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 /* Software engineering face-lift by Greg J. Badros, 11-Dec-1999,
  44    gjb@cs.washington.edu, http://www.cs.washington.edu/homes/gjb */
  45
  46 \f
  47
  48 /* regex-posix.c -- POSIX regular expression support.
  49
  50    This code was written against Henry Spencer's famous regex package.
  51    The principal reference for POSIX behavior was the man page for this
  52    library, not the 1003.2 document itself.  Ergo, other `POSIX'
  53    libraries which do not agree with the Spencer implementation may
  54    produce varying behavior.  Sigh. */
  55
  56 #include <sys/types.h>
  57
  58 #include "libguile/_scm.h"
  59
  60 /* Supposedly, this file is never compiled unless we know we have
  61    POSIX regular expressions.  But we still put this in an #ifdef so
  62    the file is CPP'able (for dependency scanning) even on systems that
  63    don't have a <regex.h> header.  */
  64 #ifdef HAVE_REGCOMP
  65 #ifdef HAVE_REGEX_H
  66 #include <regex.h>
  67 #else
  68 #ifdef HAVE_RXPOSIX_H
  69 #include <rxposix.h>            /* GNU Rx library */
  70 #else
  71 #ifdef HAVE_RX_RXPOSIX_H
  72 #include <rx/rxposix.h>         /* GNU Rx library on Linux */
  73 #endif
  74 #endif
  75 #endif
  76 #endif
  77
  78 #include "libguile/smob.h"
  79 #include "libguile/symbols.h"
  80 #include "libguile/vectors.h"
  81 #include "libguile/strports.h"
  82 #include "libguile/ports.h"
  83 #include "libguile/feature.h"
  84 #include "libguile/strings.h"
  85
  86 #include "libguile/validate.h"
  87 #include "libguile/regex-posix.h"
  88
  89 /* This is defined by some regex libraries and omitted by others. */
  90 #ifndef REG_BASIC
  91 #define REG_BASIC 0
  92 #endif
  93
  94 scm_bits_t scm_tc16_regex;
  95
  96 static size_t
  97 regex_free (SCM obj)
  98 {
  99   regfree (SCM_RGX (obj));
 100   free (SCM_RGX (obj));
 101   return sizeof(regex_t);
 102 }
 103
 104 \f
 105
 106 SCM_SYMBOL (scm_regexp_error_key, "regular-expression-syntax");
 107
 108 static char *
 109 scm_regexp_error_msg (int regerrno, regex_t *rx)
 110 {
 111   SCM errmsg;
 112   int l;
 113
 114   /* FIXME: must we wrap any external calls in SCM_DEFER_INTS...SCM_ALLOW_INTS?
 115      Or are these only necessary when a SCM object may be left in an
 116      undetermined state (half-formed)?  If the latter then I believe we
 117      may do without the critical section code. -twp */
 118
 119   /* We could simply make errmsg a char pointer, and allocate space with
 120      malloc.  But since we are about to pass the pointer to scm_error, which
 121      never returns, we would never have the opportunity to free it.  Creating
 122      it as a SCM object means that the system will GC it at some point. */
 123
 124   errmsg = scm_make_string (SCM_MAKINUM (80), SCM_UNDEFINED);
 125   SCM_DEFER_INTS;
 126   l = regerror (regerrno, rx, SCM_STRING_CHARS (errmsg), 80);
 127   if (l > 80)
 128     {
 129       errmsg = scm_make_string (SCM_MAKINUM (l), SCM_UNDEFINED);
 130       regerror (regerrno, rx, SCM_STRING_CHARS (errmsg), l);
 131     }
 132   SCM_ALLOW_INTS;
 133   return SCM_STRING_CHARS (errmsg);
 134 }
 135
 136 SCM_DEFINE (scm_regexp_p, "regexp?", 1, 0, 0,
 137             (SCM obj),
 138             "Return @code{#t} if @var{obj} is a compiled regular expression,\n"
 139             "or @code{#f} otherwise.")
