posix/fnmatch.c

   1 /* Copyright (C) 1991, 92, 93, 96, 97, 98, 99 Free Software Foundation, Inc.
   2    This file is part of the GNU C Library.
   3
   4    This library is free software; you can redistribute it and/or
   5    modify it under the terms of the GNU Library General Public License as
   6    published by the Free Software Foundation; either version 2 of the
   7    License, or (at your option) any later version.
   8
   9    This library is distributed in the hope that it will be useful,
  10    but WITHOUT ANY WARRANTY; without even the implied warranty of
  11    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  12    Library General Public License for more details.
  13
  14    You should have received a copy of the GNU Library General Public
  15    License along with this library; see the file COPYING.LIB.  If not,
  16    write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  17    Boston, MA 02111-1307, USA.  */
  18
  19 #if HAVE_CONFIG_H
  20 # include <config.h>
  21 #endif
  22
  23 /* Enable GNU extensions in fnmatch.h.  */
  24 #ifndef _GNU_SOURCE
  25 # define _GNU_SOURCE    1
  26 #endif
  27
  28 #include <errno.h>
  29 #include <fnmatch.h>
  30 #include <ctype.h>
  31
  32 #if HAVE_STRING_H || defined _LIBC
  33 # include <string.h>
  34 #else
  35 # include <strings.h>
  36 #endif
  37
  38 #if defined STDC_HEADERS || defined _LIBC
  39 # include <stdlib.h>
  40 #endif
  41
  42 /* For platform which support the ISO C amendement 1 functionality we
  43    support user defined character classes.  */
  44 #if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H)
  45 /* Solaris 2.5 has a bug: <wchar.h> must be included before <wctype.h>.  */
  46 # include <wchar.h>
  47 # include <wctype.h>
  48 #endif
  49
  50 /* Comment out all this code if we are using the GNU C Library, and are not
  51    actually compiling the library itself.  This code is part of the GNU C
  52    Library, but also included in many other GNU distributions.  Compiling
  53    and linking in this code is a waste when using the GNU C library
  54    (especially if it is a shared library).  Rather than having every GNU
  55    program understand `configure --with-gnu-libc' and omit the object files,
  56    it is simpler to just do this in the source for each such file.  */
  57
  58 #if defined _LIBC || !defined __GNU_LIBRARY__
  59
  60
  61 # if defined STDC_HEADERS || !defined isascii
  62 #  define ISASCII(c) 1
  63 # else
  64 #  define ISASCII(c) isascii(c)
  65 # endif
  66
  67 #ifdef isblank
  68 # define ISBLANK(c) (ISASCII (c) && isblank (c))
  69 #else
  70 # define ISBLANK(c) ((c) == ' ' || (c) == '\t')
  71 #endif
  72 #ifdef isgraph
  73 # define ISGRAPH(c) (ISASCII (c) && isgraph (c))
  74 #else
  75 # define ISGRAPH(c) (ISASCII (c) && isprint (c) && !isspace (c))
  76 #endif
  77
  78 #define ISPRINT(c) (ISASCII (c) && isprint (c))
  79 #define ISDIGIT(c) (ISASCII (c) && isdigit (c))
  80 #define ISALNUM(c) (ISASCII (c) && isalnum (c))
  81 #define ISALPHA(c) (ISASCII (c) && isalpha (c))
  82 #define ISCNTRL(c) (ISASCII (c) && iscntrl (c))
  83 #define ISLOWER(c) (ISASCII (c) && islower (c))
  84 #define ISPUNCT(c) (ISASCII (c) && ispunct (c))
  85 #define ISSPACE(c) (ISASCII (c) && isspace (c))
  86 #define ISUPPER(c) (ISASCII (c) && isupper (c))
  87 #define ISXDIGIT(c) (ISASCII (c) && isxdigit (c))
  88
  89 # define STREQ(s1, s2) ((strcmp (s1, s2) == 0))
  90
  91 # if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H)
  92 /* The GNU C library provides support for user-defined character classes
  93    and the functions from ISO C amendement 1.  */
  94 #  ifdef CHARCLASS_NAME_MAX
  95 #   define CHAR_CLASS_MAX_LENGTH CHARCLASS_NAME_MAX
  96 #  else
  97 /* This shouldn't happen but some implementation might still have this
  98    problem.  Use a reasonable default value.  */
  99 #   define CHAR_CLASS_MAX_LENGTH 256
