diff options
Diffstat (limited to 'lib')
| -rw-r--r-- | lib/Makefile.am | 55 | ||||
| -rw-r--r-- | lib/Makefile.in | 136 | ||||
| -rw-r--r-- | lib/alloca.c | 492 | ||||
| -rw-r--r-- | lib/dirname.c | 70 | ||||
| -rw-r--r-- | lib/error.c | 119 | ||||
| -rw-r--r-- | lib/fileblocks.c | 66 | ||||
| -rw-r--r-- | lib/filemode.c | 237 | ||||
| -rw-r--r-- | lib/fnmatch.c | 200 | ||||
| -rw-r--r-- | lib/fnmatch.h | 67 | ||||
| -rw-r--r-- | lib/getline.c | 126 | ||||
| -rw-r--r-- | lib/getopt.c | 748 | ||||
| -rw-r--r-- | lib/getopt.h | 129 | ||||
| -rw-r--r-- | lib/getopt1.c | 180 | ||||
| -rw-r--r-- | lib/idcache.c | 210 | ||||
| -rw-r--r-- | lib/listfile.c | 370 | ||||
| -rw-r--r-- | lib/memcmp.c | 32 | ||||
| -rw-r--r-- | lib/memset.c | 29 | ||||
| -rw-r--r-- | lib/mktime.c | 502 | ||||
| -rw-r--r-- | lib/modechange.c | 341 | ||||
| -rw-r--r-- | lib/modechange.h | 55 | ||||
| -rw-r--r-- | lib/modetype.h | 81 | ||||
| -rw-r--r-- | lib/nextelem.c | 37 | ||||
| -rw-r--r-- | lib/pathmax.h | 53 | ||||
| -rw-r--r-- | lib/regex.c | 5244 | ||||
| -rw-r--r-- | lib/regex.h | 487 | ||||
| -rw-r--r-- | lib/savedir.c | 131 | ||||
| -rw-r--r-- | lib/stpcpy.c | 32 | ||||
| -rw-r--r-- | lib/strdup.c | 43 | ||||
| -rw-r--r-- | lib/strftime.c | 469 | ||||
| -rw-r--r-- | lib/strspn.c | 23 | ||||
| -rw-r--r-- | lib/strstr.c | 44 | ||||
| -rw-r--r-- | lib/strtol.c | 186 | ||||
| -rw-r--r-- | lib/wait.h | 15 | ||||
| -rw-r--r-- | lib/xgetcwd.c | 78 | ||||
| -rw-r--r-- | lib/xmalloc.c | 95 | ||||
| -rw-r--r-- | lib/xstrdup.c | 36 |
36 files changed, 10804 insertions, 414 deletions
diff --git a/lib/Makefile.am b/lib/Makefile.am index 8f8319a5..568061f3 100644 --- a/lib/Makefile.am +++ b/lib/Makefile.am @@ -1,44 +1,19 @@ -## Process this file with automake to produce Makefile.in. +LIBRARIES = find +find_SOURCES = regex.c dirname.c error.c filemode.c getopt.c getopt1.c \ +idcache.c listfile.c modechange.c nextelem.c savedir.c xmalloc.c getline.c \ +xstrdup.c xgetcwd.c fnmatch.c $(find_OPT_SOURCES) -AUTOMAKE_OPTIONS = 1.5 gnits -# no-dependencies +find_OPT_SOURCES = fileblocks.c memcmp.c memset.c mktime.c stpcpy.c strdup.c \ +strftime.c strspn.c strstr.c strtol.c alloca.c -noinst_LIBRARIES = libfind.a +DIST_OTHER = fnmatch.h getopt.h modechange.h modetype.h pathmax.h \ +regex.h wait.h -if CROSS_COMPILING -# The regexprops program needs to be a native executable, so we -# can't build it with a cross-compiler. -else -noinst_PROGRAMS = regexprops -regexprops_SOURCES = regexprops.c regextype.c -endif - -libfind_a_SOURCES = gnulib-version.c findutils-version.c -EXTRA_DIST = modetype.h wait.h extendbuf.h savedirinfo.h buildcmd.h \ - gnulib-version.h gnulib-version.c findutils-version.h -BUILT_SOURCES = gnulib-version.c -SUFFIXES = -MOSTLYCLEANFILES = -CLEANFILES = -DISTCLEANFILES = -MAINTAINERCLEANFILES = - - -INCLUDES = -I../gnulib/lib -I$(top_srcdir)/gnulib/lib -LDADD = ../gnulib/lib/libgnulib.a @LIBINTL@ - -libfind_a_SOURCES += modetype.h nextelem.h printquoted.h listfile.h \ - regextype.h dircallback.h -libfind_a_SOURCES += listfile.c nextelem.c extendbuf.c buildcmd.c savedirinfo.c \ - forcefindlib.c qmark.c printquoted.c regextype.c dircallback.c - -EXTRA_DIST += waitpid.c forcefindlib.c - - -libfind_a_LIBADD = @FINDLIBOBJS@ -libfind_a_DEPENDENCIES = @FINDLIBOBJS@ -# libfind_la_LIBADD = @LTFINDLIBOBJS@ - -# libfind_a_OBJECTS += @FINDLIBOBJS@ -# libfind_la_OBJECTS += @LTFINDLIBOBJS@ +CONFIG_HEADER = ../config.h +INCLUDES = -I.. -I$(srcdir) +fnmatch.o: fnmatch.h +getopt1.o: getopt.h +listfile.o xgetcwd.o: pathmax.h +modechange.o: modechange.h +regex.o: regex.h diff --git a/lib/Makefile.in b/lib/Makefile.in new file mode 100644 index 00000000..534b294a --- /dev/null +++ b/lib/Makefile.in @@ -0,0 +1,136 @@ +# Makefile.in generated automatically by automake from Makefile.am. +# Copyright (C) 1994 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., 675 Mass Ave, Cambridge, MA 02139, USA. + +SHELL = /bin/sh + +srcdir = @srcdir@ +top_srcdir = @top_srcdir@ +VPATH = @srcdir@ +prefix = @prefix@ +exec_prefix = @exec_prefix@ + +bindir = $(exec_prefix)/bin +sbindir = $(exec_prefix)/sbin +libexecdir = $(exec_prefix)/libexec +datadir = $(prefix)/share +sysconfdir = $(prefix)/etc +sharedstatedir = $(prefix)/com +localstatedir = $(prefix)/var +libdir = $(exec_prefix)/lib +infodir = $(prefix)/info +mandir = $(prefix)/man +includedir = $(prefix)/include +oldincludedir = /usr/include + +INSTALL = @INSTALL@ +INSTALL_PROGRAM = @INSTALL_PROGRAM@ +INSTALL_DATA = @INSTALL_DATA@ +transform = @program_transform_name@ + +ALL = ${PROGRAMS} ${LIBPROGRAMS} ${SCRIPTS} ${LIBSCRIPTS} ${LIBFILES} +CC = @CC@ +LEX = @LEX@ +YACC = @YACC@ +ANSI2KNR = ./ansi2knr + +DEFS = @DEFS@ +CPPFLAGS = @CPPFLAGS@ +CFLAGS = @CFLAGS@ +LDFLAGS = @LDFLAGS@ +LIBS = @LIBS@ + +find_OBJECTS = regex.o dirname.o error.o filemode.o getopt.o getopt1.o idcache.o listfile.o modechange.o nextelem.o savedir.o xmalloc.o getline.o xstrdup.o xgetcwd.o fnmatch.o +LIBFILES = libfind.a +AR = ar +RANLIB = @RANLIB@ +ALLOCA = @ALLOCA@ +LIBOBJS = @LIBOBJS@ + +SOURCES = ${find_SOURCES} +DIST_CONF = Makefile.am Makefile.in +DIST_FILES = $(DIST_CONF) $(SOURCES) $(TEXINFOS) $(INFOS) $(MANS) $(DIST_OTHER) + +LIBRARIES = find +find_SOURCES = regex.c dirname.c error.c filemode.c getopt.c getopt1.c \ +idcache.c listfile.c modechange.c nextelem.c savedir.c xmalloc.c getline.c \ +xstrdup.c xgetcwd.c fnmatch.c $(find_OPT_SOURCES) + +find_OPT_SOURCES = fileblocks.c memcmp.c memset.c mktime.c stpcpy.c strdup.c \ +strftime.c strspn.c strstr.c strtol.c alloca.c + +DIST_OTHER = fnmatch.h getopt.h modechange.h modetype.h pathmax.h \ +regex.h wait.h + +CONFIG_HEADER = ../config.h +INCLUDES = -I.. -I$(srcdir) + +all:: ${ALL} + +.c.o: + $(CC) -c $(DEFS) $(INCLUDES) $(CPPFLAGS) $(CFLAGS) $< + +$(find_OBJECTS): ../config.h +install:: install-libraries + +install-libraries: $(LIBFILES) + +uninstall:: uninstall-libraries + +uninstall-libraries: + +libfind.a: $(find_OBJECTS) @LIBOBJS@ @ALLOCA@ + rm -f libfind.a + $(AR) cru libfind.a $(find_OBJECTS) @LIBOBJS@ @ALLOCA@ + $(RANLIB) libfind.a + +mostlyclean: + rm -f *.o core + +clean: mostlyclean + rm -f $(PROGRAMS) $(LIBPROGRAMS) $(LIBFILES) $(TEXFILES) $(CLEANFILES) + +distclean: clean + rm -f Makefile *.tab.c $(DISTCLEANFILES) + rm -f config.cache config.log config.status ${CONFIG_HEADER} stamp-h + +realclean: distclean + rm -f TAGS $(INFOS) + +dist: $(DIST_FILES) $(DIST_DIRS) + -mkdir ../`cat ../distname`/$(subdir) + @for file in $(DIST_FILES); do \ + echo linking $$file; \ + ln $(srcdir)/$$file ../`cat ../distname`/$(subdir)/$$file || \ + { echo copying $$file instead; cp -p $(srcdir)/$$file ../`cat ../distname`/$(subdir)/$$file;}; \ + done + +check dvi info install uninstall:: + +tags:: TAGS + +TAGS:: + cd $(srcdir); etags $(SOURCES) + +# Tell versions [3.59,3.63) of GNU make to not export all variables. +# Otherwise a system limit (for SysV at least) may be exceeded. +.NOEXPORT: + +fnmatch.o: fnmatch.h +getopt1.o: getopt.h +listfile.o xgetcwd.o: pathmax.h +modechange.o: modechange.h +regex.o: regex.h diff --git a/lib/alloca.c b/lib/alloca.c new file mode 100644 index 00000000..7020f32c --- /dev/null +++ b/lib/alloca.c @@ -0,0 +1,492 @@ +/* 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. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#ifdef emacs +#include "blockinput.h" +#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 +#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 + +#if __STDC__ +typedef void *pointer; +#else +typedef char *pointer; +#endif + +#define NULL 0 + +/* Different portions of Emacs need to call different versions of + malloc. The Emacs executable needs alloca to call xmalloc, because + ordinary malloc isn't protected from input signals. On the other + hand, the utilities in lib-src need alloca to call malloc; some of + them are very simple, and don't have an xmalloc routine. + + Non-Emacs programs expect this to call use xmalloc. + + Callers below should use malloc. */ + +#ifndef emacs +#define malloc xmalloc +#endif +extern pointer malloc (); + +/* 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 () +{ + 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. */ + +pointer +alloca (size) + unsigned 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 ((pointer) 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. */ + + { + register pointer new = malloc (sizeof (header) + size); + /* Address of header. */ + + ((header *) new)->h.next = last_alloca_header; + ((header *) new)->h.deep = depth; + + last_alloca_header = (header *) new; + + /* User storage begins just after header. */ + + return (pointer) ((char *) new + sizeof (header)); + } +} + +#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/lib/dirname.c b/lib/dirname.c new file mode 100644 index 00000000..15d25967 --- /dev/null +++ b/lib/dirname.c @@ -0,0 +1,70 @@ +/* dirname.c -- return all but the last element in a path + Copyright (C) 1990 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#ifdef STDC_HEADERS +#include <stdlib.h> +#else +char *malloc (); +#endif +#if defined(STDC_HEADERS) || defined(HAVE_STRING_H) +#include <string.h> +#else +#include <strings.h> +#ifndef strrchr +#define strrchr rindex +#endif +#endif + +/* Return the leading directories part of PATH, + allocated with malloc. If out of memory, return 0. + Assumes that trailing slashes have already been + removed. */ + +char * +dirname (path) + char *path; +{ + char *newpath; + char *slash; + int length; /* Length of result, not including NUL. */ + + slash = strrchr (path, '/'); + if (slash == 0) + { + /* File is in the current directory. */ + path = "."; + length = 1; + } + else + { + /* Remove any trailing slashes from the result. */ + while (slash > path && *slash == '/') + --slash; + + length = slash - path + 1; + } + newpath = (char *) malloc (length + 1); + if (newpath == 0) + return 0; + strncpy (newpath, path, length); + newpath[length] = 0; + return newpath; +} diff --git a/lib/error.c b/lib/error.c new file mode 100644 index 00000000..19c2ba88 --- /dev/null +++ b/lib/error.c @@ -0,0 +1,119 @@ +/* error.c -- error handler for noninteractive utilities + Copyright (C) 1990, 91, 92, 93, 94 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +/* Written by David MacKenzie <djm@gnu.ai.mit.edu>. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#include <stdio.h> + +#if HAVE_VPRINTF || HAVE_DOPRNT +# if __STDC__ +# include <stdarg.h> +# define VA_START(args, lastarg) va_start(args, lastarg) +# else +# include <varargs.h> +# define VA_START(args, lastarg) va_start(args) +# endif +#else +# define va_alist a1, a2, a3, a4, a5, a6, a7, a8 +# define va_dcl char *a1, *a2, *a3, *a4, *a5, *a6, *a7, *a8; +#endif + +#if STDC_HEADERS +# include <stdlib.h> +# include <string.h> +#else +void exit (); +#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) () = NULL; + +/* The calling program should define program_name and set it to the + name of the executing program. */ +extern char *program_name; + +#if HAVE_STRERROR +char *strerror (); +#else +static char * +private_strerror (errnum) + int errnum; +{ + extern char *sys_errlist[]; + extern int sys_nerr; + + if (errnum > 0 && errnum <= sys_nerr) + return sys_errlist[errnum]; + return "Unknown system error"; +} +#define strerror private_strerror +#endif + +/* 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. */ +/* VARARGS */ + +void +#if defined(VA_START) && __STDC__ +error (int status, int errnum, const char *message, ...) +#else +error (status, errnum, message, va_alist) + int status; + int errnum; + char *message; + va_dcl +#endif +{ +#ifdef VA_START + va_list args; +#endif + + if (error_print_progname) + (*error_print_progname) (); + else + { + fflush (stdout); + fprintf (stderr, "%s: ", program_name); + } + +#ifdef VA_START + VA_START (args, message); +# if HAVE_VPRINTF + vfprintf (stderr, message, args); +# else + _doprnt (message, args, stderr); +# endif + va_end (args); +#else + fprintf (stderr, message, a1, a2, a3, a4, a5, a6, a7, a8); +#endif + + if (errnum) + fprintf (stderr, ": %s", strerror (errnum)); + putc ('\n', stderr); + fflush (stderr); + if (status) + exit (status); +} diff --git a/lib/fileblocks.c b/lib/fileblocks.c new file mode 100644 index 00000000..83ac04fd --- /dev/null +++ b/lib/fileblocks.c @@ -0,0 +1,66 @@ +/* Convert file size to number of blocks on System V-like machines. + Copyright (C) 1990 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +/* Written by Brian L. Matthews, blm@6sceng.UUCP. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#if !defined (HAVE_ST_BLOCKS) && !defined(_POSIX_SOURCE) +#include <sys/types.h> +#include <sys/param.h> + +#ifndef NINDIR +/* Some SysV's, like Irix, seem to lack these. Hope they're correct. */ +/* Size of a indirect block, in bytes. */ +#ifndef BSIZE +#define BSIZE 1024 +#endif + +/* Number of inode pointers per indirect block. */ +#define NINDIR (BSIZE/sizeof(daddr_t)) +#endif /* !NINDIR */ + +/* Number of direct block addresses in an inode. */ +#define NDIR 10 + +/* Return the number of 512-byte blocks in a file of SIZE bytes. */ + +long +st_blocks (size) + long size; +{ + long datablks = (size + 512 - 1) / 512; + long indrblks = 0; + + if (datablks > NDIR) + { + indrblks = (datablks - NDIR - 1) / NINDIR + 1; + + if (datablks > NDIR + NINDIR) + { + indrblks += (datablks - NDIR - NINDIR - 1) / (NINDIR * NINDIR) + 1; + + if (datablks > NDIR + NINDIR + NINDIR * NINDIR) + indrblks++; + } + } + + return datablks + indrblks; +} +#endif diff --git a/lib/filemode.c b/lib/filemode.c new file mode 100644 index 00000000..c86ee2ff --- /dev/null +++ b/lib/filemode.c @@ -0,0 +1,237 @@ +/* filemode.c -- make a string describing file modes + Copyright (C) 1985, 1990, 1993 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#include <sys/types.h> +#include <sys/stat.h> + +#ifndef S_IREAD +#define S_IREAD S_IRUSR +#define S_IWRITE S_IWUSR +#define S_IEXEC S_IXUSR +#endif + +#ifdef STAT_MACROS_BROKEN +#undef S_ISBLK +#undef S_ISCHR +#undef S_ISDIR +#undef S_ISFIFO +#undef S_ISLNK +#undef S_ISMPB +#undef S_ISMPC +#undef S_ISNWK +#undef S_ISREG +#undef S_ISSOCK +#endif /* STAT_MACROS_BROKEN. */ + +#if !defined(S_ISBLK) && defined(S_IFBLK) +#define S_ISBLK(m) (((m) & S_IFMT) == S_IFBLK) +#endif +#if !defined(S_ISCHR) && defined(S_IFCHR) +#define S_ISCHR(m) (((m) & S_IFMT) == S_IFCHR) +#endif +#if !defined(S_ISDIR) && defined(S_IFDIR) +#define S_ISDIR(m) (((m) & S_IFMT) == S_IFDIR) +#endif +#if !defined(S_ISREG) && defined(S_IFREG) +#define S_ISREG(m) (((m) & S_IFMT) == S_IFREG) +#endif +#if !defined(S_ISFIFO) && defined(S_IFIFO) +#define S_ISFIFO(m) (((m) & S_IFMT) == S_IFIFO) +#endif +#if !defined(S_ISLNK) && defined(S_IFLNK) +#define S_ISLNK(m) (((m) & S_IFMT) == S_IFLNK) +#endif +#if !defined(S_ISSOCK) && defined(S_IFSOCK) +#define S_ISSOCK(m) (((m) & S_IFMT) == S_IFSOCK) +#endif +#if !defined(S_ISMPB) && defined(S_IFMPB) /* V7 */ +#define S_ISMPB(m) (((m) & S_IFMT) == S_IFMPB) +#define S_ISMPC(m) (((m) & S_IFMT) == S_IFMPC) +#endif +#if !defined(S_ISNWK) && defined(S_IFNWK) /* HP/UX */ +#define S_ISNWK(m) (((m) & S_IFMT) == S_IFNWK) +#endif + +void mode_string (); +static char ftypelet (); +static void rwx (); +static void setst (); + +/* filemodestring - fill in string STR with an ls-style ASCII + representation of the st_mode field of file stats block STATP. + 10 characters are stored in STR; no terminating null is added. + The characters stored in STR are: + + 0 File type. 'd' for directory, 'c' for character + special, 'b' for block special, 'm' for multiplex, + 'l' for symbolic link, 's' for socket, 'p' for fifo, + '-' for regular, '?' for any other file type + + 1 'r' if the owner may read, '-' otherwise. + + 2 'w' if the owner may write, '-' otherwise. + + 3 'x' if the owner may execute, 's' if the file is + set-user-id, '-' otherwise. + 'S' if the file is set-user-id, but the execute + bit isn't set. + + 4 'r' if group members may read, '-' otherwise. + + 5 'w' if group members may write, '-' otherwise. + + 6 'x' if group members may execute, 's' if the file is + set-group-id, '-' otherwise. + 'S' if it is set-group-id but not executable. + + 7 'r' if any user may read, '-' otherwise. + + 8 'w' if any user may write, '-' otherwise. + + 9 'x' if any user may execute, 't' if the file is "sticky" + (will be retained in swap space after execution), '-' + otherwise. + 'T' if the file is sticky but not executable. */ + +void +filemodestring (statp, str) + struct stat *statp; + char *str; +{ + mode_string (statp->st_mode, str); +} + +/* Like filemodestring, but only the relevant part of the `struct stat' + is given as an argument. */ + +void +mode_string (mode, str) + unsigned short mode; + char *str; +{ + str[0] = ftypelet ((long) mode); + rwx ((mode & 0700) << 0, &str[1]); + rwx ((mode & 0070) << 3, &str[4]); + rwx ((mode & 0007) << 6, &str[7]); + setst (mode, str); +} + +/* Return a character indicating the type of file described by + file mode BITS: + 'd' for directories + 'b' for block special files + 'c' for character special files + 'm' for multiplexor files + 'l' for symbolic links + 's' for sockets + 'p' for fifos + '-' for regular files + '?' for any other file type. */ + +static char +ftypelet (bits) + long bits; +{ +#ifdef S_ISBLK + if (S_ISBLK (bits)) + return 'b'; +#endif + if (S_ISCHR (bits)) + return 'c'; + if (S_ISDIR (bits)) + return 'd'; + if (S_ISREG (bits)) + return '-'; +#ifdef S_ISFIFO + if (S_ISFIFO (bits)) + return 'p'; +#endif +#ifdef S_ISLNK + if (S_ISLNK (bits)) + return 'l'; +#endif +#ifdef S_ISSOCK + if (S_ISSOCK (bits)) + return 's'; +#endif +#ifdef S_ISMPC + if (S_ISMPC (bits)) + return 'm'; +#endif +#ifdef S_ISNWK + if (S_ISNWK (bits)) + return 'n'; +#endif + return '?'; +} + +/* Look at read, write, and execute bits in BITS and set + flags in CHARS accordingly. */ + +static void +rwx (bits, chars) + unsigned short bits; + char *chars; +{ + chars[0] = (bits & S_IREAD) ? 'r' : '-'; + chars[1] = (bits & S_IWRITE) ? 'w' : '-'; + chars[2] = (bits & S_IEXEC) ? 'x' : '-'; +} + +/* Set the 's' and 't' flags in file attributes string CHARS, + according to the file mode BITS. */ + +static void +setst (bits, chars) + unsigned short bits; + char *chars; +{ +#ifdef S_ISUID + if (bits & S_ISUID) + { + if (chars[3] != 'x') + /* Set-uid, but not executable by owner. */ + chars[3] = 'S'; + else + chars[3] = 's'; + } +#endif +#ifdef S_ISGID + if (bits & S_ISGID) + { + if (chars[6] != 'x') + /* Set-gid, but not executable by group. */ + chars[6] = 'S'; + else + chars[6] = 's'; + } +#endif +#ifdef S_ISVTX + if (bits & S_ISVTX) + { + if (chars[9] != 'x') + /* Sticky, but not executable by others. */ + chars[9] = 'T'; + else + chars[9] = 't'; + } +#endif +} diff --git a/lib/fnmatch.c b/lib/fnmatch.c new file mode 100644 index 00000000..2fb65b52 --- /dev/null +++ b/lib/fnmatch.c @@ -0,0 +1,200 @@ +/* Copyright (C) 1991, 1992, 1993 Free Software Foundation, Inc. + +This library is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public License as +published by the Free Software Foundation; either version 2 of the +License, or (at your option) any later version. + +This library 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 +Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with this library; see the file COPYING.LIB. If +not, write to the Free Software Foundation, Inc., 675 Mass Ave, +Cambridge, MA 02139, USA. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#include <errno.h> +#include <fnmatch.h> +#include <ctype.h> + + +/* Comment out all this code if we are using the GNU C Library, and are not + actually compiling the library itself. This code is part of the GNU C + Library, but also included in many other GNU distributions. Compiling + and linking in this code is a waste when using the GNU C library + (especially if it is a shared library). Rather than having every GNU + program understand `configure --with-gnu-libc' and omit the object files, + it is simpler to just do this in the source for each such file. */ + +#if defined (_LIBC) || !defined (__GNU_LIBRARY__) + + +#if !defined(__GNU_LIBRARY__) && !defined(STDC_HEADERS) +extern int errno; +#endif + +/* Match STRING against the filename pattern PATTERN, returning zero if + it matches, nonzero if not. */ +int +fnmatch (pattern, string, flags) + const char *pattern; + const char *string; + int flags; +{ + register const char *p = pattern, *n = string; + register char c; + +/* Note that this evalutes C many times. */ +#define FOLD(c) ((flags & FNM_CASEFOLD) && isupper (c) ? tolower (c) : (c)) + + while ((c = *p++) != '\0') + { + c = FOLD (c); + + switch (c) + { + case '?': + if (*n == '\0') + return FNM_NOMATCH; + else if ((flags & FNM_FILE_NAME) && *n == '/') + return FNM_NOMATCH; + else if ((flags & FNM_PERIOD) && *n == '.' && + (n == string || ((flags & FNM_FILE_NAME) && n[-1] == '/'))) + return FNM_NOMATCH; + break; + + case '\\': + if (!(flags & FNM_NOESCAPE)) + { + c = *p++; + c = FOLD (c); + } + if (FOLD (*n) != c) + return FNM_NOMATCH; + break; + + case '*': + if ((flags & FNM_PERIOD) && *n == '.' && + (n == string || ((flags & FNM_FILE_NAME) && n[-1] == '/'))) + return FNM_NOMATCH; + + for (c = *p++; c == '?' || c == '*'; c = *p++, ++n) + if (((flags & FNM_FILE_NAME) && *n == '/') || + (c == '?' && *n == '\0')) + return FNM_NOMATCH; + + if (c == '\0') + return 0; + + { + char c1 = (!(flags & FNM_NOESCAPE) && c == '\\') ? *p : c; + c1 = FOLD (c1); + for (--p; *n != '\0'; ++n) + if ((c == '[' || FOLD (*n) == c1) && + fnmatch (p, n, flags & ~FNM_PERIOD) == 0) + return 0; + return FNM_NOMATCH; + } + + case '[': + { + /* Nonzero if the sense of the character class is inverted. */ + register int not; + + if (*n == '\0') + return FNM_NOMATCH; + + if ((flags & FNM_PERIOD) && *n == '.' && + (n == string || ((flags & FNM_FILE_NAME) && n[-1] == '/'))) + return FNM_NOMATCH; + + not = (*p == '!' || *p == '^'); + if (not) + ++p; + + c = *p++; + for (;;) + { + register char cstart = c, cend = c; + + if (!(flags & FNM_NOESCAPE) && c == '\\') + cstart = cend = *p++; + + cstart = cend = FOLD (cstart); + + if (c == '\0') + /* [ (unterminated) loses. */ + return FNM_NOMATCH; + + c = *p++; + c = FOLD (c); + + if ((flags & FNM_FILE_NAME) && c == '/') + /* [/] can never match. */ + return FNM_NOMATCH; + + if (c == '-' && *p != ']') + { + cend = *p++; + if (!(flags & FNM_NOESCAPE) && cend == '\\') + cend = *p++; + if (cend == '\0') + return FNM_NOMATCH; + cend = FOLD (cend); + + c = *p++; + } + + if (FOLD (*n) >= cstart && FOLD (*n) <= cend) + goto matched; + + if (c == ']') + break; + } + if (!not) + return FNM_NOMATCH; + break; + + matched:; + /* Skip the rest of the [...] that already matched. */ + while (c != ']') + { + if (c == '\0') + /* [... (unterminated) loses. */ + return FNM_NOMATCH; + + c = *p++; + if (!(flags & FNM_NOESCAPE) && c == '\\') + /* XXX 1003.2d11 is unclear if this is right. */ + ++p; + } + if (not) + return FNM_NOMATCH; + } + break; + + default: + if (c != FOLD (*n)) + return FNM_NOMATCH; + } + + ++n; + } + + if (*n == '\0') + return 0; + + if ((flags & FNM_LEADING_DIR) && *n == '/') + /* The FNM_LEADING_DIR flag says that "foo*" matches "foobar/frobozz". */ + return 0; + + return FNM_NOMATCH; +} + +#endif /* _LIBC or not __GNU_LIBRARY__. */ diff --git a/lib/fnmatch.h b/lib/fnmatch.h new file mode 100644 index 00000000..69eab0c0 --- /dev/null +++ b/lib/fnmatch.h @@ -0,0 +1,67 @@ +/* Copyright (C) 1991, 1992, 1993 Free Software Foundation, Inc. + +This library is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public License as +published by the Free Software Foundation; either version 2 of the +License, or (at your option) any later version. + +This library 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 +Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with this library; see the file COPYING.LIB. If +not, write to the Free Software Foundation, Inc., 675 Mass Ave, +Cambridge, MA 02139, USA. */ + +#ifndef _FNMATCH_H + +#define _FNMATCH_H 1 + +#ifdef __cplusplus +extern "C" { +#endif + +#if defined (__cplusplus) || (defined (__STDC__) && __STDC__) +#undef __P +#define __P(args) args +#else /* Not C++ or ANSI C. */ +#undef __P +#define __P(args) () +/* We can get away without defining `const' here only because in this file + it is used only inside the prototype for `fnmatch', which is elided in + non-ANSI C where `const' is problematical. */ +#endif /* C++ or ANSI C. */ + + +/* We #undef these before defining them because some losing systems + (HP-UX A.08.07 for example) define these in <unistd.h>. */ +#undef FNM_PATHNAME +#undef FNM_NOESCAPE +#undef FNM_PERIOD + +/* Bits set in the FLAGS argument to `fnmatch'. */ +#define FNM_PATHNAME (1 << 0) /* No wildcard can ever match `/'. */ +#define FNM_NOESCAPE (1 << 1) /* Backslashes don't quote special chars. */ +#define FNM_PERIOD (1 << 2) /* Leading `.' is matched only explicitly. */ + +#if !defined (_POSIX_C_SOURCE) || _POSIX_C_SOURCE < 2 || defined (_GNU_SOURCE) +#define FNM_FILE_NAME FNM_PATHNAME /* Preferred GNU name. */ +#define FNM_LEADING_DIR (1 << 3) /* Ignore `/...' after a match. */ +#define FNM_CASEFOLD (1 << 4) /* Compare without regard to case. */ +#endif + +/* Value returned by `fnmatch' if STRING does not match PATTERN. */ +#define FNM_NOMATCH 1 + +/* Match STRING against the filename pattern PATTERN, + returning zero if it matches, FNM_NOMATCH if not. */ +extern int fnmatch __P ((const char *__pattern, const char *__string, + int __flags)); + +#ifdef __cplusplus +} +#endif + +#endif /* fnmatch.h */ diff --git a/lib/getline.c b/lib/getline.c new file mode 100644 index 00000000..519ff932 --- /dev/null +++ b/lib/getline.c @@ -0,0 +1,126 @@ +/* getline.c -- Replacement for GNU C library function getline + +Copyright (C) 1993 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 of the +License, 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +/* Written by Jan Brittenson, bson@gnu.ai.mit.edu. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#include <sys/types.h> +#include <stdio.h> +#define NDEBUG +#include <assert.h> + +#if STDC_HEADERS +#include <stdlib.h> +#else +char *malloc (), *realloc (); +#endif + +/* Always add at least this many bytes when extending the buffer. */ +#define MIN_CHUNK 64 + +/* Read up to (and including) a TERMINATOR from STREAM into *LINEPTR + + OFFSET (and null-terminate it). *LINEPTR is a pointer returned from + malloc (or NULL), pointing to *N characters of space. It is realloc'd + as necessary. Return the number of characters read (not including the + null terminator), or -1 on error or EOF. */ + +int +getstr (lineptr, n, stream, terminator, offset) + char **lineptr; + size_t *n; + FILE *stream; + char terminator; + int offset; +{ + int nchars_avail; /* Allocated but unused chars in *LINEPTR. */ + char *read_pos; /* Where we're reading into *LINEPTR. */ + int ret; + + if (!lineptr || !n || !stream) + return -1; + + if (!*lineptr) + { + *n = MIN_CHUNK; + *lineptr = malloc (*n); + if (!*lineptr) + return -1; + } + + nchars_avail = *n - offset; + read_pos = *lineptr + offset; + + for (;;) + { + register int c = getc (stream); + + /* We always want at least one char left in the buffer, since we + always (unless we get an error while reading the first char) + NUL-terminate the line buffer. */ + + assert(*n - nchars_avail == read_pos - *lineptr); + if (nchars_avail < 1) + { + if (*n > MIN_CHUNK) + *n *= 2; + else + *n += MIN_CHUNK; + + nchars_avail = *n + *lineptr - read_pos; + *lineptr = realloc (*lineptr, *n); + if (!*lineptr) + return -1; + read_pos = *n - nchars_avail + *lineptr; + assert(*n - nchars_avail == read_pos - *lineptr); + } + + if (c == EOF || ferror (stream)) + { + /* Return partial line, if any. */ + if (read_pos == *lineptr) + return -1; + else + break; + } + + *read_pos++ = c; + nchars_avail--; + + if (c == terminator) + /* Return the line. */ + break; + } + + /* Done - NUL terminate and return the number of chars read. */ + *read_pos = '\0'; + + ret = read_pos - (*lineptr + offset); + return ret; +} + +int +getline (lineptr, n, stream) + char **lineptr; + size_t *n; + FILE *stream; +{ + return getstr (lineptr, n, stream, '\n', 0); +} diff --git a/lib/getopt.c b/lib/getopt.c new file mode 100644 index 00000000..43c0a6a9 --- /dev/null +++ b/lib/getopt.c @@ -0,0 +1,748 @@ +/* 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 roland@gnu.ai.mit.edu + before changing it! + + Copyright (C) 1987, 88, 89, 90, 91, 92, 93, 94 + 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, 675 Mass Ave, Cambridge, MA 02139, USA. */ + +/* This tells Alpha OSF/1 not to define a getopt prototype in <stdio.h>. + Ditto for AIX 3.2 and <stdlib.h>. */ +#ifndef _NO_PROTO +#define _NO_PROTO +#endif + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#if !defined (__STDC__) || !__STDC__ +/* This is a separate conditional since some stdc systems + reject `defined (const)'. */ +#ifndef const +#define const +#endif +#endif + +#include <stdio.h> + +/* Comment out all this code if we are using the GNU C Library, and are not + actually compiling the library itself. This code is part of the GNU C + Library, but also included in many other GNU distributions. Compiling + and linking in this code is a waste when using the GNU C library + (especially if it is a shared library). Rather than having every GNU + program understand `configure --with-gnu-libc' and omit the object files, + it is simpler to just do this in the source for each such file. */ + +#if defined (_LIBC) || !defined (__GNU_LIBRARY__) + + +/* This needs to come after some library #include + to get __GNU_LIBRARY__ defined. */ +#ifdef __GNU_LIBRARY__ +/* Don't include stdlib.h for non-GNU C libraries because some of them + contain conflicting prototypes for getopt. */ +#include <stdlib.h> +#endif /* GNU C library. */ + +/* This version of `getopt' appears to the caller like standard Unix `getopt' + but it behaves differently for the user, since it allows the user + to intersperse the options with the other arguments. + + As `getopt' 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. + + Setting the environment variable POSIXLY_CORRECT disables permutation. + Then the 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.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 = NULL; + +/* 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 EOF, 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. */ + +/* XXX 1003.2 says this must be 1 before any call. */ +int optind = 0; + +/* 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. */ + +static char *nextchar; + +/* 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 = '?'; + +/* 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. + + 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 EOF with `optind' != ARGC. */ + +static enum +{ + REQUIRE_ORDER, PERMUTE, RETURN_IN_ORDER +} ordering; + +/* Value of POSIXLY_CORRECT environment variable. */ +static char *posixly_correct; + +#ifdef __GNU_LIBRARY__ +/* We want to avoid inclusion of string.h with non-GNU libraries + because there are many ways it can cause trouble. + On some systems, it contains special magic macros that don't work + in GCC. */ +#include <string.h> +#define my_index strchr +#else + +/* Avoid depending on library functions or files + whose names are inconsistent. */ + +char *getenv (); + +static char * +my_index (str, chr) + const char *str; + int chr; +{ + while (*str) + { + if (*str == chr) + return (char *) str; + str++; + } + return 0; +} + +/* If using GCC, we can safely declare strlen this way. + If not using GCC, it is ok not to declare it. */ +#ifdef __GNUC__ +/* Note that Motorola Delta 68k R3V7 comes with GCC but not stddef.h. + That was relevant to code that was here before. */ +#if !defined (__STDC__) || !