Add tests for strfrom functions
[glibc.git] / posix / tst-rxspencer.c
blobdba40e1476d1c3b20ff73a19d1b8918ac0c0c053
1 /* Regular expression tests.
2 Copyright (C) 2003-2016 Free Software Foundation, Inc.
3 This file is part of the GNU C Library.
4 Contributed by Jakub Jelinek <jakub@redhat.com>, 2003.
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.
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.
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/>. */
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 #include <locale.h>
27 #include <getopt.h>
29 static void
30 replace_special_chars (char *str)
32 for (; (str = strpbrk (str, "NTSZ")) != NULL; ++str)
33 switch (*str)
35 case 'N': *str = '\n'; break;
36 case 'T': *str = '\t'; break;
37 case 'S': *str = ' '; break;
38 case 'Z': *str = '\0'; break;
42 static void
43 glibc_re_syntax (char *str)
45 char *p, *end = strchr (str, '\0') + 1;
47 /* Replace [[:<:]] with \< and [[:>:]] with \>. */
48 for (p = str; (p = strstr (p, "[[:")) != NULL; )
49 if ((p[3] == '<' || p[3] == '>') && strncmp (p + 4, ":]]", 3) == 0)
51 p[0] = '\\';
52 p[1] = p[3];
53 memmove (p + 2, p + 7, end - p - 7);
54 end -= 5;
55 p += 2;
57 else
58 p += 3;
61 static char *
62 mb_replace (char *dst, const char c)
64 switch (c)
66 /* Replace a with \'a and A with \'A. */
67 case 'a':
68 *dst++ = '\xc3';
69 *dst++ = '\xa1';
70 break;
71 case 'A':
72 *dst++ = '\xc3';
73 *dst++ = '\x81';
74 break;
75 /* Replace b with \v{c} and B with \v{C}. */
76 case 'b':
77 *dst++ = '\xc4';
78 *dst++ = '\x8d';
79 break;
80 case 'B':
81 *dst++ = '\xc4';
82 *dst++ = '\x8c';
83 break;
84 /* Replace c with \v{d} and C with \v{D}. */
85 case 'c':
86 *dst++ = '\xc4';
87 *dst++ = '\x8f';
88 break;
89 case 'C':
90 *dst++ = '\xc4';
91 *dst++ = '\x8e';
92 break;
93 /* Replace d with \'e and D with \'E. */
94 case 'd':
95 *dst++ = '\xc3';
96 *dst++ = '\xa9';
97 break;
98 case 'D':
99 *dst++ = '\xc3';
100 *dst++ = '\x89';
101 break;
103 return dst;
106 static char *
107 mb_frob_string (const char *str, const char *letters)
109 char *ret, *dst;
110 const char *src;
112 if (str == NULL)
113 return NULL;
115 ret = malloc (2 * strlen (str) + 1);
116 if (ret == NULL)
117 return NULL;
119 for (src = str, dst = ret; *src; ++src)
120 if (strchr (letters, *src))
121 dst = mb_replace (dst, *src);
122 else
123 *dst++ = *src;
124 *dst = '\0';
125 return ret;
128 /* Like mb_frob_string, but don't replace anything between
129 [: and :], [. and .] or [= and =] or characters escaped
130 with a backslash. */
132 static char *
133 mb_frob_pattern (const char *str, const char *letters)
135 char *ret, *dst;
136 const char *src;
137 int in_class = 0, escaped = 0;
139 if (str == NULL)
140 return NULL;
142 ret = malloc (2 * strlen (str) + 1);
143 if (ret == NULL)
144 return NULL;
146 for (src = str, dst = ret; *src; ++src)
147 if (*src == '\\')
149 escaped ^= 1;
150 *dst++ = *src;
152 else if (escaped)
154 escaped = 0;
155 *dst++ = *src;
156 continue;
158 else if (!in_class && strchr (letters, *src))
159 dst = mb_replace (dst, *src);
160 else
162 if (!in_class && *src == '[' && strchr (":.=", src[1]))
163 in_class = 1;
164 else if (in_class && *src == ']' && strchr (":.=", src[-1]))
165 in_class = 0;
166 *dst++ = *src;
168 *dst = '\0';
169 return ret;
172 static int
173 check_match (regmatch_t *rm, int idx, const char *string,
174 const char *match, const char *fail)
176 if (match[0] == '-' && match[1] == '\0')
178 if (rm[idx].rm_so == -1 && rm[idx].rm_eo == -1)
179 return 0;
180 printf ("%s rm[%d] unexpectedly matched\n", fail, idx);
181 return 1;
184 if (rm[idx].rm_so == -1 || rm[idx].rm_eo == -1)
186 printf ("%s rm[%d] unexpectedly did not match\n", fail, idx);
187 return 1;
190 if (match[0] == '@')
192 if (rm[idx].rm_so != rm[idx].rm_eo)
194 printf ("%s rm[%d] not empty\n", fail, idx);
195 return 1;
198 if (strncmp (string + rm[idx].rm_so, match + 1, strlen (match + 1) ?: 1))
