diff options
author | Alain Magloire <alainm@rcsm.ee.mcgill.ca> | 2001-03-04 05:33:12 +0000 |
---|---|---|
committer | Alain Magloire <alainm@rcsm.ee.mcgill.ca> | 2001-03-04 05:33:12 +0000 |
commit | 832757b094adc56acf2dd728de1f2459228d1fdf (patch) | |
tree | 8d317088bf2b472fffa2d75efd9db69304b17b88 /lib | |
parent | a1d4c778d77635386fcf96264a3d654b7d6beaf6 (diff) | |
download | grep-832757b094adc56acf2dd728de1f2459228d1fdf.tar.gz |
from GNU tar and fetish.
Diffstat (limited to 'lib')
-rw-r--r-- | lib/Makefile.am | 20 | ||||
-rw-r--r-- | lib/alloca.c | 503 | ||||
-rw-r--r-- | lib/atexit.c | 14 | ||||
-rw-r--r-- | lib/closeout.c | 121 | ||||
-rw-r--r-- | lib/closeout.h | 17 | ||||
-rw-r--r-- | lib/error.c | 276 | ||||
-rw-r--r-- | lib/error.h | 78 | ||||
-rw-r--r-- | lib/exclude.c | 128 | ||||
-rw-r--r-- | lib/exclude.h | 35 | ||||
-rw-r--r-- | lib/fnmatch.c | 230 | ||||
-rw-r--r-- | lib/fnmatch.h | 69 | ||||
-rw-r--r-- | lib/getopt.c | 1049 | ||||
-rw-r--r-- | lib/getopt.h | 169 | ||||
-rw-r--r-- | lib/hard-locale.c | 85 | ||||
-rw-r--r-- | lib/hard-locale.h | 18 | ||||
-rw-r--r-- | lib/isdir.c | 42 | ||||
-rw-r--r-- | lib/memchr.c | 199 | ||||
-rw-r--r-- | lib/obstack.c | 598 | ||||
-rw-r--r-- | lib/obstack.h | 593 | ||||
-rw-r--r-- | lib/posix/Makefile.am | 4 | ||||
-rw-r--r-- | lib/posix/regex.h | 542 | ||||
-rw-r--r-- | lib/quotearg.c | 619 | ||||
-rw-r--r-- | lib/quotearg.h | 110 | ||||
-rw-r--r-- | lib/regex.c | 5885 | ||||
-rw-r--r-- | lib/regex.h | 46 | ||||
-rw-r--r-- | lib/savedir.c | 179 | ||||
-rw-r--r-- | lib/savedir.h | 18 | ||||
-rw-r--r-- | lib/stpcpy.c | 50 | ||||
-rw-r--r-- | lib/strtol.c | 472 | ||||
-rw-r--r-- | lib/strtoul.c | 22 | ||||
-rw-r--r-- | lib/strtoull.c | 27 | ||||
-rw-r--r-- | lib/strtoumax.c | 81 | ||||
-rw-r--r-- | lib/xalloc.h | 87 | ||||
-rw-r--r-- | lib/xmalloc.c | 116 | ||||
-rw-r--r-- | lib/xstrtol.c | 282 | ||||
-rw-r--r-- | lib/xstrtol.h | 64 | ||||
-rw-r--r-- | lib/xstrtoumax.c | 31 |
37 files changed, 12879 insertions, 0 deletions
diff --git a/lib/Makefile.am b/lib/Makefile.am new file mode 100644 index 00000000..c7f08ec5 --- /dev/null +++ b/lib/Makefile.am @@ -0,0 +1,20 @@ +# +AUTOMAKE_OPTIONS = ../src/ansi2knr + +SUBDIRS = posix + +noinst_LIBRARIES = libgreputils.a + +INCLUDES = -I.. -I$(srcdir) -I../intl +DEFS = -DLIBDIR=\"$(libdir)\" @DEFS@ + +libgreputils_a_SOURCES = \ + closeout.c error.c exclude.c getopt.c getopt1.c hard-locale.c \ + isdir.c obstack.c quotearg.c savedir.c xmalloc.c xstrtol.c xstrtoumax.c + +libgreputils_a_LIBADD = @LIBOBJS@ @ALLOCA@ +libgreputils_a_DEPENDENCIES = $(libgreputils_a_LIBADD) +noinst_HEADERS = \ + closeout.h error.h exclude.h fnmatch.h getopt.h \ + hard-locale.h obstack.h quotearg.h regex.h savedir.h xalloc.h xstrtol.h + diff --git a/lib/alloca.c b/lib/alloca.c new file mode 100644 index 00000000..55a80677 --- /dev/null +++ b/lib/alloca.c @@ -0,0 +1,503 @@ +/* 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 + +#if HAVE_STRING_H +# include <string.h> +#endif +#if HAVE_STDLIB_H +# include <stdlib.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 + +# ifndef NULL +# define NULL 0 +# endif + +/* 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 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 (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. */ + + if (new == 0) + abort(); + + ((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/atexit.c b/lib/atexit.c new file mode 100644 index 00000000..85e992f7 --- /dev/null +++ b/lib/atexit.c @@ -0,0 +1,14 @@ +/* Wrapper to implement ANSI C's atexit using SunOS's on_exit. */ +/* This function is in the public domain. --Mike Stump. */ + +#include "config.h" + +int +atexit(f) + void (*f)(); +{ + /* If the system doesn't provide a definition for atexit, use on_exit + if the system provides that. */ + on_exit (f, 0); + return 0; +} diff --git a/lib/closeout.c b/lib/closeout.c new file mode 100644 index 00000000..4fdac32b --- /dev/null +++ b/lib/closeout.c @@ -0,0 +1,121 @@ +/* closeout.c - close standard output + Copyright (C) 1998, 1999, 2000, 2001 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include <config.h> +#endif + +#if ENABLE_NLS +# include <libintl.h> +# define _(Text) gettext (Text) +#else +# define _(Text) Text +#endif + +#if HAVE_STDLIB_H +# include <stdlib.h> +#endif +#ifndef EXIT_FAILURE +# define EXIT_FAILURE 1 +#endif + +#include <stdio.h> + +#include <errno.h> +#ifndef errno +extern int errno; +#endif + +#include "closeout.h" +#include "error.h" +#include "quotearg.h" +#if 0 +#include "__fpending.h" +#endif + +static int default_exit_status = EXIT_FAILURE; +static const char *file_name; + +/* Set the value to be used for the exit status when close_stdout is called. + This is useful when it is not convenient to call close_stdout_status, + e.g., when close_stdout is called via atexit. */ +void +close_stdout_set_status (int status) +{ + default_exit_status = status; +} + +/* Set the file name to be reported in the event an error is detected + by close_stdout_status. */ +void +close_stdout_set_file_name (const char *file) +{ + file_name = file; +} + +/* Close standard output, exiting with status STATUS on failure. + If a program writes *anything* to stdout, that program should `fflush' + stdout and make sure that it succeeds before exiting. Otherwise, + suppose that you go to the extreme of checking the return status + of every function that does an explicit write to stdout. The last + printf can succeed in writing to the internal stream buffer, and yet + the fclose(stdout) could still fail (due e.g., to a disk full error) + when it tries to write out that buffered data. Thus, you would be + left with an incomplete output file and the offending program would + exit successfully. + + FIXME: note the fflush suggested above is implicit in the fclose + we actually do below. Consider doing only the fflush and/or using + setvbuf to inhibit buffering. + + Besides, it's wasteful to check the return value from every call + that writes to stdout -- just let the internal stream state record + the failure. That's what the ferror test is checking below. + + It's important to detect such failures and exit nonzero because many + tools (most notably `make' and other build-management systems) depend + on being able to detect failure in other tools via their exit status. */ + +void +close_stdout_status (int status) +{ + int e = ferror (stdout) ? 0 : -1; + +#if 0 + if (__fpending (stdout) == 0) + return; +#endif + + if (fclose (stdout) != 0) + e = errno; + + if (0 < e) + { + char const *write_error = _("write error"); + if (file_name) + error (status, e, "%s: %s", quotearg_colon (file_name), write_error); + else + error (status, e, "%s", write_error); + } +} + +/* Close standard output, exiting with status EXIT_FAILURE on failure. */ +void +close_stdout (void) +{ + close_stdout_status (default_exit_status); +} diff --git a/lib/closeout.h b/lib/closeout.h new file mode 100644 index 00000000..80f24e45 --- /dev/null +++ b/lib/closeout.h @@ -0,0 +1,17 @@ +#ifndef CLOSEOUT_H +# define CLOSEOUT_H 1 + +# ifndef PARAMS +# if defined PROTOTYPES || (defined __STDC__ && __STDC__) +# define PARAMS(Args) Args +# else +# define PARAMS(Args) () +# endif +# endif + +void close_stdout_set_status PARAMS ((int status)); +void close_stdout_set_file_name PARAMS ((const char *file)); +void close_stdout PARAMS ((void)); +void close_stdout_status PARAMS ((int status)); + +#endif diff --git a/lib/error.c b/lib/error.c new file mode 100644 index 00000000..c78b6cff --- /dev/null +++ b/lib/error.c @@ -0,0 +1,276 @@ +/* Error handler for noninteractive utilities + Copyright (C) 1990-1998, 2000 Free Software Foundation, Inc. + + This file is part of the GNU C Library. Its master source is NOT part of + the C library, however. The master source lives in /gd/gnu/lib. + + 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., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +/* Written by David MacKenzie <djm@gnu.ai.mit.edu>. */ + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +#include <stdio.h> +#if HAVE_LIBINTL_H +# include <libintl.h> +#endif + +#if HAVE_VPRINTF || HAVE_DOPRNT || _LIBC +# 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 || _LIBC +# include <stdlib.h> +# include <string.h> +#else +void exit (); +#endif + +#include "error.h" + +#ifndef HAVE_DECL_STRERROR_R +"this configure-time declaration test was not run" +#endif +#if !HAVE_DECL_STRERROR_R +char *strerror_r (); +#endif + +#ifndef _ +# define _(String) String +#endif + +/* If NULL, error will flush stdout, then print on stderr the program + name, a colon and a space. Otherwise, error will call this + function without parameters instead. */ +void (*error_print_progname) ( +#if __STDC__ - 0 + void +#endif + ); + +/* This variable is incremented each time `error' is called. */ +unsigned int error_message_count; + +#ifdef _LIBC +/* In the GNU C library, there is a predefined variable for this. */ + +# define program_name program_invocation_name +# include <errno.h> + +/* In GNU libc we want do not want to use the common name `error' directly. + Instead make it a weak alias. */ +# define error __error +# define error_at_line __error_at_line + +# ifdef USE_IN_LIBIO +# include <libio/iolibio.h> +# define fflush(s) _IO_fflush (s) +# endif + +#else /* not _LIBC */ + +/* The calling program should define program_name and set it to the + name of the executing program. */ +extern char *program_name; + +# ifdef HAVE_STRERROR_R +# define __strerror_r strerror_r +# else +# if HAVE_STRERROR +# ifndef strerror /* On some systems, strerror is a macro */ +char *strerror (); +# endif +# 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 /* HAVE_STRERROR */ +# endif /* HAVE_STRERROR_R */ +#endif /* not _LIBC */ + +/* 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 || _LIBC + 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 + + ++error_message_count; + if (errnum) + { +#if defined HAVE_STRERROR_R || _LIBC + char errbuf[1024]; +# if HAVE_WORKING_STRERROR_R || _LIBC + fprintf (stderr, ": %s", __strerror_r (errnum, errbuf, sizeof errbuf)); +# else + /* Don't use __strerror_r's return value because on some systems + (at least DEC UNIX 4.0[A-D]) strerror_r returns `int'. */ + __strerror_r (errnum, errbuf, sizeof errbuf); + fprintf (stderr, ": %s", errbuf); +# endif +#else + fprintf (stderr, ": %s", strerror (errnum)); +#endif + } + putc ('\n', stderr); + fflush (stderr); + if (status) + exit (status); +} + +/* Sometimes we want to have at most one error per line. This + variable controls whether this mode is selected or not. */ +int error_one_per_line; + +void +#if defined VA_START && __STDC__ +error_at_line (int status, int errnum, const char *file_name, + unsigned int line_number, const char *message, ...) +#else +error_at_line (status, errnum, file_name, line_number, message, va_alist) + int status; + int errnum; + const char *file_name; + unsigned int line_number; + char *message; + va_dcl +#endif +{ +#ifdef VA_START + va_list args; +#endif + + if (error_one_per_line) + { + static const char *old_file_name; + static unsigned int old_line_number; + + if (old_line_number == line_number && + (file_name == old_file_name || !strcmp (old_file_name, file_name))) + /* Simply return and print nothing. */ + return; + + old_file_name = file_name; + old_line_number = line_number; + } + + if (error_print_progname) + (*error_print_progname) (); + else + { + fflush (stdout); + fprintf (stderr, "%s:", program_name); + } + + if (file_name != NULL) + fprintf (stderr, "%s:%d: ", file_name, line_number); + +#ifdef VA_START + VA_START (args, message); +# if HAVE_VPRINTF || _LIBC + 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 + + ++error_message_count; + if (errnum) + { +#if defined HAVE_STRERROR_R || _LIBC + char errbuf[1024]; +# if HAVE_WORKING_STRERROR_R || _LIBC + fprintf (stderr, ": %s", __strerror_r (errnum, errbuf, sizeof errbuf)); +# else + /* Don't use __strerror_r's return value because on some systems + (at least DEC UNIX 4.0[A-D]) strerror_r returns `int'. */ + __strerror_r (errnum, errbuf, sizeof errbuf); + fprintf (stderr, ": %s", errbuf); +# endif +#else + fprintf (stderr, ": %s", strerror (errnum)); +#endif + } + putc ('\n', stderr); + fflush (stderr); + if (status) + exit (status); +} + +#ifdef _LIBC +/* Make the weak alias. */ +# undef error +# undef error_at_line +weak_alias (__error, error) +weak_alias (__error_at_line, error_at_line) +#endif diff --git a/lib/error.h b/lib/error.h new file mode 100644 index 00000000..20f75824 --- /dev/null +++ b/lib/error.h @@ -0,0 +1,78 @@ +/* Declaration for error-reporting function + Copyright (C) 1995, 1996, 1997 Free Software Foundation, Inc. + + + NOTE: The canonical source of this file is maintained with the GNU C Library. + Bugs can be reported to bug-glibc@prep.ai.mit.edu. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, + USA. */ + +#ifndef _ERROR_H +#define _ERROR_H 1 + +#ifndef __attribute__ +/* This feature is available in gcc versions 2.5 and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 5) || __STRICT_ANSI__ +# define __attribute__(Spec) /* empty */ +# endif +/* The __-protected variants of `format' and `printf' attributes + are accepted by gcc versions 2.6.4 (effectively 2.7) and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 7) +# define __format__ format +# define __printf__ printf +# endif +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +#if defined (__STDC__) && __STDC__ + +/* Print a message with `fprintf (stderr, FORMAT, ...)'; + if ERRNUM is nonzero, follow it with ": " and strerror (ERRNUM). + If STATUS is nonzero, terminate the program with `exit (STATUS)'. */ + +extern void error (int status, int errnum, const char *format, ...) + __attribute__ ((__format__ (__printf__, 3, 4))); + +extern void error_at_line (int status, int errnum, const char *fname, + unsigned int lineno, const char *format, ...) + __attribute__ ((__format__ (__printf__, 5, 6))); + +/* If NULL, error will flush stdout, then print on stderr the program + name, a colon and a space. Otherwise, error will call this + function without parameters instead. */ +extern void (*error_print_progname) (void); + +#else +void error (); +void error_at_line (); +extern void (*error_print_progname) (); +#endif + +/* This variable is incremented each time `error' is called. */ +extern unsigned int error_message_count; + +/* Sometimes we want to have at most one error per line. This + variable controls whether this mode is selected or not. */ +extern int error_one_per_line; + +#ifdef __cplusplus +} +#endif + +#endif /* error.h */ diff --git a/lib/exclude.c b/lib/exclude.c new file mode 100644 index 00000000..1925a398 --- /dev/null +++ b/lib/exclude.c @@ -0,0 +1,128 @@ +/* exclude.c -- exclude file names + Copyright 1992, 1993, 1994, 1997, 1999, 2000 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert <eggert@twinsun.com> */ + +#if HAVE_CONFIG_H +# include <config.h> +#endif + +#include <errno.h> +#ifndef errno +extern int errno; +#endif +#include <exclude.h> +#include <fnmatch.h> +#include <stdio.h> +#include <sys/types.h> + +void *xmalloc PARAMS ((size_t)); +void *xrealloc PARAMS ((void *, size_t)); + +/* Keep track of excluded file name patterns. */ + +struct exclude + { + char const **exclude; + int exclude_alloc; + int exclude_count; + }; + +struct exclude * +new_exclude (void) +{ + struct exclude *ex = (struct exclude *) xmalloc (sizeof (struct exclude)); + ex->exclude_count = 0; + ex->exclude_alloc = 64; + ex->exclude = (char const **) xmalloc (ex->exclude_alloc * sizeof (char *)); + return ex; +} + +int +excluded_filename (struct exclude const *ex, char const *f, int options) +{ + char const * const *exclude = ex->exclude; + int exclude_count = ex->exclude_count; + int i; + + for (i = 0; i < exclude_count; i++) + if (fnmatch (exclude[i], f, options) == 0) + return 1; + + return 0; +} + +void +add_exclude (struct exclude *ex, char const *pattern) +{ + if (ex->exclude_alloc <= ex->exclude_count) + ex->exclude = (char const **) xrealloc (ex->exclude, + ((ex->exclude_alloc *= 2) + * sizeof (char *))); + + ex->exclude[ex->exclude_count++] = pattern; +} + +int +add_exclude_file (void (*add_func) PARAMS ((struct exclude *, char const *)), + struct exclude *ex, char const *filename, char line_end) +{ + int use_stdin = filename[0] == '-' && !filename[1]; + FILE *in; + char *buf; + char *p; + char const *pattern; + char const *lim; + size_t buf_alloc = 1024; + size_t buf_count = 0; + int c; + int e = 0; + + if (use_stdin) + in = stdin; + else if (! (in = fopen (filename, "r"))) + return -1; + + buf = xmalloc (buf_alloc); + + while ((c = getc (in)) != EOF) + { + buf[buf_count++] = c; + if (buf_count == buf_alloc) + buf = xrealloc (buf, buf_alloc *= 2); + } + + buf = xrealloc (buf, buf_count + 1); + + if (ferror (in)) + e = errno; + + if (!use_stdin && fclose (in) != 0) + e = errno; + + for (pattern = p = buf, lim = buf + buf_count; p <= lim; p++) + if (p < lim ? *p == line_end : buf < p && p[-1]) + { + *p = '\0'; + (*add_func) (ex, pattern); + pattern = p + 1; + } + + errno = e; + return e ? -1 : 0; +} diff --git a/lib/exclude.h b/lib/exclude.h new file mode 100644 index 00000000..8a48a00e --- /dev/null +++ b/lib/exclude.h @@ -0,0 +1,35 @@ +/* exclude.h -- declarations for excluding file names + Copyright 1992, 1993, 1994, 1997, 1999 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert <eggert@twinsun.com> */ + +#ifndef PARAMS +# if defined PROTOTYPES || (defined __STDC__ && __STDC__) +# define PARAMS(Args) Args +# else +# define PARAMS(Args) () +# endif +#endif + +struct exclude; + +struct exclude *new_exclude PARAMS ((void)); +void add_exclude PARAMS ((struct exclude *, char const *)); +int add_exclude_file PARAMS ((void (*) (struct exclude *, char const *), + struct exclude *, char const *, char)); +int excluded_filename PARAMS ((struct exclude const *, char const *, int)); diff --git a/lib/fnmatch.c b/lib/fnmatch.c new file mode 100644 index 00000000..9bff8c22 --- /dev/null +++ b/lib/fnmatch.c @@ -0,0 +1,230 @@ +/* Copyright 1991, 1992, 1993, 1996, 1997, 2000 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include <config.h> +#endif + +/* Enable GNU extensions in fnmatch.h. */ +#ifndef _GNU_SOURCE +# define _GNU_SOURCE 1 +#endif + +#include <errno.h> +#include <fnmatch.h> +#include <ctype.h> + +#if defined STDC_HEADERS || !defined isascii +# define IN_CTYPE_DOMAIN(c) 1 +#else +# define IN_CTYPE_DOMAIN(c) isascii (c) +#endif + +#define ISUPPER(c) (IN_CTYPE_DOMAIN (c) && isupper (c)) + + +#ifndef errno +extern int errno; +#endif + +/* Match STRING against the filename pattern PATTERN, returning zero if + it matches, nonzero if not. */ +int +fnmatch (const char *pattern, const char *string, int flags) +{ + register const char *p = pattern, *n = string; + register char c; + +/* Note that this evaluates C many times. */ +#define FOLD(c) ((flags & FNM_CASEFOLD) && ISUPPER ((unsigned char) (c)) \ + ? tolower ((unsigned char) (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++; + if (c == '\0') + /* Trailing \ loses. */ + return FNM_NOMATCH; + 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++) + { + if (c == '?') + { + /* A ? needs to match one character. */ + if (*n == '\0' || (*n == '/' && (flags & FNM_FILE_NAME))) + /* There isn't another character; no match. */ + return FNM_NOMATCH; + else + /* One character of the string is consumed in matching + this ? wildcard, so *??? won't match if there are + less than three characters. */ + ++n; + } + } + + if (c == '\0') + { + if ((flags & (FNM_FILE_NAME | FNM_LEADING_DIR)) == FNM_FILE_NAME) + for (; *n != '\0'; n++) + if (*n == '/') + return FNM_NOMATCH; + 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; + else if (*n == '/' && (flags & FNM_FILE_NAME)) + break; + 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 == '\\') + { + if (*p == '\0') + return FNM_NOMATCH; + 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 == '\\') + { + if (*p == '\0') + return FNM_NOMATCH; + /* 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; + +#undef FOLD +} diff --git a/lib/fnmatch.h b/lib/fnmatch.h new file mode 100644 index 00000000..af1dcf52 --- /dev/null +++ b/lib/fnmatch.h @@ -0,0 +1,69 @@ +/* Copyright (C) 1991, 1992, 1993 Free Software Foundation, Inc. + +NOTE: The canonical source of this file is maintained with the GNU C Library. +Bugs can be reported to bug-glibc@prep.ai.mit.edu. + +This program is free software; you can redistribute it and/or modify it +under the terms of the GNU General Public License as published by the +Free Software Foundation; either version 2, or (at your option) any +later version. + +This program is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +GNU General Public License for more details. + +You should have received a copy of the GNU General Public License +along with this program; if not, write to the Free Software Foundation, +Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef _FNMATCH_H + +#define _FNMATCH_H 1 + +#ifdef __cplusplus +extern "C" { +#endif + +#if defined (__cplusplus) || (defined (__STDC__) && __STDC__) +#undef __P +#define __P(protos) protos +#else /* Not C++ or ANSI C. */ +#undef __P +#define __P(protos) () +/* 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/getopt.c b/lib/getopt.c new file mode 100644 index 00000000..d176d3e7 --- /dev/null +++ b/lib/getopt.c @@ -0,0 +1,1049 @@ +/* Getopt for GNU. + NOTE: The canonical source of this file is maintained with the GNU + C Library. Bugs can be reported to bug-glibc@gnu.org. + + Copyright (C) 1987, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 + Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* 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> +#else +# if !defined __STDC__ || !__STDC__ +/* This is a separate conditional since some stdc systems + reject `defined (const)'. */ +# ifndef const +# define const +# endif +# 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. */ + +#define GETOPT_INTERFACE_VERSION 2 +#if !defined _LIBC && defined __GLIBC__ && __GLIBC__ >= 2 +# include <gnu-versions.h> +# if _GNU_GETOPT_INTERFACE_VERSION == GETOPT_INTERFACE_VERSION +# define ELIDE_CODE +# endif +#endif + +#ifndef ELIDE_CODE + + +/* 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> +# include <unistd.h> +#endif /* GNU C library. */ + +#ifdef VMS +# include <unixlib.h> +# if HAVE_STRING_H - 0 +# include <string.h> +# endif +#endif + +#ifndef _ +/* This is for other GNU distributions with internationalized messages. + When compiling libc, the _ macro is predefined. */ +# ifdef HAVE_LIBINTL_H +# include <libintl.h> +# define _(msgid) gettext (msgid) +# else +# define _(msgid) (msgid) +# endif +#endif + +/* 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; + +/* Index in ARGV of the next element to be scanned. + This is used for communication to and from the caller + and for communication between successive calls to `getopt'. + + On entry to `getopt', zero means this is the first call; initialize. + + When `getopt' returns -1, this is the index of the first of the + non-option elements that the caller should itself scan. + + Otherwise, `optind' communicates from one call to the next + how much of ARGV has been scanned so far. */ + +/* 1003.2 says this must be 1 before any call. */ +int optind = 1; + +/* Formerly, initialization of getopt depended on optind==0, which + causes problems with re-calling getopt as programs generally don't + know that. */ + +int __getopt_initialized; + +/* The next char to be scanned in the option-element + in which the last option character we returned was found. + This allows us to pick up the scan where we left off. + + If this is zero, or a null string, it means resume the scan + by advancing to the next ARGV-element. */ + +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 -1 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 + +# if HAVE_STRING_H +# include <string.h> +# else +# include <strings.h> +# endif + +/* Avoid depending on library functions or files + whose names are inconsistent. */ + +#ifndef getenv +extern char *getenv (); +#endif + +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__) && !defined strlen +/* 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; + +#ifdef _LIBC +/* Bash 2.0 gives us an environment variable containing flags + indicating ARGV elements that should not be considered arguments. */ + +/* Defined in getopt_init.c */ +extern char *__getopt_nonoption_flags; + +static int nonoption_flags_max_len; +static int nonoption_flags_len; + +static int original_argc; +static char *const *original_argv; + +/* Make sure the environment variable bash 2.0 puts in the environment + is valid for the getopt call we must make sure that the ARGV passed + to getopt is that one passed to the process. */ +static void +__attribute__ ((unused)) +store_args_and_env (int argc, char *const *argv) +{ + /* XXX This is no good solution. We should rather copy the args so + that we can compare them later. But we must not use malloc(3). */ + original_argc = argc; + original_argv = argv; +} +# ifdef text_set_element +text_set_element (__libc_subinit, store_args_and_env); +# endif /* text_set_element */ + +# define SWAP_FLAGS(ch1, ch2) \ + if (nonoption_flags_len > 0) \ + { \ + char __tmp = __getopt_nonoption_flags[ch1]; \ + __getopt_nonoption_flags[ch1] = __getopt_nonoption_flags[ch2]; \ + __getopt_nonoption_flags[ch2] = __tmp; \ + } +#else /* !