1 /*
2 * Copyright (c) 1989, 1993, 1994
3 * The Regents of the University of California. All rights reserved.
4 *
5 * This code is derived from software contributed to Berkeley by
6 * Guido van Rossum.
7 *
8 * Copyright (c) 2011 The FreeBSD Foundation
9 * All rights reserved.
10 * Portions of this software were developed by David Chisnall
11 * under sponsorship from the FreeBSD Foundation.
12 *
13 * Redistribution and use in source and binary forms, with or without
14 * modification, are permitted provided that the following conditions
15 * are met:
16 * 1. Redistributions of source code must retain the above copyright
17 * notice, this list of conditions and the following disclaimer.
18 * 2. Redistributions in binary form must reproduce the above copyright
19 * notice, this list of conditions and the following disclaimer in the
20 * documentation and/or other materials provided with the distribution.
21 * 3. Neither the name of the University nor the names of its contributors
22 * may be used to endorse or promote products derived from this software
23 * without specific prior written permission.
24 *
25 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
26 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
27 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
28 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
29 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
30 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
31 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
32 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
33 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
34 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
35 * SUCH DAMAGE.
36 *
37 * @(#)fnmatch.c 8.2 (Berkeley) 4/16/94
38 * $FreeBSD: head/lib/libc/gen/fnmatch.c 254091 2013-08-08 09:04:02Z ache $
39 */
40
41
42 /*
43 * Function fnmatch() as specified in POSIX 1003.2-1992, section B.6.
44 * Compares a filename or pathname to a pattern.
45 */
46
47 /*
48 * Some notes on multibyte character support:
49 * 1. Patterns with illegal byte sequences match nothing.
50 * 2. Illegal byte sequences in the "string" argument are handled by treating
51 * them as single-byte characters with a value of the first byte of the
52 * sequence cast to wchar_t.
53 * 3. Multibyte conversion state objects (mbstate_t) are passed around and
54 * used for most, but not all, conversions. Further work will be required
55 * to support state-dependent encodings.
56 */
57
58 #include <fnmatch.h>
59 #include <limits.h>
60 #include <string.h>
61 #include <wchar.h>
62 #include <wctype.h>
63
64 #include "collate.h"
65
66 #define EOS '\0'
67
68 #define RANGE_MATCH 1
69 #define RANGE_NOMATCH 0
70 #define RANGE_ERROR (-1)
71
72 static int rangematch(const char *, wchar_t, int, char **, mbstate_t *);
73 static int fnmatch1(const char *, const char *, const char *, int, mbstate_t,
74 mbstate_t);
75
76 int
fnmatch(const char * pattern,const char * string,int flags)77 fnmatch(const char *pattern, const char *string, int flags)
78 {
79 static const mbstate_t initial;
80
81 return (fnmatch1(pattern, string, string, flags, initial, initial));
82 }
83
84 static int
fnmatch1(const char * pattern,const char * string,const char * stringstart,int flags,mbstate_t patmbs,mbstate_t strmbs)85 fnmatch1(const char *pattern, const char *string, const char *stringstart,
86 int flags, mbstate_t patmbs, mbstate_t strmbs)
87 {
88 char *newp;
89 char c;
90 wchar_t pc, sc;
91 size_t pclen, sclen;
92
93 for (;;) {
94 pclen = mbrtowc(&pc, pattern, MB_LEN_MAX, &patmbs);
95 if (pclen == (size_t)-1 || pclen == (size_t)-2)
96 return (FNM_NOMATCH);
97 pattern += pclen;
98 sclen = mbrtowc(&sc, string, MB_LEN_MAX, &strmbs);
99 if (sclen == (size_t)-1 || sclen == (size_t)-2) {
100 sc = (unsigned char)*string;
101 sclen = 1;
102 memset(&strmbs, 0, sizeof(strmbs));
103 }
104 switch (pc) {
105 case EOS:
106 if ((flags & FNM_LEADING_DIR) && sc == '/')
107 return (0);
108 return (sc == EOS ? 0 : FNM_NOMATCH);
109 case '?':
110 if (sc == EOS)
111 return (FNM_NOMATCH);
112 if (sc == '/' && (flags & FNM_PATHNAME))
113 return (FNM_NOMATCH);
114 if (sc == '.' && (flags & FNM_PERIOD) &&
115 (string == stringstart ||
116 ((flags & FNM_PATHNAME) && *(string - 1) == '/')))
117 return (FNM_NOMATCH);
118 string += sclen;
119 break;
120 case '*':
121 c = *pattern;
122 /* Collapse multiple stars. */
123 while (c == '*')
124 c = *++pattern;
125
126 if (sc == '.' && (flags & FNM_PERIOD) &&
127 (string == stringstart ||
128 ((flags & FNM_PATHNAME) && *(string - 1) == '/')))
129 return (FNM_NOMATCH);
130
131 /* Optimize for pattern with * at end or before /. */
132 if (c == EOS)
133 if (flags & FNM_PATHNAME)
134 return ((flags & FNM_LEADING_DIR) ||
135 strchr(string, '/') == NULL ?
