posix/fnmatch.c

   1 /* Copyright (C) 1991, 92, 93, 96, 97, 98 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
  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 int
 133 fnmatch (pattern, string, flags)
 134      const char *pattern;
 135      const char *string;
 136      int flags;
 137 {
 138   register const char *p = pattern, *n = string;
 139   register unsigned char c;
 140
 141 /* Note that this evaluates C many times.  */
 142 # ifdef _LIBC
 143 #  define FOLD(c) ((flags & FNM_CASEFOLD) ? tolower (c) : (c))
 144 # else
 145 #  define FOLD(c) ((flags & FNM_CASEFOLD) && ISUPPER (c) ? tolower (c) : (c))
 146 # endif
 147
 148   while ((c = *p++) != '\0')
 149     {
 150       c = FOLD (c);
 151
 152       switch (c)
 153         {
 154         case '?':
 155           if (*n == '\0')
 156             return FNM_NOMATCH;
 157           else if ((flags & FNM_FILE_NAME) && *n == '/')
 158             return FNM_NOMATCH;
 159           else if ((flags & FNM_PERIOD) && *n == '.' &&
 160                    (n == string || ((flags & FNM_FILE_NAME) && n[-1] == '/')))
 161             return FNM_NOMATCH;
 162           break;
 163
 164         case '\\':
 165           if (!(flags & FNM_NOESCAPE))
 166             {
 167               c = *p++;
 168               if (c == '\0')
 169                 /* Trailing \ loses.  */
 170                 return FNM_NOMATCH;
 171               c = FOLD (c);
 172             }
 173           if (FOLD ((unsigned char) *n) != c)
 174             return FNM_NOMATCH;
 175           break;
 176
 177         case '*':
 178           if ((flags & FNM_PERIOD) && *n == '.' &&
 179               (n == string || ((flags & FNM_FILE_NAME) && n[-1] == '/')))
 180             return FNM_NOMATCH;
 181
 182           for (c = *p++; c == '?' || c == '*'; c = *p++)
 183             {
 184               if ((flags & FNM_FILE_NAME) && *n == '/')
 185                 /* A slash does not match a wildcard under FNM_FILE_NAME.  */
 186                 return FNM_NOMATCH;
 187               else if (c == '?')
 188                 {
 189                   /* A ? needs to match one character.  */
 190                   if (*n == '\0')
 191                     /* There isn't another character; no match.  */
 192                     return FNM_NOMATCH;
 193                   else
 194                     /* One character of the string is consumed in matching
 195                        this ? wildcard, so *??? won't match if there are
 196                        less than three characters.  */
 197                     ++n;
 198                 }
 199             }
 200
 201           if (c == '\0')
 202             return 0;
 203
 204           {
 205             unsigned char c1 = (!(flags & FNM_NOESCAPE) && c == '\\') ? *p : c;
 206             c1 = FOLD (c1);
 207             for (--p; *n != '\0'; ++n)
 208               if ((c == '[' || FOLD ((unsigned char) *n) == c1) &&
 209                   fnmatch (p, n, flags & ~FNM_PERIOD) == 0)
 210                 return 0;
 211             return FNM_NOMATCH;
 212           }
 213
 214         case '[':
 215           {
 216             /* Nonzero if the sense of the character class is inverted.  */
 217             static int posixly_correct;
 218             register int not;
 219             char cold;
 220
 221             if (posixly_correct == 0)
 222               posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1;
 223
 224             if (*n == '\0')
 225               return FNM_NOMATCH;
 226
 227             if (*n == '.' && (flags & FNM_PERIOD) &&
 228                 (n == string || ((flags & FNM_FILE_NAME) && n[-1] == '/')))
 229               return FNM_NOMATCH;
 230
 231             if (*n == '/' && (flags & FNM_FILE_NAME))
 232               /* `/' cannot be matched.  */
 233               return FNM_NOMATCH;
 234
 235             not = (*p == '!' || (posixly_correct < 0 && *p == '^'));
 236             if (not)
 237               ++p;
 238
 239             c = *p++;
 240             for (;;)
 241               {
 242                 unsigned char fn = FOLD ((unsigned char) *n);
 243
 244                 if (!(flags & FNM_NOESCAPE) && c == '\\')
 245                   {
 246                     if (*p == '\0')
 247                       return FNM_NOMATCH;
 248                     c = FOLD ((unsigned char) *p);
 249                     ++p;
 250
 251                     if (c == fn)
 252                       goto matched;
 253                   }
 254                 else if (c == '[' && *p == ':')
 255                   {
 256                     /* Leave room for the null.  */
 257                     char str[CHAR_CLASS_MAX_LENGTH + 1];
 258                     size_t c1 = 0;
 259 # if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H)
 260                     wctype_t wt;
