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 38 #if defined(LIBC_SCCS) && !defined(lint) 39 static char sccsid[] = "@(#)fnmatch.c 8.2 (Berkeley) 4/16/94"; 40 #endif /* LIBC_SCCS and not lint */ 41 #include <sys/cdefs.h> 42 __FBSDID("$FreeBSD$"); 43 44 /* 45 * Function fnmatch() as specified in POSIX 1003.2-1992, section B.6. 46 * Compares a filename or pathname to a pattern. 47 */ 48 49 /* 50 * Some notes on multibyte character support: 51 * 1. Patterns with illegal byte sequences match nothing. 52 * 2. Illegal byte sequences in the "string" argument are handled by treating 53 * them as single-byte characters with a value of the first byte of the 54 * sequence cast to wchar_t. 55 * 3. Multibyte conversion state objects (mbstate_t) are passed around and 56 * used for most, but not all, conversions. Further work will be required 57 * to support state-dependent encodings. 58 */ 59 60 #include <fnmatch.h> 61 #include <limits.h> 62 #include <string.h> 63 #include <wchar.h> 64 #include <wctype.h> 65 66 #include "collate.h" 67 68 #define EOS '\0' 69 70 #define RANGE_MATCH 1 71 #define RANGE_NOMATCH 0 72 #define RANGE_ERROR (-1) 73 74 static int rangematch(const char *, wchar_t, int, char **, mbstate_t *); 75 static int fnmatch1(const char *, const char *, const char *, int, mbstate_t, 76 mbstate_t); 77 78 int 79 fnmatch(const char *pattern, const char *string, int flags) 80 { 81 static const mbstate_t initial; 82 83 return (fnmatch1(pattern, string, string, flags, initial, initial)); 84 } 85 86 static int 87 fnmatch1(const char *pattern, const char *string, const char *stringstart, 88 int flags, mbstate_t patmbs, mbstate_t strmbs) 89 { 90 const char *bt_pattern, *bt_string; 91 mbstate_t bt_patmbs, bt_strmbs; 92 char *newp; 93 char c; 94 wchar_t pc, sc; 95 size_t pclen, sclen; 96 97 bt_pattern = bt_string = NULL; 98 for (;;) { 99 pclen = mbrtowc(&pc, pattern, MB_LEN_MAX, &patmbs); 100 if (pclen == (size_t)-1 || pclen == (size_t)-2) 101 return (FNM_NOMATCH); 102 pattern += pclen; 103 sclen = mbrtowc(&sc, string, MB_LEN_MAX, &strmbs); 104 if (sclen == (size_t)-1 || sclen == (size_t)-2) { 105 sc = (unsigned char)*string; 106 sclen = 1; 107 memset(&strmbs, 0, sizeof(strmbs)); 108 } 109 switch (pc) { 110 case EOS: 111 if ((flags & FNM_LEADING_DIR) && sc == '/') 112 return (0); 113 if (sc == EOS) 114 return (0); 115 goto backtrack; 116 case '?': 117 if (sc == EOS) 118 return (FNM_NOMATCH); 119 if (sc == '/' && (flags & FNM_PATHNAME)) 120 goto backtrack; 121 if (sc == '.' && (flags & FNM_PERIOD) && 122 (string == stringstart || 123 ((flags & FNM_PATHNAME) && *(string - 1) == '/'))) 124 goto backtrack; 125 string += sclen; 126 break; 127 case '*': 128 c = *pattern; 129 /* Collapse multiple stars. */ 130 while (c == '*') 131 c = *++pattern; 132 133 if (sc == '.' && (flags & FNM_PERIOD) && 134 (string == stringstart || 135 ((flags & FNM_PATHNAME) && *(string - 1) == '/'))) 136 goto backtrack; 137 138 /* Optimize for pattern with * at end or before /. */ 139 if (c == EOS) 140 if (flags & FNM_PATHNAME) 141 return ((flags & FNM_LEADING_DIR) || 142 strchr(string, '/') == NULL ? 143 0 : FNM_NOMATCH); 144 else 145 return (0); 146 else if (c == '/' && flags & FNM_PATHNAME) { 147 if ((string = strchr(string, '/')) == NULL) 148 return (FNM_NOMATCH); 149 break; 150 } 151 152 /* 153 * First try the shortest match for the '*' that 154 * could work. We can forget any earlier '*' since 155 * there is no way having it match more characters 156 * can help us, given that we are already here. 157 */ 158 bt_pattern = pattern, bt_patmbs = patmbs; 159 bt_string = string, bt_strmbs = strmbs; 160 break; 161 case '[': 162 if (sc == EOS) 163 return (FNM_NOMATCH); 164 if (sc == '/' && (flags & FNM_PATHNAME)) 165 goto backtrack; 166 if (sc == '.' && (flags & FNM_PERIOD) && 167 (string == stringstart || 168 ((flags & FNM_PATHNAME) && *(string - 1) == '/'))) 169 goto backtrack; 170 171 switch (rangematch(pattern, sc, flags, &newp, 172 &patmbs)) { 173 case RANGE_ERROR: 174 goto norm; 175 case RANGE_MATCH: 176 pattern = newp; 177 break; 178 case RANGE_NOMATCH: 179 goto backtrack; 180 } 181 string += sclen; 182 break; 183 case '\\': 184 if (!