1 /* $OpenBSD: fnmatch.c,v 1.16 2011/12/06 11:47:46 stsp Exp $ */
2
3 /* Copyright (c) 2011, VMware, Inc.
4 * All rights reserved.
5 *
6 * Redistribution and use in source and binary forms, with or without
7 * modification, are permitted provided that the following conditions are met:
8 * * Redistributions of source code must retain the above copyright
9 * notice, this list of conditions and the following disclaimer.
10 * * Redistributions in binary form must reproduce the above copyright
11 * notice, this list of conditions and the following disclaimer in the
12 * documentation and/or other materials provided with the distribution.
13 * * Neither the name of the VMware, Inc. nor the names of its contributors
14 * may be used to endorse or promote products derived from this software
15 * without specific prior written permission.
16 *
17 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
18 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 * ARE DISCLAIMED. IN NO EVENT SHALL VMWARE, INC. OR CONTRIBUTORS BE LIABLE FOR
21 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
22 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
23 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
24 * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
25 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
26 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
27 */
28
29 /*
30 * Copyright (c) 2008 Todd C. Miller <millert@openbsd.org>
31 *
32 * Permission to use, copy, modify, and distribute this software for any
33 * purpose with or without fee is hereby granted, provided that the above
34 * copyright notice and this permission notice appear in all copies.
35 *
36 * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
37 * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
38 * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
39 * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
40 * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
41 * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
42 * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
43 */
44
45 /* Authored by William A. Rowe Jr. <wrowe; apache.org, vmware.com>, April 2011
46 *
47 * Derived from The Open Group Base Specifications Issue 7, IEEE Std 1003.1-2008
48 * as described in;
49 * http://pubs.opengroup.org/onlinepubs/9699919799/functions/fnmatch.html
50 *
51 * Filename pattern matches defined in section 2.13, "Pattern Matching Notation"
52 * from chapter 2. "Shell Command Language"
53 * http://pubs.opengroup.org/onlinepubs/9699919799/utilities/V3_chap02.html#tag_18_13
54 * where; 1. A bracket expression starting with an unquoted <circumflex> '^'
55 * character CONTINUES to specify a non-matching list; 2. an explicit <period> '.'
56 * in a bracket expression matching list, e.g. "[.abc]" does NOT match a leading
57 * <period> in a filename; 3. a <left-square-bracket> '[' which does not introduce
58 * a valid bracket expression is treated as an ordinary character; 4. a differing
59 * number of consecutive slashes within pattern and string will NOT match;
60 * 5. a trailing '\' in FNM_ESCAPE mode is treated as an ordinary '\' character.
61 *
62 * Bracket expansion defined in section 9.3.5, "RE Bracket Expression",
63 * from chapter 9, "Regular Expressions"
64 * http://pubs.opengroup.org/onlinepubs/9699919799/basedefs/V1_chap09.html#tag_09_03_05
65 * with no support for collating symbols, equivalence class expressions or
66 * character class expressions. A partial range expression with a leading
67 * hyphen following a valid range expression will match only the ordinary
68 * <hyphen> and the ending character (e.g. "[a-m-z]" will match characters
69 * 'a' through 'm', a <hyphen> '-', or a 'z').
70 *
71 * Supports BSD extensions FNM_LEADING_DIR to match pattern to the end of one
72 * path segment of string, and FNM_CASEFOLD to ignore alpha case.
73 *
74 * NOTE: Only POSIX/C single byte locales are correctly supported at this time.
75 * Notably, non-POSIX locales with FNM_CASEFOLD produce undefined results,
76 * particularly in ranges of mixed case (e.g. "[A-z]") or spanning alpha and
77 * nonalpha characters within a range.
78 *
79 * XXX comments below indicate porting required for multi-byte character sets
80 * and non-POSIX locale collation orders; requires mbr* APIs to track shift
81 * state of pattern and string (rewinding pattern and string repeatedly).
82 *
83 * Certain parts of the code assume 0x00-0x3F are unique with any MBCS (e.g.
84 * UTF-8, SHIFT-JIS, etc). Any implementation allowing '\' as an alternate
85 * path delimiter must be aware that 0x5C is NOT unique within SHIFT-JIS.