__STDC__ +/* gcc with -traditional declares the built-in strlen to return int, + and has done so at least since version 2.4.5. -- rms. */ +extern int strlen (const char *); +#endif /* not __STDC__ */ +#endif /* __GNUC__ */ + +#endif /* not __GNU_LIBRARY__ */ + +/* 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. */ + +static int first_nonopt; +static int last_nonopt; + +/* 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 (argv) + char **argv; +{ + int bottom = first_nonopt; + int middle = last_nonopt; + int top = 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. */ + + 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; + } + /* 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; + } + /* Exclude the moved top segment from further swapping. */ + bottom += len; + } + } + + /* Update records for the slots the non-options now occupy. */ + + first_nonopt += (optind - last_nonopt); + last_nonopt = optind; +} + +/* Initialize the internal data when the first call is made. */ + +static const char * +_getopt_initialize (optstring) + const char *optstring; +{ + /* 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. */ + + first_nonopt = last_nonopt = optind = 1; + + nextchar = NULL; + + posixly_correct = getenv ("POSIXLY_CORRECT"); + + /* Determine how to handle the ordering of options and nonoptions. */ + + if (optstring[0] == '-') + { + ordering = RETURN_IN_ORDER; + ++optstring; + } + else if (optstring[0] == '+') + { + ordering = REQUIRE_ORDER; + ++optstring; + } + else if (posixly_correct != NULL) + ordering = REQUIRE_ORDER; + else + ordering = PERMUTE; + + 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 `EOF'. + 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. + + The elements of ARGV aren't really const, because we permute them. + But we pretend they're const in the prototype to be compatible + with other systems. + + 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. */ + +int +_getopt_internal (argc, argv, optstring, longopts, longind, long_only) + int argc; + char *const *argv; + const char *optstring; + const struct option *longopts; + int *longind; + int long_only; +{ + optarg = NULL; + + if (optind == 0) + optstring = _getopt_initialize (optstring); + + if (nextchar == NULL || *nextchar == '\0') + { + /* Advance to the next ARGV-element. */ + + if (ordering == PERMUTE) + { + /* If we have just processed some options following some non-options, + exchange them so that the options come first. */ + + if (first_nonopt != last_nonopt && last_nonopt != optind) + exchange ((char **) argv); + else if (last_nonopt != optind) + first_nonopt = optind; + + /* Skip any additional non-options + and extend the range of non-options previously skipped. */ + + while (optind < argc + && (argv[optind][0] != '-' || argv[optind][1] == '\0')) + optind++; + last_nonopt = 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 (optind != argc && !strcmp (argv[optind], "--")) + { + optind++; + + if (first_nonopt != last_nonopt && last_nonopt != optind) + exchange ((char **) argv); + else if (first_nonopt == last_nonopt) + first_nonopt = optind; + last_nonopt = argc; + + 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 (optind == argc) + { + /* Set the next-arg-index to point at the non-options + that we previously skipped, so the caller will digest them. */ + if (first_nonopt != last_nonopt) + optind = first_nonopt; + return EOF; + } + + /* 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 ((argv[optind][0] != '-' || argv[optind][1] == '\0')) + { + if (ordering == REQUIRE_ORDER) + return EOF; + optarg = argv[optind++]; + return 1; + } + + /* We have found another option-ARGV-element. + Skip the initial punctuation. */ + + nextchar = (argv[optind] + 1 + + (longopts != NULL && argv[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[optind][1] == '-' + || (long_only && (argv[optind][2] || !my_index (optstring, argv[optind][1]))))) + { + char *nameend; + const struct option *p; + const struct option *pfound = NULL; + int exact = 0; + int ambig = 0; + int indfound; + int option_index; + + for (nameend = 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, nextchar, nameend - nextchar)) + { + if (nameend - 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 (opterr) + fprintf (stderr, "%s: option `%s' is ambiguous\n", + argv[0], argv[optind]); + nextchar += strlen (nextchar); + optind++; + return '?'; + } + + if (pfound != NULL) + { + option_index = indfound; + 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) + optarg = nameend + 1; + else + { + if (opterr) + { + if (argv[optind - 1][1] == '-') + /* --option */ + fprintf (stderr, + "%s: option `--%s' doesn't allow an argument\n", + argv[0], pfound->name); + else + /* +option or -option */ + fprintf (stderr, + "%s: option `%c%s' doesn't allow an argument\n", + argv[0], argv[optind - 1][0], pfound->name); + } + nextchar += strlen (nextchar); + return '?'; + } + } + else if (pfound->has_arg == 1) + { + if (optind < argc) + optarg = argv[optind++]; + else + { + if (opterr) + fprintf (stderr, "%s: option `%s' requires an argument\n", + argv[0], argv[optind - 1]); + nextchar += strlen (nextchar); + return optstring[0] == ':' ? ':' : '?'; + } + } + nextchar += strlen (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[optind][1] == '-' + || my_index (optstring, *nextchar) == NULL) + { + if (opterr) + { + if (argv[optind][1] == '-') + /* --option */ + fprintf (stderr, "%s: unrecognized option `--%s'\n", + argv[0], nextchar); + else + /* +option or -option */ + fprintf (stderr, "%s: unrecognized option `%c%s'\n", + argv[0], argv[optind][0], nextchar); + } + nextchar = (char *) ""; + optind++; + return '?'; + } + } + + /* Look at and handle the next short option-character. */ + + { + char c = *nextchar++; + char *temp = my_index (optstring, c); + + /* Increment `optind' when we start to process its last character. */ + if (*nextchar == '\0') + ++optind; + + if (temp == NULL || c == ':') + { + if (opterr) + { + if (posixly_correct) + /* 1003.2 specifies the format of this message. */ + fprintf (stderr, "%s: illegal option -- %c\n", argv[0], c); + else + fprintf (stderr, "%s: invalid option -- %c\n", argv[0], c); + } + optopt = c; + return '?'; + } + if (temp[1] == ':') + { + if (temp[2] == ':') + { + /* This is an option that accepts an argument optionally. */ + if (*nextchar != '\0') + { + optarg = nextchar; + optind++; + } + else + optarg = NULL; + nextchar = NULL; + } + else + { + /* This is an option that requires an argument. */ + if (*nextchar != '\0') + { + optarg = nextchar; + /* If we end this ARGV-element by taking the rest as an arg, + we must advance to the next element now. */ + optind++; + } + else if (optind == argc) + { + if (opterr) + { + /* 1003.2 specifies the format of this message. */ + fprintf (stderr, "%s: option requires an argument -- %c\n", + argv[0], c); + } + optopt = c; + if (optstring[0] == ':') + c = ':'; + else + c = '?'; + } + else + /* We already incremented `optind' once; + increment it again when taking next ARGV-elt as argument. */ + optarg = argv[optind++]; + nextchar = NULL; + } + } + return c; + } +} + +int +getopt (argc, argv, optstring) + int argc; + char *const *argv; + const char *optstring; +{ + return _getopt_internal (argc, argv, optstring, + (const struct option *) 0, + (int *) 0, + 0); +} + +#endif /* _LIBC or not __GNU_LIBRARY__. */ + +#ifdef TEST + +/* Compile with -DTEST to make an executable for use in testing + the above definition of `getopt'. */ + +int +main (argc, argv) + 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 == EOF) + 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/lib/getopt.h b/lib/getopt.h new file mode 100644 index 00000000..4ac33b71 --- /dev/null +++ b/lib/getopt.h @@ -0,0 +1,129 @@ +/* Declarations for getopt. + Copyright (C) 1989, 90, 91, 92, 93, 94 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, 675 Mass Ave, Cambridge, MA 02139, USA. */ + +#ifndef _GETOPT_H +#define _GETOPT_H 1 + +#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 EOF, 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; + +/* 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 +{ +#if defined (__STDC__) && __STDC__ + const char *name; +#else + char *name; +#endif + /* 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 + +#if defined (__STDC__) && __STDC__ +#ifdef __GNU_LIBRARY__ +/* Many other libraries have conflicting prototypes for getopt, with + differences in the consts, in stdlib.h. To avoid compilation + errors, only prototype getopt for the GNU C library. */ +extern int getopt (int argc, char *const *argv, const char *shortopts); +#else /* not __GNU_LIBRARY__ */ +extern int getopt (); +#endif /* __GNU_LIBRARY__ */ +extern int getopt_long (int argc, char *const *argv, const char *shortopts, + const struct option *longopts, int *longind); +extern int getopt_long_only (int argc, char *const *argv, + const char *shortopts, + const struct option *longopts, int *longind); + +/* Internal only. Users should not call this directly. */ +extern int _getopt_internal (int argc, char *const *argv, + const char *shortopts, + const struct option *longopts, int *longind, + int long_only); +#else /* not __STDC__ */ +extern int getopt (); +extern int getopt_long (); +extern int getopt_long_only (); + +extern int _getopt_internal (); +#endif /* __STDC__ */ + +#ifdef __cplusplus +} +#endif + +#endif /* _GETOPT_H */ diff --git a/lib/getopt1.c b/lib/getopt1.c new file mode 100644 index 00000000..4580211c --- /dev/null +++ b/lib/getopt1.c @@ -0,0 +1,180 @@ +/* getopt_long and getopt_long_only entry points for GNU getopt. + Copyright (C) 1987, 88, 89, 90, 91, 92, 1993, 1994 + 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, 675 Mass Ave, Cambridge, MA 02139, USA. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#include "getopt.h" + +#if !defined (__STDC__) || !__STDC__ +/* This is a separate conditional since some stdc systems + reject `defined (const)'. */ +#ifndef const +#define const +#endif +#endif + +#include <stdio.h> + +/* Comment out all this code if we are using the GNU C Library, and are not + actually compiling the library itself. This code is part of the GNU C + Library, but also included in many other GNU distributions. Compiling + and linking in this code is a waste when using the GNU C library + (especially if it is a shared library). Rather than having every GNU + program understand `configure --with-gnu-libc' and omit the object files, + it is simpler to just do this in the source for each such file. */ + +#if defined (_LIBC) || !defined (__GNU_LIBRARY__) + + +/* This needs to come after some library #include + to get __GNU_LIBRARY__ defined. */ +#ifdef __GNU_LIBRARY__ +#include <stdlib.h> +#else +char *getenv (); +#endif + +#ifndef NULL +#define NULL 0 +#endif + +int +getopt_long (argc, argv, options, long_options, opt_index) + int argc; + char *const *argv; + const char *options; + const struct option *long_options; + int *opt_index; +{ + return _getopt_internal (argc, argv, options, long_options, opt_index, 0); +} + +/* 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 (argc, argv, options, long_options, opt_index) + int argc; + char *const *argv; + const char *options; + const struct option *long_options; + int *opt_index; +{ + return _getopt_internal (argc, argv, options, long_options, opt_index, 1); +} + + +#endif /* _LIBC or not __GNU_LIBRARY__. */ + +#ifdef TEST + +#include <stdio.h> + +int +main (argc, argv) + 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 == EOF) + 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/lib/idcache.c b/lib/idcache.c new file mode 100644 index 00000000..34dcc07c --- /dev/null +++ b/lib/idcache.c @@ -0,0 +1,210 @@ +/* idcache.c -- map user and group IDs, cached for speed + Copyright (C) 1985, 1988, 1989, 1990 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#include <stdio.h> +#include <sys/types.h> +#include <pwd.h> +#include <grp.h> + +#if defined(STDC_HEADERS) || defined(HAVE_STRING_H) +#include <string.h> +#else +#include <strings.h> +#endif + +#ifdef HAVE_UNISTD_H +#include <unistd.h> +#endif +#ifndef _POSIX_VERSION +struct passwd *getpwuid (); +struct passwd *getpwnam (); +struct group *getgrgid (); +struct group *getgrnam (); +#endif + +char *xmalloc (); +char *xstrdup (); + +struct userid +{ + union + { + uid_t u; + gid_t g; + } id; + char *name; + struct userid *next; +}; + +static struct userid *user_alist; + +/* The members of this list have names not in the local passwd file. */ +static struct userid *nouser_alist; + +/* Translate UID to a login name or a stringified number, + with cache. */ + +char * +getuser (uid) + uid_t uid; +{ + register struct userid *tail; + struct passwd *pwent; + char usernum_string[20]; + + for (tail = user_alist; tail; tail = tail->next) + if (tail->id.u == uid) + return tail->name; + + pwent = getpwuid (uid); + tail = (struct userid *) xmalloc (sizeof (struct userid)); + tail->id.u = uid; + if (pwent == 0) + { + sprintf (usernum_string, "%u", (unsigned) uid); + tail->name = xstrdup (usernum_string); + } + else + tail->name = xstrdup (pwent->pw_name); + + /* Add to the head of the list, so most recently used is first. */ + tail->next = user_alist; + user_alist = tail; + return tail->name; +} + +/* Translate USER to a UID, with cache. + Return NULL if there is no such user. + (We also cache which user names have no passwd entry, + so we don't keep looking them up.) */ + +uid_t * +getuidbyname (user) + char *user; +{ + register struct userid *tail; + struct passwd *pwent; + + for (tail = user_alist; tail; tail = tail->next) + /* Avoid a function call for the most common case. */ + if (*tail->name == *user && !strcmp (tail->name, user)) + return &tail->id.u; + + for (tail = nouser_alist; tail; tail = tail->next) + /* Avoid a function call for the most common case. */ + if (*tail->name == *user && !strcmp (tail->name, user)) + return 0; + + pwent = getpwnam (user); + + tail = (struct userid *) xmalloc (sizeof (struct userid)); + tail->name = xstrdup (user); + + /* Add to the head of the list, so most recently used is first. */ + if (pwent) + { + tail->id.u = pwent->pw_uid; + tail->next = user_alist; + user_alist = tail; + return &tail->id.u; + } + + tail->next = nouser_alist; + nouser_alist = tail; + return 0; +} + +/* Use the same struct as for userids. */ +static struct userid *group_alist; +static struct userid *nogroup_alist; + +/* Translate GID to a group name or a stringified number, + with cache. */ + +char * +getgroup (gid) + gid_t gid; +{ + register struct userid *tail; + struct group *grent; + char groupnum_string[20]; + + for (tail = group_alist; tail; tail = tail->next) + if (tail->id.g == gid) + return tail->name; + + grent = getgrgid (gid); + tail = (struct userid *) xmalloc (sizeof (struct userid)); + tail->id.g = gid; + if (grent == 0) + { + sprintf (groupnum_string, "%u", (unsigned int) gid); + tail->name = xstrdup (groupnum_string); + } + else + tail->name = xstrdup (grent->gr_name); + + /* Add to the head of the list, so most recently used is first. */ + tail->next = group_alist; + group_alist = tail; + return tail->name; +} + +/* Translate GROUP to a UID, with cache. + Return NULL if there is no such group. + (We also cache which group names have no group entry, + so we don't keep looking them up.) */ + +gid_t * +getgidbyname (group) + char *group; +{ + register struct userid *tail; + struct group *grent; + + for (tail = group_alist; tail; tail = tail->next) + /* Avoid a function call for the most common case. */ + if (*tail->name == *group && !strcmp (tail->name, group)) + return &tail->id.g; + + for (tail = nogroup_alist; tail; tail = tail->next) + /* Avoid a function call for the most common case. */ + if (*tail->name == *group && !strcmp (tail->name, group)) + return 0; + + grent = getgrnam (group); + + tail = (struct userid *) xmalloc (sizeof (struct userid)); + tail->name = xstrdup (group); + + /* Add to the head of the list, so most recently used is first. */ + if (grent) + { + tail->id.g = grent->gr_gid; + tail->next = group_alist; + group_alist = tail; + return &tail->id.g; + } + + tail->next = nogroup_alist; + nogroup_alist = tail; + return 0; +} diff --git a/lib/listfile.c b/lib/listfile.c index 9b4ed20d..7e32079f 100644 --- a/lib/listfile.c +++ b/lib/listfile.c @@ -1,46 +1,44 @@ /* listfile.c -- display a long listing of a file - Copyright (C) 1991, 1993, 2000, 2004, 2005, 2007 Free Software Foundation, Inc. + Copyright (C) 1991, 1993 Free Software Foundation, Inc. - This program is free software: you can redistribute it and/or modify + 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 3 of the License, or - (at your option) any later version. - + 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, see <http://www.gnu.org/licenses/>. -*/ + along with this program; if not, write to the Free Software + Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ +#ifdef HAVE_CONFIG_H #include <config.h> - -#include <alloca.h> +#endif #include <sys/types.h> #include <sys/stat.h> #include <stdio.h> -#include <stdlib.h> #include <pwd.h> #include <grp.h> #include <time.h> #include <errno.h> -#include <fcntl.h> -#include <string.h> -#include <unistd.h> /* for readlink() */ -#include <openat.h> - -#include "human.h" -#include "xalloc.h" #include "pathmax.h" -#include "error.h" -#include "filemode.h" -#include "dircallback.h" -#include "idcache.h" -#include "listfile.h" +#if defined(HAVE_STRING_H) || defined(STDC_HEADERS) +#include <string.h> +#else +#include <strings.h> +#endif +#ifdef STDC_HEADERS +#include <stdlib.h> +#else +char *getenv (); +extern int errno; +#endif /* Since major is a function on SVR4, we can't use `ifndef major'. */ #ifdef MAJOR_IN_MKDEV @@ -52,31 +50,6 @@ #define HAVE_MAJOR #endif - - - - -#ifdef HAVE_LOCALE_H -#include <locale.h> -#endif - -#if ENABLE_NLS -# include <libintl.h> -# define _(Text) gettext (Text) -#else -# define _(Text) Text -#define textdomain(Domain) -#define bindtextdomain(Package, Directory) -#endif -#ifdef gettext_noop -# define N_(String) gettext_noop (String) -#else -/* See locate.c for explanation as to why not use (String) */ -# define N_(String) String -#endif - - - #ifdef STAT_MACROS_BROKEN #undef S_ISCHR #undef S_ISBLK @@ -89,69 +62,40 @@ #ifndef S_ISBLK #define S_ISBLK(m) (((m) & S_IFMT) == S_IFBLK) #endif -#if defined S_IFLNK && !defined S_ISLNK +#if defined(S_IFLNK) && !defined(S_ISLNK) #define S_ISLNK(m) (((m) & S_IFMT) == S_IFLNK) #endif -/* Get or fake the disk device blocksize. - Usually defined by sys/param.h (if at all). */ -#ifndef DEV_BSIZE -# ifdef BSIZE -# define DEV_BSIZE BSIZE -# else /* !BSIZE */ -# define DEV_BSIZE 4096 -# endif /* !BSIZE */ -#endif /* !DEV_BSIZE */ +#if defined(S_ISLNK) +int readlink (); +#endif /* Extract or fake data from a `struct stat'. - ST_BLKSIZE: Preferred I/O blocksize for the file, in bytes. - ST_NBLOCKS: Number of blocks in the file, including indirect blocks. - ST_NBLOCKSIZE: Size of blocks used when calculating ST_NBLOCKS. */ -#ifndef HAVE_STRUCT_STAT_ST_BLOCKS -# define ST_BLKSIZE(statbuf) DEV_BSIZE -# if defined _POSIX_SOURCE || !defined BSIZE /* fileblocks.c uses BSIZE. */ -# define ST_NBLOCKS(statbuf) \ - (S_ISREG ((statbuf).st_mode) \ - || S_ISDIR ((statbuf).st_mode) \ - ? (statbuf).st_size / ST_NBLOCKSIZE + ((statbuf).st_size % ST_NBLOCKSIZE != 0) : 0) -# else /* !_POSIX_SOURCE && BSIZE */ -# define ST_NBLOCKS(statbuf) \ - (S_ISREG ((statbuf).st_mode) \ - || S_ISDIR ((statbuf).st_mode) \ - ? st_blocks ((statbuf).st_size) : 0) -# endif /* !_POSIX_SOURCE && BSIZE */ -#else /* HAVE_STRUCT_STAT_ST_BLOCKS */ -/* Some systems, like Sequents, return st_blksize of 0 on pipes. */ -# define ST_BLKSIZE(statbuf) ((statbuf).st_blksize > 0 \ - ? (statbuf).st_blksize : DEV_BSIZE) -# if defined hpux || defined __hpux__ || defined __hpux -/* HP-UX counts st_blocks in 1024-byte units. - This loses when mixing HP-UX and BSD filesystems with NFS. */ -# define ST_NBLOCKSIZE 1024 -# else /* !hpux */ -# if defined _AIX && defined _I386 -/* AIX PS/2 counts st_blocks in 4K units. */ -# define ST_NBLOCKSIZE (4 * 1024) -# else /* not AIX PS/2 */ -# if defined _CRAY -# define ST_NBLOCKS(statbuf) \ - (S_ISREG ((statbuf).st_mode) \ - || S_ISDIR ((statbuf).st_mode) \ - ? (statbuf).st_blocks * ST_BLKSIZE(statbuf)/ST_NBLOCKSIZE : 0) -# endif /* _CRAY */ -# endif /* not AIX PS/2 */ -# endif /* !hpux */ -#endif /* HAVE_STRUCT_STAT_ST_BLOCKS */ - -#ifndef ST_NBLOCKS -# define ST_NBLOCKS(statbuf) \ - (S_ISREG ((statbuf).st_mode) \ - || S_ISDIR ((statbuf).st_mode) \ - ? (statbuf).st_blocks : 0) + ST_NBLOCKS: Number of 512-byte blocks in the file + (including indirect blocks). + HP-UX, perhaps uniquely, counts st_blocks in 1024-byte units. + This workaround loses when mixing HP-UX and 4BSD filesystems, though. */ +#ifdef _POSIX_SOURCE +# define ST_NBLOCKS(statp) (((statp)->st_size + 512 - 1) / 512) +#else +# ifndef HAVE_ST_BLOCKS +# define ST_NBLOCKS(statp) (st_blocks ((statp)->st_size)) +# else +# if defined(hpux) || defined(__hpux__) +# define ST_NBLOCKS(statp) ((statp)->st_blocks * 2) +# else +# define ST_NBLOCKS(statp) ((statp)->st_blocks) +# endif +# endif #endif -#ifndef ST_NBLOCKSIZE -# define ST_NBLOCKSIZE 512 +/* Convert B 512-byte blocks to kilobytes if K is nonzero, + otherwise return it unchanged. */ +#define convert_blocks(b, k) ((k) ? ((b) + 1) / 2 : (b)) + +#ifndef _POSIX_VERSION +struct passwd *getpwuid (); +struct group *getgrgid (); #endif #ifdef major /* Might be defined in sys/types.h. */ @@ -163,148 +107,86 @@ #endif #undef HAVE_MAJOR +char *xmalloc (); +void error (); +void mode_string (); -static void print_name (register const char *p, FILE *stream, int literal_control_chars); - - -size_t -file_blocksize(const struct stat *p) -{ - return ST_NBLOCKSIZE; -} - - +char *get_link_name (); +char *getgroup (); +char *getuser (); +void print_name_with_quoting (); /* NAME is the name to print. RELNAME is the path to access it from the current directory. STATP is the results of stat or lstat on it. - Use CURRENT_TIME to decide whether to print yyyy or hh:mm. - Use OUTPUT_BLOCK_SIZE to determine how to print file block counts - and sizes. STREAM is the stdio stream to print on. */ void -list_file (const char *name, - int dirfd, - char *relname, - const struct stat *statp, - time_t current_time, - int output_block_size, - int literal_control_chars, - FILE *stream) +list_file (name, relname, statp, stream) + char *name; + char *relname; + struct stat *statp; + FILE *stream; { - char modebuf[12]; - struct tm const *when_local; - char const *user_name; - char const *group_name; - char hbuf[LONGEST_HUMAN_READABLE + 1]; - -#if HAVE_ST_DM_MODE - /* Cray DMF: look at the file's migrated, not real, status */ - strmode (statp->st_dm_mode, modebuf); -#else - strmode (statp->st_mode, modebuf); -#endif + static int kilobytes = -1; /* -1 = uninitialized, 0 = 512, 1 = 1024. */ + char modebuf[20]; + char timebuf[40]; + time_t current_time = time ((time_t *) 0); - fprintf (stream, "%6s ", - human_readable ((uintmax_t) statp->st_ino, hbuf, - human_ceiling, - 1, 1)); + if (kilobytes == -1) + kilobytes = getenv ("POSIXLY_CORRECT") == 0; + + mode_string (statp->st_mode, modebuf); + modebuf[10] = '\0'; + + strcpy (timebuf, ctime (&statp->st_mtime)); + if (current_time > statp->st_mtime + 6L * 30L * 24L * 60L * 60L /* Old. */ + || current_time < statp->st_mtime - 60L * 60L) /* In the future. */ + { + /* The file is fairly old or in the future. + POSIX says the cutoff is 6 months old; + approximate this by 6*30 days. + Allow a 1 hour slop factor for what is considered "the future", + to allow for NFS server/client clock disagreement. + Show the year instead of the time of day. */ + strcpy (timebuf + 11, timebuf + 19); + } + timebuf[16] = 0; - fprintf (stream, "%4s ", - human_readable ((uintmax_t) ST_NBLOCKS (*statp), hbuf, - human_ceiling, - ST_NBLOCKSIZE, output_block_size)); + fprintf (stream, "%6lu ", statp->st_ino); + fprintf (stream, "%4u ", convert_blocks (ST_NBLOCKS (statp), kilobytes)); - /* modebuf includes the space between the mode and the number of links, - as the POSIX "optional alternate access method flag". */ - fprintf (stream, "%s%3lu ", modebuf, (unsigned long) statp->st_nlink); + /* The space between the mode and the number of links is the POSIX + "optional alternate access method flag". */ + fprintf (stream, "%s %3u ", modebuf, statp->st_nlink); - user_name = getuser (statp->st_uid); - if (user_name) - fprintf (stream, "%-8s ", user_name); - else - fprintf (stream, "%-8lu ", (unsigned long) statp->st_uid); + fprintf (stream, "%-8.8s ", getuser (statp->st_uid)); - group_name = getgroup (statp->st_gid); - if (group_name) - fprintf (stream, "%-8s ", group_name); - else - fprintf (stream, "%-8lu ", (unsigned long) statp->st_gid); + fprintf (stream, "%-8.8s ", getgroup (statp->st_gid)); if (S_ISCHR (statp->st_mode) || S_ISBLK (statp->st_mode)) #ifdef HAVE_ST_RDEV - fprintf (stream, "%3lu, %3lu ", - (unsigned long) major (statp->st_rdev), - (unsigned long) minor (statp->st_rdev)); + fprintf (stream, "%3u, %3u ", major (statp->st_rdev), minor (statp->st_rdev)); #else fprintf (stream, " "); #endif else - fprintf (stream, "%8s ", - human_readable ((uintmax_t) statp->st_size, hbuf, - human_ceiling, - 1, - output_block_size < 0 ? output_block_size : 1)); + fprintf (stream, "%8lu ", statp->st_size); - if ((when_local = localtime (&statp->st_mtime))) - { - char init_bigbuf[256]; - char *buf = init_bigbuf; - size_t bufsize = sizeof init_bigbuf; + fprintf (stream, "%s ", timebuf + 4); - /* Use strftime rather than ctime, because the former can produce - locale-dependent names for the month (%b). - - Output the year if the file is fairly old or in the future. - POSIX says the cutoff is 6 months old; - approximate this by 6*30 days. - Allow a 1 hour slop factor for what is considered "the future", - to allow for NFS server/client clock disagreement. */ - char const *fmt = - ((current_time - 6 * 30 * 24 * 60 * 60 <= statp->st_mtime - && statp->st_mtime <= current_time + 60 * 60) - ? "%b %e %H:%M" - : "%b %e %Y"); - - while (!strftime (buf, bufsize, fmt, when_local)) - buf = alloca (bufsize *= 2); - - fprintf (stream, "%s ", buf); - } - else - { - /* The time cannot be represented as a local time; - print it as a huge integer number of seconds. */ - int width = 12; - - if (statp->st_mtime < 0) - { - char const *num = human_readable (- (uintmax_t) statp->st_mtime, - hbuf, human_ceiling, 1, 1); - int sign_width = width - strlen (num); - fprintf (stream, "%*s%s ", - sign_width < 0 ? 0 : sign_width, "-", num); - } - else - fprintf (stream, "%*s ", width, - human_readable ((uintmax_t) statp->st_mtime, hbuf, - human_ceiling, - 1, 1)); - } - - print_name (name, stream, literal_control_chars); + print_name_with_quoting (name, stream); #ifdef S_ISLNK if (S_ISLNK (statp->st_mode)) { - char *linkname = get_link_name_at (name, dirfd, relname); + char *linkname = get_link_name (name, relname); if (linkname) { fputs (" -> ", stream); - print_name (linkname, stream, literal_control_chars); + print_name_with_quoting (linkname, stream); free (linkname); } } @@ -312,16 +194,10 @@ list_file (const char *name, putc ('\n', stream); } - -static void -print_name_without_quoting (const char *p, FILE *stream) -{ - fprintf(stream, "%s", p); -} - - -static void -print_name_with_quoting (register const char *p, FILE *stream) +void +print_name_with_quoting (p, stream) + register char *p; + FILE *stream; { register unsigned char c; @@ -370,17 +246,11 @@ print_name_with_quoting (register const char *p, FILE *stream) } } -static void print_name (register const char *p, FILE *stream, int literal_control_chars) -{ - if (literal_control_chars) - print_name_without_quoting(p, stream); - else - print_name_with_quoting(p, stream); -} - #ifdef S_ISLNK -static char * -get_link_name (const char *name, char *relname) +char * +get_link_name (name, relname) + char *name; + char *relname; { register char *linkname; register int linklen; @@ -389,7 +259,7 @@ get_link_name (const char *name, char *relname) mount points with some automounters. So allocate a pessimistic PATH_MAX + 1 bytes. */ #define LINK_BUF PATH_MAX - linkname = xmalloc (LINK_BUF + 1); + linkname = (char *) xmalloc (LINK_BUF + 1); linklen = readlink (relname, linkname, LINK_BUF); if (linklen < 0) { @@ -400,34 +270,4 @@ get_link_name (const char *name, char *relname) linkname[linklen] = '\0'; return linkname; } - -struct link_name_args -{ - const char *name; - char *relname; - char *result; -}; - -static int -get_link_name_cb(void *context) -{ - struct link_name_args *args = context; - args->result = get_link_name(args->name, args->relname); - return 0; -} - -char * -get_link_name_at (const char *name, int dirfd, char *relname) -{ - struct link_name_args args; - args.result = NULL; - args.name = name; - args.relname = relname; - if (0 == run_in_dir(dirfd, get_link_name_cb, &args)) - return args.result; - else - return NULL; -} - - #endif diff --git a/lib/memcmp.c b/lib/memcmp.c new file mode 100644 index 00000000..339b5818 --- /dev/null +++ b/lib/memcmp.c @@ -0,0 +1,32 @@ +/* memcmp.c -- compare memory. + Return: + <0 if S1 < S2, + 0 if strings are identical, + >0 if S1 > S2. + Stops looking after N characters. Doesn't stop at nulls. + In the public domain. + By David MacKenzie <djm@gnu.ai.mit.edu>. */ + +#include <sys/types.h> + +int +#if __STDC__ +memcmp (void const *v1, void const *v2, size_t n) +{ + register char *s1 = (char *) v1, *s2 = (char *) v2; +#else +memcmp (s1, s2, n) + register char *s1, *s2; + register unsigned n; +{ +#endif + register int diff; + + while (n--) + { + diff = *s1++ - *s2++; + if (diff) + return diff; + } + return 0; +} diff --git a/lib/memset.c b/lib/memset.c new file mode 100644 index 00000000..0e819f20 --- /dev/null +++ b/lib/memset.c @@ -0,0 +1,29 @@ +/* memset.c -- set an area of memory to a given value + Copyright (C) 1991 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +char * +memset (str, c, len) + char *str; + int c; + unsigned len; +{ + register char *st = str; + + while (len-- > 0) + *st++ = c; + return str; +} diff --git a/lib/mktime.c b/lib/mktime.c new file mode 100644 index 00000000..d441d8d8 --- /dev/null +++ b/lib/mktime.c @@ -0,0 +1,502 @@ +/* Copyright (C) 1993, 1994 Free Software Foundation, Inc. + Contributed by Noel Cragg (noel@cs.oberlin.edu), with fixes by + Michael E. Calwas (calwas@ttd.teradyne.com) and + Wade Hampton (tasi029@tmn.com). + +This file is part of the GNU C Library. + +The GNU C Library is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public License as +published by the Free Software Foundation; either version 2 of the +License, or (at your option) any later version. + +The GNU C Library 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 +Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with the GNU C Library; see the file COPYING.LIB. If +not, write to the Free Software Foundation, Inc., 675 Mass Ave, +Cambridge, MA 02139, USA. */ + +/* Define this to have a standalone program to test this implementation of + mktime. */ +/* #define DEBUG */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#include <sys/types.h> /* Some systems define `time_t' here. */ +#include <time.h> + + +#ifndef __isleap +/* Nonzero if YEAR is a leap year (every 4 years, + except every 100th isn't, and every 400th is). */ +#define __isleap(year) \ + ((year) % 4 == 0 && ((year) % 100 != 0 || (year) % 400 == 0)) +#endif + +#ifndef __P +#if defined (__GNUC__) || (defined (__STDC__) && __STDC__) +#define __P(args) args +#else +#define __P(args) () +#endif /* GCC. */ +#endif /* Not __P. */ + +/* How many days are in each month. */ +const unsigned short int __mon_lengths[2][12] = + { + /* Normal years. */ + { 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 }, + /* Leap years. */ + { 31, 29, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 } + }; + + +static int times_through_search; /* This library routine should never + hang -- make sure we always return + when we're searching for a value */ + + +#ifdef DEBUG + +#include <stdio.h> +#include <ctype.h> + +int debugging_enabled = 0; + +/* Print the values in a `struct tm'. */ +static void +printtm (it) + struct tm *it; +{ + printf ("%02d/%02d/%04d %02d:%02d:%02d (%s) yday:%03d dst:%d gmtoffset:%ld", + it->tm_mon + 1, + it->tm_mday, + it->tm_year + 1900, + it->tm_hour, + it->tm_min, + it->tm_sec, + it->tm_zone, + it->tm_yday, + it->tm_isdst, + it->tm_gmtoff); +} +#endif + + +static time_t +dist_tm (t1, t2) + struct tm *t1; + struct tm *t2; +{ + time_t distance = 0; + unsigned long int v1, v2; + int diff_flag = 0; + + v1 = v2 = 0; + +#define doit(x, secs) \ + v1 += t1->x * secs; \ + v2 += t2->x * secs; \ + if (!diff_flag) \ + { \ + if (t1->x < t2->x) \ + diff_flag = -1; \ + else if (t1->x > t2->x) \ + diff_flag = 1; \ + } + + doit (tm_year, 31536000); /* Okay, not all years have 365 days. */ + doit (tm_mon, 2592000); /* Okay, not all months have 30 days. */ + doit (tm_mday, 86400); + doit (tm_hour, 3600); + doit (tm_min, 60); + doit (tm_sec, 1); + +#undef doit + + /* We should also make sure that the sign of DISTANCE is correct -- if + DIFF_FLAG is positive, the distance should be positive and vice versa. */ + + distance = (v1 > v2) ? (v1 - v2) : (v2 - v1); + if (diff_flag < 0) + distance = -distance; + + if (times_through_search > 20) /* Arbitrary # of calls, but makes sure we + never hang if there's a problem with + this algorithm. */ + { + distance = diff_flag; + } + + /* We