compat/fnmatch/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 platforms which support the ISO C amendment 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 amendment 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 /* This function doesn't exist on most systems.  */
 131
 132 # if !defined HAVE___STRCHRNUL && !defined _LIBC
 133 static char *
 134 __strchrnul (s, c)
 135      const char *s;
 136      int c;
 137 {
 138   char *result = strchr (s, c);
 139   if (result == NULL)
 140     result = strchr (s, '\0');
 141   return result;
 142 }
 143 # endif
 144
 145 # ifndef internal_function
 146 /* Inside GNU libc we mark some function in a special way.  In other
 147    environments simply ignore the marking.  */
 148 #  define internal_function
 149 # endif
 150
 151 /* Match STRING against the filename pattern PATTERN, returning zero if
 152    it matches, nonzero if not.  */
 153 static int internal_fnmatch __P ((const char *pattern, const char *string,
 154                                   int no_leading_period, int flags))
 155      internal_function;
 156 static int
 157 internal_function
 158 internal_fnmatch (pattern, string, no_leading_period, flags)
 159      const char *pattern;
 160      const char *string;
 161      int no_leading_period;
 162      int flags;
 163 {
 164   register const char *p = pattern, *n = string;
 165   register unsigned char c;
 166
 167 /* Note that this evaluates C many times.  */
 168 # ifdef _LIBC
 169 #  define FOLD(c) ((flags & FNM_CASEFOLD) ? tolower (c) : (c))
 170 # else
 171 #  define FOLD(c) ((flags & FNM_CASEFOLD) && ISUPPER (c) ? tolower (c) : (c))
 172 # endif
 173
 174   while ((c = *p++) != '\0')
 175     {
 176       c = FOLD (c);
 177
 178       switch (c)
 179         {
 180         case '?':
 181           if (*n == '\0')
 182             return FNM_NOMATCH;
 183           else if (*n == '/' && (flags & FNM_FILE_NAME))
 184             return FNM_NOMATCH;
 185           else if (*n == '.' && no_leading_period
 186                    && (n == string
 187                        || (n[-1] == '/' && (flags & FNM_FILE_NAME))))
 188             return FNM_NOMATCH;
 189           break;
 190
 191         case '\\':
 192           if (!(flags & FNM_NOESCAPE))
 193             {
 194               c = *p++;
 195               if (c == '\0')
 196                 /* Trailing \ loses.  */
 197                 return FNM_NOMATCH;
 198               c = FOLD (c);
 199             }
 200           if (FOLD ((unsigned char) *n) != c)
 201             return FNM_NOMATCH;
 202           break;
 203
 204         case '*':
 205           if (*n == '.' && no_leading_period
 206               && (n == string
 207                   || (n[-1] == '/' && (flags & FNM_FILE_NAME))))
 208             return FNM_NOMATCH;
 209
 210           for (c = *p++; c == '?' || c == '*'; c = *p++)
 211             {
 212               if (*n == '/' && (flags & FNM_FILE_NAME))
 213                 /* A slash does not match a wildcard under FNM_FILE_NAME.  */
 214                 return FNM_NOMATCH;
 215               else if (c == '?')
 216                 {
 217                   /* A ? needs to match one character.  */
 218                   if (*n == '\0')
 219                     /* There isn't another character; no match.  */
 220                     return FNM_NOMATCH;
 221                   else
 222                     /* One character of the string is consumed in matching
 223                        this ? wildcard, so *??? won't match if there are
 224                        less than three characters.  */
 225                     ++n;
 226                 }
 227             }
 228
 229           if (c == '\0')
 230             /* The wildcard(s) is/are the last element of the pattern.