86 */
87
88 #include <fnmatch.h>
89 #include <string.h>
90 #include <ctype.h>
91 #include <limits.h>
92
93 #include "charclass.h"
94
95 #define RANGE_MATCH 1
96 #define RANGE_NOMATCH 0
97 #define RANGE_ERROR (-1)
98
99 static int
classmatch(const char * pattern,char test,int foldcase,const char ** ep)100 classmatch(const char *pattern, char test, int foldcase, const char **ep)
101 {
102 struct cclass *cc;
103 const char *colon;
104 size_t len;
105 int rval = RANGE_NOMATCH;
106 const char * const mismatch = pattern;
107
108 if (*pattern != '[' || pattern[1] != ':') {
109 *ep = mismatch;
110 return(RANGE_ERROR);
111 }
112
113 pattern += 2;
114
115 if ((colon = strchr(pattern, ':')) == NULL || colon[1] != ']') {
116 *ep = mismatch;
117 return(RANGE_ERROR);
118 }
119 *ep = colon + 2;
120 len = (size_t)(colon - pattern);
121
122 if (foldcase && strncmp(pattern, "upper:]", 7) == 0)
123 pattern = "lower:]";
124 for (cc = cclasses; cc->name != NULL; cc++) {
125 if (!strncmp(pattern, cc->name, len) && cc->name[len] == '\0') {
126 if (cc->isctype((unsigned char)test))
127 rval = RANGE_MATCH;
128 break;
129 }
130 }
131 if (cc->name == NULL) {
132 /* invalid character class, treat as normal text */
133 *ep = mismatch;
134 rval = RANGE_ERROR;
135 }
136 return(rval);
137 }
138
139 /* Most MBCS/collation/case issues handled here. Wildcard '*' is not handled.
140 * EOS '\0' and the FNM_PATHNAME '/' delimiters are not advanced over,
141 * however the "\/" sequence is advanced to '/'.
142 *
143 * Both pattern and string are **char to support pointer increment of arbitrary
144 * multibyte characters for the given locale, in a later iteration of this code
145 */
fnmatch_ch(const char ** pattern,const char ** string,int flags)146 static int fnmatch_ch(const char **pattern, const char **string, int flags)
147 {
148 const char * const mismatch = *pattern;
149 const int nocase = !!(flags & FNM_CASEFOLD);
150 const int escape = !(flags & FNM_NOESCAPE);
151 const int slash = !!(flags & FNM_PATHNAME);
152 int result = FNM_NOMATCH;
153 const char *startch;
154 int negate;
155
156 if (**pattern == '[')
157 {
158 ++*pattern;
159
160 /* Handle negation, either leading ! or ^ operators (never both) */
161 negate = ((**pattern == '!') || (**pattern == '^'));
162 if (negate)
163 ++*pattern;
164
165 /* ']' is an ordinary character at the start of the range pattern */
166 if (**pattern == ']')
167 goto leadingclosebrace;
168
169 while (**pattern)
170 {
171 if (**pattern == ']') {
172 ++*pattern;
173 /* XXX: Fix for MBCS character width */
174 ++*string;
175 return (result ^ negate);
176 }
177
178 if (escape && (**pattern == '\\')) {
179 ++*pattern;
180
181 /* Patterns must be terminated with ']', not EOS */
182 if (!**pattern)
183 break;
184 }
185
186 /* Patterns must be terminated with ']' not '/' */
187 if (slash && (**pattern == '/'))
188 break;
189
190 /* Match character classes. */
191 if (classmatch(*pattern, **string, nocase, pattern)
192 == RANGE_MATCH) {
193 result = 0;
194 continue;
195 }
196
197 leadingclosebrace:
198 /* Look at only well-formed range patterns;
199 * "x-]" is not allowed unless escaped ("x-\]")
200 * XXX: Fix for locale/MBCS character width
201 */
202 if (((*pattern)[1] == '-') && ((*pattern)[2] != ']'))
203 {
204 startch = *pattern;
205 *pattern += (escape && ((*pattern)[2] == '\\')) ? 3 : 2;
206
