +#undef INTMAX_MAX +#define INTMAX_MIN (~ INTMAX_MAX) +#ifdef INT64_MAX +# define INTMAX_MAX INT64_MAX +#else +# define INTMAX_MAX INT32_MAX +#endif + +#undef UINTMAX_MAX +#ifdef UINT64_MAX +# define UINTMAX_MAX UINT64_MAX +#else +# define UINTMAX_MAX UINT32_MAX +#endif + +/* 7.18.3. Limits of other integer types */ + +/* ptrdiff_t limits */ +#undef PTRDIFF_MIN +#undef PTRDIFF_MAX +#define PTRDIFF_MIN \ + _STDINT_MIN (1, @BITSIZEOF_PTRDIFF_T@, 0@PTRDIFF_T_SUFFIX@) +#define PTRDIFF_MAX \ + _STDINT_MAX (1, @BITSIZEOF_PTRDIFF_T@, 0@PTRDIFF_T_SUFFIX@) + +/* sig_atomic_t limits */ +#undef SIG_ATOMIC_MIN +#undef SIG_ATOMIC_MAX +#define SIG_ATOMIC_MIN \ + _STDINT_MIN (@HAVE_SIGNED_SIG_ATOMIC_T@, @BITSIZEOF_SIG_ATOMIC_T@, \ + 0@SIG_ATOMIC_T_SUFFIX@) +#define SIG_ATOMIC_MAX \ + _STDINT_MAX (@HAVE_SIGNED_SIG_ATOMIC_T@, @BITSIZEOF_SIG_ATOMIC_T@, \ + 0@SIG_ATOMIC_T_SUFFIX@) + + +/* size_t limit */ +#undef SIZE_MAX +#define SIZE_MAX _STDINT_MAX (0, @BITSIZEOF_SIZE_T@, 0@SIZE_T_SUFFIX@) + +/* wchar_t limits */ +#undef WCHAR_MIN +#undef WCHAR_MAX +#define WCHAR_MIN \ + _STDINT_MIN (@HAVE_SIGNED_WCHAR_T@, @BITSIZEOF_WCHAR_T@, 0@WCHAR_T_SUFFIX@) +#define WCHAR_MAX \ + _STDINT_MAX (@HAVE_SIGNED_WCHAR_T@, @BITSIZEOF_WCHAR_T@, 0@WCHAR_T_SUFFIX@) + +/* wint_t limits */ +#undef WINT_MIN +#undef WINT_MAX +#define WINT_MIN \ + _STDINT_MIN (@HAVE_SIGNED_WINT_T@, @BITSIZEOF_WINT_T@, 0@WINT_T_SUFFIX@) +#define WINT_MAX \ + _STDINT_MAX (@HAVE_SIGNED_WINT_T@, @BITSIZEOF_WINT_T@, 0@WINT_T_SUFFIX@) + +#endif /* !defined __cplusplus || defined __STDC_LIMIT_MACROS */ + +/* 7.18.4. Macros for integer constants */ + +#if ! defined __cplusplus || defined __STDC_CONSTANT_MACROS + +/* 7.18.4.1. Macros for minimum-width integer constants */ +/* According to ISO C 99 Technical Corrigendum 1 */ + +/* Here we assume a standard architecture where the hardware integer + types have 8, 16, 32, optionally 64 bits, and int is 32 bits. */ + +#undef INT8_C +#undef UINT8_C +#define INT8_C(x) x +#define UINT8_C(x) x + +#undef INT16_C +#undef UINT16_C +#define INT16_C(x) x +#define UINT16_C(x) x + +#undef INT32_C +#undef UINT32_C +#define INT32_C(x) x +#define UINT32_C(x) x ## U + +#undef INT64_C +#undef UINT64_C +#if LONG_MAX >> 31 >> 31 == 1 +# define INT64_C(x) x##L +#elif defined _MSC_VER +# define INT64_C(x) x##i64 +#elif @HAVE_LONG_LONG_INT@ +# define INT64_C(x) x##LL +#endif +#if ULONG_MAX >> 31 >> 31 >> 1 == 1 +# define UINT64_C(x) x##UL +#elif defined _MSC_VER +# define UINT64_C(x) x##ui64 +#elif @HAVE_UNSIGNED_LONG_LONG_INT@ +# define UINT64_C(x) x##ULL +#endif + +/* 7.18.4.2. Macros for greatest-width integer constants */ + +#undef INTMAX_C +#if @HAVE_LONG_LONG_INT@ && LONG_MAX >> 30 == 1 +# define INTMAX_C(x) x##LL +#elif defined int64_t +# define INTMAX_C(x) INT64_C(x) +#else +# define INTMAX_C(x) x##L +#endif + +#undef UINTMAX_C +#if @HAVE_UNSIGNED_LONG_LONG_INT@ && ULONG_MAX >> 31 == 1 +# define UINTMAX_C(x) x##ULL +#elif defined uint64_t +# define UINTMAX_C(x) UINT64_C(x) +#else +# define UINTMAX_C(x) x##UL +#endif + +#endif /* !defined __cplusplus || defined __STDC_CONSTANT_MACROS */ + +#endif /* _GL_STDINT_H */ diff --git a/gl/strcase.h b/gl/strcase.h new file mode 100644 index 00000000..07d4c9a4 --- /dev/null +++ b/gl/strcase.h @@ -0,0 +1,56 @@ +/* Case-insensitive string comparison functions. + Copyright (C) 1995-1996, 2001, 2003, 2005-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _STRCASE_H +#define _STRCASE_H + +#include <stddef.h> +/* Include header files with a possibly conflicting declarations of strcasecmp + before we define it as a macro, so that they will be no-ops if included + after strcasecmp is defined as a macro. */ +#include <string.h> + +#ifdef __cplusplus +extern "C" { +#endif + + +/* No known system has a strcasecmp() function that works correctly in + multibyte locales. Therefore we use our version always. */ +#define strcasecmp rpl_strcasecmp +/* Compare strings S1 and S2, ignoring case, returning less than, equal to or + greater than zero if S1 is lexicographically less than, equal to or greater + than S2. + Note: This function may, in multibyte locales, return 0 for strings of + different lengths! */ +extern int strcasecmp (const char *s1, const char *s2); + +/* Compare no more than N characters of strings S1 and S2, ignoring case, + returning less than, equal to or greater than zero if S1 is + lexicographically less than, equal to or greater than S2. + Note: This function can not work correctly in multibyte locales. */ +#if ! HAVE_DECL_STRNCASECMP +extern int strncasecmp (const char *s1, const char *s2, size_t n); +#endif + + +#ifdef __cplusplus +} +#endif + + +#endif /* _STRCASE_H */ diff --git a/gl/strcasecmp.c b/gl/strcasecmp.c new file mode 100644 index 00000000..99d5dd22 --- /dev/null +++ b/gl/strcasecmp.c @@ -0,0 +1,103 @@ +/* Case-insensitive string comparison function. + Copyright (C) 1998-1999, 2005-2006 Free Software Foundation, Inc. + Written by Bruno Haible <bruno@clisp.org>, 2005, + based on earlier glibc code. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include "strcase.h" + +#include <ctype.h> +#include <limits.h> + +#if HAVE_MBRTOWC +# include "mbuiter.h" +#endif + +#define TOLOWER(Ch) (isupper (Ch) ? tolower (Ch) : (Ch)) + +/* Compare strings S1 and S2, ignoring case, returning less than, equal to or + greater than zero if S1 is lexicographically less than, equal to or greater + than S2. + Note: This function may, in multibyte locales, return 0 for strings of + different lengths! */ +int +strcasecmp (const char *s1, const char *s2) +{ + if (s1 == s2) + return 0; + + /* Be careful not to look at the entire extent of s1 or s2 until needed. + This is useful because when two strings differ, the difference is + most often already in the very few first characters. */ +#if HAVE_MBRTOWC + if (MB_CUR_MAX > 1) + { + mbui_iterator_t iter1; + mbui_iterator_t iter2; + + mbui_init (iter1, s1); + mbui_init (iter2, s2); + + while (mbui_avail (iter1) && mbui_avail (iter2)) + { + int cmp = mb_casecmp (mbui_cur (iter1), mbui_cur (iter2)); + + if (cmp != 0) + return cmp; + + mbui_advance (iter1); + mbui_advance (iter2); + } + if (mbui_avail (iter1)) + /* s2 terminated before s1. */ + return 1; + if (mbui_avail (iter2)) + /* s1 terminated before s2. */ + return -1; + return 0; + } + else +#endif + { + const unsigned char *p1 = (const unsigned char *) s1; + const unsigned char *p2 = (const unsigned char *) s2; + unsigned char c1, c2; + + do + { + c1 = TOLOWER (*p1); + c2 = TOLOWER (*p2); + + if (c1 == '\0') + break; + + ++p1; + ++p2; + } + while (c1 == c2); + + if (UCHAR_MAX <= INT_MAX) + return c1 - c2; + else + /* On machines where 'char' and 'int' are types of the same size, the + difference of two 'unsigned char' values - including the sign bit - + doesn't fit in an 'int'. */ + return (c1 > c2 ? 