 140 #define FUNC_NAME s_scm_regexp_p
 141 {
 142   return SCM_BOOL(SCM_RGXP (obj));
 143 }
 144 #undef FUNC_NAME
 145
 146 SCM_DEFINE (scm_make_regexp, "make-regexp", 1, 0, 1,
 147             (SCM pat, SCM flags),
 148             "Compile the regular expression described by @var{pat}, and\n"
 149             "return the compiled regexp structure.  If @var{pat} does not\n"
 150             "describe a legal regular expression, @code{make-regexp} throws\n"
 151             "a @code{regular-expression-syntax} error.\n"
 152             "\n"
 153             "The @var{flags} arguments change the behavior of the compiled\n"
 154             "regular expression.  The following flags may be supplied:\n"
 155             "\n"
 156             "@table @code\n"
 157             "@item regexp/icase\n"
 158             "Consider uppercase and lowercase letters to be the same when\n"
 159             "matching.\n"
 160             "@item regexp/newline\n"
 161             "If a newline appears in the target string, then permit the\n"
 162             "@samp{^} and @samp{$} operators to match immediately after or\n"
 163             "immediately before the newline, respectively.  Also, the\n"
 164             "@samp{.} and @samp{[^...]} operators will never match a newline\n"
 165             "character.  The intent of this flag is to treat the target\n"
 166             "string as a buffer containing many lines of text, and the\n"
 167             "regular expression as a pattern that may match a single one of\n"
 168             "those lines.\n"
 169             "@item regexp/basic\n"
 170             "Compile a basic (``obsolete'') regexp instead of the extended\n"
 171             "(``modern'') regexps that are the default.  Basic regexps do\n"
 172             "not consider @samp{|}, @samp{+} or @samp{?} to be special\n"
 173             "characters, and require the @samp{@{...@}} and @samp{(...)}\n"
 174             "metacharacters to be backslash-escaped (@pxref{Backslash\n"
 175             "Escapes}).  There are several other differences between basic\n"
 176             "and extended regular expressions, but these are the most\n"
 177             "significant.\n"
 178             "@item regexp/extended\n"
 179             "Compile an extended regular expression rather than a basic\n"
 180             "regexp.  This is the default behavior; this flag will not\n"
 181             "usually be needed.  If a call to @code{make-regexp} includes\n"
 182             "both @code{regexp/basic} and @code{regexp/extended} flags, the\n"
 183             "one which comes last will override the earlier one.\n"
 184             "@end table")
 185 #define FUNC_NAME s_scm_make_regexp
 186 {
 187   SCM flag;
 188   regex_t *rx;
 189   int status, cflags;
 190
 191   SCM_VALIDATE_STRING (1, pat);
 192   SCM_VALIDATE_REST_ARGUMENT (flags);
 193   SCM_STRING_COERCE_0TERMINATION_X (pat);
 194
 195   /* Examine list of regexp flags.  If REG_BASIC is supplied, then
 196      turn off REG_EXTENDED flag (on by default). */
 197   cflags = REG_EXTENDED;
 198   flag = flags;
 199   while (!SCM_NULLP (flag))
 200     {
 201       if (SCM_INUM (SCM_CAR (flag)) == REG_BASIC)
 202         cflags &= ~REG_EXTENDED;
 203       else
 204         cflags |= SCM_INUM (SCM_CAR (flag));
 205       flag = SCM_CDR (flag);
 206     }
 207
 208   rx = SCM_MUST_MALLOC_TYPE (regex_t);
 209   status = regcomp (rx, SCM_STRING_CHARS (pat),
 210                     /* Make sure they're not passing REG_NOSUB;
 211                        regexp-exec assumes we're getting match data.  */
 212                     cflags & ~REG_NOSUB);
 213   if (status != 0)
 214     {
 215       scm_error (scm_regexp_error_key,
 216                  FUNC_NAME,
 217                  scm_regexp_error_msg (status, rx),
 218                  SCM_BOOL_F,
 219                  SCM_BOOL_F);
 220       /* never returns */
 221     }
 222   SCM_RETURN_NEWSMOB (scm_tc16_regex, rx);
 223 }
 224 #undef FUNC_NAME
 225