 100 #  endif
 101
 102 #  ifdef _LIBC
 103 #   define IS_CHAR_CLASS(string) __wctype (string)
 104 #  else
 105 #   define IS_CHAR_CLASS(string) wctype (string)
 106 #  endif
 107 # else
 108 #  define CHAR_CLASS_MAX_LENGTH  6 /* Namely, `xdigit'.  */
 109
 110 #  define IS_CHAR_CLASS(string)                                               \
 111    (STREQ (string, "alpha") || STREQ (string, "upper")                        \
 112     || STREQ (string, "lower") || STREQ (string, "digit")                     \
 113     || STREQ (string, "alnum") || STREQ (string, "xdigit")                    \
 114     || STREQ (string, "space") || STREQ (string, "print")                     \
 115     || STREQ (string, "punct") || STREQ (string, "graph")                     \
 116     || STREQ (string, "cntrl") || STREQ (string, "blank"))
 117 # endif
 118
 119 /* Avoid depending on library functions or files
 120    whose names are inconsistent.  */
 121
 122 # if !defined _LIBC && !defined getenv
 123 extern char *getenv ();
 124 # endif
 125
 126 # ifndef errno
 127 extern int errno;
 128 # endif
 129
 130 /* Match STRING against the filename pattern PATTERN, returning zero if
 131    it matches, nonzero if not.  */
 132 static int
 133 #ifdef _LIBC
 134 internal_function
 135 #endif
 136 internal_fnmatch (const char *pattern, const char *string,
 137                   int no_leading_period, int flags)
 138 {
 139   register const char *p = pattern, *n = string;
 140   register unsigned char c;
 141
 142 /* Note that this evaluates C many times.  */
 143 # ifdef _LIBC
 144 #  define FOLD(c) ((flags & FNM_CASEFOLD) ? tolower (c) : (c))
 145 # else
 146 #  define FOLD(c) ((flags & FNM_CASEFOLD) && ISUPPER (c) ? tolower (c) : (c))
 147 # endif
 148
 149   while ((c = *p++) != '\0')
 150     {
 151       c = FOLD (c);
 152
 153       switch (c)
 154         {
 155         case '?':
 156           if (*n == '\0')
 157             return FNM_NOMATCH;
 158           else if (*n == '/' && (flags & FNM_FILE_NAME))
 159             return FNM_NOMATCH;
 160           else if (*n == '.' && no_leading_period
 161                    && (n == string
 162                        || (n[-1] == '/' && (flags & FNM_FILE_NAME))))
 163             return FNM_NOMATCH;
 164           break;
 165
 166         case '\\':
 167           if (!(flags & FNM_NOESCAPE))
 168             {
 169               c = *p++;
 170               if (c == '\0')
 171                 /* Trailing \ loses.  */
 172                 return FNM_NOMATCH;
 173               c = FOLD (c);
 174             }
 175           if (FOLD ((unsigned char) *n) != c)
 176             return FNM_NOMATCH;
 177           break;
 178
 179         case '*':
 180           if (*n == '.' && no_leading_period
 181               && (n == string
 182                   || (n[-1] == '/' && (flags & FNM_FILE_NAME))))
 183             return FNM_NOMATCH;
 184
 185           for (c = *p++; c == '?' || c == '*'; c = *p++)
 186             {
 187               if (*n == '/' && (flags & FNM_FILE_NAME))
 188                 /* A slash does not match a wildcard under FNM_FILE_NAME.  */
 189                 return FNM_NOMATCH;
 190               else if (c == '?')
 191                 {
 192                   /* A ? needs to match one character.  */
 193                   if (*n == '\0')
 194                     /* There isn't another character; no match.  */
 195                     return FNM_NOMATCH;
 196                   else
 197                     /* One character of the string is consumed in matching
 198                        this ? wildcard, so *??? won't match if there are
 199                        less than three characters.  */
 200                     ++n;
 201                 }
 202             }
 203
 204           if (c == '\0')
 205             /* The wildcard(s) is/are the last element of the pattern.