need this DIFF_FLAG business because it is forseeable that the + distance may be zero when, in actuality, the two structures are + different. This is usually the case when the dates are 366 days apart + and one of the years is a leap year. */ + + if (distance == 0 && diff_flag) + distance = 86400 * diff_flag; + + return distance; +} + + +/* MKTIME converts the values in a struct tm to a time_t. The values + in tm_wday and tm_yday are ignored; other values can be put outside + of legal ranges since they will be normalized. This routine takes + care of that normalization. */ + +void +do_normalization (tmptr) + struct tm *tmptr; +{ + +#define normalize(foo,x,y,bar); \ + while (tmptr->foo < x) \ + { \ + tmptr->bar--; \ + tmptr->foo = (y - (x - tmptr->foo) + 1); \ + } \ + while (tmptr->foo > y) \ + { \ + tmptr->foo = (x + (tmptr->foo - y) - 1); \ + tmptr->bar++; \ + } + + normalize (tm_sec, 0, 59, tm_min); + normalize (tm_min, 0, 59, tm_hour); + normalize (tm_hour, 0, 23, tm_mday); + + /* Do the month first, so day range can be found. */ + normalize (tm_mon, 0, 11, tm_year); + + /* Since the day range modifies the month, we should be careful how + we reference the array of month lengths -- it is possible that + the month will go negative, hence the modulo... + + Also, tm_year is the year - 1900, so we have to 1900 to have it + work correctly. */ + + normalize (tm_mday, 1, + __mon_lengths[__isleap (tmptr->tm_year + 1900)] + [((tmptr->tm_mon < 0) + ? (12 + (tmptr->tm_mon % 12)) + : (tmptr->tm_mon % 12)) ], + tm_mon); + + /* Do the month again, because the day may have pushed it out of range. */ + normalize (tm_mon, 0, 11, tm_year); + + /* Do the day again, because the month may have changed the range. */ + normalize (tm_mday, 1, + __mon_lengths[__isleap (tmptr->tm_year + 1900)] + [((tmptr->tm_mon < 0) + ? (12 + (tmptr->tm_mon % 12)) + : (tmptr->tm_mon % 12)) ], + tm_mon); + +#ifdef DEBUG + if (debugging_enabled) + { + printf (" After normalizing:\n "); + printtm (tmptr); + putchar ('\n'); + } +#endif + +} + + +/* Here's where the work gets done. */ + +#define BAD_STRUCT_TM ((time_t) -1) + +time_t +_mktime_internal (timeptr, producer) + struct tm *timeptr; + struct tm *(*producer) __P ((const time_t *)); +{ + struct tm our_tm; /* our working space */ + struct tm *me = &our_tm; /* a pointer to the above */ + time_t result; /* the value we return */ + + *me = *timeptr; /* copy the struct tm that was passed + in by the caller */ + + + /***************************/ + /* Normalize the structure */ + /***************************/ + + /* This routine assumes that the value of TM_ISDST is -1, 0, or 1. + If the user didn't pass it in that way, fix it. */ + + if (me->tm_isdst > 0) + me->tm_isdst = 1; + else if (me->tm_isdst < 0) + me->tm_isdst = -1; + + do_normalization (me); + + /* Get out of here if it's not possible to represent this struct. + If any of the values in the normalized struct tm are negative, + our algorithms won't work. Luckily, we only need to check the + year at this point; normalization guarantees that all values will + be in correct ranges EXCEPT the year. */ + + if (me->tm_year < 0) + return BAD_STRUCT_TM; + + /*************************************************/ + /* Find the appropriate time_t for the structure */ + /*************************************************/ + + /* Modified b-search -- make intelligent guesses as to where the + time might lie along the timeline, assuming that our target time + lies a linear distance (w/o considering time jumps of a + particular region). + + Assume that time does not fluctuate at all along the timeline -- + e.g., assume that a day will always take 86400 seconds, etc. -- + and come up with a hypothetical value for the time_t + representation of the struct tm TARGET, in relation to the guess + variable -- it should be pretty close! + + After testing this, the maximum number of iterations that I had + on any number that I tried was 3! Not bad. + + The reason this is not a subroutine is that we will modify some + fields in the struct tm (yday and mday). I've never felt good + about side-effects when writing structured code... */ + + { + struct tm *guess_tm; + time_t guess = 0; + time_t distance = 0; + time_t last_distance = 0; + + times_through_search = 0; + + do + { + guess += distance; + + times_through_search++; + + guess_tm = (*producer) (&guess); + +#ifdef DEBUG + if (debugging_enabled) + { + printf (" Guessing time_t == %d\n ", (int) guess); + printtm (guess_tm); + putchar ('\n'); + } +#endif + + /* How far is our guess from the desired struct tm? */ + distance = dist_tm (me, guess_tm); + + /* Handle periods of time where a period of time is skipped. + For example, 2:15 3 April 1994 does not exist, because DST + is in effect. The distance function will alternately + return values of 3600 and -3600, because it doesn't know + that the requested time doesn't exist. In these situations + (even if the skip is not exactly an hour) the distances + returned will be the same, but alternating in sign. We + want the later time, so check to see that the distance is + oscillating and we've chosen the correct of the two + possibilities. + + Useful: 3 Apr 94 765356300, 30 Oct 94 783496000 */ + + if ((distance == -last_distance) && (distance < last_distance)) + { + /* If the caller specified that the DST flag was off, it's + not possible to represent this time. */ + if (me->tm_isdst == 0) + { +#ifdef DEBUG + printf (" Distance is oscillating -- dst flag nixes struct!\n"); +#endif + return BAD_STRUCT_TM; + } + +#ifdef DEBUG + printf (" Distance is oscillating -- chose the later time.\n"); +#endif + distance = 0; + } + + if ((distance == 0) && (me->tm_isdst != -1) + && (me->tm_isdst != guess_tm->tm_isdst)) + { + /* If we're in this code, we've got the right time but the + wrong daylight savings flag. We need to move away from + the time that we have and approach the other time from + the other direction. That is, if I've requested the + non-DST version of a time and I get the DST version + instead, I want to put us forward in time and search + backwards to get the other time. I checked all of the + configuration files for the tz package -- no entry + saves more than two hours, so I think we'll be safe by + moving 24 hours in one direction. IF THE AMOUNT OF + TIME SAVED IN THE CONFIGURATION FILES CHANGES, THIS + VALUE MAY NEED TO BE ADJUSTED. Luckily, we can never + have more than one level of overlaps, or this would + never work. */ + +#define SKIP_VALUE 86400 + + if (guess_tm->tm_isdst == 0) + /* we got the later one, but want the earlier one */ + distance = -SKIP_VALUE; + else + distance = SKIP_VALUE; + +#ifdef DEBUG + printf (" Got the right time, wrong DST value -- adjusting\n"); +#endif + } + + last_distance = distance; + + } while (distance != 0); + + /* Check to see that the dst flag matches */ + + if (me->tm_isdst != -1) + { + if (me->tm_isdst != guess_tm->tm_isdst) + { +#ifdef DEBUG + printf (" DST flag doesn't match! FIXME?\n"); +#endif + return BAD_STRUCT_TM; + } + } + + result = guess; /* Success! */ + + /* On successful completion, the values of tm_wday and tm_yday + have to be set appropriately. */ + + /* me->tm_yday = guess_tm->tm_yday; + me->tm_mday = guess_tm->tm_mday; */ + + *me = *guess_tm; + } + + /* Update the caller's version of the structure */ + + *timeptr = *me; + + return result; +} + +time_t +#ifdef DEBUG /* make it work even if the system's + libc has it's own mktime routine */ +my_mktime (timeptr) +#else +mktime (timeptr) +#endif + struct tm *timeptr; +{ + return _mktime_internal (timeptr, localtime); +} + +#ifdef DEBUG +void +main (argc, argv) + int argc; + char *argv[]; +{ + int time; + int result_time; + struct tm *tmptr; + + if (argc == 1) + { + long q; + + printf ("starting long test...\n"); + + for (q = 10000000; q < 1000000000; q += 599) + { + struct tm *tm = localtime ((time_t *) &q); + if ((q % 10000) == 0) { printf ("%ld\n", q); fflush (stdout); } + if (q != my_mktime (tm)) + { printf ("failed for %ld\n", q); fflush (stdout); } + } + + printf ("test finished\n"); + + exit (0); + } + + if (argc != 2) + { + printf ("wrong # of args\n"); + exit (0); + } + + debugging_enabled = 1; /* We want to see the info */ + + ++argv; + time = atoi (*argv); + + tmptr = localtime ((time_t *) &time); + printf ("Localtime tells us that a time_t of %d represents\n ", time); + printtm (tmptr); + putchar ('\n'); + + printf (" Given localtime's return val, mktime returns %d which is\n ", + (int) my_mktime (tmptr)); + printtm (tmptr); + putchar ('\n'); + +#if 0 + tmptr->tm_sec -= 20; + tmptr->tm_min -= 20; + tmptr->tm_hour -= 20; + tmptr->tm_mday -= 20; + tmptr->tm_mon -= 20; + tmptr->tm_year -= 20; + tmptr->tm_gmtoff -= 20000; /* This has no effect! */ + tmptr->tm_zone = NULL; /* Nor does this! */ + tmptr->tm_isdst = -1; +#endif + + tmptr->tm_hour += 1; + tmptr->tm_isdst = -1; + + printf ("\n\nchanged ranges: "); + printtm (tmptr); + putchar ('\n'); + + result_time = my_mktime (tmptr); + printf ("\nmktime: %d\n", result_time); + + tmptr->tm_isdst = 0; + + printf ("\n\nchanged ranges: "); + printtm (tmptr); + putchar ('\n'); + + result_time = my_mktime (tmptr); + printf ("\nmktime: %d\n", result_time); +} +#endif /* DEBUG */ + + +/* +Local Variables: +compile-command: "gcc -g mktime.c -o mktime -DDEBUG" +End: +*/ diff --git a/lib/modechange.c b/lib/modechange.c new file mode 100644 index 00000000..3862599e --- /dev/null +++ b/lib/modechange.c @@ -0,0 +1,341 @@ +/* modechange.c -- file mode manipulation + Copyright (C) 1989, 1990 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +/* Written by David MacKenzie <djm@ai.mit.edu> */ + +/* The ASCII mode string is compiled into a linked list of `struct + modechange', which can then be applied to each file to be changed. + We do this instead of re-parsing the ASCII string for each file + because the compiled form requires less computation to use; when + changing the mode of many files, this probably results in a + performance gain. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#include <sys/types.h> +#include <sys/stat.h> +#include "modechange.h" + +#ifdef STDC_HEADERS +#include <stdlib.h> +#else +char *malloc (); +#endif + +#ifndef NULL +#define NULL 0 +#endif + +#ifdef STAT_MACROS_BROKEN +#undef S_ISDIR +#endif /* STAT_MACROS_BROKEN. */ + +#if !defined(S_ISDIR) && defined(S_IFDIR) +#define S_ISDIR(m) (((m) & S_IFMT) == S_IFDIR) +#endif + +/* Return newly allocated memory to hold one element of type TYPE. */ +#define talloc(type) ((type *) malloc (sizeof (type))) + +#define isodigit(c) ((c) >= '0' && (c) <= '7') + +static int oatoi (); + +/* Return a linked list of file mode change operations created from + MODE_STRING, an ASCII string that contains either an octal number + specifying an absolute mode, or symbolic mode change operations with + the form: + [ugoa...][[+-=][rwxXstugo...]...][,...] + MASKED_OPS is a bitmask indicating which symbolic mode operators (=+-) + should not affect bits set in the umask when no users are given. + Operators not selected in MASKED_OPS ignore the umask. + + Return MODE_INVALID if `mode_string' does not contain a valid + representation of file mode change operations; + return MODE_MEMORY_EXHAUSTED if there is insufficient memory. */ + +struct mode_change * +mode_compile (mode_string, masked_ops) + register char *mode_string; + unsigned masked_ops; +{ + struct mode_change *head; /* First element of the linked list. */ + struct mode_change *change; /* An element of the linked list. */ + int i; /* General purpose temporary. */ + unsigned short umask_value; /* The umask value (surprise). */ + unsigned short affected_bits; /* Which bits in the mode are operated on. */ + unsigned short affected_masked; /* `affected_bits' modified by umask. */ + unsigned ops_to_mask; /* Operators to actually use umask on. */ + + i = oatoi (mode_string); + if (i >= 0) + { + if (i > 07777) + return MODE_INVALID; + head = talloc (struct mode_change); + if (head == NULL) + return MODE_MEMORY_EXHAUSTED; + head->next = NULL; + head->op = '='; + head->flags = 0; + head->value = i; + head->affected = 07777; /* Affect all permissions. */ + return head; + } + + umask_value = umask (0); + umask (umask_value); /* Restore the old value. */ + + head = NULL; +#ifdef lint + change = NULL; +#endif + --mode_string; + + /* One loop iteration for each "ugoa...=+-rwxXstugo...[=+-rwxXstugo...]". */ + do + { + affected_bits = 0; + ops_to_mask = 0; + /* Turn on all the bits in `affected_bits' for each group given. */ + for (++mode_string;; ++mode_string) + switch (*mode_string) + { + case 'u': + affected_bits |= 04700; + break; + case 'g': + affected_bits |= 02070; + break; + case 'o': + affected_bits |= 01007; + break; + case 'a': + affected_bits |= 07777; + break; + default: + goto no_more_affected; + } + + no_more_affected: + /* If none specified, affect all bits, except perhaps those + set in the umask. */ + if (affected_bits == 0) + { + affected_bits = 07777; + ops_to_mask = masked_ops; + } + + while (*mode_string == '=' || *mode_string == '+' || *mode_string == '-') + { + /* Add the element to the tail of the list, so the operations + are performed in the correct order. */ + if (head == NULL) + { + head = talloc (struct mode_change); + if (head == NULL) + return MODE_MEMORY_EXHAUSTED; + change = head; + } + else + { + change->next = talloc (struct mode_change); + if (change->next == NULL) + { + mode_free (change); + return MODE_MEMORY_EXHAUSTED; + } + change = change->next; + } + + change->next = NULL; + change->op = *mode_string; /* One of "=+-". */ + affected_masked = affected_bits; + if (ops_to_mask & (*mode_string == '=' ? MODE_MASK_EQUALS + : *mode_string == '+' ? MODE_MASK_PLUS + : MODE_MASK_MINUS)) + affected_masked &= ~umask_value; + change->affected = affected_masked; + change->value = 0; + change->flags = 0; + + /* Set `value' according to the bits set in `affected_masked'. */ + for (++mode_string;; ++mode_string) + switch (*mode_string) + { + case 'r': + change->value |= 00444 & affected_masked; + break; + case 'w': + change->value |= 00222 & affected_masked; + break; + case 'X': + change->flags |= MODE_X_IF_ANY_X; + /* Fall through. */ + case 'x': + change->value |= 00111 & affected_masked; + break; + case 's': + /* Set the setuid/gid bits if `u' or `g' is selected. */ + change->value |= 06000 & affected_masked; + break; + case 't': + /* Set the "save text image" bit if `o' is selected. */ + change->value |= 01000 & affected_masked; + break; + case 'u': + /* Set the affected bits to the value of the `u' bits + on the same file. */ + if (change->value) + goto invalid; + change->value = 00700; + change->flags |= MODE_COPY_EXISTING; + break; + case 'g': + /* Set the affected bits to the value of the `g' bits + on the same file. */ + if (change->value) + goto invalid; + change->value = 00070; + change->flags |= MODE_COPY_EXISTING; + break; + case 'o': + /* Set the affected bits to the value of the `o' bits + on the same file. */ + if (change->value) + goto invalid; + change->value = 00007; + change->flags |= MODE_COPY_EXISTING; + break; + default: + goto no_more_values; + } + no_more_values:; + } + } while (*mode_string == ','); + if (*mode_string == 0) + return head; +invalid: + mode_free (head); + return MODE_INVALID; +} + +/* Return file mode OLDMODE, adjusted as indicated by the list of change + operations CHANGES. If OLDMODE is a directory, the type `X' + change affects it even if no execute bits were set in OLDMODE. + The returned value has the S_IFMT bits cleared. */ + +unsigned short +mode_adjust (oldmode, changes) + unsigned oldmode; + register struct mode_change *changes; +{ + unsigned short newmode; /* The adjusted mode and one operand. */ + unsigned short value; /* The other operand. */ + + newmode = oldmode & 07777; + + for (; changes; changes = changes->next) + { + if (changes->flags & MODE_COPY_EXISTING) + { + /* Isolate in `value' the bits in `newmode' to copy, given in + the mask `changes->value'. */ + value = newmode & changes->value; + + if (changes->value & 00700) + /* Copy `u' permissions onto `g' and `o'. */ + value |= (value >> 3) | (value >> 6); + else if (changes->value & 00070) + /* Copy `g' permissions onto `u' and `o'. */ + value |= (value << 3) | (value >> 3); + else + /* Copy `o' permissions onto `u' and `g'. */ + value |= (value << 3) | (value << 6); + + /* In order to change only `u', `g', or `o' permissions, + or some combination thereof, clear unselected bits. + This can not be done in mode_compile because the value + to which the `changes->affected' mask is applied depends + on the old mode of each file. */ + value &= changes->affected; + } + else + { + value = changes->value; + /* If `X', do not affect the execute bits if the file is not a + directory and no execute bits are already set. */ + if ((changes->flags & MODE_X_IF_ANY_X) + && !S_ISDIR (oldmode) + && (newmode & 00111) == 0) + value &= ~00111; /* Clear the execute bits. */ + } + + switch (changes->op) + { + case '=': + /* Preserve the previous values in `newmode' of bits that are + not affected by this change operation. */ + newmode = (newmode & ~changes->affected) | value; + break; + case '+': + newmode |= value; + break; + case '-': + newmode &= ~value; + break; + } + } + return newmode; +} + +/* Free the memory used by the list of file mode change operations + CHANGES. */ + +void +mode_free (changes) + register struct mode_change *changes; +{ + register struct mode_change *next; + + while (changes) + { + next = changes->next; + free (changes); + changes = next; + } +} + +/* Return a positive integer containing the value of the ASCII + octal number S. If S is not an octal number, return -1. */ + +static int +oatoi (s) + char *s; +{ + register int i; + + if (*s == 0) + return -1; + for (i = 0; isodigit (*s); ++s) + i = i * 8 + *s - '0'; + if (*s) + return -1; + return i; +} diff --git a/lib/modechange.h b/lib/modechange.h new file mode 100644 index 00000000..4a298830 --- /dev/null +++ b/lib/modechange.h @@ -0,0 +1,55 @@ +/* modechange.h -- definitions for file mode manipulation + Copyright (C) 1989, 1990 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +/* Masks for the `flags' field in a `struct mode_change'. */ + +/* Affect the execute bits only if at least one execute bit is set already, + or if the file is a directory. */ +#define MODE_X_IF_ANY_X 01 + +/* If set, copy some existing permissions for u, g, or o onto the other two. + Which of u, g, or o is copied is determined by which bits are set in the + `value' field. */ +#define MODE_COPY_EXISTING 02 + +struct mode_change +{ + char op; /* One of "=+-". */ + char flags; /* Special operations. */ + unsigned short affected; /* Set for u/g/o/s/s/t, if to be affected. */ + unsigned short value; /* Bits to add/remove. */ + struct mode_change *next; /* Link to next change in list. */ +}; + +/* Masks for mode_compile argument. */ +#define MODE_MASK_EQUALS 1 +#define MODE_MASK_PLUS 2 +#define MODE_MASK_MINUS 4 + +/* Error return values for mode_compile. */ +#define MODE_INVALID (struct mode_change *) 0 +#define MODE_MEMORY_EXHAUSTED (struct mode_change *) 1 + +#ifdef __STDC__ +struct mode_change *mode_compile (char *, unsigned); +unsigned short mode_adjust (unsigned, struct mode_change *); +void mode_free (struct mode_change *); +#else +struct mode_change *mode_compile (); +unsigned short mode_adjust (); +void mode_free (); +#endif diff --git a/lib/modetype.h b/lib/modetype.h index 2c99ecaf..fd4092bf 100644 --- a/lib/modetype.h +++ b/lib/modetype.h @@ -1,28 +1,25 @@ /* modetype.h -- file type bits definitions for POSIX systems Requires sys/types.h sys/stat.h. - Copyright (C) 1990, 2007 Free Software Foundation, Inc. + Copyright (C) 1990 Free Software Foundation, Inc. - This program is free software: you can redistribute it and/or modify + 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 3 of the License, or - (at your option) any later version. - + 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, see <http://www.gnu.org/licenses/>. -*/ + along with this program; if not, write to the Free Software + Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ /* POSIX.1 doesn't mention the S_IFMT bits; instead, it uses S_IStype test macros. To make storing file types more convenient, define them; the values don't need to correspond to what the kernel uses, because of the way we use them. */ -#ifndef INC_MODETYPE_H -#define INC_MODETYPE_H 1 - #ifndef S_IFMT /* Doesn't have traditional Unix macros. */ #define S_IFBLK 1 #define S_IFCHR 2 @@ -37,9 +34,6 @@ #ifdef S_ISSOCK #define S_IFSOCK 64 #endif -#ifdef S_ISDOOR -#define S_IFDOOR 128 -#endif #endif /* !S_IFMT */ #ifdef STAT_MACROS_BROKEN @@ -50,7 +44,6 @@ #undef S_ISFIFO #undef S_ISLNK #undef S_ISSOCK -#undef S_ISDOOR #undef S_ISMPB #undef S_ISMPC #undef S_ISNWK @@ -79,64 +72,10 @@ #if !defined(S_ISSOCK) && defined(S_IFSOCK) #define S_ISSOCK(m) (((m) & S_IFMT) == S_IFSOCK) #endif -#if !defined(S_ISDOOR) && defined(S_IFDOOR) -#define S_ISDOOR(m) (((m) & S_IFMT) == S_IFDOOR) -#endif #if !defined(S_ISMPB) && defined(S_IFMPB) /* V7 */ -/* Also available on Coherent, according to - * Albert D. Cahalan (acahalan@cs.uml.edu) - */ -#define S_ISMPB(m) (((m) & S_IFMT) == S_IFMPB) /* multiplexed block device */ -#define S_ISMPC(m) (((m) & S_IFMT) == S_IFMPC) /* multiplexed char device */ -/* GNU BFD library source uses type letter 'm' for these */ +#define S_ISMPB(m) (((m) & S_IFMT) == S_IFMPB) +#define S_ISMPC(m) (((m) & S_IFMT) == S_IFMPC) #endif - #if !defined(S_ISNWK) && defined(S_IFNWK) /* HP/UX */ -/* Apparently HPUX ls gives 'n' as the type letter for these. */ #define S_ISNWK(m) (((m) & S_IFMT) == S_IFNWK) #endif - -#endif - -/* The above macros don't handle - * /bin/ls letters Mode What is it? - * S_IFNAM (Xenix "name files") - * H S_ISCDF (HPUX Context Dependent Files) - * S_IFCMP - * S_IFSHAD - */ - -/* -In message <199907051927.PAA01106@jupiter.cs.uml.edu> -Albert Cahalan wrote:- - -BTW, I believe many of these can't actually exist on disk. -Some of these (like S_IFSHAD AFAIK) are not seen by userspace. - -hex name ls octal description -0000 000000 SCO out-of-service inode, BSD unknown type -1000 S_IFIFO p| 010000 fifo (named pipe) -2000 S_IFCHR c 020000 character special -3000 S_IFMPC 030000 multiplexed character device (Coherent) -4000 S_IFDIR d/ 040000 directory -5000 S_IFNAM 050000 XENIX special named file -6000 S_IFBLK b 060000 block special -7000 S_IFMPB 070000 multiplexed block device (Coherent) -8000 S_IFREG - 100000 regular -9000 S_IFCMP 110000 VxFS compressed (file?) -9000 S_IFNWK 110000 HP-UX network special -a000 S_IFLNK l@ 120000 symbolic link -b000 S_IFSHAD 130000 Solaris shadow inode for ACL -c000 S_IFSOCK s= 140000 socket (also "S_IFSOC" on VxFS) -d000 S_IFDOOR D 150000 Solaris door -e000 S_IFWHT w% 160000 BSD whiteout (not used for inode) -f000 S_IFMT 170000 mask (not used for inode) -hex name ls octal description -0200 S_ISVTX 001000 save swapped text even after use -0400 S_ISGID 002000 set group ID on execution -0400 S_ENFMT 002000 SysV forced file locking (shared w/ S_ISGID) -0800 S_CDF 004000 HP-UX hidden directory -0800 S_ISUID 004000 set user ID on execution - - -*/ diff --git a/lib/nextelem.c b/lib/nextelem.c index a4bd8dd6..4b9ddc1f 100644 --- a/lib/nextelem.c +++ b/lib/nextelem.c @@ -1,28 +1,29 @@ /* Return the next element of a path. - Copyright (C) 1992,2005 Free Software Foundation, Inc. + Copyright (C) 1992 Free Software Foundation, Inc. - This program is free software: you can redistribute it and/or modify + 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 3 of the License, or - (at your option) any later version. - + 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, see <http://www.gnu.org/licenses/>. -*/ + along with this program; if not, write to the Free Software + Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ -/* Written by David MacKenzie <djm@gnu.org>, +/* Written by David MacKenzie <djm@gnu.ai.mit.edu>, inspired by John P. Rouillard <rouilj@cs.umb.edu>. */ +#ifdef HAVE_CONFIG_H #include <config.h> - +#endif #include <stdio.h> -#if defined HAVE_STRING_H || defined STDC_HEADERS +#if defined(HAVE_STRING_H) || defined(STDC_HEADERS) #include <string.h> #else #include <strings.h> @@ -30,12 +31,9 @@ #define strchr index #endif #endif -#if defined STDC_HEADERS -#include <stdlib.h> -#endif - -#include "nextelem.h" +char *strdup (); +void free (); /* Return the next element of a colon-separated path. A null entry in the path is equivalent to "." (the current directory). @@ -45,7 +43,8 @@ return NULL if there are no more elements. */ char * -next_element (const char *new_path, int curdir_ok) +next_element (new_path) + char *new_path; { static char *path = NULL; /* Freshly allocated copy of NEW_PATH. */ static char *end; /* Start of next element to return. */ @@ -66,7 +65,7 @@ next_element (const char *new_path, int curdir_ok) if (final_colon) { final_colon = 0; - return curdir_ok ? "." : ""; + return "."; } return NULL; } @@ -79,7 +78,7 @@ next_element (const char *new_path, int curdir_ok) { /* An empty path element. */ *end++ = '\0'; - return curdir_ok ? "." : ""; + return "."; } else if (end == NULL) { diff --git a/lib/pathmax.h b/lib/pathmax.h new file mode 100644 index 00000000..00fb0b51 --- /dev/null +++ b/lib/pathmax.h @@ -0,0 +1,53 @@ +/* Define PATH_MAX somehow. Requires sys/types.h. + Copyright (C) 1992 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +#ifndef _PATHMAX_H +#define _PATHMAX_H + +#ifdef HAVE_UNISTD_H +#include <unistd.h> +#endif + +/* Non-POSIX BSD systems might have gcc's limits.h, which doesn't define + PATH_MAX but might cause redefinition warnings when sys/param.h is + later included (as on MORE/BSD 4.3). */ +#if defined(_POSIX_VERSION) || (defined(HAVE_LIMITS_H) && !defined(__GNUC__)) +#include <limits.h> +#endif + +#ifndef _POSIX_PATH_MAX +#define _POSIX_PATH_MAX 255 +#endif + +#if !defined(PATH_MAX) && defined(_PC_PATH_MAX) +#define PATH_MAX (pathconf ("/", _PC_PATH_MAX) < 1 ? 1024 : pathconf ("/", _PC_PATH_MAX)) +#endif + +/* Don't include sys/param.h if it already has been. */ +#if defined(HAVE_SYS_PARAM_H) && !defined(PATH_MAX) && !defined(MAXPATHLEN) +#include <sys/param.h> +#endif + +#if !defined(PATH_MAX) && defined(MAXPATHLEN) +#define PATH_MAX MAXPATHLEN +#endif + +#ifndef PATH_MAX +#define PATH_MAX _POSIX_PATH_MAX +#endif + +#endif /* _PATHMAX_H */ diff --git a/lib/regex.c b/lib/regex.c new file mode 100644 index 00000000..33c7ea24 --- /dev/null +++ b/lib/regex.c @@ -0,0 +1,5244 @@ +/* Extended regular expression matching and search library, + version 0.12. + (Implements POSIX draft P10003.2/D11.2, except for + internationalization features.) + + Copyright (C) 1993, 1994 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +/* AIX requires this to be the first thing in the file. */ +#if defined (_AIX) && !defined (REGEX_MALLOC) + #pragma alloca +#endif + +#define _GNU_SOURCE + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +/* We need this for `regex.h', and perhaps for the Emacs include files. */ +#include <sys/types.h> + +/* The `emacs' switch turns on certain matching commands + that make sense only in Emacs. */ +#ifdef emacs + +#include "lisp.h" +#include "buffer.h" +#include "syntax.h" + +/* Emacs uses `NULL' as a predicate. */ +#undef NULL + +#else /* not emacs */ + +#ifdef STDC_HEADERS +#include <stdlib.h> +#else +char *malloc (); +char *realloc (); +#endif + + +/* We used to test for `BSTRING' here, but only GCC and Emacs define + `BSTRING', as far as I know, and neither of them use this code. */ +#ifndef INHIBIT_STRING_HEADER +#if HAVE_STRING_H || STDC_HEADERS +#include <string.h> +#ifndef bcmp +#define bcmp(s1, s2, n) memcmp ((s1), (s2), (n)) +#endif +#ifndef bcopy +#define bcopy(s, d, n) memcpy ((d), (s), (n)) +#endif +#ifndef bzero +#define bzero(s, n) memset ((s), 0, (n)) +#endif +#else +#include <strings.h> +#endif +#endif + +/* Define the syntax stuff for \<, \>, etc. */ + +/* This must be nonzero for the wordchar and notwordchar pattern + commands in re_match_2. */ +#ifndef Sword +#define Sword 1 +#endif + +#ifdef SYNTAX_TABLE + +extern char *re_syntax_table; + +#else /* not SYNTAX_TABLE */ + +/* How many characters in the character set. */ +#define CHAR_SET_SIZE 256 + +static char re_syntax_table[CHAR_SET_SIZE]; + +static void +init_syntax_once () +{ + register int c; + static int done = 0; + + if (done) + return; + + bzero (re_syntax_table, sizeof re_syntax_table); + + for (c = 'a'; c <= 'z'; c++) + re_syntax_table[c] = Sword; + + for (c = 'A'; c <= 'Z'; c++) + re_syntax_table[c] = Sword; + + for (c = '0'; c <= '9'; c++) + re_syntax_table[c] = Sword; + + re_syntax_table['_'] = Sword; + + done = 1; +} + +#endif /* not SYNTAX_TABLE */ + +#define SYNTAX(c) re_syntax_table[c] + +#endif /* not emacs */ + +/* Get the interface, including the syntax bits. */ +#include "regex.h" + +/* isalpha etc. are used for the character classes. */ +#include <ctype.h> + +/* Jim Meyering writes: + + "... Some ctype macros are valid only for character codes that + isascii says are ASCII (SGI's IRIX-4.0.5 is one such system --when + using /bin/cc or gcc but without giving an ansi option). So, all + ctype uses should be through macros like ISPRINT... If + STDC_HEADERS is defined, then autoconf has verified that the ctype + macros don't need to be guarded with references to isascii. ... + Defining isascii to 1 should let any compiler worth its salt + eliminate the && through constant folding." */ + +#if defined (STDC_HEADERS) || (!defined (isascii) && !defined (HAVE_ISASCII)) +#define ISASCII(c) 1 +#else +#define ISASCII(c) isascii(c) +#endif + +#ifdef isblank +#define ISBLANK(c) (ISASCII (c) && isblank (c)) +#else +#define ISBLANK(c) ((c) == ' ' || (c) == '\t') +#endif +#ifdef isgraph +#define ISGRAPH(c) (ISASCII (c) && isgraph (c)) +#else +#define ISGRAPH(c) (ISASCII (c) && isprint (c) && !isspace (c)) +#endif + +#define ISPRINT(c) (ISASCII (c) && isprint (c)) +#define ISDIGIT(c) (ISASCII (c) && isdigit (c)) +#define ISALNUM(c) (ISASCII (c) && isalnum (c)) +#define ISALPHA(c) (ISASCII (c) && isalpha (c)) +#define ISCNTRL(c) (ISASCII (c) && iscntrl (c)) +#define ISLOWER(c) (ISASCII (c) && islower (c)) +#define ISPUNCT(c) (ISASCII (c) && ispunct (c)) +#define ISSPACE(c) (ISASCII (c) && isspace (c)) +#define ISUPPER(c) (ISASCII (c) && isupper (c)) +#define ISXDIGIT(c) (ISASCII (c) && isxdigit (c)) + +#ifndef NULL +#define NULL 0 +#endif + +/* We remove any previous definition of `SIGN_EXTEND_CHAR', + since ours (we hope) works properly with all combinations of + machines, compilers, `char' and `unsigned char' argument types. + (Per Bothner suggested the basic approach.) */ +#undef SIGN_EXTEND_CHAR +#if __STDC__ +#define SIGN_EXTEND_CHAR(c) ((signed char) (c)) +#else /* not __STDC__ */ +/* As in Harbison and Steele. */ +#define SIGN_EXTEND_CHAR(c) ((((unsigned char) (c)) ^ 128) - 128) +#endif + +/* Should we use malloc or alloca? If REGEX_MALLOC is not defined, we + use `alloca' instead of `malloc'. This is because using malloc in + re_search* or re_match* could cause memory leaks when C-g is used in + Emacs; also, malloc is slower and causes storage fragmentation. On + the other hand, malloc is more portable, and easier to debug. + + Because we sometimes use alloca, some routines have to be macros, + not functions -- `alloca'-allocated space disappears at the end of the + function it is called in. */ + +#ifdef REGEX_MALLOC + +#define REGEX_ALLOCATE malloc +#define REGEX_REALLOCATE(source, osize, nsize) realloc (source, nsize) + +#else /* not REGEX_MALLOC */ + +/* Emacs already defines alloca, sometimes. */ +#ifndef alloca + +/* Make alloca work the best possible way. */ +#ifdef __GNUC__ +#define alloca __builtin_alloca +#else /* not __GNUC__ */ +#if HAVE_ALLOCA_H +#include <alloca.h> +#else /* not __GNUC__ or HAVE_ALLOCA_H */ +#ifndef _AIX /* Already did AIX, up at the top. */ +char *alloca (); +#endif /* not _AIX */ +#endif /* not HAVE_ALLOCA_H */ +#endif /* not __GNUC__ */ + +#endif /* not alloca */ + +#define REGEX_ALLOCATE alloca + +/* Assumes a `char *destination' variable. */ +#define REGEX_REALLOCATE(source, osize, nsize) \ + (destination = (char *) alloca (nsize), \ + bcopy (source, destination, osize), \ + destination) + +#endif /* not REGEX_MALLOC */ + + +/* True if `size1' is non-NULL and PTR is pointing anywhere inside + `string1' or just past its end. This works if PTR is NULL, which is + a good thing. */ +#define FIRST_STRING_P(ptr) \ + (size1 && string1 <= (ptr) && (ptr) <= string1 + size1) + +/* (Re)Allocate N items of type T using malloc, or fail. */ +#define