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 # ifndef NULL
 131 #  define NULL 0
 132 # endif
 133
 134 /* This function doesn't exist on most systems.  */
 135
 136 # if !defined HAVE___STRCHRNUL && !defined _LIBC
 137 static char *
 138 __strchrnul (const char *s, int c)
 139
 140
 141 {
 142   char *result = strchr (s, c);
 143   if (result == NULL)
 144     result = strchr (s, '\0');
 145   return result;
 146 }
 147 # endif
 148
 149 # ifndef internal_function
 150 /* Inside GNU libc we mark some function in a special way.  In other
 151    environments simply ignore the marking.  */
 152 #  define internal_function
 153 # endif
 154
 155 /* Match STRING against the filename pattern PATTERN, returning zero if
 156    it matches, nonzero if not.  */
 157 static int internal_fnmatch __P ((const char *pattern, const char *string,
 158                                   int no_leading_period, int flags))
 159      internal_function;
 160 static int
 161 internal_function
 162 internal_fnmatch (const char *pattern, const char *string, int no_leading_period, int flags)
 163
 164
 165
 166
 167 {
 168   register const char *p = pattern, *n = string;
 169   register unsigned char c;
 170
 171 /* Note that this evaluates C many times.  */
 172 # ifdef _LIBC
 173 #  define FOLD(c) ((flags & FNM_CASEFOLD) ? tolower (c) : (c))
 174 # else
 175 #  define FOLD(c) ((flags & FNM_CASEFOLD) && ISUPPER (c) ? tolower (c) : (c))
 176 # endif
 177
 178   while ((c = *p++) != '\0')
 179     {
 180       c = FOLD (c);
 181
 182       switch (c)
 183         {
 184         case '?':
 185           if (*n == '\0')
 186             return FNM_NOMATCH;
 187           else if (*n == '/' && (flags & FNM_FILE_NAME))
 188             return FNM_NOMATCH;
 189           else if (*n == '.' && no_leading_period
 190                    && (n == string
 191                        || (n[-1] == '/' && (flags & FNM_FILE_NAME))))
 192             return FNM_NOMATCH;
 193           break;
 194
 195         case '\\':
 196           if (!(flags & FNM_NOESCAPE))
 197             {
 198               c = *p++;
 199               if (c == '\0')
 200                 /* Trailing \ loses.  */
 201                 return FNM_NOMATCH;
 202               c = FOLD (c);
 203             }
 204           if (FOLD ((unsigned char) *n) != c)
 205             return FNM_NOMATCH;
 206           break;
 207
 208         case '*':
 209           if (*n == '.' && no_leading_period
 210               && (n == string
 211                   || (n[-1] == '/' && (flags & FNM_FILE_NAME))))
 212             return FNM_NOMATCH;
 213
 214           for (c = *p++; c == '?' || c == '*'; c = *p++)
 215             {
 216               if (*n == '/' && (flags & FNM_FILE_NAME))
 217                 /* A slash does not match a wildcard under FNM_FILE_NAME.  */
 218                 return FNM_NOMATCH;
 219               else if (c == '?')
 220                 {
 221                   /* A ? needs to match one character.  */
 222                   if (*n == '\0')
 223                     /* There isn't another character; no match.  */
 224                     return FNM_NOMATCH;
 225                   else
 226                     /* One character of the string is consumed in matching
 227                        this ? wildcard, so *??? won't match if there are
 228                        less than three characters.  */
 229                     ++n;
 230                 }
 231             }
 232
 233           if (c == '\0')
 234             /* The wildcard(s) is/are the last element of the pattern.