 206                If the name is a file name and contains another slash
 207                this does mean it cannot match.  */
 208             return ((flags & FNM_FILE_NAME) && strchr (n, '/') != NULL
 209                     ? FNM_NOMATCH : 0);
 210           else
 211             {
 212               const char *endp;
 213
 214               endp = __strchrnul (n, (flags & FNM_FILE_NAME) ? '/' : '\0');
 215
 216               if (c == '[')
 217                 {
 218                   int flags2 = ((flags & FNM_FILE_NAME)
 219                                 ? flags : (flags & ~FNM_PERIOD));
 220
 221                   for (--p; n < endp; ++n)
 222                     if (internal_fnmatch (p, n,
 223                                           (no_leading_period
 224                                            && (n == string
 225                                                || (n[-1] == '/'
 226                                                    && (flags
 227                                                        & FNM_FILE_NAME)))),
 228                                           flags2)
 229                         == 0)
 230                       return 0;
 231                 }
 232               else if (c == '/' && (flags & FNM_FILE_NAME))
 233                 {
 234                   while (*n != '\0' && *n != '/')
 235                     ++n;
 236                   if (*n == '/'
 237                       && (internal_fnmatch (p, n + 1, flags & FNM_PERIOD,
 238                                             flags) == 0))
 239                     return 0;
 240                 }
 241               else
 242                 {
 243                   int flags2 = ((flags & FNM_FILE_NAME)
 244                                 ? flags : (flags & ~FNM_PERIOD));
 245
 246                   if (c == '\\' && !(flags & FNM_NOESCAPE))
 247                     c = *p;
 248                   c = FOLD (c);
 249                   for (--p; n < endp; ++n)
 250                     if (FOLD ((unsigned char) *n) == c
 251                         && (internal_fnmatch (p, n,
 252                                               (no_leading_period
 253                                                && (n == string
 254                                                    || (n[-1] == '/'
 255                                                        && (flags
 256                                                            & FNM_FILE_NAME)))),
 257                                               flags2) == 0))
 258                       return 0;
 259                 }
 260             }
 261
 262           /* If we come here no match is possible with the wildcard.  */
 263           return FNM_NOMATCH;
 264
 265         case '[':
 266           {
 267             /* Nonzero if the sense of the character class is inverted.  */
 268             static int posixly_correct;
 269             register int not;
 270             char cold;
 271
 272             if (posixly_correct == 0)
 273               posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1;
 274
 275             if (*n == '\0')
 276               return FNM_NOMATCH;
 277
 278             if (*n == '.' && no_leading_period && (n == string
 279                                                    || (n[-1] == '/'
 280                                                        && (flags
 281                                                            & FNM_FILE_NAME))))
 282               return FNM_NOMATCH;
 283
 284             if (*n == '/' && (flags & FNM_FILE_NAME))
 285               /* `/' cannot be matched.  */
 286               return FNM_NOMATCH;
 287
 288             not = (*p == '!' || (posixly_correct < 0 && *p == '^'));
 289             if (not)
 290               ++p;
 291
 292             c = *p++;
 293             for (;;)
 294               {
 295                 unsigned char fn = FOLD ((unsigned char) *n);
 296
 297                 if (!(flags & FNM_NOESCAPE) && c == '\\')
 298                   {
 299                     if (*p == '\0')
 300                       return FNM_NOMATCH;
 301                     c = FOLD ((unsigned char) *p);
 302                     ++p;
 303
 304                     if (c == fn)
 305                       goto matched;
 306                   }
 307                 else if (c == '[' && *p == ':')
 308                   {
 309                     /* Leave room for the null.  */
 310                     char str[CHAR_CLASS_MAX_LENGTH + 1];
 311                     size_t c1 = 0;
 312 # if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H)
 313                     wctype_t wt;
 314 # endif
 315                     const char *startp = p;
 316
 317                     for (;;)
 318                       {
 319                         if (c1 == CHAR_CLASS_MAX_LENGTH)
 320                           /* The name is too long and therefore the pattern
 321                              is ill-formed.  */
 322                           return FNM_NOMATCH;
 323
 324                         c = *++p;
 325                         if (c == ':' && p[1] == ']')
 326                           {
 327                             p += 2;
 328                             break;
 329                           }
 330                         if (c < 'a' || c >= 'z')
 331                           {
 332                             /* This cannot possibly be a character class name.