TALLOC(n, t) ((t *) malloc ((n) * sizeof (t))) +#define RETALLOC(addr, n, t) ((addr) = (t *) realloc (addr, (n) * sizeof (t))) +#define RETALLOC_IF(addr, n, t) \ + if (addr) RETALLOC((addr), (n), t); else (addr) = TALLOC ((n), t) +#define REGEX_TALLOC(n, t) ((t *) REGEX_ALLOCATE ((n) * sizeof (t))) + +#define BYTEWIDTH 8 /* In bits. */ + +#define STREQ(s1, s2) ((strcmp (s1, s2) == 0)) + +#undef MAX +#undef MIN +#define MAX(a, b) ((a) > (b) ? (a) : (b)) +#define MIN(a, b) ((a) < (b) ? (a) : (b)) + +typedef char boolean; +#define false 0 +#define true 1 + +static int re_match_2_internal (); + +/* These are the command codes that appear in compiled regular + expressions. Some opcodes are followed by argument bytes. A + command code can specify any interpretation whatsoever for its + arguments. Zero bytes may appear in the compiled regular expression. */ + +typedef enum +{ + no_op = 0, + + /* Followed by one byte giving n, then by n literal bytes. */ + exactn, + + /* Matches any (more or less) character. */ + anychar, + + /* Matches any one char belonging to specified set. First + following byte is number of bitmap bytes. Then come bytes + for a bitmap saying which chars are in. Bits in each byte + are ordered low-bit-first. A character is in the set if its + bit is 1. A character too large to have a bit in the map is + automatically not in the set. */ + charset, + + /* Same parameters as charset, but match any character that is + not one of those specified. */ + charset_not, + + /* Start remembering the text that is matched, for storing in a + register. Followed by one byte with the register number, in + the range 0 to one less than the pattern buffer's re_nsub + field. Then followed by one byte with the number of groups + inner to this one. (This last has to be part of the + start_memory only because we need it in the on_failure_jump + of re_match_2.) */ + start_memory, + + /* Stop remembering the text that is matched and store it in a + memory register. Followed by one byte with the register + number, in the range 0 to one less than `re_nsub' in the + pattern buffer, and one byte with the number of inner groups, + just like `start_memory'. (We need the number of inner + groups here because we don't have any easy way of finding the + corresponding start_memory when we're at a stop_memory.) */ + stop_memory, + + /* Match a duplicate of something remembered. Followed by one + byte containing the register number. */ + duplicate, + + /* Fail unless at beginning of line. */ + begline, + + /* Fail unless at end of line. */ + endline, + + /* Succeeds if at beginning of buffer (if emacs) or at beginning + of string to be matched (if not). */ + begbuf, + + /* Analogously, for end of buffer/string. */ + endbuf, + + /* Followed by two byte relative address to which to jump. */ + jump, + + /* Same as jump, but marks the end of an alternative. */ + jump_past_alt, + + /* Followed by two-byte relative address of place to resume at + in case of failure. */ + on_failure_jump, + + /* Like on_failure_jump, but pushes a placeholder instead of the + current string position when executed. */ + on_failure_keep_string_jump, + + /* Throw away latest failure point and then jump to following + two-byte relative address. */ + pop_failure_jump, + + /* Change to pop_failure_jump if know won't have to backtrack to + match; otherwise change to jump. This is used to jump + back to the beginning of a repeat. If what follows this jump + clearly won't match what the repeat does, such that we can be + sure that there is no use backtracking out of repetitions + already matched, then we change it to a pop_failure_jump. + Followed by two-byte address. */ + maybe_pop_jump, + + /* Jump to following two-byte address, and push a dummy failure + point. This failure point will be thrown away if an attempt + is made to use it for a failure. A `+' construct makes this + before the first repeat. Also used as an intermediary kind + of jump when compiling an alternative. */ + dummy_failure_jump, + + /* Push a dummy failure point and continue. Used at the end of + alternatives. */ + push_dummy_failure, + + /* Followed by two-byte relative address and two-byte number n. + After matching N times, jump to the address upon failure. */ + succeed_n, + + /* Followed by two-byte relative address, and two-byte number n. + Jump to the address N times, then fail. */ + jump_n, + + /* Set the following two-byte relative address to the + subsequent two-byte number. The address *includes* the two + bytes of number. */ + set_number_at, + + wordchar, /* Matches any word-constituent character. */ + notwordchar, /* Matches any char that is not a word-constituent. */ + + wordbeg, /* Succeeds if at word beginning. */ + wordend, /* Succeeds if at word end. */ + + wordbound, /* Succeeds if at a word boundary. */ + notwordbound /* Succeeds if not at a word boundary. */ + +#ifdef emacs + ,before_dot, /* Succeeds if before point. */ + at_dot, /* Succeeds if at point. */ + after_dot, /* Succeeds if after point. */ + + /* Matches any character whose syntax is specified. Followed by + a byte which contains a syntax code, e.g., Sword. */ + syntaxspec, + + /* Matches any character whose syntax is not that specified. */ + notsyntaxspec +#endif /* emacs */ +} re_opcode_t; + +/* Common operations on the compiled pattern. */ + +/* Store NUMBER in two contiguous bytes starting at DESTINATION. */ + +#define STORE_NUMBER(destination, number) \ + do { \ + (destination)[0] = (number) & 0377; \ + (destination)[1] = (number) >> 8; \ + } while (0) + +/* Same as STORE_NUMBER, except increment DESTINATION to + the byte after where the number is stored. Therefore, DESTINATION + must be an lvalue. */ + +#define STORE_NUMBER_AND_INCR(destination, number) \ + do { \ + STORE_NUMBER (destination, number); \ + (destination) += 2; \ + } while (0) + +/* Put into DESTINATION a number stored in two contiguous bytes starting + at SOURCE. */ + +#define EXTRACT_NUMBER(destination, source) \ + do { \ + (destination) = *(source) & 0377; \ + (destination) += SIGN_EXTEND_CHAR (*((source) + 1)) << 8; \ + } while (0) + +#ifdef DEBUG +static void +extract_number (dest, source) + int *dest; + unsigned char *source; +{ + int temp = SIGN_EXTEND_CHAR (*(source + 1)); + *dest = *source & 0377; + *dest += temp << 8; +} + +#ifndef EXTRACT_MACROS /* To debug the macros. */ +#undef EXTRACT_NUMBER +#define EXTRACT_NUMBER(dest, src) extract_number (&dest, src) +#endif /* not EXTRACT_MACROS */ + +#endif /* DEBUG */ + +/* Same as EXTRACT_NUMBER, except increment SOURCE to after the number. + SOURCE must be an lvalue. */ + +#define EXTRACT_NUMBER_AND_INCR(destination, source) \ + do { \ + EXTRACT_NUMBER (destination, source); \ + (source) += 2; \ + } while (0) + +#ifdef DEBUG +static void +extract_number_and_incr (destination, source) + int *destination; + unsigned char **source; +{ + extract_number (destination, *source); + *source += 2; +} + +#ifndef EXTRACT_MACROS +#undef EXTRACT_NUMBER_AND_INCR +#define EXTRACT_NUMBER_AND_INCR(dest, src) \ + extract_number_and_incr (&dest, &src) +#endif /* not EXTRACT_MACROS */ + +#endif /* DEBUG */ + +/* If DEBUG is defined, Regex prints many voluminous messages about what + it is doing (if the variable `debug' is nonzero). If linked with the + main program in `iregex.c', you can enter patterns and strings + interactively. And if linked with the main program in `main.c' and + the other test files, you can run the already-written tests. */ + +#ifdef DEBUG + +/* We use standard I/O for debugging. */ +#include <stdio.h> + +/* It is useful to test things that ``must'' be true when debugging. */ +#include <assert.h> + +static int debug = 0; + +#define DEBUG_STATEMENT(e) e +#define DEBUG_PRINT1(x) if (debug) printf (x) +#define DEBUG_PRINT2(x1, x2) if (debug) printf (x1, x2) +#define DEBUG_PRINT3(x1, x2, x3) if (debug) printf (x1, x2, x3) +#define DEBUG_PRINT4(x1, x2, x3, x4) if (debug) printf (x1, x2, x3, x4) +#define DEBUG_PRINT_COMPILED_PATTERN(p, s, e) \ + if (debug) print_partial_compiled_pattern (s, e) +#define DEBUG_PRINT_DOUBLE_STRING(w, s1, sz1, s2, sz2) \ + if (debug) print_double_string (w, s1, sz1, s2, sz2) + + +extern void printchar (); + +/* Print the fastmap in human-readable form. */ + +void +print_fastmap (fastmap) + char *fastmap; +{ + unsigned was_a_range = 0; + unsigned i = 0; + + while (i < (1 << BYTEWIDTH)) + { + if (fastmap[i++]) + { + was_a_range = 0; + printchar (i - 1); + while (i < (1 << BYTEWIDTH) && fastmap[i]) + { + was_a_range = 1; + i++; + } + if (was_a_range) + { + printf ("-"); + printchar (i - 1); + } + } + } + putchar ('\n'); +} + + +/* Print a compiled pattern string in human-readable form, starting at + the START pointer into it and ending just before the pointer END. */ + +void +print_partial_compiled_pattern (start, end) + unsigned char *start; + unsigned char *end; +{ + int mcnt, mcnt2; + unsigned char *p = start; + unsigned char *pend = end; + + if (start == NULL) + { + printf ("(null)\n"); + return; + } + + /* Loop over pattern commands. */ + while (p < pend) + { + printf ("%d:\t", p - start); + + switch ((re_opcode_t) *p++) + { + case no_op: + printf ("/no_op"); + break; + + case exactn: + mcnt = *p++; + printf ("/exactn/%d", mcnt); + do + { + putchar ('/'); + printchar (*p++); + } + while (--mcnt); + break; + + case start_memory: + mcnt = *p++; + printf ("/start_memory/%d/%d", mcnt, *p++); + break; + + case stop_memory: + mcnt = *p++; + printf ("/stop_memory/%d/%d", mcnt, *p++); + break; + + case duplicate: + printf ("/duplicate/%d", *p++); + break; + + case anychar: + printf ("/anychar"); + break; + + case charset: + case charset_not: + { + register int c, last = -100; + register int in_range = 0; + + printf ("/charset [%s", + (re_opcode_t) *(p - 1) == charset_not ? "^" : ""); + + assert (p + *p < pend); + + for (c = 0; c < 256; c++) + if (c / 8 < *p + && (p[1 + (c/8)] & (1 << (c % 8)))) + { + /* Are we starting a range? */ + if (last + 1 == c && ! in_range) + { + putchar ('-'); + in_range = 1; + } + /* Have we broken a range? */ + else if (last + 1 != c && in_range) + { + printchar (last); + in_range = 0; + } + + if (! in_range) + printchar (c); + + last = c; + } + + if (in_range) + printchar (last); + + putchar (']'); + + p += 1 + *p; + } + break; + + case begline: + printf ("/begline"); + break; + + case endline: + printf ("/endline"); + break; + + case on_failure_jump: + extract_number_and_incr (&mcnt, &p); + printf ("/on_failure_jump to %d", p + mcnt - start); + break; + + case on_failure_keep_string_jump: + extract_number_and_incr (&mcnt, &p); + printf ("/on_failure_keep_string_jump to %d", p + mcnt - start); + break; + + case dummy_failure_jump: + extract_number_and_incr (&mcnt, &p); + printf ("/dummy_failure_jump to %d", p + mcnt - start); + break; + + case push_dummy_failure: + printf ("/push_dummy_failure"); + break; + + case maybe_pop_jump: + extract_number_and_incr (&mcnt, &p); + printf ("/maybe_pop_jump to %d", p + mcnt - start); + break; + + case pop_failure_jump: + extract_number_and_incr (&mcnt, &p); + printf ("/pop_failure_jump to %d", p + mcnt - start); + break; + + case jump_past_alt: + extract_number_and_incr (&mcnt, &p); + printf ("/jump_past_alt to %d", p + mcnt - start); + break; + + case jump: + extract_number_and_incr (&mcnt, &p); + printf ("/jump to %d", p + mcnt - start); + break; + + case succeed_n: + extract_number_and_incr (&mcnt, &p); + extract_number_and_incr (&mcnt2, &p); + printf ("/succeed_n to %d, %d times", p + mcnt - start, mcnt2); + break; + + case jump_n: + extract_number_and_incr (&mcnt, &p); + extract_number_and_incr (&mcnt2, &p); + printf ("/jump_n to %d, %d times", p + mcnt - start, mcnt2); + break; + + case set_number_at: + extract_number_and_incr (&mcnt, &p); + extract_number_and_incr (&mcnt2, &p); + printf ("/set_number_at location %d to %d", p + mcnt - start, mcnt2); + break; + + case wordbound: + printf ("/wordbound"); + break; + + case notwordbound: + printf ("/notwordbound"); + break; + + case wordbeg: + printf ("/wordbeg"); + break; + + case wordend: + printf ("/wordend"); + +#ifdef emacs + case before_dot: + printf ("/before_dot"); + break; + + case at_dot: + printf ("/at_dot"); + break; + + case after_dot: + printf ("/after_dot"); + break; + + case syntaxspec: + printf ("/syntaxspec"); + mcnt = *p++; + printf ("/%d", mcnt); + break; + + case notsyntaxspec: + printf ("/notsyntaxspec"); + mcnt = *p++; + printf ("/%d", mcnt); + break; +#endif /* emacs */ + + case wordchar: + printf ("/wordchar"); + break; + + case notwordchar: + printf ("/notwordchar"); + break; + + case begbuf: + printf ("/begbuf"); + break; + + case endbuf: + printf ("/endbuf"); + break; + + default: + printf ("?%d", *(p-1)); + } + + putchar ('\n'); + } + + printf ("%d:\tend of pattern.\n", p - start); +} + + +void +print_compiled_pattern (bufp) + struct re_pattern_buffer *bufp; +{ + unsigned char *buffer = bufp->buffer; + + print_partial_compiled_pattern (buffer, buffer + bufp->used); + printf ("%d bytes used/%d bytes allocated.\n", bufp->used, bufp->allocated); + + if (bufp->fastmap_accurate && bufp->fastmap) + { + printf ("fastmap: "); + print_fastmap (bufp->fastmap); + } + + printf ("re_nsub: %d\t", bufp->re_nsub); + printf ("regs_alloc: %d\t", bufp->regs_allocated); + printf ("can_be_null: %d\t", bufp->can_be_null); + printf ("newline_anchor: %d\n", bufp->newline_anchor); + printf ("no_sub: %d\t", bufp->no_sub); + printf ("not_bol: %d\t", bufp->not_bol); + printf ("not_eol: %d\t", bufp->not_eol); + printf ("syntax: %d\n", bufp->syntax); + /* Perhaps we should print the translate table? */ +} + + +void +print_double_string (where, string1, size1, string2, size2) + const char *where; + const char *string1; + const char *string2; + int size1; + int size2; +{ + unsigned this_char; + + if (where == NULL) + printf ("(null)"); + else + { + if (FIRST_STRING_P (where)) + { + for (this_char = where - string1; this_char < size1; this_char++) + printchar (string1[this_char]); + + where = string2; + } + + for (this_char = where - string2; this_char < size2; this_char++) + printchar (string2[this_char]); + } +} + +#else /* not DEBUG */ + +#undef assert +#define assert(e) + +#define DEBUG_STATEMENT(e) +#define DEBUG_PRINT1(x) +#define DEBUG_PRINT2(x1, x2) +#define DEBUG_PRINT3(x1, x2, x3) +#define DEBUG_PRINT4(x1, x2, x3, x4) +#define DEBUG_PRINT_COMPILED_PATTERN(p, s, e) +#define DEBUG_PRINT_DOUBLE_STRING(w, s1, sz1, s2, sz2) + +#endif /* not DEBUG */ + +/* 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. */ +reg_syntax_t re_syntax_options = RE_SYNTAX_EMACS; + + +/* 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; +} + +/* 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. */ + +static const char *re_error_msg[] = + { NULL, /* REG_NOERROR */ + "No match", /* REG_NOMATCH */ + "Invalid regular expression", /* REG_BADPAT */ + "Invalid collation character", /* REG_ECOLLATE */ + "Invalid character class name", /* REG_ECTYPE */ + "Trailing backslash", /* REG_EESCAPE */ + "Invalid back reference", /* REG_ESUBREG */ + "Unmatched [ or [^", /* REG_EBRACK */ + "Unmatched ( or \\(", /* REG_EPAREN */ + "Unmatched \\{", /* REG_EBRACE */ + "Invalid content of \\{\\}", /* REG_BADBR */ + "Invalid range end", /* REG_ERANGE */ + "Memory exhausted", /* REG_ESPACE */ + "Invalid preceding regular expression", /* REG_BADRPT */ + "Premature end of regular expression", /* REG_EEND */ + "Regular expression too big", /* REG_ESIZE */ + "Unmatched ) or \\)", /* REG_ERPAREN */ + }; + +/* Avoiding alloca during matching, to placate r_alloc. */ + +/* Define MATCH_MAY_ALLOCATE unless we need to make sure that the + searching and matching functions should not call alloca. On some + systems, alloca is implemented in terms of malloc, and if we're + using the relocating allocator routines, then malloc could cause a + relocation, which might (if the strings being searched are in the + ralloc heap) shift the data out from underneath the regexp + routines. + + Here's another reason to avoid allocation: Emacs + processes input from X in a signal handler; processing X input may + call malloc; if input arrives while a matching routine is calling + malloc, then we're scrod. But Emacs can't just block input while + calling matching routines; then we don't notice interrupts when + they come in. So, Emacs blocks input around all regexp calls + except the matching calls, which it leaves unprotected, in the + faith that they will not malloc. */ + +/* Normally, this is fine. */ +#define MATCH_MAY_ALLOCATE + +/* The match routines may not allocate if (1) they would do it with malloc + and (2) it's not safe for them to use malloc. */ +#if (defined (C_ALLOCA) || defined (REGEX_MALLOC)) && (defined (emacs) || defined (REL_ALLOC)) +#undef MATCH_MAY_ALLOCATE +#endif + + +/* Failure stack declarations and macros; both re_compile_fastmap and + re_match_2 use a failure stack. These have to be macros because of + REGEX_ALLOCATE. */ + + +/* Number of failure points for which to initially allocate space + when matching. If this number is exceeded, we allocate more + space, so it is not a hard limit. */ +#ifndef INIT_FAILURE_ALLOC +#define INIT_FAILURE_ALLOC 5 +#endif + +/* Roughly the maximum number of failure points on the stack. Would be + exactly that if always used MAX_FAILURE_SPACE each time we failed. + This is a variable only so users of regex can assign to it; we never + change it ourselves. */ +int re_max_failures = 2000; + +typedef unsigned char *fail_stack_elt_t; + +typedef struct +{ + fail_stack_elt_t *stack; + unsigned size; + unsigned avail; /* Offset of next open position. */ +} fail_stack_type; + +#define FAIL_STACK_EMPTY() (fail_stack.avail == 0) +#define FAIL_STACK_PTR_EMPTY() (fail_stack_ptr->avail == 0) +#define FAIL_STACK_FULL() (fail_stack.avail == fail_stack.size) +#define FAIL_STACK_TOP() (fail_stack.stack[fail_stack.avail]) + + +/* Initialize `fail_stack'. Do `return -2' if the alloc fails. */ + +#ifdef MATCH_MAY_ALLOCATE +#define INIT_FAIL_STACK() \ + do { \ + fail_stack.stack = (fail_stack_elt_t *) \ + REGEX_ALLOCATE (INIT_FAILURE_ALLOC * sizeof (fail_stack_elt_t)); \ + \ + if (fail_stack.stack == NULL) \ + return -2; \ + \ + fail_stack.size = INIT_FAILURE_ALLOC; \ + fail_stack.avail = 0; \ + } while (0) +#else +#define INIT_FAIL_STACK() \ + do { \ + fail_stack.avail = 0; \ + } while (0) +#endif + + +/* Double the size of FAIL_STACK, up to approximately `re_max_failures' items. + + Return 1 if succeeds, and 0 if either ran out of memory + allocating space for it or it was already too large. + + REGEX_REALLOCATE requires `destination' be declared. */ + +#define DOUBLE_FAIL_STACK(fail_stack) \ + ((fail_stack).size > re_max_failures * MAX_FAILURE_ITEMS \ + ? 0 \ + : ((fail_stack).stack = (fail_stack_elt_t *) \ + REGEX_REALLOCATE ((fail_stack).stack, \ + (fail_stack).size * sizeof (fail_stack_elt_t), \ + ((fail_stack).size << 1) * sizeof (fail_stack_elt_t)), \ + \ + (fail_stack).stack == NULL \ + ? 0 \ + : ((fail_stack).size <<= 1, \ + 1))) + + +/* Push PATTERN_OP on FAIL_STACK. + + Return 1 if was able to do so and 0 if ran out of memory allocating + space to do so. */ +#define PUSH_PATTERN_OP(pattern_op, fail_stack) \ + ((FAIL_STACK_FULL () \ + && !DOUBLE_FAIL_STACK (fail_stack)) \ + ? 0 \ + : ((fail_stack).stack[(fail_stack).avail++] = pattern_op, \ + 1)) + +/* This pushes an item onto the failure stack. Must be a four-byte + value. Assumes the variable `fail_stack'. Probably should only + be called from within `PUSH_FAILURE_POINT'. */ +#define PUSH_FAILURE_ITEM(item) \ + fail_stack.stack[fail_stack.avail++] = (fail_stack_elt_t) item + +/* The complement operation. Assumes `fail_stack' is nonempty. */ +#define POP_FAILURE_ITEM() fail_stack.stack[--fail_stack.avail] + +/* Used to omit pushing failure point id's when we're not debugging. */ +#ifdef DEBUG +#define DEBUG_PUSH PUSH_FAILURE_ITEM +#define DEBUG_POP(item_addr) *(item_addr) = POP_FAILURE_ITEM () +#else +#define DEBUG_PUSH(item) +#define DEBUG_POP(item_addr) +#endif + + +/* Push the information about the state we will need + if we ever fail back to it. + + Requires variables fail_stack, regstart, regend, reg_info, and + num_regs be declared. DOUBLE_FAIL_STACK requires `destination' be + declared. + + Does `return FAILURE_CODE' if runs out of memory. */ + +#define PUSH_FAILURE_POINT(pattern_place, string_place, failure_code) \ + do { \ + char *destination; \ + /* Must be int, so when we don't save any registers, the arithmetic \ + of 0 + -1 isn't done as unsigned. */ \ + int this_reg; \ + \ + DEBUG_STATEMENT (failure_id++); \ + DEBUG_STATEMENT (nfailure_points_pushed++); \ + DEBUG_PRINT2 ("\nPUSH_FAILURE_POINT #%u:\n", failure_id); \ + DEBUG_PRINT2 (" Before push, next avail: %d\n", (fail_stack).avail);\ + DEBUG_PRINT2 (" size: %d\n", (fail_stack).size);\ + \ + DEBUG_PRINT2 (" slots needed: %d\n", NUM_FAILURE_ITEMS); \ + DEBUG_PRINT2 (" available: %d\n", REMAINING_AVAIL_SLOTS); \ + \ + /* Ensure we have enough space allocated for what we will push. */ \ + while (REMAINING_AVAIL_SLOTS < NUM_FAILURE_ITEMS) \ + { \ + if (!DOUBLE_FAIL_STACK (fail_stack)) \ + return failure_code; \ + \ + DEBUG_PRINT2 ("\n Doubled stack; size now: %d\n", \ + (fail_stack).size); \ + DEBUG_PRINT2 (" slots available: %d\n", REMAINING_AVAIL_SLOTS);\ + } \ + \ + /* Push the info, starting with the registers. */ \ + DEBUG_PRINT1 ("\n"); \ + \ + for (this_reg = lowest_active_reg; this_reg <= highest_active_reg; \ + this_reg++) \ + { \ + DEBUG_PRINT2 (" Pushing reg: %d\n", this_reg); \ + DEBUG_STATEMENT (num_regs_pushed++); \ + \ + DEBUG_PRINT2 (" start: 0x%x\n", regstart[this_reg]); \ + PUSH_FAILURE_ITEM (regstart[this_reg]); \ + \ + DEBUG_PRINT2 (" end: 0x%x\n", regend[this_reg]); \ + PUSH_FAILURE_ITEM (regend[this_reg]); \ + \ + DEBUG_PRINT2 (" info: 0x%x\n ", reg_info[this_reg]); \ + DEBUG_PRINT2 (" match_null=%d", \ + REG_MATCH_NULL_STRING_P (reg_info[this_reg])); \ + DEBUG_PRINT2 (" active=%d", IS_ACTIVE (reg_info[this_reg])); \ + DEBUG_PRINT2 (" matched_something=%d", \ + MATCHED_SOMETHING (reg_info[this_reg])); \ + DEBUG_PRINT2 (" ever_matched=%d", \ + EVER_MATCHED_SOMETHING (reg_info[this_reg])); \ + DEBUG_PRINT1 ("\n"); \ + PUSH_FAILURE_ITEM (reg_info[this_reg].word); \ + } \ + \ + DEBUG_PRINT2 (" Pushing low active reg: %d\n", lowest_active_reg);\ + PUSH_FAILURE_ITEM (lowest_active_reg); \ + \ + DEBUG_PRINT2 (" Pushing high active reg: %d\n", highest_active_reg);\ + PUSH_FAILURE_ITEM (highest_active_reg); \ + \ + DEBUG_PRINT2 (" Pushing pattern 0x%x: ", pattern_place); \ + DEBUG_PRINT_COMPILED_PATTERN (bufp, pattern_place, pend); \ + PUSH_FAILURE_ITEM (pattern_place); \ + \ + DEBUG_PRINT2 (" Pushing string 0x%x: `", string_place); \ + DEBUG_PRINT_DOUBLE_STRING (string_place, string1, size1, string2, \ + size2); \ + DEBUG_PRINT1 ("'\n"); \ + PUSH_FAILURE_ITEM (string_place); \ + \ + DEBUG_PRINT2 (" Pushing failure id: %u\n", failure_id); \ + DEBUG_PUSH (failure_id); \ + } while (0) + +/* This is the number of items that are pushed and popped on the stack + for each register. */ +#define NUM_REG_ITEMS 3 + +/* Individual items aside from the registers. */ +#ifdef DEBUG +#define NUM_NONREG_ITEMS 5 /* Includes failure point id. */ +#else +#define NUM_NONREG_ITEMS 4 +#endif + +/* We push at most this many items on the stack. */ +#define MAX_FAILURE_ITEMS ((num_regs - 1) * NUM_REG_ITEMS + NUM_NONREG_ITEMS) + +/* We actually push this many items. */ +#define NUM_FAILURE_ITEMS \ + ((highest_active_reg - lowest_active_reg + 1) * NUM_REG_ITEMS \ + + NUM_NONREG_ITEMS) + +/* How many items can still be added to the stack without overflowing it. */ +#define REMAINING_AVAIL_SLOTS ((fail_stack).size - (fail_stack).avail) + + +/* Pops what PUSH_FAIL_STACK pushes. + + We restore into the parameters, all of which should be lvalues: + STR -- the saved data position. + PAT -- the saved pattern position. + LOW_REG, HIGH_REG -- the highest and lowest active registers. + REGSTART, REGEND -- arrays of string positions. + REG_INFO -- array of information about each subexpression. + + Also assumes the variables `fail_stack' and (if debugging), `bufp', + `pend', `string1', `size1', `string2', and `size2'. */ + +#define POP_FAILURE_POINT(str, pat, low_reg, high_reg, regstart, regend, reg_info)\ +{ \ + DEBUG_STATEMENT (fail_stack_elt_t failure_id;) \ + int this_reg; \ + const unsigned char *string_temp; \ + \ + assert (!FAIL_STACK_EMPTY ()); \ + \ + /* Remove failure points and point to how many regs pushed. */ \ + DEBUG_PRINT1 ("POP_FAILURE_POINT:\n"); \ + DEBUG_PRINT2 (" Before pop, next avail: %d\n", fail_stack.avail); \ + DEBUG_PRINT2 (" size: %d\n", fail_stack.size); \ + \ + assert (fail_stack.avail >= NUM_NONREG_ITEMS); \ + \ + DEBUG_POP (&failure_id); \ + DEBUG_PRINT2 (" Popping failure id: %u\n", failure_id); \ + \ + /* If the saved string location is NULL, it came from an \ + on_failure_keep_string_jump opcode, and we want to throw away the \ + saved NULL, thus retaining our current position in the string. */ \ + string_temp = POP_FAILURE_ITEM (); \ + if (string_temp != NULL) \ + str = (const char *) string_temp; \ + \ + DEBUG_PRINT2 (" Popping string 0x%x: `", str); \ + DEBUG_PRINT_DOUBLE_STRING (str, string1, size1, string2, size2); \ + DEBUG_PRINT1 ("'\n"); \ + \ + pat = (unsigned char *) POP_FAILURE_ITEM (); \ + DEBUG_PRINT2 (" Popping pattern 0x%x: ", pat); \ + DEBUG_PRINT_COMPILED_PATTERN (bufp, pat, pend); \ + \ + /* Restore register info. */ \ + high_reg = (unsigned) POP_FAILURE_ITEM (); \ + DEBUG_PRINT2 (" Popping high active reg: %d\n", high_reg); \ + \ + low_reg = (unsigned) POP_FAILURE_ITEM (); \ + DEBUG_PRINT2 (" Popping low active reg: %d\n", low_reg); \ + \ + for (this_reg = high_reg; this_reg >= low_reg; this_reg--) \ + { \ + DEBUG_PRINT2 (" Popping reg: %d\n", this_reg); \ + \ + reg_info[this_reg].word = POP_FAILURE_ITEM (); \ + DEBUG_PRINT2 (" info: 0x%x\n", reg_info[this_reg]); \ + \ + regend[this_reg] = (const char *) POP_FAILURE_ITEM (); \ + DEBUG_PRINT2 (" end: 0x%x\n", regend[this_reg]); \ + \ + regstart[this_reg] = (const char *) POP_FAILURE_ITEM (); \ + DEBUG_PRINT2 (" start: 0x%x\n", regstart[this_reg]); \ + } \ + \ + DEBUG_STATEMENT (nfailure_points_popped++); \ +} /* POP_FAILURE_POINT */ + + + +/* Structure for per-register (a.k.a. per-group) information. + This must not be longer than one word, because we push this value + onto the failure stack. Other register information, such as the + starting and ending positions (which are addresses), and the list of + inner groups (which is a bits list) are maintained in separate + variables. + + We are making a (strictly speaking) nonportable assumption here: that + the compiler will pack our bit fields into something that fits into + the type of `word', i.e., is something that fits into one item on the + failure stack. */ +typedef union +{ + fail_stack_elt_t word; + struct + { + /* This field is one if this group can match the empty string, + zero if not. If not yet determined, `MATCH_NULL_UNSET_VALUE'. */ +#define MATCH_NULL_UNSET_VALUE 3 + unsigned match_null_string_p : 2; + unsigned is_active : 1; + unsigned matched_something : 1; + unsigned ever_matched_something : 1; + } bits; +} register_info_type; + +#define REG_MATCH_NULL_STRING_P(R) ((R).bits.match_null_string_p) +#define IS_ACTIVE(R) ((R).bits.is_active) +#define MATCHED_SOMETHING(R) ((R).bits.matched_something) +#define EVER_MATCHED_SOMETHING(R) ((R).bits.ever_matched_something) + + +/* Call this when have matched a real character; it sets `matched' flags + for the subexpressions which we are currently inside. Also records + that those subexprs have matched. */ +#define SET_REGS_MATCHED() \ + do \ + { \ + unsigned r; \ + for (r = lowest_active_reg; r <= highest_active_reg; r++) \ + { \ + MATCHED_SOMETHING (reg_info[r]) \ + = EVER_MATCHED_SOMETHING (reg_info[r]) \ + = 1; \ + } \ + } \ + while (0) + + +/* Registers are set to a sentinel when they haven't yet matched. */ +#define REG_UNSET_VALUE ((char *) -1) +#define REG_UNSET(e) ((e) == REG_UNSET_VALUE) + + + +/* How do we implement a missing MATCH_MAY_ALLOCATE? + We make the fail stack a global thing, and then grow it to + re_max_failures when we compile. */ +#ifndef MATCH_MAY_ALLOCATE +static fail_stack_type fail_stack; + +static const char ** regstart, ** regend; +static const char ** old_regstart, ** old_regend; +static const char **best_regstart, **best_regend; +static register_info_type *reg_info; +static const char **reg_dummy; +static register_info_type *reg_info_dummy; +#endif + + +/* Subroutine declarations and macros for regex_compile. */ + +static void store_op1 (), store_op2 (); +static void insert_op1 (), insert_op2 (); +static boolean at_begline_loc_p (), at_endline_loc_p (); +static boolean group_in_compile_stack (); +static reg_errcode_t compile_range (); + +/* Fetch the next character in the uncompiled pattern---translating it + if necessary. Also cast from a signed character in the constant + string passed to us by the user to an unsigned char that we can use + as an array index (in, e.g., `translate'). */ +#define PATFETCH(c) \ + do {if (p == pend) return REG_EEND; \ + c = (unsigned char) *p++; \ + if (translate) c = translate[c]; \ + } while (0) + +/* Fetch the next character in the uncompiled pattern, with no + translation. */ +#define PATFETCH_RAW(c) \ + do {if (p == pend) return REG_EEND; \ + c = (unsigned char) *p++; \ + } while (0) + +/* Go backwards one character in the pattern. */ +#define PATUNFETCH p-- + + +/* If `translate' is non-null, return translate[D], else just D. We + cast the subscript to translate because some data is declared as + `char *', to avoid warnings when a string constant is passed. But + when we use a character as a subscript we must make it unsigned. */ +#define TRANSLATE(d) (translate ? translate[(unsigned char) (d)] : (d)) + + +/* Macros for outputting the compiled pattern into `buffer'. */ + +/* If the buffer isn't allocated when it comes in, use this. */ +#define INIT_BUF_SIZE 32 + +/* Make sure we have at least N more bytes of space in buffer. */ +#define GET_BUFFER_SPACE(n) \ + while (b - bufp->buffer + (n) > bufp->allocated) \ + EXTEND_BUFFER () + +/* Make sure we have one more byte of buffer space and then add C to it. */ +#define BUF_PUSH(c) \ + do { \ + GET_BUFFER_SPACE (1); \ + *b++ = (unsigned char) (c); \ + } while (0) + + +/* Ensure we have two more bytes of buffer space and then append C1 and C2. */ +#define BUF_PUSH_2(c1, c2) \ + do { \ + GET_BUFFER_SPACE (2); \ + *b++ = (unsigned char) (c1); \ + *b++ = (unsigned char) (c2); \ + } while (0) + + +/* As with BUF_PUSH_2, except for three bytes. */ +#define BUF_PUSH_3(c1, c2, c3) \ + do { \ + GET_BUFFER_SPACE (3); \ + *b++ = (unsigned char) (c1); \ + *b++ = (unsigned char) (c2); \ + *b++ = (unsigned char) (c3); \ + } while (0) + + +/* Store a jump with opcode OP at LOC to location TO. We store a + relative address offset by the three bytes the jump itself occupies. */ +#define STORE_JUMP(op, loc, to) \ + store_op1 (op, loc, (to) - (loc) - 3) + +/* Likewise, for a two-argument jump. */ +#define STORE_JUMP2(op, loc, to, arg) \ + store_op2 (op, loc, (to) - (loc) - 3, arg) + +/* Like `STORE_JUMP', but for inserting. Assume `b' is the buffer end. */ +#define INSERT_JUMP(op, loc, to) \ + insert_op1 (op, loc, (to) - (loc) - 3, b) + +/* Like `STORE_JUMP2', but for inserting. Assume `b' is the buffer end. */ +#define INSERT_JUMP2(op, loc, to, arg) \ + insert_op2 (op, loc, (to) - (loc) - 3, arg, b) + + +/* This is not an arbitrary limit: the arguments which represent offsets + into the pattern are two bytes long. So if 2^16 bytes turns out to + be too small, many things would have to change. */ +#define MAX_BUF_SIZE (1L << 16) + + +/* Extend the buffer by twice its current size via realloc and + reset the pointers that pointed into the old block to point to the + correct places in the new one. If extending the buffer results in it + being larger than MAX_BUF_SIZE, then flag memory exhausted. */ +#define EXTEND_BUFFER() \ + do { \ + unsigned char *old_buffer = bufp->buffer; \ + if (bufp->allocated == MAX_BUF_SIZE) \ + return REG_ESIZE; \ + bufp->allocated <<= 1; \ + if (bufp->allocated > MAX_BUF_SIZE) \ + bufp->allocated = MAX_BUF_SIZE; \ + bufp->buffer = (unsigned char *) realloc (bufp->buffer, bufp->allocated);\ + if (bufp->buffer == NULL) \ + return REG_ESPACE; \ + /* If the buffer moved, move all the pointers into it. */ \ + if (old_buffer != bufp->buffer) \ + { \ + b = (b - old_buffer) + bufp->buffer; \ + begalt = (begalt - old_buffer) + bufp->buffer; \ + if (fixup_alt_jump) \ + fixup_alt_jump = (fixup_alt_jump - old_buffer) + bufp->buffer;\ + if (laststart) \ + laststart = (laststart - old_buffer) + bufp->buffer; \ + if (pending_exact) \ + pending_exact = (pending_exact - old_buffer) + bufp->buffer; \ + } \ + } while (0) + + +/* Since we have one byte reserved for the register number argument to + {start,stop}_memory, the maximum number of groups we can report + things about is what fits in that byte. */ +#define MAX_REGNUM 255 + +/* But patterns can have more than `MAX_REGNUM' registers. We just + ignore the excess. */ +typedef unsigned regnum_t; + + +/* Macros for the compile stack. */ + +/* Since offsets can go either forwards or backwards, this type needs to + be able to hold values from -(MAX_BUF_SIZE - 1) to MAX_BUF_SIZE - 1. */ +typedef int pattern_offset_t; + +typedef struct +{ + pattern_offset_t begalt_offset; + pattern_offset_t fixup_alt_jump; + pattern_offset_t inner_group_offset; + pattern_offset_t laststart_offset; + regnum_t regnum; +} compile_stack_elt_t; + + +typedef struct +{ + compile_stack_elt_t *stack; + unsigned size; + unsigned avail; /* Offset of next open position. */ +} compile_stack_type; + + +#define INIT_COMPILE_STACK_SIZE 32 + +#define COMPILE_STACK_EMPTY (compile_stack.avail == 0) +#define COMPILE_STACK_FULL (compile_stack.avail == compile_stack.size) + +/* The next available element. */ +#define COMPILE_STACK_TOP (compile_stack.stack[compile_stack.avail]) + + +/* Set the bit for character C in a list. */ +#define SET_LIST_BIT(c) \ + (b[((unsigned char) (c)) / BYTEWIDTH] \ + |= 1 << (((unsigned char) c) % BYTEWIDTH)) + + +/* Get the next unsigned number in the uncompiled pattern. */ +#define GET_UNSIGNED_NUMBER(num) \ + { if (p != pend) \ + { \ + PATFETCH (c); \ + while (ISDIGIT (c)) \ + { \ + if (num < 0) \ + num = 0; \ + num = num * 10 + c - '0'; \ + if (p == pend) \ + break; \ + PATFETCH (c); \ + } \ + } \ + } + +#define CHAR_CLASS_MAX_LENGTH 6 /* Namely, `xdigit'. */ + +#define IS_CHAR_CLASS(string) \ + (STREQ (string, "alpha") || STREQ (string, "upper") \ + || STREQ (string, "lower") || STREQ (string, "digit") \ + || STREQ (string, "alnum") || STREQ (string, "xdigit") \ + || STREQ (string, "space") || STREQ (string, "print") \ + || STREQ (string, "punct") || STREQ (string, "graph") \ + || STREQ (string, "cntrl") || STREQ (string, "blank")) + +/* `regex_compile' compiles PATTERN (of length SIZE) according to SYNTAX. + Returns one of error codes defined in `regex.h', or zero for success. + + Assumes the `allocated' (and perhaps `buffer') and `translate' + fields are set in BUFP on entry. + + If it succeeds, results are put in BUFP (if it returns an error, the + contents of BUFP are undefined): + `buffer' is the compiled pattern; + `syntax' is set to SYNTAX; + `used' is set to the length of the compiled pattern; + `fastmap_accurate' is zero; + `re_nsub' is the number of subexpressions in PATTERN; + `not_bol' and `not_eol' are zero; + + The `fastmap' and `newline_anchor' fields are neither + examined nor set. */ + +/* Return, freeing storage we allocated. */ +#define FREE_STACK_RETURN(value) \ + return (free (compile_stack.stack), value) + +static reg_errcode_t +regex_compile (pattern, size, syntax, bufp) + const char *pattern; + int size; + reg_syntax_t syntax; + struct re_pattern_buffer *bufp; +{ + /* We fetch characters from PATTERN here. Even though PATTERN is + `char *' (i.e., signed), we declare these variables as unsigned, so + they can be reliably used as array indices. */ + register unsigned char c, c1; + + /* A random temporary spot in PATTERN. */ + const char *p1; + + /* Points to the end of the buffer, where we should append. */ + register unsigned char *b; + + /* Keeps track of unclosed groups. */ + compile_stack_type compile_stack; + + /* Points to the current (ending) position in the pattern. */ + const char *p = pattern; + const char *pend = pattern + size; + + /* How to translate the characters in the pattern. */ + char *translate = bufp->translate; + + /* Address of the count-byte of the most recently inserted `exactn' + command. This makes it possible to tell if a new exact-match + character can be added to that command or if the character requires + a new `exactn' command. */ + unsigned char *pending_exact = 0; + + /* Address of start of the most recently finished expression. + This tells, e.g., postfix * where to find the start of its + operand. Reset at the beginning of groups and alternatives. */ + unsigned char *laststart = 0; + + /* Address of beginning of regexp, or inside of last group. */ + unsigned char *begalt; + + /* Place in the uncompiled pattern (i.e., the {) to + which to go back if the interval is invalid. */ + const char *beg_interval; + + /* Address of the place where a forward jump should go to the end of + the containing expression. Each alternative of an `or' -- except the + last -- ends with a forward jump of this sort. */ + unsigned char *fixup_alt_jump = 0; + + /* Counts open-groups as they are encountered. Remembered for the + matching close-group on the compile stack, so the same register + number is put in the stop_memory as the start_memory. */ + regnum_t regnum = 0; + +#ifdef DEBUG + DEBUG_PRINT1 ("\nCompiling pattern: "); + if (debug) + { + unsigned debug_count; + + for (debug_count = 0; debug_count < size; debug_count++) + printchar (pattern[debug_count]); + putchar ('\n'); + } +#endif /* DEBUG */ + + /* Initialize the compile stack. */ + compile_stack.stack = TALLOC (INIT_COMPILE_STACK_SIZE, compile_stack_elt_t); + if (compile_stack.stack == NULL) + return REG_ESPACE; + + compile_stack.size = INIT_COMPILE_STACK_SIZE; + compile_stack.avail = 0; + + /* Initialize the pattern buffer. */ + bufp->syntax = syntax; + bufp->fastmap_accurate = 0; + bufp->not_bol = bufp->not_eol = 0; + + /* Set `used' to zero, so that if we return an error, the pattern + printer (for debugging) will think there's no pattern. We reset it + at the end. */ + bufp->used = 0; + + /* Always count groups, whether or not bufp->no_sub is set. */ + bufp->re_nsub = 0; + +#if !defined (emacs) && !defined (SYNTAX_TABLE) + /* Initialize the syntax table. */ + init_syntax_once (); +#endif + + if (bufp->allocated == 0) + { + if (bufp->buffer) + { /* 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. */ + RETALLOC (bufp->buffer, INIT_BUF_SIZE, unsigned char); + } + else + { /* Caller did not allocate a buffer. Do it for them. */ + bufp->buffer = TALLOC (INIT_BUF_SIZE, unsigned char); + } + if (!bufp->buffer) FREE_STACK_RETURN (REG_ESPACE); + + bufp->allocated = INIT_BUF_SIZE; + } + + begalt = b = bufp->buffer; + + /* Loop through the uncompiled pattern until we're at the end. */ + while (p != pend) + { + PATFETCH (c); + + switch (c) + { + case '^': + { + if ( /* If at start of pattern, it's an operator. */ + p == pattern + 1 + /* If context independent, it's an operator. */ + || syntax & RE_CONTEXT_INDEP_ANCHORS + /* Otherwise, depends on what's come before. */ + || at_begline_loc_p (pattern, p, syntax)) + BUF_PUSH (begline); + else + goto normal_char; + } + break; + + + case '$': + { + if ( /* If at end of pattern, it's an operator. */ + p == pend + /* If context independent, it's an operator. */ + || syntax & RE_CONTEXT_INDEP_ANCHORS + /* Otherwise, depends on what's next. */ + || at_endline_loc_p (p, pend, syntax)) + BUF_PUSH (endline); + else + goto normal_char; + } + break; + + + case '+': + case '?': + if ((syntax & RE_BK_PLUS_QM) + || (syntax & RE_LIMITED_OPS)) + goto normal_char; + handle_plus: + case '*': + /* If there is no previous pattern... */ + if (!laststart) + { + if (syntax & RE_CONTEXT_INVALID_OPS) + FREE_STACK_RETURN (REG_BADRPT); + else if (!(syntax & RE_CONTEXT_INDEP_OPS)) + goto normal_char; + } + + { + /* Are we optimizing this jump? */ + boolean keep_string_p = false; + + /* 1 means zero (many) matches is allowed. */ + char zero_times_ok = 0, many_times_ok = 0; + + /* If there is a sequence of repetition chars, collapse it + down to just one (the right one). We can't combine + interval operators with these because of, e.g., `a{2}*', + which should only match an even number of `a's. */ + + for (;;) + { + zero_times_ok |= c != '+'; + many_times_ok |= c != '?'; + + if (p == pend) + break; + + PATFETCH (c); + + if (c == '*' + || (!(syntax & RE_BK_PLUS_QM) && (c == '+' || c == '?'))) + ; + + else if (syntax & RE_BK_PLUS_QM && c == '\\') + { + if (p == pend) FREE_STACK_RETURN (REG_EESCAPE); + + PATFETCH (c1); + if (!(c1 == '+' || c1 == '?')) + { + PATUNFETCH; + PATUNFETCH; + break; + } + + c = c1; + } + else + { + PATUNFETCH; + break; + } + + /* If we get here, we found another repeat character. */ + } + + /* Star, etc. applied to an empty pattern is equivalent + to an empty pattern. */ + if (!laststart) + break; + + /* Now we know whether or not zero matches is allowed + and also whether or not two or more matches is allowed. */ + if (many_times_ok) + { /* More than one repetition is allowed, so put in at the + end a backward relative jump from `b' to before the next + jump we're going to put in below (which jumps from + laststart to after this jump). + + But if we are at the `*' in the exact sequence `.*\n', + insert an unconditional jump backwards to the ., + instead of the beginning of the loop. This way we only + push a failure point once, instead of every time + through the loop. */ + assert (p - 1 > pattern); + + /* Allocate the space for the jump. */ + GET_BUFFER_SPACE (3); + + /* We know we are not at the first character of the pattern, + because laststart was nonzero. And we've already + incremented `p', by the way, to be the character after + the `*'. Do we have to do something analogous here + for null bytes, because of RE_DOT_NOT_NULL? */ + if (TRANSLATE (*(p - 2)) == TRANSLATE ('.') + && zero_times_ok + && p < pend && TRANSLATE (*p) == TRANSLATE ('\n') + && !(syntax & RE_DOT_NEWLINE)) + { /* We have .*\n. */ + STORE_JUMP (jump, b, laststart); + keep_string_p = true; + } + else + /* Anything else. */ + STORE_JUMP (maybe_pop_jump, b, laststart - 3); + + /* We've added more stuff to the buffer. */ + b += 3; + } + + /* On failure, jump from laststart to b + 3, which will be the + end of the buffer after this jump is inserted. */ + GET_BUFFER_SPACE (3); + INSERT_JUMP (keep_string_p ? on_failure_keep_string_jump + : on_failure_jump, + laststart, b + 3); + pending_exact = 0; + b += 3; + + if (!zero_times_ok) + { + /* At least one repetition is required, so insert a + `dummy_failure_jump' before the initial + `on_failure_jump' instruction of the loop. This + effects a skip over that instruction the first time + we hit that loop. */ + GET_BUFFER_SPACE (3); + INSERT_JUMP (dummy_failure_jump, laststart, laststart + 6); + b += 3; + } + } + break; + + + case '.': + laststart = b; + BUF_PUSH (anychar); + break; + + + case '[': + { + boolean had_char_class = false; + + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + /* Ensure that we have enough space to push a charset: the + opcode, the length count, and the bitset; 34 bytes in all. */ + GET_BUFFER_SPACE (34); + + laststart = b; + + /* We test `*p == '^' twice, instead of using an if + statement, so we only need one BUF_PUSH. */ + BUF_PUSH (*p == '^' ? charset_not : charset); + if (*p == '^') + p++; + + /* Remember the first position in the bracket expression. */ + p1 = p; + + /* Push the number of bytes in the bitmap. */ + BUF_PUSH ((1 << BYTEWIDTH) / BYTEWIDTH); + + /* Clear the whole map. */ + bzero (b, (1 << BYTEWIDTH) / BYTEWIDTH); + + /* charset_not matches newline according to a syntax bit. */ + if ((re_opcode_t) b[-2] == charset_not + && (syntax & RE_HAT_LISTS_NOT_NEWLINE)) + SET_LIST_BIT ('\n'); + + /* Read in characters and ranges, setting map bits. */ + for (;;) + { + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + PATFETCH (c); + + /* \ might escape characters inside [...] and [^...]. */ + if ((syntax & RE_BACKSLASH_ESCAPE_IN_LISTS) && c == '\\') + { + if (p == pend) FREE_STACK_RETURN (REG_EESCAPE); + + PATFETCH (c1); + SET_LIST_BIT (c1); + continue; + } + + /* Could be the end of the bracket expression. If it's + not (i.e., when the bracket expression is `[]' so + far), the ']' character bit gets set way below. */ + if (c == ']' && p != p1 + 1) + break; + + /* Look ahead to see if it's a range when the last thing + was a character class. */ + if (had_char_class && c == '-' && *p != ']') + FREE_STACK_RETURN (REG_ERANGE); + + /* Look ahead to see if it's a range when the last thing + was a character: if this is a hyphen not at the + beginning or the end of a list, then it's the range + operator. */ + if (c == '-' + && !(p - 2 >= pattern && p[-2] == '[') + && !(p - 3 >= pattern && p[-3] == '[' && p[-2] == '^') + && *p != ']') + { + reg_errcode_t ret + = compile_range (&p, pend, translate, syntax, b); + if (ret != REG_NOERROR) FREE_STACK_RETURN (ret); + } + + else if (p[0] == '-' && p[1] != ']') + { /* This handles ranges made up of characters only. */ + reg_errcode_t ret; + + /* Move past the `-'. */ + PATFETCH (c1); + + ret = compile_range (&p, pend, translate, syntax, b); + if (ret != REG_NOERROR) FREE_STACK_RETURN (ret); + } + + /* See if we're at the beginning of a possible character + class. */ + + else if (syntax & RE_CHAR_CLASSES && c == '[' && *p == ':') + { /* Leave room for the null. */ + char str[CHAR_CLASS_MAX_LENGTH + 1]; + + PATFETCH (c); + c1 = 0; + + /* If pattern is `[[:'. */ + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (;;) + { + PATFETCH (c); + if (c == ':' || c == ']' || p == pend + || c1 == CHAR_CLASS_MAX_LENGTH) + break; + str[c1++] = c; + } + str[c1] = '\0'; + + /* If isn't a word bracketed by `[:' and:`]': + undo the ending character, the letters, and leave + the leading `:' and `[' (but set bits for them). */ + if (c == ':' && *p == ']') + { + int ch; + boolean is_alnum = STREQ (str, "alnum"); + boolean is_alpha = STREQ (str, "alpha"); + boolean is_blank = STREQ (str, "blank"); + boolean is_cntrl = STREQ (str, "cntrl"); + boolean is_digit = STREQ (str, "digit"); + boolean is_graph = STREQ (str, "graph"); + boolean is_lower = STREQ (str, "lower"); + boolean is_print = STREQ (str, "print"); + boolean is_punct = STREQ (str, "punct"); + boolean is_space = STREQ (str, "space"); + boolean is_upper = STREQ (str, "upper"); + boolean is_xdigit = STREQ (str, "xdigit"); + + if (!IS_CHAR_CLASS (str)) + FREE_STACK_RETURN (REG_ECTYPE); + + /* Throw away the ] at the end of the character + class. */ + PATFETCH (c); + + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (ch = 0; ch < 1 << BYTEWIDTH; ch++) + { + /* This was split into 3 if's to + avoid an arbitrary limit in some compiler. */ + if ( (is_alnum && ISALNUM (ch)) + || (is_alpha && ISALPHA (ch)) + || (is_blank && ISBLANK (ch)) + || (is_cntrl && ISCNTRL (ch))) + SET_LIST_BIT (ch); + if ( (is_digit && ISDIGIT (ch)) + || (is_graph && ISGRAPH (ch)) + || (is_lower && ISLOWER (ch)) + || (is_print && ISPRINT (ch))) + SET_LIST_BIT (ch); + if ( (is_punct && ISPUNCT (ch)) + || (is_space && ISSPACE (ch)) + || (is_upper && ISUPPER (ch)) + || (is_xdigit && ISXDIGIT (ch))) + SET_LIST_BIT (ch); + } + had_char_class = true; + } + else + { + c1++; + while (c1--) + PATUNFETCH; + SET_LIST_BIT ('['); + SET_LIST_BIT (':'); + had_char_class = false; + } + } + else + { + had_char_class = false; + SET_LIST_BIT (c); + } + } + + /* Discard any (non)matching list bytes that are all 0 at the + end of the map. Decrease the map-length byte too. */ + while ((int) b[-1] > 0 && b[b[-1] - 1] == 0) + b[-1]--; + b += b[-1]; + } + break; + + + case '(': + if (syntax & RE_NO_BK_PARENS) + goto handle_open; + else + goto normal_char; + + + case ')': + if (syntax & RE_NO_BK_PARENS) + goto handle_close; + else + goto normal_char; + + + case '\n': + if (syntax & RE_NEWLINE_ALT) + goto handle_alt; + else + goto normal_char; + + + case '|': + if (syntax & RE_NO_BK_VBAR) + goto handle_alt; + else + goto normal_char; + + + case '{': + if (syntax & RE_INTERVALS && syntax & RE_NO_BK_BRACES) + goto handle_interval; + else + goto normal_char; + + + case '\\': + if (p == pend) FREE_STACK_RETURN (REG_EESCAPE); + + /* Do not translate the character after the \, so that we can + distinguish, e.g., \B from \b, even if we normally would + translate, e.g., B to b. */ + PATFETCH_RAW (c); + + switch (c) + { + case '(': + if (syntax & RE_NO_BK_PARENS) + goto normal_backslash; + + handle_open: + bufp->re_nsub++; + regnum++; + + if (COMPILE_STACK_FULL) + { + RETALLOC (compile_stack.stack, compile_stack.size << 1, + compile_stack_elt_t); + if (compile_stack.stack == NULL) return REG_ESPACE; + + compile_stack.size <<= 1; + } + + /* These are the values to restore when we hit end of this + group. They are all relative offsets, so that if the + whole pattern moves because of realloc, they will still + be valid. */ + COMPILE_STACK_TOP.begalt_offset = begalt - bufp->buffer; + COMPILE_STACK_TOP.fixup_alt_jump + = fixup_alt_jump ? fixup_alt_jump - bufp->buffer + 1 : 0; + COMPILE_STACK_TOP.laststart_offset = b - bufp->buffer; + COMPILE_STACK_TOP.regnum = regnum; + + /* We will eventually replace the 0 with the number of + groups inner to this one. But do not push a + start_memory for groups beyond the last one we can + represent in the compiled pattern. */ + if (regnum <= MAX_REGNUM) + { + COMPILE_STACK_TOP.inner_group_offset = b - bufp->buffer + 2; + BUF_PUSH_3 (start_memory, regnum, 0); + } + + compile_stack.avail++; + + fixup_alt_jump = 0; + laststart = 0; + begalt = b; + /* If we've reached MAX_REGNUM groups, then this open + won't actually generate any code, so we'll have to + clear pending_exact explicitly. */ + pending_exact = 0; + break; + + + case ')': + if (syntax & RE_NO_BK_PARENS) goto normal_backslash; + + if (COMPILE_STACK_EMPTY) + if (syntax & RE_UNMATCHED_RIGHT_PAREN_ORD) + goto normal_backslash; + else + FREE_STACK_RETURN (REG_ERPAREN); + + handle_close: + if (fixup_alt_jump) + { /* Push a dummy failure point at the end of the + alternative for a possible future + `pop_failure_jump' to pop. See comments at + `push_dummy_failure' in `re_match_2'. */ + BUF_PUSH (push_dummy_failure); + + /* We allocated space for this jump when we assigned + to `fixup_alt_jump', in the `handle_alt' case below. */ + STORE_JUMP (jump_past_alt, fixup_alt_jump, b - 1); + } + + /* See similar code for backslashed left paren above. */ + if (COMPILE_STACK_EMPTY) + if (syntax & RE_UNMATCHED_RIGHT_PAREN_ORD) + goto normal_char; + else + FREE_STACK_RETURN (REG_ERPAREN); + + /* Since we just checked for an empty stack above, this + ``can't happen''. */ + assert (compile_stack.avail != 0); + { + /* We don't just want to restore into `regnum', because + later groups should continue to be numbered higher, + as in `(ab)c(de)' -- the second group is #2. */ + regnum_t this_group_regnum; + + compile_stack.avail--; + begalt = bufp->buffer + COMPILE_STACK_TOP.begalt_offset; + fixup_alt_jump + = COMPILE_STACK_TOP.fixup_alt_jump + ? bufp->buffer + COMPILE_STACK_TOP.fixup_alt_jump - 1 + : 0; + laststart = bufp->buffer + COMPILE_STACK_TOP.laststart_offset; + this_group_regnum = COMPILE_STACK_TOP.regnum; + /* If we've reached MAX_REGNUM groups, then this open + won't actually generate any code, so we'll have to + clear pending_exact explicitly. */ + pending_exact = 0; + + /* We're at the end of the group, so now we know how many + groups were inside this one. */ + if (this_group_regnum <= MAX_REGNUM) + { + unsigned char *inner_group_loc + = bufp->buffer + COMPILE_STACK_TOP.inner_group_offset; + + *inner_group_loc = regnum - this_group_regnum; + BUF_PUSH_3 (stop_memory, this_group_regnum, + regnum - this_group_regnum); + } + } + break; + + + case '|': /* `\|'. */ + if (syntax & RE_LIMITED_OPS || syntax & RE_NO_BK_VBAR) + goto normal_backslash; + handle_alt: + if (syntax & RE_LIMITED_OPS) + goto normal_char; + + /* Insert before the previous alternative a jump which + jumps to this alternative if the former fails. */ + GET_BUFFER_SPACE (3); + INSERT_JUMP (on_failure_jump, begalt, b + 6); + pending_exact = 0; + b += 3; + + /* The alternative before this one has a jump after it + which gets executed if it gets matched. Adjust that + jump so it will jump to this alternative's analogous + jump (put in below, which in turn will jump to the next + (if any) alternative's such jump, etc.). The last such + jump jumps to the correct final destination. A picture: + _____ _____ + | | | | + | v | v + a | b | c + + If we are at `b', then fixup_alt_jump right now points to a + three-byte space after `a'. We'll put in the jump, set + fixup_alt_jump to right after `b', and leave behind three + bytes which we'll fill in when we get to after `c'. */ + + if (fixup_alt_jump) + STORE_JUMP (jump_past_alt, fixup_alt_jump, b); + + /* Mark and leave space for a jump after this alternative, + to be filled in later either by next alternative or + when know we're at the end of a series of alternatives. */ + fixup_alt_jump = b; + GET_BUFFER_SPACE (3); + b += 3; + + laststart = 0; + begalt = b; + break; + + + case '{': + /* If \{ is a literal. */ + if (!(syntax & RE_INTERVALS) + /* If we're at `\{' and it's not the open-interval + operator. */ + || ((syntax & RE_INTERVALS) && (syntax & RE_NO_BK_BRACES)) + || (p - 2 == pattern && p == pend)) + goto normal_backslash; + + handle_interval: + { + /* If got here, then the syntax allows intervals. */ + + /* At least (most) this many matches must be made. */ + int lower_bound = -1, upper_bound = -1; + + beg_interval = p - 1; + + if (p == pend) + { + if (syntax & RE_NO_BK_BRACES) + goto unfetch_interval; + else + FREE_STACK_RETURN (REG_EBRACE); + } + + GET_UNSIGNED_NUMBER (lower_bound); + + if (c == ',') + { + GET_UNSIGNED_NUMBER (upper_bound); + if (upper_bound < 0) upper_bound = RE_DUP_MAX; + } + else + /* Interval such as `{1}' => match exactly once. */ + upper_bound = lower_bound; + + if (lower_bound < 0 || upper_bound > RE_DUP_MAX + || lower_bound > upper_bound) + { + if (syntax & RE_NO_BK_BRACES) + goto unfetch_interval; + else + FREE_STACK_RETURN (REG_BADBR); + } + + if (!(syntax & RE_NO_BK_BRACES)) + { + if (c != '\\') FREE_STACK_RETURN (REG_EBRACE); + + PATFETCH (c); + } + + if (c != '}') + { + if (syntax & RE_NO_BK_BRACES) + goto unfetch_interval; + else + FREE_STACK_RETURN (REG_BADBR); + } + + /* We just parsed a valid interval. */ + + /* If it's invalid to have no preceding re. */ + if (!laststart) + { + if (syntax & RE_CONTEXT_INVALID_OPS) + FREE_STACK_RETURN (REG_BADRPT); + else if (syntax & RE_CONTEXT_INDEP_OPS) + laststart = b; + else + goto unfetch_interval; + } + + /* If the upper bound is zero, don't want to succeed at + all; jump from `laststart' to `b + 3', which will be + the end of the buffer after we insert the jump. */ + if (upper_bound == 0) + { + GET_BUFFER_SPACE (3); + INSERT_JUMP (jump, laststart, b + 3); + b += 3; + } + + /* Otherwise, we have a nontrivial interval. When + we're all done, the pattern will look like: + set_number_at <jump count> <upper bound> + set_number_at <succeed_n count> <lower bound> + succeed_n <after jump addr> <succeed_n count> + <body of loop> + jump_n <succeed_n addr> <jump count> + (The upper bound and `jump_n' are omitted if + `upper_bound' is 1, though.) */ + else + { /* If the upper bound is > 1, we need to insert + more at the end of the loop. */ + unsigned nbytes = 10 + (upper_bound > 1) * 10; + + GET_BUFFER_SPACE (nbytes); + + /* Initialize lower bound of the `succeed_n', even + though it will be set during matching by its + attendant `set_number_at' (inserted next), + because `re_compile_fastmap' needs to know. + Jump to the `jump_n' we might insert below. */ + INSERT_JUMP2 (succeed_n, laststart, + b + 5 + (upper_bound > 1) * 5, + lower_bound); + b += 5; + + /* Code to initialize the lower bound. Insert + before the `succeed_n'. The `5' is the last two + bytes of this `set_number_at', plus 3 bytes of + the following `succeed_n'. */ + insert_op2 (set_number_at, laststart, 5, lower_bound, b); + b += 5; + + if (upper_bound > 1) + { /* More than one repetition is allowed, so + append a backward jump to the `succeed_n' + that starts this interval. + + When we've reached this during matching, + we'll have matched the interval once, so + jump back only `upper_bound - 1' times. */ + STORE_JUMP2 (jump_n, b, laststart + 5, + upper_bound - 1); + b += 5; + + /* The location we want to set is the second + parameter of the `jump_n'; that is `b-2' as + an absolute address. `laststart' will be + the `set_number_at' we're about to insert; + `laststart+3' the number to set, the source + for the relative address. But we are + inserting into the middle of the pattern -- + so everything is getting moved up by 5. + Conclusion: (b - 2) - (laststart + 3) + 5, + i.e., b - laststart. + + We insert this at the beginning of the loop + so that if we fail during matching, we'll + reinitialize the bounds. */ + insert_op2 (set_number_at, laststart, b - laststart, + upper_bound - 1, b); + b += 5; + } + } + pending_exact = 0; + beg_interval = NULL; + } + break; + + unfetch_interval: + /* If an invalid interval, match the characters as literals. */ + assert (beg_interval); + p = beg_interval; + beg_interval = NULL; + + /* normal_char and normal_backslash need `c'. */ + PATFETCH (c); + + if (!(syntax & RE_NO_BK_BRACES)) + { + if (p > pattern && p[-1] == '\\') + goto normal_backslash; + } + goto normal_char; + +#ifdef emacs + /* There is no way to specify the before_dot and after_dot + operators. rms says this is ok. --karl */ + case '=': + BUF_PUSH (at_dot); + break; + + case 's': + laststart = b; + PATFETCH (c); + BUF_PUSH_2 (syntaxspec, syntax_spec_code[c]); + break; + + case 'S': + laststart = b; + PATFETCH (c); + BUF_PUSH_2 (notsyntaxspec, syntax_spec_code[c]); + break; +#endif /* emacs */ + + + case 'w': + laststart = b; + BUF_PUSH (wordchar); + break; + + + case 'W': + laststart = b; + BUF_PUSH (notwordchar); + break; + + + case '<': + BUF_PUSH (wordbeg); + break; + + case '>': + BUF_PUSH (wordend); + break; + + case 'b': + BUF_PUSH (wordbound); + break; + + case 'B': + BUF_PUSH (notwordbound); + break; + + case '`': + BUF_PUSH (begbuf); + break; + + case '\'': + BUF_PUSH (endbuf); + 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) + goto normal_char; + + c1 = c - '0'; + + if (c1 > regnum) + FREE_STACK_RETURN (REG_ESUBREG); + + /* Can't back reference to a subexpression if inside of it. */ + if (group_in_compile_stack (compile_stack, c1)) + goto normal_char; + + laststart = b; + BUF_PUSH_2 (duplicate, c1); + break; + + + case '+': + case '?': + if (syntax & RE_BK_PLUS_QM) + goto handle_plus; + else + goto normal_backslash; + + default: + normal_backslash: + /* You might think it would be useful for \ to mean + not to translate; but if we don't translate it + it will never match anything. */ + c = TRANSLATE (c); + goto normal_char; + } + break; + + + default: + /* Expects the character in `c'. */ + normal_char: + /* If no exactn currently being built. */ + if (!pending_exact + + /* If last exactn not at current position. */ + || pending_exact + *pending_exact + 1 != b + + /* We have only one byte following the exactn for the count. */ + || *pending_exact == (1 << BYTEWIDTH) - 1 + + /* If followed by a repetition operator. */ + || *p == '*' || *p == '^' + || ((syntax & RE_BK_PLUS_QM) + ? *p == '\\' && (p[1] == '+' || p[1] == '?') + : (*p == '+' || *p == '?')) + || ((syntax & RE_INTERVALS) + && ((syntax & RE_NO_BK_BRACES) + ? *p == '{' + : (p[0] == '\\' && p[1] == '{')))) + { + /* Start building a new exactn. */ + + laststart = b; + + BUF_PUSH_2 (exactn, 0); + pending_exact = b - 1; + } + + BUF_PUSH (c); + (*pending_exact)++; + break; + } /* switch (c) */ + } /* while p != pend */ + + + /* Through the pattern now. */ + + if (fixup_alt_jump) + STORE_JUMP (jump_past_alt, fixup_alt_jump, b); + + if (!COMPILE_STACK_EMPTY) + FREE_STACK_RETURN (REG_EPAREN); + + free (compile_stack.stack); + + /* We have succeeded; set the length of the buffer. */ + bufp->used = b - bufp->buffer; + +#ifdef DEBUG + if (debug) + { + DEBUG_PRINT1 ("\nCompiled pattern: \n"); + print_compiled_pattern (bufp); + } +#endif /* DEBUG */ + +#ifndef MATCH_MAY_ALLOCATE + /* Initialize the failure stack to the largest possible stack. This + isn't necessary unless we're trying to avoid calling alloca in + the search and match routines. */ + { + int num_regs = bufp->re_nsub + 1; + + /* Since DOUBLE_FAIL_STACK refuses to double only if the current size + is strictly greater than re_max_failures, the largest possible stack + is 2 * re_max_failures failure points. */ + if (fail_stack.size < (2 * re_max_failures * MAX_FAILURE_ITEMS)) + { + fail_stack.size = (2 * re_max_failures * MAX_FAILURE_ITEMS); + +#ifdef emacs + if (! fail_stack.stack) + fail_stack.stack + = (fail_stack_elt_t *) xmalloc (fail_stack.size + * sizeof (fail_stack_elt_t)); + else + fail_stack.stack + = (fail_stack_elt_t *) xrealloc (fail_stack.stack, + (fail_stack.size + * sizeof (fail_stack_elt_t))); +#else /* not emacs */ + if (! fail_stack.stack) + fail_stack.stack + = (fail_stack_elt_t *) malloc (fail_stack.size + * sizeof (fail_stack_elt_t)); + else + fail_stack.stack + = (fail_stack_elt_t *) realloc (fail_stack.stack, + (fail_stack.size + * sizeof (fail_stack_elt_t))); +#endif /* not emacs */ + } + + /* Initialize some other variables the matcher uses. */ + RETALLOC_IF (regstart, num_regs, const char *); + RETALLOC_IF (regend, num_regs, const char *); + RETALLOC_IF (old_regstart, num_regs, const char *); + RETALLOC_IF (old_regend, num_regs, const char *); + RETALLOC_IF (best_regstart, num_regs, const char *); + RETALLOC_IF (best_regend, num_regs, const char *); + RETALLOC_IF (reg_info, num_regs, register_info_type); + RETALLOC_IF (reg_dummy, num_regs, const char *); + RETALLOC_IF (reg_info_dummy, num_regs, register_info_type); + } +#endif + + return REG_NOERROR; +} /* regex_compile */ + +/* Subroutines for `regex_compile'. */ + +/* Store OP at LOC followed by two-byte integer parameter ARG. */ + +static void +store_op1 (op, loc, arg) + re_opcode_t op; + unsigned char *loc; + int arg; +{ + *loc = (unsigned char) op; + STORE_NUMBER (loc + 1, arg); +} + + +/* Like `store_op1', but for two two-byte parameters ARG1 and ARG2. */ + +static void +store_op2 (op, loc, arg1, arg2) + re_opcode_t op; + unsigned char *loc; + int arg1, arg2; +{ + *loc = (unsigned char) op; + STORE_NUMBER (loc + 1, arg1); + STORE_NUMBER (loc + 3, arg2); +} + + +/* Copy the bytes from LOC to END to open up three bytes of space at LOC + for OP followed by two-byte integer parameter ARG. */ + +static void +insert_op1 (op, loc, arg, end) + re_opcode_t op; + unsigned char *loc; + int arg; + unsigned char *end; +{ + register unsigned char *pfrom = end; + register unsigned char *pto = end + 3; + + while (pfrom != loc) + *--pto = *--pfrom; + + store_op1 (op, loc, arg); +} + + +/* Like `insert_op1', but for two two-byte parameters ARG1 and ARG2. */ + +static void +insert_op2 (op, loc, arg1, arg2, end) + re_opcode_t op; + unsigned char *loc; + int arg1, arg2; + unsigned char *end; +{ + register unsigned char *pfrom = end; + register unsigned char *pto = end + 5; + + while (pfrom != loc) + *--pto = *--pfrom; + + store_op2 (op, loc, arg1, arg2); +} + + +/* P points to just after a ^ in PATTERN. Return true if that ^ comes + after an alternative or a begin-subexpression. We assume there is at + least one character before the ^. */ + +static boolean +at_begline_loc_p (pattern, p, syntax) + const char *pattern, *p; + reg_syntax_t syntax; +{ + const char *prev = p - 2; + boolean prev_prev_backslash = prev > pattern && prev[-1] == '\\'; + + return + /* After a subexpression? */ + (*prev == '(' && (syntax & RE_NO_BK_PARENS || prev_prev_backslash)) + /* After an alternative? */ + || (*prev == '|' && (syntax & RE_NO_BK_VBAR || prev_prev_backslash)); +} + + +/* The dual of at_begline_loc_p. This one is for $. We assume there is + at least one character after the $, i.e., `P < PEND'. */ + +static boolean +at_endline_loc_p (p, pend, syntax) + const char *p, *pend; + int syntax; +{ + const char *next = p; + boolean next_backslash = *next == '\\'; + const char *next_next = p + 1 < pend ? p + 1 : NULL; + + return + /* Before a subexpression? */ + (syntax & RE_NO_BK_PARENS ? *next == ')' + : next_backslash && next_next && *next_next == ')') + /* Before an alternative? */ + || (syntax & RE_NO_BK_VBAR ? *next == '|' + : next_backslash && next_next && *next_next == '|'); +} + + +/* Returns true if REGNUM is in one of COMPILE_STACK's elements and + false if it's not. */ + +static boolean +group_in_compile_stack (compile_stack, regnum) + compile_stack_type compile_stack; + regnum_t regnum; +{ + int this_element; + + for (this_element = compile_stack.avail - 1; + this_element >= 0; + this_element--) + if (compile_stack.stack[this_element].regnum == regnum) + return true; + + return false; +} + + +/* Read the ending character of a range (in a bracket expression) from the + uncompiled pattern *P_PTR (which ends at PEND). We assume the + starting character is in `P[-2]'. (`P[-1]' is the character `-'.) + Then we set the translation of all bits between the starting and + ending characters (inclusive) in the compiled pattern B. + + Return an error code. + + We use these short variable names so we can use the same macros as + `regex_compile' itself. */ + +static reg_errcode_t +compile_range (p_ptr, pend, translate, syntax, b) + const char **p_ptr, *pend; + char *translate; + reg_syntax_t syntax; + unsigned char *b; +{ + unsigned this_char; + + const char *p = *p_ptr; + int range_start, range_end; + + if (p == pend) + return REG_ERANGE; + + /* Even though the pattern is a signed `char *', we need to fetch + with unsigned char *'s; if the high bit of the pattern character + is set, the range endpoints will be negative if we fetch using a + signed char *. + + We also want to fetch the endpoints without translating them; the + appropriate translation is done in the bit-setting loop below. */ + /* The SVR4 compiler on the 3B2 had trouble with unsigned const char *. */ + range_start = ((const unsigned char *) p)[-2]; + range_end = ((const unsigned char *) p)[0]; + + /* Have to increment the pointer into the pattern string, so the + caller isn't still at the ending character. */ + (*p_ptr)++; + + /* If the start is after the end, the range is empty. */ + if (range_start > range_end) + return syntax & RE_NO_EMPTY_RANGES ? REG_ERANGE : REG_NOERROR; + + /* Here we see why `this_char' has to be larger than an `unsigned + char' -- the range is inclusive, so if `range_end' == 0xff + (assuming 8-bit characters), we would otherwise go into an infinite + loop, since all characters <= 0xff. */ + for (this_char = range_start; this_char <= range_end; this_char++) + { + SET_LIST_BIT (TRANSLATE (this_char)); + } + + return REG_NOERROR; +} + +/* re_compile_fastmap computes a ``fastmap'' for the compiled pattern in + BUFP. A fastmap records which of the (1 << BYTEWIDTH) possible + characters can start a string that matches the pattern. This fastmap + is used by re_search to skip quickly over impossible starting points. + + The caller must supply the address of a (1 << BYTEWIDTH)-byte data + area as BUFP->fastmap. + + We set the `fastmap', `fastmap_accurate', and `can_be_null' fields in + the pattern buffer. + + Returns 0 if we succeed, -2 if an internal error. */ + +int +re_compile_fastmap (bufp) + struct re_pattern_buffer *bufp; +{ + int j, k; +#ifdef MATCH_MAY_ALLOCATE + fail_stack_type fail_stack; +#endif +#ifndef REGEX_MALLOC + char *destination; +#endif + /* We don't push any register information onto the failure stack. */ + unsigned num_regs = 0; + + register char *fastmap = bufp->fastmap; + unsigned char *pattern = bufp->buffer; + unsigned long size = bufp->used; + unsigned char *p = pattern; + register unsigned char *pend = pattern + size; + + /* Assume that each path through the pattern can be null until + proven otherwise. We set this false at the bottom of switch + statement, to which we get only if a particular path doesn't + match the empty string. */ + boolean path_can_be_null = true; + + /* We aren't doing a `succeed_n' to begin with. */ + boolean succeed_n_p = false; + + assert (fastmap != NULL && p != NULL); + + INIT_FAIL_STACK (); + bzero (fastmap, 1 << BYTEWIDTH); /* Assume nothing's valid. */ + bufp->fastmap_accurate = 1; /* It will be when we're done. */ + bufp->can_be_null = 0; + + while (p != pend || !FAIL_STACK_EMPTY ()) + { + if (p == pend) + { + bufp->can_be_null |= path_can_be_null; + + /* Reset for next path. */ + path_can_be_null = true; + + p = fail_stack.stack[--fail_stack.avail]; + } + + /* We should never be about to go beyond the end of the pattern. */ + assert (p < pend); + +#ifdef SWITCH_ENUM_BUG + switch ((int) ((re_opcode_t) *p++)) +#else + switch ((re_opcode_t) *p++) +#endif + { + + /* I guess the idea here is to simply not bother with a fastmap + if a backreference is used, since it's too hard to figure out + the fastmap for the corresponding group. Setting + `can_be_null' stops `re_search_2' from using the fastmap, so + that is all we do. */ + case duplicate: + bufp->can_be_null = 1; + return 0; + + + /* Following are the cases which match a character. These end + with `break'. */ + + case exactn: + fastmap[p[1]] = 1; + break; + + + case charset: + for (j = *p++ * BYTEWIDTH - 1; j >= 0; j--) + if (p[j / BYTEWIDTH] & (1 << (j % BYTEWIDTH))) + fastmap[j] = 1; + break; + + + case charset_not: + /* Chars beyond end of map must be allowed. */ + for (j = *p * BYTEWIDTH; j < (1 << BYTEWIDTH); j++) + fastmap[j] = 1; + + for (j = *p++ * BYTEWIDTH - 1; j >= 0; j--) + if (!(p[j / BYTEWIDTH] & (1 << (j % BYTEWIDTH)))) + fastmap[j] = 1; + break; + + + case wordchar: + for (j = 0; j < (1 << BYTEWIDTH); j++) + if (SYNTAX (j) == Sword) + fastmap[j] = 1; + break; + + + case notwordchar: + for (j = 0; j < (1 << BYTEWIDTH); j++) + if (SYNTAX (j) != Sword) + fastmap[j] = 1; + break; + + + case anychar: + { + int fastmap_newline = fastmap['\n']; + + /* `.' matches anything ... */ + for (j = 0; j < (1 << BYTEWIDTH); j++) + fastmap[j] = 1; + + /* ... except perhaps newline. */ + if (!(bufp->syntax & RE_DOT_NEWLINE)) + fastmap['\n'] = fastmap_newline; + + /* Return if we have already set `can_be_null'; if we have, + then the fastmap is irrelevant. Something's wrong here. */ + else if (bufp->can_be_null) + return 0; + + /* Otherwise, have to check alternative paths. */ + break; + } + +#ifdef emacs + case syntaxspec: + k = *p++; + for (j = 0; j < (1 << BYTEWIDTH); j++) + if (SYNTAX (j) == (enum syntaxcode) k) + fastmap[j] = 1; + break; + + + case notsyntaxspec: + k = *p++; + for (j = 0; j < (1 << BYTEWIDTH); j++) + if (SYNTAX (j) != (enum syntaxcode) k) + fastmap[j] = 1; + break; + + + /* All cases after this match the empty string. These end with + `continue'. */ + + + case before_dot: + case at_dot: + case after_dot: + continue; +#endif /* not emacs */ + + + case no_op: + case begline: + case endline: + case begbuf: + case endbuf: + case wordbound: + case notwordbound: + case wordbeg: + case wordend: + case push_dummy_failure: + continue; + + + case jump_n: + case pop_failure_jump: + case maybe_pop_jump: + case jump: + case jump_past_alt: + case dummy_failure_jump: + EXTRACT_NUMBER_AND_INCR (j, p); + p += j; + if (j > 0) + continue; + + /* Jump backward implies we just went through the body of a + loop and matched nothing. Opcode jumped to should be + `on_failure_jump' or `succeed_n'. Just treat it like an + ordinary jump. For a * loop, it has pushed its failure + point already; if so, discard that as redundant. */ + if ((re_opcode_t) *p != on_failure_jump + && (re_opcode_t) *p != succeed_n) + continue; + + p++; + EXTRACT_NUMBER_AND_INCR (j, p); + p += j; + + /* If what's on the stack is where we are now, pop it. */ + if (!FAIL_STACK_EMPTY () + && fail_stack.stack[fail_stack.avail - 1] == p) + fail_stack.avail--; + + continue; + + + case on_failure_jump: + case on_failure_keep_string_jump: + handle_on_failure_jump: + EXTRACT_NUMBER_AND_INCR (j, p); + + /* For some patterns, e.g., `(a?)?', `p+j' here points to the + end of the pattern. We don't want to push such a point, + since when we restore it above, entering the switch will + increment `p' past the end of the pattern. We don't need + to push such a point since we obviously won't find any more + fastmap entries beyond `pend'. Such a pattern can match + the null string, though. */ + if (p + j < pend) + { + if (!PUSH_PATTERN_OP (p + j, fail_stack)) + return -2; + } + else + bufp->can_be_null = 1; + + if (succeed_n_p) + { + EXTRACT_NUMBER_AND_INCR (k, p); /* Skip the n. */ + succeed_n_p = false; + } + + continue; + + + case succeed_n: + /* Get to the number of times to succeed. */ + p += 2; + + /* Increment p past the n for when k != 0. */ + EXTRACT_NUMBER_AND_INCR (k, p); + if (k == 0) + { + p -= 4; + succeed_n_p = true; /* Spaghetti code alert. */ + goto handle_on_failure_jump; + } + continue; + + + case set_number_at: + p += 4; + continue; + + + case start_memory: + case stop_memory: + p += 2; + continue; + + + default: + abort (); /* We have listed all the cases. */ + } /* switch *p++ */ + + /* Getting here means we have found the possible starting + characters for one path of the pattern -- and that the empty + string does not match. We need not follow this path further. + Instead, look at the next alternative (remembered on the + stack), or quit if no more. The test at the top of the loop + does these things. */ + path_can_be_null = false; + p = pend; + } /* while p */ + + /* Set `can_be_null' for the last path (also the first path, if the + pattern is empty). */ + bufp->can_be_null |= path_can_be_null; + return 0; +} /* re_compile_fastmap */ + +/* 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; + unsigned 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 = (regoff_t *) 0; + } +} + +/* Searching routines. */ + +/* Like re_search_2, below, but only one string is specified, and + doesn't let you say where to stop matching. */ + +int +re_search (bufp, string, size, startpos, range, regs) + struct re_pattern_buffer *bufp; + const char *string; + int size, startpos, range; + struct re_registers *regs; +{ + return re_search_2 (bufp, NULL, 0, string, size, startpos, range, + regs, size); +} + + +/* Using the compiled pattern in BUFP->buffer, first tries to match the + virtual concatenation of STRING1 and STRING2, starting first at index + STARTPOS, then at STARTPOS + 1, and so on. + + STRING1 and STRING2 have length SIZE1 and SIZE2, respectively. + + RANGE is how far to scan while trying to match. RANGE = 0 means try + only at STARTPOS; in general, the last start tried is STARTPOS + + RANGE. + + In REGS, return the indices of the virtual concatenation of STRING1 + and STRING2 that matched the entire BUFP->buffer and its contained + subexpressions. + + Do not consider matching one past the index STOP in the virtual + concatenation of STRING1 and STRING2. + + We return either the position in the strings at which the match was + found, -1 if no match, or -2 if error (such as failure + stack overflow). */ + +int +re_search_2 (bufp, string1, size1, string2, size2, startpos, range, regs, stop) + struct re_pattern_buffer *bufp; + const char *string1, *string2; + int size1, size2; + int startpos; + int range; + struct re_registers *regs; + int stop; +{ + int val; + register char *fastmap = bufp->fastmap; + register char *translate = bufp->translate; + int total_size = size1 + size2; + int endpos = startpos + range; + + /* Check for out-of-range STARTPOS. */ + if (startpos < 0 || startpos > total_size) + return -1; + + /* Fix up RANGE if it might eventually take us outside + the virtual concatenation of STRING1 and STRING2. */ + if (endpos < -1) + range = -1 - startpos; + else if (endpos > total_size) + range = total_size - startpos; + + /* If the search isn't to be a backwards one, don't waste time in a + search for a pattern that must be anchored. */ + if (bufp->used > 0 && (re_opcode_t) bufp->buffer[0] == begbuf && range > 0) + { + if (startpos > 0) + return -1; + else + range = 1; + } + + /* Update the fastmap now if not correct already. */ + if (fastmap && !bufp->fastmap_accurate) + if (re_compile_fastmap (bufp) == -2) + return -2; + + /* Loop through the string, looking for a place to start matching. */ + for (;;) + { + /* If a fastmap is supplied, skip quickly over characters that + cannot be the start of a match. If the pattern can match the + null string, however, we don't need to skip characters; we want + the first null string. */ + if (fastmap && startpos < total_size && !bufp->can_be_null) + { + if (range > 0) /* Searching forwards. */ + { + register const char *d; + register int lim = 0; + int irange = range; + + if (startpos < size1 && startpos + range >= size1) + lim = range - (size1 - startpos); + + d = (startpos >= size1 ? string2 - size1 : string1) + startpos; + + /* Written out as an if-else to avoid testing `translate' + inside the loop. */ + if (translate) + while (range > lim + && !fastmap[(unsigned char) + translate[(unsigned char) *d++]]) + range--; + else + while (range > lim && !fastmap[(unsigned char) *d++]) + range--; + + startpos += irange - range; + } + else /* Searching backwards. */ + { + register char c = (size1 == 0 || startpos >= size1 + ? string2[startpos - size1] + : string1[startpos]); + + if (!fastmap[(unsigned char) TRANSLATE (c)]) + goto advance; + } + } + + /* If can't match the null string, and that's all we have left, fail. */ + if (range >= 0 && startpos == total_size && fastmap + && !bufp->can_be_null) + return -1; + + val = re_match_2_internal (bufp, string1, size1, string2, size2, + startpos, regs, stop); +#ifndef REGEX_MALLOC +#ifdef C_ALLOCA + alloca (0); +#endif +#endif + + if (val >= 0) + return startpos; + + if (val == -2) + return -2; + + advance: + if (!range) + break; + else if (range > 0) + { + range--; + startpos++; + } + else + { + range++; + startpos--; + } + } + return -1; +} /* re_search_2 */ + +/* Declarations and macros for re_match_2. */ + +static int bcmp_translate (); +static boolean alt_match_null_string_p (), + common_op_match_null_string_p (), + group_match_null_string_p (); + +/* This converts PTR, a pointer into one of the search strings `string1' + and `string2' into an offset from the beginning of that string. */ +#define POINTER_TO_OFFSET(ptr) \ + (FIRST_STRING_P (ptr) \ + ? ((regoff_t) ((ptr) - string1)) \ + : ((regoff_t) ((ptr) - string2 + size1))) + +/* Macros for dealing with the split strings in re_match_2. */ + +#define MATCHING_IN_FIRST_STRING (dend == end_match_1) + +/* Call before fetching a character with *d. This switches over to + string2 if necessary. */ +#define PREFETCH() \ + while (d == dend) \ + { \ + /* End of string2 => fail. */ \ + if (dend == end_match_2) \ + goto fail; \ + /* End of string1 => advance to string2. */ \ + d = string2; \ + dend = end_match_2; \ + } + + +/* Test if at very beginning or at very end of the virtual concatenation + of `string1' and `string2'. If only one string, it's `string2'. */ +#define AT_STRINGS_BEG(d) ((d) == (size1 ? string1 : string2) || !size2) +#define AT_STRINGS_END(d) ((d) == end2) + + +/* Test if D points to a character which is word-constituent. We have + two special cases to check for: if past the end of string1, look at + the first character in string2; and if before the beginning of + string2, look at the last character in string1. */ +#define WORDCHAR_P(d) \ + (SYNTAX ((d) == end1 ? *string2 \ + : (d) == string2 - 1 ? *(end1 - 1) : *(d)) \ + == Sword) + +/* Test if the character before D and the one at D differ with respect + to being word-constituent. */ +#define AT_WORD_BOUNDARY(d) \ + (AT_STRINGS_BEG (d) || AT_STRINGS_END (d) \ + || WORDCHAR_P (d - 1) != WORDCHAR_P (d)) + + +/* Free everything we malloc. */ +#ifdef MATCH_MAY_ALLOCATE +#ifdef REGEX_MALLOC +#define FREE_VAR(var) if (var) free (var); var = NULL +#define FREE_VARIABLES() \ + do { \ + FREE_VAR (fail_stack.stack); \ + FREE_VAR (regstart); \ + FREE_VAR (regend); \ + FREE_VAR (old_regstart); \ + FREE_VAR (old_regend); \ + FREE_VAR (best_regstart); \ + FREE_VAR (best_regend); \ + FREE_VAR (reg_info); \ + FREE_VAR (reg_dummy); \ + FREE_VAR (reg_info_dummy); \ + } while (0) +#else /* not REGEX_MALLOC */ +/* This used to do alloca (0), but now we do that in the caller. */ +#define FREE_VARIABLES() /* Nothing */ +#endif /* not REGEX_MALLOC */ +#else +#define FREE_VARIABLES() /* Do nothing! */ +#endif /* not MATCH_MAY_ALLOCATE */ + +/* These values must meet several constraints. They must not be valid + register values; since we have a limit of 255 registers (because + we use only one byte in the pattern for the register number), we can + use numbers larger than 255. They must differ by 1, because of + NUM_FAILURE_ITEMS above. And the value for the lowest register must + be larger than the value for the highest register, so we do not try + to actually save any registers when none are active. */ +#define NO_HIGHEST_ACTIVE_REG (1 << BYTEWIDTH) +#define NO_LOWEST_ACTIVE_REG (NO_HIGHEST_ACTIVE_REG + 1) + +/* Matching routines. */ + +#ifndef emacs /* Emacs never uses this. */ +/* re_match is like re_match_2 except it takes only a single string. */ + +int +re_match (bufp, string, size, pos, regs) + struct re_pattern_buffer *bufp; + const char *string; + int size, pos; + struct re_registers *regs; +{ + int result = re_match_2_internal (bufp, NULL, 0, string, size, + pos, regs, size); + alloca (0); + return result; +} +#endif /* not emacs */ + + +/* re_match_2 matches the compiled pattern in BUFP against the + the (virtual) concatenation of STRING1 and STRING2 (of length SIZE1 + and SIZE2, respectively). We start matching at POS, and stop + matching at STOP. + + If REGS is non-null and the `no_sub' field of BUFP is nonzero, we + store offsets for the substring each group matched in REGS. See the + documentation for exactly how many groups we fill. + + We return -1 if no match, -2 if an internal error (such as the + failure stack overflowing). Otherwise, we return the length of the + matched substring. */ + +int +re_match_2 (bufp, string1, size1, string2, size2, pos, regs, stop) + struct re_pattern_buffer *bufp; + const char *string1, *string2; + int size1, size2; + int pos; + struct re_registers *regs; + int stop; +{ + int result = re_match_2_internal (bufp, string1, size1, string2, size2, + pos, regs, stop); + alloca (0); + return result; +} + +/* This is a separate function so that we can force an alloca cleanup + afterwards. */ +static int +re_match_2_internal (bufp, string1, size1, string2, size2, pos, regs, stop) + struct re_pattern_buffer *bufp; + const char *string1, *string2; + int size1, size2; + int pos; + struct re_registers *regs; + int stop; +{ + /* General temporaries. */ + int mcnt; + unsigned char *p1; + + /* Just past the end of the corresponding string. */ + const char *end1, *end2; + + /* Pointers into string1 and string2, just past the last characters in + each to consider matching. */ + const char *end_match_1, *end_match_2; + + /* Where we are in the data, and the end of the current string. */ + const char *d, *dend; + + /* Where we are in the pattern, and the end of the pattern. */ + unsigned char *p = bufp->buffer; + register unsigned char *pend = p + bufp->used; + + /* Mark the opcode just after a start_memory, so we can test for an + empty subpattern when we get to the stop_memory. */ + unsigned char *just_past_start_mem = 0; + + /* We use this to map every character in the string. */ + char *translate = bufp->translate; + + /* Failure point stack. Each place that can handle a failure further + down the line pushes a failure point on this stack. It consists of + restart, regend, and reg_info for all registers corresponding to + the subexpressions we're currently inside, plus the number of such + registers, and, finally, two char *'s. The first char * is where + to resume scanning the pattern; the second one is where to resume + scanning the strings. If the latter is zero, the failure point is + a ``dummy''; if a failure happens and the failure point is a dummy, + it gets discarded and the next next one is tried. */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, this is global. */ + fail_stack_type fail_stack; +#endif +#ifdef DEBUG + static unsigned failure_id = 0; + unsigned nfailure_points_pushed = 0, nfailure_points_popped = 0; +#endif + + /* We fill all the registers internally, independent of what we + return, for use in backreferences. The number here includes + an element for register zero. */ + unsigned num_regs = bufp->re_nsub + 1; + + /* The currently active registers. */ + unsigned lowest_active_reg = NO_LOWEST_ACTIVE_REG; + unsigned highest_active_reg = NO_HIGHEST_ACTIVE_REG; + + /* Information on the contents of registers. These are pointers into + the input strings; they record just what was matched (on this + attempt) by a subexpression part of the pattern, that is, the + regnum-th regstart pointer points to where in the pattern we began + matching and the regnum-th regend points to right after where we + stopped matching the regnum-th subexpression. (The zeroth register + keeps track of what the whole pattern matches.) */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, these are global. */ + const char **regstart, **regend; +#endif + + /* If a group that's operated upon by a repetition operator fails to + match anything, then the register for its start will need to be + restored because it will have been set to wherever in the string we + are when we last see its open-group operator. Similarly for a + register's end. */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, these are global. */ + const char **old_regstart, **old_regend; +#endif + + /* The is_active field of reg_info helps us keep track of which (possibly + nested) subexpressions we are currently in. The matched_something + field of reg_info[reg_num] helps us tell whether or not we have + matched any of the pattern so far this time through the reg_num-th + subexpression. These two fields get reset each time through any + loop their register is in. */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, this is global. */ + register_info_type *reg_info; +#endif + + /* The following record the register info as found in the above + variables when we find a match better than any we've seen before. + This happens as we backtrack through the failure points, which in + turn happens only if we have not yet matched the entire string. */ + unsigned best_regs_set = false; +#ifdef MATCH_MAY_ALLOCATE /* otherwise, these are global. */ + const char **best_regstart, **best_regend; +#endif + + /* Logically, this is `best_regend[0]'. But we don't want to have to + allocate space for that if we're not allocating space for anything + else (see below). Also, we never need info about register 0 for + any of the other register vectors, and it seems rather a kludge to + treat `best_regend' differently than the rest. So we keep track of + the end of the best match so far in a separate variable. We + initialize this to NULL so that when we backtrack the first time + and need to test it, it's not garbage. */ + const char *match_end = NULL; + + /* Used when we pop values we don't care about. */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, these are global. */ + const char **reg_dummy; + register_info_type *reg_info_dummy; +#endif + +#ifdef DEBUG + /* Counts the total number of registers pushed. */ + unsigned num_regs_pushed = 0; +#endif + + DEBUG_PRINT1 ("\n\nEntering re_match_2.\n"); + + INIT_FAIL_STACK (); + +#ifdef MATCH_MAY_ALLOCATE + /* Do not bother to initialize all the register variables if there are + no groups in the pattern, as it takes a fair amount of time. If + there are groups, we include space for register 0 (the whole + pattern), even though we never use it, since it simplifies the + array indexing. We should fix this. */ + if (bufp->re_nsub) + { + regstart = REGEX_TALLOC (num_regs, const char *); + regend = REGEX_TALLOC (num_regs, const char *); + old_regstart = REGEX_TALLOC (num_regs, const char *); + old_regend = REGEX_TALLOC (num_regs, const char *); + best_regstart = REGEX_TALLOC (num_regs, const char *); + best_regend = REGEX_TALLOC (num_regs, const char *); + reg_info = REGEX_TALLOC (num_regs, register_info_type); + reg_dummy = REGEX_TALLOC (num_regs, const char *); + reg_info_dummy = REGEX_TALLOC (num_regs, register_info_type); + + if (!(regstart && regend && old_regstart && old_regend && reg_info + && best_regstart && best_regend && reg_dummy && reg_info_dummy)) + { + FREE_VARIABLES (); + return -2; + } + } +#if defined (REGEX_MALLOC) + else + { + /* We must initialize all our variables to NULL, so that + `FREE_VARIABLES' doesn't try to free them. */ + regstart = regend = old_regstart = old_regend = best_regstart + = best_regend = reg_dummy = NULL; + reg_info = reg_info_dummy = (register_info_type *) NULL; + } +#endif /* REGEX_MALLOC */ +#endif /* MATCH_MAY_ALLOCATE */ + + /* The starting position is bogus. */ + if (pos < 0 || pos > size1 + size2) + { + FREE_VARIABLES (); + return -1; + } + + /* Initialize subexpression text positions to -1 to mark ones that no + start_memory/stop_memory has been seen for. Also initialize the + register information struct. */ + for (mcnt = 1; mcnt < num_regs; mcnt++) + { + regstart[mcnt] = regend[mcnt] + = old_regstart[mcnt] = old_regend[mcnt] = REG_UNSET_VALUE; + + REG_MATCH_NULL_STRING_P (reg_info[mcnt]) = MATCH_NULL_UNSET_VALUE; + IS_ACTIVE (reg_info[mcnt]) = 0; + MATCHED_SOMETHING (reg_info[mcnt]) = 0; + EVER_MATCHED_SOMETHING (reg_info[mcnt]) = 0; + } + + /* We move `string1' into `string2' if the latter's empty -- but not if + `string1' is null. */ + if (size2 == 0 && string1 != NULL) + { + string2 = string1; + size2 = size1; + string1 = 0; + size1 = 0; + } + end1 = string1 + size1; + end2 = string2 + size2; + + /* Compute where to stop matching, within the two strings. */ + if (stop <= size1) + { + end_match_1 = string1 + stop; + end_match_2 = string2; + } + else + { + end_match_1 = end1; + end_match_2 = string2 + stop - size1; + } + + /* `p' scans through the pattern as `d' scans through the data. + `dend' is the end of the input string that `d' points within. `d' + is advanced into the following input string whenever necessary, but + this happens before fetching; therefore, at the beginning of the + loop, `d' can be pointing at the end of a string, but it cannot + equal `string2'. */ + if (size1 > 0 && pos <= size1) + { + d = string1 + pos; + dend = end_match_1; + } + else + { + d = string2 + pos - size1; + dend = end_match_2; + } + + DEBUG_PRINT1 ("The compiled pattern is: "); + DEBUG_PRINT_COMPILED_PATTERN (bufp, p, pend); + DEBUG_PRINT1 ("The string to match is: `"); + DEBUG_PRINT_DOUBLE_STRING (d, string1, size1, string2, size2); + DEBUG_PRINT1 ("'\n"); + + /* This loops over pattern commands. It exits by returning from the + function if the match is complete, or it drops through if the match + fails at this starting point in the input data. */ + for (;;) + { + DEBUG_PRINT2 ("\n0x%x: ", p); + + if (p == pend) + { /* End of pattern means we might have succeeded. */ + DEBUG_PRINT1 ("end of pattern ... "); + + /* If we haven't matched the entire string, and we want the + longest match, try backtracking. */ + if (d != end_match_2) + { + /* 1 if this match ends in the same string (string1 or string2) + as the best previous match. */ + boolean same_str_p = (FIRST_STRING_P (match_end) + == MATCHING_IN_FIRST_STRING); + /* 1 if this match is the best seen so far. */ + boolean best_match_p; + + /* AIX compiler got confused when this was combined + with the previous declaration. */ + if (same_str_p) + best_match_p = d > match_end; + else + best_match_p = !MATCHING_IN_FIRST_STRING; + + DEBUG_PRINT1 ("backtracking.\n"); + + if (!FAIL_STACK_EMPTY ()) + { /* More failure points to try. */ + + /* If exceeds best match so far, save it. */ + if (!best_regs_set || best_match_p) + { + best_regs_set = true; + match_end = d; + + DEBUG_PRINT1 ("\nSAVING match as best so far.\n"); + + for (mcnt = 1; mcnt < num_regs; mcnt++) + { + best_regstart[mcnt] = regstart[mcnt]; + best_regend[mcnt] = regend[mcnt]; + } + } + goto fail; + } + + /* If no failure points, don't restore garbage. And if + last match is real best match, don't restore second + best one. */ + else if (best_regs_set && !best_match_p) + { + restore_best_regs: + /* Restore best match. It may happen that `dend == + end_match_1' while the restored d is in string2. + For example, the pattern `x.*y.*z' against the + strings `x-' and `y-z-', if the two strings are + not consecutive in memory. */ + DEBUG_PRINT1 ("Restoring best registers.\n"); + + d = match_end; + dend = ((d >= string1 && d <= end1) + ? end_match_1 : end_match_2); + + for (mcnt = 1; mcnt < num_regs; mcnt++) + { + regstart[mcnt] = best_regstart[mcnt]; + regend[mcnt] = best_regend[mcnt]; + } + } + } /* d != end_match_2 */ + + DEBUG_PRINT1 ("Accepting match.\n"); + + /* If caller wants register contents data back, do it. */ + if (regs && !bufp->no_sub) + { + /* Have the register data arrays been allocated? */ + if (bufp->regs_allocated == REGS_UNALLOCATED) + { /* No. So allocate them with malloc. We need one + extra element beyond `num_regs' for the `-1' marker + GNU code uses. */ + regs->num_regs = MAX (RE_NREGS, num_regs + 1); + regs->start = TALLOC (regs->num_regs, regoff_t); + regs->end = TALLOC (regs->num_regs, regoff_t); + if (regs->start == NULL || regs->end == NULL) + return -2; + bufp->regs_allocated = REGS_REALLOCATE; + } + else if (bufp->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 (regs->num_regs < num_regs + 1) + { + regs->num_regs = num_regs + 1; + RETALLOC (regs->start, regs->num_regs, regoff_t); + RETALLOC (regs->end, regs->num_regs, regoff_t); + if (regs->start == NULL || regs->end == NULL) + return -2; + } + } + else + { + /* These braces fend off a "empty body in an else-statement" + warning under GCC when assert expands to nothing. */ + assert (bufp->regs_allocated == REGS_FIXED); + } + + /* Convert the pointer data in `regstart' and `regend' to + indices. Register zero has to be set differently, + since we haven't kept track of any info for it. */ + if (regs->num_regs > 0) + { + regs->start[0] = pos; + regs->end[0] = (MATCHING_IN_FIRST_STRING + ? ((regoff_t) (d - string1)) + : ((regoff_t) (d - string2 + size1))); + } + + /* Go through the first `min (num_regs, regs->num_regs)' + registers, since that is all we initialized. */ + for (mcnt = 1; mcnt < MIN (num_regs, regs->num_regs); mcnt++) + { + if (REG_UNSET (regstart[mcnt]) || REG_UNSET (regend[mcnt])) + regs->start[mcnt] = regs->end[mcnt] = -1; + else + { + regs->start[mcnt] + = (regoff_t) POINTER_TO_OFFSET (regstart[mcnt]); + regs->end[mcnt] + = (regoff_t) POINTER_TO_OFFSET (regend[mcnt]); + } + } + + /* If the regs structure we return has more elements than + were in the pattern, set the extra elements to -1. If + we (re)allocated the registers, this is the case, + because we always allocate enough to have at least one + -1 at the end. */ + for (mcnt = num_regs; mcnt < regs->num_regs; mcnt++) + regs->start[mcnt] = regs->end[mcnt] = -1; + } /* regs && !bufp->no_sub */ + + FREE_VARIABLES (); + DEBUG_PRINT4 ("%u failure points pushed, %u popped (%u remain).\n", + nfailure_points_pushed, nfailure_points_popped, + nfailure_points_pushed - nfailure_points_popped); + DEBUG_PRINT2 ("%u registers pushed.\n", num_regs_pushed); + + mcnt = d - pos - (MATCHING_IN_FIRST_STRING + ? string1 + : string2 - size1); + + DEBUG_PRINT2 ("Returning %d from re_match_2.\n", mcnt); + + return mcnt; + } + + /* Otherwise match next pattern command. */ +#ifdef SWITCH_ENUM_BUG + switch ((int) ((re_opcode_t) *p++)) +#else + switch ((re_opcode_t) *p++) +#endif + { + /* Ignore these. Used to ignore the n of succeed_n's which + currently have n == 0. */ + case no_op: + DEBUG_PRINT1 ("EXECUTING no_op.\n"); + break; + + + /* Match the next n pattern characters exactly. The following + byte in the pattern defines n, and the n bytes after that + are the characters to match. */ + case exactn: + mcnt = *p++; + DEBUG_PRINT2 ("EXECUTING exactn %d.\n", mcnt); + + /* This is written out as an if-else so we don't waste time + testing `translate' inside the loop. */ + if (translate) + { + do + { + PREFETCH (); + if (translate[(unsigned char) *d++] != (char) *p++) + goto fail; + } + while (--mcnt); + } + else + { + do + { + PREFETCH (); + if (*d++ != (char) *p++) goto fail; + } + while (--mcnt); + } + SET_REGS_MATCHED (); + break; + + + /* Match any character except possibly a newline or a null. */ + case anychar: + DEBUG_PRINT1 ("EXECUTING anychar.\n"); + + PREFETCH (); + + if ((!(bufp->syntax & RE_DOT_NEWLINE) && TRANSLATE (*d) == '\n') + || (bufp->syntax & RE_DOT_NOT_NULL && TRANSLATE (*d) == '\000')) + goto fail; + + SET_REGS_MATCHED (); + DEBUG_PRINT2 (" Matched `%d'.\n", *d); + d++; + break; + + + case charset: + case charset_not: + { + register unsigned char c; + boolean not = (re_opcode_t) *(p - 1) == charset_not; + + DEBUG_PRINT2 ("EXECUTING charset%s.