 261 # endif
 262
 263                     for (;;)
 264                       {
 265                         if (c1 == CHAR_CLASS_MAX_LENGTH)
 266                           /* The name is too long and therefore the pattern
 267                              is ill-formed.  */
 268                           return FNM_NOMATCH;
 269
 270                         c = *++p;
 271                         if (c == ':' && p[1] == ']')
 272                           {
 273                             p += 2;
 274                             break;
 275                           }
 276                         str[c1++] = 'c';
 277                       }
 278                     str[c1] = '\0';
 279
 280 # if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H)
 281                     wt = IS_CHAR_CLASS (str);
 282                     if (wt == 0)
 283                       /* Invalid character class name.  */
 284                       return FNM_NOMATCH;
 285
 286                     if (__iswctype (__btowc ((unsigned char) *n), wt))
 287                       goto matched;
 288 # else
 289                     if ((STREQ (str, "alnum") && ISALNUM ((unsigned char) *n))
 290                         || (STREQ (str, "alpha") && ISALPHA ((unsigned char) *n))
 291                         || (STREQ (str, "blank") && ISBLANK ((unsigned char) *n))
 292                         || (STREQ (str, "cntrl") && ISCNTRL ((unsigned char) *n))
 293                         || (STREQ (str, "digit") && ISDIGIT ((unsigned char) *n))
 294                         || (STREQ (str, "graph") && ISGRAPH ((unsigned char) *n))
 295                         || (STREQ (str, "lower") && ISLOWER ((unsigned char) *n))
 296                         || (STREQ (str, "print") && ISPRINT ((unsigned char) *n))
 297                         || (STREQ (str, "punct") && ISPUNCT ((unsigned char) *n))
 298                         || (STREQ (str, "space") && ISSPACE ((unsigned char) *n))
 299                         || (STREQ (str, "upper") && ISUPPER ((unsigned char) *n))
 300                         || (STREQ (str, "xdigit") && ISXDIGIT ((unsigned char) *n)))
 301                       goto matched;
 302 # endif
 303                   }
 304                 else if (c == '\0')
 305                   /* [ (unterminated) loses.  */
 306                   return FNM_NOMATCH;
 307                 else if (FOLD (c) == fn)
 308                   goto matched;
 309
 310                 cold = c;
 311                 c = *p++;
 312
 313                 if (c == '-' && *p != ']')
 314                   {
 315                     /* It is a range.  */
 316                     unsigned char cend = *p++;
 317                     if (!(flags & FNM_NOESCAPE) && cend == '\\')
 318                       cend = *p++;
 319                     if (cend == '\0')
 320                       return FNM_NOMATCH;
 321
 322                     if (cold <= fn && fn <= FOLD (cend))
 323                       goto matched;
 324
 325                     c = *p++;
 326                   }
 327                 if (c == ']')
 328                   break;
 329               }
 330
 331             if (!not)
 332               return FNM_NOMATCH;
 333             break;
 334
 335           matched:
 336             /* Skip the rest of the [...] that already matched.  */
 337             while (c != ']')
 338               {
 339                 if (c == '\0')
 340                   /* [... (unterminated) loses.  */
 341                   return FNM_NOMATCH;
 342
 343                 c = *p++;
 344                 if (!(flags & FNM_NOESCAPE) && c == '\\')
 345                   {
 346                     if (*p == '\0')
 347                       return FNM_NOMATCH;
 348                     /* XXX 1003.2d11 is unclear if this is right.  */
 349                     ++p;
 350                   }
 351                 else if (c == '[' && *p == ':')
 352                   {
 353                     do
 354                       if (*++p == '\0')
 355                         return FNM_NOMATCH;
 356                     while (*p != ':' || p[1] == ']');
 357                     p += 2;
 358                     c = *p;
 359                   }
 360               }
 361             if (not)
 362               return FNM_NOMATCH;
 363           }
 364           break;
 365
 366         default:
 367           if (c != FOLD ((unsigned char) *n))
 368             return FNM_NOMATCH;
 369         }
 370
 371       ++n;
 372     }
 373
 374   if (*n == '\0')
 375     return 0;
 376
 377   if ((flags & FNM_LEADING_DIR) && *n == '/')
 378     /* The FNM_LEADING_DIR flag says that "foo*" matches "foobar/frobozz".  */
 379     return 0;
 380
 381   return FNM_NOMATCH;
 382
 383 # undef FOLD
 384 }
 385
 386 #endif  /* _LIBC or not __GNU_LIBRARY__.  */