_LIBC */ +# define SWAP_FLAGS(ch1, ch2) +#endif /* _LIBC */ + +/* Exchange two adjacent subsequences of ARGV. + One subsequence is elements [first_nonopt,last_nonopt) + which contains all the non-options that have been skipped so far. + The other is elements [last_nonopt,optind), which contains all + the options processed since those non-options were skipped. + + `first_nonopt' and `last_nonopt' are relocated so that they describe + the new indices of the non-options in ARGV after they are moved. */ + +#if defined __STDC__ && __STDC__ +static void exchange (char **); +#endif + +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. */ + +#ifdef _LIBC + /* First make sure the handling of the `__getopt_nonoption_flags' + string can work normally. Our top argument must be in the range + of the string. */ + if (nonoption_flags_len > 0 && top >= nonoption_flags_max_len) + { + /* We must extend the array. The user plays games with us and + presents new arguments. */ + char *new_str = malloc (top + 1); + if (new_str == NULL) + nonoption_flags_len = nonoption_flags_max_len = 0; + else + { + memset (__mempcpy (new_str, __getopt_nonoption_flags, + nonoption_flags_max_len), + '\0', top + 1 - nonoption_flags_max_len); + nonoption_flags_max_len = top + 1; + __getopt_nonoption_flags = new_str; + } + } +#endif + + while (top > middle && middle > bottom) + { + if (top - middle > middle - bottom) + { + /* Bottom segment is the short one. */ + int len = middle - bottom; + register int i; + + /* Swap it with the top part of the top segment. */ + for (i = 0; i < len; i++) + { + tem = argv[bottom + i]; + argv[bottom + i] = argv[top - (middle - bottom) + i]; + argv[top - (middle - bottom) + i] = tem; + SWAP_FLAGS (bottom + i, top - (middle - bottom) + i); + } + /* Exclude the moved bottom segment from further swapping. */ + top -= len; + } + else + { + /* Top segment is the short one. */ + int len = top - middle; + register int i; + + /* Swap it with the bottom part of the bottom segment. */ + for (i = 0; i < len; i++) + { + tem = argv[bottom + i]; + argv[bottom + i] = argv[middle + i]; + argv[middle + i] = tem; + SWAP_FLAGS (bottom + i, middle + i); + } + /* Exclude the moved top segment from further swapping. */ + bottom += len; + } + } + + /* Update records for the slots the non-options now occupy. */ + + first_nonopt += (optind - last_nonopt); + last_nonopt = optind; +} + +/* Initialize the internal data when the first call is made. */ + +#if defined __STDC__ && __STDC__ +static const char *_getopt_initialize (int, char *const *, const char *); +#endif +static const char * +_getopt_initialize (argc, argv, optstring) + int argc; + char *const *argv; + 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; + + 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; + +#ifdef _LIBC + if (posixly_correct == NULL + && argc == original_argc && argv == original_argv) + { + if (nonoption_flags_max_len == 0) + { + if (__getopt_nonoption_flags == NULL + || __getopt_nonoption_flags[0] == '\0') + nonoption_flags_max_len = -1; + else + { + const char *orig_str = __getopt_nonoption_flags; + int len = nonoption_flags_max_len = strlen (orig_str); + if (nonoption_flags_max_len < argc) + nonoption_flags_max_len = argc; + __getopt_nonoption_flags = + (char *) malloc (nonoption_flags_max_len); + if (__getopt_nonoption_flags == NULL) + nonoption_flags_max_len = -1; + else + memset (__mempcpy (__getopt_nonoption_flags, orig_str, len), + '\0', nonoption_flags_max_len - len); + } + } + nonoption_flags_len = nonoption_flags_max_len; + } + else + nonoption_flags_len = 0; +#endif + + return optstring; +} + +/* Scan elements of ARGV (whose length is ARGC) for option characters + given in OPTSTRING. + + If an element of ARGV starts with '-', and is not exactly "-" or "--", + then it is an option element. The characters of this element + (aside from the initial '-') are option characters. If `getopt' + is called repeatedly, it returns successively each of the option characters + from each of the option elements. + + If `getopt' finds another option character, it returns that character, + updating `optind' and `nextchar' so that the next call to `getopt' can + resume the scan with the following option character or ARGV-element. + + If there are no more option characters, `getopt' returns -1. + Then `optind' is the index in ARGV of the first ARGV-element + that is not an option. (The ARGV-elements have been permuted + so that those that are not options now come last.) + + OPTSTRING is a string containing the legitimate option characters. + If an option character is seen that is not listed in OPTSTRING, + return '?' after printing an error message. If you set `opterr' to + zero, the error message is suppressed but we still return '?'. + + If a char in OPTSTRING is followed by a colon, that means it wants an arg, + so the following text in the same ARGV-element, or the text of the following + ARGV-element, is returned in `optarg'. Two colons mean an option that + wants an optional arg; if there is text in the current ARGV-element, + it is returned in `optarg', otherwise `optarg' is set to zero. + + If OPTSTRING starts with `-' or `+', it requests different methods of + handling the non-option ARGV-elements. + See the comments about RETURN_IN_ORDER and REQUIRE_ORDER, above. + + Long-named options begin with `--' instead of `-'. + Their names may be abbreviated as long as the abbreviation is unique + or is an exact match for some defined option. If they have an + argument, it follows the option name in the same ARGV-element, separated + from the option name by a `=', or else the in next ARGV-element. + When `getopt' finds a long-named option, it returns 0 if that option's + `flag' field is nonzero, the value of the option's `val' field + if the `flag' field is zero. + + 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 || !__getopt_initialized) + { + if (optind == 0) + optind = 1; /* Don't scan ARGV[0], the program name. */ + optstring = _getopt_initialize (argc, argv, optstring); + __getopt_initialized = 1; + } + + /* Test whether ARGV[optind] points to a non-option argument. + Either it does not have option syntax, or there is an environment flag + from the shell indicating it is not an option. The later information + is only used when the used in the GNU libc. */ +#ifdef _LIBC +# define NONOPTION_P (argv[optind][0] != '-' || argv[optind][1] == '\0' \ + || (optind < nonoption_flags_len \ + && __getopt_nonoption_flags[optind] == '1')) +#else +# define NONOPTION_P (argv[optind][0] != '-' || argv[optind][1] == '\0') +#endif + + if (nextchar == NULL || *nextchar == '\0') + { + /* Advance to the next ARGV-element. */ + + /* Give FIRST_NONOPT & LAST_NONOPT rational values if OPTIND has been + moved back by the user (who may also have changed the arguments). */ + if (last_nonopt > optind) + last_nonopt = optind; + if (first_nonopt > optind) + first_nonopt = optind; + + 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 && NONOPTION_P) + 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 -1; + } + + /* If we have come to a non-option and did not permute it, + either stop the scan or describe it to the caller and pass it by. */ + + if (NONOPTION_P) + { + if (ordering == REQUIRE_ORDER) + return -1; + 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 = -1; + 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 ((unsigned int) (nameend - nextchar) + == (unsigned int) strlen (p->name)) + { + /* Exact match found. */ + pfound = p; + indfound = option_index; + exact = 1; + break; + } + else if (pfound == NULL) + { + /* First nonexact match found. */ + pfound = p; + indfound = option_index; + } + else + /* 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++; + optopt = 0; + 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); + + optopt = pfound->val; + 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); + optopt = pfound->val; + 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++; + optopt = 0; + 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 '?'; + } + /* Convenience. Treat POSIX -W foo same as long option --foo */ + if (temp[0] == 'W' && temp[1] == ';') + { + char *nameend; + const struct option *p; + const struct option *pfound = NULL; + int exact = 0; + int ambig = 0; + int indfound = 0; + int option_index; + + /* This is an option that requires an argument. */ + if (*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 = '?'; + return c; + } + else + /* We already incremented `optind' once; + increment it again when taking next ARGV-elt as argument. */ + optarg = argv[optind++]; + + /* optarg is now the argument, see if it's in the + table of longopts. */ + + for (nextchar = nameend = optarg; *nameend && *nameend != '='; nameend++) + /* Do nothing. */ ; + + /* Test all long options for either exact match + or abbreviated matches. */ + for (p = longopts, option_index = 0; p->name; p++, option_index++) + if (!strncmp (p->name, nextchar, nameend - nextchar)) + { + if ((unsigned int) (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 `-W %s' is ambiguous\n"), + argv[0], argv[optind]); + nextchar += strlen (nextchar); + optind++; + return '?'; + } + if (pfound != NULL) + { + option_index = indfound; + if (*nameend) + { + /* Don't test has_arg with >, because some C compilers don't + allow it to be used on enums. */ + if (pfound->has_arg) + optarg = nameend + 1; + else + { + if (opterr) + fprintf (stderr, _("\ +%s: option `-W %s' doesn't allow an argument\n"), + argv[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; + } + nextchar = NULL; + return 'W'; /* Let the application handle it. */ + } + if (temp[1] == ':') + { + if (temp[2] == ':') + { + /* This is an option that accepts an argument optionally. */ + if (*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 /* Not ELIDE_CODE. */ + +#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 == -1) + break; + + switch (c) + { + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + if (digit_optind != 0 && digit_optind != this_option_optind) + printf ("digits occur in two different argv-elements.\n"); + digit_optind = this_option_optind; + printf ("option %c\n", c); + break; + + case 'a': + printf ("option a\n"); + break; + + case 'b': + printf ("option b\n"); + break; + + case 'c': + printf ("option c with value `%s'\n", optarg); + break; + + case '?': + break; + + default: + printf ("?? getopt returned character code 0%o ??\n", c); + } + } + + if (optind < argc) + { + printf ("non-option ARGV-elements: "); + while (optind < argc) + printf ("%s ", argv[optind++]); + printf ("\n"); + } + + exit (0); +} + +#endif /* TEST */ diff --git a/lib/getopt.h b/lib/getopt.h new file mode 100644 index 00000000..39f4a168 --- /dev/null +++ b/lib/getopt.h @@ -0,0 +1,169 @@ +/* Declarations for getopt. + Copyright (C) 1989,90,91,92,93,94,96,97,98 Free Software Foundation, Inc. + NOTE: The canonical source of this file is maintained with the GNU C Library. + Bugs can be reported to bug-glibc@gnu.org. + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, + USA. */ + +#ifndef _GETOPT_H + +#ifndef __need_getopt +# define _GETOPT_H 1 +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +/* For communication from `getopt' to the caller. + When `getopt' finds an option that takes an argument, + the argument value is returned here. + Also, when `ordering' is RETURN_IN_ORDER, + each non-option ARGV-element is returned here. */ + +extern char *optarg; + +/* Index in ARGV of the next element to be scanned. + This is used for communication to and from the caller + and for communication between successive calls to `getopt'. + + On entry to `getopt', zero means this is the first call; initialize. + + When `getopt' returns -1, this is the index of the first of the + non-option elements that the caller should itself scan. + + Otherwise, `optind' communicates from one call to the next + how much of ARGV has been scanned so far. */ + +extern int optind; + +/* Callers store zero here to inhibit the error message `getopt' prints + for unrecognized options. */ + +extern int opterr; + +/* Set to an option character which was unrecognized. */ + +extern int optopt; + +#ifndef __need_getopt +/* Describe the long-named options requested by the application. + The LONG_OPTIONS argument to getopt_long or getopt_long_only is a vector + of `struct option' terminated by an element containing a name which is + zero. + + The field `has_arg' is: + no_argument (or 0) if the option does not take an argument, + required_argument (or 1) if the option requires an argument, + optional_argument (or 2) if the option takes an optional argument. + + If the field `flag' is not NULL, it points to a variable that is set + to the value given in the field `val' when the option is found, but + left unchanged if the option is not found. + + To have a long-named option do something other than set an `int' to + a compiled-in constant, such as set a value from `optarg', set the + option's `flag' field to zero and its `val' field to a nonzero + value (the equivalent single-letter option character, if there is + one). For long options that have a zero `flag' field, `getopt' + returns the contents of the `val' field. */ + +struct option +{ +# 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 +#endif /* need getopt */ + + +/* Get definitions and prototypes for functions to process the + arguments in ARGV (ARGC of them, minus the program name) for + options given in OPTS. + + Return the option character from OPTS just read. Return -1 when + there are no more options. For unrecognized options, or options + missing arguments, `optopt' is set to the option letter, and '?' is + returned. + + The OPTS string is a list of characters which are recognized option + letters, optionally followed by colons, specifying that that letter + takes an argument, to be placed in `optarg'. + + If a letter in OPTS is followed by two colons, its argument is + optional. This behavior is specific to the GNU `getopt'. + + The argument `--' causes premature termination of argument + scanning, explicitly telling `getopt' that there are no more + options. + + If OPTS begins with `--', then non-option arguments are treated as + arguments to the option '\0'. This behavior is specific to the GNU + `getopt'. */ + +#if (defined __STDC__ && __STDC__) || defined PROTOTYPES +# 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__ */ + +# ifndef __need_getopt +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); +# endif +#else /* not ((defined __STDC__ && __STDC__) || defined PROTOTYPES) */ +extern int getopt (); +# ifndef __need_getopt +extern int getopt_long (); +extern int getopt_long_only (); + +extern int _getopt_internal (); +# endif +#endif /* (defined __STDC__ && __STDC__) || defined PROTOTYPES */ + +#ifdef __cplusplus +} +#endif + +/* Make sure we later can get all the definitions and declarations. */ +#undef __need_getopt + +#endif /* getopt.h */ diff --git a/lib/hard-locale.c b/lib/hard-locale.c new file mode 100644 index 00000000..1c75b390 --- /dev/null +++ b/lib/hard-locale.c @@ -0,0 +1,85 @@ +/* hard-locale.c -- Determine whether a locale is hard. + Copyright 1997, 1998, 1999 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include <config.h> +#endif + +#ifndef __GNUC__ +# ifdef HAVE_ALLOCA_H +# include <alloca.h> +# else +# ifdef _AIX + # pragma alloca +# else +# ifdef _WIN32 +# include <malloc.h> +# include <io.h> +# else +# ifndef alloca +char *alloca (); +# endif +# endif +# endif +# endif +#endif + +#if HAVE_LOCALE_H +# include <locale.h> +#endif + +#if HAVE_STRING_H +# include <string.h> +#endif + +/* Return nonzero if the current CATEGORY locale is hard, i.e. if you + can't get away with assuming traditional C or POSIX behavior. */ +int +hard_locale (int category) +{ +#if ! (defined ENABLE_NLS && HAVE_SETLOCALE) + return 0; +#else + + int hard = 1; + char const *p = setlocale (category, 0); + + if (p) + { +# if defined __GLIBC__ && __GLIBC__ >= 2 + if (strcmp (p, "C") == 0 || strcmp (p, "POSIX") == 0) + hard = 0; +# else + char *locale = alloca (strlen (p) + 1); + strcpy (locale, p); + + /* Temporarily set the locale to the "C" and "POSIX" locales to + find their names, so that we can determine whether one or the + other is the caller's locale. */ + if (((p = setlocale (category, "C")) && strcmp (p, locale) == 0) + || ((p = setlocale (category, "POSIX")) && strcmp (p, locale) == 0)) + hard = 0; + + /* Restore the caller's locale. */ + setlocale (category, locale); +# endif + } + + return hard; + +#endif +} diff --git a/lib/hard-locale.h b/lib/hard-locale.h new file mode 100644 index 00000000..5b054d9a --- /dev/null +++ b/lib/hard-locale.h @@ -0,0 +1,18 @@ +#ifndef HARD_LOCALE_H_ +# define HARD_LOCALE_H_ 1 + +# if HAVE_CONFIG_H +# include <config.h> +# endif + +# ifndef PARAMS +# if defined PROTOTYPES || (defined __STDC__ && __STDC__) +# define PARAMS(Args) Args +# else +# define PARAMS(Args) () +# endif +# endif + +int hard_locale PARAMS ((int)); + +#endif /* HARD_LOCALE_H_ */ diff --git a/lib/isdir.c b/lib/isdir.c new file mode 100644 index 00000000..01cf8d5e --- /dev/null +++ b/lib/isdir.c @@ -0,0 +1,42 @@ +/* isdir.c -- determine whether a directory exists + Copyright (C) 1990, 1998 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include <config.h> +#endif + +#include <sys/types.h> +#include <sys/stat.h> + +#if STAT_MACROS_BROKEN +# undef S_ISDIR +#endif + +#if !defined S_ISDIR && defined S_IFDIR +# define S_ISDIR(Mode) (((Mode) & S_IFMT) == S_IFDIR) +#endif + +/* If PATH is an existing directory or symbolic link to a directory, + return nonzero, else 0. */ + +int +isdir (const char *path) +{ + struct stat stats; + + return stat (path, &stats) == 0 && S_ISDIR (stats.st_mode); +} diff --git a/lib/memchr.c b/lib/memchr.c new file mode 100644 index 00000000..8c3b59c1 --- /dev/null +++ b/lib/memchr.c @@ -0,0 +1,199 @@ +/* Copyright (C) 1991, 1993, 1997, 1999 Free Software Foundation, Inc. + Based on strlen implementation by Torbjorn Granlund (tege@sics.se), + with help from Dan Sahlin (dan@sics.se) and + commentary by Jim Blandy (jimb@ai.mit.edu); + adaptation to memchr suggested by Dick Karpinski (dick@cca.ucsf.edu), + and implemented by Roland McGrath (roland@ai.mit.edu). + +NOTE: The canonical source of this file is maintained with the GNU C Library. +Bugs can be reported to bug-glibc@prep.ai.mit.edu. + +This program is free software; you can redistribute it and/or modify it +under the terms of the GNU General Public License as published by the +Free Software Foundation; either version 2, or (at your option) any +later version. + +This program is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +GNU General Public License for more details. + +You should have received a copy of the GNU General Public License +along with this program; if not, write to the Free Software +Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, +USA. */ + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +#undef __ptr_t +#if defined (__cplusplus) || (defined (__STDC__) && __STDC__) +# define __ptr_t void * +#else /* Not C++ or ANSI C. */ +# define __ptr_t char * +#endif /* C++ or ANSI C. */ + +#if defined (_LIBC) +# include <string.h> +#endif + +#if defined (HAVE_LIMITS_H) || defined (_LIBC) +# include <limits.h> +#endif + +#define LONG_MAX_32_BITS 2147483647 + +#ifndef LONG_MAX +# define LONG_MAX LONG_MAX_32_BITS +#endif + +#include <sys/types.h> + + +/* Search no more than N bytes of S for C. */ + +__ptr_t +memchr (s, c, n) + const __ptr_t s; + int c; + size_t n; +{ + const unsigned char *char_ptr; + const unsigned long int *longword_ptr; + unsigned long int longword, magic_bits, charmask; + + c = (unsigned char) c; + + /* Handle the first few characters by reading one character at a time. + Do this until CHAR_PTR is aligned on a longword boundary. */ + for (char_ptr = (const unsigned char *) s; + n > 0 && ((unsigned long int) char_ptr + & (sizeof (longword) - 1)) != 0; + --n, ++char_ptr) + if (*char_ptr == c) + return (__ptr_t) char_ptr; + + /* All these elucidatory comments refer to 4-byte longwords, + but the theory applies equally well to 8-byte longwords. */ + + longword_ptr = (unsigned long int *) char_ptr; + + /* Bits 31, 24, 16, and 8 of this number are zero. Call these bits + the "holes." Note that there is a hole just to the left of + each byte, with an extra at the end: + + bits: 01111110 11111110 11111110 11111111 + bytes: AAAAAAAA BBBBBBBB CCCCCCCC DDDDDDDD + + The 1-bits make sure that carries propagate to the next 0-bit. + The 0-bits provide holes for carries to fall into. */ + + if (sizeof (longword) != 4 && sizeof (longword) != 8) + abort (); + +#if LONG_MAX <= LONG_MAX_32_BITS + magic_bits = 0x7efefeff; +#else + magic_bits = ((unsigned long int) 0x7efefefe << 32) | 0xfefefeff; +#endif + + /* Set up a longword, each of whose bytes is C. */ + charmask = c | (c << 8); + charmask |= charmask << 16; +#if LONG_MAX > LONG_MAX_32_BITS + charmask |= charmask << 32; +#endif + + /* Instead of the traditional loop which tests each character, + we will test a longword at a time. The tricky part is testing + if *any of the four* bytes in the longword in question are zero. */ + while (n >= sizeof (longword)) + { + /* We tentatively exit the loop if adding MAGIC_BITS to + LONGWORD fails to change any of the hole bits of LONGWORD. + + 1) Is this safe? Will it catch all the zero bytes? + Suppose there is a byte with all zeros. Any carry bits + propagating from its left will fall into the hole at its + least significant bit and stop. Since there will be no + carry from its most significant bit, the LSB of the + byte to the left will be unchanged, and the zero will be + detected. + + 2) Is this worthwhile? Will it ignore everything except + zero bytes? Suppose every byte of LONGWORD has a bit set + somewhere. There will be a carry into bit 8. If bit 8 + is set, this will carry into bit 16. If bit 8 is clear, + one of bits 9-15 must be set, so there will be a carry + into bit 16. Similarly, there will be a carry into bit + 24. If one of bits 24-30 is set, there will be a carry + into bit 31, so all of the hole bits will be changed. + + The one misfire occurs when bits 24-30 are clear and bit + 31 is set; in this case, the hole at bit 31 is not + changed. If we had access to the processor carry flag, + we could close this loophole by putting the fourth hole + at bit 32! + + So it ignores everything except 128's, when they're aligned + properly. + + 3) But wait! Aren't we looking for C, not zero? + Good point. So what we do is XOR LONGWORD with a longword, + each of whose bytes is C. This turns each byte that is C + into a zero. */ + + longword = *longword_ptr++ ^ charmask; + + /* Add MAGIC_BITS to LONGWORD. */ + if ((((longword + magic_bits) + + /* Set those bits that were unchanged by the addition. */ + ^ ~longword) + + /* Look at only the hole bits. If any of the hole bits + are unchanged, most likely one of the bytes was a + zero. */ + & ~magic_bits) != 0) + { + /* Which of the bytes was C? If none of them were, it was + a misfire; continue the search. */ + + const unsigned char *cp = (const unsigned char *) (longword_ptr - 1); + + if (cp[0] == c) + return (__ptr_t) cp; + if (cp[1] == c) + return (__ptr_t) &cp[1]; + if (cp[2] == c) + return (__ptr_t) &cp[2]; + if (cp[3] == c) + return (__ptr_t) &cp[3]; +#if LONG_MAX > 2147483647 + if (cp[4] == c) + return (__ptr_t) &cp[4]; + if (cp[5] == c) + return (__ptr_t) &cp[5]; + if (cp[6] == c) + return (__ptr_t) &cp[6]; + if (cp[7] == c) + return (__ptr_t) &cp[7]; +#endif + } + + n -= sizeof (longword); + } + + char_ptr = (const unsigned char *) longword_ptr; + + while (n-- > 0) + { + if (*char_ptr == c) + return (__ptr_t) char_ptr; + else + ++char_ptr; + } + + return 0; +} diff --git a/lib/obstack.c b/lib/obstack.c new file mode 100644 index 00000000..17c63134 --- /dev/null +++ b/lib/obstack.c @@ -0,0 +1,598 @@ +/* obstack.c - subroutines used implicitly by object stack macros + Copyright (C) 1988-1994,96,97,98,99 Free Software Foundation, Inc. + + This file is part of the GNU C Library. Its master source is NOT part of + the C library, however. The master source lives in /gd/gnu/lib. + + 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., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#ifdef HAVE_CONFIG_H +#include <config.h> +#endif + +#include "obstack.h" + +/* NOTE BEFORE MODIFYING THIS FILE: This version number must be + incremented whenever callers compiled using an old obstack.h can no + longer properly call the functions in this obstack.c. */ +#define OBSTACK_INTERFACE_VERSION 1 + +/* Comment out all this code if we are using the GNU C Library, and are not + actually compiling the library itself, and the installed library + supports the same library interface we do. 