1 : c1 < c2 ? -1 : 0); + } +} diff --git a/gl/strdup.c b/gl/strdup.c new file mode 100644 index 00000000..17d40d62 --- /dev/null +++ b/gl/strdup.c @@ -0,0 +1,55 @@ +/* Copyright (C) 1991, 1996, 1997, 1998, 2002, 2003, 2004, 2006 Free + Software Foundation, Inc. + + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _LIBC +# include <config.h> +/* Get specification. */ +# include "strdup.h" +#endif + +#include <stdlib.h> +#include <string.h> + +#undef __strdup +#ifdef _LIBC +# undef strdup +#endif + +#ifndef weak_alias +# define __strdup strdup +#endif + +/* Duplicate S, returning an identical malloc'd string. */ +char * +__strdup (const char *s) +{ + size_t len = strlen (s) + 1; + void *new = malloc (len); + + if (new == NULL) + return NULL; + + return (char *) memcpy (new, s, len); +} +#ifdef libc_hidden_def +libc_hidden_def (__strdup) +#endif +#ifdef weak_alias +weak_alias (__strdup, strdup) +#endif diff --git a/gl/strdup.h b/gl/strdup.h new file mode 100644 index 00000000..73e66e3d --- /dev/null +++ b/gl/strdup.h @@ -0,0 +1,39 @@ +/* strdup.h -- duplicate a string + Copyright (C) 2004, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef STRDUP_H_ +#define STRDUP_H_ + +/* Get strdup declaration, if available. */ +#include <string.h> + +#ifdef __cplusplus +extern "C" { +#endif + + +#if defined HAVE_DECL_STRDUP && !HAVE_DECL_STRDUP && !defined strdup +/* Duplicate S, returning an identical malloc'd string. */ +extern char *strdup (const char *s); +#endif + + +#ifdef __cplusplus +} +#endif + +#endif /* STRDUP_H_ */ diff --git a/gl/stripslash.c b/gl/stripslash.c new file mode 100644 index 00000000..342d497c --- /dev/null +++ b/gl/stripslash.c @@ -0,0 +1,45 @@ +/* stripslash.c -- remove redundant trailing slashes from a file name + + Copyright (C) 1990, 2001, 2003-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "dirname.h" + +/* Remove trailing slashes from FILE. Return true if a trailing slash + was removed. This is useful when using file name completion from a + shell that adds a "/" after directory names (such as tcsh and + bash), because on symlinks to directories, several system calls + have different semantics according to whether a trailing slash is + present. */ + +bool +strip_trailing_slashes (char *file) +{ + char *base = last_component (file); + char *base_lim; + bool had_slash; + + /* last_component returns "" for file system roots, but we need to turn + `///' into `/'. */ + if (! *base) + base = file; + base_lim = base + base_len (base); + had_slash = (*base_lim != '\0'); + *base_lim = '\0'; + return had_slash; +} diff --git a/gl/strncasecmp.c b/gl/strncasecmp.c new file mode 100644 index 00000000..f59b953a --- /dev/null +++ b/gl/strncasecmp.c @@ -0,0 +1,63 @@ +/* strncasecmp.c -- case insensitive string comparator + Copyright (C) 1998-1999, 2005-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include "strcase.h" + +#include <ctype.h> +#include <limits.h> + +#define TOLOWER(Ch) (isupper (Ch) ? tolower (Ch) : (Ch)) + +/* Compare no more than N bytes of strings S1 and S2, + ignoring case, returning less than, equal to or + greater than zero if S1 is lexicographically less + than, equal to or greater than S2. */ + +int +strncasecmp (const char *s1, const char *s2, size_t n) +{ + register const unsigned char *p1 = (const unsigned char *) s1; + register const unsigned char *p2 = (const unsigned char *) s2; + unsigned char c1, c2; + + if (p1 == p2 || n == 0) + return 0; + + do + { + c1 = TOLOWER (*p1); + c2 = TOLOWER (*p2); + + if (--n == 0 || c1 == '\0') + break; + + ++p1; + ++p2; + } + while (c1 == c2); + + if (UCHAR_MAX <= INT_MAX) + return c1 - c2; + else + /* On machines where 'char' and 'int' are types of the same size, the + difference of two 'unsigned char' values - including the sign bit - + doesn't fit in an 'int'. */ + return (c1 > c2 ? 1 : c1 < c2 ? -1 : 0); +} diff --git a/gl/strndup.c b/gl/strndup.c new file mode 100644 index 00000000..290f494f --- /dev/null +++ b/gl/strndup.c @@ -0,0 +1,39 @@ +/* A replacement function, for systems that lack strndup. + + Copyright (C) 1996, 1997, 1998, 2001, 2002, 2003, 2005, 2006 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> +#include "strndup.h" + +#include <stdlib.h> +#include <string.h> + +#include "strnlen.h" + +char * +strndup (char const *s, size_t n) +{ + size_t len = strnlen (s, n); + char *new = malloc (len + 1); + + if (new == NULL) + return NULL; + + new[len] = '\0'; + return memcpy (new, s, len); +} diff --git a/gl/strndup.h b/gl/strndup.h new file mode 100644 index 00000000..b983dd22 --- /dev/null +++ b/gl/strndup.h @@ -0,0 +1,32 @@ +/* Duplicate a size-bounded string. + Copyright (C) 2003, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Get size_t. */ +#include <stddef.h> +/* If HAVE_STRNDUP, get the strndup declaration. + If !HAVE_STRNDUP, include <string.h> now so that it doesn't cause + trouble if included later. */ +#include <string.h> + +#if !HAVE_STRNDUP +# undef strndup +# define strndup rpl_strndup +# if !HAVE_DECL_STRNDUP /* Don't risk conflicting declarations. */ +/* Return a newly allocated copy of at most N bytes of STRING. */ +extern char *strndup (const char *string, size_t n); +# endif +#endif diff --git a/gl/strnlen.c b/gl/strnlen.c new file mode 100644 index 00000000..593fd1b7 --- /dev/null +++ b/gl/strnlen.c @@ -0,0 +1,31 @@ +/* Find the length of STRING, but scan at most MAXLEN characters. + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + Written by Simon Josefsson. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "strnlen.h" + +/* Find the length of STRING, but scan at most MAXLEN characters. + If no '\0' terminator is found in that many characters, return MAXLEN. */ + +size_t +strnlen (const char *string, size_t maxlen) +{ + const char *end = memchr (string, '\0', maxlen); + return end ? (size_t) (end - string) : maxlen; +} diff --git a/gl/strnlen.h b/gl/strnlen.h new file mode 100644 index 00000000..ba74dba7 --- /dev/null +++ b/gl/strnlen.h @@ -0,0 +1,32 @@ +/* Find the length of STRING, but scan at most MAXLEN characters. + Copyright (C) 2005 Free Software Foundation, Inc. + Written by Simon Josefsson. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef STRNLEN_H +#define STRNLEN_H + +/* Get strnlen declaration, if available. */ +#include <string.h> + +#if defined HAVE_DECL_STRNLEN && !HAVE_DECL_STRNLEN +/* Find the length (number of bytes) of STRING, but scan at most + MAXLEN bytes. If no '\0' terminator is found in that many bytes, + return MAXLEN. */ +extern size_t strnlen(const char *string, size_t maxlen); +#endif + +#endif /* STRNLEN_H */ diff --git a/gl/strnlen1.c b/gl/strnlen1.c new file mode 100644 index 00000000..422ed9e8 --- /dev/null +++ b/gl/strnlen1.c @@ -0,0 +1,36 @@ +/* Find the length of STRING + 1, but scan at most MAXLEN bytes. + Copyright (C) 2005-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include "strnlen1.h" + +#include <string.h> + +/* Find the length of STRING + 1, but scan at most MAXLEN bytes. + If no '\0' terminator is found in that many characters, return MAXLEN. */ +/* This is the same as strnlen (string, maxlen - 1) + 1. */ +size_t +strnlen1 (const char *string, size_t maxlen) +{ + const char *end = (const char *) memchr (string, '\0', maxlen); + if (end != NULL) + return end - string + 1; + else + return maxlen; +} diff --git a/gl/strnlen1.h b/gl/strnlen1.h new file mode 100644 index 00000000..7ce7d0c8 --- /dev/null +++ b/gl/strnlen1.h @@ -0,0 +1,40 @@ +/* Find the length of STRING + 1, but scan at most MAXLEN bytes. + Copyright (C) 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _STRNLEN1_H +#define _STRNLEN1_H + +#include <stddef.h> + + +#ifdef __cplusplus +extern "C" { +#endif + + +/* Find the length of STRING + 1, but scan at most MAXLEN bytes. + If no '\0' terminator is found in that many characters, return MAXLEN. */ +/* This is the same as strnlen (string, maxlen - 1) + 1. */ +extern size_t strnlen1 (const char *string, size_t maxlen); + + +#ifdef __cplusplus +} +#endif + + +#endif /* _STRNLEN1_H */ diff --git a/gl/unistd--.h b/gl/unistd--.h new file mode 100644 index 00000000..1fe6ce8b --- /dev/null +++ b/gl/unistd--.h @@ -0,0 +1,28 @@ +/* Like unistd.h, but redefine some names to avoid glitches. + + Copyright (C) 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <unistd.h> +#include "unistd-safer.h" + +#undef dup +#define dup dup_safer + +#undef pipe +#define pipe pipe_safer diff --git a/gl/unistd-safer.h b/gl/unistd-safer.h new file mode 100644 index 00000000..f95999d3 --- /dev/null +++ b/gl/unistd-safer.h @@ -0,0 +1,23 @@ +/* Invoke unistd-like functions, but avoid some glitches. + + Copyright (C) 2001, 2003, 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +int dup_safer (int); +int fd_safer (int); +int pipe_safer (int[2]); diff --git a/gl/unistd_.h b/gl/unistd_.h new file mode 100644 index 00000000..36fa6731 --- /dev/null +++ b/gl/unistd_.h @@ -0,0 +1,52 @@ +/* Substitute for and wrapper around <unistd.h>. + Copyright (C) 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _GL_UNISTD_H +#define _GL_UNISTD_H + +#if HAVE_UNISTD_H +# include @ABSOLUTE_UNISTD_H@ +#endif + + +/* Declare overridden functions. */ + +#ifdef __cplusplus +extern "C" { +#endif + +#ifdef FCHDIR_REPLACEMENT + +/* Change the process' current working directory to the directory on which + the given file descriptor is open. */ +extern int fchdir (int /*fd*/); + +# define close rpl_close +extern int close (int); +# define dup rpl_dup +extern int dup (int); +# define dup2 rpl_dup2 +extern int dup2 (int, int); + +#endif + +#ifdef __cplusplus +} +#endif + + +#endif /* _GL_UNISTD_H */ diff --git a/gl/vasnprintf.c b/gl/vasnprintf.c new file mode 100644 index 00000000..0fe2aada --- /dev/null +++ b/gl/vasnprintf.c @@ -0,0 +1,889 @@ +/* vsprintf with automatic memory allocation. + Copyright (C) 1999, 2002-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Tell glibc's <stdio.h> to provide a prototype for snprintf(). + This must come before <config.h> because <config.h> may include + <features.h>, and once <features.h> has been included, it's too late. */ +#ifndef _GNU_SOURCE +# define _GNU_SOURCE 1 +#endif + +#include <config.h> +#ifndef IN_LIBINTL +# include <alloca.h> +#endif + +/* Specification. */ +#if WIDE_CHAR_VERSION +# include "vasnwprintf.h" +#else +# include "vasnprintf.h" +#endif + +#include <stdio.h> /* snprintf(), sprintf() */ +#include <stdlib.h> /* abort(), malloc(), realloc(), free() */ +#include <string.h> /* memcpy(), strlen() */ +#include <errno.h> /* errno */ +#include <limits.h> /* CHAR_BIT */ +#include <float.h> /* DBL_MAX_EXP, LDBL_MAX_EXP */ +#if WIDE_CHAR_VERSION +# include "wprintf-parse.h" +#else +# include "printf-parse.h" +#endif + +/* Checked size_t computations. */ +#include "xsize.h" + +#ifdef HAVE_WCHAR_T +# ifdef HAVE_WCSLEN +# define local_wcslen wcslen +# else + /* Solaris 2.5.1 has wcslen() in a separate library libw.so. To avoid + a dependency towards this library, here is a local substitute. + Define this substitute only once, even if this file is included + twice in the same compilation unit. */ +# ifndef local_wcslen_defined +# define local_wcslen_defined 1 +static size_t +local_wcslen (const wchar_t *s) +{ + const wchar_t *ptr; + + for (ptr = s; *ptr != (wchar_t) 0; ptr++) + ; + return ptr - s; +} +# endif +# endif +#endif + +#if WIDE_CHAR_VERSION +# define VASNPRINTF vasnwprintf +# define CHAR_T wchar_t +# define DIRECTIVE wchar_t_directive +# define DIRECTIVES wchar_t_directives +# define PRINTF_PARSE wprintf_parse +# define USE_SNPRINTF 1 +# if HAVE_DECL__SNWPRINTF + /* On Windows, the function swprintf() has a different signature than + on Unix; we use the _snwprintf() function instead. */ +# define SNPRINTF _snwprintf +# else + /* Unix. */ +# define SNPRINTF swprintf +# endif +#else +# define VASNPRINTF vasnprintf +# define CHAR_T char +# define DIRECTIVE char_directive +# define DIRECTIVES char_directives +# define PRINTF_PARSE printf_parse +# define USE_SNPRINTF (HAVE_DECL__SNPRINTF || HAVE_SNPRINTF) +# if HAVE_DECL__SNPRINTF + /* Windows. */ +# define SNPRINTF _snprintf +# else + /* Unix. */ +# define SNPRINTF snprintf +# endif +#endif + +CHAR_T * +VASNPRINTF (CHAR_T *resultbuf, size_t *lengthp, const CHAR_T *format, va_list args) +{ + DIRECTIVES d; + arguments a; + + if (PRINTF_PARSE (format, &d, &a) < 0) + { + errno = EINVAL; + return NULL; + } + +#define CLEANUP() \ + free (d.dir); \ + if (a.arg) \ + free (a.arg); + + if (printf_fetchargs (args, &a) < 0) + { + CLEANUP (); + errno = EINVAL; + return NULL; + } + + { + size_t buf_neededlength; + CHAR_T *buf; + CHAR_T *buf_malloced; + const CHAR_T *cp; + size_t i; + DIRECTIVE *dp; + /* Output string accumulator. */ + CHAR_T *result; + size_t allocated; + size_t length; + + /* Allocate a small buffer that will hold a directive passed to + sprintf or snprintf. */ + buf_neededlength = + xsum4 (7, d.max_width_length, d.max_precision_length, 6); +#if HAVE_ALLOCA + if (buf_neededlength < 4000 / sizeof (CHAR_T)) + { + buf = (CHAR_T *) alloca (buf_neededlength * sizeof (CHAR_T)); + buf_malloced = NULL; + } + else +#endif + { + size_t buf_memsize = xtimes (buf_neededlength, sizeof (CHAR_T)); + if (size_overflow_p (buf_memsize)) + goto out_of_memory_1; + buf = (CHAR_T *) malloc (buf_memsize); + if (buf == NULL) + goto out_of_memory_1; + buf_malloced = buf; + } + + if (resultbuf != NULL) + { + result = resultbuf; + allocated = *lengthp; + } + else + { + result = NULL; + allocated = 0; + } + length = 0; + /* Invariants: + result is either == resultbuf or == NULL or malloc-allocated. + If length > 0, then result != NULL. */ + + /* Ensures that allocated >= needed. Aborts through a jump to + out_of_memory if needed is SIZE_MAX or otherwise too big. */ +#define ENSURE_ALLOCATION(needed) \ + if ((needed) > allocated) \ + { \ + size_t memory_size; \ + CHAR_T *memory; \ + \ + allocated = (allocated > 0 ? xtimes (allocated, 2) : 12); \ + if ((needed) > allocated) \ + allocated = (needed); \ + memory_size = xtimes (allocated, sizeof (CHAR_T)); \ + if (size_overflow_p (memory_size)) \ + goto out_of_memory; \ + if (result == resultbuf || result == NULL) \ + memory = (CHAR_T *) malloc (memory_size); \ + else \ + memory = (CHAR_T *) realloc (result, memory_size); \ + if (memory == NULL) \ + goto out_of_memory; \ + if (result == resultbuf && length > 0) \ + memcpy (memory, result, length * sizeof (CHAR_T)); \ + result = memory; \ + } + + for (cp = format, i = 0, dp = &d.dir[0]; ; cp = dp->dir_end, i++, dp++) + { + if (cp != dp->dir_start) + { + size_t n = dp->dir_start - cp; + size_t augmented_length = xsum (length, n); + + ENSURE_ALLOCATION (augmented_length); + memcpy (result + length, cp, n * sizeof (CHAR_T)); + length = augmented_length; + } + if (i == d.count) + break; + + /* Execute a single directive. */ + if (dp->conversion == '%') + { + size_t augmented_length; + + if (!(dp->arg_index == ARG_NONE)) + abort (); + augmented_length = xsum (length, 1); + ENSURE_ALLOCATION (augmented_length); + result[length] = '%'; + length = augmented_length; + } + else + { + if (!(dp->arg_index != ARG_NONE)) + abort (); + + if (dp->conversion == 'n') + { + switch (a.arg[dp->arg_index].type) + { + case TYPE_COUNT_SCHAR_POINTER: + *a.arg[dp->arg_index].a.a_count_schar_pointer = length; + break; + case TYPE_COUNT_SHORT_POINTER: + *a.arg[dp->arg_index].a.a_count_short_pointer = length; + break; + case TYPE_COUNT_INT_POINTER: + *a.arg[dp->arg_index].a.a_count_int_pointer = length; + break; + case TYPE_COUNT_LONGINT_POINTER: + *a.arg[dp->arg_index].a.a_count_longint_pointer = length; + break; +#ifdef HAVE_LONG_LONG_INT + case TYPE_COUNT_LONGLONGINT_POINTER: + *a.arg[dp->arg_index].a.a_count_longlongint_pointer = length; + break; +#endif + default: + abort (); + } + } + else + { + arg_type type = a.arg[dp->arg_index].type; + CHAR_T *p; + unsigned int prefix_count; + int prefixes[2]; +#if !USE_SNPRINTF + size_t tmp_length; + CHAR_T tmpbuf[700]; + CHAR_T *tmp; + + /* Allocate a temporary buffer of sufficient size for calling + sprintf. */ + { + size_t width; + size_t precision; + + width = 0; + if (dp->width_start != dp->width_end) + { + if (dp->width_arg_index != ARG_NONE) + { + int arg; + + if (!(a.arg[dp->width_arg_index].type == TYPE_INT)) + abort (); + arg = a.arg[dp->width_arg_index].a.a_int; + width = (arg < 0 ? (unsigned int) (-arg) : arg); + } + else + { + const CHAR_T *digitp = dp->width_start; + + do + width = xsum (xtimes (width, 10), *digitp++ - '0'); + while (digitp != dp->width_end); + } + } + + precision = 6; + if (dp->precision_start != dp->precision_end) + { + if (dp->precision_arg_index != ARG_NONE) + { + int arg; + + if (!