\n", not ? "_not" : ""); + + PREFETCH (); + c = TRANSLATE (*d); /* The character to match. */ + + /* Cast to `unsigned' instead of `unsigned char' in case the + bit list is a full 32 bytes long. */ + if (c < (unsigned) (*p * BYTEWIDTH) + && p[1 + c / BYTEWIDTH] & (1 << (c % BYTEWIDTH))) + not = !not; + + p += 1 + *p; + + if (!not) goto fail; + + SET_REGS_MATCHED (); + d++; + break; + } + + + /* The beginning of a group is represented by start_memory. + The arguments are the register number in the next byte, and the + number of groups inner to this one in the next. The text + matched within the group is recorded (in the internal + registers data structure) under the register number. */ + case start_memory: + DEBUG_PRINT3 ("EXECUTING start_memory %d (%d):\n", *p, p[1]); + + /* Find out if this group can match the empty string. */ + p1 = p; /* To send to group_match_null_string_p. */ + + if (REG_MATCH_NULL_STRING_P (reg_info[*p]) == MATCH_NULL_UNSET_VALUE) + REG_MATCH_NULL_STRING_P (reg_info[*p]) + = group_match_null_string_p (&p1, pend, reg_info); + + /* Save the position in the string where we were the last time + we were at this open-group operator in case the group is + operated upon by a repetition operator, e.g., with `(a*)*b' + against `ab'; then we want to ignore where we are now in + the string in case this attempt to match fails. */ + old_regstart[*p] = REG_MATCH_NULL_STRING_P (reg_info[*p]) + ? REG_UNSET (regstart[*p]) ? d : regstart[*p] + : regstart[*p]; + DEBUG_PRINT2 (" old_regstart: %d\n", + POINTER_TO_OFFSET (old_regstart[*p])); + + regstart[*p] = d; + DEBUG_PRINT2 (" regstart: %d\n", POINTER_TO_OFFSET (regstart[*p])); + + IS_ACTIVE (reg_info[*p]) = 1; + MATCHED_SOMETHING (reg_info[*p]) = 0; + + /* This is the new highest active register. */ + highest_active_reg = *p; + + /* If nothing was active before, this is the new lowest active + register. */ + if (lowest_active_reg == NO_LOWEST_ACTIVE_REG) + lowest_active_reg = *p; + + /* Move past the register number and inner group count. */ + p += 2; + just_past_start_mem = p; + break; + + + /* The stop_memory opcode represents the end of a group. Its + arguments are the same as start_memory's: the register + number, and the number of inner groups. */ + case stop_memory: + DEBUG_PRINT3 ("EXECUTING stop_memory %d (%d):\n", *p, p[1]); + + /* We need to save the string position the last time we were at + this close-group operator in case the group is operated + upon by a repetition operator, e.g., with `((a*)*(b*)*)*' + against `aba'; then we want to ignore where we are now in + the string in case this attempt to match fails. */ + old_regend[*p] = REG_MATCH_NULL_STRING_P (reg_info[*p]) + ? REG_UNSET (regend[*p]) ? d : regend[*p] + : regend[*p]; + DEBUG_PRINT2 (" old_regend: %d\n", + POINTER_TO_OFFSET (old_regend[*p])); + + regend[*p] = d; + DEBUG_PRINT2 (" regend: %d\n", POINTER_TO_OFFSET (regend[*p])); + + /* This register isn't active anymore. */ + IS_ACTIVE (reg_info[*p]) = 0; + + /* If this was the only register active, nothing is active + anymore. */ + if (lowest_active_reg == highest_active_reg) + { + lowest_active_reg = NO_LOWEST_ACTIVE_REG; + highest_active_reg = NO_HIGHEST_ACTIVE_REG; + } + else + { /* We must scan for the new highest active register, since + it isn't necessarily one less than now: consider + (a(b)c(d(e)f)g). When group 3 ends, after the f), the + new highest active register is 1. */ + unsigned char r = *p - 1; + while (r > 0 && !IS_ACTIVE (reg_info[r])) + r--; + + /* If we end up at register zero, that means that we saved + the registers as the result of an `on_failure_jump', not + a `start_memory', and we jumped to past the innermost + `stop_memory'. For example, in ((.)*) we save + registers 1 and 2 as a result of the *, but when we pop + back to the second ), we are at the stop_memory 1. + Thus, nothing is active. */ + if (r == 0) + { + lowest_active_reg = NO_LOWEST_ACTIVE_REG; + highest_active_reg = NO_HIGHEST_ACTIVE_REG; + } + else + highest_active_reg = r; + } + + /* If just failed to match something this time around with a + group that's operated on by a repetition operator, try to + force exit from the ``loop'', and restore the register + information for this group that we had before trying this + last match. */ + if ((!MATCHED_SOMETHING (reg_info[*p]) + || just_past_start_mem == p - 1) + && (p + 2) < pend) + { + boolean is_a_jump_n = false; + + p1 = p + 2; + mcnt = 0; + switch ((re_opcode_t) *p1++) + { + case jump_n: + is_a_jump_n = true; + case pop_failure_jump: + case maybe_pop_jump: + case jump: + case dummy_failure_jump: + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + if (is_a_jump_n) + p1 += 2; + break; + + default: + /* do nothing */ ; + } + p1 += mcnt; + + /* If the next operation is a jump backwards in the pattern + to an on_failure_jump right before the start_memory + corresponding to this stop_memory, exit from the loop + by forcing a failure after pushing on the stack the + on_failure_jump's jump in the pattern, and d. */ + if (mcnt < 0 && (re_opcode_t) *p1 == on_failure_jump + && (re_opcode_t) p1[3] == start_memory && p1[4] == *p) + { + /* If this group ever matched anything, then restore + what its registers were before trying this last + failed match, e.g., with `(a*)*b' against `ab' for + regstart[1], and, e.g., with `((a*)*(b*)*)*' + against `aba' for regend[3]. + + Also restore the registers for inner groups for, + e.g., `((a*)(b*))*' against `aba' (register 3 would + otherwise get trashed). */ + + if (EVER_MATCHED_SOMETHING (reg_info[*p])) + { + unsigned r; + + EVER_MATCHED_SOMETHING (reg_info[*p]) = 0; + + /* Restore this and inner groups' (if any) registers. */ + for (r = *p; r < *p + *(p + 1); r++) + { + regstart[r] = old_regstart[r]; + + /* xx why this test? */ + if ((int) old_regend[r] >= (int) regstart[r]) + regend[r] = old_regend[r]; + } + } + p1++; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + PUSH_FAILURE_POINT (p1 + mcnt, d, -2); + + goto fail; + } + } + + /* Move past the register number and the inner group count. */ + p += 2; + break; + + + /* \<digit> has been turned into a `duplicate' command which is + followed by the numeric value of <digit> as the register number. */ + case duplicate: + { + register const char *d2, *dend2; + int regno = *p++; /* Get which register to match against. */ + DEBUG_PRINT2 ("EXECUTING duplicate %d.\n", regno); + + /* Can't back reference a group which we've never matched. */ + if (REG_UNSET (regstart[regno]) || REG_UNSET (regend[regno])) + goto fail; + + /* Where in input to try to start matching. */ + d2 = regstart[regno]; + + /* Where to stop matching; if both the place to start and + the place to stop matching are in the same string, then + set to the place to stop, otherwise, for now have to use + the end of the first string. */ + + dend2 = ((FIRST_STRING_P (regstart[regno]) + == FIRST_STRING_P (regend[regno])) + ? regend[regno] : end_match_1); + for (;;) + { + /* If necessary, advance to next segment in register + contents. */ + while (d2 == dend2) + { + if (dend2 == end_match_2) break; + if (dend2 == regend[regno]) break; + + /* End of string1 => advance to string2. */ + d2 = string2; + dend2 = regend[regno]; + } + /* At end of register contents => success */ + if (d2 == dend2) break; + + /* If necessary, advance to next segment in data. */ + PREFETCH (); + + /* How many characters left in this segment to match. */ + mcnt = dend - d; + + /* Want how many consecutive characters we can match in + one shot, so, if necessary, adjust the count. */ + if (mcnt > dend2 - d2) + mcnt = dend2 - d2; + + /* Compare that many; failure if mismatch, else move + past them. */ + if (translate + ? bcmp_translate (d, d2, mcnt, translate) + : bcmp (d, d2, mcnt)) + goto fail; + d += mcnt, d2 += mcnt; + } + } + break; + + + /* begline matches the empty string at the beginning of the string + (unless `not_bol' is set in `bufp'), and, if + `newline_anchor' is set, after newlines. */ + case begline: + DEBUG_PRINT1 ("EXECUTING begline.\n"); + + if (AT_STRINGS_BEG (d)) + { + if (!bufp->not_bol) break; + } + else if (d[-1] == '\n' && bufp->newline_anchor) + { + break; + } + /* In all other cases, we fail. */ + goto fail; + + + /* endline is the dual of begline. */ + case endline: + DEBUG_PRINT1 ("EXECUTING endline.\n"); + + if (AT_STRINGS_END (d)) + { + if (!bufp->not_eol) break; + } + + /* We have to ``prefetch'' the next character. */ + else if ((d == end1 ? *string2 : *d) == '\n' + && bufp->newline_anchor) + { + break; + } + goto fail; + + + /* Match at the very beginning of the data. */ + case begbuf: + DEBUG_PRINT1 ("EXECUTING begbuf.\n"); + if (AT_STRINGS_BEG (d)) + break; + goto fail; + + + /* Match at the very end of the data. */ + case endbuf: + DEBUG_PRINT1 ("EXECUTING endbuf.\n"); + if (AT_STRINGS_END (d)) + break; + goto fail; + + + /* on_failure_keep_string_jump is used to optimize `.*\n'. It + pushes NULL as the value for the string on the stack. Then + `pop_failure_point' will keep the current value for the + string, instead of restoring it. To see why, consider + matching `foo\nbar' against `.*\n'. The .* matches the foo; + then the . fails against the \n. But the next thing we want + to do is match the \n against the \n; if we restored the + string value, we would be back at the foo. + + Because this is used only in specific cases, we don't need to + check all the things that `on_failure_jump' does, to make + sure the right things get saved on the stack. Hence we don't + share its code. The only reason to push anything on the + stack at all is that otherwise we would have to change + `anychar's code to do something besides goto fail in this + case; that seems worse than this. */ + case on_failure_keep_string_jump: + DEBUG_PRINT1 ("EXECUTING on_failure_keep_string_jump"); + + EXTRACT_NUMBER_AND_INCR (mcnt, p); + DEBUG_PRINT3 (" %d (to 0x%x):\n", mcnt, p + mcnt); + + PUSH_FAILURE_POINT (p + mcnt, NULL, -2); + break; + + + /* Uses of on_failure_jump: + + Each alternative starts with an on_failure_jump that points + to the beginning of the next alternative. Each alternative + except the last ends with a jump that in effect jumps past + the rest of the alternatives. (They really jump to the + ending jump of the following alternative, because tensioning + these jumps is a hassle.) + + Repeats start with an on_failure_jump that points past both + the repetition text and either the following jump or + pop_failure_jump back to this on_failure_jump. */ + case on_failure_jump: + on_failure: + DEBUG_PRINT1 ("EXECUTING on_failure_jump"); + + EXTRACT_NUMBER_AND_INCR (mcnt, p); + DEBUG_PRINT3 (" %d (to 0x%x)", mcnt, p + mcnt); + + /* If this on_failure_jump comes right before a group (i.e., + the original * applied to a group), save the information + for that group and all inner ones, so that if we fail back + to this point, the group's information will be correct. + For example, in \(a*\)*\1, we need the preceding group, + and in \(\(a*\)b*\)\2, we need the inner group. */ + + /* We can't use `p' to check ahead because we push + a failure point to `p + mcnt' after we do this. */ + p1 = p; + + /* We need to skip no_op's before we look for the + start_memory in case this on_failure_jump is happening as + the result of a completed succeed_n, as in \(a\)\{1,3\}b\1 + against aba. */ + while (p1 < pend && (re_opcode_t) *p1 == no_op) + p1++; + + if (p1 < pend && (re_opcode_t) *p1 == start_memory) + { + /* We have a new highest active register now. This will + get reset at the start_memory we are about to get to, + but we will have saved all the registers relevant to + this repetition op, as described above. */ + highest_active_reg = *(p1 + 1) + *(p1 + 2); + if (lowest_active_reg == NO_LOWEST_ACTIVE_REG) + lowest_active_reg = *(p1 + 1); + } + + DEBUG_PRINT1 (":\n"); + PUSH_FAILURE_POINT (p + mcnt, d, -2); + break; + + + /* A smart repeat ends with `maybe_pop_jump'. + We change it to either `pop_failure_jump' or `jump'. */ + case maybe_pop_jump: + EXTRACT_NUMBER_AND_INCR (mcnt, p); + DEBUG_PRINT2 ("EXECUTING maybe_pop_jump %d.\n", mcnt); + { + register unsigned char *p2 = p; + + /* Compare the beginning of the repeat with what in the + pattern follows its end. If we can establish that there + is nothing that they would both match, i.e., that we + would have to backtrack because of (as in, e.g., `a*a') + then we can change to pop_failure_jump, because we'll + never have to backtrack. + + This is not true in the case of alternatives: in + `(a|ab)*' we do need to backtrack to the `ab' alternative + (e.g., if the string was `ab'). But instead of trying to + detect that here, the alternative has put on a dummy + failure point which is what we will end up popping. */ + + /* Skip over open/close-group commands. + If what follows this loop is a ...+ construct, + look at what begins its body, since we will have to + match at least one of that. */ + while (1) + { + if (p2 + 2 < pend + && ((re_opcode_t) *p2 == stop_memory + || (re_opcode_t) *p2 == start_memory)) + p2 += 3; + else if (p2 + 6 < pend + && (re_opcode_t) *p2 == dummy_failure_jump) + p2 += 6; + else + break; + } + + p1 = p + mcnt; + /* p1[0] ... p1[2] are the `on_failure_jump' corresponding + to the `maybe_finalize_jump' of this case. Examine what + follows. */ + + /* If we're at the end of the pattern, we can change. */ + if (p2 == pend) + { + /* Consider what happens when matching ":\(.*\)" + against ":/". I don't really understand this code + yet. */ + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 + (" End of pattern: change to `pop_failure_jump'.\n"); + } + + else if ((re_opcode_t) *p2 == exactn + || (bufp->newline_anchor && (re_opcode_t) *p2 == endline)) + { + register unsigned char c + = *p2 == (unsigned char) endline ? '\n' : p2[2]; + + if ((re_opcode_t) p1[3] == exactn && p1[5] != c) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT3 (" %c != %c => pop_failure_jump.\n", + c, p1[5]); + } + + else if ((re_opcode_t) p1[3] == charset + || (re_opcode_t) p1[3] == charset_not) + { + int not = (re_opcode_t) p1[3] == charset_not; + + if (c < (unsigned char) (p1[4] * BYTEWIDTH) + && p1[5 + c / BYTEWIDTH] & (1 << (c % BYTEWIDTH))) + not = !not; + + /* `not' is equal to 1 if c would match, which means + that we can't change to pop_failure_jump. */ + if (!not) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 (" No match => pop_failure_jump.\n"); + } + } + } + else if ((re_opcode_t) *p2 == charset) + { +#ifdef DEBUG + register unsigned char c + = *p2 == (unsigned char) endline ? '\n' : p2[2]; +#endif + + if ((re_opcode_t) p1[3] == exactn + && ! ((int) p2[1] * BYTEWIDTH > (int) p1[4] + && (p2[1 + p1[4] / BYTEWIDTH] + & (1 << (p1[4] % BYTEWIDTH))))) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT3 (" %c != %c => pop_failure_jump.\n", + c, p1[5]); + } + + else if ((re_opcode_t) p1[3] == charset_not) + { + int idx; + /* We win if the charset_not inside the loop + lists every character listed in the charset after. */ + for (idx = 0; idx < (int) p2[1]; idx++) + if (! (p2[2 + idx] == 0 + || (idx < (int) p1[4] + && ((p2[2 + idx] & ~ p1[5 + idx]) == 0)))) + break; + + if (idx == p2[1]) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 (" No match => pop_failure_jump.\n"); + } + } + else if ((re_opcode_t) p1[3] == charset) + { + int idx; + /* We win if the charset inside the loop + has no overlap with the one after the loop. */ + for (idx = 0; + idx < (int) p2[1] && idx < (int) p1[4]; + idx++) + if ((p2[2 + idx] & p1[5 + idx]) != 0) + break; + + if (idx == p2[1] || idx == p1[4]) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 (" No match => pop_failure_jump.\n"); + } + } + } + } + p -= 2; /* Point at relative address again. */ + if ((re_opcode_t) p[-1] != pop_failure_jump) + { + p[-1] = (unsigned char) jump; + DEBUG_PRINT1 (" Match => jump.\n"); + goto unconditional_jump; + } + /* Note fall through. */ + + + /* The end of a simple repeat has a pop_failure_jump back to + its matching on_failure_jump, where the latter will push a + failure point. The pop_failure_jump takes off failure + points put on by this pop_failure_jump's matching + on_failure_jump; we got through the pattern to here from the + matching on_failure_jump, so didn't fail. */ + case pop_failure_jump: + { + /* We need to pass separate storage for the lowest and + highest registers, even though we don't care about the + actual values. Otherwise, we will restore only one + register from the stack, since lowest will == highest in + `pop_failure_point'. */ + unsigned dummy_low_reg, dummy_high_reg; + unsigned char *pdummy; + const char *sdummy; + + DEBUG_PRINT1 ("EXECUTING pop_failure_jump.\n"); + POP_FAILURE_POINT (sdummy, pdummy, + dummy_low_reg, dummy_high_reg, + reg_dummy, reg_dummy, reg_info_dummy); + } + /* Note fall through. */ + + + /* Unconditionally jump (without popping any failure points). */ + case jump: + unconditional_jump: + EXTRACT_NUMBER_AND_INCR (mcnt, p); /* Get the amount to jump. */ + DEBUG_PRINT2 ("EXECUTING jump %d ", mcnt); + p += mcnt; /* Do the jump. */ + DEBUG_PRINT2 ("(to 0x%x).\n", p); + break; + + + /* We need this opcode so we can detect where alternatives end + in `group_match_null_string_p' et al. */ + case jump_past_alt: + DEBUG_PRINT1 ("EXECUTING jump_past_alt.\n"); + goto unconditional_jump; + + + /* Normally, the on_failure_jump pushes a failure point, which + then gets popped at pop_failure_jump. We will end up at + pop_failure_jump, also, and with a pattern of, say, `a+', we + are skipping over the on_failure_jump, so we have to push + something meaningless for pop_failure_jump to pop. */ + case dummy_failure_jump: + DEBUG_PRINT1 ("EXECUTING dummy_failure_jump.\n"); + /* It doesn't matter what we push for the string here. What + the code at `fail' tests is the value for the pattern. */ + PUSH_FAILURE_POINT (0, 0, -2); + goto unconditional_jump; + + + /* At the end of an alternative, we need to push a dummy failure + point in case we are followed by a `pop_failure_jump', because + we don't want the failure point for the alternative to be + popped. For example, matching `(a|ab)*' against `aab' + requires that we match the `ab' alternative. */ + case push_dummy_failure: + DEBUG_PRINT1 ("EXECUTING push_dummy_failure.\n"); + /* See comments just above at `dummy_failure_jump' about the + two zeroes. */ + PUSH_FAILURE_POINT (0, 0, -2); + break; + + /* Have to succeed matching what follows at least n times. + After that, handle like `on_failure_jump'. */ + case succeed_n: + EXTRACT_NUMBER (mcnt, p + 2); + DEBUG_PRINT2 ("EXECUTING succeed_n %d.\n", mcnt); + + assert (mcnt >= 0); + /* Originally, this is how many times we HAVE to succeed. */ + if (mcnt > 0) + { + mcnt--; + p += 2; + STORE_NUMBER_AND_INCR (p, mcnt); + DEBUG_PRINT3 (" Setting 0x%x to %d.\n", p, mcnt); + } + else if (mcnt == 0) + { + DEBUG_PRINT2 (" Setting two bytes from 0x%x to no_op.\n", p+2); + p[2] = (unsigned char) no_op; + p[3] = (unsigned char) no_op; + goto on_failure; + } + break; + + case jump_n: + EXTRACT_NUMBER (mcnt, p + 2); + DEBUG_PRINT2 ("EXECUTING jump_n %d.\n", mcnt); + + /* Originally, this is how many times we CAN jump. */ + if (mcnt) + { + mcnt--; + STORE_NUMBER (p + 2, mcnt); + goto unconditional_jump; + } + /* If don't have to jump any more, skip over the rest of command. */ + else + p += 4; + break; + + case set_number_at: + { + DEBUG_PRINT1 ("EXECUTING set_number_at.\n"); + + EXTRACT_NUMBER_AND_INCR (mcnt, p); + p1 = p + mcnt; + EXTRACT_NUMBER_AND_INCR (mcnt, p); + DEBUG_PRINT3 (" Setting 0x%x to %d.\n", p1, mcnt); + STORE_NUMBER (p1, mcnt); + break; + } + + case wordbound: + DEBUG_PRINT1 ("EXECUTING wordbound.\n"); + if (AT_WORD_BOUNDARY (d)) + break; + goto fail; + + case notwordbound: + DEBUG_PRINT1 ("EXECUTING notwordbound.\n"); + if (AT_WORD_BOUNDARY (d)) + goto fail; + break; + + case wordbeg: + DEBUG_PRINT1 ("EXECUTING wordbeg.\n"); + if (WORDCHAR_P (d) && (AT_STRINGS_BEG (d) || !WORDCHAR_P (d - 1))) + break; + goto fail; + + case wordend: + DEBUG_PRINT1 ("EXECUTING wordend.\n"); + if (!AT_STRINGS_BEG (d) && WORDCHAR_P (d - 1) + && (!WORDCHAR_P (d) || AT_STRINGS_END (d))) + break; + goto fail; + +#ifdef emacs + case before_dot: + DEBUG_PRINT1 ("EXECUTING before_dot.\n"); + if (PTR_CHAR_POS ((unsigned char *) d) >= point) + goto fail; + break; + + case at_dot: + DEBUG_PRINT1 ("EXECUTING at_dot.\n"); + if (PTR_CHAR_POS ((unsigned char *) d) != point) + goto fail; + break; + + case after_dot: + DEBUG_PRINT1 ("EXECUTING after_dot.\n"); + if (PTR_CHAR_POS ((unsigned char *) d) <= point) + goto fail; + break; +#if 0 /* not emacs19 */ + case at_dot: + DEBUG_PRINT1 ("EXECUTING at_dot.\n"); + if (PTR_CHAR_POS ((unsigned char *) d) + 1 != point) + goto fail; + break; +#endif /* not emacs19 */ + + case syntaxspec: + DEBUG_PRINT2 ("EXECUTING syntaxspec %d.\n", mcnt); + mcnt = *p++; + goto matchsyntax; + + case wordchar: + DEBUG_PRINT1 ("EXECUTING Emacs wordchar.\n"); + mcnt = (int) Sword; + matchsyntax: + PREFETCH (); + /* Can't use *d++ here; SYNTAX may be an unsafe macro. */ + d++; + if (SYNTAX (d[-1]) != (enum syntaxcode) mcnt) + goto fail; + SET_REGS_MATCHED (); + break; + + case notsyntaxspec: + DEBUG_PRINT2 ("EXECUTING notsyntaxspec %d.\n", mcnt); + mcnt = *p++; + goto matchnotsyntax; + + case notwordchar: + DEBUG_PRINT1 ("EXECUTING Emacs notwordchar.\n"); + mcnt = (int) Sword; + matchnotsyntax: + PREFETCH (); + /* Can't use *d++ here; SYNTAX may be an unsafe macro. */ + d++; + if (SYNTAX (d[-1]) == (enum syntaxcode) mcnt) + goto fail; + SET_REGS_MATCHED (); + break; + +#else /* not emacs */ + case wordchar: + DEBUG_PRINT1 ("EXECUTING non-Emacs wordchar.\n"); + PREFETCH (); + if (!WORDCHAR_P (d)) + goto fail; + SET_REGS_MATCHED (); + d++; + break; + + case notwordchar: + DEBUG_PRINT1 ("EXECUTING non-Emacs notwordchar.\n"); + PREFETCH (); + if (WORDCHAR_P (d)) + goto fail; + SET_REGS_MATCHED (); + d++; + break; +#endif /* not emacs */ + + default: + abort (); + } + continue; /* Successfully executed one pattern command; keep going. */ + + + /* We goto here if a matching operation fails. */ + fail: + if (!FAIL_STACK_EMPTY ()) + { /* A restart point is known. Restore to that state. */ + DEBUG_PRINT1 ("\nFAIL:\n"); + POP_FAILURE_POINT (d, p, + lowest_active_reg, highest_active_reg, + regstart, regend, reg_info); + + /* If this failure point is a dummy, try the next one. */ + if (!p) + goto fail; + + /* If we failed to the end of the pattern, don't examine *p. */ + assert (p <= pend); + if (p < pend) + { + boolean is_a_jump_n = false; + + /* If failed to a backwards jump that's part of a repetition + loop, need to pop this failure point and use the next one. */ + switch ((re_opcode_t) *p) + { + case jump_n: + is_a_jump_n = true; + case maybe_pop_jump: + case pop_failure_jump: + case jump: + p1 = p + 1; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + p1 += mcnt; + + if ((is_a_jump_n && (re_opcode_t) *p1 == succeed_n) + || (!is_a_jump_n + && (re_opcode_t) *p1 == on_failure_jump)) + goto fail; + break; + default: + /* do nothing */ ; + } + } + + if (d >= string1 && d <= end1) + dend = end_match_1; + } + else + break; /* Matching at this starting point really fails. */ + } /* for (;;) */ + + if (best_regs_set) + goto restore_best_regs; + + FREE_VARIABLES (); + + return -1; /* Failure to match. */ +} /* re_match_2 */ + +/* Subroutine definitions for re_match_2. */ + + +/* We are passed P pointing to a register number after a start_memory. + + Return true if the pattern up to the corresponding stop_memory can + match the empty string, and false otherwise. + + If we find the matching stop_memory, sets P to point to one past its number. + Otherwise, sets P to an undefined byte less than or equal to END. + + We don't handle duplicates properly (yet). */ + +static boolean +group_match_null_string_p (p, end, reg_info) + unsigned char **p, *end; + register_info_type *reg_info; +{ + int mcnt; + /* Point to after the args to the start_memory. */ + unsigned char *p1 = *p + 2; + + while (p1 < end) + { + /* Skip over opcodes that can match nothing, and return true or + false, as appropriate, when we get to one that can't, or to the + matching stop_memory. */ + + switch ((re_opcode_t) *p1) + { + /* Could be either a loop or a series of alternatives. */ + case on_failure_jump: + p1++; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + + /* If the next operation is not a jump backwards in the + pattern. */ + + if (mcnt >= 0) + { + /* Go through the on_failure_jumps of the alternatives, + seeing if any of the alternatives cannot match nothing. + The last alternative starts with only a jump, + whereas the rest start with on_failure_jump and end + with a jump, e.g., here is the pattern for `a|b|c': + + /on_failure_jump/0/6/exactn/1/a/jump_past_alt/0/6 + /on_failure_jump/0/6/exactn/1/b/jump_past_alt/0/3 + /exactn/1/c + + So, we have to first go through the first (n-1) + alternatives and then deal with the last one separately. */ + + + /* Deal with the first (n-1) alternatives, which start + with an on_failure_jump (see above) that jumps to right + past a jump_past_alt. */ + + while ((re_opcode_t) p1[mcnt-3] == jump_past_alt) + { + /* `mcnt' holds how many bytes long the alternative + is, including the ending `jump_past_alt' and + its number. */ + + if (!alt_match_null_string_p (p1, p1 + mcnt - 3, + reg_info)) + return false; + + /* Move to right after this alternative, including the + jump_past_alt. */ + p1 += mcnt; + + /* Break if it's the beginning of an n-th alternative + that doesn't begin with an on_failure_jump. */ + if ((re_opcode_t) *p1 != on_failure_jump) + break; + + /* Still have to check that it's not an n-th + alternative that starts with an on_failure_jump. */ + p1++; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + if ((re_opcode_t) p1[mcnt-3] != jump_past_alt) + { + /* Get to the beginning of the n-th alternative. */ + p1 -= 3; + break; + } + } + + /* Deal with the last alternative: go back and get number + of the `jump_past_alt' just before it. `mcnt' contains + the length of the alternative. */ + EXTRACT_NUMBER (mcnt, p1 - 2); + + if (!alt_match_null_string_p (p1, p1 + mcnt, reg_info)) + return false; + + p1 += mcnt; /* Get past the n-th alternative. */ + } /* if mcnt > 0 */ + break; + + + case stop_memory: + assert (p1[1] == **p); + *p = p1 + 2; + return true; + + + default: + if (!common_op_match_null_string_p (&p1, end, reg_info)) + return false; + } + } /* while p1 < end */ + + return false; +} /* group_match_null_string_p */ + + +/* Similar to group_match_null_string_p, but doesn't deal with alternatives: + It expects P to be the first byte of a single alternative and END one + byte past the last. The alternative can contain groups. */ + +static boolean +alt_match_null_string_p (p, end, reg_info) + unsigned char *p, *end; + register_info_type *reg_info; +{ + int mcnt; + unsigned char *p1 = p; + + while (p1 < end) + { + /* Skip over opcodes that can match nothing, and break when we get + to one that can't. */ + + switch ((re_opcode_t) *p1) + { + /* It's a loop. */ + case on_failure_jump: + p1++; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + p1 += mcnt; + break; + + default: + if (!common_op_match_null_string_p (&p1, end, reg_info)) + return false; + } + } /* while p1 < end */ + + return true; +} /* alt_match_null_string_p */ + + +/* Deals with the ops common to group_match_null_string_p and + alt_match_null_string_p. + + Sets P to one after the op and its arguments, if any. */ + +static boolean +common_op_match_null_string_p (p, end, reg_info) + unsigned char **p, *end; + register_info_type *reg_info; +{ + int mcnt; + boolean ret; + int reg_no; + unsigned char *p1 = *p; + + switch ((re_opcode_t) *p1++) + { + case no_op: + case begline: + case endline: + case begbuf: + case endbuf: + case wordbeg: + case wordend: + case wordbound: + case notwordbound: +#ifdef emacs + case before_dot: + case at_dot: + case after_dot: +#endif + break; + + case start_memory: + reg_no = *p1; + assert (reg_no > 0 && reg_no <= MAX_REGNUM); + ret = group_match_null_string_p (&p1, end, reg_info); + + /* Have to set this here in case we're checking a group which + contains a group and a back reference to it. */ + + if (REG_MATCH_NULL_STRING_P (reg_info[reg_no]) == MATCH_NULL_UNSET_VALUE) + REG_MATCH_NULL_STRING_P (reg_info[reg_no]) = ret; + + if (!ret) + return false; + break; + + /* If this is an optimized succeed_n for zero times, make the jump. */ + case jump: + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + if (mcnt >= 0) + p1 += mcnt; + else + return false; + break; + + case succeed_n: + /* Get to the number of times to succeed. */ + p1 += 2; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + + if (mcnt == 0) + { + p1 -= 4; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + p1 += mcnt; + } + else + return false; + break; + + case duplicate: + if (!REG_MATCH_NULL_STRING_P (reg_info[*p1])) + return false; + break; + + case set_number_at: + p1 += 4; + + default: + /* All other opcodes mean we cannot match the empty string. */ + return false; + } + + *p = p1; + return true; +} /* common_op_match_null_string_p */ + + +/* Return zero if TRANSLATE[S1] and TRANSLATE[S2] are identical for LEN + bytes; nonzero otherwise. */ + +static int +bcmp_translate (s1, s2, len, translate) + unsigned char *s1, *s2; + register int len; + char *translate; +{ + register unsigned char *p1 = s1, *p2 = s2; + while (len) + { + if (translate[*p1++] != translate[*p2++]) return 1; + len--; + } + return 0; +} + +/* Entry points for GNU code. */ + +/* re_compile_pattern is the GNU regular expression compiler: it + compiles PATTERN (of length SIZE) 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. + + We call regex_compile to do the actual compilation. */ + +const char * +re_compile_pattern (pattern, length, bufp) + const char *pattern; + int length; + struct re_pattern_buffer *bufp; +{ + reg_errcode_t ret; + + /* GNU code is written to assume at least RE_NREGS registers will be set + (and at least one extra will be -1). */ + bufp->regs_allocated = REGS_UNALLOCATED; + + /* 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. */ + bufp->no_sub = 0; + + /* Match anchors at newline. */ + bufp->newline_anchor = 1; + + ret = regex_compile (pattern, length, re_syntax_options, bufp); + + return re_error_msg[(int) ret]; +} + +/* Entry points compatible with 4.2 BSD regex library. We don't define + them unless specifically requested. */ + +#ifdef _REGEX_RE_COMP + +/* BSD has one and only one pattern buffer. */ +static struct re_pattern_buffer re_comp_buf; + +char * +re_comp (s) + const char *s; +{ + reg_errcode_t ret; + + if (!s) + { + if (!re_comp_buf.buffer) + return "No previous regular expression"; + return 0; + } + + if (!re_comp_buf.buffer) + { + re_comp_buf.buffer = (unsigned char *) malloc (200); + if (re_comp_buf.buffer == NULL) + return "Memory exhausted"; + re_comp_buf.allocated = 200; + + re_comp_buf.fastmap = (char *) malloc (1 << BYTEWIDTH); + if (re_comp_buf.fastmap == NULL) + return "Memory exhausted"; + } + + /* 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 = regex_compile (s, strlen (s), re_syntax_options, &re_comp_buf); + + /* Yes, we're discarding `const' here. */ + return (char *) re_error_msg[(int) ret]; +} + + +int +re_exec (s) + const char *s; +{ + const int len = strlen (s); + return + 0 <= re_search (&re_comp_buf, s, len, 0, len, (struct re_registers *) 0); +} +#endif /* _REGEX_RE_COMP */ + +/* POSIX.2 functions. Don't define these for Emacs. */ + +#ifndef emacs + +/* 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' and `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 *preg; + const char *pattern; + int cflags; +{ + reg_errcode_t ret; + unsigned syntax + = (cflags & REG_EXTENDED) ? + RE_SYNTAX_POSIX_EXTENDED : RE_SYNTAX_POSIX_BASIC; + + /* regex_compile will allocate the space for the compiled pattern. */ + preg->buffer = 0; + preg->allocated = 0; + preg->used = 0; + + /* Don't bother to use a fastmap when searching. This simplifies the + REG_NEWLINE case: if we used a fastmap, we'd have to put all the + characters after newlines into the fastmap. This way, we just try + every character. */ + preg->fastmap = 0; + + if (cflags & REG_ICASE) + { + unsigned i; + + preg->translate = (char *) malloc (CHAR_SET_SIZE); + if (preg->translate == NULL) + return (int) REG_ESPACE; + + /* Map uppercase characters to corresponding lowercase ones. */ + for (i = 0; i < CHAR_SET_SIZE; i++) + preg->translate[i] = ISUPPER (i) ? tolower (i) : i; + } + else + preg->translate = NULL; + + /* 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); + + /* POSIX says a null character in the pattern terminates it, so we + can use strlen here in compiling the pattern. */ + ret = regex_compile (pattern, strlen (pattern), syntax, preg); + + /* 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; + + return (int) ret; +} + + +/* 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 *preg; + const char *string; + size_t nmatch; + regmatch_t pmatch[]; + int eflags; +{ + int ret; + struct re_registers regs; + regex_t private_preg; + int len = strlen (string); + boolean want_reg_info = !preg->no_sub && nmatch > 0; + + private_preg = *preg; + + private_preg.not_bol = !!