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. */ + +#include <stdio.h> /* Random thing to get __GNU_LIBRARY__. */ +#if !defined (_LIBC) && defined (__GNU_LIBRARY__) && __GNU_LIBRARY__ > 1 +#include <gnu-versions.h> +#if _GNU_OBSTACK_INTERFACE_VERSION == OBSTACK_INTERFACE_VERSION +#define ELIDE_CODE +#endif +#endif + + +#ifndef ELIDE_CODE + + +#if defined (__STDC__) && __STDC__ +#define POINTER void * +#else +#define POINTER char * +#endif + +/* Determine default alignment. */ +struct fooalign {char x; double d;}; +#define DEFAULT_ALIGNMENT \ + ((PTR_INT_TYPE) ((char *) &((struct fooalign *) 0)->d - (char *) 0)) +/* If malloc were really smart, it would round addresses to DEFAULT_ALIGNMENT. + But in fact it might be less smart and round addresses to as much as + DEFAULT_ROUNDING. So we prepare for it to do that. */ +union fooround {long x; double d;}; +#define DEFAULT_ROUNDING (sizeof (union fooround)) + +/* When we copy a long block of data, this is the unit to do it with. + On some machines, copying successive ints does not work; + in such a case, redefine COPYING_UNIT to `long' (if that works) + or `char' as a last resort. */ +#ifndef COPYING_UNIT +#define COPYING_UNIT int +#endif + + +/* The functions allocating more room by calling `obstack_chunk_alloc' + jump to the handler pointed to by `obstack_alloc_failed_handler'. + This can be set to a user defined function which should either + abort gracefully or use longjump - but shouldn't return. This + variable by default points to the internal function + `print_and_abort'. */ +#if defined (__STDC__) && __STDC__ +static void print_and_abort (void); +void (*obstack_alloc_failed_handler) (void) = print_and_abort; +#else +static void print_and_abort (); +void (*obstack_alloc_failed_handler) () = print_and_abort; +#endif + +/* Exit value used when `print_and_abort' is used. */ +#if defined __GNU_LIBRARY__ || defined HAVE_STDLIB_H +#include <stdlib.h> +#endif +#ifndef EXIT_FAILURE +#define EXIT_FAILURE 1 +#endif +int obstack_exit_failure = EXIT_FAILURE; + +/* The non-GNU-C macros copy the obstack into this global variable + to avoid multiple evaluation. */ + +struct obstack *_obstack; + +/* Define a macro that either calls functions with the traditional malloc/free + calling interface, or calls functions with the mmalloc/mfree interface + (that adds an extra first argument), based on the state of use_extra_arg. + For free, do not use ?:, since some compilers, like the MIPS compilers, + do not allow (expr) ? void : void. */ + +#if defined (__STDC__) && __STDC__ +#define CALL_CHUNKFUN(h, size) \ + (((h) -> use_extra_arg) \ + ? (*(h)->chunkfun) ((h)->extra_arg, (size)) \ + : (*(struct _obstack_chunk *(*) (long)) (h)->chunkfun) ((size))) + +#define CALL_FREEFUN(h, old_chunk) \ + do { \ + if ((h) -> use_extra_arg) \ + (*(h)->freefun) ((h)->extra_arg, (old_chunk)); \ + else \ + (*(void (*) (void *)) (h)->freefun) ((old_chunk)); \ + } while (0) +#else +#define CALL_CHUNKFUN(h, size) \ + (((h) -> use_extra_arg) \ + ? (*(h)->chunkfun) ((h)->extra_arg, (size)) \ + : (*(struct _obstack_chunk *(*) ()) (h)->chunkfun) ((size))) + +#define CALL_FREEFUN(h, old_chunk) \ + do { \ + if ((h) -> use_extra_arg) \ + (*(h)->freefun) ((h)->extra_arg, (old_chunk)); \ + else \ + (*(void (*) ()) (h)->freefun) ((old_chunk)); \ + } while (0) +#endif + + +/* Initialize an obstack H for use. Specify chunk size SIZE (0 means default). + Objects start on multiples of ALIGNMENT (0 means use default). + CHUNKFUN is the function to use to allocate chunks, + and FREEFUN the function to free them. + + Return nonzero if successful, calls obstack_alloc_failed_handler if + allocation fails. */ + +int +_obstack_begin (h, size, alignment, chunkfun, freefun) + struct obstack *h; + int size; + int alignment; +#if defined (__STDC__) && __STDC__ + POINTER (*chunkfun) (long); + void (*freefun) (void *); +#else + POINTER (*chunkfun) (); + void (*freefun) (); +#endif +{ + register struct _obstack_chunk *chunk; /* points to new chunk */ + + if (alignment == 0) + alignment = (int) DEFAULT_ALIGNMENT; + if (size == 0) + /* Default size is what GNU malloc can fit in a 4096-byte block. */ + { + /* 12 is sizeof (mhead) and 4 is EXTRA from GNU malloc. + Use the values for range checking, because if range checking is off, + the extra bytes won't be missed terribly, but if range checking is on + and we used a larger request, a whole extra 4096 bytes would be + allocated. + + These number are irrelevant to the new GNU malloc. I suspect it is + less sensitive to the size of the request. */ + int extra = ((((12 + DEFAULT_ROUNDING - 1) & ~(DEFAULT_ROUNDING - 1)) + + 4 + DEFAULT_ROUNDING - 1) + & ~(DEFAULT_ROUNDING - 1)); + size = 4096 - extra; + } + +#if defined (__STDC__) && __STDC__ + h->chunkfun = (struct _obstack_chunk * (*)(void *, long)) chunkfun; + h->freefun = (void (*) (void *, struct _obstack_chunk *)) freefun; +#else + h->chunkfun = (struct _obstack_chunk * (*)()) chunkfun; + h->freefun = freefun; +#endif + h->chunk_size = size; + h->alignment_mask = alignment - 1; + h->use_extra_arg = 0; + + chunk = h->chunk = CALL_CHUNKFUN (h, h -> chunk_size); + if (!chunk) + (*obstack_alloc_failed_handler) (); + h->next_free = h->object_base = chunk->contents; + h->chunk_limit = chunk->limit + = (char *) chunk + h->chunk_size; + chunk->prev = 0; + /* The initial chunk now contains no empty object. */ + h->maybe_empty_object = 0; + h->alloc_failed = 0; + return 1; +} + +int +_obstack_begin_1 (h, size, alignment, chunkfun, freefun, arg) + struct obstack *h; + int size; + int alignment; +#if defined (__STDC__) && __STDC__ + POINTER (*chunkfun) (POINTER, long); + void (*freefun) (POINTER, POINTER); +#else + POINTER (*chunkfun) (); + void (*freefun) (); +#endif + POINTER arg; +{ + register struct _obstack_chunk *chunk; /* points to new chunk */ + + if (alignment == 0) + alignment = (int) DEFAULT_ALIGNMENT; + if (size == 0) + /* Default size is what GNU malloc can fit in a 4096-byte block. */ + { + /* 12 is sizeof (mhead) and 4 is EXTRA from GNU malloc. + Use the values for range checking, because if range checking is off, + the extra bytes won't be missed terribly, but if range checking is on + and we used a larger request, a whole extra 4096 bytes would be + allocated. + + These number are irrelevant to the new GNU malloc. I suspect it is + less sensitive to the size of the request. */ + int extra = ((((12 + DEFAULT_ROUNDING - 1) & ~(DEFAULT_ROUNDING - 1)) + + 4 + DEFAULT_ROUNDING - 1) + & ~(DEFAULT_ROUNDING - 1)); + size = 4096 - extra; + } + +#if defined(__STDC__) && __STDC__ + h->chunkfun = (struct _obstack_chunk * (*)(void *,long)) chunkfun; + h->freefun = (void (*) (void *, struct _obstack_chunk *)) freefun; +#else + h->chunkfun = (struct _obstack_chunk * (*)()) chunkfun; + h->freefun = freefun; +#endif + h->chunk_size = size; + h->alignment_mask = alignment - 1; + h->extra_arg = arg; + h->use_extra_arg = 1; + + chunk = h->chunk = CALL_CHUNKFUN (h, h -> chunk_size); + if (!chunk) + (*obstack_alloc_failed_handler) (); + h->next_free = h->object_base = chunk->contents; + h->chunk_limit = chunk->limit + = (char *) chunk + h->chunk_size; + chunk->prev = 0; + /* The initial chunk now contains no empty object. */ + h->maybe_empty_object = 0; + h->alloc_failed = 0; + return 1; +} + +/* Allocate a new current chunk for the obstack *H + on the assumption that LENGTH bytes need to be added + to the current object, or a new object of length LENGTH allocated. + Copies any partial object from the end of the old chunk + to the beginning of the new one. */ + +void +_obstack_newchunk (h, length) + struct obstack *h; + int length; +{ + register struct _obstack_chunk *old_chunk = h->chunk; + register struct _obstack_chunk *new_chunk; + register long new_size; + register long obj_size = h->next_free - h->object_base; + register long i; + long already; + + /* Compute size for new chunk. */ + new_size = (obj_size + length) + (obj_size >> 3) + 100; + if (new_size < h->chunk_size) + new_size = h->chunk_size; + + /* Allocate and initialize the new chunk. */ + new_chunk = CALL_CHUNKFUN (h, new_size); + if (!new_chunk) + (*obstack_alloc_failed_handler) (); + h->chunk = new_chunk; + new_chunk->prev = old_chunk; + new_chunk->limit = h->chunk_limit = (char *) new_chunk + new_size; + + /* Move the existing object to the new chunk. + Word at a time is fast and is safe if the object + is sufficiently aligned. */ + if (h->alignment_mask + 1 >= DEFAULT_ALIGNMENT) + { + for (i = obj_size / sizeof (COPYING_UNIT) - 1; + i >= 0; i--) + ((COPYING_UNIT *)new_chunk->contents)[i] + = ((COPYING_UNIT *)h->object_base)[i]; + /* We used to copy the odd few remaining bytes as one extra COPYING_UNIT, + but that can cross a page boundary on a machine + which does not do strict alignment for COPYING_UNITS. */ + already = obj_size / sizeof (COPYING_UNIT) * sizeof (COPYING_UNIT); + } + else + already = 0; + /* Copy remaining bytes one by one. */ + for (i = already; i < obj_size; i++) + new_chunk->contents[i] = h->object_base[i]; + + /* If the object just copied was the only data in OLD_CHUNK, + free that chunk and remove it from the chain. + But not if that chunk might contain an empty object. */ + if (h->object_base == old_chunk->contents && ! h->maybe_empty_object) + { + new_chunk->prev = old_chunk->prev; + CALL_FREEFUN (h, old_chunk); + } + + h->object_base = new_chunk->contents; + h->next_free = h->object_base + obj_size; + /* The new chunk certainly contains no empty object yet. */ + h->maybe_empty_object = 0; +} + +/* Return nonzero if object OBJ has been allocated from obstack H. + This is here for debugging. + If you use it in a program, you are probably losing. */ + +#if defined (__STDC__) && __STDC__ +/* Suppress -Wmissing-prototypes warning. We don't want to declare this in + obstack.h because it is just for debugging. */ +int _obstack_allocated_p (struct obstack *h, POINTER obj); +#endif + +int +_obstack_allocated_p (h, obj) + struct obstack *h; + POINTER obj; +{ + register struct _obstack_chunk *lp; /* below addr of any objects in this chunk */ + register struct _obstack_chunk *plp; /* point to previous chunk if any */ + + lp = (h)->chunk; + /* We use >= rather than > since the object cannot be exactly at + the beginning of the chunk but might be an empty object exactly + at the end of an adjacent chunk. */ + while (lp != 0 && ((POINTER) lp >= obj || (POINTER) (lp)->limit < obj)) + { + plp = lp->prev; + lp = plp; + } + return lp != 0; +} + +/* Free objects in obstack H, including OBJ and everything allocate + more recently than OBJ. If OBJ is zero, free everything in H. */ + +#undef obstack_free + +/* This function has two names with identical definitions. + This is the first one, called from non-ANSI code. */ + +void +_obstack_free (h, obj) + struct obstack *h; + POINTER obj; +{ + register struct _obstack_chunk *lp; /* below addr of any objects in this chunk */ + register struct _obstack_chunk *plp; /* point to previous chunk if any */ + + lp = h->chunk; + /* We use >= because there cannot be an object at the beginning of a chunk. + But there can be an empty object at that address + at the end of another chunk. */ + while (lp != 0 && ((POINTER) lp >= obj || (POINTER) (lp)->limit < obj)) + { + plp = lp->prev; + CALL_FREEFUN (h, lp); + lp = plp; + /* If we switch chunks, we can't tell whether the new current + chunk contains an empty object, so assume that it may. */ + h->maybe_empty_object = 1; + } + if (lp) + { + h->object_base = h->next_free = (char *) (obj); + h->chunk_limit = lp->limit; + h->chunk = lp; + } + else if (obj != 0) + /* obj is not in any of the chunks! */ + abort (); +} + +/* This function is used from ANSI code. */ + +void +obstack_free (h, obj) + struct obstack *h; + POINTER obj; +{ + register struct _obstack_chunk *lp; /* below addr of any objects in this chunk */ + register struct _obstack_chunk *plp; /* point to previous chunk if any */ + + lp = h->chunk; + /* We use >= because there cannot be an object at the beginning of a chunk. + But there can be an empty object at that address + at the end of another chunk. */ + while (lp != 0 && ((POINTER) lp >= obj || (POINTER) (lp)->limit < obj)) + { + plp = lp->prev; + CALL_FREEFUN (h, lp); + lp = plp; + /* If we switch chunks, we can't tell whether the new current + chunk contains an empty object, so assume that it may. */ + h->maybe_empty_object = 1; + } + if (lp) + { + h->object_base = h->next_free = (char *) (obj); + h->chunk_limit = lp->limit; + h->chunk = lp; + } + else if (obj != 0) + /* obj is not in any of the chunks! */ + abort (); +} + +int +_obstack_memory_used (h) + struct obstack *h; +{ + register struct _obstack_chunk* lp; + register int nbytes = 0; + + for (lp = h->chunk; lp != 0; lp = lp->prev) + { + nbytes += lp->limit - (char *) lp; + } + return nbytes; +} + +/* Define the error handler. */ +#ifndef _ +# ifdef HAVE_LIBINTL_H +# include <libintl.h> +# ifndef _ +# define _(Str) gettext (Str) +# endif +# else +# define _(Str) (Str) +# endif +#endif +#if defined _LIBC && defined USE_IN_LIBIO +# include <libio/iolibio.h> +# define fputs(s, f) _IO_fputs (s, f) +#endif + +static void +print_and_abort () +{ + fputs (_("memory exhausted"), stderr); + fputc ('\n', stderr); + exit (obstack_exit_failure); +} + +#if 0 +/* These are now turned off because the applications do not use it + and it uses bcopy via obstack_grow, which causes trouble on sysV. */ + +/* Now define the functional versions of the obstack macros. + Define them to simply use the corresponding macros to do the job. */ + +#if defined (__STDC__) && __STDC__ +/* These function definitions do not work with non-ANSI preprocessors; + they won't pass through the macro names in parentheses. */ + +/* The function names appear in parentheses in order to prevent + the macro-definitions of the names from being expanded there. */ + +POINTER (obstack_base) (obstack) + struct obstack *obstack; +{ + return obstack_base (obstack); +} + +POINTER (obstack_next_free) (obstack) + struct obstack *obstack; +{ + return obstack_next_free (obstack); +} + +int (obstack_object_size) (obstack) + struct obstack *obstack; +{ + return obstack_object_size (obstack); +} + +int (obstack_room) (obstack) + struct obstack *obstack; +{ + return obstack_room (obstack); +} + +int (obstack_make_room) (obstack, length) + struct obstack *obstack; + int length; +{ + return obstack_make_room (obstack, length); +} + +void (obstack_grow) (obstack, pointer, length) + struct obstack *obstack; + POINTER pointer; + int length; +{ + obstack_grow (obstack, pointer, length); +} + +void (obstack_grow0) (obstack, pointer, length) + struct obstack *obstack; + POINTER pointer; + int length; +{ + obstack_grow0 (obstack, pointer, length); +} + +void (obstack_1grow) (obstack, character) + struct obstack *obstack; + int character; +{ + obstack_1grow (obstack, character); +} + +void (obstack_blank) (obstack, length) + struct obstack *obstack; + int length; +{ + obstack_blank (obstack, length); +} + +void (obstack_1grow_fast) (obstack, character) + struct obstack *obstack; + int character; +{ + obstack_1grow_fast (obstack, character); +} + +void (obstack_blank_fast) (obstack, length) + struct obstack *obstack; + int length; +{ + obstack_blank_fast (obstack, length); +} + +POINTER (obstack_finish) (obstack) + struct obstack *obstack; +{ + return obstack_finish (obstack); +} + +POINTER (obstack_alloc) (obstack, length) + struct obstack *obstack; + int length; +{ + return obstack_alloc (obstack, length); +} + +POINTER (obstack_copy) (obstack, pointer, length) + struct obstack *obstack; + POINTER pointer; + int length; +{ + return obstack_copy (obstack, pointer, length); +} + +POINTER (obstack_copy0) (obstack, pointer, length) + struct obstack *obstack; + POINTER pointer; + int length; +{ + return obstack_copy0 (obstack, pointer, length); +} + +#endif /* __STDC__ */ + +#endif /* 0 */ + +#endif /* !ELIDE_CODE */ diff --git a/lib/obstack.h b/lib/obstack.h new file mode 100644 index 00000000..4d49ce02 --- /dev/null +++ b/lib/obstack.h @@ -0,0 +1,593 @@ +/* obstack.h - object stack macros + Copyright (C) 1988,89,90,91,92,93,94,96,97,98,99 Free Software Foundation, Inc. + + This file is part of the GNU C Library. Its master source is NOT part of + the C library, however. The master source lives in /gd/gnu/lib. + + 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., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +/* Summary: + +All the apparent functions defined here are macros. The idea +is that you would use these pre-tested macros to solve a +very specific set of problems, and they would run fast. +Caution: no side-effects in arguments please!! They may be +evaluated MANY times!! + +These macros operate a stack of objects. Each object starts life +small, and may grow to maturity. (Consider building a word syllable +by syllable.) An object can move while it is growing. Once it has +been "finished" it never changes address again. So the "top of the +stack" is typically an immature growing object, while the rest of the +stack is of mature, fixed size and fixed address objects. + +These routines grab large chunks of memory, using a function you +supply, called `obstack_chunk_alloc'. On occasion, they free chunks, +by calling `obstack_chunk_free'. You must define them and declare +them before using any obstack macros. + +Each independent stack is represented by a `struct obstack'. +Each of the obstack macros expects a pointer to such a structure +as the first argument. + +One motivation for this package is the problem of growing char strings +in symbol tables. Unless you are "fascist pig with a read-only mind" +--Gosper's immortal quote from HAKMEM item 154, out of context--you +would not like to put any arbitrary upper limit on the length of your +symbols. + +In practice this often means you will build many short symbols and a +few long symbols. At the time you are reading a symbol you don't know +how long it is. One traditional method is to read a symbol into a +buffer, realloc()ating the buffer every time you try to read a symbol +that is longer than the buffer. This is beaut, but you still will +want to copy the symbol from the buffer to a more permanent +symbol-table entry say about half the time. + +With obstacks, you can work differently. Use one obstack for all symbol +names. As you read a symbol, grow the name in the obstack gradually. +When the name is complete, finalize it. Then, if the symbol exists already, +free the newly read name. + +The way we do this is to take a large chunk, allocating memory from +low addresses. When you want to build a symbol in the chunk you just +add chars above the current "high water mark" in the chunk. When you +have finished adding chars, because you got to the end of the symbol, +you know how long the chars are, and you can create a new object. +Mostly the chars will not burst over the highest address of the chunk, +because you would typically expect a chunk to be (say) 100 times as +long as an average object. + +In case that isn't clear, when we have enough chars to make up +the object, THEY ARE ALREADY CONTIGUOUS IN THE CHUNK (guaranteed) +so we just point to it where it lies. No moving of chars is +needed and this is the second win: potentially long strings need +never be explicitly shuffled. Once an object is formed, it does not +change its address during its lifetime. + +When the chars burst over a chunk boundary, we allocate a larger +chunk, and then copy the partly formed object from the end of the old +chunk to the beginning of the new larger chunk. We then carry on +accreting characters to the end of the object as we normally would. + +A special macro is provided to add a single char at a time to a +growing object. This allows the use of register variables, which +break the ordinary 'growth' macro. + +Summary: + We allocate large chunks. + We carve out one object at a time from the current chunk. + Once carved, an object never moves. + We are free to append data of any size to the currently + growing object. + Exactly one object is growing in an obstack at any one time. + You can run one obstack per control block. + You may have as many control blocks as you dare. + Because of the way we do it, you can `unwind' an obstack + back to a previous state. (You may remove objects much + as you would with a stack.) +*/ + + +/* Don't do the contents of this file more than once. */ + +#ifndef _OBSTACK_H +#define _OBSTACK_H 1 + +#ifdef __cplusplus +extern "C" { +#endif + +/* We use subtraction of (char *) 0 instead of casting to int + because on word-addressable machines a simple cast to int + may ignore the byte-within-word field of the pointer. */ + +#ifndef __PTR_TO_INT +# define __PTR_TO_INT(P) ((P) - (char *) 0) +#endif + +#ifndef __INT_TO_PTR +# define __INT_TO_PTR(P) ((P) + (char *) 0) +#endif + +/* We need the type of the resulting object. If __PTRDIFF_TYPE__ is + defined, as with GNU C, use that; that way we don't pollute the + namespace with <stddef.h>'s symbols. Otherwise, if <stddef.h> is + available, include it and use ptrdiff_t. In traditional C, long is + the best that we can do. */ + +#ifdef __PTRDIFF_TYPE__ +# define PTR_INT_TYPE __PTRDIFF_TYPE__ +#else +# ifdef HAVE_STDDEF_H +# include <stddef.h> +# define PTR_INT_TYPE ptrdiff_t +# else +# define PTR_INT_TYPE long +# endif +#endif + +#if defined _LIBC || defined HAVE_STRING_H +# include <string.h> +# define _obstack_memcpy(To, From, N) memcpy ((To), (From), (N)) +#else +# ifdef memcpy +# define _obstack_memcpy(To, From, N) memcpy ((To), (From), (N)) +# else +# define _obstack_memcpy(To, From, N) bcopy ((From), (To), (N)) +# endif +#endif + +struct _obstack_chunk /* Lives at front of each chunk. */ +{ + char *limit; /* 1 past end of this chunk */ + struct _obstack_chunk *prev; /* address of prior chunk or NULL */ + char contents[4]; /* objects begin here */ +}; + +struct obstack /* control current object in current chunk */ +{ + long