posix/tst-pcre.c

   1 /* Regular expression tests.
   2    Copyright (C) 2003 Free Software Foundation, Inc.
   3    This file is part of the GNU C Library.
   4    Contributed by Jakub Jelinek <jakub@redhat.com>, 2003.
   5
   6    The GNU C Library is free software; you can redistribute it and/or
   7    modify it under the terms of the GNU Lesser General Public
   8    License as published by the Free Software Foundation; either
   9    version 2.1 of the License, or (at your option) any later version.
  10
  11    The GNU C Library is distributed in the hope that it will be useful,
  12    but WITHOUT ANY WARRANTY; without even the implied warranty of
  13    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14    Lesser General Public License for more details.
  15
  16    You should have received a copy of the GNU Lesser General Public
  17    License along with the GNU C Library; if not, see
  18    <http://www.gnu.org/licenses/>.  */
  19
  20 #include <sys/types.h>
  21 #include <mcheck.h>
  22 #include <regex.h>
  23 #include <stdio.h>
  24 #include <stdlib.h>
  25 #include <string.h>
  26
  27 int
  28 main (int argc, char **argv)
  29 {
  30   int ret = 0;
  31   char *line = NULL;
  32   size_t line_len = 0;
  33   ssize_t len;
  34   FILE *f;
  35   char *pattern = NULL, *string = NULL;
  36   regmatch_t rm[20];
  37   size_t pattern_alloced = 0, string_alloced = 0;
  38   int ignorecase = 0;
  39   int pattern_valid = 0, rm_valid = 0;
  40   size_t linenum;
  41
  42   mtrace ();
  43
  44   if (argc < 2)
  45     {
  46       fprintf (stderr, "Missing test filename\n");
  47       return 1;
  48     }
  49
  50   f = fopen (argv[1], "r");
  51   if (f == NULL)
  52     {
  53       fprintf (stderr, "Couldn't open %s\n", argv[1]);
  54       return 1;
  55     }
  56
  57   if ((len = getline (&line, &line_len, f)) <= 0
  58       || strncmp (line, "# PCRE", 6) != 0)
  59     {
  60       fprintf (stderr, "Not a PCRE test file\n");
  61       fclose (f);
  62       free (line);
  63       return 1;
  64     }
  65
  66   linenum = 1;
  67
  68   while ((len = getline (&line, &line_len, f)) > 0)
  69     {
  70       char *p;
  71       unsigned long num;
  72
  73       ++linenum;
  74
  75       if (line[len - 1] == '\n')
  76         line[--len] = '\0';
  77
  78       if (line[0] == '#')
  79         continue;
  80
  81       if (line[0] == '\0')
  82         {
  83           /* End of test.  */
  84           ignorecase = 0;
  85           pattern_valid = 0;
  86           rm_valid = 0;
  87           continue;
  88         }
  89
  90       if (line[0] == '/')
  91         {
  92           /* Pattern.  */
  93           p = strrchr (line + 1, '/');
  94
  95           pattern_valid = 0;
  96           rm_valid = 0;
  97           if (p == NULL)
  98             {
  99               printf ("%zd: Invalid pattern line: %s\n", linenum, line);
 100               ret = 1;
 101               continue;
 102             }
 103
 104           if (p[1] == 'i' && p[2] == '\0')
 105             ignorecase = 1;
 106           else if (p[1] != '\0')
 107             {
 108               printf ("%zd: Invalid pattern line: %s\n", linenum, line);
 109               ret = 1;
 110               continue;
 111             }
 112
 113           if (pattern_alloced < (size_t) (p - line))
 114             {
 115               pattern = realloc (pattern, p - line);
 116               if (pattern == NULL)
 117                 {
 118                   printf ("%zd: Cannot record pattern: %m\n", linenum);
 119                   ret = 1;
 120                   break;
 121                 }
 122               pattern_alloced = p - line;
 123             }
 124
 125           memcpy (pattern, line + 1, p - line - 1);
 126           pattern[p - line - 1] = '\0';
 127           pattern_valid = 1;
 128           continue;
 129         }
 130
 131       if (strncmp (line, "    ", 4) == 0)
 132         {
 133           regex_t re;
 134           int n;
 135
 136           if (!pattern_valid)
 137             {
 138               printf ("%zd: No previous valid pattern %s\n", linenum, line);
 139               continue;
 140             }
 141
 142           if (string_alloced < (size_t) (len - 3))
 143             {
 144               string = realloc (string, len - 3);
 145               if (string == NULL)
 146                 {
 147                   printf ("%zd: Cannot record search string: %m\n", linenum);
 148                   ret = 1;
 149                   break;
 150                 }
 151               string_alloced = len - 3;
 152             }
 153
 154           memcpy (string, line + 4, len - 3);
 155
 156           n = regcomp (&re, pattern,
 157                        REG_EXTENDED | (ignorecase ? REG_ICASE : 0));
 158           if (n != 0)
 159             {
 160               char buf[500];
 161               regerror (n, &re, buf, sizeof (buf));
 162               printf ("%zd: regcomp failed for %s: %s\n",
 163                       linenum, pattern, buf);
 164               ret = 1;
 165               continue;
 166             }
 167
 168           if (regexec (&re, string, 20, rm, 0))
 169             {
 170               rm[0].rm_so = -1;
 171               rm[0].rm_eo = -1;
 172             }
 173
 174           regfree (&re);
 175           rm_valid = 1;
 176           continue;
 177         }
 178
 179       if (!rm_valid)
 180         {
 181           printf ("%zd: No preceeding pattern or search string\n", linenum);
 182           ret = 1;
 183           continue;
 184         }
 185
 186       if (strcmp (line, "No match") == 0)
 187         {
 188           if (rm[0].rm_so != -1 || rm[0].rm_eo != -1)
 189             {
 190               printf ("%zd: /%s/ on %s unexpectedly matched %d..%d\n",
 191                       linenum, pattern, string, rm[0].rm_so, rm[0].rm_eo);
 192               ret = 1;
 193             }
 194
 195           continue;
 196         }
 197
 198       p = line;
 199       if (*p == ' ')
 200         ++p;
 201
 202       num = strtoul (p, &p, 10);
 203       if (num >= 20 || *p != ':' || p[1] != ' ')
 204         {
 205           printf ("%zd: Invalid line %s\n", linenum, line);
 206           ret = 1;
 207           continue;
 208         }
 209
 210       if (rm[num].rm_so == -1 || rm[num].rm_eo == -1)
 211         {
 212           if (strcmp (p + 2, "<unset>") != 0)
 213             {
 214               printf ("%zd: /%s/ on %s unexpectedly failed to match register %ld %d..%d\n",
 215                       linenum, pattern, string, num,
 216                       rm[num].rm_so, rm[num].rm_eo);
 217               ret = 1;
 218             }
 219           continue;
 220         }
 221
 222       if (rm[num].rm_eo < rm[num].rm_so
 223           || rm[num].rm_eo - rm[num].rm_so != len - (p + 2 - line)
 224           || strncmp (p + 2, string + rm[num].rm_so,
 225                       rm[num].rm_eo - rm[num].rm_so) != 0)
 226         {
 227           printf ("%zd: /%s/ on %s unexpectedly failed to match %s for register %ld %d..%d\n",
 228                   linenum, pattern, string, p + 2, num,
 229                   rm[num].rm_so, rm[num].rm_eo);
 230           ret = 1;
 231           continue;
 232         }
 233     }
 234
 235   free (pattern);
 236   free (string);
 237   free (line);
 238   fclose (f);
 239   return ret;
 240 }