207 /* NOT a properly balanced [expr] pattern, EOS terminated
208 * or ranges containing a slash in FNM_PATHNAME mode pattern
209 * fall out to to the rewind and test '[' literal code path
210 */
211 if (!**pattern || (slash && (**pattern == '/')))
212 break;
213
214 /* XXX: handle locale/MBCS comparison, advance by MBCS char width */
215 if ((**string >= *startch) && (**string <= **pattern))
216 result = 0;
217 else if (nocase && (isupper(**string) || isupper(*startch)
218 || isupper(**pattern))
219 && (tolower(**string) >= tolower(*startch))
220 && (tolower(**string) <= tolower(**pattern)))
221 result = 0;
222
223 ++*pattern;
224 continue;
225 }
226
227 /* XXX: handle locale/MBCS comparison, advance by MBCS char width */
228 if ((**string == **pattern))
229 result = 0;
230 else if (nocase && (isupper(**string) || isupper(**pattern))
231 && (tolower(**string) == tolower(**pattern)))
232 result = 0;
233
234 ++*pattern;
235 }
236
237 /* NOT a properly balanced [expr] pattern; Rewind
238 * and reset result to test '[' literal
239 */
240 *pattern = mismatch;
241 result = FNM_NOMATCH;
242 }
243 else if (**pattern == '?') {
244 /* Optimize '?' match before unescaping **pattern */
245 if (!**string || (slash && (**string == '/')))
246 return FNM_NOMATCH;
247 result = 0;
248 goto fnmatch_ch_success;
249 }
250 else if (escape && (**pattern == '\\') && (*pattern)[1]) {
251 ++*pattern;
252 }
253
254 /* XXX: handle locale/MBCS comparison, advance by the MBCS char width */
255 if (**string == **pattern)
256 result = 0;
257 else if (nocase && (isupper(**string) || isupper(**pattern))
258 && (tolower(**string) == tolower(**pattern)))
259 result = 0;
260
261 /* Refuse to advance over trailing slash or nulls
262 */
263 if (!**string || !**pattern || (slash && ((**string == '/') || (**pattern == '/'))))
264 return result;
265
266 fnmatch_ch_success:
267 ++*pattern;
268 ++*string;
269 return result;
270 }
271
272
fnmatch(const char * pattern,const char * string,int flags)273 int fnmatch(const char *pattern, const char *string, int flags)
274 {
275 static const char dummystring[2] = {' ', 0};
276 const int escape = !(flags & FNM_NOESCAPE);
277 const int slash = !!(flags & FNM_PATHNAME);
278 const int leading_dir = !!(flags & FNM_LEADING_DIR);
279 const char *strendseg;
280 const char *dummyptr;
281 const char *matchptr;
282 int wild;
283 /* For '*' wild processing only; surpress 'used before initialization'
284 * warnings with dummy initialization values;
285 */
286 const char *strstartseg = NULL;
287 const char *mismatch = NULL;
288 int matchlen = 0;
289
290 if (strnlen(pattern, PATH_MAX) == PATH_MAX ||
291 strnlen(string, PATH_MAX) == PATH_MAX)
292 return (FNM_NOMATCH);
293
294 if (*pattern == '*')
295 goto firstsegment;
296
297 while (*pattern && *string)
298 {
299 /* Pre-decode "\/" which has no special significance, and
300 * match balanced slashes, starting a new segment pattern
301 */
302 if (slash && escape && (*pattern == '\\') && (pattern[1] == '/'))
303 ++pattern;
304 if (slash && (*pattern == '/') && (*string == '/')) {
305 ++pattern;
306 ++string;
307 }
308
309 firstsegment:
310 /* At the beginning of each segment, validate leading period behavior.
311 */
312 if ((flags & FNM_PERIOD) && (*string == '.'))
313 {
314 if (*pattern == '.')
315 ++pattern;
316 else if (escape && (*pattern == '\\') && (pattern[1] == '.'))