chunk_size; /* preferred size to allocate chunks in */ + struct _obstack_chunk *chunk; /* address of current struct obstack_chunk */ + char *object_base; /* address of object we are building */ + char *next_free; /* where to add next char to current object */ + char *chunk_limit; /* address of char after current chunk */ + PTR_INT_TYPE temp; /* Temporary for some macros. */ + int alignment_mask; /* Mask of alignment for each object. */ +#if defined __STDC__ && __STDC__ + /* These prototypes vary based on `use_extra_arg', and we use + casts to the prototypeless function type in all assignments, + but having prototypes here quiets -Wstrict-prototypes. */ + struct _obstack_chunk *(*chunkfun) (void *, long); + void (*freefun) (void *, struct _obstack_chunk *); + void *extra_arg; /* first arg for chunk alloc/dealloc funcs */ +#else + struct _obstack_chunk *(*chunkfun) (); /* User's fcn to allocate a chunk. */ + void (*freefun) (); /* User's function to free a chunk. */ + char *extra_arg; /* first arg for chunk alloc/dealloc funcs */ +#endif + unsigned use_extra_arg:1; /* chunk alloc/dealloc funcs take extra arg */ + unsigned maybe_empty_object:1;/* There is a possibility that the current + chunk contains a zero-length object. This + prevents freeing the chunk if we allocate + a bigger chunk to replace it. */ + unsigned alloc_failed:1; /* No longer used, as we now call the failed + handler on error, but retained for binary + compatibility. */ +}; + +/* Declare the external functions we use; they are in obstack.c. */ + +#if defined __STDC__ && __STDC__ +extern void _obstack_newchunk (struct obstack *, int); +extern void _obstack_free (struct obstack *, void *); +extern int _obstack_begin (struct obstack *, int, int, + void *(*) (long), void (*) (void *)); +extern int _obstack_begin_1 (struct obstack *, int, int, + void *(*) (void *, long), + void (*) (void *, void *), void *); +extern int _obstack_memory_used (struct obstack *); +#else +extern void _obstack_newchunk (); +extern void _obstack_free (); +extern int _obstack_begin (); +extern int _obstack_begin_1 (); +extern int _obstack_memory_used (); +#endif + +#if defined __STDC__ && __STDC__ + +/* Do the function-declarations after the structs + but before defining the macros. */ + +void obstack_init (struct obstack *obstack); + +void * obstack_alloc (struct obstack *obstack, int size); + +void * obstack_copy (struct obstack *obstack, void *address, int size); +void * obstack_copy0 (struct obstack *obstack, void *address, int size); + +void obstack_free (struct obstack *obstack, void *block); + +void obstack_blank (struct obstack *obstack, int size); + +void obstack_grow (struct obstack *obstack, void *data, int size); +void obstack_grow0 (struct obstack *obstack, void *data, int size); + +void obstack_1grow (struct obstack *obstack, int data_char); +void obstack_ptr_grow (struct obstack *obstack, void *data); +void obstack_int_grow (struct obstack *obstack, int data); + +void * obstack_finish (struct obstack *obstack); + +int obstack_object_size (struct obstack *obstack); + +int obstack_room (struct obstack *obstack); +void obstack_make_room (struct obstack *obstack, int size); +void obstack_1grow_fast (struct obstack *obstack, int data_char); +void obstack_ptr_grow_fast (struct obstack *obstack, void *data); +void obstack_int_grow_fast (struct obstack *obstack, int data); +void obstack_blank_fast (struct obstack *obstack, int size); + +void * obstack_base (struct obstack *obstack); +void * obstack_next_free (struct obstack *obstack); +int obstack_alignment_mask (struct obstack *obstack); +int obstack_chunk_size (struct obstack *obstack); +int obstack_memory_used (struct obstack *obstack); + +#endif /* __STDC__ */ + +/* Non-ANSI C cannot really support alternative functions for these macros, + so we do not declare them. */ + +/* Error handler called when `obstack_chunk_alloc' failed to allocate + more memory. This can be set to a user defined function which + should either abort gracefully or use longjump - but shouldn't + return. The default action is to print a message and abort. */ +#if defined __STDC__ && __STDC__ +extern void (*obstack_alloc_failed_handler) (void); +#else +extern void (*obstack_alloc_failed_handler) (); +#endif + +/* Exit value used when `print_and_abort' is used. */ +extern int obstack_exit_failure; + +/* Pointer to beginning of object being allocated or to be allocated next. + Note that this might not be the final address of the object + because a new chunk might be needed to hold the final size. */ + +#define obstack_base(h) ((h)->object_base) + +/* Size for allocating ordinary chunks. */ + +#define obstack_chunk_size(h) ((h)->chunk_size) + +/* Pointer to next byte not yet allocated in current chunk. */ + +#define obstack_next_free(h) ((h)->next_free) + +/* Mask specifying low bits that should be clear in address of an object. */ + +#define obstack_alignment_mask(h) ((h)->alignment_mask) + +/* To prevent prototype warnings provide complete argument list in + standard C version. */ +#if defined __STDC__ && __STDC__ + +# define obstack_init(h) \ + _obstack_begin ((h), 0, 0, \ + (void *(*) (long)) obstack_chunk_alloc, (void (*) (void *)) obstack_chunk_free) + +# define obstack_begin(h, size) \ + _obstack_begin ((h), (size), 0, \ + (void *(*) (long)) obstack_chunk_alloc, (void (*) (void *)) obstack_chunk_free) + +# define obstack_specify_allocation(h, size, alignment, chunkfun, freefun) \ + _obstack_begin ((h), (size), (alignment), \ + (void *(*) (long)) (chunkfun), (void (*) (void *)) (freefun)) + +# define obstack_specify_allocation_with_arg(h, size, alignment, chunkfun, freefun, arg) \ + _obstack_begin_1 ((h), (size), (alignment), \ + (void *(*) (void *, long)) (chunkfun), \ + (void (*) (void *, void *)) (freefun), (arg)) + +# define obstack_chunkfun(h, newchunkfun) \ + ((h) -> chunkfun = (struct _obstack_chunk *(*)(void *, long)) (newchunkfun)) + +# define obstack_freefun(h, newfreefun) \ + ((h) -> freefun = (void (*)(void *, struct _obstack_chunk *)) (newfreefun)) + +#else + +# define obstack_init(h) \ + _obstack_begin ((h), 0, 0, \ + (void *(*) ()) obstack_chunk_alloc, (void (*) ()) obstack_chunk_free) + +# define obstack_begin(h, size) \ + _obstack_begin ((h), (size), 0, \ + (void *(*) ()) obstack_chunk_alloc, (void (*) ()) obstack_chunk_free) + +# define obstack_specify_allocation(h, size, alignment, chunkfun, freefun) \ + _obstack_begin ((h), (size), (alignment), \ + (void *(*) ()) (chunkfun), (void (*) ()) (freefun)) + +# define obstack_specify_allocation_with_arg(h, size, alignment, chunkfun, freefun, arg) \ + _obstack_begin_1 ((h), (size), (alignment), \ + (void *(*) ()) (chunkfun), (void (*) ()) (freefun), (arg)) + +# define obstack_chunkfun(h, newchunkfun) \ + ((h) -> chunkfun = (struct _obstack_chunk *(*)()) (newchunkfun)) + +# define obstack_freefun(h, newfreefun) \ + ((h) -> freefun = (void (*)()) (newfreefun)) + +#endif + +#define obstack_1grow_fast(h,achar) (*((h)->next_free)++ = achar) + +#define obstack_blank_fast(h,n) ((h)->next_free += (n)) + +#define obstack_memory_used(h) _obstack_memory_used (h) + +#if defined __GNUC__ && defined __STDC__ && __STDC__ +/* NextStep 2.0 cc is really gcc 1.93 but it defines __GNUC__ = 2 and + does not implement __extension__. But that compiler doesn't define + __GNUC_MINOR__. */ +# if __GNUC__ < 2 || (__NeXT__ && !__GNUC_MINOR__) +# define __extension__ +# endif + +/* For GNU C, if not -traditional, + we can define these macros to compute all args only once + without using a global variable. + Also, we can avoid using the `temp' slot, to make faster code. */ + +# define obstack_object_size(OBSTACK) \ + __extension__ \ + ({ struct obstack *__o = (OBSTACK); \ + (unsigned) (__o->next_free - __o->object_base); }) + +# define obstack_room(OBSTACK) \ + __extension__ \ + ({ struct obstack *__o = (OBSTACK); \ + (unsigned) (__o->chunk_limit - __o->next_free); }) + +# define obstack_make_room(OBSTACK,length) \ +__extension__ \ +({ struct obstack *__o = (OBSTACK); \ + int __len = (length); \ + if (__o->chunk_limit - __o->next_free < __len) \ + _obstack_newchunk (__o, __len); \ + (void) 0; }) + +# define obstack_empty_p(OBSTACK) \ + __extension__ \ + ({ struct obstack *__o = (OBSTACK); \ + (__o->chunk->prev == 0 && __o->next_free - __o->chunk->contents == 0); }) + +# define obstack_grow(OBSTACK,where,length) \ +__extension__ \ +({ struct obstack *__o = (OBSTACK); \ + int __len = (length); \ + if (__o->next_free + __len > __o->chunk_limit) \ + _obstack_newchunk (__o, __len); \ + _obstack_memcpy (__o->next_free, (char *) (where), __len); \ + __o->next_free += __len; \ + (void) 0; }) + +# define obstack_grow0(OBSTACK,where,length) \ +__extension__ \ +({ struct obstack *__o = (OBSTACK); \ + int __len = (length); \ + if (__o->next_free + __len + 1 > __o->chunk_limit) \ + _obstack_newchunk (__o, __len + 1); \ + _obstack_memcpy (__o->next_free, (char *) (where), __len); \ + __o->next_free += __len; \ + *(__o->next_free)++ = 0; \ + (void) 0; }) + +# define obstack_1grow(OBSTACK,datum) \ +__extension__ \ +({ struct obstack *__o = (OBSTACK); \ + if (__o->next_free + 1 > __o->chunk_limit) \ + _obstack_newchunk (__o, 1); \ + *(__o->next_free)++ = (datum); \ + (void) 0; }) + +/* These assume that the obstack alignment is good enough for pointers or ints, + and that the data added so far to the current object + shares that much alignment. */ + +# define obstack_ptr_grow(OBSTACK,datum) \ +__extension__ \ +({ struct obstack *__o = (OBSTACK); \ + if (__o->next_free + sizeof (void *) > __o->chunk_limit) \ + _obstack_newchunk (__o, sizeof (void *)); \ + *((void **)__o->next_free)++ = ((void *)datum); \ + (void) 0; }) + +# define obstack_int_grow(OBSTACK,datum) \ +__extension__ \ +({ struct obstack *__o = (OBSTACK); \ + if (__o->next_free + sizeof (int) > __o->chunk_limit) \ + _obstack_newchunk (__o, sizeof (int)); \ + *((int *)__o->next_free)++ = ((int)datum); \ + (void) 0; }) + +# define obstack_ptr_grow_fast(h,aptr) (*((void **) (h)->next_free)++ = (void *)aptr) +# define obstack_int_grow_fast(h,aint) (*((int *) (h)->next_free)++ = (int) aint) + +# define obstack_blank(OBSTACK,length) \ +__extension__ \ +({ struct obstack *__o = (OBSTACK); \ + int __len = (length); \ + if (__o->chunk_limit - __o->next_free < __len) \ + _obstack_newchunk (__o, __len); \ + __o->next_free += __len; \ + (void) 0; }) + +# define obstack_alloc(OBSTACK,length) \ +__extension__ \ +({ struct obstack *__h = (OBSTACK); \ + obstack_blank (__h, (length)); \ + obstack_finish (__h); }) + +# define obstack_copy(OBSTACK,where,length) \ +__extension__ \ +({ struct obstack *__h = (OBSTACK); \ + obstack_grow (__h, (where), (length)); \ + obstack_finish (__h); }) + +# define obstack_copy0(OBSTACK,where,length) \ +__extension__ \ +({ struct obstack *__h = (OBSTACK); \ + obstack_grow0 (__h, (where), (length)); \ + obstack_finish (__h); }) + +/* The local variable is named __o1 to avoid a name conflict + when obstack_blank is called. */ +# define obstack_finish(OBSTACK) \ +__extension__ \ +({ struct obstack *__o1 = (OBSTACK); \ + void *value; \ + value = (void *) __o1->object_base; \ + if (__o1->next_free == value) \ + __o1->maybe_empty_object = 1; \ + __o1->next_free \ + = __INT_TO_PTR ((__PTR_TO_INT (__o1->next_free)+__o1->alignment_mask)\ + & ~ (__o1->alignment_mask)); \ + if (__o1->next_free - (char *)__o1->chunk \ + > __o1->chunk_limit - (char *)__o1->chunk) \ + __o1->next_free = __o1->chunk_limit; \ + __o1->object_base = __o1->next_free; \ + value; }) + +# define obstack_free(OBSTACK, OBJ) \ +__extension__ \ +({ struct obstack *__o = (OBSTACK); \ + void *__obj = (OBJ); \ + if (__obj > (void *)__o->chunk && __obj < (void *)__o->chunk_limit) \ + __o->next_free = __o->object_base = (char *)__obj; \ + else (obstack_free) (__o, __obj); }) + +#else /* not __GNUC__ or not __STDC__ */ + +# define obstack_object_size(h) \ + (unsigned) ((h)->next_free - (h)->object_base) + +# define obstack_room(h) \ + (unsigned) ((h)->chunk_limit - (h)->next_free) + +# define obstack_empty_p(h) \ + ((h)->chunk->prev == 0 && (h)->next_free - (h)->chunk->contents == 0) + +/* Note that the call to _obstack_newchunk is enclosed in (..., 0) + so that we can avoid having void expressions + in the arms of the conditional expression. + Casting the third operand to void was tried before, + but some compilers won't accept it. */ + +# define obstack_make_room(h,length) \ +( (h)->temp = (length), \ + (((h)->next_free + (h)->temp > (h)->chunk_limit) \ + ? (_obstack_newchunk ((h), (h)->temp), 0) : 0)) + +# define obstack_grow(h,where,length) \ +( (h)->temp = (length), \ + (((h)->next_free + (h)->temp > (h)->chunk_limit) \ + ? (_obstack_newchunk ((h), (h)->temp), 0) : 0), \ + _obstack_memcpy ((h)->next_free, (char *) (where), (h)->temp), \ + (h)->next_free += (h)->temp) + +# define obstack_grow0(h,where,length) \ +( (h)->temp = (length), \ + (((h)->next_free + (h)->temp + 1 > (h)->chunk_limit) \ + ? (_obstack_newchunk ((h), (h)->temp + 1), 0) : 0), \ + _obstack_memcpy ((h)->next_free, (char *) (where), (h)->temp), \ + (h)->next_free += (h)->temp, \ + *((h)->next_free)++ = 0) + +# define obstack_1grow(h,datum) \ +( (((h)->next_free + 1 > (h)->chunk_limit) \ + ? (_obstack_newchunk ((h), 1), 0) : 0), \ + (*((h)->next_free)++ = (datum))) + +# define obstack_ptr_grow(h,datum) \ +( (((h)->next_free + sizeof (char *) > (h)->chunk_limit) \ + ? (_obstack_newchunk ((h), sizeof (char *)), 0) : 0), \ + (*((char **) (((h)->next_free+=sizeof(char *))-sizeof(char *))) = ((char *) datum))) + +# define obstack_int_grow(h,datum) \ +( (((h)->next_free + sizeof (int) > (h)->chunk_limit) \ + ? (_obstack_newchunk ((h), sizeof (int)), 0) : 0), \ + (*((int *) (((h)->next_free+=sizeof(int))-sizeof(int))) = ((int) datum))) + +# define obstack_ptr_grow_fast(h,aptr) (*((char **) (h)->next_free)++ = (char *) aptr) +# define obstack_int_grow_fast(h,aint) (*((int *) (h)->next_free)++ = (int) aint) + +# define obstack_blank(h,length) \ +( (h)->temp = (length), \ + (((h)->chunk_limit - (h)->next_free < (h)->temp) \ + ? (_obstack_newchunk ((h), (h)->temp), 0) : 0), \ + ((h)->next_free += (h)->temp)) + +# define obstack_alloc(h,length) \ + (obstack_blank ((h), (length)), obstack_finish ((h))) + +# define obstack_copy(h,where,length) \ + (obstack_grow ((h), (where), (length)), obstack_finish ((h))) + +# define obstack_copy0(h,where,length) \ + (obstack_grow0 ((h), (where), (length)), obstack_finish ((h))) + +# define obstack_finish(h) \ +( ((h)->next_free == (h)->object_base \ + ? (((h)->maybe_empty_object = 1), 0) \ + : 0), \ + (h)->temp = __PTR_TO_INT ((h)->object_base), \ + (h)->next_free \ + = __INT_TO_PTR ((__PTR_TO_INT ((h)->next_free)+(h)->alignment_mask) \ + & ~ ((h)->alignment_mask)), \ + (((h)->next_free - (char *) (h)->chunk \ + > (h)->chunk_limit - (char *) (h)->chunk) \ + ? ((h)->next_free = (h)->chunk_limit) : 0), \ + (h)->object_base = (h)->next_free, \ + __INT_TO_PTR ((h)->temp)) + +# if defined __STDC__ && __STDC__ +# define obstack_free(h,obj) \ +( (h)->temp = (char *) (obj) - (char *) (h)->chunk, \ + (((h)->temp > 0 && (h)->temp < (h)->chunk_limit - (char *) (h)->chunk)\ + ? (int) ((h)->next_free = (h)->object_base \ + = (h)->temp + (char *) (h)->chunk) \ + : (((obstack_free) ((h), (h)->temp + (char *) (h)->chunk), 0), 0))) +# else +# define obstack_free(h,obj) \ +( (h)->temp = (char *) (obj) - (char *) (h)->chunk, \ + (((h)->temp > 0 && (h)->temp < (h)->chunk_limit - (char *) (h)->chunk)\ + ? (int) ((h)->next_free = (h)->object_base \ + = (h)->temp + (char *) (h)->chunk) \ + : (_obstack_free ((h), (h)->temp + (char *) (h)->chunk), 0))) +# endif + +#endif /* not __GNUC__ or not __STDC__ */ + +#ifdef __cplusplus +} /* C++ */ +#endif + +#endif /* obstack.h */ diff --git a/lib/posix/Makefile.am b/lib/posix/Makefile.am new file mode 100644 index 00000000..610ef229 --- /dev/null +++ b/lib/posix/Makefile.am @@ -0,0 +1,4 @@ +## Process this file with automake to create Makefile.in +AUTOMAKE_OPTIONS=no-dependencies + +EXTRA_DIST = regex.h diff --git a/lib/posix/regex.h b/lib/posix/regex.h new file mode 100644 index 00000000..d88ab92b --- /dev/null +++ b/lib/posix/regex.h @@ -0,0 +1,542 @@ +/* Definitions for data structures and routines for the regular + expression library, version 0.12. + Copyright (C) 1985,89,90,91,92,93,95,96,97,98 Free Software Foundation, Inc. + + This file is part of the GNU C Library. Its master source is NOT part of + the C library, however. The master source lives in /gd/gnu/lib. + + 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., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#ifndef _REGEX_H +#define _REGEX_H 1 + +/* Allow the use in C++ code. */ +#ifdef __cplusplus +extern "C" { +#endif + +/* POSIX says that <sys/types.h> must be included (by the caller) before + <regex.h>. */ + +#if !defined _POSIX_C_SOURCE && !defined _POSIX_SOURCE && defined 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 two types have to be signed and unsigned integer type + wide enough to hold a value of a pointer. For most ANSI compilers + ptrdiff_t and size_t should be likely OK. Still size of these two + types is 2 for Microsoft C. Ugh... */ +typedef long int s_reg_t; +typedef unsigned long int active_reg_t; + +/* The following bits are used to determine the regexp syntax we + recognize. The set/not-set meanings are chosen so that Emacs syntax + remains the value 0. The bits are given in alphabetical order, and + the definitions shifted by one from the previous bit; thus, when we + add or remove a bit, only one other definition need change. */ +typedef unsigned long int reg_syntax_t; + +/* If this bit is not set, then \ inside a bracket expression is literal. + If set, then such a \ quotes the following character. */ +#define RE_BACKSLASH_ESCAPE_IN_LISTS ((unsigned long int) 1) + +/* If this bit is not set, then + and ? are operators, and \+ and \? are + literals. + If set, then \+ and \? are operators and + and ? are literals. */ +#define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1) + +/* If this bit is set, then character classes are supported. They are: + [:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:], + [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:]. + If not set, then character classes are not supported. */ +#define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1) + +/* If this bit is set, then ^ and $ are always anchors (outside bracket + expressions, of course). + If this bit is not set, then it depends: + ^ is an anchor if it is at the beginning of a regular + expression or after an open-group or an alternation operator; + $ is an anchor if it is at the end of a regular expression, or + before a close-group or an alternation operator. + + This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because + POSIX draft 11.2 says that * etc. in leading positions is undefined. + We already implemented a previous draft which made those constructs + invalid, though, so we haven't changed the code back. */ +#define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1) + +/* If this bit is set, then special characters are always special + regardless of where they are in the pattern. + If this bit is not set, then special characters are special only in + some contexts; otherwise they are ordinary. Specifically, + * + ? and intervals are only special when not after the beginning, + open-group, or alternation operator. */ +#define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1) + +/* If this bit is set, then *, +, ?, and { cannot be first in an re or + immediately after an alternation or begin-group operator. */ +#define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1) + +/* If this bit is set, then . matches newline. + If not set, then it doesn't. */ +#define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1) + +/* If this bit is set, then . doesn't match NUL. + If not set, then it does. */ +#define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1) + +/* If this bit is set, nonmatching lists [^...] do not match newline. + If not set, they do. */ +#define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1) + +/* If this bit is set, either \{...\} or {...} defines an + interval, depending on RE_NO_BK_BRACES. + If not set, \{, \}, {, and } are literals. */ +#define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1) + +/* If this bit is set, +, ? and | aren't recognized as operators. + If not set, they are. */ +#define RE_LIMITED_OPS (RE_INTERVALS << 1) + +/* If this bit is set, newline is an alternation operator. + If not set, newline is literal. */ +#define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1) + +/* If this bit is set, then `{...}' defines an interval, and \{ and \} + are literals. + If not set, then `\{...\}' defines an interval. */ +#define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1) + +/* If this bit is set, (...) defines a group, and \( and \) are literals. + If not set, \(...