200 printf ("%s rm[%d] not matching %s\n", fail, idx, match);
201 return 1;
203 return 0;
206 if (rm[idx].rm_eo - rm[idx].rm_so != strlen (match)
207 || strncmp (string + rm[idx].rm_so, match,
208 rm[idx].rm_eo - rm[idx].rm_so))
210 printf ("%s rm[%d] not matching %s\n", fail, idx, match);
211 return 1;
214 return 0;
217 static int
218 test (const char *pattern, int cflags, const char *string, int eflags,
219 char *expect, char *matches, const char *fail)
221 regex_t re;
222 regmatch_t rm[10];
223 int n, ret = 0;
225 n = regcomp (&re, pattern, cflags);
226 if (n != 0)
228 char buf[500];
229 if (eflags == -1)
231 static struct { reg_errcode_t code; const char *name; } codes []
232 #define C(x) { REG_##x, #x }
233 = { C(NOERROR), C(NOMATCH), C(BADPAT), C(ECOLLATE),
234 C(ECTYPE), C(EESCAPE), C(ESUBREG), C(EBRACK),
235 C(EPAREN), C(EBRACE), C(BADBR), C(ERANGE),
236 C(ESPACE), C(BADRPT) };
238 for (int i = 0; i < sizeof (codes) / sizeof (codes[0]); ++i)
239 if (n == codes[i].code)
241 if (strcmp (string, codes[i].name))
243 printf ("%s regcomp returned REG_%s (expected REG_%s)\n",
244 fail, codes[i].name, string);
245 return 1;
247 return 0;
250 printf ("%s regcomp return value REG_%d\n", fail, n);
251 return 1;
254 regerror (n, &re, buf, sizeof (buf));
255 printf ("%s regcomp failed: %s\n", fail, buf);
256 return 1;
259 if (eflags == -1)
261 regfree (&re);
263 /* The test case file assumes something only guaranteed by the
264 rxspencer regex implementation. Namely that for empty
265 expressions regcomp() return REG_EMPTY. This is not the case
266 for us and so we ignore this error. */
267 if (strcmp (string, "EMPTY") == 0)
268 return 0;
270 printf ("%s regcomp unexpectedly succeeded\n", fail);
271 return 1;
274 if (regexec (&re, string, 10, rm, eflags))
276 regfree (&re);
277 if (expect == NULL)
278 return 0;
279 printf ("%s regexec failed\n", fail);
280 return 1;
283 regfree (&re);
285 if (expect == NULL)
287 printf ("%s regexec unexpectedly succeeded\n", fail);
288 return 1;
291 if (cflags & REG_NOSUB)
292 return 0;
294 ret = check_match (rm, 0, string, expect, fail);
295 if (matches == NULL)
296 return ret;
298 for (n = 1; ret == 0 && n < 10; ++n)
300 char *p = NULL;
302 if (matches)
304 p = strchr (matches, ',');
305 if (p != NULL)
306 *p = '\0';
308 ret = check_match (rm, n, string, matches ?: "-", fail);
309 if (p)
311 *p = ',';
312 matches = p + 1;
314 else
315 matches = NULL;
318 return ret;
321 static int
322 mb_test (const char *pattern, int cflags, const char *string, int eflags,
323 char *expect, const char *matches, const char *letters,
324 const char *fail)
326 char *pattern_mb = mb_frob_pattern (pattern, letters);
327 const char *string_mb
328 = eflags == -1 ? string : mb_frob_string (string, letters);
329 char *expect_mb = mb_frob_string (expect, letters);
330 char *matches_mb = mb_frob_string (matches, letters);
331 int ret = 0;
333 if (!pattern_mb || !string_mb
334 || (expect && !expect_mb) || (matches && !matches_mb))
336 printf ("%s %m", fail);
337 ret = 1;
339 else
340 ret = test (pattern_mb, cflags, string_mb, eflags, expect_mb,
341 matches_mb, fail);
343 free (matches_mb);
344 free (expect_mb);
345 if (string_mb != string)
346 free ((char *) string_mb);
347 free (pattern_mb);
348 return ret;
351 static int
352 mb_tests (const char *pattern, int cflags, const char *string, int eflags,
353 char *expect, const char *matches)
355 int ret = 0;
356 int i;
357 char letters[9], fail[20];
359 /* The tests aren't supposed to work with xdigit, since a-dA-D are
360 hex digits while \'a \'A \v{c}\v{C}\v{d}\v{D}\'e \'E are not. */
361 if (strstr (pattern, "[:xdigit:]"))
362 return 0;
364 /* XXX: regex ATM handles only single byte equivalence classes. */
365 if (strstr (pattern, "[[=b=]]"))
366 return 0;
368 for (i = 1; i < 16; ++i)
370 char *p = letters;
371 if (i & 1)