317 pattern += 2;
318 else
319 return FNM_NOMATCH;
320 ++string;
321 }
322
323 /* Determine the end of string segment
324 *
325 * Presumes '/' character is unique, not composite in any MBCS encoding
326 */
327 if (slash) {
328 strendseg = strchr(string, '/');
329 if (!strendseg)
330 strendseg = strchr(string, '\0');
331 }
332 else {
333 strendseg = strchr(string, '\0');
334 }
335
336 /* Allow pattern '*' to be consumed even with no remaining string to match
337 */
338 while (*pattern)
339 {
340 if ((string > strendseg)
341 || ((string == strendseg) && (*pattern != '*')))
342 break;
343
344 if (slash && ((*pattern == '/')
345 || (escape && (*pattern == '\\')
346 && (pattern[1] == '/'))))
347 break;
348
349 /* Reduce groups of '*' and '?' to n '?' matches
350 * followed by one '*' test for simplicity
351 */
352 for (wild = 0; ((*pattern == '*') || (*pattern == '?')); ++pattern)
353 {
354 if (*pattern == '*') {
355 wild = 1;
356 }
357 else if (string < strendseg) { /* && (*pattern == '?') */
358 /* XXX: Advance 1 char for MBCS locale */
359 ++string;
360 }
361 else { /* (string >= strendseg) && (*pattern == '?') */
362 return FNM_NOMATCH;
363 }
364 }
365
366 if (wild)
367 {
368 strstartseg = string;
369 mismatch = pattern;
370
371 /* Count fixed (non '*') char matches remaining in pattern
372 * excluding '/' (or "\/") and '*'
373 */
374 for (matchptr = pattern, matchlen = 0; 1; ++matchlen)
375 {
376 if ((*matchptr == '\0')
377 || (slash && ((*matchptr == '/')
378 || (escape && (*matchptr == '\\')
379 && (matchptr[1] == '/')))))
380 {
381 /* Compare precisely this many trailing string chars,
382 * the resulting match needs no wildcard loop
383 */
384 /* XXX: Adjust for MBCS */
385 if (string + matchlen > strendseg)
386 return FNM_NOMATCH;
387
388 string = strendseg - matchlen;
389 wild = 0;
390 break;
391 }
392
393 if (*matchptr == '*')
394 {
395 /* Ensure at least this many trailing string chars remain
396 * for the first comparison
397 */
398 /* XXX: Adjust for MBCS */
399 if (string + matchlen > strendseg)
400 return FNM_NOMATCH;
401
402 /* Begin first wild comparison at the current position */
403 break;
404 }
405
406 /* Skip forward in pattern by a single character match
407 * Use a dummy fnmatch_ch() test to count one "[range]" escape
408 */
409 /* XXX: Adjust for MBCS */
410 if (escape && (*matchptr == '\\') && matchptr[1]) {
411 matchptr += 2;
412 }
413 else if (*matchptr == '[') {
414 dummyptr = dummystring;
415 fnmatch_ch(&matchptr, &dummyptr, flags);
416 }
417 else {
418 ++matchptr;
419 }
420 }
421 }
422
423 /* Incrementally match string against the pattern
424 */
425 while (*pattern && (string < strendseg))
426 {
427 /* Success; begin a new wild pattern search
428 */
429 if (*pattern == '*')
430 break;
431
432 if (slash && ((*string == '/')
433 || (*pattern == '/')
434 || (escape && (*pattern == '\\')
435 && (pattern[1] == '/'))))
436 break;
437
438 /* Compare ch's (the pattern is advanced over "\/" to the '/',
439 * but slashes will mismatch, and are not consumed)
440 */
441 if (!fnmatch_ch(&pattern, &string, flags))
442 continue;
443
444 /* Failed to match, loop against next char offset of string segment
445 * until not enough string chars remain to match the fixed pattern
446 */
447 if (wild) {
448 /* XXX: Advance 1 char for MBCS locale */
449 string = ++strstartseg;
450 if (string + matchlen > strendseg)
451 return FNM_NOMATCH;
452
453 pattern = mismatch;
454 continue;
455 }
456 else
457 return FNM_NOMATCH;
458 }
459 }
460
461 if (*string && !((slash || leading_dir) && (*string == '/')))
462 return FNM_NOMATCH;
463
464 if (*pattern && !(slash && ((*pattern == '/')
465 || (escape && (*pattern == '\\')
466 && (pattern[1] == '/')))))
467 return FNM_NOMATCH;
468
469 if (leading_dir && !*pattern && *string == '/')
470 return 0;
471 }
472
473 /* Where both pattern and string are at EOS, declare success
474 */
475 if (!*string && !*pattern)
476 return 0;
477
478 /* pattern didn't match to the end of string */
479 return FNM_NOMATCH;
480 }
481