(flags & FNM_NOESCAPE)) { 185 pclen = mbrtowc(&pc, pattern, MB_LEN_MAX, 186 &patmbs); 187 if (pclen == 0 || pclen == (size_t)-1 || 188 pclen == (size_t)-2) 189 return (FNM_NOMATCH); 190 pattern += pclen; 191 } 192 /* FALLTHROUGH */ 193 default: 194 norm: 195 string += sclen; 196 if (pc == sc) 197 ; 198 else if ((flags & FNM_CASEFOLD) && 199 (towlower(pc) == towlower(sc))) 200 ; 201 else { 202 backtrack: 203 /* 204 * If we have a mismatch (other than hitting 205 * the end of the string), go back to the last 206 * '*' seen and have it match one additional 207 * character. 208 */ 209 if (bt_pattern == NULL) 210 return (FNM_NOMATCH); 211 sclen = mbrtowc(&sc, bt_string, MB_LEN_MAX, 212 &bt_strmbs); 213 if (sclen == (size_t)-1 || 214 sclen == (size_t)-2) { 215 sc = (unsigned char)*bt_string; 216 sclen = 1; 217 memset(&bt_strmbs, 0, 218 sizeof(bt_strmbs)); 219 } 220 if (sc == EOS) 221 return (FNM_NOMATCH); 222 if (sc == '/' && flags & FNM_PATHNAME) 223 return (FNM_NOMATCH); 224 bt_string += sclen; 225 pattern = bt_pattern, patmbs = bt_patmbs; 226 string = bt_string, strmbs = bt_strmbs; 227 } 228 break; 229 } 230 } 231 /* NOTREACHED */ 232 } 233 234 static int 235 rangematch(const char *pattern, wchar_t test, int flags, char **newp, 236 mbstate_t *patmbs) 237 { 238 int negate, ok; 239 wchar_t c, c2; 240 size_t pclen; 241 const char *origpat; 242 struct xlocale_collate *table = 243 (struct xlocale_collate*)__get_locale()->components[XLC_COLLATE]; 244 245 /* 246 * A bracket expression starting with an unquoted circumflex 247 * character produces unspecified results (IEEE 1003.2-1992, 248 * 3.13.2). This implementation treats it like '!', for 249 * consistency with the regular expression syntax. 250 * J.T. Conklin (conklin@ngai.kaleida.com) 251 */ 252 if ((negate = (*pattern == '!' || *pattern == '^'))) 253 ++pattern; 254 255 if (flags & FNM_CASEFOLD) 256 test = towlower(test); 257 258 /* 259 * A right bracket shall lose its special meaning and represent 260 * itself in a bracket expression if it occurs first in the list. 261 * -- POSIX.2 2.8.3.2 262 */ 263 ok = 0; 264 origpat = pattern; 265 for (;;) { 266 if (*pattern == ']' && pattern > origpat) { 267 pattern++; 268 break; 269 } else if (*pattern == '\0') { 270 return (RANGE_ERROR); 271 } else if (*pattern == '/' && (flags & FNM_PATHNAME)) { 272 return (RANGE_NOMATCH); 273 } else if (*pattern == '\\' && !(flags & FNM_NOESCAPE)) 274 pattern++; 275 pclen = mbrtowc(&c, pattern, MB_LEN_MAX, patmbs); 276 if (pclen == (size_t)-1 || pclen == (size_t)-2) 277 return (RANGE_NOMATCH); 278 pattern += pclen; 279 280 if (flags & FNM_CASEFOLD) 281 c = towlower(c); 282 283 if (*pattern == '-' && *(pattern + 1) != EOS && 284 *(pattern + 1) != ']') { 285 if (*++pattern == '\\' && !(flags & FNM_NOESCAPE)) 286 if (*pattern != EOS) 287 pattern++; 288 pclen = mbrtowc(&c2, pattern, MB_LEN_MAX, patmbs); 289 if (pclen == (size_t)-1 || pclen == (size_t)-2) 290 return (RANGE_NOMATCH); 291 pattern += pclen; 292 if (c2 == EOS) 293 return (RANGE_ERROR); 294 295 if (flags & FNM_CASEFOLD) 296 c2 = towlower(c2); 297 298 if (table->__collate_load_error ? 299 c <= test && test <= c2 : 300 __wcollate_range_cmp(c, test) <= 0 301 && __wcollate_range_cmp(test, c2) <= 0 302 ) 303 ok = 1; 304 } else if (c == test) 305 ok = 1; 306 } 307 308 *newp = (char *)pattern; 309 return (ok == negate ? RANGE_NOMATCH : RANGE_MATCH); 310 } 311