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 <stddef.h>
  29 #include <errno.h>
  30 #include <fnmatch.h>
  31 #include <ctype.h>
  32
  33 #if HAVE_STRING_H || defined _LIBC
  34 # include <string.h>
  35 #else
  36 # include <strings.h>
  37 #endif
  38
  39 #if defined STDC_HEADERS || defined _LIBC
  40 # include <stdlib.h>
  41 #endif
  42
  43 /* For platforms which support the ISO C amendment 1 functionality we
  44    support user defined character classes.  */
  45 #if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H)
  46 /* Solaris 2.5 has a bug: <wchar.h> must be included before <wctype.h>.  */
  47 # include <wchar.h>
  48 # include <wctype.h>
  49 #endif
  50
  51 /* Comment out all this code if we are using the GNU C Library, and are not
  52    actually compiling the library itself.  This code is part of the GNU C
  53    Library, but also included in many other GNU distributions.  Compiling
  54    and linking in this code is a waste when using the GNU C library
  55    (especially if it is a shared library).  Rather than having every GNU
  56    program understand `configure --with-gnu-libc' and omit the object files,
  57    it is simpler to just do this in the source for each such file.  */
  58
  59 #if defined NO_FNMATCH || defined NO_FNMATCH_CASEFOLD || \
  60     defined _LIBC || !defined __GNU_LIBRARY__
  61
  62
  63 # if defined STDC_HEADERS || !defined isascii
  64 #  define ISASCII(c) 1
  65 # else
  66 #  define ISASCII(c) isascii(c)
  67 # endif
  68
  69 # ifdef isblank
  70 #  define ISBLANK(c) (ISASCII (c) && isblank (c))
  71 # else
  72 #  define ISBLANK(c) ((c) == ' ' || (c) == '\t')
  73 # endif
  74 # ifdef isgraph
  75 #  define ISGRAPH(c) (ISASCII (c) && isgraph (c))
  76 # else
  77 #  define ISGRAPH(c) (ISASCII (c) && isprint (c) && !isspace (c))
  78 # endif
  79
  80 # define ISPRINT(c) (ISASCII (c) && isprint (c))
  81 # define ISDIGIT(c) (ISASCII (c) && isdigit (c))
  82 # define ISALNUM(c) (ISASCII (c) && isalnum (c))
  83 # define ISALPHA(c) (ISASCII (c) && isalpha (c))
  84 # define ISCNTRL(c) (ISASCII (c) && iscntrl (c))
  85 # define ISLOWER(c) (ISASCII (c) && islower (c))
  86 # define ISPUNCT(c) (ISASCII (c) && ispunct (c))
  87 # define ISSPACE(c) (ISASCII (c) && isspace (c))
  88 # define ISUPPER(c) (ISASCII (c) && isupper (c))
  89 # define ISXDIGIT(c) (ISASCII (c) && isxdigit (c))
  90
  91 # define STREQ(s1, s2) ((strcmp (s1, s2) == 0))
  92
  93 # if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H)
  94 /* The GNU C library provides support for user-defined character classes
  95    and the functions from ISO C amendment 1.  */
  96 #  ifdef CHARCLASS_NAME_MAX
  97 #   define CHAR_CLASS_MAX_LENGTH CHARCLASS_NAME_MAX
  98 #  else
  99 /* This shouldn't happen but some implementation might still have this
 100    problem.  Use a reasonable default value.  */
 101 #   define CHAR_CLASS_MAX_LENGTH 256
 102 #  endif
 103
 104 #  ifdef _LIBC
 105 #   define IS_CHAR_CLASS(string) __wctype (string)
 106 #  else
 107 #   define IS_CHAR_CLASS(string) wctype (string)
 108 #  endif
 109 # else
 110 #  define CHAR_CLASS_MAX_LENGTH  6 /* Namely, `xdigit'.  */
 111
 112 #  define IS_CHAR_CLASS(string)                                               \
 113    (STREQ (string, "alpha") || STREQ (string, "upper")                        \
 114     || STREQ (string, "lower") || STREQ (string, "digit")                     \
 115     || STREQ (string, "alnum") || STREQ (string, "xdigit")                    \
 116     || STREQ (string, "space") || STREQ (string, "print")                     \
 117     || STREQ (string, "punct") || STREQ (string, "graph")                     \
 118     || STREQ (string, "cntrl") || STREQ (string, "blank"))
 119 # endif
 120