(a.arg[dp->precision_arg_index].type == TYPE_INT)) + abort (); + arg = a.arg[dp->precision_arg_index].a.a_int; + precision = (arg < 0 ? 0 : arg); + } + else + { + const CHAR_T *digitp = dp->precision_start + 1; + + precision = 0; + while (digitp != dp->precision_end) + precision = xsum (xtimes (precision, 10), *digitp++ - '0'); + } + } + + switch (dp->conversion) + { + + case 'd': case 'i': case 'u': +# ifdef HAVE_LONG_LONG_INT + if (type == TYPE_LONGLONGINT || type == TYPE_ULONGLONGINT) + tmp_length = + (unsigned int) (sizeof (unsigned long long) * CHAR_BIT + * 0.30103 /* binary -> decimal */ + ) + + 1; /* turn floor into ceil */ + else +# endif + if (type == TYPE_LONGINT || type == TYPE_ULONGINT) + tmp_length = + (unsigned int) (sizeof (unsigned long) * CHAR_BIT + * 0.30103 /* binary -> decimal */ + ) + + 1; /* turn floor into ceil */ + else + tmp_length = + (unsigned int) (sizeof (unsigned int) * CHAR_BIT + * 0.30103 /* binary -> decimal */ + ) + + 1; /* turn floor into ceil */ + if (tmp_length < precision) + tmp_length = precision; + /* Multiply by 2, as an estimate for FLAG_GROUP. */ + tmp_length = xsum (tmp_length, tmp_length); + /* Add 1, to account for a leading sign. */ + tmp_length = xsum (tmp_length, 1); + break; + + case 'o': +# ifdef HAVE_LONG_LONG_INT + if (type == TYPE_LONGLONGINT || type == TYPE_ULONGLONGINT) + tmp_length = + (unsigned int) (sizeof (unsigned long long) * CHAR_BIT + * 0.333334 /* binary -> octal */ + ) + + 1; /* turn floor into ceil */ + else +# endif + if (type == TYPE_LONGINT || type == TYPE_ULONGINT) + tmp_length = + (unsigned int) (sizeof (unsigned long) * CHAR_BIT + * 0.333334 /* binary -> octal */ + ) + + 1; /* turn floor into ceil */ + else + tmp_length = + (unsigned int) (sizeof (unsigned int) * CHAR_BIT + * 0.333334 /* binary -> octal */ + ) + + 1; /* turn floor into ceil */ + if (tmp_length < precision) + tmp_length = precision; + /* Add 1, to account for a leading sign. */ + tmp_length = xsum (tmp_length, 1); + break; + + case 'x': case 'X': +# ifdef HAVE_LONG_LONG_INT + if (type == TYPE_LONGLONGINT || type == TYPE_ULONGLONGINT) + tmp_length = + (unsigned int) (sizeof (unsigned long long) * CHAR_BIT + * 0.25 /* binary -> hexadecimal */ + ) + + 1; /* turn floor into ceil */ + else +# endif + if (type == TYPE_LONGINT || type == TYPE_ULONGINT) + tmp_length = + (unsigned int) (sizeof (unsigned long) * CHAR_BIT + * 0.25 /* binary -> hexadecimal */ + ) + + 1; /* turn floor into ceil */ + else + tmp_length = + (unsigned int) (sizeof (unsigned int) * CHAR_BIT + * 0.25 /* binary -> hexadecimal */ + ) + + 1; /* turn floor into ceil */ + if (tmp_length < precision) + tmp_length = precision; + /* Add 2, to account for a leading sign or alternate form. */ + tmp_length = xsum (tmp_length, 2); + break; + + case 'f': case 'F': +# ifdef HAVE_LONG_DOUBLE + if (type == TYPE_LONGDOUBLE) + tmp_length = + (unsigned int) (LDBL_MAX_EXP + * 0.30103 /* binary -> decimal */ + * 2 /* estimate for FLAG_GROUP */ + ) + + 1 /* turn floor into ceil */ + + 10; /* sign, decimal point etc. */ + else +# endif + tmp_length = + (unsigned int) (DBL_MAX_EXP + * 0.30103 /* binary -> decimal */ + * 2 /* estimate for FLAG_GROUP */ + ) + + 1 /* turn floor into ceil */ + + 10; /* sign, decimal point etc. */ + tmp_length = xsum (tmp_length, precision); + break; + + case 'e': case 'E': case 'g': case 'G': + case 'a': case 'A': + tmp_length = + 12; /* sign, decimal point, exponent etc. */ + tmp_length = xsum (tmp_length, precision); + break; + + case 'c': +# if defined HAVE_WINT_T && !WIDE_CHAR_VERSION + if (type == TYPE_WIDE_CHAR) + tmp_length = MB_CUR_MAX; + else +# endif + tmp_length = 1; + break; + + case 's': +# ifdef HAVE_WCHAR_T + if (type == TYPE_WIDE_STRING) + { + tmp_length = + local_wcslen (a.arg[dp->arg_index].a.a_wide_string); + +# if !WIDE_CHAR_VERSION + tmp_length = xtimes (tmp_length, MB_CUR_MAX); +# endif + } + else +# endif + tmp_length = strlen (a.arg[dp->arg_index].a.a_string); + break; + + case 'p': + tmp_length = + (unsigned int) (sizeof (void *) * CHAR_BIT + * 0.25 /* binary -> hexadecimal */ + ) + + 1 /* turn floor into ceil */ + + 2; /* account for leading 0x */ + break; + + default: + abort (); + } + + if (tmp_length < width) + tmp_length = width; + + tmp_length = xsum (tmp_length, 1); /* account for trailing NUL */ + } + + if (tmp_length <= sizeof (tmpbuf) / sizeof (CHAR_T)) + tmp = tmpbuf; + else + { + size_t tmp_memsize = xtimes (tmp_length, sizeof (CHAR_T)); + + if (size_overflow_p (tmp_memsize)) + /* Overflow, would lead to out of memory. */ + goto out_of_memory; + tmp = (CHAR_T *) malloc (tmp_memsize); + if (tmp == NULL) + /* Out of memory. */ + goto out_of_memory; + } +#endif + + /* Construct the format string for calling snprintf or + sprintf. */ + p = buf; + *p++ = '%'; + if (dp->flags & FLAG_GROUP) + *p++ = '\''; + if (dp->flags & FLAG_LEFT) + *p++ = '-'; + if (dp->flags & FLAG_SHOWSIGN) + *p++ = '+'; + if (dp->flags & FLAG_SPACE) + *p++ = ' '; + if (dp->flags & FLAG_ALT) + *p++ = '#'; + if (dp->flags & FLAG_ZERO) + *p++ = '0'; + if (dp->width_start != dp->width_end) + { + size_t n = dp->width_end - dp->width_start; + memcpy (p, dp->width_start, n * sizeof (CHAR_T)); + p += n; + } + if (dp->precision_start != dp->precision_end) + { + size_t n = dp->precision_end - dp->precision_start; + memcpy (p, dp->precision_start, n * sizeof (CHAR_T)); + p += n; + } + + switch (type) + { +#ifdef HAVE_LONG_LONG_INT + case TYPE_LONGLONGINT: + case TYPE_ULONGLONGINT: + *p++ = 'l'; + /*FALLTHROUGH*/ +#endif + case TYPE_LONGINT: + case TYPE_ULONGINT: +#ifdef HAVE_WINT_T + case TYPE_WIDE_CHAR: +#endif +#ifdef HAVE_WCHAR_T + case TYPE_WIDE_STRING: +#endif + *p++ = 'l'; + break; +#ifdef HAVE_LONG_DOUBLE + case TYPE_LONGDOUBLE: + *p++ = 'L'; + break; +#endif + default: + break; + } + *p = dp->conversion; +#if USE_SNPRINTF + p[1] = '%'; + p[2] = 'n'; + p[3] = '\0'; +#else + p[1] = '\0'; +#endif + + /* Construct the arguments for calling snprintf or sprintf. */ + prefix_count = 0; + if (dp->width_arg_index != ARG_NONE) + { + if (!(a.arg[dp->width_arg_index].type == TYPE_INT)) + abort (); + prefixes[prefix_count++] = a.arg[dp->width_arg_index].a.a_int; + } + if (dp->precision_arg_index != ARG_NONE) + { + if (!(a.arg[dp->precision_arg_index].type == TYPE_INT)) + abort (); + prefixes[prefix_count++] = a.arg[dp->precision_arg_index].a.a_int; + } + +#if USE_SNPRINTF + /* Prepare checking whether snprintf returns the count + via %n. */ + ENSURE_ALLOCATION (xsum (length, 1)); + result[length] = '\0'; +#endif + + for (;;) + { + size_t maxlen; + int count; + int retcount; + + maxlen = allocated - length; + count = -1; + retcount = 0; + +#if USE_SNPRINTF +# define SNPRINTF_BUF(arg) \ + switch (prefix_count) \ + { \ + case 0: \ + retcount = SNPRINTF (result + length, maxlen, buf, \ + arg, &count); \ + break; \ + case 1: \ + retcount = SNPRINTF (result + length, maxlen, buf, \ + prefixes[0], arg, &count); \ + break; \ + case 2: \ + retcount = SNPRINTF (result + length, maxlen, buf, \ + prefixes[0], prefixes[1], arg, \ + &count); \ + break; \ + default: \ + abort (); \ + } +#else +# define SNPRINTF_BUF(arg) \ + switch (prefix_count) \ + { \ + case 0: \ + count = sprintf (tmp, buf, arg); \ + break; \ + case 1: \ + count = sprintf (tmp, buf, prefixes[0], arg); \ + break; \ + case 2: \ + count = sprintf (tmp, buf, prefixes[0], prefixes[1],\ + arg); \ + break; \ + default: \ + abort (); \ + } +#endif + + switch (type) + { + case TYPE_SCHAR: + { + int arg = a.arg[dp->arg_index].a.a_schar; + SNPRINTF_BUF (arg); + } + break; + case TYPE_UCHAR: + { + unsigned int arg = a.arg[dp->arg_index].a.a_uchar; + SNPRINTF_BUF (arg); + } + break; + case TYPE_SHORT: + { + int arg = a.arg[dp->arg_index].a.a_short; + SNPRINTF_BUF (arg); + } + break; + case TYPE_USHORT: + { + unsigned int arg = a.arg[dp->arg_index].a.a_ushort; + SNPRINTF_BUF (arg); + } + break; + case