libguile/regex-posix.c

   1 /*      Copyright (C) 1997, 1998, 1999, 2000, 2001 Free Software Foundation, Inc.
   2  *
   3  * This program is free software; you can redistribute it and/or modify
   4  * it under the terms of the GNU General Public License as published by
   5  * the Free Software Foundation; either version 2, or (at your option)
   6  * any later version.
   7  *
   8  * This program is distributed in the hope that it will be useful,
   9  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  10  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  11  * GNU General Public License for more details.
  12  *
  13  * You should have received a copy of the GNU General Public License
  14  * along with this software; see the file COPYING.  If not, write to
  15  * the Free Software Foundation, Inc., 59 Temple Place, Suite 330,
  16  * Boston, MA 02111-1307 USA
  17  *
  18  * As a special exception, the Free Software Foundation gives permission
  19  * for additional uses of the text contained in its release of GUILE.
  20  *
  21  * The exception is that, if you link the GUILE library with other files
  22  * to produce an executable, this does not by itself cause the
  23  * resulting executable to be covered by the GNU General Public License.
  24  * Your use of that executable is in no way restricted on account of
  25  * linking the GUILE library code into it.
  26  *
  27  * This exception does not however invalidate any other reasons why
  28  * the executable file might be covered by the GNU General Public License.
  29  *
  30  * This exception applies only to the code released by the
  31  * Free Software Foundation under the name GUILE.  If you copy
  32  * code from other Free Software Foundation releases into a copy of
  33  * GUILE, as the General Public License permits, the exception does
  34  * not apply to the code that you add in this way.  To avoid misleading
  35  * anyone as to the status of such modified files, you must delete
  36  * this exception notice from them.
  37  *
  38  * If you write modifications of your own for GUILE, it is your choice
  39  * whether to permit this exception to apply to your modifications.
  40  * If you do not wish that, delete this exception notice.
  41  */
  42
  43
  44 \f
  45
  46 /* regex-posix.c -- POSIX regular expression support.
  47
  48    This code was written against Henry Spencer's famous regex package.
  49    The principal reference for POSIX behavior was the man page for this
  50    library, not the 1003.2 document itself.  Ergo, other `POSIX'
  51    libraries which do not agree with the Spencer implementation may
  52    produce varying behavior.  Sigh. */
  53
  54 #include <sys/types.h>
  55
  56 #include "libguile/_scm.h"
  57
  58 /* Supposedly, this file is never compiled unless we know we have
  59    POSIX regular expressions.  But we still put this in an #ifdef so
  60    the file is CPP'able (for dependency scanning) even on systems that
  61    don't have a <regex.h> header.  */
  62 #ifdef HAVE_REGCOMP
  63 #ifdef HAVE_REGEX_H
  64 #include <regex.h>
  65 #else
  66 #ifdef HAVE_RXPOSIX_H
  67 #include <rxposix.h>            /* GNU Rx library */
  68 #else
  69 #ifdef HAVE_RX_RXPOSIX_H
  70 #include <rx/rxposix.h>         /* GNU Rx library on Linux */
  71 #endif
  72 #endif
  73 #endif
  74 #endif
  75
  76 #include "libguile/smob.h"
  77 #include "libguile/symbols.h"
  78 #include "libguile/vectors.h"
  79 #include "libguile/strports.h"
  80 #include "libguile/ports.h"
  81 #include "libguile/feature.h"
  82 #include "libguile/strings.h"
  83
  84 #include "libguile/validate.h"
  85 #include "libguile/regex-posix.h"
  86
  87 /* This is defined by some regex libraries and omitted by others. */
  88 #ifndef REG_BASIC
  89 #define REG_BASIC 0
  90 #endif
  91
  92 scm_t_bits scm_tc16_regex;
  93
  94 static size_t
  95 regex_free (SCM obj)
  96 {
  97   regfree (SCM_RGX (obj));
  98   free (SCM_RGX (obj));
  99   return sizeof(regex_t);
 100 }
 101
 102 \f
 103
 104 SCM_SYMBOL (scm_regexp_error_key, "regular-expression-syntax");
 105
 106 static char *
 107 scm_regexp_error_msg (int regerrno, regex_t *rx)
 108 {
 109   SCM errmsg;
 110   int l;
 111
 112   /* FIXME: must we wrap any external calls in SCM_DEFER_INTS...SCM_ALLOW_INTS?