 121 /* Avoid depending on library functions or files
 122    whose names are inconsistent.  */
 123
 124 # if !defined _LIBC && !defined getenv
 125 extern char *getenv (const char *name);
 126 # endif
 127
 128 # ifndef errno
 129 extern int errno;
 130 # endif
 131
 132 # ifndef NULL
 133 #  define NULL 0
 134 # endif
 135
 136 /* This function doesn't exist on most systems.  */
 137
 138 # if !defined HAVE___STRCHRNUL && !defined _LIBC
 139 static char *
 140 __strchrnul (const char *s, int c)
 141
 142
 143 {
 144   char *result = strchr (s, c);
 145   if (result == NULL)
 146     result = strchr (s, '\0');
 147   return result;
 148 }
 149 # endif
 150
 151 # ifndef internal_function
 152 /* Inside GNU libc we mark some function in a special way.  In other
 153    environments simply ignore the marking.  */
 154 #  define internal_function
 155 # endif
 156
 157 /* Match STRING against the filename pattern PATTERN, returning zero if
 158    it matches, nonzero if not.  */
 159 static int internal_fnmatch __P ((const char *pattern, const char *string,
 160                                   int no_leading_period, int flags))
 161      internal_function;
 162 static int
 163 internal_function
 164 internal_fnmatch (const char *pattern, const char *string, int no_leading_period, int flags)
 165
 166
 167
 168
 169 {
 170   register const char *p = pattern, *n = string;
 171   register unsigned char c;
 172
 173 /* Note that this evaluates C many times.  */
 174 # ifdef _LIBC
 175 #  define FOLD(c) ((flags & FNM_CASEFOLD) ? tolower (c) : (c))
 176 # else
 177 #  define FOLD(c) ((flags & FNM_CASEFOLD) && ISUPPER (c) ? tolower (c) : (c))
 178 # endif
 179
 180   while ((c = *p++) != '\0')
 181     {
 182       c = FOLD (c);
 183
 184       switch (c)
 185         {
 186         case '?':
 187           if (*n == '\0')
 188             return FNM_NOMATCH;
 189           else if (*n == '/' && (flags & FNM_FILE_NAME))
 190             return FNM_NOMATCH;
 191           else if (*n == '.' && no_leading_period
 192                    && (n == string
 193                        || (n[-1] == '/' && (flags & FNM_FILE_NAME))))
 194             return FNM_NOMATCH;
 195           break;
 196
 197         case '\\':
 198           if (!(flags & FNM_NOESCAPE))
 199             {
 200               c = *p++;
 201               if (c == '\0')
 202                 /* Trailing \ loses.  */
 203                 return FNM_NOMATCH;
 204               c = FOLD (c);
 205             }
 206           if (FOLD ((unsigned char) *n) != c)
 207             return FNM_NOMATCH;
 208           break;
 209
 210         case '*':
 211           if (*n == '.' && no_leading_period
 212               && (n == string
 213                   || (n[-1] == '/' && (flags & FNM_FILE_NAME))))
 214             return FNM_NOMATCH;
 215
 216           for (c = *p++; c == '?' || c == '*'; c = *p++)
 217             {
 218               if (*n == '/' && (flags & FNM_FILE_NAME))
 219                 /* A slash does not match a wildcard under FNM_FILE_NAME.  */
 220                 return FNM_NOMATCH;
 221               else if (c == '?')
 222                 {
 223                   /* A ? needs to match one character.  */
 224                   if (*n == '\0')
 225                     /* There isn't another character; no match.  */
 226                     return FNM_NOMATCH;
 227                   else
 228                     /* One character of the string is consumed in matching
 229                        this ? wildcard, so *??? won't match if there are
 230                        less than three characters.  */
 231                     ++n;
 232                 }
 233             }
 234
 235           if (c == '\0')
 236             /* The wildcard(s) is/are the last element of the pattern.