 235                If the name is a file name and contains another slash
 236                this does mean it cannot match.  */
 237             return ((flags & FNM_FILE_NAME) && strchr (n, '/') != NULL
 238                     ? FNM_NOMATCH : 0);
 239           else
 240             {
 241               const char *endp;
 242
 243               endp = __strchrnul (n, (flags & FNM_FILE_NAME) ? '/' : '\0');
 244
 245               if (c == '[')
 246                 {
 247                   int flags2 = ((flags & FNM_FILE_NAME)
 248                                 ? flags : (flags & ~FNM_PERIOD));
 249
 250                   for (--p; n < endp; ++n)
 251                     if (internal_fnmatch (p, n,
 252                                           (no_leading_period
 253                                            && (n == string
 254                                                || (n[-1] == '/'
 255                                                    && (flags
 256                                                        & FNM_FILE_NAME)))),
 257                                           flags2)
 258                         == 0)
 259                       return 0;
 260                 }
 261               else if (c == '/' && (flags & FNM_FILE_NAME))
 262                 {
 263                   while (*n != '\0' && *n != '/')
 264                     ++n;
 265                   if (*n == '/'
 266                       && (internal_fnmatch (p, n + 1, flags & FNM_PERIOD,
 267                                             flags) == 0))
 268                     return 0;
 269                 }
 270               else
 271                 {
 272                   int flags2 = ((flags & FNM_FILE_NAME)
 273                                 ? flags : (flags & ~FNM_PERIOD));
 274
 275                   if (c == '\\' && !(flags & FNM_NOESCAPE))
 276                     c = *p;
 277                   c = FOLD (c);
 278                   for (--p; n < endp; ++n)
 279                     if (FOLD ((unsigned char) *n) == c
 280                         && (internal_fnmatch (p, n,
 281                                               (no_leading_period
 282                                                && (n == string
 283                                                    || (n[-1] == '/'
 284                                                        && (flags
 285                                                            & FNM_FILE_NAME)))),
 286                                               flags2) == 0))
 287                       return 0;
 288                 }
 289             }
 290
 291           /* If we come here no match is possible with the wildcard.  */
 292           return FNM_NOMATCH;
 293
 294         case '[':
 295           {
 296             /* Nonzero if the sense of the character class is inverted.  */
 297             static int posixly_correct;
 298             register int not;
 299             char cold;
 300
 301             if (posixly_correct == 0)
 302               posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1;
 303
 304             if (*n == '\0')
 305               return FNM_NOMATCH;
 306
 307             if (*n == '.' && no_leading_period && (n == string
 308                                                    || (n[-1] == '/'
 309                                                        && (flags
 310                                                            & FNM_FILE_NAME))))
 311               return FNM_NOMATCH;
 312
 313             if (*n == '/' && (flags & FNM_FILE_NAME))
 314               /* `/' cannot be matched.  */
 315               return FNM_NOMATCH;
 316
 317             not = (*p == '!' || (posixly_correct < 0 && *p == '^'));
 318             if (not)
 319               ++p;
 320
 321             c = *p++;
 322             for (;;)
 323               {
 324                 unsigned char fn = FOLD ((unsigned char) *n);
 325
 326                 if (!(flags & FNM_NOESCAPE) && c == '\\')
 327                   {
 328                     if (*p == '\0')
 329                       return FNM_NOMATCH;
 330                     c = FOLD ((unsigned char) *p);
 331                     ++p;
 332
 333                     if (c == fn)
 334                       goto matched;
 335                   }
 336                 else if (c == '[' && *p == ':')
 337                   {
 338                     /* Leave room for the null.  */
 339                     char str[CHAR_CLASS_MAX_LENGTH + 1];
 340                     size_t c1 = 0;
 341 # if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H)
 342                     wctype_t wt;
 343 # endif
 344                     const char *startp = p;
 345
 346                     for (;;)
 347                       {
 348                         if (c1 > CHAR_CLASS_MAX_LENGTH)
 349                           /* The name is too long and therefore the pattern
 350                              is ill-formed.  */
 351                           return FNM_NOMATCH;
 352
 353                         c = *++p;
 354                         if (c == ':' && p[1] == ']')
 355                           {
 356                             p += 2;
 357                             break;
 358                           }
 359                         if (c < 'a' || c >= 'z')
 360                           {
 361                             /* This cannot possibly be a character class name.