 226 SCM_DEFINE (scm_regexp_exec, "regexp-exec", 2, 2, 0,
 227             (SCM rx, SCM str, SCM start, SCM flags),
 228             "Match the compiled regular expression @var{rx} against\n"
 229             "@code{str}.  If the optional integer @var{start} argument is\n"
 230             "provided, begin matching from that position in the string.\n"
 231             "Return a match structure describing the results of the match,\n"
 232             "or @code{#f} if no match could be found.\n"
 233             "\n"
 234             "The @var{flags} arguments change the matching behavior.\n"
 235             "The following flags may be supplied:\n"
 236             "\n"
 237             "@table @code\n"
 238             "@item regexp/notbol\n"
 239             "Operator @samp{^} always fails (unless @code{regexp/newline}\n"
 240             "is used).  Use this when the beginning of the string should\n"
 241             "not be considered the beginning of a line.\n"
 242             "@item regexp/noteol\n"
 243             "Operator @samp{$} always fails (unless @code{regexp/newline}\n"
 244             "is used).  Use this when the end of the string should not be\n"
 245             "considered the end of a line.\n"
 246             "@end table")
 247 #define FUNC_NAME s_scm_regexp_exec
 248 {
 249   int status, nmatches, offset;
 250   regmatch_t *matches;
 251   SCM mvec = SCM_BOOL_F;
 252
 253   SCM_VALIDATE_RGXP (1,rx);
 254   SCM_VALIDATE_STRING (2, str);
 255   SCM_VALIDATE_INUM_DEF_COPY (3,start,0,offset);
 256   SCM_ASSERT_RANGE (3,start, offset >= 0 && offset <= SCM_STRING_LENGTH (str));
 257   if (SCM_UNBNDP (flags))
 258     flags = SCM_INUM0;
 259   SCM_VALIDATE_INUM (4,flags);
 260   SCM_STRING_COERCE_0TERMINATION_X (str);
 261
 262   /* re_nsub doesn't account for the `subexpression' representing the
 263      whole regexp, so add 1 to nmatches. */
 264
 265   nmatches = SCM_RGX(rx)->re_nsub + 1;
 266   SCM_DEFER_INTS;
 267   matches = SCM_MUST_MALLOC_TYPE_NUM (regmatch_t,nmatches);
 268   status = regexec (SCM_RGX (rx), SCM_STRING_CHARS (str) + offset,
 269                     nmatches, matches,
 270                     SCM_INUM (flags));
 271   if (!status)
 272     {
 273       int i;
 274       /* The match vector must include a cell for the string that was matched,
 275          so add 1. */
 276       mvec = scm_c_make_vector (nmatches + 1, SCM_UNSPECIFIED);
 277       SCM_VELTS(mvec)[0] = str;
 278       for (i = 0; i < nmatches; ++i)
 279         if (matches[i].rm_so == -1)
 280           SCM_VELTS(mvec)[i+1] = scm_cons (SCM_MAKINUM (-1), SCM_MAKINUM (-1));
 281         else
 282           SCM_VELTS(mvec)[i+1]
 283             = scm_cons(SCM_MAKINUM(matches[i].rm_so + offset),
 284                        SCM_MAKINUM(matches[i].rm_eo + offset));
 285     }
 286   scm_must_free ((char *) matches);
 287   SCM_ALLOW_INTS;
 288
 289   if (status != 0 && status != REG_NOMATCH)
 290     scm_error (scm_regexp_error_key,
 291                FUNC_NAME,
 292                scm_regexp_error_msg (status, SCM_RGX (rx)),
 293                SCM_BOOL_F,
 294                SCM_BOOL_F);
 295   return mvec;
 296 }
 297 #undef FUNC_NAME
 298
 299 void
 300 scm_init_regex_posix ()
 301 {
 302   scm_tc16_regex = scm_make_smob_type ("regexp", sizeof (regex_t));
 303   scm_set_smob_free (scm_tc16_regex, regex_free);
 304
 305   /* Compilation flags.  */
 306   scm_c_define ("regexp/basic", scm_long2num (REG_BASIC));
 307   scm_c_define ("regexp/extended", scm_long2num (REG_EXTENDED));
 308   scm_c_define ("regexp/icase", scm_long2num (REG_ICASE));
 309   scm_c_define ("regexp/newline", scm_long2num (REG_NEWLINE));
 310
 311   /* Execution flags.  */
 312   scm_c_define ("regexp/notbol", scm_long2num (REG_NOTBOL));
 313   scm_c_define ("regexp/noteol", scm_long2num (REG_NOTEOL));
 314
 315 #ifndef SCM_MAGIC_SNARFER
 316 #include "libguile/regex-posix.x"
 317 #endif
 318
 319   scm_add_feature ("regex");
 320 }
 321
 322 /*
 323   Local Variables:
 324   c-file-style: "gnu"
 325   End:
 326 */