 237                If the name is a file name and contains another slash
 238                this does mean it cannot match.  */
 239             return ((flags & FNM_FILE_NAME) && strchr (n, '/') != NULL
 240                     ? FNM_NOMATCH : 0);
 241           else
 242             {
 243               const char *endp;
 244
 245               endp = __strchrnul (n, (flags & FNM_FILE_NAME) ? '/' : '\0');
 246
 247               if (c == '[')
 248                 {
 249                   int flags2 = ((flags & FNM_FILE_NAME)
 250                                 ? flags : (flags & ~FNM_PERIOD));
 251
 252                   for (--p; n < endp; ++n)
 253                     if (internal_fnmatch (p, n,
 254                                           (no_leading_period
 255                                            && (n == string
 256                                                || (n[-1] == '/'
 257                                                    && (flags
 258                                                        & FNM_FILE_NAME)))),
 259                                           flags2)
 260                         == 0)
 261                       return 0;
 262                 }
 263               else if (c == '/' && (flags & FNM_FILE_NAME))
 264                 {
 265                   while (*n != '\0' && *n != '/')
 266                     ++n;
 267                   if (*n == '/'
 268                       && (internal_fnmatch (p, n + 1, flags & FNM_PERIOD,
 269                                             flags) == 0))
 270                     return 0;
 271                 }
 272               else
 273                 {
 274                   int flags2 = ((flags & FNM_FILE_NAME)
 275                                 ? flags : (flags & ~FNM_PERIOD));
 276
 277                   if (c == '\\' && !(flags & FNM_NOESCAPE))
 278                     c = *p;
 279                   c = FOLD (c);
 280                   for (--p; n < endp; ++n)
 281                     if (FOLD ((unsigned char) *n) == c
 282                         && (internal_fnmatch (p, n,
 283                                               (no_leading_period
 284                                                && (n == string
 285                                                    || (n[-1] == '/'
 286                                                        && (flags
 287                                                            & FNM_FILE_NAME)))),
 288                                               flags2) == 0))
 289                       return 0;
 290                 }
 291             }
 292
 293           /* If we come here no match is possible with the wildcard.  */
 294           return FNM_NOMATCH;
 295
 296         case '[':
 297           {
 298             /* Nonzero if the sense of the character class is inverted.  */
 299             static int posixly_correct;
 300             register int not;
 301             char cold;
 302
 303             if (posixly_correct == 0)
 304               posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1;
 305
 306             if (*n == '\0')
 307               return FNM_NOMATCH;
 308
 309             if (*n == '.' && no_leading_period && (n == string
 310                                                    || (n[-1] == '/'
 311                                                        && (flags
 312                                                            & FNM_FILE_NAME))))
 313               return FNM_NOMATCH;
 314
 315             if (*n == '/' && (flags & FNM_FILE_NAME))
 316               /* `/' cannot be matched.  */
 317               return FNM_NOMATCH;
 318
 319             not = (*p == '!' || (posixly_correct < 0 && *p == '^'));
 320             if (not)
 321               ++p;
 322
 323             c = *p++;
 324             for (;;)
 325               {
 326                 unsigned char fn = FOLD ((unsigned char) *n);
 327
 328                 if (!(flags & FNM_NOESCAPE) && c == '\\')
 329                   {
 330                     if (*p == '\0')
 331                       return FNM_NOMATCH;
 332                     c = FOLD ((unsigned char) *p);
 333                     ++p;
 334
 335                     if (c == fn)
 336                       goto matched;
 337                   }
 338                 else if (c == '[' && *p == ':')
 339                   {
 340                     /* Leave room for the null.  */
 341                     char str[CHAR_CLASS_MAX_LENGTH + 1];
 342                     size_t c1 = 0;
 343 # if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H)
 344                     wctype_t wt;
 345 # endif
 346                     const char *startp = p;
 347
 348                     for (;;)
 349                       {
 350                         if (c1 > CHAR_CLASS_MAX_LENGTH)
 351                           /* The name is too long and therefore the pattern
 352                              is ill-formed.  */
 353                           return FNM_NOMATCH;
 354
 355                         c = *++p;
 356                         if (c == ':' && p[1] == ']')
 357                           {
 358                             p += 2;
 359                             break;
 360                           }
 361                         if (c < 'a' || c >= 'z')
 362                           {
 363                             /* This cannot possibly be a character class name.