(eflags & REG_NOTBOL); + private_preg.not_eol = !!(eflags & REG_NOTEOL); + + /* The user has told us exactly how many registers to return + information about, via `nmatch'. We have to pass that on to the + matching routines. */ + private_preg.regs_allocated = REGS_FIXED; + + if (want_reg_info) + { + regs.num_regs = nmatch; + regs.start = TALLOC (nmatch, regoff_t); + regs.end = TALLOC (nmatch, regoff_t); + if (regs.start == NULL || regs.end == NULL) + return (int) REG_NOMATCH; + } + + /* Perform the searching operation. */ + ret = re_search (&private_preg, string, len, + /* start: */ 0, /* range: */ len, + want_reg_info ? ®s : (struct re_registers *) 0); + + /* Copy the register information to the POSIX structure. */ + if (want_reg_info) + { + if (ret >= 0) + { + unsigned r; + + for (r = 0; r < nmatch; r++) + { + pmatch[r].rm_so = regs.start[r]; + pmatch[r].rm_eo = regs.end[r]; + } + } + + /* If we needed the temporary register info, free the space now. */ + free (regs.start); + free (regs.end); + } + + /* We want zero return to mean success, unlike `re_search'. */ + return ret >= 0 ? (int) REG_NOERROR : (int) REG_NOMATCH; +} + + +/* Returns a message corresponding to an error code, ERRCODE, returned + from either regcomp or regexec. We don't use PREG here. */ + +size_t +regerror (errcode, preg, errbuf, errbuf_size) + int errcode; + const regex_t *preg; + char *errbuf; + size_t errbuf_size; +{ + const char *msg; + size_t msg_size; + + if (errcode < 0 + || errcode >= (sizeof (re_error_msg) / sizeof (re_error_msg[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 = re_error_msg[errcode]; + + /* POSIX doesn't require that we do anything in this case, but why + not be nice. */ + if (! msg) + msg = "Success"; + + msg_size = strlen (msg) + 1; /* Includes the null. */ + + if (errbuf_size != 0) + { + if (msg_size > errbuf_size) + { + strncpy (errbuf, msg, errbuf_size - 1); + errbuf[errbuf_size - 1] = 0; + } + else + strcpy (errbuf, msg); + } + + return msg_size; +} + + +/* Free dynamically allocated space used by PREG. */ + +void +regfree (preg) + regex_t *preg; +{ + if (preg->buffer != NULL) + free (preg->buffer); + preg->buffer = NULL; + + preg->allocated = 0; + preg->used = 0; + + if (preg->fastmap != NULL) + free (preg->fastmap); + preg->fastmap = NULL; + preg->fastmap_accurate = 0; + + if (preg->translate != NULL) + free (preg->translate); + preg->translate = NULL; +} + +#endif /* not emacs */ + +/* +Local variables: +make-backup-files: t +version-control: t +trim-versions-without-asking: nil +End: +*/ diff --git a/lib/regex.h b/lib/regex.h new file mode 100644 index 00000000..55927f62 --- /dev/null +++ b/lib/regex.h @@ -0,0 +1,487 @@ +/* Definitions for data structures and routines for the regular + expression library, version 0.12. + + Copyright (C) 1985, 89, 90, 91, 92, 1993 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +#ifndef __REGEXP_LIBRARY_H__ +#define __REGEXP_LIBRARY_H__ + +/* POSIX says that <sys/types.h> must be included (by the caller) before + <regex.h>. */ + +#ifdef VMS +/* VMS doesn't have `size_t' in <sys/types.h>, even though POSIX says it + should be there. */ +#include <stddef.h> +#endif + + +/* 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 reg_syntax_t; + +/* 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 (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) + +/* 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; + +/* 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_UNMATCHED_RIGHT_PAREN_ORD) + +#define RE_SYNTAX_POSIX_AWK \ + (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS) + +#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) + +/* 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) + +/* 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_UNMATCHED_RIGHT_PAREN_ORD) + +/* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INVALID_OPS + replaces RE_CONTEXT_INDEP_OPS 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]]] */ + +/* Maximum number of duplicates an interval can allow. Some systems + (erroneously) define this in other header files, but we want our + value, so remove any previous define. */ +#ifdef RE_DUP_MAX +#undef RE_DUP_MAX +#endif +#define RE_DUP_MAX ((1 << 15) - 1) + + +/* 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 (REG_EXTENDED << 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 (REG_ICASE << 1) + +/* 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 (REG_NEWLINE << 1) + + +/* 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) + + +/* If any error codes are removed, changed, or added, update the + `re_error_msg' table in regex.c. */ +typedef enum +{ + 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, /* Not implemented. */ + 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; + +/* 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 +{ +/* [[[begin 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 *buffer; + + /* Number of bytes to which `buffer' points. */ + unsigned long allocated; + + /* Number of bytes actually used in `buffer'. */ + unsigned long used; + + /* Syntax setting with which the pattern was compiled. */ + reg_syntax_t 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 *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. */ + char *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 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. */ +#define REGS_UNALLOCATED 0 +#define REGS_REALLOCATE 1 +#define REGS_FIXED 2 + unsigned 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 fastmap_accurate : 1; + + /* If set, `re_match_2' does not return information about + subexpressions. */ + unsigned no_sub : 1; + + /* If set, a beginning-of-line anchor doesn't match at the + beginning of the string. */ + unsigned not_bol : 1; + + /* Similarly for an end-of-line anchor. */ + unsigned not_eol : 1; + + /* If true, an anchor at a newline matches. */ + unsigned newline_anchor : 1; + +/* [[[end pattern_buffer]]] */ +}; + +typedef struct re_pattern_buffer regex_t; + +/* Type for byte offsets within the string. POSIX mandates this. */ +typedef int regoff_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 +{ + unsigned num_regs; + regoff_t *start; + regoff_t *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. */ +#ifndef RE_NREGS +#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. */ + +/* To avoid duplicating every routine declaration -- once with a + prototype (if we are ANSI), and once without (if we aren't) -- we + use the following macro to declare argument types. This + unfortunately clutters up the declarations a bit, but I think it's + worth it. */ + +#if __STDC__ + +#define _RE_ARGS(args) args + +#else /* not __STDC__ */ + +#define _RE_ARGS(args) () + +#endif /* not __STDC__ */ + +/* 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 _RE_ARGS ((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 + _RE_ARGS ((const char *pattern, int 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 _RE_ARGS ((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 int re_search + _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string, + int length, int start, int 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 int re_search_2 + _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1, + int length1, const char *string2, int length2, + int start, int range, struct re_registers *regs, int stop)); + + +/* Like `re_search', but return how many characters in STRING the regexp + in BUFFER matched, starting at position START. */ +extern int re_match + _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string, + int length, int start, struct re_registers *regs)); + + +/* Relates to `re_match' as `re_search_2' relates to `re_search'. */ +extern int re_match_2 + _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1, + int length1, const char *string2, int length2, + int start, struct re_registers *regs, int 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 + _RE_ARGS ((struct re_pattern_buffer *buffer, struct re_registers *regs, + unsigned num_regs, regoff_t *starts, regoff_t *ends)); + +#ifdef _REGEX_RE_COMP +/* 4.2 bsd compatibility. */ +extern char *re_comp _RE_ARGS ((const char *)); +extern int re_exec _RE_ARGS ((const char *)); +#endif + +/* POSIX compatibility. */ +extern int regcomp _RE_ARGS ((regex_t *preg, const char *pattern, int cflags)); +extern int regexec + _RE_ARGS ((const regex_t *preg, const char *string, size_t nmatch, + regmatch_t pmatch[], int eflags)); +extern size_t regerror + _RE_ARGS ((int errcode, const regex_t *preg, char *errbuf, + size_t errbuf_size)); +extern void regfree _RE_ARGS ((regex_t *preg)); + +#endif /* not __REGEXP_LIBRARY_H__ */ + +/* +Local variables: +make-backup-files: t +version-control: t +trim-versions-without-asking: nil +End: +*/ diff --git a/lib/savedir.c b/lib/savedir.c new file mode 100644 index 00000000..1992cf55 --- /dev/null +++ b/lib/savedir.c @@ -0,0 +1,131 @@ +/* savedir.c -- save the list of files in a directory in a string + Copyright (C) 1990 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +/* Written by David MacKenzie <djm@gnu.ai.mit.edu>. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#include <sys/types.h> + +#ifdef HAVE_UNISTD_H +#include <unistd.h> +#endif + +#if HAVE_DIRENT_H +# include <dirent.h> +# define NAMLEN(dirent) strlen((dirent)->d_name) +#else +# define dirent direct +# define NAMLEN(dirent) (dirent)->d_namlen +# if HAVE_SYS_NDIR_H +# include <sys/ndir.h> +# endif +# if HAVE_SYS_DIR_H +# include <sys/dir.h> +# endif +# if HAVE_NDIR_H +# include <ndir.h> +# endif +#endif + +#ifdef CLOSEDIR_VOID +/* Fake a return value. */ +#define CLOSEDIR(d) (closedir (d), 0) +#else +#define CLOSEDIR(d) closedir (d) +#endif + +#ifdef STDC_HEADERS +#include <stdlib.h> +#include <string.h> +#else +char *malloc (); +char *realloc (); +#endif +#ifndef NULL +#define NULL 0 +#endif + +char *stpcpy (); + +/* Return a freshly allocated string containing the filenames + in directory DIR, separated by '\0' characters; + the end is marked by two '\0' characters in a row. + NAME_SIZE is the number of bytes to initially allocate + for the string; it will be enlarged as needed. + Return NULL if DIR cannot be opened or if out of memory. */ + +char * +savedir (dir, name_size) + char *dir; + unsigned name_size; +{ + DIR *dirp; + struct dirent *dp; + char *name_space; + char *namep; + + dirp = opendir (dir); + if (dirp == NULL) + return NULL; + + name_space = (char *) malloc (name_size); + if (name_space == NULL) + { + closedir (dirp); + return NULL; + } + namep = name_space; + + while ((dp = readdir (dirp)) != NULL) + { + /* Skip "." and ".." (some NFS filesystems' directories lack them). */ + if (dp->d_name[0] != '.' + || (dp->d_name[1] != '\0' + && (dp->d_name[1] != '.' || dp->d_name[2] != '\0'))) + { + unsigned size_needed = (namep - name_space) + NAMLEN (dp) + 2; + + if (size_needed > name_size) + { + char *new_name_space; + + while (size_needed > name_size) + name_size += 1024; + + new_name_space = realloc (name_space, name_size); + if (new_name_space == NULL) + { + closedir (dirp); + return NULL; + } + namep += new_name_space - name_space; + name_space = new_name_space; + } + namep = stpcpy (namep, dp->d_name) + 1; + } + } + *namep = '\0'; + if (CLOSEDIR (dirp)) + { + free (name_space); + return NULL; + } + return name_space; +} diff --git a/lib/stpcpy.c b/lib/stpcpy.c new file mode 100644 index 00000000..5ca0a2e3 --- /dev/null +++ b/lib/stpcpy.c @@ -0,0 +1,32 @@ +/* stpcpy.c -- copy a string and return pointer to end of new string + Copyright (C) 1989, 1990 Free Software Foundation. + + 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +/* Copy SRC to DEST, returning the address of the terminating '\0' in DEST. */ + +char * +stpcpy (dest, src) + char *dest; + const char *src; +{ + while ((*dest++ = *src++) != '\0') + /* Do nothing. */ ; + return dest - 1; +} diff --git a/lib/strdup.c b/lib/strdup.c new file mode 100644 index 00000000..1d60f139 --- /dev/null +++ b/lib/strdup.c @@ -0,0 +1,43 @@ +/* strdup.c -- return a newly allocated copy of a string + Copyright (C) 1990 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#ifdef STDC_HEADERS +#include <string.h> +#include <stdlib.h> +#else +char *malloc (); +char *strcpy (); +#endif + +/* Return a newly allocated copy of STR, + or 0 if out of memory. */ + +char * +strdup (str) + const char *str; +{ + char *newstr; + + newstr = (char *) malloc (strlen (str) + 1); + if (newstr) + strcpy (newstr, str); + return newstr; +} diff --git a/lib/strftime.c b/lib/strftime.c new file mode 100644 index 00000000..484852a7 --- /dev/null +++ b/lib/strftime.c @@ -0,0 +1,469 @@ +/* strftime - custom formatting of date and/or time + Copyright (C) 1989, 1991, 1992 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +/* Note: this version of strftime lacks locale support, + but it is standalone. + + Performs `%' substitutions similar to those in printf. Except + where noted, substituted fields have a fixed size; numeric fields are + padded if necessary. Padding is with zeros by default; for fields + that display a single number, padding can be changed or inhibited by + following the `%' with one of the modifiers described below. Unknown + field specifiers are copied as normal characters. All other + characters are copied to the output without change. + + Supports a superset of the ANSI C field specifiers. + + Literal character fields: + % % + n newline + t tab + + Numeric modifiers (a nonstandard extension): + - do not pad the field + _ pad the field with spaces + + Time fields: + %H hour (00..23) + %I hour (01..12) + %k hour ( 0..23) + %l hour ( 1..12) + %M minute (00..59) + %p locale's AM or PM + %r time, 12-hour (hh:mm:ss [AP]M) + %R time, 24-hour (hh:mm) + %s time in seconds since 00:00:00, Jan 1, 1970 (a nonstandard extension) + %S second (00..61) + %T time, 24-hour (hh:mm:ss) + %X locale's time representation (%H:%M:%S) + %Z time zone (EDT), or nothing if no time zone is determinable + + Date fields: + %a locale's abbreviated weekday name (Sun..Sat) + %A locale's full weekday name, variable length (Sunday..Saturday) + %b locale's abbreviated month name (Jan..Dec) + %B locale's full month name, variable length (January..December) + %c locale's date and time (Sat Nov 04 12:02:33 EST 1989) + %C century (00..99) + %d day of month (01..31) + %e day of month ( 1..31) + %D date (mm/dd/yy) + %h same as %b + %j day of year (001..366) + %m month (01..12) + %U week number of year with Sunday as first day of week (00..53) + %w day of week (0..6) + %W week number of year with Monday as first day of week (00..53) + %x locale's date representation (mm/dd/yy) + %y last two digits of year (00..99) + %Y year (1970...) + + David MacKenzie <djm@gnu.ai.mit.edu> */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#include <stdio.h> +#include <sys/types.h> +#if defined(TM_IN_SYS_TIME) || (!defined(HAVE_TM_ZONE) && !defined(HAVE_TZNAME)) +#include <sys/time.h> +#else +#include <time.h> +#endif + +#ifndef STDC_HEADERS +time_t mktime (); +#endif + +#if defined(HAVE_TZNAME) +extern char *tzname[2]; +#endif + +/* Types of padding for numbers in date and time. */ +enum padding +{ + none, blank, zero +}; + +static char const* const days[] = +{ + "Sunday", "Monday", "Tuesday", "Wednesday", "Thursday", "Friday", "Saturday" +}; + +static char const * const months[] = +{ + "January", "February", "March", "April", "May", "June", + "July", "August", "September", "October", "November", "December" +}; + +/* Add character C to STRING and increment LENGTH, + unless LENGTH would exceed MAX. */ + +#define add_char(c) \ + do \ + { \ + if (length + 1 <= max) \ + string[length++] = (c); \ + } \ + while (0) + +/* Add a 2 digit number to STRING, padding if specified. + Return the number of characters added, up to MAX. */ + +static int +add_num2 (string, num, max, pad) + char *string; + int num; + int max; + enum padding pad; +{ + int top = num / 10; + int length = 0; + + if (top == 0 && pad == blank) + add_char (' '); + else if (top != 0 || pad == zero) + add_char (top + '0'); + add_char (num % 10 + '0'); + return length; +} + +/* Add a 3 digit number to STRING, padding if specified. + Return the number of characters added, up to MAX. */ + +static int +add_num3 (string, num, max, pad) + char *string; + int num; + int max; + enum padding pad; +{ + int top = num / 100; + int mid = (num - top * 100) / 10; + int length = 0; + + if (top == 0 && pad == blank) + add_char (' '); + else if (top != 0 || pad == zero) + add_char (top + '0'); + if (mid == 0 && top == 0 && pad == blank) + add_char (' '); + else if (mid != 0 || top != 0 || pad == zero) + add_char (mid + '0'); + add_char (num % 10 + '0'); + return length; +} + +/* Like strncpy except return the number of characters copied. */ + +static int +add_str (to, from, max) + char *to; + const char *from; + int max; +{ + int i; + + for (i = 0; from[i] && i <= max; ++i) + to[i] = from[i]; + return i; +} + +static int +add_num_time_t (string, max, num) + char *string; + int max; + time_t num; +{ + /* This buffer is large enough to hold the character representation + (including the trailing NUL) of any unsigned decimal quantity + whose binary representation fits in 128 bits. */ + char buf[40]; + int length; + + if (sizeof (num) > 16) + abort (); + sprintf (buf, "%lu", (unsigned long) num); + length = add_str (string, buf, max); + return length; +} + +/* Return the week in the year of the time in TM, with the weeks + starting on Sundays. */ + +static int +sun_week (tm) + struct tm *tm; +{ + int dl; + + /* Set `dl' to the day in the year of the last day of the week previous + to the one containing the day specified in TM. If the day specified + in TM is in the first week of the year, `dl' will be negative or 0. + Otherwise, calculate the number of complete weeks before our week + (dl / 7) and add any partial week at the start of the year (dl % 7). */ + dl = tm->tm_yday - tm->tm_wday; + return dl <= 0 ? 0 : dl / 7 + (dl % 7 != 0); +} + +/* Return the week in the year of the time in TM, with the weeks + starting on Mondays. */ + +static int +mon_week (tm) + struct tm *tm; +{ + int dl, wday; + + if (tm->tm_wday == 0) + wday = 6; + else + wday = tm->tm_wday - 1; + dl = tm->tm_yday - wday; + return dl <= 0 ? 0 : dl / 7 + (dl % 7 != 0); +} + +#if !defined(HAVE_TM_ZONE) && !defined(HAVE_TZNAME) +char * +zone_name (tp) + struct tm *tp; +{ + char *timezone (); + struct timeval tv; + struct timezone tz; + + gettimeofday (&tv, &tz); + return timezone (tz.tz_minuteswest, tp->tm_isdst); +} +#endif + +/* Format the time given in TM according to FORMAT, and put the + results in STRING. + Return the number of characters (not including terminating null) + that were put into STRING, or 0 if the length would have + exceeded MAX. */ + +size_t +strftime (string, max, format, tm) + char *string; + size_t max; + const char *format; + const struct tm *tm; +{ + enum padding pad; /* Type of padding to apply. */ + size_t length = 0; /* Characters put in STRING so far. */ + + for (; *format && length < max; ++format) + { + if (*format != '%') + add_char (*format); + else + { + ++format; + /* Modifiers: */ + if (*format == '-') + { + pad = none; + ++format; + } + else if (*format == '_') + { + pad = blank; + ++format; + } + else + pad = zero; + + switch (*format) + { + /* Literal character fields: */ + case 0: + case '%': + add_char ('%'); + break; + case 'n': + add_char ('\n'); + break; + case 't': + add_char ('\t'); + break; + default: + add_char (*format); + break; + + /* Time fields: */ + case 'H': + case 'k': + length += + add_num2 (&string[length], tm->tm_hour, max - length, + *format == 'H' ? pad : blank); + break; + case 'I': + case 'l': + { + int hour12; + + if (tm->tm_hour == 0) + hour12 = 12; + else if (tm->tm_hour > 12) + hour12 = tm->tm_hour - 12; + else + hour12 = tm->tm_hour; + length += + add_num2 (&string[length], hour12, max - length, + *format == 'I' ? pad : blank); + } + break; + case 'M': + length += + add_num2 (&string[length], tm->tm_min, max - length, pad); + break; + case 'p': + if (tm->tm_hour < 12) + add_char ('A'); + else + add_char ('P'); + add_char ('M'); + break; + case 'r': + length += + strftime (&string[length], max - length, "%I:%M:%S %p", tm); + break; + case 'R': + length += + strftime (&string[length], max - length, "%H:%M", tm); + break; + + case 's': + { + struct tm writable_tm; + writable_tm = *tm; + length += add_num_time_t (&string[length], max - length, + mktime (&writable_tm)); + } + break; + + case 'S': + length += + add_num2 (&string[length], tm->tm_sec, max - length, pad); + break; + case 'T': + length += + strftime (&string[length], max - length, "%H:%M:%S", tm); + break; + case 'X': + length += + strftime (&string[length], max - length, "%H:%M:%S", tm); + break; + case 'Z': +#ifdef HAVE_TM_ZONE + length += add_str (&string[length], tm->tm_zone, max - length); +#else +#ifdef HAVE_TZNAME + if (tm->tm_isdst && tzname[1] && *tzname[1]) + length += add_str (&string[length], tzname[1], max - length); + else + length += add_str (&string[length], tzname[0], max - length); +#else + length += add_str (&string[length], zone_name (tm), max - length); +#endif +#endif + break; + + /* Date fields: */ + case 'a': + add_char (days[tm->tm_wday][0]); + add_char (days[tm->tm_wday][1]); + add_char (days[tm->tm_wday][2]); + break; + case 'A': + length += + add_str (&string[length], days[tm->tm_wday], max - length); + break; + case 'b': + case 'h': + add_char (months[tm->tm_mon][0]); + add_char (months[tm->tm_mon][1]); + add_char (months[tm->tm_mon][2]); + break; + case 'B': + length += + add_str (&string[length], months[tm->tm_mon], max - length); + break; + case 'c': + length += + strftime (&string[length], max - length, + "%a %b %d %H:%M:%S %Z %Y", tm); + break; + case 'C': + length += + add_num2 (&string[length], (tm->tm_year + 1900) / 100, + max - length, pad); + break; + case 'd': + length += + add_num2 (&string[length], tm->tm_mday, max - length, pad); + break; + case 'e': + length += + add_num2 (&string[length], tm->tm_mday, max - length, blank); + break; + case 'D': + length += + strftime (&string[length], max - length, "%m/%d/%y", tm); + break; + case 'j': + length += + add_num3 (&string[length], tm->tm_yday + 1, max - length, pad); + break; + case 'm': + length += + add_num2 (&string[length], tm->tm_mon + 1, max - length, pad); + break; + case 'U': + length += + add_num2 (&string[length], sun_week (tm), max - length, pad); + break; + case 'w': + add_char (tm->tm_wday + '0'); + break; + case 'W': + length += + add_num2 (&string[length], mon_week (tm), max - length, pad); + break; + case 'x': + length += + strftime (&string[length], max - length, "%m/%d/%y", tm); + break; + case 'y': + length += + add_num2 (&string[length], tm->tm_year % 100, + max - length, pad); + break; + case 'Y': + add_char ((tm->tm_year + 1900) / 1000 + '0'); + length += + add_num3 (&string[length], + (1900 + tm->tm_year) % 1000, max - length, zero); + break; + } + } + } + add_char (0); + return length - 1; +} diff --git a/lib/strspn.c b/lib/strspn.c index 53f3a1fd..01d8d0fc 100644 --- a/lib/strspn.c +++ b/lib/strspn.c @@ -1,24 +1,25 @@ /* strspn.c -- return numbers of chars at start of string in a class - Copyright (C) 1987, 1990, 2004 Free Software Foundation, Inc. + Copyright (C) 1987, 1990 Free Software Foundation, Inc. - This program is free software: you can redistribute it and/or modify + 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 3 of the License, or - (at your option) any later version. - + 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, see <http://www.gnu.org/licenses/>. -*/ + along with this program; if not, write to the Free Software + Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ +#ifdef HAVE_CONFIG_H #include <config.h> +#endif - -#if defined HAVE_STRING_H +#if defined(HAVE_STRING_H) #include <string.h> #else #include <strings.h> @@ -27,7 +28,6 @@ #endif #endif -#if !defined(HAVE_STRSPN) int strspn (str, class) char *str, *class; @@ -38,4 +38,3 @@ strspn (str, class) ++st; return st - str; } -#endif diff --git a/lib/strstr.c b/lib/strstr.c new file mode 100644 index 00000000..16b748b6 --- /dev/null +++ b/lib/strstr.c @@ -0,0 +1,44 @@ +/* strstr.c -- return the offset of one string within another + Copyright (C) 1989, 1990 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +/* Written by Mike Rendell <michael@cs.mun.ca>. */ + +/* Return the starting address of string S2 in S1; + return 0 if it is not found. */ + +char * +strstr (s1, s2) + char *s1; + char *s2; +{ + int i; + char *p1; + char *p2; + char *s = s1; + + for (p2 = s2, i = 0; *s; p2 = s2, i++, s++) + { + for (p1 = s; *p1 && *p2 && *p1 == *p2; p1++, p2++) + ; + if (!*p2) + break; + } + if (!*p2) + return s1 + i; + + return 0; +} diff --git a/lib/strtol.c b/lib/strtol.c new file mode 100644 index 00000000..08ef0a47 --- /dev/null +++ b/lib/strtol.c @@ -0,0 +1,186 @@ +/* Copyright (C) 1991, 1992 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#include <ctype.h> +#include <errno.h> + +#if HAVE_LIMITS_H +#include <limits.h> +#endif + +#ifndef ULONG_MAX +#define ULONG_MAX ((unsigned long) ~(unsigned long) 0) +#endif + +#ifndef LONG_MAX +#define LONG_MAX (~(1 << (sizeof (long) * 8 - 1))) +#endif + +#ifndef LONG_MIN +#define LONG_MIN (-LONG_MAX - 1) +#endif + +#if STDC_HEADERS +#include <stddef.h> +#include <stdlib.h> +#else +#define NULL 0 +extern int errno; +#endif + +#ifndef UNSIGNED +#define UNSIGNED 0 +#endif + +/* Convert NPTR to an `unsigned long int' or `long int' in base BASE. + If BASE is 0 the base is determined by the presence of a leading + zero, indicating octal or a leading "0x" or "0X", indicating hexadecimal. + If BASE is < 2 or > 36, it is reset to 10. + If ENDPTR is not NULL, a pointer to the character after the last + one converted is stored in *ENDPTR. */ +#if UNSIGNED +unsigned long int +#define strtol strtoul +#else +long int +#endif +strtol (nptr, endptr, base) + const char *nptr; + char **endptr; + int base; +{ + int negative; + register unsigned long int cutoff; + register unsigned int cutlim; + register unsigned long int i; + register const char *s; + register unsigned char c; + const char *save; + int overflow; + + if (base < 0 || base == 1 || base > 36) + base = 10; + + s = nptr; + + /* Skip white space. */ + while (isspace (*s)) + ++s; + if (*s == '\0') + goto noconv; + + /* Check for a sign. */ + if (*s == '-') + { + negative = 1; + ++s; + } + else if (*s == '+') + { + negative = 0; + ++s; + } + else + negative = 0; + + if (base == 16 && s[0] == '0' && toupper (s[1]) == 'X') + s += 2; + + /* If BASE is zero, figure it out ourselves. */ + if (base == 0) + { + if (*s == '0') + { + if (toupper (s[1]) == 'X') + { + s += 2; + base = 16; + } + else + base = 8; + } + else + base = 10; + } + + /* Save the pointer so we can check later if anything happened. */ + save = s; + + cutoff = ULONG_MAX / (unsigned long int) base; + cutlim = ULONG_MAX % (unsigned long int) base; + + overflow = 0; + i = 0; + for (c = *s; c != '\0'; c = *++s) + { + if (isdigit (c)) + c -= '0'; + else if (isalpha (c)) + c = toupper (c) - 'A' + 10; + else + break; + if (c >= base) + break; + /* Check for overflow. */ + if (i > cutoff || (i == cutoff && c > cutlim)) + overflow = 1; + else + { + i *= (unsigned long int) base; + i += c; + } + } + + /* Check if anything actually happened. */ + if (s == save) + goto noconv; + + /* Store in ENDPTR the address of one character + past the last character we converted. */ + if (endptr != NULL) + *endptr = (char *) s; + +#if !UNSIGNED + /* Check for a value that is within the range of + `unsigned long int', but outside the range of `long int'. */ + if (i > (negative ? + -(unsigned long int) LONG_MIN : (unsigned long int) LONG_MAX)) + overflow = 1; +#endif + + if (overflow) + { + errno = ERANGE; +#if UNSIGNED + return ULONG_MAX; +#else + return negative ? LONG_MIN : LONG_MAX; +#endif + } + + /* Return the result of the appropriate sign. */ + return (negative ? -i : i); + +noconv:; + /* There was no number to convert. */ + if (endptr != NULL) + *endptr = (char *) nptr; + return 0L; +} @@ -1,19 +1,20 @@ /* wait.h -- POSIX macros for evaluating exit statuses Copyright (C) 1990 Free Software Foundation, Inc. - This program is free software: you can redistribute it and/or modify + 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 3 of the License, or - (at your option) any later version. - + 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, see <http://www.gnu.org/licenses/>. -*/ + along with this program; if not, write to the Free Software + Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ + #include <sys/types.h> /* For pid_t. */ #if HAVE_SYS_WAIT_H #include <sys/wait.h> diff --git a/lib/xgetcwd.c b/lib/xgetcwd.c new file mode 100644 index 00000000..1c1a7bd7 --- /dev/null +++ b/lib/xgetcwd.c @@ -0,0 +1,78 @@ +/* xgetcwd.c -- return current directory with unlimited length + Copyright (C) 1992 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +/* Written by David MacKenzie <djm@gnu.ai.mit.edu>. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#include <stdio.h> +#include <errno.h> +#ifndef errno +extern int errno; +#endif +#include <sys/types.h> +#include "pathmax.h" + +#ifndef HAVE_GETCWD +char *getwd (); +#define getcwd(buf, max) getwd (buf) +#else +char *getcwd (); +#endif + +/* Amount to increase buffer size by in each try. */ +#define PATH_INCR 32 + +char *xmalloc (); +char *xrealloc (); +void free (); + +/* Return the current directory, newly allocated, arbitrarily long. + Return NULL and set errno on error. */ + +char * +xgetcwd () +{ + char *cwd; + char *ret; + unsigned path_max; + + errno = 0; + path_max = (unsigned) PATH_MAX; + path_max += 2; /* The getcwd docs say to do this. */ + + cwd = xmalloc (path_max); + + errno = 0; + while ((ret = getcwd (cwd, path_max)) == NULL && errno == ERANGE) + { + path_max += PATH_INCR; + cwd = xrealloc (cwd, path_max); + errno = 0; + } + + if (ret == NULL) + { + int save_errno = errno; + free (cwd); + errno = save_errno; + return NULL; + } + return cwd; +} diff --git a/lib/xmalloc.c b/lib/xmalloc.c new file mode 100644 index 00000000..9f701111 --- /dev/null +++ b/lib/xmalloc.c @@ -0,0 +1,95 @@ +/* xmalloc.c -- malloc with out of memory checking + Copyright (C) 1990, 91, 92, 93, 94 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#if __STDC__ +#define VOID void +#else +#define VOID char +#endif + +#include <sys/types.h> + +#if STDC_HEADERS +#include <stdlib.h> +#else +VOID *malloc (); +VOID *realloc (); +void free (); +#endif + +#ifndef EXIT_FAILURE +#define EXIT_FAILURE 1 +#endif + +/* Exit value when the requested amount of memory is not available. + The caller may set it to some other value. */ +int xmalloc_exit_failure = EXIT_FAILURE; + +#if __STDC__ && (HAVE_VPRINTF || HAVE_DOPRNT) +void error (int, int, const char *, ...); +#else +void error (); +#endif + +static VOID * +fixup_null_alloc (n) + size_t n; +{ + VOID *p; + + p = 0; + if (n == 0) + p = malloc ((size_t) 1); + if (p == 0) + error (xmalloc_exit_failure, 0, "memory exhausted"); + return p; +} + +/* Allocate N bytes of memory dynamically, with error checking. */ + +VOID * +xmalloc (n) + size_t n; +{ + VOID *p; + + p = malloc (n); + if (p == 0) + p = fixup_null_alloc (n); + return p; +} + +/* Change the size of an allocated block of memory P to N bytes, + with error checking. + If P is NULL, run xmalloc. */ + +VOID * +xrealloc (p, n) + VOID *p; + size_t n; +{ + if (p == 0) + return xmalloc (n); + p = realloc (p, n); + if (p == 0) + p = fixup_null_alloc (n); + return p; +} diff --git a/lib/xstrdup.c b/lib/xstrdup.c new file mode 100644 index 00000000..27cd0c67 --- /dev/null +++ b/lib/xstrdup.c @@ -0,0 +1,36 @@ +/* xstrdup.c -- copy a string with out of memory checking + Copyright (C) 1990 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., 675 Mass Ave, Cambridge, MA 02139, USA. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#if defined(STDC_HEADERS) || defined(HAVE_STRING_H) +#include <string.h> +#else +#include <strings.h> +#endif +char *xmalloc (); + +/* Return a newly allocated copy of STRING. */ + +char * +xstrdup (string) + char *string; +{ + return strcpy (xmalloc (strlen (string) + 1), string); +} |