 113      Or are these only necessary when a SCM object may be left in an
 114      undetermined state (half-formed)?  If the latter then I believe we
 115      may do without the critical section code. -twp */
 116
 117   /* We could simply make errmsg a char pointer, and allocate space with
 118      malloc.  But since we are about to pass the pointer to scm_error, which
 119      never returns, we would never have the opportunity to free it.  Creating
 120      it as a SCM object means that the system will GC it at some point. */
 121
 122   errmsg = scm_make_string (SCM_MAKINUM (80), SCM_UNDEFINED);
 123   SCM_DEFER_INTS;
 124   l = regerror (regerrno, rx, SCM_STRING_CHARS (errmsg), 80);
 125   if (l > 80)
 126     {
 127       errmsg = scm_make_string (SCM_MAKINUM (l), SCM_UNDEFINED);
 128       regerror (regerrno, rx, SCM_STRING_CHARS (errmsg), l);
 129     }
 130   SCM_ALLOW_INTS;
 131   return SCM_STRING_CHARS (errmsg);
 132 }
 133
 134 SCM_DEFINE (scm_regexp_p, "regexp?", 1, 0, 0,
 135             (SCM obj),
 136             "Return @code{#t} if @var{obj} is a compiled regular expression,\n"
 137             "or @code{#f} otherwise.")
 138 #define FUNC_NAME s_scm_regexp_p
 139 {
 140   return SCM_BOOL(SCM_RGXP (obj));
 141 }
 142 #undef FUNC_NAME
 143
 144 SCM_DEFINE (scm_make_regexp, "make-regexp", 1, 0, 1,
 145             (SCM pat, SCM flags),
 146             "Compile the regular expression described by @var{pat}, and\n"
 147             "return the compiled regexp structure.  If @var{pat} does not\n"
 148             "describe a legal regular expression, @code{make-regexp} throws\n"
 149             "a @code{regular-expression-syntax} error.\n"
 150             "\n"
 151             "The @var{flags} arguments change the behavior of the compiled\n"
 152             "regular expression.  The following flags may be supplied:\n"
 153             "\n"
 154             "@table @code\n"
 155             "@item regexp/icase\n"
 156             "Consider uppercase and lowercase letters to be the same when\n"
 157             "matching.\n"
 158             "@item regexp/newline\n"
 159             "If a newline appears in the target string, then permit the\n"
 160             "@samp{^} and @samp{$} operators to match immediately after or\n"
 161             "immediately before the newline, respectively.  Also, the\n"
 162             "@samp{.} and @samp{[^...]} operators will never match a newline\n"
 163             "character.  The intent of this flag is to treat the target\n"
 164             "string as a buffer containing many lines of text, and the\n"
 165             "regular expression as a pattern that may match a single one of\n"
 166             "those lines.\n"
 167             "@item regexp/basic\n"
 168             "Compile a basic (``obsolete'') regexp instead of the extended\n"
 169             "(``modern'') regexps that are the default.  Basic regexps do\n"
 170             "not consider @samp{|}, @samp{+} or @samp{?} to be special\n"
 171             "characters, and require the @samp{@{...@}} and @samp{(...)}\n"
 172             "metacharacters to be backslash-escaped (@pxref{Backslash\n"
 173             "Escapes}).  There are several other differences between basic\n"
 174             "and extended regular expressions, but these are the most\n"
 175             "significant.\n"
 176             "@item regexp/extended\n"
 177             "Compile an extended regular expression rather than a basic\n"
 178             "regexp.  This is the default behavior; this flag will not\n"
 179             "usually be needed.  If a call to @code{make-regexp} includes\n"
 180             "both @code{regexp/basic} and @code{regexp/extended} flags, the\n"
 181             "one which comes last will override the earlier one.\n"
 182             "@end table")
 183 #define FUNC_NAME s_scm_make_regexp
 184 {
 185   SCM flag;
 186   regex_t *rx;
 187   int status, cflags;
 188
 189   SCM_VALIDATE_STRING (1, pat);
 190   SCM_VALIDATE_REST_ARGUMENT (flags);
 191   SCM_STRING_COERCE_0TERMINATION_X (pat);
 192
 193   /* Examine list of regexp flags.  If REG_BASIC is supplied, then
 194      turn off REG_EXTENDED flag (on by default). */
 195   cflags = REG_EXTENDED;
 196   flag = flags;
 197   while (!SCM_NULLP (flag))
 198     {
 199       if (SCM_INUM (SCM_CAR (flag)) == REG_BASIC)
 200         cflags &= ~REG_EXTENDED;
 201       else
 202         cflags |= SCM_INUM (SCM_CAR (flag));
 203       flag = SCM_CDR (flag);
 204     }
 205
 206   rx = SCM_MUST_MALLOC_TYPE (regex_t);
 207   status = regcomp (rx, SCM_STRING_CHARS (pat),