TYPE_INT: + { + int arg = a.arg[dp->arg_index].a.a_int; + SNPRINTF_BUF (arg); + } + break; + case TYPE_UINT: + { + unsigned int arg = a.arg[dp->arg_index].a.a_uint; + SNPRINTF_BUF (arg); + } + break; + case TYPE_LONGINT: + { + long int arg = a.arg[dp->arg_index].a.a_longint; + SNPRINTF_BUF (arg); + } + break; + case TYPE_ULONGINT: + { + unsigned long int arg = a.arg[dp->arg_index].a.a_ulongint; + SNPRINTF_BUF (arg); + } + break; +#ifdef HAVE_LONG_LONG_INT + case TYPE_LONGLONGINT: + { + long long int arg = a.arg[dp->arg_index].a.a_longlongint; + SNPRINTF_BUF (arg); + } + break; + case TYPE_ULONGLONGINT: + { + unsigned long long int arg = a.arg[dp->arg_index].a.a_ulonglongint; + SNPRINTF_BUF (arg); + } + break; +#endif + case TYPE_DOUBLE: + { + double arg = a.arg[dp->arg_index].a.a_double; + SNPRINTF_BUF (arg); + } + break; +#ifdef HAVE_LONG_DOUBLE + case TYPE_LONGDOUBLE: + { + long double arg = a.arg[dp->arg_index].a.a_longdouble; + SNPRINTF_BUF (arg); + } + break; +#endif + case TYPE_CHAR: + { + int arg = a.arg[dp->arg_index].a.a_char; + SNPRINTF_BUF (arg); + } + break; +#ifdef HAVE_WINT_T + case TYPE_WIDE_CHAR: + { + wint_t arg = a.arg[dp->arg_index].a.a_wide_char; + SNPRINTF_BUF (arg); + } + break; +#endif + case TYPE_STRING: + { + const char *arg = a.arg[dp->arg_index].a.a_string; + SNPRINTF_BUF (arg); + } + break; +#ifdef HAVE_WCHAR_T + case TYPE_WIDE_STRING: + { + const wchar_t *arg = a.arg[dp->arg_index].a.a_wide_string; + SNPRINTF_BUF (arg); + } + break; +#endif + case TYPE_POINTER: + { + void *arg = a.arg[dp->arg_index].a.a_pointer; + SNPRINTF_BUF (arg); + } + break; + default: + abort (); + } + +#if USE_SNPRINTF + /* Portability: Not all implementations of snprintf() + are ISO C 99 compliant. Determine the number of + bytes that snprintf() has produced or would have + produced. */ + if (count >= 0) + { + /* Verify that snprintf() has NUL-terminated its + result. */ + if (count < maxlen && result[length + count] != '\0') + abort (); + /* Portability hack. */ + if (retcount > count) + count = retcount; + } + else + { + /* snprintf() doesn't understand the '%n' + directive. */ + if (p[1] != '\0') + { + /* Don't use the '%n' directive; instead, look + at the snprintf() return value. */ + p[1] = '\0'; + continue; + } + else + { + /* Look at the snprintf() return value. */ + if (retcount < 0) + { + /* HP-UX 10.20 snprintf() is doubly deficient: + It doesn't understand the '%n' directive, + *and* it returns -1 (rather than the length + that would have been required) when the + buffer is too small. */ + size_t bigger_need = + xsum (xtimes (allocated, 2), 12); + ENSURE_ALLOCATION (bigger_need); + continue; + } + else + count = retcount; + } + } +#endif + + /* Attempt to handle failure. */ + if (count < 0) + { + if (!(result == resultbuf || result == NULL)) + free (result); + if (buf_malloced != NULL) + free (buf_malloced); + CLEANUP (); + errno = EINVAL; + return NULL; + } + +#if !USE_SNPRINTF + if (count >= tmp_length) + /* tmp_length was incorrectly calculated - fix the + code above! */ + abort (); +#endif + + /* Make room for the result. */ + if (count >= maxlen) + { + /* Need at least count bytes. But allocate + proportionally, to avoid looping eternally if + snprintf() reports a too small count. */ + size_t n = + xmax (xsum (length, count), xtimes (allocated, 2)); + + ENSURE_ALLOCATION (n); +#if USE_SNPRINTF + continue; +#endif + } + +#if USE_SNPRINTF + /* The snprintf() result did fit. */ +#else + /* Append the sprintf() result. */ + memcpy (result + length, tmp, count * sizeof (CHAR_T)); + if (tmp != tmpbuf) + free (tmp); +#endif + + length += count; + break; + } + } + } + } + + /* Add the final NUL. */ + ENSURE_ALLOCATION (xsum (length, 1)); + result[length] = '\0'; + + if (result != resultbuf && length + 1 < allocated) + { + /* Shrink the allocated memory if possible. */ + CHAR_T *memory; + + memory = (CHAR_T *) realloc (result, (length + 1) * sizeof (CHAR_T)); + if (memory != NULL) + result = memory; + } + + if (buf_malloced != NULL) + free (buf_malloced); + CLEANUP (); + *lengthp = length; + /* Note that we can produce a big string of a length > INT_MAX. POSIX + says that snprintf() fails with errno = EOVERFLOW in this case, but + that's only because snprintf() returns an 'int'. This function does + not have this limitation. */ + return result; + + out_of_memory: + if (!(result == resultbuf || result == NULL)) + free (result); + if (buf_malloced != NULL) + free (buf_malloced); + out_of_memory_1: + CLEANUP (); + errno = ENOMEM; + return NULL; + } +} + +#undef SNPRINTF +#undef USE_SNPRINTF +#undef PRINTF_PARSE +#undef DIRECTIVES +#undef DIRECTIVE +#undef CHAR_T +#undef VASNPRINTF diff --git a/gl/vasnprintf.h b/gl/vasnprintf.h new file mode 100644 index 00000000..894008ca --- /dev/null +++ b/gl/vasnprintf.h @@ -0,0 +1,77 @@ +/* vsprintf with automatic memory allocation. + Copyright (C) 2002-2004 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _VASNPRINTF_H +#define _VASNPRINTF_H + +/* Get va_list. */ +#include <stdarg.h> + +/* Get size_t. */ +#include <stddef.h> + +#ifndef __attribute__ +/* This feature is available in gcc versions 2.5 and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 5) || __STRICT_ANSI__ +# define __attribute__(Spec) /* empty */ +# endif +/* The __-protected variants of `format' and `printf' attributes + are accepted by gcc versions 2.6.4 (effectively 2.7) and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 7) +# define __format__ format +# define __printf__ printf +# endif +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +/* Write formatted output to a string dynamically allocated with malloc(). + You can pass a preallocated buffer for the result in RESULTBUF and its + size in *LENGTHP; otherwise you pass RESULTBUF = NULL. + If successful, return the address of the string (this may be = RESULTBUF + if no dynamic memory allocation was necessary) and set *LENGTHP to the + number of resulting bytes, excluding the trailing NUL. Upon error, set + errno and return NULL. + + When dynamic memory allocation occurs, the preallocated buffer is left + alone (with possibly modified contents). This makes it possible to use + a statically allocated or stack-allocated buffer, like this: + + char buf[100]; + size_t len = sizeof (buf); + char *output = vasnprintf (buf, &len, format, args); + if (output == NULL) + ... error handling ...; + else + { + ... use the output string ...; + if (output != buf) + free (output); + } + */ +extern char * asnprintf (char *resultbuf, size_t *lengthp, const char *format, ...) + __attribute__ ((__format__ (__printf__, 3, 4))); +extern char * vasnprintf (char *resultbuf, size_t *lengthp, const char *format, va_list args) + __attribute__ ((__format__ (__printf__, 3, 0))); + +#ifdef __cplusplus +} +#endif + +#endif /* _VASNPRINTF_H */ diff --git a/gl/vasprintf.c b/gl/vasprintf.c new file mode 100644 index 00000000..82470734 --- /dev/null +++ b/gl/vasprintf.c @@ -0,0 +1,52 @@ +/* Formatted output to strings. + Copyright (C) 1999, 2002, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include "vasprintf.h" + +#include <errno.h> +#include <limits.h> +#include <stdlib.h> + +#include "vasnprintf.h" + +/* Some systems, like OSF/1 4.0 and Woe32, don't have EOVERFLOW. */ +#ifndef EOVERFLOW +# define EOVERFLOW E2BIG +#endif + +int +vasprintf (char **resultp, const char *format, va_list args) +{ + size_t length; + char *result = vasnprintf (NULL, &length, format, args); + if (result == NULL) + return -1; + + if (length > INT_MAX) + { + free (result); + errno = EOVERFLOW; + return -1; + } + + *resultp = result; + /* Return the number of resulting bytes, excluding the trailing NUL. */ + return length; +} diff --git a/gl/vasprintf.h b/gl/vasprintf.h new file mode 100644 index 00000000..ab5515ce --- /dev/null +++ b/gl/vasprintf.h @@ -0,0 +1,63 @@ +/* vsprintf with automatic memory allocation. + Copyright (C) 2002-2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _VASPRINTF_H +#define _VASPRINTF_H + +#if HAVE_VASPRINTF + +/* Get asprintf(), vasprintf() declarations. */ +#include <stdio.h> + +#else + +/* Get va_list. */ +#include <stdarg.h> + +#ifndef __attribute__ +/* This feature is available in gcc versions 2.5 and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 5) || __STRICT_ANSI__ +# define __attribute__(Spec) /* empty */ +# endif +/* The __-protected variants of `format' and `printf' attributes + are accepted by gcc versions 2.6.4 (effectively 2.7) and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 7) +# define __format__ format +# define __printf__ printf +# endif +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +/* Write formatted output to a string dynamically allocated with malloc(). + If the memory allocation succeeds, store the address of the string in + *RESULT and return the number of resulting bytes, excluding the trailing + NUL. Upon memory allocation error, or some other error, return -1. */ +extern int asprintf (char **result, const char *format, ...) + __attribute__ ((__format__ (__printf__, 2, 3))); +extern int vasprintf (char **result, const char *format, va_list args) + __attribute__ ((__format__ (__printf__, 2, 0))); + +#ifdef __cplusplus +} +#endif + +#endif + +#endif /* _VASPRINTF_H */ diff --git a/gl/vsnprintf.c b/gl/vsnprintf.c new file mode 100644 index 00000000..af567d2c --- /dev/null +++ b/gl/vsnprintf.c @@ -0,0 +1,77 @@ +/* Formatted output to strings. + Copyright (C) 2004, 2006 Free Software Foundation, Inc. + Written by Simon Josefsson and Yoann Vandoorselaere <yoann@prelude-ids.org>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +/* Specification. */ +#include "vsnprintf.h" + +#include <errno.h> +#include <limits.h> +#include <stdarg.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +#include "vasnprintf.h" + +/* Some systems, like OSF/1 4.0 and Woe32, don't have EOVERFLOW. */ +#ifndef EOVERFLOW +# define EOVERFLOW E2BIG +#endif + +/* Print formatted output to string STR. Similar to vsprintf, but + additional length SIZE limit how much is written into STR. Returns + string length of formatted string (which may be larger than SIZE). + STR may be NULL, in which case nothing will be written. On error, + return a negative value. */ +int +vsnprintf (char *str, size_t size, const char *format, va_list args) +{ + char *output; + size_t len; + size_t lenbuf = size; + + output = vasnprintf (str, &lenbuf, format, args); + len = lenbuf; + + if (!output) + return -1; + + if (output != str) + { + if (size) + { + size_t pruned_len = (len < size ? len : size - 1); + memcpy (str, output, pruned_len); + str[pruned_len] = '\0'; + } + + free (output); + } + + if (len > INT_MAX) + { + errno = EOVERFLOW; + return -1; + } + + return len; +} diff --git a/gl/vsnprintf.h b/gl/vsnprintf.h new file mode 100644 index 00000000..f80c77a9 --- /dev/null +++ b/gl/vsnprintf.h @@ -0,0 +1,31 @@ +/* Formatted output to strings. + Copyright (C) 2004 Free Software Foundation, Inc. + Written by Simon Josefsson and Yoann Vandoorselaere <yoann@prelude-ids.org>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef VSNPRINTF_H +#define VSNPRINTF_H + +#include <stdarg.h> + +/* Get vsnprintf declaration, if available. */ +#include <stdio.h> + +#if defined HAVE_DECL_VSNPRINTF && !HAVE_DECL_VSNPRINTF +int vsnprintf (char *str, size_t size, const char *format, va_list args); +#endif + +#endif /* VSNPRINTF_H */ diff --git a/gl/wchar_.h b/gl/wchar_.h new file mode 100644 index 00000000..6813a211 --- /dev/null +++ b/gl/wchar_.h @@ -0,0 +1,42 @@ +/* A substitute for ISO C99 <wchar.h>, for platforms that have issues. + + Copyright (C) 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Eric Blake. */ + +/* + * ISO C 99 <wchar.h> for platforms that have issues. + * <http://www.opengroup.org/susv3xbd/wchar.h.html> + * + * For now, this just ensures proper prerequisite inclusion order. + */ + +#ifndef _GL_WCHAR_H +#define _GL_WCHAR_H + +/* Tru64 with Desktop Toolkit C has a bug: <stdio.h> must be included before + <wchar.h>. + BSD/OS 4.0.1 has a bug: <stddef.h>, <stdio.h> and <time.h> must be + included before <wchar.h>. */ +#include <stddef.h> +#include <stdio.h> +#include <time.h> + +/* Include the original <wchar.h>. */ +#include @ABSOLUTE_WCHAR_H@ + +#endif /* _GL_WCHAR_H */ diff --git a/gl/wctype_.h b/gl/wctype_.h new file mode 100644 index 00000000..1297c61e --- /dev/null +++ b/gl/wctype_.h @@ -0,0 +1,154 @@ +/* A substitute for ISO C99 <wctype.h>, for platforms that lack it. + + Copyright (C) 2006, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Bruno Haible and Paul Eggert. */ + +/* + * ISO C 99 <wctype.h> for platforms that lack it. + * <http://www.opengroup.org/susv3xbd/wctype.h.html> + * + * iswctype, towctrans, towlower, towupper, wctrans, wctype, + * wctrans_t, and wctype_t are not yet implemented. + */ + +#ifndef _GL_WCTYPE_H +#define _GL_WCTYPE_H + +#if @HAVE_WINT_T@ +/* Solaris 2.5 has a bug: <wchar.h> must be included before <wctype.h>. + Tru64 with Desktop Toolkit C has a bug: <stdio.h> must be included before + <wchar.h>. + BSD/OS 4.0.1 has a bug: <stddef.h>, <stdio.h> and <time.h> must be + included before <wchar.h>. */ +# include <stddef.h> +# include <stdio.h> +# include <time.h> +# include <wchar.h> +typedef wint_t __wctype_wint_t; +#else +typedef int __wctype_wint_t; +#endif + +/* Include the original <wctype.h> if it exists. + BeOS 5 has the functions but no <wctype.h>. */ +#if @HAVE_WCTYPE_H@ +# include @ABSOLUTE_WCTYPE_H@ +#endif + +/* FreeBSD 4.4 to 4.11 has <wctype.h> but lacks the functions. + Assume all 12 functions are implemented the same way, or not at all. */ +#if ! HAVE_ISWCNTRL + +/* IRIX 5.3 has macros but no functions, its isw* macros refer to an + undefined variable _ctmp_ and to <ctype.h> macros like _P, and they + refer to system functions like _iswctype that are not in the + standard C library. Rather than try to get ancient buggy + implementations like this to work, just disable them. */ +# undef iswalnum +# undef iswalpha +# undef iswblank +# undef iswcntrl +# undef iswdigit +# undef iswgraph +# undef iswlower +# undef iswprint +# undef iswpunct +# undef iswspace +# undef iswupper +# undef iswxdigit + +static inline int +iswalnum (__wctype_wint_t wc) +{ + return ((wc >= '0' && wc <= '9') + || ((wc & ~0x20) >= 'A' && (wc & ~0x20) <= 'Z')); +} + +static inline int +iswalpha (__wctype_wint_t wc) +{ + return (wc & ~0x20) >= 'A' && (wc & ~0x20) <= 'Z'; +} + +static inline int +iswblank (__wctype_wint_t wc) +{ + return wc == ' ' || wc == '\t'; +} + +static inline int +iswcntrl (__wctype_wint_t wc) +{ + return (wc & ~0x1f) == 0 || wc == 0x7f; +} + +static inline int +iswdigit (__wctype_wint_t wc) +{ + return wc >= '0' && wc <= '9'; +} + +static inline int +iswgraph (__wctype_wint_t wc) +{ + return wc >= '!' && wc <= '~'; +} + +static inline int +iswlower (__wctype_wint_t wc) +{ + return wc >= 'a' && wc <= 'z'; +} + +static inline int +iswprint (__wctype_wint_t wc) +{ + return wc >= ' ' && wc <= '~'; +} + +static inline int +iswpunct (__wctype_wint_t wc) +{ + return (wc >= '!' && wc <= '~' + && !