\) defines a group, and ( and ) are literals. */ +#define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1) + +/* If this bit is set, then \<digit> matches <digit>. + If not set, then \<digit> is a back-reference. */ +#define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1) + +/* If this bit is set, then | is an alternation operator, and \| is literal. + If not set, then \| is an alternation operator, and | is literal. */ +#define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1) + +/* If this bit is set, then an ending range point collating higher + than the starting range point, as in [z-a], is invalid. + If not set, then when ending range point collates higher than the + starting range point, the range is ignored. */ +#define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1) + +/* If this bit is set, then an unmatched ) is ordinary. + If not set, then an unmatched ) is invalid. */ +#define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1) + +/* If this bit is set, succeed as soon as we match the whole pattern, + without further backtracking. */ +#define RE_NO_POSIX_BACKTRACKING (RE_UNMATCHED_RIGHT_PAREN_ORD << 1) + +/* If this bit is set, do not process the GNU regex operators. + If not set, then the GNU regex operators are recognized. */ +#define RE_NO_GNU_OPS (RE_NO_POSIX_BACKTRACKING << 1) + +/* If this bit is set, turn on internal regex debugging. + If not set, and debugging was on, turn it off. + This only works if regex.c is compiled -DDEBUG. + We define this bit always, so that all that's needed to turn on + debugging is to recompile regex.c; the calling code can always have + this bit set, and it won't affect anything in the normal case. */ +#define RE_DEBUG (RE_NO_GNU_OPS << 1) + +/* 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_DOT_NEWLINE | RE_CONTEXT_INDEP_ANCHORS \ + | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_GNU_OPS) + +#define RE_SYNTAX_GNU_AWK \ + ((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DEBUG) \ + & ~(RE_DOT_NOT_NULL | RE_INTERVALS | RE_CONTEXT_INDEP_OPS)) + +#define RE_SYNTAX_POSIX_AWK \ + (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \ + | RE_INTERVALS | RE_NO_GNU_OPS) + +#define RE_SYNTAX_GREP \ + (RE_BK_PLUS_QM | RE_CHAR_CLASSES \ + | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS \ + | RE_NEWLINE_ALT) + +#define RE_SYNTAX_EGREP \ + (RE_CHAR_CLASSES | RE_CONTEXT_INDEP_ANCHORS \ + | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE \ + | RE_NEWLINE_ALT | RE_NO_BK_PARENS \ + | RE_NO_BK_VBAR) + +#define RE_SYNTAX_POSIX_EGREP \ + (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES) + +/* 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 +/* If sizeof(int) == 2, then ((1 << 15) - 1) overflows. */ +#define RE_DUP_MAX (0x7fff) + + +/* 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 +{ +#ifdef _XOPEN_SOURCE + REG_ENOSYS = -1, /* This will never happen for this implementation. */ +#endif + + 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. */ + +#ifndef RE_TRANSLATE_TYPE +# define RE_TRANSLATE_TYPE char * +#endif + +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 int allocated; + + /* Number of bytes actually used in `buffer'. */ + unsigned long int 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. */ + RE_TRANSLATE_TYPE 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, size_t length, + struct re_pattern_buffer *buffer)); + + +/* Compile a fastmap for the compiled pattern in BUFFER; used to + accelerate searches. Return 0 if successful and -2 if was an + internal error. */ +extern int re_compile_fastmap _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)); + +#if defined _REGEX_RE_COMP || defined _LIBC +# ifndef _CRAY +/* 4.2 bsd compatibility. */ +extern char *re_comp _RE_ARGS ((const char *)); +extern int re_exec _RE_ARGS ((const char *)); +# endif +#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)); + + +#ifdef __cplusplus +} +#endif /* C++ */ + +#endif /* regex.h */ + +/* +Local variables: +make-backup-files: t +version-control: t +trim-versions-without-asking: nil +End: +*/ diff --git a/lib/quotearg.c b/lib/quotearg.c new file mode 100644 index 00000000..a668a68b --- /dev/null +++ b/lib/quotearg.c @@ -0,0 +1,619 @@ +/* quotearg.c - quote arguments for output + Copyright (C) 1998, 1999, 2000, 2001 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert <eggert@twinsun.com> */ + +#if HAVE_CONFIG_H +# include <config.h> +#endif + +#if HAVE_STDDEF_H +# include <stddef.h> /* For the definition of size_t on windows w/MSVC. */ +#endif +#include <sys/types.h> +#include <quotearg.h> +#include <xalloc.h> + +#include <ctype.h> + +#if ENABLE_NLS +# include <libintl.h> +# define _(text) gettext (text) +#else +# define _(text) text +#endif +#define N_(text) text + +#if HAVE_LIMITS_H +# include <limits.h> +#endif +#ifndef CHAR_BIT +# define CHAR_BIT 8 +#endif +#ifndef UCHAR_MAX +# define UCHAR_MAX ((unsigned char) -1) +#endif + +#if HAVE_C_BACKSLASH_A +# define ALERT_CHAR '\a' +#else +# define ALERT_CHAR '\7' +#endif + +#if HAVE_STDLIB_H +# include <stdlib.h> +#endif + +#if HAVE_STRING_H +# include <string.h> +#endif + +#if HAVE_WCHAR_H +# include <wchar.h> +#endif + +#if HAVE_MBRTOWC +size_t mbrtowc (); +# ifdef mbstate_t +# define mbrtowc(pwc, s, n, ps) (mbrtowc) (pwc, s, n, 0) +# define mbsinit(ps) 1 +# endif +#else +/* Disable multibyte processing entirely. Since MB_CUR_MAX is 1, the + other macros are defined only for documentation and to satisfy C + syntax. */ +# undef MB_CUR_MAX +# define MB_CUR_MAX 1 +# define mbrtowc(pwc, s, n, ps) ((*(pwc) = *(s)) != 0) +# define mbsinit(ps) 1 +# define iswprint(wc) ISPRINT ((unsigned char) (wc)) +#endif + +#ifndef iswprint +# if HAVE_WCTYPE_H +# include <wctype.h> +# endif +# if !defined iswprint && !HAVE_ISWPRINT +# define iswprint(wc) 1 +# endif +#endif + +#define INT_BITS (sizeof (int) * CHAR_BIT) + +#if defined (STDC_HEADERS) || (!defined (isascii) && !defined (HAVE_ISASCII)) +# define IN_CTYPE_DOMAIN(c) 1 +#else +# define IN_CTYPE_DOMAIN(c) isascii(c) +#endif + +/* Undefine to protect against the definition in wctype.h of solaris2.6. */ +#undef ISPRINT +#define ISPRINT(c) (IN_CTYPE_DOMAIN (c) && isprint (c)) + +struct quoting_options +{ + /* Basic quoting style. */ + enum quoting_style style; + + /* Quote the characters indicated by this bit vector even if the + quoting style would not normally require them to be quoted. */ + int quote_these_too[(UCHAR_MAX / INT_BITS) + 1]; +}; + +/* Names of quoting styles. */ +char const *const quoting_style_args[] = +{ + "literal", + "shell", + "shell-always", + "c", + "escape", + "locale", + "clocale", + 0 +}; + +/* Correspondences to quoting style names. */ +enum quoting_style const quoting_style_vals[] = +{ + literal_quoting_style, + shell_quoting_style, + shell_always_quoting_style, + c_quoting_style, + escape_quoting_style, + locale_quoting_style, + clocale_quoting_style +}; + +/* The default quoting options. */ +static struct quoting_options default_quoting_options; + +/* Allocate a new set of quoting options, with contents initially identical + to O if O is not null, or to the default if O is null. + It is the caller's responsibility to free the result. */ +struct quoting_options * +clone_quoting_options (struct quoting_options *o) +{ + struct quoting_options *p + = (struct quoting_options *) xmalloc (sizeof (struct quoting_options)); + *p = *(o ? o : &default_quoting_options); + return p; +} + +/* Get the value of O's quoting style. If O is null, use the default. */ +enum quoting_style +get_quoting_style (struct quoting_options *o) +{ + return (o ? o : &default_quoting_options)->style; +} + +/* In O (or in the default if O is null), + set the value of the quoting style to S. */ +void +set_quoting_style (struct quoting_options *o, enum quoting_style s) +{ + (o ? o : &default_quoting_options)->style = s; +} + +/* In O (or in the default if O is null), + set the value of the quoting options for character C to I. + Return the old value. Currently, the only values defined for I are + 0 (the default) and 1 (which means to quote the character even if + it would not otherwise be quoted). */ +int +set_char_quoting (struct quoting_options *o, char c, int i) +{ + unsigned char uc = c; + int *p = (o ? o : &default_quoting_options)->quote_these_too + uc / INT_BITS; + int shift = uc % INT_BITS; + int r = (*p >> shift) & 1; + *p ^= ((i & 1) ^ r) << shift; + return r; +} + +/* MSGID approximates a quotation mark. Return its translation if it + has one; otherwise, return either it or "\"", depending on S. */ +static char const * +gettext_quote (char const *msgid, enum quoting_style s) +{ + char const *translation = _(msgid); + if (translation == msgid && s == clocale_quoting_style) + translation = "\""; + return translation; +} + +/* Place into buffer BUFFER (of size BUFFERSIZE) a quoted version of + argument ARG (of size ARGSIZE), using QUOTING_STYLE and the + non-quoting-style part of O to control quoting. + Terminate the output with a null character, and return the written + size of the output, not counting the terminating null. + If BUFFERSIZE is too small to store the output string, return the + value that would have been returned had BUFFERSIZE been large enough. + If ARGSIZE is -1, use the string length of the argument for ARGSIZE. + + This function acts like quotearg_buffer (BUFFER, BUFFERSIZE, ARG, + ARGSIZE, O), except it uses QUOTING_STYLE instead of the quoting + style specified by O, and O may not be null. */ + +static size_t +quotearg_buffer_restyled (char *buffer, size_t buffersize, + char const *arg, size_t argsize, + enum quoting_style quoting_style, + struct quoting_options const *o) +{ + size_t i; + size_t len = 0; + char const *quote_string = 0; + size_t quote_string_len = 0; + int backslash_escapes = 0; + int unibyte_locale = MB_CUR_MAX == 1; + +#define STORE(c) \ + do \ + { \ + if (len < buffersize) \ + buffer[len] = (c); \ + len++; \ + } \ + while (0) + + switch (quoting_style) + { + case c_quoting_style: + STORE ('"'); + backslash_escapes = 1; + quote_string = "\""; + quote_string_len = 1; + break; + + case escape_quoting_style: + backslash_escapes = 1; + break; + + case locale_quoting_style: + case clocale_quoting_style: + { + /* Get translations for open and closing quotation marks. + + The message catalog should translate "`" to a left + quotation mark suitable for the locale, and similarly for + "'". If the catalog has no translation, + locale_quoting_style quotes `like this', and + clocale_quoting_style quotes "like this". + + For example, an American English Unicode locale should + translate "`" to U+201C (LEFT DOUBLE QUOTATION MARK), and + should translate "'" to U+201D (RIGHT DOUBLE QUOTATION + MARK). A British English Unicode locale should instead + translate these to U+2018 (LEFT SINGLE QUOTATION MARK) and + U+2019 (RIGHT SINGLE QUOTATION MARK), respectively. */ + + char const *left = gettext_quote (N_("`"), quoting_style); + char const *right = gettext_quote (N_("'"), quoting_style); + for (quote_string = left; *quote_string; quote_string++) + STORE (*quote_string); + backslash_escapes = 1; + quote_string = right; + quote_string_len = strlen (quote_string); + } + break; + + case shell_always_quoting_style: + STORE ('\''); + quote_string = "'"; + quote_string_len = 1; + break; + + default: + break; + } + + for (i = 0; ! (argsize == (size_t) -1 ? arg[i] == '\0' : i == argsize); i++) + { + unsigned char c; + unsigned char esc; + + if (backslash_escapes + && quote_string_len + && i + quote_string_len <= argsize + && memcmp (arg + i, quote_string, quote_string_len) == 0) + STORE ('\\'); + + c = arg[i]; + switch (c) + { + case '?': + switch (quoting_style) + { + case shell_quoting_style: + goto use_shell_always_quoting_style; + + case c_quoting_style: + if (i + 2 < argsize && arg[i + 1] == '?') + switch (arg[i + 2]) + { + case '!': case '\'': + case '(': case ')': case '-': case '/': + case '<': case '=': case '>': + /* Escape the second '?' in what would otherwise be + a trigraph. */ + i += 2; + c = arg[i + 2]; + STORE ('?'); + STORE ('\\'); + STORE ('?'); + break; + } + break; + + default: + break; + } + break; + + case ALERT_CHAR: esc = 'a'; goto c_escape; + case '\b': esc = 'b'; goto c_escape; + case '\f': esc = 'f'; goto c_escape; + case '\n': esc = 'n'; goto c_and_shell_escape; + case '\r': esc = 'r'; goto c_and_shell_escape; + case '\t': esc = 't'; goto c_and_shell_escape; + case '\v': esc = 'v'; goto c_escape; + case '\\': esc = c; goto c_and_shell_escape; + + c_and_shell_escape: + if (quoting_style == shell_quoting_style) + goto use_shell_always_quoting_style; + c_escape: + if (backslash_escapes) + { + c = esc; + goto store_escape; + } + break; + + case '#': case '~': + if (i != 0) + break; + /* Fall through. */ + case ' ': + case '!': /* special in bash */ + case '"': case '$': case '&': + case '(': case ')': case '*': case ';': + case '<': case '>': case '[': + case '^': /* special in old /bin/sh, e.g. SunOS 4.1.4 */ + case '`': case '|': + /* A shell special character. In theory, '$' and '`' could + be the first bytes of multibyte characters, which means + we should check them with mbrtowc, but in practice this + doesn't happen so it's not worth worrying about. */ + if (quoting_style == shell_quoting_style) + goto use_shell_always_quoting_style; + break; + + case '\'': + switch (quoting_style) + { + case shell_quoting_style: + goto use_shell_always_quoting_style; + + case shell_always_quoting_style: + STORE ('\''); + STORE ('\\'); + STORE ('\''); + break; + + default: + break; + } + break; + + case '%': case '+': case ',': case '-': case '.': case '/': + case '0': case '1': case '2': case '3': case '4': case '5': + case '6': case '7': case '8': case '9': case ':': case '=': + case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': + case 'G': case 'H': case 'I': case 'J': case 'K': case 'L': + case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R': + case 'S': case 'T': case 'U': case 'V': case 'W': case 'X': + case 'Y': case 'Z': case ']': case '_': case 'a': case 'b': + case 'c': case 'd': case 'e': case 'f': case 'g': case 'h': + case 'i': case 'j': case 'k': case 'l': case 'm': case 'n': + case 'o': case 'p': case 'q': case 'r': case 's': case 't': + case 'u': case 'v': case 'w': case 'x': case 'y': case 'z': + case '{': case '}': + /* These characters don't cause problems, no matter what the + quoting style is. They cannot start multibyte sequences. */ + break; + + default: + /* If we have a multibyte sequence, copy it until we reach + its end, find an error, or come back to the initial shift + state. For C-like styles, if the sequence has + unprintable characters, escape the whole sequence, since + we can't easily escape single characters within it. */ + { + /* Length of multibyte sequence found so far. */ + size_t m; + + int printable; + + if (unibyte_locale) + { + m = 1; + printable = ISPRINT (c); + } + else + { + mbstate_t mbstate; + memset (&mbstate, 0, sizeof mbstate); + + m = 0; + printable = 1; + if (argsize == (size_t) -1) + argsize = strlen (arg); + + do + { + wchar_t w; + size_t bytes = mbrtowc (&w, &arg[i + m], + argsize - (i + m), &mbstate); + if (bytes == 0) + break; + else if (bytes == (size_t) -1) + { + printable = 0; + break; + } + else if (bytes == (size_t) -2) + { + printable = 0; + while (i + m < argsize && arg[i + m]) + m++; + break; + } + else + { + if (! iswprint (w)) + printable = 0; + m += bytes; + } + } + while (! mbsinit (&mbstate)); + } + + if (1 < m || (backslash_escapes && ! printable)) + { + /* Output a multibyte sequence, or an escaped + unprintable unibyte character. */ + size_t ilim = i + m; + + for (;;) + { + if (backslash_escapes && ! printable) + { + STORE ('\\'); + STORE ('0' + (c >> 6)); + STORE ('0' + ((c >> 3) & 7)); + c = '0' + (c & 7); + } + if (ilim <= i + 1) + break; + STORE (c); + c = arg[++i]; + } + + goto store_c; + } + } + } + + if (! (backslash_escapes + && o->quote_these_too[c / INT_BITS] & (1 << (c % INT_BITS)))) + goto store_c; + + store_escape: + STORE ('\\'); + + store_c: + STORE (c); + } + + if (quote_string) + for (; *quote_string; quote_string++) + STORE (*quote_string); + + if (len < buffersize) + buffer[len] = '\0'; + return len; + + use_shell_always_quoting_style: + return quotearg_buffer_restyled (buffer, buffersize, arg, argsize, + shell_always_quoting_style, o); +} + +/* Place into buffer BUFFER (of size BUFFERSIZE) a quoted version of + argument ARG (of size ARGSIZE), using O to control quoting. + If O is null, use the default. + Terminate the output with a null character, and return the written + size of the output, not counting the terminating null. + If BUFFERSIZE is too small to store the output string, return the + value that would have been returned had BUFFERSIZE been large enough. + If ARGSIZE is -1, use the string length of the argument for ARGSIZE. */ +size_t +quotearg_buffer (char *buffer, size_t buffersize, + char const *arg, size_t argsize, + struct quoting_options const *o) +{ + struct quoting_options const *p = o ? o : &default_quoting_options; + return quotearg_buffer_restyled (buffer, buffersize, arg, argsize, + p->style, p); +} + +/* Use storage slot N to return a quoted version of the string ARG. + OPTIONS specifies the quoting options. + The returned value points to static storage that can be + reused by the next call to this function with the same value of N. + N must be nonnegative. N is deliberately declared with type "int" + to allow for future extensions (using negative values). */ +static char * +quotearg_n_options (int n, char const *arg, + struct quoting_options const *options) +{ + /* Preallocate a slot 0 buffer, so that the caller can always quote + one small component of a "memory exhausted" message in slot 0. */ + static char slot0[256]; + static unsigned int nslots = 1; + struct slotvec + { + size_t size; + char *val; + }; + static struct slotvec slotvec0 = {sizeof slot0, slot0}; + static struct slotvec *slotvec = &slotvec0; + + if (nslots <= n) + { + int n1 = n + 1; + size_t s = n1 * sizeof (struct slotvec); + if (! (0 < n1 && n1 == s / sizeof (struct slotvec))) + abort (); + if (slotvec == &slotvec0) + { + slotvec = (struct slotvec *) xmalloc (sizeof (struct slotvec)); + *slotvec = slotvec0; + } + slotvec = (struct slotvec *) xrealloc (slotvec, s); + memset (slotvec + nslots, 0, (n1 - nslots) * sizeof (struct slotvec)); + nslots = n; + } + + { + size_t size = slotvec[n].size; + char *val = slotvec[n].val; + size_t qsize = quotearg_buffer (val, size, arg, (size_t) -1, options); + + if (size <= qsize) + { + slotvec[n].size = size = qsize + 1; + slotvec[n].val = val = xrealloc (val == slot0 ? 0 : val, size); + quotearg_buffer (val, size, arg, (size_t) -1, options); + } + + return val; + } +} + +char * +quotearg_n (unsigned int n, char const *arg) +{ + return quotearg_n_options (n, arg, &default_quoting_options); +} + +char * +quotearg (char const *arg) +{ + return quotearg_n (0, arg); +} + +char * +quotearg_n_style (unsigned int n, enum quoting_style s, char const *arg) +{ + struct quoting_options o; + o.style = s; + memset (o.quote_these_too, 0, sizeof o.quote_these_too); + return quotearg_n_options (n, arg, &o); +} + +char * +quotearg_style (enum quoting_style s, char const *arg) +{ + return quotearg_n_style (0, s, arg); +} + +char * +quotearg_char (char const *arg, char ch) +{ + struct quoting_options options; + options = default_quoting_options; + set_char_quoting (&options, ch, 1); + return quotearg_n_options (0, arg, &options); +} + +char * +quotearg_colon (char const *arg) +{ + return quotearg_char (arg, ':'); +} diff --git a/lib/quotearg.h b/lib/quotearg.h new file mode 100644 index 00000000..f6463b1a --- /dev/null +++ b/lib/quotearg.h @@ -0,0 +1,110 @@ +/* quotearg.h - quote arguments for output + Copyright (C) 1998, 1999, 2000 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert <eggert@twinsun.com> */ + +/* Basic quoting styles. */ +enum quoting_style + { + literal_quoting_style, /* --quoting-style=literal */ + shell_quoting_style, /* --quoting-style=shell */ + shell_always_quoting_style, /* --quoting-style=shell-always */ + c_quoting_style, /* --quoting-style=c */ + escape_quoting_style, /* --quoting-style=escape */ + locale_quoting_style, /* --quoting-style=locale */ + clocale_quoting_style /* --quoting-style=clocale */ + }; + +/* For now, --quoting-style=literal is the default, but this may change. */ +#ifndef DEFAULT_QUOTING_STYLE +# define DEFAULT_QUOTING_STYLE literal_quoting_style +#endif + +/* Names of quoting styles and their corresponding values. */ +extern char const *const quoting_style_args[]; +extern enum quoting_style const quoting_style_vals[]; + +struct quoting_options; + +#ifndef PARAMS +# if defined PROTOTYPES || defined __STDC__ +# define PARAMS(Args) Args +# else +# define PARAMS(Args) () +# endif +#endif + +/* The functions listed below set and use a hidden variable + that contains the default quoting style options. */ + +/* Allocate a new set of quoting options, with contents initially identical + to O if O is not null, or to the default if O is null. + It is the caller's responsibility to free the result. */ +struct quoting_options *clone_quoting_options + PARAMS ((struct quoting_options *o)); + +/* Get the value of O's quoting style. If O is null, use the default. */ +enum quoting_style get_quoting_style PARAMS ((struct quoting_options *o)); + +/* In O (or in the default if O is null), + set the value of the quoting style to S. */ +void set_quoting_style PARAMS ((struct quoting_options *o, + enum quoting_style s)); + +/* In O (or in the default if O is null), + set the value of the quoting options for character C to I. + Return the old value. Currently, the only values defined for I are + 0 (the default) and 1 (which means to quote the character even if + it would not otherwise be quoted). */ +int set_char_quoting PARAMS ((struct quoting_options *o, char c, int i)); + +/* Place into buffer BUFFER (of size BUFFERSIZE) a quoted version of + argument ARG (of size ARGSIZE), using O to control quoting. + If O is null, use the default. + Terminate the output with a null character, and return the written + size of the output, not counting the terminating null. + If BUFFERSIZE is too small to store the output string, return the + value that would have been returned had BUFFERSIZE been large enough. + If ARGSIZE is -1, use the string length of the argument for ARGSIZE. */ +size_t quotearg_buffer PARAMS ((char *buffer, size_t buffersize, + char const *arg, size_t argsize, + struct quoting_options const *o)); + +/* Use storage slot N to return a quoted version of the string ARG. + Use the default quoting options. + The returned value points to static storage that can be + reused by the next call to this function with the same value of N. + N must be nonnegative. */ +char *quotearg_n PARAMS ((unsigned int n, char const *arg)); + +/* Equivalent to quotearg_n (0, ARG). */ +char *quotearg PARAMS ((char const *arg)); + +/* Use style S and storage slot N to return a quoted version of the string ARG. + This is like quotearg_n (N, ARG), except that it uses S with no other + options to specify the quoting method. */ +char *quotearg_n_style PARAMS ((unsigned int n, enum quoting_style s, + char const *arg)); + +/* Equivalent to quotearg_n_style (0, S, ARG). */ +char *quotearg_style PARAMS ((enum quoting_style s, char const *arg)); + +/* Like quotearg (ARG), except also quote any instances of CH. */ +char *quotearg_char PARAMS ((char const *arg, char ch)); + +/* Equivalent to quotearg_char (ARG, ':'). */ +char *quotearg_colon PARAMS ((char const *arg)); diff --git a/lib/regex.c b/lib/regex.c new file mode 100644 index 00000000..640e3fe5 --- /dev/null +++ b/lib/regex.c @@ -0,0 +1,5885 @@ +/* Extended regular expression matching and search library, + version 0.12. + (Implements POSIX draft P1003.2/D11.2, except for some of the + internationalization features.) + Copyright (C) 1993, 94, 95, 96, 97, 98, 99 Free Software Foundation, Inc. + + 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., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +/* AIX requires this to be the first thing in the file. */ +#if defined _AIX && !defined REGEX_MALLOC + #pragma alloca +#endif + +#undef _GNU_SOURCE +#define _GNU_SOURCE + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +#ifndef PARAMS +# if defined __GNUC__ || (defined __STDC__ && __STDC__) +# define PARAMS(args) args +# else +# define PARAMS(args) () +# endif /* GCC. */ +#endif /* Not PARAMS. */ + +#if defined STDC_HEADERS && !defined emacs +# include <stddef.h> +#else +/* We need this for `regex.h', and perhaps for the Emacs include files. */ +# include <sys/types.h> +#endif + +#define WIDE_CHAR_SUPPORT (HAVE_WCTYPE_H && HAVE_WCHAR_H && HAVE_BTOWC) + +/* For platform which support the ISO C amendement 1 functionality we + support user defined character classes. */ +#if defined _LIBC || WIDE_CHAR_SUPPORT +/* Solaris 2.5 has a bug: <wchar.h> must be included before <wctype.h>. */ +# include <wchar.h> +# include <wctype.h> +#endif + +#ifdef _LIBC +/* We have to keep the namespace clean. */ +# define regfree(preg) __regfree (preg) +# define regexec(pr, st, nm, pm, ef) __regexec (pr, st, nm, pm, ef) +# define regcomp(preg, pattern, cflags) __regcomp (preg, pattern, cflags) +# define regerror(errcode, preg, errbuf, errbuf_size) \ + __regerror(errcode, preg, errbuf, errbuf_size) +# define re_set_registers(bu, re, nu, st, en) \ + __re_set_registers (bu, re, nu, st, en) +# define re_match_2(bufp, string1, size1, string2, size2, pos, regs, stop) \ + __re_match_2 (bufp, string1, size1, string2, size2, pos, regs, stop) +# define re_match(bufp, string, size, pos, regs) \ + __re_match (bufp, string, size, pos, regs) +# define re_search(bufp, string, size, startpos, range, regs) \ + __re_search (bufp, string, size, startpos, range, regs) +# define re_compile_pattern(pattern, length, bufp) \ + __re_compile_pattern (pattern, length, bufp) +# define re_set_syntax(syntax) __re_set_syntax (syntax) +# define re_search_2(bufp, st1, s1, st2, s2, startpos, range, regs, stop) \ + __re_search_2 (bufp, st1, s1, st2, s2, startpos, range, regs, stop) +# define re_compile_fastmap(bufp) __re_compile_fastmap (bufp) + +#define btowc __btowc +#endif + + +#ifndef _ +/* This is for other GNU distributions with internationalized messages. + When compiling libc, the _ and N_ macros are predefined. */ +# ifdef HAVE_LIBINTL_H +# include <libintl.h> +# else +# define gettext(msgid) (msgid) +# endif +# define N_(msgid) (msgid) +#endif + +/* 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" + +#else /* not emacs */ + +/* If we are not linking with Emacs proper, + we can't use the relocating allocator + even if config.h says that we can. */ +# undef REL_ALLOC + +# if defined STDC_HEADERS || defined _LIBC +# include <stdlib.h> +# else +char *malloc (); +char *realloc (); +# endif + +/* When used in Emacs's lib-src, we need to get bzero and bcopy somehow. + If nothing else has been done, use the method below. */ +# ifdef INHIBIT_STRING_HEADER +# if !