 333                                Match it as a normal range.  */
 334                             p = startp;
 335                             c = '[';
 336                             goto normal_bracket;
 337                           }
 338                         str[c1++] = c;
 339                       }
 340                     str[c1] = '\0';
 341
 342 # if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H)
 343                     wt = IS_CHAR_CLASS (str);
 344                     if (wt == 0)
 345                       /* Invalid character class name.  */
 346                       return FNM_NOMATCH;
 347
 348                     if (__iswctype (__btowc ((unsigned char) *n), wt))
 349                       goto matched;
 350 # else
 351                     if ((STREQ (str, "alnum") && ISALNUM ((unsigned char) *n))
 352                         || (STREQ (str, "alpha") && ISALPHA ((unsigned char) *n))
 353                         || (STREQ (str, "blank") && ISBLANK ((unsigned char) *n))
 354                         || (STREQ (str, "cntrl") && ISCNTRL ((unsigned char) *n))
 355                         || (STREQ (str, "digit") && ISDIGIT ((unsigned char) *n))
 356                         || (STREQ (str, "graph") && ISGRAPH ((unsigned char) *n))
 357                         || (STREQ (str, "lower") && ISLOWER ((unsigned char) *n))
 358                         || (STREQ (str, "print") && ISPRINT ((unsigned char) *n))
 359                         || (STREQ (str, "punct") && ISPUNCT ((unsigned char) *n))
 360                         || (STREQ (str, "space") && ISSPACE ((unsigned char) *n))
 361                         || (STREQ (str, "upper") && ISUPPER ((unsigned char) *n))
 362                         || (STREQ (str, "xdigit") && ISXDIGIT ((unsigned char) *n)))
 363                       goto matched;
 364 # endif
 365                   }
 366                 else if (c == '\0')
 367                   /* [ (unterminated) loses.  */
 368                   return FNM_NOMATCH;
 369                 else
 370                   {
 371                   normal_bracket:
 372                     if (FOLD (c) == fn)
 373                       goto matched;
 374
 375                     cold = c;
 376                     c = *p++;
 377
 378                     if (c == '-' && *p != ']')
 379                       {
 380                         /* It is a range.  */
 381                         unsigned char cend = *p++;
 382                         if (!(flags & FNM_NOESCAPE) && cend == '\\')
 383                           cend = *p++;
 384                         if (cend == '\0')
 385                           return FNM_NOMATCH;
 386
 387                         if (cold <= fn && fn <= FOLD (cend))
 388                           goto matched;
 389
 390                         c = *p++;
 391                       }
 392                   }
 393
 394                 if (c == ']')
 395                   break;
 396               }
 397
 398             if (!not)
 399               return FNM_NOMATCH;
 400             break;
 401
 402           matched:
 403             /* Skip the rest of the [...] that already matched.  */
 404             while (c != ']')
 405               {
 406                 if (c == '\0')
 407                   /* [... (unterminated) loses.  */
 408                   return FNM_NOMATCH;
 409
 410                 c = *p++;
 411                 if (!(flags & FNM_NOESCAPE) && c == '\\')
 412                   {
 413                     if (*p == '\0')
 414                       return FNM_NOMATCH;
 415                     /* XXX 1003.2d11 is unclear if this is right.  */
 416                     ++p;
 417                   }
 418                 else if (c == '[' && *p == ':')
 419                   {
 420                     do
 421                       if (*++p == '\0')
 422                         return FNM_NOMATCH;
 423                     while (*p != ':' || p[1] == ']');
 424                     p += 2;
 425                     c = *p;
 426                   }
 427               }
 428             if (not)
 429               return FNM_NOMATCH;
 430           }
 431           break;
 432
 433         default:
 434           if (c != FOLD ((unsigned char) *n))
 435             return FNM_NOMATCH;
 436         }
 437
 438       ++n;
 439     }
 440
 441   if (*n == '\0')
 442     return 0;
 443
 444   if ((flags & FNM_LEADING_DIR) && *n == '/')
 445     /* The FNM_LEADING_DIR flag says that "foo*" matches "foobar/frobozz".  */
 446     return 0;
 447
 448   return FNM_NOMATCH;
 449
 450 # undef FOLD
 451 }
 452
 453
 454 int
 455 fnmatch (pattern, string, flags)
 456      const char *pattern;
 457      const char *string;
 458      int flags;
 459 {
 460   return internal_fnmatch (pattern, string, flags & FNM_PERIOD, flags);
 461 }
 462
 463 #endif  /* _LIBC or not __GNU_LIBRARY__.  */