((wc >= '0' && wc <= '9') + || ((wc & ~0x20) >= 'A' && (wc & ~0x20) <= 'Z'))); +} + +static inline int +iswspace (__wctype_wint_t wc) +{ + return (wc == ' ' || wc == '\t' + || wc == '\n' || wc == '\v' || wc == '\f' || wc == '\r'); +} + +static inline int +iswupper (__wctype_wint_t wc) +{ + return wc >= 'A' && wc <= 'Z'; +} + +static inline int +iswxdigit (__wctype_wint_t wc) +{ + return ((wc >= '0' && wc <= '9') + || ((wc & ~0x20) >= 'A' && (wc & ~0x20) <= 'F')); +} + +# endif /* ! HAVE_ISWCNTRL */ + +#endif /* _GL_WCTYPE_H */ diff --git a/gl/wcwidth.h b/gl/wcwidth.h new file mode 100644 index 00000000..8ed5ff8c --- /dev/null +++ b/gl/wcwidth.h @@ -0,0 +1,57 @@ +/* Determine the number of screen columns needed for a character. + Copyright (C) 2006, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _gl_WCWIDTH_H +#define _gl_WCWIDTH_H + +#if HAVE_WCHAR_T + +/* Get wcwidth if available, along with wchar_t. */ +# include <wchar.h> + +/* Get iswprint. */ +# include <wctype.h> + +# ifndef HAVE_DECL_WCWIDTH +"this configure-time declaration test was not run" +# endif +# ifndef wcwidth +# if !HAVE_WCWIDTH + +/* wcwidth doesn't exist, so assume all printable characters have + width 1. */ +static inline int +wcwidth (wchar_t wc) +{ + return wc == 0 ? 0 : iswprint (wc) ? 1 : -1; +} + +# elif !HAVE_DECL_WCWIDTH + +/* wcwidth exists but is not declared. */ +extern +# ifdef __cplusplus +"C" +# endif +int wcwidth (int /* actually wchar_t */); + +# endif +# endif + +#endif /* HAVE_WCHAR_T */ + +#endif /* _gl_WCWIDTH_H */ diff --git a/gl/xalloc-die.c b/gl/xalloc-die.c new file mode 100644 index 00000000..090f060d --- /dev/null +++ b/gl/xalloc-die.c @@ -0,0 +1,42 @@ +/* Report a memory allocation failure and exit. + + Copyright (C) 1997, 1998, 1999, 2000, 2002, 2003, 2004, 2006 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "xalloc.h" + +#include <stdlib.h> + +#include "error.h" +#include "exitfail.h" + +#include "gettext.h" +#define _(msgid) gettext (msgid) + +void +xalloc_die (void) +{ + error (exit_failure, 0, "%s", _("memory exhausted")); + + /* The `noreturn' cannot be given to error, since it may return if + its first argument is 0. To help compilers understand the + xalloc_die does not return, call abort. Also, the abort is a + safety feature if exit_failure is 0 (which shouldn't happen). */ + abort (); +} diff --git a/gl/xalloc.h b/gl/xalloc.h new file mode 100644 index 00000000..17ab5142 --- /dev/null +++ b/gl/xalloc.h @@ -0,0 +1,267 @@ +/* xalloc.h -- malloc with out-of-memory checking + + Copyright (C) 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, + 1999, 2000, 2003, 2004, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef XALLOC_H_ +# define XALLOC_H_ + +# include <stddef.h> + + +# ifdef __cplusplus +extern "C" { +# endif + + +# ifndef __attribute__ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 8) || __STRICT_ANSI__ +# define __attribute__(x) +# endif +# endif + +# ifndef ATTRIBUTE_NORETURN +# define ATTRIBUTE_NORETURN __attribute__ ((__noreturn__)) +# endif + +/* This function is always triggered when memory is exhausted. + It must be defined by the application, either explicitly + or by using gnulib's xalloc-die module. This is the + function to call when one wants the program to die because of a + memory allocation failure. */ +extern void xalloc_die (void) ATTRIBUTE_NORETURN; + +void *xmalloc (size_t s); +void *xzalloc (size_t s); +void *xcalloc (size_t n, size_t s); +void *xrealloc (void *p, size_t s); +void *x2realloc (void *p, size_t *pn); +void *xmemdup (void const *p, size_t s); +char *xstrdup (char const *str); + +/* Return 1 if an array of N objects, each of size S, cannot exist due + to size arithmetic overflow. S must be positive and N must be + nonnegative. This is a macro, not an inline function, so that it + works correctly even when SIZE_MAX < N. + + By gnulib convention, SIZE_MAX represents overflow in size + calculations, so the conservative dividend to use here is + SIZE_MAX - 1, since SIZE_MAX might represent an overflowed value. + However, malloc (SIZE_MAX) fails on all known hosts where + sizeof (ptrdiff_t) <= sizeof (size_t), so do not bother to test for + exactly-SIZE_MAX allocations on such hosts; this avoids a test and + branch when S is known to be 1. */ +# define xalloc_oversized(n, s) \ + ((size_t) (sizeof (ptrdiff_t) <= sizeof (size_t) ? -1 : -2) / (s) < (n)) + + +/* In the following macros, T must be an elementary or structure/union or + typedef'ed type, or a pointer to such a type. To apply one of the + following macros to a function pointer or array type, you need to typedef + it first and use the typedef name. */ + +/* Allocate an object of type T dynamically, with error checking. */ +/* extern t *XMALLOC (typename t); */ +# define XMALLOC(t) ((t *) xmalloc (sizeof (t))) + +/* Allocate memory for N elements of type T, with error checking. */ +/* extern t *XNMALLOC (size_t n, typename t); */ +# define XNMALLOC(n, t) \ + ((t *) (sizeof (t) == 1 ? xmalloc (n) : xnmalloc (n, sizeof (t)))) + +/* Allocate an object of type T dynamically, with error checking, + and zero it. */ +/* extern t *XZALLOC (typename t); */ +# define XZALLOC(t) ((t *) xzalloc (sizeof (t))) + +/* Allocate memory for N elements of type T, with error checking, + and zero it. */ +/* extern t *XCALLOC (size_t n, typename t); */ +# define XCALLOC(n, t) \ + ((t *) (sizeof (t) == 1 ? xzalloc (n) : xcalloc (n, sizeof (t)))) + + +# if HAVE_INLINE +# define static_inline static inline +# else + void *xnmalloc (size_t n, size_t s); + void *xnrealloc (void *p, size_t n, size_t s); + void *x2nrealloc (void *p, size_t *pn, size_t s); + char *xcharalloc (size_t n); +# endif + +# ifdef static_inline + +/* Allocate an array of N objects, each with S bytes of memory, + dynamically, with error checking. S must be nonzero. */ + +static_inline void * +xnmalloc (size_t n, size_t s) +{ + if (xalloc_oversized (n, s)) + xalloc_die (); + return xmalloc (n * s); +} + +/* Change the size of an allocated block of memory P to an array of N + objects each of S bytes, with error checking. S must be nonzero. */ + +static_inline void * +xnrealloc (void *p, size_t n, size_t s) +{ + if (xalloc_oversized (n, s)) + xalloc_die (); + return xrealloc (p, n * s); +} + +/* If P is null, allocate a block of at least *PN such objects; + otherwise, reallocate P so that it contains more than *PN objects + each of S bytes. *PN must be nonzero unless P is null, and S must + be nonzero. Set *PN to the new number of objects, and return the + pointer to the new block. *PN is never set to zero, and the + returned pointer is never null. + + Repeated reallocations are guaranteed to make progress, either by + allocating an initial block with a nonzero size, or by allocating a + larger block. + + In the following implementation, nonzero sizes are doubled so that + repeated reallocations have O(N log N) overall cost rather than + O(N**2) cost, but the specification for this function does not + guarantee that sizes are doubled. + + Here is an example of use: + + int *p = NULL; + size_t used = 0; + size_t allocated = 0; + + void + append_int (int value) + { + if (used == allocated) + p = x2nrealloc (p, &allocated, sizeof *p); + p[used++] = value; + } + + This causes x2nrealloc to allocate a block of some nonzero size the + first time it is called. + + To have finer-grained control over the initial size, set *PN to a + nonzero value before calling this function with P == NULL. For + example: + + int *p = NULL; + size_t used = 0; + size_t allocated = 0; + size_t allocated1 = 1000; + + void + append_int (int value) + { + if (used == allocated) + { + p = x2nrealloc (p, &allocated1, sizeof *p); + allocated = allocated1; + } + p[used++] = value; + } + + */ + +static_inline void * +x2nrealloc (void *p, size_t *pn, size_t s) +{ + size_t n = *pn; + + if (! p) + { + if (! n) + { + /* The approximate size to use for initial small allocation + requests, when the invoking code specifies an old size of + zero. 64 bytes is the largest "small" request for the + GNU C library malloc. */ + enum { DEFAULT_MXFAST = 64 }; + + n = DEFAULT_MXFAST / s; + n += !n; + } + } + else + { + if (((size_t) -1) / 2 / s < n) + xalloc_die (); + n *= 2; + } + + *pn = n; + return xrealloc (p, n * s); +} + +/* Return a pointer to a new buffer of N bytes. This is like xmalloc, + except it returns char *. */ + +static_inline char * +xcharalloc (size_t