(defined HAVE_BZERO && defined HAVE_BCOPY) +# if !defined bzero && !defined bcopy +# undef INHIBIT_STRING_HEADER +# endif +# endif +# endif + +/* This is the normal way of making sure we have a bcopy and a bzero. + This is used in most programs--a few other programs avoid this + by defining INHIBIT_STRING_HEADER. */ +# ifndef INHIBIT_STRING_HEADER +# if defined HAVE_STRING_H || defined STDC_HEADERS || defined _LIBC +# include <string.h> +# ifndef bzero +# ifndef _LIBC +# define bzero(s, n) (memset (s, '\0', n), (s)) +# else +# define bzero(s, n) __bzero (s, n) +# endif +# endif +# else +# include <strings.h> +# ifndef memcmp +# define memcmp(s1, s2, n) bcmp (s1, s2, n) +# endif +# ifndef memcpy +# define memcpy(d, s, n) (bcopy (s, d, n), (d)) +# endif +# 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 SWITCH_ENUM_BUG +# define SWITCH_ENUM_CAST(x) ((int)(x)) +# else +# define SWITCH_ENUM_CAST(x) (x) +# endif + +#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." + Solaris defines some of these symbols so we must undefine them first. */ + +#undef ISASCII +#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 + +#undef ISPRINT +#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)) + +#ifdef _tolower +# define TOLOWER(c) _tolower(c) +#else +# define TOLOWER(c) tolower(c) +#endif + +#ifndef NULL +# define NULL (void *)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 + +#ifndef emacs +/* How many characters in the character set. */ +# define CHAR_SET_SIZE 256 + +# ifdef SYNTAX_TABLE + +extern char *re_syntax_table; + +# else /* not SYNTAX_TABLE */ + +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 = 0; c < CHAR_SET_SIZE; ++c) + if (ISALNUM (c)) + re_syntax_table[c] = Sword; + + re_syntax_table['_'] = Sword; + + done = 1; +} + +# endif /* not SYNTAX_TABLE */ + +# define SYNTAX(c) re_syntax_table[((c) & 0xFF)] + +#endif /* emacs */ + +/* 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) +# define REGEX_FREE free + +#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> +# endif /* 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), \ + memcpy (destination, source, osize)) + +/* No need to do anything to free, after alloca. */ +# define REGEX_FREE(arg) ((void)0) /* Do nothing! But inhibit gcc warning. */ + +#endif /* not REGEX_MALLOC */ + +/* Define how to allocate the failure stack. */ + +#if defined REL_ALLOC && defined REGEX_MALLOC + +# define REGEX_ALLOCATE_STACK(size) \ + r_alloc (&failure_stack_ptr, (size)) +# define REGEX_REALLOCATE_STACK(source, osize, nsize) \ + r_re_alloc (&failure_stack_ptr, (nsize)) +# define REGEX_FREE_STACK(ptr) \ + r_alloc_free (&failure_stack_ptr) + +#else /* not using relocating allocator */ + +# ifdef REGEX_MALLOC + +# define REGEX_ALLOCATE_STACK malloc +# define REGEX_REALLOCATE_STACK(source, osize, nsize) realloc (source, nsize) +# define REGEX_FREE_STACK free + +# else /* not REGEX_MALLOC */ + +# define REGEX_ALLOCATE_STACK alloca + +# define REGEX_REALLOCATE_STACK(source, osize, nsize) \ + REGEX_REALLOCATE (source, osize, nsize) +/* No need to explicitly free anything. */ +# define REGEX_FREE_STACK(arg) + +# endif /* not REGEX_MALLOC */ +#endif /* not using relocating allocator */ + + +/* 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 PARAMS ((struct re_pattern_buffer *bufp, + const char *string1, int size1, + const char *string2, int size2, + int pos, + struct re_registers *regs, + int stop)); + +/* 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, + + /* Succeed right away--no more backtracking. */ + succeed, + + /* 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 _RE_ARGS ((int *dest, unsigned char *source)); +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 _RE_ARGS ((int *destination, + unsigned char **source)); +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; + +# 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) + + +/* 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; + putchar (i - 1); + while (i < (1 << BYTEWIDTH) && fastmap[i]) + { + was_a_range = 1; + i++; + } + if (was_a_range) + { + printf ("-"); + putchar (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 *p1; + 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 ('/'); + putchar (*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) + { + putchar (last); + in_range = 0; + } + + if (! in_range) + putchar (c); + + last = c; + } + + if (in_range) + putchar (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); + p1 = p + mcnt; + extract_number_and_incr (&mcnt2, &p); + printf ("/succeed_n to %d, %d times", p1 - start, mcnt2); + break; + + case jump_n: + extract_number_and_incr (&mcnt, &p); + p1 = p + mcnt; + extract_number_and_incr (&mcnt2, &p); + printf ("/jump_n to %d, %d times", p1 - start, mcnt2); + break; + + case set_number_at: + extract_number_and_incr (&mcnt, &p); + p1 = p + mcnt; + extract_number_and_incr (&mcnt2, &p); + printf ("/set_number_at location %d to %d", p1 - 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 ("%ld bytes used/%ld 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: %lx\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; +{ + int this_char; + + if (where == NULL) + printf ("(null)"); + else + { + if (FIRST_STRING_P (where)) + { + for (this_char = where - string1; this_char < size1; this_char++) + putchar (string1[this_char]); + + where = string2; + } + + for (this_char = where - string2; this_char < size2; this_char++) + putchar (string2[this_char]); + } +} + +void +printchar (c) + int c; +{ + putc (c, stderr); +} + +#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. */ +/* This has no initializer because initialized variables in Emacs + become read-only after dumping. */ +reg_syntax_t re_syntax_options; + + +/* Specify the precise syntax of regexps for compilation. This provides + for compatibility for various utilities which historically have + different, incompatible syntaxes. + + The argument SYNTAX is a bit mask comprised of the various bits + defined in regex.h. We return the old syntax. */ + +reg_syntax_t +re_set_syntax (syntax) + reg_syntax_t syntax; +{ + reg_syntax_t ret = re_syntax_options; + + re_syntax_options = syntax; +#ifdef DEBUG + if (syntax & RE_DEBUG) + debug = 1; + else if (debug) /* was on but now is not */ + debug = 0; +#endif /* DEBUG */ + return ret; +} +#ifdef _LIBC +weak_alias (__re_set_syntax, re_set_syntax) +#endif + +/* This table gives an error message for each of the error codes listed + in regex.h. Obviously the order here has to be same as there. + POSIX doesn't require that we do anything for REG_NOERROR, + but why not be nice? */ + +#if 0 + /* This section is for xgettext; it sees the strings wrapped inside + N_() and marks them as needing translation. They should match + the strings in re_error_msgid. We can't use the usual string + concatenation trick to initialize re_error_msgid, since other GNU + distributions use this file with traditional C, and traditional C + lacks string concatenation. */ + N_("Success") /* REG_NOERROR */ + N_("No match") /* REG_NOMATCH */ + N_("Invalid regular expression") /* REG_BADPAT */ + N_("Invalid collation character") /* REG_ECOLLATE */ + N_("Invalid character class name") /* REG_ECTYPE */ + N_("Trailing backslash") /* REG_EESCAPE */ + N_("Invalid back reference") /* REG_ESUBREG */ + N_("Unmatched [ or [^") /* REG_EBRACK */ + N_("Unmatched ( or \\(") /* REG_EPAREN */ + N_("Unmatched \\{") /* REG_EBRACE */ + N_("Invalid content of \\{\\}") /* REG_BADBR */ + N_("Invalid range end") /* REG_ERANGE */ + N_("Memory exhausted") /* REG_ESPACE */ + N_("Invalid preceding regular expression") /* REG_BADRPT */ + N_("Premature end of regular expression") /* REG_EEND */ + N_("Regular expression too big") /* REG_ESIZE */ + N_("Unmatched ) or \\)") /* REG_ERPAREN */ +#endif + +static const char re_error_msgid[] = "\ +Success\0\ +No match\0\ +Invalid regular expression\0\ +Invalid collation character\0\ +Invalid character class name\0\ +Trailing backslash\0\ +Invalid back reference\0\ +Unmatched [ or [^\0\ +Unmatched ( or \\(\0\ +Unmatched \\{\0\ +Invalid content of \\{\\}\0\ +Invalid range end\0\ +Memory exhausted\0\ +Invalid preceding regular expression\0\ +Premature end of regular expression\0\ +Regular expression too big\0\ +Unmatched ) or \\)"; + +#define REG_NOERROR_IDX 0 +#define REG_NOMATCH_IDX (REG_NOERROR_IDX + sizeof "Success") +#define REG_BADPAT_IDX (REG_NOMATCH_IDX + sizeof "No match") +#define REG_ECOLLATE_IDX (REG_BADPAT_IDX + sizeof "Invalid regular expression") +#define REG_ECTYPE_IDX (REG_ECOLLATE_IDX + sizeof "Invalid collation character") +#define REG_EESCAPE_IDX (REG_ECTYPE_IDX + sizeof "Invalid character class name") +#define REG_ESUBREG_IDX (REG_EESCAPE_IDX + sizeof "Trailing backslash") +#define REG_EBRACK_IDX (REG_ESUBREG_IDX + sizeof "Invalid back reference") +#define REG_EPAREN_IDX (REG_EBRACK_IDX + sizeof "Unmatched [ or [^") +#define REG_EBRACE_IDX (REG_EPAREN_IDX + sizeof "Unmatched ( or \\(") +#define REG_BADBR_IDX (REG_EBRACE_IDX + sizeof "Unmatched \\{") +#define REG_ERANGE_IDX (REG_BADBR_IDX + sizeof "Invalid content of \\{\\}") +#define REG_ESPACE_IDX (REG_ERANGE_IDX + sizeof "Invalid range end") +#define REG_BADRPT_IDX (REG_ESPACE_IDX + sizeof "Memory exhausted") +#define REG_EEND_IDX (REG_BADRPT_IDX + sizeof "Invalid preceding regular expression") +#define REG_ESIZE_IDX (REG_EEND_IDX + sizeof "Premature end of regular expression") +#define REG_ERPAREN_IDX (REG_ESIZE_IDX + sizeof "Regular expression too big") + +static const size_t re_error_msgid_idx[] = + { + REG_NOERROR_IDX, + REG_NOMATCH_IDX, + REG_BADPAT_IDX, + REG_ECOLLATE_IDX, + REG_ECTYPE_IDX, + REG_EESCAPE_IDX, + REG_ESUBREG_IDX, + REG_EBRACK_IDX, + REG_EPAREN_IDX, + REG_EBRACE_IDX, + REG_BADBR_IDX, + REG_ERANGE_IDX, + REG_ESPACE_IDX, + REG_BADRPT_IDX, + REG_EEND_IDX, + REG_ESIZE_IDX, + REG_ERPAREN_IDX + }; + +/* 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 + +/* When using GNU C, we are not REALLY using the C alloca, no matter + what config.h may say. So don't take precautions for it. */ +#ifdef __GNUC__ +# undef C_ALLOCA +#endif + +/* 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. + Note that if REL_ALLOC is defined, matching would not use malloc for the + failure stack, but we would still use it for the register vectors; + so REL_ALLOC should not affect this. */ +#if (defined C_ALLOCA || defined REGEX_MALLOC) && defined emacs +# 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_STACK. */ + + +/* 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_ITEMS items each time we failed. + This is a variable only so users of regex can assign to it; we never + change it ourselves. */ + +#ifdef INT_IS_16BIT + +# if defined MATCH_MAY_ALLOCATE +/* 4400 was enough to cause a crash on Alpha OSF/1, + whose default stack limit is 2mb. */ +long int re_max_failures = 4000; +# else +long int re_max_failures = 2000; +# endif + +union fail_stack_elt +{ + unsigned char *pointer; + long int integer; +}; + +typedef union fail_stack_elt fail_stack_elt_t; + +typedef struct +{ + fail_stack_elt_t *stack; + unsigned long int size; + unsigned long int avail; /* Offset of next open position. */ +} fail_stack_type; + +#else /* not INT_IS_16BIT */ + +# if defined MATCH_MAY_ALLOCATE +/* 4400 was enough to cause a crash on Alpha OSF/1, + whose default stack limit is 2mb. */ +int re_max_failures = 20000; +# else +int re_max_failures = 2000; +# endif + +union fail_stack_elt +{ + unsigned char *pointer; + int integer; +}; + +typedef union fail_stack_elt fail_stack_elt_t; + +typedef struct +{ + fail_stack_elt_t *stack; + unsigned size; + unsigned avail; /* Offset of next open position. */ +} fail_stack_type; + +#endif /* INT_IS_16BIT */ + +#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 macros to initialize and free the failure 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_STACK (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) + +# define RESET_FAIL_STACK() REGEX_FREE_STACK (fail_stack.stack) +#else +# define INIT_FAIL_STACK() \ + do { \ + fail_stack.avail = 0; \ + } while (0) + +# define RESET_FAIL_STACK() +#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_STACK requires `destination' be declared. */ + +#define DOUBLE_FAIL_STACK(fail_stack) \ + ((fail_stack).size > (unsigned) (re_max_failures * MAX_FAILURE_ITEMS) \ + ? 0 \ + : ((fail_stack).stack = (fail_stack_elt_t *) \ + REGEX_REALLOCATE_STACK ((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 pointer POINTER 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(POINTER, FAIL_STACK) \ + ((FAIL_STACK_FULL () \ + && !DOUBLE_FAIL_STACK (FAIL_STACK)) \ + ? 0 \ + : ((FAIL_STACK).stack[(FAIL_STACK).avail++].pointer = POINTER, \ + 1)) + +/* Push a pointer value onto the failure stack. + Assumes the variable `fail_stack'. Probably should only + be called from within `PUSH_FAILURE_POINT'. */ +#define PUSH_FAILURE_POINTER(item) \ + fail_stack.stack[fail_stack.avail++].pointer = (unsigned char *) (item) + +/* This pushes an integer-valued item onto the failure stack. + Assumes the variable `fail_stack'. Probably should only + be called from within `PUSH_FAILURE_POINT'. */ +#define PUSH_FAILURE_INT(item) \ + fail_stack.stack[fail_stack.avail++].integer = (item) + +/* Push a fail_stack_elt_t value onto the failure stack. + Assumes the variable `fail_stack'. Probably should only + be called from within `PUSH_FAILURE_POINT'. */ +#define PUSH_FAILURE_ELT(item) \ + fail_stack.stack[fail_stack.avail++] = (item) + +/* These three POP... operations complement the three PUSH... operations. + All assume that `fail_stack' is nonempty. */ +#define POP_FAILURE_POINTER() fail_stack.stack[--fail_stack.avail].pointer +#define POP_FAILURE_INT() fail_stack.stack[--fail_stack.avail].integer +#define POP_FAILURE_ELT() 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_INT +# define DEBUG_POP(item_addr) *(item_addr) = POP_FAILURE_INT () +#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_pushed 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. */ \ + /* Can't be int, since there is not a shred of a guarantee that int \ + is wide enough to hold a value of something to which pointer can \ + be assigned */ \ + active_reg_t 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: %ld\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"); \ + \ + if (1) \ + for (this_reg = lowest_active_reg; this_reg <= highest_active_reg; \ + this_reg++) \ + { \ + DEBUG_PRINT2 (" Pushing reg: %lu\n", this_reg); \ + DEBUG_STATEMENT (num_regs_pushed++); \ + \ + DEBUG_PRINT2 (" start: %p\n", regstart[this_reg]); \ + PUSH_FAILURE_POINTER (regstart[this_reg]); \ + \ + DEBUG_PRINT2 (" end: %p\n", regend[this_reg]); \ + PUSH_FAILURE_POINTER (regend[this_reg]); \ + \ + DEBUG_PRINT2 (" info: %p\n ", \ + reg_info[this_reg].word.pointer); \ + 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_ELT (reg_info[this_reg].word); \ + } \ + \ + DEBUG_PRINT2 (" Pushing low active reg: %ld\n", lowest_active_reg);\ + PUSH_FAILURE_INT (lowest_active_reg); \ + \ + DEBUG_PRINT2 (" Pushing high active reg: %ld\n", highest_active_reg);\ + PUSH_FAILURE_INT (highest_active_reg); \ + \ + DEBUG_PRINT2 (" Pushing pattern %p:\n", pattern_place); \ + DEBUG_PRINT_COMPILED_PATTERN (bufp, pattern_place, pend); \ + PUSH_FAILURE_POINTER (pattern_place); \ + \ + DEBUG_PRINT2 (" Pushing string %p: `", string_place); \ + DEBUG_PRINT_DOUBLE_STRING (string_place, string1, size1, string2, \ + size2); \ + DEBUG_PRINT1 ("'\n"); \ + PUSH_FAILURE_POINTER (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. */ +/* We used to use (num_regs - 1), which is the number of registers + this regexp will save; but that was changed to 5 + to avoid stack overflow for a regexp with lots of parens. */ +#define MAX_FAILURE_ITEMS (5 * NUM_REG_ITEMS + NUM_NONREG_ITEMS) + +/* We actually push this many items. */ +#define NUM_FAILURE_ITEMS \ + (((0 \ + ? 0 : 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 (unsigned failure_id;) \ + active_reg_t 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_POINTER (); \ + if (string_temp != NULL) \ + str = (const char *) string_temp; \ + \ + DEBUG_PRINT2 (" Popping string %p: `", str); \ + DEBUG_PRINT_DOUBLE_STRING (str, string1, size1, string2, size2); \ + DEBUG_PRINT1 ("'\n"); \ + \ + pat = (unsigned char *) POP_FAILURE_POINTER (); \ + DEBUG_PRINT2 (" Popping pattern %p:\n", pat); \ + DEBUG_PRINT_COMPILED_PATTERN (bufp, pat, pend); \ + \ + /* Restore register info. */ \ + high_reg = (active_reg_t) POP_FAILURE_INT (); \ + DEBUG_PRINT2 (" Popping high active reg: %ld\n", high_reg); \ + \ + low_reg = (active_reg_t) POP_FAILURE_INT (); \ + DEBUG_PRINT2 (" Popping low active reg: %ld\n", low_reg); \ + \ + if (1) \ + for (this_reg = high_reg; this_reg >= low_reg; this_reg--) \ + { \ + DEBUG_PRINT2 (" Popping reg: %ld\n", this_reg); \ + \ + reg_info[this_reg].word = POP_FAILURE_ELT (); \ + DEBUG_PRINT2 (" info: %p\n", \ + reg_info[this_reg].word.pointer); \ + \ + regend[this_reg] = (const char *) POP_FAILURE_POINTER (); \ + DEBUG_PRINT2 (" end: %p\n", regend[this_reg]); \ + \ + regstart[this_reg] = (const char *) POP_FAILURE_POINTER (); \ + DEBUG_PRINT2 (" start: %p\n", regstart[this_reg]); \ + } \ + else \ + { \ + for (this_reg = highest_active_reg; this_reg > high_reg; this_reg--) \ + { \ + reg_info[this_reg].word.integer = 0; \ + regend[this_reg] = 0; \ + regstart[this_reg] = 0; \ + } \ + highest_active_reg = high_reg; \ + } \ + \ + set_regs_matched_done = 0; \ + DEBUG_STATEMENT (nfailure_points_popped++); \ +} /* POP_FAILURE_POINT */ + + + +/* Structure for per-register (a.k.a. per-group) information. + 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. */ + + +/* Declarations and macros for re_match_2. */ + +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 \ + { \ + if (!set_regs_matched_done) \ + { \ + active_reg_t r; \ + set_regs_matched_done = 1; \ + 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. */ +static char reg_unset_dummy; +#define REG_UNSET_VALUE (®_unset_dummy) +#define REG_UNSET(e) ((e) == REG_UNSET_VALUE) + +/* Subroutine declarations and macros for regex_compile. */ + +static reg_errcode_t regex_compile _RE_ARGS ((const char *pattern, size_t size, + reg_syntax_t syntax, + struct re_pattern_buffer *bufp)); +static void store_op1 _RE_ARGS ((re_opcode_t op, unsigned char *loc, int arg)); +static void store_op2 _RE_ARGS ((re_opcode_t op, unsigned char *loc, + int arg1, int arg2)); +static void insert_op1 _RE_ARGS ((re_opcode_t op, unsigned char *loc, + int arg, unsigned char *end)); +static void insert_op2 _RE_ARGS ((re_opcode_t op, unsigned char *loc, + int arg1, int arg2, unsigned char *end)); +static boolean at_begline_loc_p _RE_ARGS ((const char *pattern, const char *p, + reg_syntax_t syntax)); +static boolean at_endline_loc_p _RE_ARGS ((const char *p, const char *pend, + reg_syntax_t syntax)); +static reg_errcode_t compile_range _RE_ARGS ((const char **p_ptr, + const char *pend, + char *translate, + reg_syntax_t syntax, + unsigned char *b)); + +/* 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'). */ +#ifndef PATFETCH +# define PATFETCH(c) \ + do {if (p == pend) return REG_EEND; \ + c = (unsigned char) *p++; \ + if (translate) c = (unsigned char) translate[c]; \ + } while (0) +#endif + +/* 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. */ +#ifndef TRANSLATE +# define TRANSLATE(d) \ + (translate ? (char) translate[(unsigned char) (d)] : (d)) +#endif + + +/* 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 ((unsigned long) (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, (int) ((to) - (loc) - 3)) + +/* Likewise, for a two-argument jump. */ +#define STORE_JUMP2(op, loc, to, arg) \ + store_op2 (op, loc, (int) ((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, (int) ((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, (int) ((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. */ +/* Any other compiler which, like MSC, has allocation limit below 2^16 + bytes will have to use approach similar to what was done below for + MSC and drop MAX_BUF_SIZE a bit. Otherwise you may end up + reallocating to 0 bytes. Such thing is not going to work too well. + You have been warned!! */ +#if defined _MSC_VER && !defined WIN32 +/* Microsoft C 16-bit versions limit malloc to approx 65512 bytes. + The REALLOC define eliminates a flurry of conversion warnings, + but is not required. */ +# define MAX_BUF_SIZE 65500L +# define REALLOC(p,s) realloc ((p), (size_t) (s)) +#else +# define MAX_BUF_SIZE (1L << 16) +# define REALLOC(p,s) realloc ((p), (s)) +#endif + +/* 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. */ +/* int may be not enough when sizeof(int) == 2. */ +typedef long 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 ('0' <= c && c <= '9') \ + { \ + if (num < 0) \ + num = 0; \ + num = num * 10 + c - '0'; \ + if (p == pend) \ + break; \ + PATFETCH (c); \ + } \ + } \ + } + +#if defined _LIBC || WIDE_CHAR_SUPPORT +/* The GNU C library provides support for user-defined character classes + and the functions from ISO C amendement 1. */ +# ifdef CHARCLASS_NAME_MAX +# define CHAR_CLASS_MAX_LENGTH CHARCLASS_NAME_MAX +# else +/* This shouldn't happen but some implementation might still have this + problem. Use a reasonable default value. */ +# define CHAR_CLASS_MAX_LENGTH 256 +# endif + +# ifdef _LIBC +# define IS_CHAR_CLASS(string) __wctype (string) +# else +# define IS_CHAR_CLASS(string) wctype (string) +# endif +#else +# 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")) +#endif + +#ifndef MATCH_MAY_ALLOCATE + +/* If we cannot allocate large objects within re_match_2_internal, + we make the fail stack and register vectors global. + The fail stack, we grow to the maximum size when a regexp + is compiled. + The register vectors, we adjust in size each time we + compile a regexp, according to the number of registers it needs. */ + +static fail_stack_type fail_stack; + +/* Size with which the following vectors are currently allocated. + That is so we can make them bigger as needed, + but never make them smaller. */ +static int regs_allocated_size; + +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; + +/* Make the register vectors big enough for NUM_REGS registers, + but don't make them smaller. */ + +static +regex_grow_registers (num_regs) + int num_regs; +{ + if (num_regs > regs_allocated_size) + { + 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); + + regs_allocated_size = num_regs; + } +} + +#endif /* not MATCH_MAY_ALLOCATE */ + +static boolean group_in_compile_stack _RE_ARGS ((compile_stack_type + compile_stack, + regnum_t regnum)); + +/* `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; + size_t 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. */ + RE_TRANSLATE_TYPE 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++) + putchar (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 == ':' && *p == ']') || p == pend) + break; + if (c1 < CHAR_CLASS_MAX_LENGTH) + str[c1++] = c; + else + /* This is in any case an invalid class name. */ + str[0] = '\0'; + } + 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 == ']') + { +#if defined _LIBC || WIDE_CHAR_SUPPORT + boolean is_lower = STREQ (str, "lower"); + boolean is_upper = STREQ (str, "upper"); + wctype_t wt; + int ch; + + wt = IS_CHAR_CLASS (str); + if (wt == 0) + 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) + { +# ifdef _LIBC + if (__iswctype (__btowc (ch), wt)) + SET_LIST_BIT (ch); +# else + if (iswctype (btowc (ch), wt)) + SET_LIST_BIT (ch); +# endif + + if (translate && (is_upper || is_lower) + && (ISUPPER (ch) || ISLOWER (ch))) + SET_LIST_BIT (ch); + } + + had_char_class = true; +#else + 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); + if ( translate && (is_upper || is_lower) + && (ISUPPER (ch) || ISLOWER (ch))) + SET_LIST_BIT (ch); + } + had_char_class = true; +#endif /* libc || wctype.h */ + } + 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': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + laststart = b; + BUF_PUSH (wordchar); + break; + + + case 'W': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + laststart = b; + BUF_PUSH (notwordchar); + break; + + + case '<': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (wordbeg); + break; + + case '>': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (wordend); + break; + + case 'b': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (wordbound); + break; + + case 'B': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (notwordbound); + break; + + case '`': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (begbuf); + break; + + case '\'': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + 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, (regnum_t) 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); + + /* If we don't want backtracking, force success + the first time we reach the end of the compiled pattern. */ + if (syntax & RE_NO_POSIX_BACKTRACKING) + BUF_PUSH (succeed); + + 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 */ + } + + regex_grow_registers (num_regs); + } +#endif /* not MATCH_MAY_ALLOCATE */ + + 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; + reg_syntax_t syntax; +{ + const char *next = p; + boolean next_backslash = *next == '\\'; + const char *next_next = p + 1 < pend ? p + 1 : 0; + + 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; + RE_TRANSLATE_TYPE translate; + reg_syntax_t syntax; + unsigned char *b; +{ + unsigned this_char; + + const char *p = *p_ptr; + reg_errcode_t ret; + char range_start[2]; + char range_end[2]; + + if (p == pend) + return REG_ERANGE; + + /* Fetch the endpoints without translating them; the + appropriate translation is done in the bit-setting loop below. */ + range_start[0] = p[-2]; range_start[1] = '\0'; + range_end[0] = p[ 0]; range_end[1] = '\0'; + + /* Have to increment the pointer into the pattern string, so the + caller isn't still at the ending character. */ + (*p_ptr)++; + + /* Report an error if the range is empty and the syntax prohibits this. */ + ret = syntax & RE_NO_EMPTY_RANGES ? REG_ERANGE : REG_NOERROR; + + /* Here we see why `this_char' has to be larger than an `unsigned + char' -- we would otherwise go into an infinite + loop, since all characters <= 0xff. */ + for (this_char = 0; this_char <= (unsigned char) -1; this_char++) + { + char ch[2]; + ch[0] = this_char; ch[1] = '\0'; + if (strcoll (range_start, ch) <= 0 && strcoll (ch, range_end) <= 0) + { + SET_LIST_BIT (TRANSLATE (this_char)); + ret = REG_NOERROR; + } + } + + return ret; +} + +/* 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 + + register char *fastmap = bufp->fastmap; + unsigned char *pattern = bufp->buffer; + unsigned char *p = pattern; + register unsigned char *pend = pattern + bufp->used; + +#ifdef REL_ALLOC + /* This holds the pointer to the failure stack, when + it is allocated relocatably. */ + fail_stack_elt_t *failure_stack_ptr; +#endif + + /* 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 (1) + { + if (p == pend || *p == succeed) + { + /* We have reached the (effective) end of pattern. */ + if (!FAIL_STACK_EMPTY ()) + { + bufp->can_be_null |= path_can_be_null; + + /* Reset for next path. */ + path_can_be_null = true; + + p = fail_stack.stack[--fail_stack.avail].pointer; + + continue; + } + else + break; + } + + /* We should never be about to go beyond the end of the pattern. */ + assert (p < pend); + + switch (SWITCH_ENUM_CAST ((re_opcode_t) *p++)) + { + + /* 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; + goto done; + + + /* 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) + goto done; + + /* 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 /* 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].pointer == 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)) + { + RESET_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; + + done: + RESET_FAIL_STACK (); + return 0; +} /* re_compile_fastmap */ +#ifdef _LIBC +weak_alias (__re_compile_fastmap, re_compile_fastmap) +#endif + +/* 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; + } +} +#ifdef _LIBC +weak_alias (__re_set_registers, re_set_registers) +#endif + +/* 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); +} +#ifdef _LIBC +weak_alias (__re_search, re_search) +#endif + + +/* 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 RE_TRANSLATE_TYPE 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. + Make sure we won't move STARTPOS below 0 or above TOTAL_SIZE. */ + if (endpos < 0) + range = 0 - 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 && range > 0 + && ((re_opcode_t) bufp->buffer[0] == begbuf + /* `begline' is like `begbuf' if it cannot match at newlines. */ + || ((re_opcode_t) bufp->buffer[0] == begline + && !bufp->newline_anchor))) + { + if (startpos > 0) + return -1; + else + range = 1; + } + +#ifdef emacs + /* In a forward search for something that starts with \=. + don't keep searching past point. */ + if (bufp->used > 0 && (re_opcode_t) bufp->buffer[0] == at_dot && range > 0) + { + range = PT - startpos; + if (range <= 0) + return -1; + } +#endif /* emacs */ + + /* 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 */ +#ifdef _LIBC +weak_alias (__re_search_2, re_search_2) +#endif + +/* 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) + +/* Disabled due to a compiler bug -- see comment at case wordbound */ +#if 0 +/* 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)) +#endif + +/* Free everything we malloc. */ +#ifdef MATCH_MAY_ALLOCATE +# define FREE_VAR(var) if (var) REGEX_FREE (var); var = NULL +# define FREE_VARIABLES() \ + do { \ + REGEX_FREE_STACK (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 +# define FREE_VARIABLES() ((void)0) /* Do nothing! But inhibit gcc warning. */ +#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); +# ifndef REGEX_MALLOC +# ifdef C_ALLOCA + alloca (0); +# endif +# endif + return result; +} +# ifdef _LIBC +weak_alias (__re_match, re_match) +# endif +#endif /* not emacs */ + +static boolean group_match_null_string_p _RE_ARGS ((unsigned char **p, + unsigned char *end, + register_info_type *reg_info)); +static boolean alt_match_null_string_p _RE_ARGS ((unsigned char *p, + unsigned char *end, + register_info_type *reg_info)); +static boolean common_op_match_null_string_p _RE_ARGS ((unsigned char **p, + unsigned char *end, + register_info_type *reg_info)); +static int bcmp_translate _RE_ARGS ((const char *s1, const char *s2, + int len, char *translate)); + +/* 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); +#ifndef REGEX_MALLOC +# ifdef C_ALLOCA + alloca (0); +# endif +#endif + return result; +} +#ifdef _LIBC +weak_alias (__re_match_2, re_match_2) +#endif + +/* 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. */ + RE_TRANSLATE_TYPE 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; + unsigned nfailure_points_pushed = 0, nfailure_points_popped = 0; +#endif + +#ifdef REL_ALLOC + /* This holds the pointer to the failure stack, when + it is allocated relocatably. */ + fail_stack_elt_t *failure_stack_ptr; +#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. */ + size_t num_regs = bufp->re_nsub + 1; + + /* The currently active registers. */ + active_reg_t lowest_active_reg = NO_LOWEST_ACTIVE_REG; + active_reg_t 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; + + /* This helps SET_REGS_MATCHED avoid doing redundant work. */ + int set_regs_matched_done = 0; + + /* 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; + } + } + 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 /* 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; (unsigned) 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:\n"); + 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 (;;) + { +#ifdef _LIBC + DEBUG_PRINT2 ("\n%p: ", p); +#else + DEBUG_PRINT2 ("\n0x%x: ", p); +#endif + + 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; (unsigned) 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; (unsigned) mcnt < num_regs; mcnt++) + { + regstart[mcnt] = best_regstart[mcnt]; + regend[mcnt] = best_regend[mcnt]; + } + } + } /* d != end_match_2 */ + + succeed_label: + 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) + { + FREE_VARIABLES (); + 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) + { + FREE_VARIABLES (); + 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; (unsigned) 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; (unsigned) mcnt < regs->num_regs; mcnt++) + regs->start[mcnt] = regs->end[mcnt] = -1; + } /* regs && !bufp->no_sub */ + + 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); + + FREE_VARIABLES (); + return mcnt; + } + + /* Otherwise match next pattern command. */ + switch (SWITCH_ENUM_CAST ((re_opcode_t) *p++)) + { + /* 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; + + case succeed: + DEBUG_PRINT1 ("EXECUTING succeed.\n"); + goto succeed_label; + + /* 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 ((unsigned char) translate[(unsigned char) *d++] + != (unsigned 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; + + /* Clear this whenever we change the register activity status. */ + set_regs_matched_done = 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; + + /* Clear this whenever we change the register activity status. */ + set_regs_matched_done = 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 < (unsigned) *p + (unsigned) *(p + 1); + r++) + { + regstart[r] = old_regstart[r]; + + /* xx why this test? */ + if (old_regend[r] >= 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) + : memcmp (d, d2, mcnt)) + goto fail; + d += mcnt, d2 += mcnt; + + /* Do this because we've match some characters. */ + SET_REGS_MATCHED (); + } + } + 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); +#ifdef _LIBC + DEBUG_PRINT3 (" %d (to %p):\n", mcnt, p + mcnt); +#else + DEBUG_PRINT3 (" %d (to 0x%x):\n", mcnt, p + mcnt); +#endif + + 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); +#ifdef _LIBC + DEBUG_PRINT3 (" %d (to %p)", mcnt, p + mcnt); +#else + DEBUG_PRINT3 (" %d (to 0x%x)", mcnt, p + mcnt); +#endif + + /* 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 \(zz\(a*\)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) + { + /* We win if the first character of the loop is not part + of the charset. */ + if ((re_opcode_t) p1[3] == exactn + && ! ((int) p2[1] * BYTEWIDTH > (int) p1[5] + && (p2[2 + p1[5] / BYTEWIDTH] + & (1 << (p1[5] % BYTEWIDTH))))) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 (" No match => pop_failure_jump.\n"); + } + + 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'. */ + active_reg_t 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. */ + + unconditional_jump: +#ifdef _LIBC + DEBUG_PRINT2 ("\n%p: ", p); +#else + DEBUG_PRINT2 ("\n0x%x: ", p); +#endif + /* Note fall through. */ + + /* Unconditionally jump (without popping any failure points). */ + case jump: + EXTRACT_NUMBER_AND_INCR (mcnt, p); /* Get the amount to jump. */ + DEBUG_PRINT2 ("EXECUTING jump %d ", mcnt); + p += mcnt; /* Do the jump. */ +#ifdef _LIBC + DEBUG_PRINT2 ("(to %p).\n", p); +#else + DEBUG_PRINT2 ("(to 0x%x).\n", p); +#endif + 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 (NULL, NULL, -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 (NULL, NULL, -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); +#ifdef _LIBC + DEBUG_PRINT3 (" Setting %p to %d.\n", p - 2, mcnt); +#else + DEBUG_PRINT3 (" Setting 0x%x to %d.\n", p - 2, mcnt); +#endif + } + else if (mcnt == 0) + { +#ifdef _LIBC + DEBUG_PRINT2 (" Setting two bytes from %p to no_op.\n", p+2); +#else + DEBUG_PRINT2 (" Setting two bytes from 0x%x to no_op.\n", p+2); +#endif + 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); +#ifdef _LIBC + DEBUG_PRINT3 (" Setting %p to %d.\n", p + 2, mcnt); +#else + DEBUG_PRINT3 (" Setting 0x%x to %d.\n", p + 2, mcnt); +#endif + 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); +#ifdef _LIBC + DEBUG_PRINT3 (" Setting %p to %d.\n", p1, mcnt); +#else + DEBUG_PRINT3 (" Setting 0x%x to %d.\n", p1, mcnt); +#endif + STORE_NUMBER (p1, mcnt); + break; + } + +#if 0 + /* The DEC Alpha C compiler 3.x generates incorrect code for the + test WORDCHAR_P (d - 1) != WORDCHAR_P (d) in the expansion of + AT_WORD_BOUNDARY, so this code is disabled. Expanding the + macro and introducing temporary variables works around the bug. */ + + 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; +#else + case wordbound: + { + boolean prevchar, thischar; + + DEBUG_PRINT1 ("EXECUTING wordbound.\n"); + if (AT_STRINGS_BEG (d) || AT_STRINGS_END (d)) + break; + + prevchar = WORDCHAR_P (d - 1); + thischar = WORDCHAR_P (d); + if (prevchar != thischar) + break; + goto fail; + } + + case notwordbound: + { + boolean prevchar, thischar; + + DEBUG_PRINT1 ("EXECUTING notwordbound.\n"); + if (AT_STRINGS_BEG (d) || AT_STRINGS_END (d)) + goto fail; + + prevchar = WORDCHAR_P (d - 1); + thischar = WORDCHAR_P (d); + if (prevchar != thischar) + goto fail; + break; + } +#endif + + 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; + + 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) + const char *s1, *s2; + register int len; + RE_TRANSLATE_TYPE translate; +{ + register const unsigned char *p1 = (const unsigned char *) s1; + register const unsigned char *p2 = (const unsigned char *) 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; + size_t 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); + + if (!ret) + return NULL; + return gettext (re_error_msgid + re_error_msgid_idx[(int) ret]); +} +#ifdef _LIBC +weak_alias (__re_compile_pattern, re_compile_pattern) +#endif + +/* Entry points compatible with 4.2 BSD regex library. We don't define + them unless specifically requested. */ + +#if defined _REGEX_RE_COMP || defined _LIBC + +/* BSD has one and only one pattern buffer. */ +static struct re_pattern_buffer re_comp_buf; + +char * +#ifdef _LIBC +/* Make these definitions weak in libc, so POSIX programs can redefine + these names if they don't use our functions, and still use + regcomp/regexec below without link errors. */ +weak_function +#endif +re_comp (s) + const char *s; +{ + reg_errcode_t ret; + + if (!s) + { + if (!re_comp_buf.buffer) + return gettext ("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 (char *) gettext (re_error_msgid + + re_error_msgid_idx[(int) REG_ESPACE]); + re_comp_buf.allocated = 200; + + re_comp_buf.fastmap = (char *) malloc (1 << BYTEWIDTH); + if (re_comp_buf.fastmap == NULL) + return (char *) gettext (re_error_msgid + + re_error_msgid_idx[(int) REG_ESPACE]); + } + + /* Since `re_exec' always passes NULL for the `regs' argument, we + don't need to initialize the pattern buffer fields which affect it. */ + + /* Match anchors at newlines. */ + re_comp_buf.newline_anchor = 1; + + ret = regex_compile (s, strlen (s), re_syntax_options, &re_comp_buf); + + if (!ret) + return NULL; + + /* Yes, we're discarding `const' here if !HAVE_LIBINTL. */ + return (char *) gettext (re_error_msgid + re_error_msgid_idx[(int) ret]); +} + + +int +#ifdef _LIBC +weak_function +#endif +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' to an allocated space for the fastmap; + `fastmap_accurate' to zero; + `re_nsub' to the number of subexpressions in PATTERN. + + PATTERN is the address of the pattern string. + + CFLAGS is a series of bits which affect compilation. + + If REG_EXTENDED is set, we use POSIX extended syntax; otherwise, we + use POSIX basic syntax. + + If REG_NEWLINE is set, then . and [^...] don't match newline. + Also, regexec will try a match beginning after every newline. + + If REG_ICASE is set, then we considers upper- and lowercase + versions of letters to be equivalent when matching. + + If REG_NOSUB is set, then when PREG is passed to regexec, that + routine will report only success or failure, and nothing about the + registers. + + It returns 0 if it succeeds, nonzero if it doesn't. (See regex.h for + the return codes and their meanings.) */ + +int +regcomp (preg, pattern, cflags) + regex_t *preg; + const char *pattern; + int cflags; +{ + reg_errcode_t ret; + reg_syntax_t 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; + + /* Try to allocate space for the fastmap. */ + preg->fastmap = (char *) malloc (1 << BYTEWIDTH); + + if (cflags & REG_ICASE) + { + unsigned i; + + preg->translate + = (RE_TRANSLATE_TYPE) malloc (CHAR_SET_SIZE + * sizeof (*(RE_TRANSLATE_TYPE)0)); + 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; + + if (ret == REG_NOERROR && preg->fastmap) + { + /* Compute the fastmap now, since regexec cannot modify the pattern + buffer. */ + if (re_compile_fastmap (preg) == -2) + { + /* Some error occured while computing the fastmap, just forget + about it. */ + free (preg->fastmap); + preg->fastmap = NULL; + } + } + + return (int) ret; +} +#ifdef _LIBC +weak_alias (__regcomp, regcomp) +#endif + + +/* 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 * 2, regoff_t); + if (regs.start == NULL) + return (int) REG_NOMATCH; + regs.end = regs.start + nmatch; + } + + /* 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); + } + + /* We want zero return to mean success, unlike `re_search'. */ + return ret >= 0 ? (int) REG_NOERROR : (int) REG_NOMATCH; +} +#ifdef _LIBC +weak_alias (__regexec, regexec) +#endif + + +/* 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 >= (int) (sizeof (re_error_msgid_idx) + / sizeof (re_error_msgid_idx[0]))) + /* Only error codes returned by the rest of the code should be passed + to this routine. If we are given anything else, or if other regex + code generates an invalid error code, then the program has a bug. + Dump core so we can fix it. */ + abort (); + + msg = gettext (re_error_msgid + re_error_msgid_idx[errcode]); + + msg_size = strlen (msg) + 1; /* Includes the null. */ + + if (errbuf_size != 0) + { + if (msg_size > errbuf_size) + { +#if defined HAVE_MEMPCPY || defined _LIBC + *((char *) __mempcpy (errbuf, msg, errbuf_size - 1)) = '\0'; +#else + memcpy (errbuf, msg, errbuf_size - 1); + errbuf[errbuf_size - 1] = 0; +#endif + } + else + memcpy (errbuf, msg, msg_size); + } + + return msg_size; +} +#ifdef _LIBC +weak_alias (__regerror, regerror) +#endif + + +/* 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; +} +#ifdef _LIBC +weak_alias (__regfree, regfree) +#endif + +#endif /* not emacs */ diff --git a/lib/regex.h b/lib/regex.h new file mode 100644 index 00000000..81789be8 --- /dev/null +++ b/lib/regex.h @@ -0,0 +1,46 @@ +#ifndef _REGEX_H +#include <posix/regex.h> + +/* Document internal interfaces. */ +extern reg_syntax_t __re_set_syntax _RE_ARGS ((reg_syntax_t syntax)); + +extern const char *__re_compile_pattern + _RE_ARGS ((const char *pattern, size_t length, + struct re_pattern_buffer *buffer)); + +extern int __re_compile_fastmap _RE_ARGS ((struct re_pattern_buffer *buffer)); + +extern int __re_search + _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string, + int length, int start, int range, struct re_registers *regs)); + +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)); + +extern int __re_match + _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string, + int length, int start, struct re_registers *regs)); + +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)); + +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)); + +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 diff --git a/lib/savedir.c b/lib/savedir.c new file mode 100644 index 00000000..9357cada --- /dev/null +++ b/lib/savedir.c @@ -0,0 +1,179 @@ +/* savedir.c -- save the list of files in a directory in a string + Copyright (C) 1990, 1997, 1998, 1999, 2000, 2001 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by David MacKenzie <djm@gnu.ai.mit.edu>. */ + +#if HAVE_CONFIG_H +# include <config.h> +#endif + +#include <sys/types.h> + +#if 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 + +#ifndef stpcpy +char *stpcpy (); +#endif + +#include <fnmatch.h> +#include "savedir.h" + +char *path; +size_t pathlen; + +static int +isdir1 (const char *dir, const char *file) +{ + int status; + int slash; + size_t dirlen = strlen (dir); + size_t filelen = strlen (file); + if ((dirlen + filelen + 2) > pathlen) + { + path = calloc (dirlen + 1 + filelen + 1, sizeof (*path)); + pathlen = dirlen + filelen + 2; + } + strcpy (path, dir); + slash = (path[dirlen] != '/'); + path[dirlen] = '/'; + strcpy (path + dirlen + slash , file); + status = isdir (path); + return status; +} + +/* 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 (const char *dir, off_t name_size, struct exclude *included_patterns, + struct exclude *excluded_patterns) +{ + DIR *dirp; + struct dirent *dp; + char *name_space; + char *namep; + + dirp = opendir (dir); + if (dirp == NULL) + return NULL; + + /* Be sure name_size is at least `1' so there's room for + the final NUL byte. */ + if (name_size <= 0) + name_size = 1; + + 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'))) + { + off_t size_needed = (namep - name_space) + NAMLEN (dp) + 2; + + if ((included_patterns || excluded_patterns) + && !isdir1 (dir, dp->d_name)) + { + if (included_patterns + && !excluded_filename (included_patterns, dp->d_name, 0)) + continue; + if (excluded_patterns + && excluded_filename (excluded_patterns, dp->d_name, 0)) + continue; + } + + 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; + } + if (path) + { + free (path); + path = NULL; + pathlen = 0; + } + return name_space; +} diff --git a/lib/savedir.h b/lib/savedir.h new file mode 100644 index 00000000..49c331b5 --- /dev/null +++ b/lib/savedir.h @@ -0,0 +1,18 @@ +#if !defined SAVEDIR_H_ +# define SAVEDIR_H_ + +#include "exclude.h" + +# ifndef PARAMS +# if defined PROTOTYPES || (defined __STDC__ && __STDC__) +# define PARAMS(Args) Args +# else +# define PARAMS(Args) () +# endif +# endif + +extern char * +savedir PARAMS ((const char *dir, off_t name_size, + struct exclude *, struct exclude *)); + +#endif diff --git a/lib/stpcpy.c b/lib/stpcpy.c new file mode 100644 index 00000000..a01636cd --- /dev/null +++ b/lib/stpcpy.c @@ -0,0 +1,50 @@ +/* stpcpy.c -- copy a string and return pointer to end of new string + Copyright (C) 1992, 1995, 1997, 1998 Free Software Foundation, Inc. + + NOTE: The canonical source of this file is maintained with the GNU C Library. + Bugs can be reported to bug-glibc@prep.ai.mit.edu. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, + USA. */ + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +#include <string.h> + +#undef __stpcpy +#undef stpcpy + +#ifndef weak_alias +# define __stpcpy stpcpy +#endif + +/* Copy SRC to DEST, returning the address of the terminating '\0' in DEST. */ +char * +__stpcpy (char *dest, const char *src) +{ + register char *d = dest; + register const char *s = src; + + do + *d++ = *s; + while (*s++ != '\0'); + + return d - 1; +} +#ifdef weak_alias +weak_alias (__stpcpy, stpcpy) +#endif diff --git a/lib/strtol.c b/lib/strtol.c new file mode 100644 index 00000000..0c9c2767 --- /dev/null +++ b/lib/strtol.c @@ -0,0 +1,472 @@ +/* Convert string representation of a number into an integer value. + Copyright (C) 1991, 92, 94, 95, 96, 97, 98, 99 Free Software Foundation, Inc. + NOTE: The canonical source of this file is maintained with the GNU C + Library. Bugs can be reported to bug-glibc@gnu.org. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include <config.h> +#endif + +#ifdef _LIBC +# define USE_NUMBER_GROUPING +# define STDC_HEADERS +# define HAVE_LIMITS_H +#endif + +#include <ctype.h> +#include <errno.h> +#ifndef errno +extern int errno; +#endif +#ifndef __set_errno +# define __set_errno(Val) errno = (Val) +#endif + +#ifdef HAVE_LIMITS_H +# include <limits.h> +#endif + +#ifdef STDC_HEADERS +# include <stddef.h> +# include <stdlib.h> +# include <string.h> +#else +# ifndef NULL +# define NULL 0 +# endif +#endif + +#ifdef USE_NUMBER_GROUPING +# include "../locale/localeinfo.h" +#endif + +/* Nonzero if we are defining `strtoul' or `strtoull', operating on + unsigned integers. */ +#ifndef UNSIGNED +# define UNSIGNED 0 +# define INT LONG int +#else +# define INT unsigned LONG int +#endif + +/* Determine the name. */ +#ifdef USE_IN_EXTENDED_LOCALE_MODEL +# if UNSIGNED +# ifdef USE_WIDE_CHAR +# ifdef QUAD +# define strtol __wcstoull_l +# else +# define strtol __wcstoul_l +# endif +# else +# ifdef QUAD +# define strtol __strtoull_l +# else +# define strtol __strtoul_l +# endif +# endif +# else +# ifdef USE_WIDE_CHAR +# ifdef QUAD +# define strtol __wcstoll_l +# else +# define strtol __wcstol_l +# endif +# else +# ifdef QUAD +# define strtol __strtoll_l +# else +# define strtol __strtol_l +# endif +# endif +# endif +#else +# if UNSIGNED +# ifdef USE_WIDE_CHAR +# ifdef QUAD +# define strtol wcstoull +# else +# define strtol wcstoul +# endif +# else +# ifdef QUAD +# define strtol strtoull +# else +# define strtol strtoul +# endif +# endif +# else +# ifdef USE_WIDE_CHAR +# ifdef QUAD +# define strtol wcstoll +# else +# define strtol wcstol +# endif +# else +# ifdef QUAD +# define strtol strtoll +# endif +# endif +# endif +#endif + +/* If QUAD is defined, we are defining `strtoll' or `strtoull', + operating on `long long int's. */ +#ifdef QUAD +# define LONG long long +# define STRTOL_LONG_MIN LONG_LONG_MIN +# define STRTOL_LONG_MAX LONG_LONG_MAX +# define STRTOL_ULONG_MAX ULONG_LONG_MAX + +/* The extra casts work around common compiler bugs, + e.g. Cray C 5.0.3.0 when t == time_t. */ +# ifndef TYPE_SIGNED +# define TYPE_SIGNED(t) (! ((t) 0 < (t) -1)) +# endif +# ifndef TYPE_MINIMUM +# define TYPE_MINIMUM(t) ((t) (TYPE_SIGNED (t) \ + ? ~ (t) 0 << (sizeof (t) * CHAR_BIT - 1) \ + : (t) 0)) +# endif +# ifndef TYPE_MAXIMUM +# define TYPE_MAXIMUM(t) ((t) (~ (t) 0 - TYPE_MINIMUM (t))) +# endif + +# ifndef ULONG_LONG_MAX +# define ULONG_LONG_MAX TYPE_MAXIMUM (unsigned long long) +# endif +# ifndef LONG_LONG_MAX +# define LONG_LONG_MAX TYPE_MAXIMUM (long long int) +# endif +# ifndef LONG_LONG_MIN +# define LONG_LONG_MIN TYPE_MINIMUM (long long int) +# endif + +# if __GNUC__ == 2 && __GNUC_MINOR__ < 7 + /* Work around gcc bug with using this constant. */ + static const unsigned long long int maxquad = ULONG_LONG_MAX; +# undef STRTOL_ULONG_MAX +# define STRTOL_ULONG_MAX maxquad +# endif +#else +# define LONG long + +# ifndef ULONG_MAX +# define ULONG_MAX ((unsigned long) ~(unsigned long) 0) +# endif +# ifndef LONG_MAX +# define LONG_MAX ((long int) (ULONG_MAX >> 1)) +# endif +# define STRTOL_LONG_MIN LONG_MIN +# define STRTOL_LONG_MAX LONG_MAX +# define STRTOL_ULONG_MAX ULONG_MAX +#endif + + +/* We use this code also for the extended locale handling where the + function gets as an additional argument the locale which has to be + used. To access the values we have to redefine the _NL_CURRENT + macro. */ +#ifdef USE_IN_EXTENDED_LOCALE_MODEL +# undef _NL_CURRENT +# define _NL_CURRENT(category, item) \ + (current->values[_NL_ITEM_INDEX (item)].string) +# define LOCALE_PARAM , loc +# define LOCALE_PARAM_DECL __locale_t loc; +#else +# define LOCALE_PARAM +# define LOCALE_PARAM_DECL +#endif + +#if defined _LIBC || defined HAVE_WCHAR_H +# include <wchar.h> +#endif + +#ifdef USE_WIDE_CHAR +# include <wctype.h> +# define L_(Ch) L##Ch +# define UCHAR_TYPE wint_t +# define STRING_TYPE wchar_t +# ifdef USE_IN_EXTENDED_LOCALE_MODEL +# define ISSPACE(Ch) __iswspace_l ((Ch), loc) +# define ISALPHA(Ch) __iswalpha_l ((Ch), loc) +# define TOUPPER(Ch) __towupper_l ((Ch), loc) +# else +# define ISSPACE(Ch) iswspace (Ch) +# define ISALPHA(Ch) iswalpha (Ch) +# define TOUPPER(Ch) towupper (Ch) +# endif +#else +# if defined STDC_HEADERS || (!defined isascii && !defined HAVE_ISASCII) +# define IN_CTYPE_DOMAIN(c) 1 +# else +# define IN_CTYPE_DOMAIN(c) isascii(c) +# endif +# define L_(Ch) Ch +# define UCHAR_TYPE unsigned char +# define STRING_TYPE char +# ifdef USE_IN_EXTENDED_LOCALE_MODEL +# define ISSPACE(Ch) __isspace_l ((Ch), loc) +# define ISALPHA(Ch) __isalpha_l ((Ch), loc) +# define TOUPPER(Ch) __toupper_l ((Ch), loc) +# else +# define ISSPACE(Ch) (IN_CTYPE_DOMAIN (Ch) && isspace (Ch)) +# define ISALPHA(Ch) (IN_CTYPE_DOMAIN (Ch) && isalpha (Ch)) +# define TOUPPER(Ch) (IN_CTYPE_DOMAIN (Ch) ? toupper (Ch) : (Ch)) +# endif +#endif + +/* For compilers which are ansi but don't define __STDC__, like SGI + Irix-4.0.5 cc, also check whether PROTOTYPES is defined. */ +#if defined (__STDC__) || defined (PROTOTYPES) +# define INTERNAL(X) INTERNAL1(X) +# define INTERNAL1(X) __##X##_internal +# define WEAKNAME(X) WEAKNAME1(X) +#else +# define INTERNAL(X) __/**/X/**/_internal +#endif + +#ifdef USE_NUMBER_GROUPING +/* This file defines a function to check for correct grouping. */ +# include "grouping.h" +#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. */ + +INT +INTERNAL (strtol) (nptr, endptr, base, group LOCALE_PARAM) + const STRING_TYPE *nptr; + STRING_TYPE **endptr; + int base; + int group; + LOCALE_PARAM_DECL +{ + int negative; + register unsigned LONG int cutoff; + register unsigned int cutlim; + register unsigned LONG int i; + register const STRING_TYPE *s; + register UCHAR_TYPE c; + const STRING_TYPE *save, *end; + int overflow; + +#ifdef USE_NUMBER_GROUPING +# ifdef USE_IN_EXTENDED_LOCALE_MODEL + struct locale_data *current = loc->__locales[LC_NUMERIC]; +# endif + /* The thousands character of the current locale. */ + wchar_t thousands = L'\0'; + /* The numeric grouping specification of the current locale, + in the format described in <locale.h>. */ + const char *grouping; + + if (group) + { + grouping = _NL_CURRENT (LC_NUMERIC, GROUPING); + if (*grouping <= 0 || *grouping == CHAR_MAX) + grouping = NULL; + else + { + /* Figure out the thousands separator character. */ +# if defined _LIBC || defined _HAVE_BTOWC + thousands = __btowc (*_NL_CURRENT (LC_NUMERIC, THOUSANDS_SEP)); + if (thousands == WEOF) + thousands = L'\0'; +# endif + if (thousands == L'\0') + grouping = NULL; + } + } + else + grouping = NULL; +#endif + + if (base < 0 || base == 1 || base > 36) + { + __set_errno (EINVAL); + return 0; + } + + save = s = nptr; + + /* Skip white space. */ + while (ISSPACE (*s)) + ++s; + if (*s == L_('\0')) + goto noconv; + + /* Check for a sign. */ + if (*s == L_('-')) + { + negative = 1; + ++s; + } + else if (*s == L_('+')) + { + negative = 0; + ++s; + } + else + negative = 0; + + /* Recognize number prefix and if BASE is zero, figure it out ourselves. */ + if (*s == L_('0')) + { + if ((base == 0 || base == 16) && TOUPPER (s[1]) == L_('X')) + { + s += 2; + base = 16; + } + else if (base == 0) + base = 8; + } + else if (base == 0) + base = 10; + + /* Save the pointer so we can check later if anything happened. */ + save = s; + +#ifdef USE_NUMBER_GROUPING + if (group) + { + /* Find the end of the digit string and check its grouping. */ + end = s; + for (c = *end; c != L_('\0'); c = *++end) + if ((wchar_t) c != thousands + && ((wchar_t) c < L_('0') || (wchar_t) c > L_('9')) + && (!ISALPHA (c) || (int) (TOUPPER (c) - L_('A') + 10) >= base)) + break; + if (*s == thousands) + end = s; + else + end = correctly_grouped_prefix (s, end, thousands, grouping); + } + else +#endif + end = NULL; + + cutoff = STRTOL_ULONG_MAX / (unsigned LONG int) base; + cutlim = STRTOL_ULONG_MAX % (unsigned LONG int) base; + + overflow = 0; + i = 0; + for (c = *s; c != L_('\0'); c = *++s) + { + if (s == end) + break; + if (c >= L_('0') && c <= L_('9')) + c -= L_('0'); + else if (ISALPHA (c)) + c = TOUPPER (c) - L_('A') + 10; + else + break; + if ((int) 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 = (STRING_TYPE *) s; + +#if !UNSIGNED + /* Check for a value that is within the range of + `unsigned LONG int', but outside the range of `LONG int'. */ + if (overflow == 0 + && i > (negative + ? -((unsigned LONG int) (STRTOL_LONG_MIN + 1)) + 1 + : (unsigned LONG int) STRTOL_LONG_MAX)) + overflow = 1; +#endif + + if (overflow) + { + __set_errno (ERANGE); +#if UNSIGNED + return STRTOL_ULONG_MAX; +#else + return negative ? STRTOL_LONG_MIN : STRTOL_LONG_MAX; +#endif + } + + /* Return the result of the appropriate sign. */ + return negative ? -i : i; + +noconv: + /* We must handle a special case here: the base is 0 or 16 and the + first two characters are '0' and 'x', but the rest are no + hexadecimal digits. This is no error case. We return 0 and + ENDPTR points to the `x`. */ + if (endptr != NULL) + { + if (save - nptr >= 2 && TOUPPER (save[-1]) == L_('X') + && save[-2] == L_('0')) + *endptr = (STRING_TYPE *) &save[-1]; + else + /* There was no number to convert. */ + *endptr = (STRING_TYPE *) nptr; + } + + return 0L; +} + +/* External user entry point. */ + +#if _LIBC - 0 == 0 +# undef PARAMS +# if defined (__STDC__) && __STDC__ +# define PARAMS(Args) Args +# else +# define PARAMS(Args) () +# endif + +/* Prototype. */ +INT strtol PARAMS ((const STRING_TYPE *nptr, STRING_TYPE **endptr, int base)); +#endif + + +INT +#ifdef weak_function +weak_function +#endif +strtol (nptr, endptr, base LOCALE_PARAM) + const STRING_TYPE *nptr; + STRING_TYPE **endptr; + int base; + LOCALE_PARAM_DECL +{ + return INTERNAL (strtol) (nptr, endptr, base, 0 LOCALE_PARAM); +} diff --git a/lib/strtoul.c b/lib/strtoul.c new file mode 100644 index 00000000..298ae9ea --- /dev/null +++ b/lib/strtoul.c @@ -0,0 +1,22 @@ +/* Copyright (C) 1991, 1999 Free Software Foundation, Inc. + +NOTE: The canonical source of this file is maintained with the GNU C Library. +Bugs can be reported to bug-glibc@prep.ai.mit.edu. + +This program is free software; you can redistribute it and/or modify it +under the terms of the GNU General Public License as published by the +Free Software Foundation; either version 2, or (at your option) any +later version. + +This program is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +GNU General Public License for more details. + +You should have received a copy of the GNU General Public License +along with this program; if not, write to the Free Software Foundation, +Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#define UNSIGNED 1 + +#include <strtol.c> diff --git a/lib/strtoull.c b/lib/strtoull.c new file mode 100644 index 00000000..d6aa1f80 --- /dev/null +++ b/lib/strtoull.c @@ -0,0 +1,27 @@ +/* Function to parse an `unsigned long long int' from text. + Copyright (C) 1995, 1996, 1997, 1999 Free Software Foundation, Inc. + NOTE: The canonical source of this file is maintained with the GNU C + Library. Bugs can be reported to bug-glibc@gnu.org. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#define QUAD 1 + +#include "strtoul.c" + +#ifdef _LIBC +strong_alias (__strtoull_internal, __strtouq_internal) +weak_alias (strtoull, strtouq) +#endif diff --git a/lib/strtoumax.c b/lib/strtoumax.c new file mode 100644 index 00000000..a147604a --- /dev/null +++ b/lib/strtoumax.c @@ -0,0 +1,81 @@ +/* Convert string representation of a number into an uintmax_t value. + Copyright 1999 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert. */ + +#if HAVE_CONFIG_H +# include <config.h> +#endif + +#if HAVE_INTTYPES_H +# include <inttypes.h> +#endif + +#if HAVE_STDLIB_H +# include <stdlib.h> +#endif + +#ifndef PARAMS +# if defined PROTOTYPES || defined __STDC__ +# define PARAMS(Args) Args +# else +# define PARAMS(Args) () +# endif +#endif + +#ifndef HAVE_DECL_STRTOUL +"this configure-time declaration test was not run" +#endif +#if !HAVE_DECL_STRTOUL +unsigned long strtoul PARAMS ((char const *, char **, int)); +#endif + +#ifndef HAVE_DECL_STRTOULL +"this configure-time declaration test was not run" +#endif +#if !HAVE_DECL_STRTOULL && HAVE_UNSIGNED_LONG_LONG +unsigned long long strtoull PARAMS ((char const *, char **, int)); +#endif + +uintmax_t +strtoumax (char const *ptr, char **endptr, int base) +{ +#define USE_IF_EQUIVALENT(function) \ + if (sizeof (uintmax_t) == sizeof function (ptr, endptr, base)) \ + return function (ptr, endptr, base); + +#if HAVE_UNSIGNED_LONG_LONG + USE_IF_EQUIVALENT (strtoull) +#endif + + USE_IF_EQUIVALENT (strtoul) + + abort (); +} + +#ifdef TESTING +# include <stdio.h> +int +main () +{ + char *p, *endptr; + printf ("sizeof uintmax_t: %d\n", sizeof (uintmax_t)); + printf ("sizeof strtoull(): %d\n", sizeof strtoull(p, &endptr, 10)); + printf ("sizeof strtoul(): %d\n", sizeof strtoul(p, &endptr, 10)); + exit (0); +} +#endif diff --git a/lib/xalloc.h b/lib/xalloc.h new file mode 100644 index 00000000..098a6c2e --- /dev/null +++ b/lib/xalloc.h @@ -0,0 +1,87 @@ +/* xalloc.h -- malloc with out-of-memory checking + Copyright (C) 1990-1998, 1999, 2000 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef XALLOC_H_ +# define XALLOC_H_ + +# ifndef PARAMS +# if defined PROTOTYPES || (defined __STDC__ && __STDC__) +# define PARAMS(Args) Args +# else +# define PARAMS(Args) () +# endif +# endif + +# ifndef __attribute__ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 8) || __STRICT_ANSI__ +# define __attribute__(x) +# endif +# endif + +# ifndef ATTRIBUTE_NORETURN +# define ATTRIBUTE_NORETURN __attribute__ ((__noreturn__)) +# endif + +/* Exit value when the requested amount of memory is not available. + It is initialized to EXIT_FAILURE, but the caller may set it to + some other value. */ +extern int xalloc_exit_failure; + +/* If this pointer is non-zero, run the specified function upon each + allocation failure. It is initialized to zero. */ +extern void (*xalloc_fail_func) PARAMS ((void)); + +/* If XALLOC_FAIL_FUNC is undefined or a function that returns, this + message is output. It is translated via gettext. + Its value is "memory exhausted". */ +extern char const xalloc_msg_memory_exhausted[]; + +/* This function is always triggered when memory is exhausted. It is + in charge of honoring the three previous items. This is the + function to call when one wants the program to die because of a + memory allocation failure. */ +extern void xalloc_die PARAMS ((void)) ATTRIBUTE_NORETURN; + +void *xmalloc PARAMS ((size_t n)); +void *xcalloc PARAMS ((size_t n, size_t s)); +void *xrealloc PARAMS ((void *p, size_t n)); +char *xstrdup PARAMS ((const char *str)); + +# define XMALLOC(Type, N_items) ((Type *) xmalloc (sizeof (Type) * (N_items))) +# define XCALLOC(Type, N_items) ((Type *) xcalloc (sizeof (Type), (N_items))) +# define XREALLOC(Ptr, Type, N_items) \ + ((Type *) xrealloc ((void *) (Ptr), sizeof (Type) * (N_items))) + +/* Declare and alloc memory for VAR of type TYPE. */ +# define NEW(Type, Var) Type *(Var) = XMALLOC (Type, 1) + +/* Free VAR only if non NULL. */ +# define XFREE(Var) \ + do { \ + if (Var) \ + free (Var); \ + } while (0) + +/* Return a pointer to a malloc'ed copy of the array SRC of NUM elements. */ +# define CCLONE(Src, Num) \ + (memcpy (xmalloc (sizeof (*Src) * (Num)), (Src), sizeof (*Src) * (Num))) + +/* Return a malloc'ed copy of SRC. */ +# define CLONE(Src) CCLONE (Src, 1) + + +#endif /* !XALLOC_H_ */ diff --git a/lib/xmalloc.c b/lib/xmalloc.c new file mode 100644 index 00000000..2f103d60 --- /dev/null +++ b/lib/xmalloc.c @@ -0,0 +1,116 @@ +/* xmalloc.c -- malloc with out of memory checking + Copyright (C) 1990-1999, 2000 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include <config.h> +#endif + +#include <sys/types.h> + +#if STDC_HEADERS +# include <stdlib.h> +#else +void *calloc (); +void *malloc (); +void *realloc (); +void free (); +#endif + +#if ENABLE_NLS +# include <libintl.h> +# define _(Text) gettext (Text) +#else +# define textdomain(Domain) +# define _(Text) Text +#endif +#define N_(Text) Text + +#include "error.h" +#include "xalloc.h" + +#ifndef EXIT_FAILURE +# define EXIT_FAILURE 1 +#endif + +#ifndef HAVE_DONE_WORKING_MALLOC_CHECK +"you must run the autoconf test for a properly working malloc -- see malloc.m4" +#endif + +#ifndef HAVE_DONE_WORKING_REALLOC_CHECK +"you must run the autoconf test for a properly working realloc --see realloc.m4" +#endif + +/* Exit value when the requested amount of memory is not available. + The caller may set it to some other value. */ +int xalloc_exit_failure = EXIT_FAILURE; + +/* If non NULL, call this function when memory is exhausted. */ +void (*xalloc_fail_func) PARAMS ((void)) = 0; + +/* If XALLOC_FAIL_FUNC is NULL, or does return, display this message + before exiting when memory is exhausted. Goes through gettext. */ +char const xalloc_msg_memory_exhausted[] = N_("memory exhausted"); + +void +xalloc_die (void) +{ + if (xalloc_fail_func) + (*xalloc_fail_func) (); + error (xalloc_exit_failure, 0, "%s", _(xalloc_msg_memory_exhausted)); + /* The `noreturn' cannot be given to error, since it may return if + its first argument is 0. To help compilers understand the + xalloc_die does terminate, call exit. */ + exit (EXIT_FAILURE); +} + +/* Allocate N bytes of memory dynamically, with error checking. */ + +void * +xmalloc (size_t n) +{ + void *p; + + p = malloc (n); + if (p == 0) + xalloc_die (); + return p; +} + +/* Change the size of an allocated block of memory P to N bytes, + with error checking. */ + +void * +xrealloc (void *p, size_t n) +{ + p = realloc (p, n); + if (p == 0) + xalloc_die (); + return p; +} + +/* Allocate memory for N elements of S bytes, with error checking. */ + +void * +xcalloc (size_t n, size_t s) +{ + void *p; + + p = calloc (n, s); + if (p == 0) + xalloc_die (); + return p; +} diff --git a/lib/xstrtol.c b/lib/xstrtol.c new file mode 100644 index 00000000..07023d9f --- /dev/null +++ b/lib/xstrtol.c @@ -0,0 +1,282 @@ +/* A more useful interface to strtol. + Copyright (C) 1995, 1996, 1998-2000 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Jim Meyering. */ + +#if HAVE_CONFIG_H +# include <config.h> +#endif + +#ifndef __strtol +# define __strtol strtol +# define __strtol_t long int +# define __xstrtol xstrtol +#endif + +/* Some pre-ANSI implementations (e.g. SunOS 4) + need stderr defined if assertion checking is enabled. */ +#include <stdio.h> + +#if STDC_HEADERS +# include <stdlib.h> +#endif + +#if HAVE_STRING_H +# include <string.h> +#else +# include <strings.h> +# ifndef strchr +# define strchr index +# endif +#endif + +#include <assert.h> +#include <ctype.h> + +#include <errno.h> +#ifndef errno +extern int errno; +#endif + +#if HAVE_LIMITS_H +# include <limits.h> +#endif + +#ifndef CHAR_BIT +# define CHAR_BIT 8 +#endif + +/* The extra casts work around common compiler bugs. */ +#define TYPE_SIGNED(t) (! ((t) 0 < (t) -1)) +/* The outer cast is needed to work around a bug in Cray C 5.0.3.0. + It is necessary at least when t == time_t. */ +#define TYPE_MINIMUM(t) ((t) (TYPE_SIGNED (t) \ + ? ~ (t) 0 << (sizeof (t) * CHAR_BIT - 1) : (t) 0)) +#define TYPE_MAXIMUM(t) (~ (t) 0 - TYPE_MINIMUM (t)) + +#if defined (STDC_HEADERS) || (!defined (isascii) && !defined (HAVE_ISASCII)) +# define IN_CTYPE_DOMAIN(c) 1 +#else +# define IN_CTYPE_DOMAIN(c) isascii(c) +#endif + +#define ISSPACE(c) (IN_CTYPE_DOMAIN (c) && isspace (c)) + +#include "xstrtol.h" + +#ifndef strtol +long int strtol (); +#endif + +#ifndef strtoul +unsigned long int strtoul (); +#endif + +#ifndef strtoumax +uintmax_t strtoumax (); +#endif + +static int +bkm_scale (__strtol_t *x, int scale_factor) +{ + __strtol_t product = *x * scale_factor; + if (*x != product / scale_factor) + return 1; + *x = product; + return 0; +} + +static int +bkm_scale_by_power (__strtol_t *x, int base, int power) +{ + while (power--) + if (bkm_scale (x, base)) + return 1; + + return 0; +} + +/* FIXME: comment. */ + +strtol_error +__xstrtol (const char *s, char **ptr, int strtol_base, + __strtol_t *val, const char *valid_suffixes) +{ + char *t_ptr; + char **p; + __strtol_t tmp; + + assert (0 <= strtol_base && strtol_base <= 36); + + p = (ptr ? ptr : &t_ptr); + + if (! TYPE_SIGNED (__strtol_t)) + { + const char *q = s; + while (ISSPACE ((unsigned char) *q)) + ++q; + if (*q == '-') + return LONGINT_INVALID; + } + + errno = 0; + tmp = __strtol (s, p, strtol_base); + if (errno != 0) + return LONGINT_OVERFLOW; + if (*p == s) + return LONGINT_INVALID; + + /* Let valid_suffixes == NULL mean `allow any suffix'. */ + /* FIXME: update all callers except the ones that allow suffixes + after the number, changing last parameter NULL to `""'. */ + if (!valid_suffixes) + { + *val = tmp; + return LONGINT_OK; + } + + if (**p != '\0') + { + int base = 1024; + int suffixes = 1; + int overflow; + + if (!strchr (valid_suffixes, **p)) + { + *val = tmp; + return LONGINT_INVALID_SUFFIX_CHAR; + } + + if (strchr (valid_suffixes, '0')) + { + /* The ``valid suffix'' '0' is a special flag meaning that + an optional second suffix is allowed, which can change + the base, e.g. "100MD" for 100 megabytes decimal. */ + + switch (p[0][1]) + { + case 'B': + suffixes++; + break; + + case 'D': + base = 1000; + suffixes++; + break; + } + } + + switch (**p) + { + case 'b': + overflow = bkm_scale (&tmp, 512); + break; + + case 'B': + overflow = bkm_scale (&tmp, 1024); + break; + + case 'c': + overflow = 0; + break; + + case 'E': /* Exa */ + overflow = bkm_scale_by_power (&tmp, base, 6); + break; + + case 'G': /* Giga */ + overflow = bkm_scale_by_power (&tmp, base, 3); + break; + + case 'k': /* kilo */ + overflow = bkm_scale_by_power (&tmp, base, 1); + break; + + case 'M': /* Mega */ + case 'm': /* 'm' is undocumented; for backward compatibility only */ + overflow = bkm_scale_by_power (&tmp, base, 2); + break; + + case 'P': /* Peta */ + overflow = bkm_scale_by_power (&tmp, base, 5); + break; + + case 'T': /* Tera */ + overflow = bkm_scale_by_power (&tmp, base, 4); + break; + + case 'w': + overflow = bkm_scale (&tmp, 2); + break; + + case 'Y': /* Yotta */ + overflow = bkm_scale_by_power (&tmp, base, 8); + break; + + case 'Z': /* Zetta */ + overflow = bkm_scale_by_power (&tmp, base, 7); + break; + + default: + *val = tmp; + return LONGINT_INVALID_SUFFIX_CHAR; + break; + } + + if (overflow) + return LONGINT_OVERFLOW; + + (*p) += suffixes; + } + + *val = tmp; + return LONGINT_OK; +} + +#ifdef TESTING_XSTRTO + +# include <stdio.h> +# include "error.h" + +char *program_name; + +int +main (int argc, char** argv) +{ + strtol_error s_err; + int i; + + program_name = argv[0]; + for (i=1; i<argc; i++) + { + char *p; + __strtol_t val; + + s_err = __xstrtol (argv[i], &p, 0, &val, "bckmw"); + if (s_err == LONGINT_OK) + { + printf ("%s->%lu (%s)\n", argv[i], val, p); + } + else + { + STRTOL_FATAL_ERROR (argv[i], "arg", s_err); + } + } + exit (0); +} + +#endif /* TESTING_XSTRTO */ diff --git a/lib/xstrtol.h b/lib/xstrtol.h new file mode 100644 index 00000000..7a9a0244 --- /dev/null +++ b/lib/xstrtol.h @@ -0,0 +1,64 @@ +#ifndef XSTRTOL_H_ +# define XSTRTOL_H_ 1 + +# if HAVE_INTTYPES_H +# include <inttypes.h> /* for uintmax_t */ +# endif + +# ifndef PARAMS +# if defined PROTOTYPES || (defined __STDC__ && __STDC__) +# define PARAMS(Args) Args +# else +# define PARAMS(Args) () +# endif +# endif + +# ifndef _STRTOL_ERROR +enum strtol_error + { + LONGINT_OK, LONGINT_INVALID, LONGINT_INVALID_SUFFIX_CHAR, LONGINT_OVERFLOW + }; +typedef enum strtol_error strtol_error; +# endif + +# define _DECLARE_XSTRTOL(name, type) \ + strtol_error \ + name PARAMS ((const char *s, char **ptr, int base, \ + type *val, const char *valid_suffixes)); +_DECLARE_XSTRTOL (xstrtol, long int) +_DECLARE_XSTRTOL (xstrtoul, unsigned long int) +_DECLARE_XSTRTOL (xstrtoumax, uintmax_t) + +# define _STRTOL_ERROR(Exit_code, Str, Argument_type_string, Err) \ + do \ + { \ + switch ((Err)) \ + { \ + case LONGINT_OK: \ + abort (); \ + \ + case LONGINT_INVALID: \ + error ((Exit_code), 0, "invalid %s `%s'", \ + (Argument_type_string), (Str)); \ + break; \ + \ + case LONGINT_INVALID_SUFFIX_CHAR: \ + error ((Exit_code), 0, "invalid character following %s `%s'", \ + (Argument_type_string), (Str)); \ + break; \ + \ + case LONGINT_OVERFLOW: \ + error ((Exit_code), 0, "%s `%s' too large", \ + (Argument_type_string), (Str)); \ + break; \ + } \ + } \ + while (0) + +# define STRTOL_FATAL_ERROR(Str, Argument_type_string, Err) \ + _STRTOL_ERROR (2, Str, Argument_type_string, Err) + +# define STRTOL_FAIL_WARN(Str, Argument_type_string, Err) \ + _STRTOL_ERROR (0, Str, Argument_type_string, Err) + +#endif /* not XSTRTOL_H_ */ diff --git a/lib/xstrtoumax.c b/lib/xstrtoumax.c new file mode 100644 index 00000000..04d7cf98 --- /dev/null +++ b/lib/xstrtoumax.c @@ -0,0 +1,31 @@ +/* xstrtoumax.c -- A more useful interface to strtoumax. + Copyright 1999 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert. */ + +#if HAVE_CONFIG_H +# include <config.h> +#endif + +#if HAVE_INTTYPES_H +# include <inttypes.h> +#endif + +#define __strtol strtoumax +#define __strtol_t uintmax_t +#define __xstrtol xstrtoumax +#include "xstrtol.c" |