 362                                Match it as a normal range.  */
 363                             p = startp;
 364                             c = '[';
 365                             goto normal_bracket;
 366                           }
 367                         str[c1++] = c;
 368                       }
 369                     str[c1] = '\0';
 370
 371 # if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H)
 372                     wt = IS_CHAR_CLASS (str);
 373                     if (wt == 0)
 374                       /* Invalid character class name.  */
 375                       return FNM_NOMATCH;
 376
 377                     if (__iswctype (__btowc ((unsigned char) *n), wt))
 378                       goto matched;
 379 # else
 380                     if ((STREQ (str, "alnum") && ISALNUM ((unsigned char) *n))
 381                         || (STREQ (str, "alpha") && ISALPHA ((unsigned char) *n))
 382                         || (STREQ (str, "blank") && ISBLANK ((unsigned char) *n))
 383                         || (STREQ (str, "cntrl") && ISCNTRL ((unsigned char) *n))
 384                         || (STREQ (str, "digit") && ISDIGIT ((unsigned char) *n))
 385                         || (STREQ (str, "graph") && ISGRAPH ((unsigned char) *n))
 386                         || (STREQ (str, "lower") && ISLOWER ((unsigned char) *n))
 387                         || (STREQ (str, "print") && ISPRINT ((unsigned char) *n))
 388                         || (STREQ (str, "punct") && ISPUNCT ((unsigned char) *n))
 389                         || (STREQ (str, "space") && ISSPACE ((unsigned char) *n))
 390                         || (STREQ (str, "upper") && ISUPPER ((unsigned char) *n))
 391                         || (STREQ (str, "xdigit") && ISXDIGIT ((unsigned char) *n)))
 392                       goto matched;
 393 # endif
 394                   }
 395                 else if (c == '\0')
 396                   /* [ (unterminated) loses.  */
 397                   return FNM_NOMATCH;
 398                 else
 399                   {
 400                   normal_bracket:
 401                     if (FOLD (c) == fn)
 402                       goto matched;
 403
 404                     cold = c;
 405                     c = *p++;
 406
 407                     if (c == '-' && *p != ']')
 408                       {
 409                         /* It is a range.  */
 410                         unsigned char cend = *p++;
 411                         if (!(flags & FNM_NOESCAPE) && cend == '\\')
 412                           cend = *p++;
 413                         if (cend == '\0')
 414                           return FNM_NOMATCH;
 415
 416                         if (cold <= fn && fn <= FOLD (cend))
 417                           goto matched;
 418
 419                         c = *p++;
 420                       }
 421                   }
 422
 423                 if (c == ']')
 424                   break;
 425               }
 426
 427             if (!not)
 428               return FNM_NOMATCH;
 429             break;
 430
 431           matched:
 432             /* Skip the rest of the [...] that already matched.  */
 433             while (c != ']')
 434               {
 435                 if (c == '\0')
 436                   /* [... (unterminated) loses.  */
 437                   return FNM_NOMATCH;
 438
 439                 c = *p++;
 440                 if (!(flags & FNM_NOESCAPE) && c == '\\')
 441                   {
 442                     if (*p == '\0')
 443                       return FNM_NOMATCH;
 444                     /* XXX 1003.2d11 is unclear if this is right.  */
 445                     ++p;
 446                   }
 447                 else if (c == '[' && *p == ':')
 448                   {
 449                     do
 450                       if (*++p == '\0')
 451                         return FNM_NOMATCH;
 452                     while (*p != ':' || p[1] == ']');
 453                     p += 2;
 454                     c = *p;
 455                   }
 456               }
 457             if (not)
 458               return FNM_NOMATCH;
 459           }
 460           break;
 461
 462         default:
 463           if (c != FOLD ((unsigned char) *n))
 464             return FNM_NOMATCH;
 465         }
 466
 467       ++n;
 468     }
 469
 470   if (*n == '\0')
 471     return 0;
 472
 473   if ((flags & FNM_LEADING_DIR) && *n == '/')
 474     /* The FNM_LEADING_DIR flag says that "foo*" matches "foobar/frobozz".  */
 475     return 0;
 476
 477   return FNM_NOMATCH;
 478
 479 # undef FOLD
 480 }
 481
 482
 483 int
 484 fnmatch (const char *pattern, const char *string, int flags)
 485
 486
 487
 488 {
 489   return internal_fnmatch (pattern, string, flags & FNM_PERIOD, flags);
 490 }
 491
 492 #endif  /* _LIBC or not __GNU_LIBRARY__.  */