 231                If the name is a file name and contains another slash
 232                this does mean it cannot match.  */
 233             return ((flags & FNM_FILE_NAME) && strchr (n, '/') != NULL
 234                     ? FNM_NOMATCH : 0);
 235           else
 236             {
 237               const char *endp;
 238
 239               endp = __strchrnul (n, (flags & FNM_FILE_NAME) ? '/' : '\0');
 240
 241               if (c == '[')
 242                 {
 243                   int flags2 = ((flags & FNM_FILE_NAME)
 244                                 ? flags : (flags & ~FNM_PERIOD));
 245
 246                   for (--p; n < endp; ++n)
 247                     if (internal_fnmatch (p, n,
 248                                           (no_leading_period
 249                                            && (n == string
 250                                                || (n[-1] == '/'
 251                                                    && (flags
 252                                                        & FNM_FILE_NAME)))),
 253                                           flags2)
 254                         == 0)
 255                       return 0;
 256                 }
 257               else if (c == '/' && (flags & FNM_FILE_NAME))
 258                 {
 259                   while (*n != '\0' && *n != '/')
 260                     ++n;
 261                   if (*n == '/'
 262                       && (internal_fnmatch (p, n + 1, flags & FNM_PERIOD,
 263                                             flags) == 0))
 264                     return 0;
 265                 }
 266               else
 267                 {
 268                   int flags2 = ((flags & FNM_FILE_NAME)
 269                                 ? flags : (flags & ~FNM_PERIOD));
 270
 271                   if (c == '\\' && !(flags & FNM_NOESCAPE))
 272                     c = *p;
 273                   c = FOLD (c);
 274                   for (--p; n < endp; ++n)
 275                     if (FOLD ((unsigned char) *n) == c
 276                         && (internal_fnmatch (p, n,
 277                                               (no_leading_period
 278                                                && (n == string
 279                                                    || (n[-1] == '/'
 280                                                        && (flags
 281                                                            & FNM_FILE_NAME)))),
 282                                               flags2) == 0))
 283                       return 0;
 284                 }
 285             }
 286
 287           /* If we come here no match is possible with the wildcard.  */
 288           return FNM_NOMATCH;
 289
 290         case '[':
 291           {
 292             /* Nonzero if the sense of the character class is inverted.  */
 293             static int posixly_correct;
 294             register int not;
 295             char cold;
 296
 297             if (posixly_correct == 0)
 298               posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1;
 299
 300             if (*n == '\0')
 301               return FNM_NOMATCH;
 302
 303             if (*n == '.' && no_leading_period && (n == string
 304                                                    || (n[-1] == '/'
 305                                                        && (flags
 306                                                            & FNM_FILE_NAME))))
 307               return FNM_NOMATCH;
 308
 309             if (*n == '/' && (flags & FNM_FILE_NAME))
 310               /* `/' cannot be matched.  */
 311               return FNM_NOMATCH;
 312
 313             not = (*p == '!' || (posixly_correct < 0 && *p == '^'));
 314             if (not)
 315               ++p;
 316
 317             c = *p++;
 318             for (;;)
 319               {
 320                 unsigned char fn = FOLD ((unsigned char) *n);
 321
 322                 if (!(flags & FNM_NOESCAPE) && c == '\\')
 323                   {
 324                     if (*p == '\0')
 325                       return FNM_NOMATCH;
 326                     c = FOLD ((unsigned char) *p);
 327                     ++p;
 328
 329                     if (c == fn)
 330                       goto matched;
 331                   }
 332                 else if (c == '[' && *p == ':')
 333                   {
 334                     /* Leave room for the null.  */
 335                     char str[CHAR_CLASS_MAX_LENGTH + 1];
 336                     size_t c1 = 0;
 337 # if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H)
 338                     wctype_t wt;
 339 # endif
 340                     const char *startp = p;
 341
 342                     for (;;)
 343                       {
 344                         if (c1 == CHAR_CLASS_MAX_LENGTH)
 345                           /* The name is too long and therefore the pattern
 346                              is ill-formed.  */
 347                           return FNM_NOMATCH;
 348
 349                         c = *++p;
 350                         if (c == ':' && p[1] == ']')
 351                           {
 352                             p += 2;
 353                             break;
 354                           }
 355                         if (c < 'a' || c >= 'z')
 356                           {
 357                             /* This cannot possibly be a character class name.