 364                                Match it as a normal range.  */
 365                             p = startp;
 366                             c = '[';
 367                             goto normal_bracket;
 368                           }
 369                         str[c1++] = c;
 370                       }
 371                     str[c1] = '\0';
 372
 373 # if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H)
 374                     wt = IS_CHAR_CLASS (str);
 375                     if (wt == 0)
 376                       /* Invalid character class name.  */
 377                       return FNM_NOMATCH;
 378
 379                     if (__iswctype (__btowc ((unsigned char) *n), wt))
 380                       goto matched;
 381 # else
 382                     if ((STREQ (str, "alnum") && ISALNUM ((unsigned char) *n))
 383                         || (STREQ (str, "alpha") && ISALPHA ((unsigned char) *n))
 384                         || (STREQ (str, "blank") && ISBLANK ((unsigned char) *n))
 385                         || (STREQ (str, "cntrl") && ISCNTRL ((unsigned char) *n))
 386                         || (STREQ (str, "digit") && ISDIGIT ((unsigned char) *n))
 387                         || (STREQ (str, "graph") && ISGRAPH ((unsigned char) *n))
 388                         || (STREQ (str, "lower") && ISLOWER ((unsigned char) *n))
 389                         || (STREQ (str, "print") && ISPRINT ((unsigned char) *n))
 390                         || (STREQ (str, "punct") && ISPUNCT ((unsigned char) *n))
 391                         || (STREQ (str, "space") && ISSPACE ((unsigned char) *n))
 392                         || (STREQ (str, "upper") && ISUPPER ((unsigned char) *n))
 393                         || (STREQ (str, "xdigit") && ISXDIGIT ((unsigned char) *n)))
 394                       goto matched;
 395 # endif
 396                   }
 397                 else if (c == '\0')
 398                   /* [ (unterminated) loses.  */
 399                   return FNM_NOMATCH;
 400                 else
 401                   {
 402                   normal_bracket:
 403                     if (FOLD (c) == fn)
 404                       goto matched;
 405
 406                     cold = c;
 407                     c = *p++;
 408
 409                     if (c == '-' && *p != ']')
 410                       {
 411                         /* It is a range.  */
 412                         unsigned char cend = *p++;
 413                         if (!(flags & FNM_NOESCAPE) && cend == '\\')
 414                           cend = *p++;
 415                         if (cend == '\0')
 416                           return FNM_NOMATCH;
 417
 418                         if (cold <= fn && fn <= FOLD (cend))
 419                           goto matched;
 420
 421                         c = *p++;
 422                       }
 423                   }
 424
 425                 if (c == ']')
 426                   break;
 427               }
 428
 429             if (!not)
 430               return FNM_NOMATCH;
 431             break;
 432
 433           matched:
 434             /* Skip the rest of the [...] that already matched.  */
 435             while (c != ']')
 436               {
 437                 if (c == '\0')
 438                   /* [... (unterminated) loses.  */
 439                   return FNM_NOMATCH;
 440
 441                 c = *p++;
 442                 if (!(flags & FNM_NOESCAPE) && c == '\\')
 443                   {
 444                     if (*p == '\0')
 445                       return FNM_NOMATCH;
 446                     /* XXX 1003.2d11 is unclear if this is right.  */
 447                     ++p;
 448                   }
 449                 else if (c == '[' && *p == ':')
 450                   {
 451                     do
 452                       if (*++p == '\0')
 453                         return FNM_NOMATCH;
 454                     while (*p != ':' || p[1] == ']');
 455                     p += 2;
 456                     c = *p;
 457                   }
 458               }
 459             if (not)
 460               return FNM_NOMATCH;
 461           }
 462           break;
 463
 464         default:
 465           if (c != FOLD ((unsigned char) *n))
 466             return FNM_NOMATCH;
 467         }
 468
 469       ++n;
 470     }
 471
 472   if (*n == '\0')
 473     return 0;
 474
 475   if ((flags & FNM_LEADING_DIR) && *n == '/')
 476     /* The FNM_LEADING_DIR flag says that "foo*" matches "foobar/frobozz".  */
 477     return 0;
 478
 479   return FNM_NOMATCH;
 480
 481 # undef FOLD
 482 }
 483
 484
 485 int
 486 fnmatch (const char *pattern, const char *string, int flags)
 487
 488
 489
 490 {
 491   return internal_fnmatch (pattern, string, flags & FNM_PERIOD, flags);
 492 }
 493
 494 #endif  /* _LIBC or not __GNU_LIBRARY__.  */