posix/bug-regex33.c

   1 /* Test re_search with multi-byte characters in EUC-JP.
   2    Copyright (C) 2012-2019 Free Software Foundation, Inc.
   3    This file is part of the GNU C Library.
   4    Contributed by Stanislav Brabec <sbrabec@suse.cz>, 2012.
   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 #define _GNU_SOURCE 1
  21 #include <locale.h>
  22 #include <regex.h>
  23 #include <stdio.h>
  24 #include <stdlib.h>
  25 #include <string.h>
  26
  27 static int
  28 do_test (void)
  29 {
  30   struct re_pattern_buffer r;
  31   struct re_registers s;
  32   int e, rc = 0;
  33   if (setlocale (LC_CTYPE, "ja_JP.EUC-JP") == NULL)
  34     {
  35       puts ("setlocale failed");
  36       return 1;
  37     }
  38   memset (&r, 0, sizeof (r));
  39   memset (&s, 0, sizeof (s));
  40
  41   /* The bug cannot be reproduced without initialized fastmap (it is SBC_MAX
  42      value from regex_internal.h).  */
  43   r.fastmap = malloc (UCHAR_MAX + 1);
  44
  45                      /* 圭 */
  46   re_compile_pattern ("\xb7\xbd", 2, &r);
  47
  48                 /* aaaaa件a新処, \xb7\xbd constitutes a false match */
  49   e = re_search (&r, "\x61\x61\x61\x61\x61\xb7\xef\x61\xbf\xb7\xbd\xe8",
  50                  12, 0, 12, &s);
  51   if (e != -1)
  52     {
  53       printf ("bug-regex33.1: false match or error: re_search() returned %d, should return -1\n", e);
  54       rc = 1;
  55     }
  56
  57                 /* aaaa件a新処, \xb7\xbd constitutes a false match,
  58                  * this is a reproducer of BZ #13637 */
  59   e = re_search (&r, "\x61\x61\x61\x61\xb7\xef\x61\xbf\xb7\xbd\xe8",
  60                  11, 0, 11, &s);
  61   if (e != -1)
  62     {
  63       printf ("bug-regex33.2: false match or error: re_search() returned %d, should return -1\n", e);
  64       rc = 1;
  65     }
  66
  67                 /* aaa件a新処, \xb7\xbd constitutes a false match,
  68                  * this is a reproducer of BZ #13637 */
  69   e = re_search (&r, "\x61\x61\x61\xb7\xef\x61\xbf\xb7\xbd\xe8",
  70                  10, 0, 10, &s);
  71   if (e != -1)
  72     {
  73       printf ("bug-regex33.3: false match or error: re_search() returned %d, should return -1\n", e);
  74       rc = 1;
  75     }
  76
  77                 /* aa件a新処, \xb7\xbd constitutes a false match */
  78   e = re_search (&r, "\x61\x61\xb7\xef\x61\xbf\xb7\xbd\xe8",
  79                  9, 0, 9, &s);
  80   if (e != -1)
  81     {
  82       printf ("bug-regex33.4: false match or error: re_search() returned %d, should return -1\n", e);
  83       rc = 1;
  84     }
  85
  86                 /* a件a新処, \xb7\xbd constitutes a false match */
  87   e = re_search (&r, "\x61\xb7\xef\x61\xbf\xb7\xbd\xe8",
  88                  8, 0, 8, &s);
  89   if (e != -1)
  90     {
  91       printf ("bug-regex33.5: false match or error: re_search() returned %d, should return -1\n", e);
  92       rc = 1;
  93     }
  94
  95                 /* 新処圭新処, \xb7\xbd here really matches 圭, but second occurrence is a false match,
  96                  * this is a reproducer of bug-regex25 and BZ #13637 */
  97   e = re_search (&r, "\xbf\xb7\xbd\xe8\xb7\xbd\xbf\xb7\xbd\xe8",
  98                  10, 0, 10, &s);
  99   if (e != 4)
 100     {
 101       printf ("bug-regex33.6: no match or false match: re_search() returned %d, should return 4\n", e);
 102       rc = 1;
 103     }
 104
 105                 /* 新処圭新, \xb7\xbd here really matches 圭,
 106                  * this is a reproducer of bug-regex25 */
 107   e = re_search (&r, "\xbf\xb7\xbd\xe8\xb7\xbd\xbf\xb7",
 108                  10, 0, 10, &s);
 109   if (e != 4)
 110     {
 111       printf ("bug-regex33.7: no match or false match: re_search() returned %d, should return 4\n", e);
 112       rc = 1;
 113     }
 114
 115   return rc;
 116 }
 117
 118 #define TEST_FUNCTION do_test ()
 119 #include "../test-skeleton.c"