136 0 : FNM_NOMATCH);
137 else
138 return (0);
139 else if (c == '/' && flags & FNM_PATHNAME) {
140 if ((string = strchr(string, '/')) == NULL)
141 return (FNM_NOMATCH);
142 break;
143 }
144
145 /* General case, use recursion. */
146 while (sc != EOS) {
147 if (!fnmatch1(pattern, string, stringstart,
148 flags, patmbs, strmbs))
149 return (0);
150 sclen = mbrtowc(&sc, string, MB_LEN_MAX,
151 &strmbs);
152 if (sclen == (size_t)-1 ||
153 sclen == (size_t)-2) {
154 sc = (unsigned char)*string;
155 sclen = 1;
156 memset(&strmbs, 0, sizeof(strmbs));
157 }
158 if (sc == '/' && flags & FNM_PATHNAME)
159 break;
160 string += sclen;
161 }
162 return (FNM_NOMATCH);
163 case '[':
164 if (sc == EOS)
165 return (FNM_NOMATCH);
166 if (sc == '/' && (flags & FNM_PATHNAME))
167 return (FNM_NOMATCH);
168 if (sc == '.' && (flags & FNM_PERIOD) &&
169 (string == stringstart ||
170 ((flags & FNM_PATHNAME) && *(string - 1) == '/')))
171 return (FNM_NOMATCH);
172
173 switch (rangematch(pattern, sc, flags, &newp,
174 &patmbs)) {
175 case RANGE_ERROR:
176 goto norm;
177 case RANGE_MATCH:
178 pattern = newp;
179 break;
180 case RANGE_NOMATCH:
181 return (FNM_NOMATCH);
182 }
183 string += sclen;
184 break;
185 case '\\':
186 if (!(flags & FNM_NOESCAPE)) {
187 pclen = mbrtowc(&pc, pattern, MB_LEN_MAX,
188 &patmbs);
189 if (pclen == (size_t)-1 || pclen == (size_t)-2)
190 return (FNM_NOMATCH);
191 pattern += pclen;
192 }
193 /* FALLTHROUGH */
194 default:
195 norm:
196 if (pc == sc)
197 ;
198 else if ((flags & FNM_CASEFOLD) &&
199 (towlower(pc) == towlower(sc)))
200 ;
201 else
202 return (FNM_NOMATCH);
203 string += sclen;
204 break;
205 }
206 }
207 /* NOTREACHED */
208 }
209
210 static int
rangematch(const char * pattern,wchar_t test,int flags,char ** newp,mbstate_t * patmbs)211 rangematch(const char *pattern, wchar_t test, int flags, char **newp,
212 mbstate_t *patmbs)
213 {
214 int negate, ok;
215 wchar_t c, c2;
216 size_t pclen;
217 const char *origpat;
218 struct xlocale_collate *table =
219 (struct xlocale_collate*)__get_locale()->components[XLC_COLLATE];
220
221 /*
222 * A bracket expression starting with an unquoted circumflex
223 * character produces unspecified results (IEEE 1003.2-1992,
224 * 3.13.2). This implementation treats it like '!', for
225 * consistency with the regular expression syntax.
226 * J.T. Conklin (conklin@ngai.kaleida.com)
227 */
228 if ((negate = (*pattern == '!' || *pattern == '^')))
229 ++pattern;
230
231 if (flags & FNM_CASEFOLD)
232 test = towlower(test);
233
234 /*
235 * A right bracket shall lose its special meaning and represent
236 * itself in a bracket expression if it occurs first in the list.
237 * -- POSIX.2 2.8.3.2
238 */
239 ok = 0;
240 origpat = pattern;
241 for (;;) {
242 if (*pattern == ']' && pattern > origpat) {
243 pattern++;
244 break;
245 } else if (*pattern == '\0') {
246 return (RANGE_ERROR);
247 } else if (*pattern == '/' && (flags & FNM_PATHNAME)) {
248 return (RANGE_NOMATCH);
249 } else if (*pattern == '\\' && !(flags & FNM_NOESCAPE))
250 pattern++;
251 pclen = mbrtowc(&c, pattern, MB_LEN_MAX, patmbs);
252 if (pclen == (size_t)-1 || pclen == (size_t)-2)
253 return (RANGE_NOMATCH);
254 pattern += pclen;
255
256 if (flags & FNM_CASEFOLD)
257 c = towlower(c);
258
259 if (*pattern == '-' && *(pattern + 1) != EOS &&
260 *(pattern + 1) != ']') {
261 if (*++pattern == '\\' && !(flags & FNM_NOESCAPE))
262 if (*pattern != EOS)
263 pattern++;
264 pclen = mbrtowc(&c2, pattern, MB_LEN_MAX, patmbs);
265 if (pclen == (size_t)-1 || pclen == (size_t)-2)
266 return (RANGE_NOMATCH);
267 pattern += pclen;
268 if (c2 == EOS)
269 return (RANGE_ERROR);
270
271 if (flags & FNM_CASEFOLD)
272 c2 = towlower(c2);
273
274 if (table->__collate_load_error ?
275 c <= test && test <= c2 :
276 __wcollate_range_cmp(c, test) <= 0
277 && __wcollate_range_cmp(test, c2) <= 0
278 )
279 ok = 1;
280 } else if (c == test)
281 ok = 1;
282 }
283
284 *newp = (char *)pattern;
285 return (ok == negate ? RANGE_NOMATCH : RANGE_MATCH);
286 }
287