373 if (!strchr (pattern, 'a') && !strchr (string, 'a')
374 && !strchr (pattern, 'A') && !strchr (string, 'A'))
375 continue;
376 *p++ = 'a', *p++ = 'A';
378 if (i & 2)
380 if (!strchr (pattern, 'b') && !strchr (string, 'b')
381 && !strchr (pattern, 'B') && !strchr (string, 'B'))
382 continue;
383 *p++ = 'b', *p++ = 'B';
385 if (i & 4)
387 if (!strchr (pattern, 'c') && !strchr (string, 'c')
388 && !strchr (pattern, 'C') && !strchr (string, 'C'))
389 continue;
390 *p++ = 'c', *p++ = 'C';
392 if (i & 8)
394 if (!strchr (pattern, 'd') && !strchr (string, 'd')
395 && !strchr (pattern, 'D') && !strchr (string, 'D'))
396 continue;
397 *p++ = 'd', *p++ = 'D';
399 *p++ = '\0';
400 sprintf (fail, "UTF-8 %s FAIL", letters);
401 ret |= mb_test (pattern, cflags, string, eflags, expect, matches,
402 letters, fail);
404 return ret;
408 main (int argc, char **argv)
410 int ret = 0;
411 char *line = NULL;
412 size_t line_len = 0;
413 ssize_t len;
414 FILE *f;
415 static int test_utf8 = 0;
416 static const struct option options[] =
418 {"utf8", no_argument, &test_utf8, 1},
419 {NULL, 0, NULL, 0 }
422 mtrace ();
424 while (getopt_long (argc, argv, "", options, NULL) >= 0);
426 if (optind + 1 != argc)
428 fprintf (stderr, "Missing test filename\n");
429 return 1;
432 f = fopen (argv[optind], "r");
433 if (f == NULL)
435 fprintf (stderr, "Couldn't open %s\n", argv[optind]);
436 return 1;
439 while ((len = getline (&line, &line_len, f)) > 0)
441 char *pattern, *flagstr, *string, *expect, *matches, *p;
442 int cflags = REG_EXTENDED, eflags = 0, try_bre_ere = 0;
444 if (line[len - 1] == '\n')
445 line[len - 1] = '\0';
447 /* Skip comments and empty lines. */
448 if (*line == '#' || *line == '\0')
449 continue;
451 puts (line);
452 fflush (stdout);
454 pattern = strtok (line, "\t");
455 if (pattern == NULL)
456 continue;
458 if (strcmp (pattern, "\"\"") == 0)
459 pattern += 2;
461 flagstr = strtok (NULL, "\t");
462 if (flagstr == NULL)
463 continue;
465 string = strtok (NULL, "\t");
466 if (string == NULL)
467 continue;
469 if (strcmp (string, "\"\"") == 0)
470 string += 2;
472 for (p = flagstr; *p; ++p)
473 switch (*p)
475 case '-':
476 break;
477 case 'b':
478 cflags &= ~REG_EXTENDED;
479 break;
480 case '&':
481 try_bre_ere = 1;
482 break;
483 case 'C':
484 eflags = -1;
485 break;
486 case 'i':
487 cflags |= REG_ICASE;
488 break;
489 case 's':
490 cflags |= REG_NOSUB;
491 break;
492 case 'n':
493 cflags |= REG_NEWLINE;
494 break;
495 case '^':
496 eflags |= REG_NOTBOL;
497 break;
498 case '$':
499 eflags |= REG_NOTEOL;
500 break;
501 case 'm':
502 case 'p':
503 case '#':
504 /* Not supported. */
505 flagstr = NULL;
506 break;
509 if (flagstr == NULL)
510 continue;
512 replace_special_chars (pattern);
513 glibc_re_syntax (pattern);
514 if (eflags != -1)
515 replace_special_chars (string);
517 expect = strtok (NULL, "\t");
518 matches = NULL;
519 if (expect != NULL)
521 replace_special_chars (expect);
522 matches = strtok (NULL, "\t");
523 if (matches != NULL)
524 replace_special_chars (matches);
527 if (setlocale (LC_ALL, "C") == NULL)
529 puts ("setlocale C failed");
530 ret = 1;
532 if (test (pattern, cflags, string, eflags, expect, matches, "FAIL")
533 || (try_bre_ere
534 && test (pattern, cflags & ~REG_EXTENDED, string, eflags,
535 expect, matches, "FAIL")))
536 ret = 1;
537 else if (test_utf8)
539 if (setlocale (LC_ALL, "cs_CZ.UTF-8") == NULL)
541 puts ("setlocale cs_CZ.UTF-8 failed");
542 ret = 1;
544 else if (test (pattern, cflags, string, eflags, expect, matches,
545 "UTF-8 FAIL")
546 || (try_bre_ere
547 && test (pattern, cflags & ~REG_EXTENDED, string,
548 eflags, expect, matches, "UTF-8 FAIL")))
549 ret = 1;
550 else if (mb_tests (pattern, cflags, string, eflags, expect, matches)
551 || (try_bre_ere
552 && mb_tests (pattern, cflags & ~REG_EXTENDED, string,
553 eflags, expect, matches)))
554 ret = 1;
558 free (line);
559 fclose (f);
560 return ret;