 358                                Match it as a normal range.  */
 359                             p = startp;
 360                             c = '[';
 361                             goto normal_bracket;
 362                           }
 363                         str[c1++] = c;
 364                       }
 365                     str[c1] = '\0';
 366
 367 # if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H)
 368                     wt = IS_CHAR_CLASS (str);
 369                     if (wt == 0)
 370                       /* Invalid character class name.  */
 371                       return FNM_NOMATCH;
 372
 373                     if (__iswctype (__btowc ((unsigned char) *n), wt))
 374                       goto matched;
 375 # else
 376                     if ((STREQ (str, "alnum") && ISALNUM ((unsigned char) *n))
 377                         || (STREQ (str, "alpha") && ISALPHA ((unsigned char) *n))
 378                         || (STREQ (str, "blank") && ISBLANK ((unsigned char) *n))
 379                         || (STREQ (str, "cntrl") && ISCNTRL ((unsigned char) *n))
 380                         || (STREQ (str, "digit") && ISDIGIT ((unsigned char) *n))
 381                         || (STREQ (str, "graph") && ISGRAPH ((unsigned char) *n))
 382                         || (STREQ (str, "lower") && ISLOWER ((unsigned char) *n))
 383                         || (STREQ (str, "print") && ISPRINT ((unsigned char) *n))
 384                         || (STREQ (str, "punct") && ISPUNCT ((unsigned char) *n))
 385                         || (STREQ (str, "space") && ISSPACE ((unsigned char) *n))
 386                         || (STREQ (str, "upper") && ISUPPER ((unsigned char) *n))
 387                         || (STREQ (str, "xdigit") && ISXDIGIT ((unsigned char) *n)))
 388                       goto matched;
 389 # endif
 390                   }
 391                 else if (c == '\0')
 392                   /* [ (unterminated) loses.  */
 393                   return FNM_NOMATCH;
 394                 else
 395                   {
 396                   normal_bracket:
 397                     if (FOLD (c) == fn)
 398                       goto matched;
 399
 400                     cold = c;
 401                     c = *p++;
 402
 403                     if (c == '-' && *p != ']')
 404                       {
 405                         /* It is a range.  */
 406                         unsigned char cend = *p++;
 407                         if (!(flags & FNM_NOESCAPE) && cend == '\\')
 408                           cend = *p++;
 409                         if (cend == '\0')
 410                           return FNM_NOMATCH;
 411
 412                         if (cold <= fn && fn <= FOLD (cend))
 413                           goto matched;
 414
 415                         c = *p++;
 416                       }
 417                   }
 418
 419                 if (c == ']')
 420                   break;
 421               }
 422
 423             if (!not)
 424               return FNM_NOMATCH;
 425             break;
 426
 427           matched:
 428             /* Skip the rest of the [...] that already matched.  */
 429             while (c != ']')
 430               {
 431                 if (c == '\0')
 432                   /* [... (unterminated) loses.  */
 433                   return FNM_NOMATCH;
 434
 435                 c = *p++;
 436                 if (!(flags & FNM_NOESCAPE) && c == '\\')
 437                   {
 438                     if (*p == '\0')
 439                       return FNM_NOMATCH;
 440                     /* XXX 1003.2d11 is unclear if this is right.  */
 441                     ++p;
 442                   }
 443                 else if (c == '[' && *p == ':')
 444                   {
 445                     do
 446                       if (*++p == '\0')
 447                         return FNM_NOMATCH;
 448                     while (*p != ':' || p[1] == ']');
 449                     p += 2;
 450                     c = *p;
 451                   }
 452               }
 453             if (not)
 454               return FNM_NOMATCH;
 455           }
 456           break;
 457
 458         default:
 459           if (c != FOLD ((unsigned char) *n))
 460             return FNM_NOMATCH;
 461         }
 462
 463       ++n;
 464     }
 465
 466   if (*n == '\0')
 467     return 0;
 468
 469   if ((flags & FNM_LEADING_DIR) && *n == '/')
 470     /* The FNM_LEADING_DIR flag says that "foo*" matches "foobar/frobozz".  */
 471     return 0;
 472
 473   return FNM_NOMATCH;
 474
 475 # undef FOLD
 476 }
 477
 478
 479 int
 480 fnmatch (pattern, string, flags)
 481      const char *pattern;
 482      const char *string;
 483      int flags;
 484 {
 485   return internal_fnmatch (pattern, string, flags & FNM_PERIOD, flags);
 486 }
 487
 488 #endif  /* _LIBC or not __GNU_LIBRARY__.  */