n) +{ + return XNMALLOC (n, char); +} + +# endif + +# ifdef __cplusplus +} + +/* C++ does not allow conversions from void * to other pointer types + without a cast. Use templates to work around the problem when + possible. */ + +template <typename T> inline T * +xrealloc (T *p, size_t s) +{ + return (T *) xrealloc ((void *) p, s); +} + +template <typename T> inline T * +xnrealloc (T *p, size_t n, size_t s) +{ + return (T *) xnrealloc ((void *) p, n, s); +} + +template <typename T> inline T * +x2realloc (T *p, size_t *pn) +{ + return (T *) x2realloc ((void *) p, pn); +} + +template <typename T> inline T * +x2nrealloc (T *p, size_t *pn, size_t s) +{ + return (T *) x2nrealloc ((void *) p, pn, s); +} + +template <typename T> inline T * +xmemdup (T const *p, size_t s) +{ + return (T *) xmemdup ((void const *) p, s); +} + +# endif + + +#endif /* !XALLOC_H_ */ diff --git a/gl/xmalloc.c b/gl/xmalloc.c new file mode 100644 index 00000000..318e0ddb --- /dev/null +++ b/gl/xmalloc.c @@ -0,0 +1,123 @@ +/* xmalloc.c -- malloc with out of memory checking + + Copyright (C) 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, + 1999, 2000, 2002, 2003, 2004, 2005, 2006 Free Software Foundation, + Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#if ! HAVE_INLINE +# define static_inline +#endif +#include "xalloc.h" +#undef static_inline + +#include <stdlib.h> +#include <string.h> + +#ifndef SIZE_MAX +# define SIZE_MAX ((size_t) -1) +#endif + +/* 1 if calloc is known to be compatible with GNU calloc. This + matters if we are not also using the calloc module, which defines + HAVE_CALLOC and supports the GNU API even on non-GNU platforms. */ +#if defined HAVE_CALLOC || defined __GLIBC__ +enum { HAVE_GNU_CALLOC = 1 }; +#else +enum { HAVE_GNU_CALLOC = 0 }; +#endif + +/* Allocate N bytes of memory dynamically, with error checking. */ + +void * +xmalloc (size_t n) +{ + void *p = malloc (n); + if (!p && n != 0) + xalloc_die (); + return p; +} + +/* Change the size of an allocated block of memory P to N bytes, + with error checking. */ + +void * +xrealloc (void *p, size_t n) +{ + p = realloc (p, n); + if (!p && n != 0) + xalloc_die (); + return p; +} + +/* If P is null, allocate a block of at least *PN bytes; otherwise, + reallocate P so that it contains more than *PN bytes. *PN must be + nonzero unless P is null. Set *PN to the new block's size, and + return the pointer to the new block. *PN is never set to zero, and + the returned pointer is never null. */ + +void * +x2realloc (void *p, size_t *pn) +{ + return x2nrealloc (p, pn, 1); +} + +/* Allocate S bytes of zeroed memory dynamically, with error checking. + There's no need for xnzalloc (N, S), since it would be equivalent + to xcalloc (N, S). */ + +void * +xzalloc (size_t s) +{ + return memset (xmalloc (s), 0, s); +} + +/* Allocate zeroed memory for N elements of S bytes, with error + checking. S must be nonzero. */ + +void * +xcalloc (size_t n, size_t s) +{ + void *p; + /* Test for overflow, since some calloc implementations don't have + proper overflow checks. But omit overflow and size-zero tests if + HAVE_GNU_CALLOC, since GNU calloc catches overflow and never + returns NULL if successful. */ + if ((! HAVE_GNU_CALLOC && xalloc_oversized (n, s)) + || (! (p = calloc (n, s)) && (HAVE_GNU_CALLOC || n != 0))) + xalloc_die (); + return p; +} + +/* Clone an object P of size S, with error checking. There's no need + for xnmemdup (P, N, S), since xmemdup (P, N * S) works without any + need for an arithmetic overflow check. */ + +void * +xmemdup (void const *p, size_t s) +{ + return memcpy (xmalloc (s), p, s); +} + +/* Clone STRING. */ + +char * +xstrdup (char const *string) +{ + return xmemdup (string, strlen (string) + 1); +} diff --git a/gl/xsize.h b/gl/xsize.h new file mode 100644 index 00000000..341fb16c --- /dev/null +++ b/gl/xsize.h @@ -0,0 +1,108 @@ +/* xsize.h -- Checked size_t computations. + + Copyright (C) 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _XSIZE_H +#define _XSIZE_H + +/* Get size_t. */ +#include <stddef.h> + +/* Get SIZE_MAX. */ +#include <limits.h> +#if HAVE_STDINT_H +# include <stdint.h> +#endif + +/* The size of memory objects is often computed through expressions of + type size_t. Example: + void* p = malloc (header_size + n * element_size). + These computations can lead to overflow. When this happens, malloc() + returns a piece of memory that is way too small, and the program then + crashes while attempting to fill the memory. + To avoid this, the functions and macros in this file check for overflow. + The convention is that SIZE_MAX represents overflow. + malloc (SIZE_MAX) is not guaranteed to fail -- think of a malloc + implementation that uses mmap --, it's recommended to use size_overflow_p() + or size_in_bounds_p() before invoking malloc(). + The example thus becomes: + size_t size = xsum (header_size, xtimes (n, element_size)); + void *p = (size_in_bounds_p (size) ? malloc (size) : NULL); +*/ + +/* Convert an arbitrary value >= 0 to type size_t. */ +#define xcast_size_t(N) \ + ((N) <= SIZE_MAX ? (size_t) (N) : SIZE_MAX) + +/* Sum of two sizes, with overflow check. */ +static inline size_t +#if __GNUC__ >= 3 +__attribute__ ((__pure__)) +#endif +xsum (size_t size1, size_t size2) +{ + size_t sum = size1 + size2; + return (sum >= size1 ? sum : SIZE_MAX); +} + +/* Sum of three sizes, with overflow check. */ +static inline size_t +#if __GNUC__ >= 3 +__attribute__ ((__pure__)) +#endif +xsum3 (size_t size1, size_t size2, size_t size3) +{ + return xsum (xsum (size1, size2), size3); +} + +/* Sum of four sizes, with overflow check. */ +static inline size_t +#if __GNUC__ >= 3 +__attribute__ ((__pure__)) +#endif +xsum4 (size_t size1, size_t size2, size_t size3, size_t size4) +{ + return xsum (xsum (xsum (size1, size2), size3), size4); +} + +/* Maximum of two sizes, with overflow check. */ +static inline size_t +#if __GNUC__ >= 3 +__attribute__ ((__pure__)) +#endif +xmax (size_t size1, size_t size2) +{ + /* No explicit check is needed here, because for any n: + max (SIZE_MAX, n) == SIZE_MAX and max (n, SIZE_MAX) == SIZE_MAX. */ + return (size1 >= size2 ? size1 : size2); +} + +/* Multiplication of a count with an element size, with overflow check. + The count must be >= 0 and the element size must be > 0. + This is a macro, not an inline function, so that it works correctly even + when N is of a wider tupe and N > SIZE_MAX. */ +#define xtimes(N, ELSIZE) \ + ((N) <= SIZE_MAX / (ELSIZE) ? (size_t) (N) * (ELSIZE) : SIZE_MAX) + +/* Check for overflow. */ +#define size_overflow_p(SIZE) \ + ((SIZE) == SIZE_MAX) +/* Check against overflow. */ +#define size_in_bounds_p(SIZE) \ + ((SIZE) != SIZE_MAX) + +#endif /* _XSIZE_H */ diff --git a/gl/xstrndup.c b/gl/xstrndup.c new file mode 100644 index 00000000..afcbc1ab --- /dev/null +++ b/gl/xstrndup.c @@ -0,0 +1,37 @@ +/* Duplicate a bounded initial segment of a string, with out-of-memory + checking. + Copyright (C) 2003, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include "xstrndup.h" + +#include "strndup.h" +#include "xalloc.h" + +/* Return a newly allocated copy of at most N bytes of STRING. + In other words, return a copy of the initial segment of length N of + STRING. */ +char * +xstrndup (const char *string, size_t n) +{ + char *s = strndup (string, n); + if (! s) + xalloc_die (); + return s; +} diff --git a/gl/xstrndup.h b/gl/xstrndup.h new file mode 100644 index 00000000..88354cfd --- /dev/null +++ b/gl/xstrndup.h @@ -0,0 +1,24 @@ +/* Duplicate a bounded initial segment of a string, with out-of-memory + checking. + Copyright (C) 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <stddef.h> + +/* Return a newly allocated copy of at most N bytes of STRING. + In other words, return a copy of the initial segment of length N of + STRING. */ +extern char *xstrndup (const char *string, size_t n); |