 208                     /* Make sure they're not passing REG_NOSUB;
 209                        regexp-exec assumes we're getting match data.  */
 210                     cflags & ~REG_NOSUB);
 211   if (status != 0)
 212     {
 213       scm_error (scm_regexp_error_key,
 214                  FUNC_NAME,
 215                  scm_regexp_error_msg (status, rx),
 216                  SCM_BOOL_F,
 217                  SCM_BOOL_F);
 218       /* never returns */
 219     }
 220   SCM_RETURN_NEWSMOB (scm_tc16_regex, rx);
 221 }
 222 #undef FUNC_NAME
 223
 224 SCM_DEFINE (scm_regexp_exec, "regexp-exec", 2, 2, 0,
 225             (SCM rx, SCM str, SCM start, SCM flags),
 226             "Match the compiled regular expression @var{rx} against\n"
 227             "@code{str}.  If the optional integer @var{start} argument is\n"
 228             "provided, begin matching from that position in the string.\n"
 229             "Return a match structure describing the results of the match,\n"
 230             "or @code{#f} if no match could be found.\n"
 231             "\n"
 232             "The @var{flags} arguments change the matching behavior.\n"
 233             "The following flags may be supplied:\n"
 234             "\n"
 235             "@table @code\n"
 236             "@item regexp/notbol\n"
 237             "Operator @samp{^} always fails (unless @code{regexp/newline}\n"
 238             "is used).  Use this when the beginning of the string should\n"
 239             "not be considered the beginning of a line.\n"
 240             "@item regexp/noteol\n"
 241             "Operator @samp{$} always fails (unless @code{regexp/newline}\n"
 242             "is used).  Use this when the end of the string should not be\n"
 243             "considered the end of a line.\n"
 244             "@end table")
 245 #define FUNC_NAME s_scm_regexp_exec
 246 {
 247   int status, nmatches, offset;
 248   regmatch_t *matches;
 249   SCM mvec = SCM_BOOL_F;
 250
 251   SCM_VALIDATE_RGXP (1,rx);
 252   SCM_VALIDATE_STRING (2, str);
 253   SCM_VALIDATE_INUM_DEF_COPY (3,start,0,offset);
 254   SCM_ASSERT_RANGE (3,start, offset >= 0 && offset <= SCM_STRING_LENGTH (str));
 255   if (SCM_UNBNDP (flags))
 256     flags = SCM_INUM0;
 257   SCM_VALIDATE_INUM (4,flags);
 258   SCM_STRING_COERCE_0TERMINATION_X (str);
 259
 260   /* re_nsub doesn't account for the `subexpression' representing the
 261      whole regexp, so add 1 to nmatches. */
 262
 263   nmatches = SCM_RGX(rx)->re_nsub + 1;
 264   SCM_DEFER_INTS;
 265   matches = SCM_MUST_MALLOC_TYPE_NUM (regmatch_t,nmatches);
 266   status = regexec (SCM_RGX (rx), SCM_STRING_CHARS (str) + offset,
 267                     nmatches, matches,
 268                     SCM_INUM (flags));
 269   if (!status)
 270     {
 271       int i;
 272       /* The match vector must include a cell for the string that was matched,
 273          so add 1. */
 274       mvec = scm_c_make_vector (nmatches + 1, SCM_UNSPECIFIED);
 275       SCM_VELTS(mvec)[0] = str;
 276       for (i = 0; i < nmatches; ++i)
 277         if (matches[i].rm_so == -1)
 278           SCM_VELTS(mvec)[i+1] = scm_cons (SCM_MAKINUM (-1), SCM_MAKINUM (-1));
 279         else
 280           SCM_VELTS(mvec)[i+1]
 281             = scm_cons (scm_long2num (matches[i].rm_so + offset),
 282                         scm_long2num (matches[i].rm_eo + offset));
 283     }
 284   scm_must_free ((char *) matches);
 285   SCM_ALLOW_INTS;
 286
 287   if (status != 0 && status != REG_NOMATCH)
 288     scm_error (scm_regexp_error_key,
 289                FUNC_NAME,
 290                scm_regexp_error_msg (status, SCM_RGX (rx)),
 291                SCM_BOOL_F,
 292                SCM_BOOL_F);
 293   return mvec;
 294 }
 295 #undef FUNC_NAME
 296
 297 void
 298 scm_init_regex_posix ()
 299 {
 300   scm_tc16_regex = scm_make_smob_type ("regexp", sizeof (regex_t));
 301   scm_set_smob_free (scm_tc16_regex, regex_free);
 302
 303   /* Compilation flags.  */
 304   scm_c_define ("regexp/basic", scm_long2num (REG_BASIC));
 305   scm_c_define ("regexp/extended", scm_long2num (REG_EXTENDED));
 306   scm_c_define ("regexp/icase", scm_long2num (REG_ICASE));
 307   scm_c_define ("regexp/newline", scm_long2num (REG_NEWLINE));
 308
 309   /* Execution flags.  */
 310   scm_c_define ("regexp/notbol", scm_long2num (REG_NOTBOL));
 311   scm_c_define ("regexp/noteol", scm_long2num (REG_NOTEOL));
 312
 313 #ifndef SCM_MAGIC_SNARFER
 314 #include "libguile/regex-posix.x"
 315 #endif
 316
 317   scm_add_feature ("regex");
 318 }
 319
 320 /*
 321   Local Variables:
 322   c-file-style: "gnu"
 323   End:
 324 */