lib/fnmatch.c

   1 /* Copyright (C) 1991-1993, 1996-2007, 2009-2017 Free Software Foundation, Inc.
   2
   3    This program is free software; you can redistribute it and/or modify
   4    it under the terms of the GNU General Public License as published by
   5    the Free Software Foundation; either version 2, or (at your option)
   6    any later version.
   7
   8    This program is distributed in the hope that it will be useful,
   9    but WITHOUT ANY WARRANTY; without even the implied warranty of
  10    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  11    GNU General Public License for more details.
  12
  13    You should have received a copy of the GNU General Public License
  14    along with this program; if not, see <http://www.gnu.org/licenses/>.  */
  15
  16 #ifndef _LIBC
  17 # include <config.h>
  18 #endif
  19
  20 /* Enable GNU extensions in fnmatch.h.  */
  21 #ifndef _GNU_SOURCE
  22 # define _GNU_SOURCE    1
  23 #endif
  24
  25 #include <fnmatch.h>
  26
  27 #include <alloca.h>
  28 #include <assert.h>
  29 #include <ctype.h>
  30 #include <errno.h>
  31 #include <stddef.h>
  32 #include <stdbool.h>
  33 #include <stdlib.h>
  34 #include <string.h>
  35
  36 #define WIDE_CHAR_SUPPORT \
  37   (HAVE_WCTYPE_H && HAVE_BTOWC && HAVE_ISWCTYPE \
  38    && HAVE_WMEMCHR && (HAVE_WMEMCPY || HAVE_WMEMPCPY))
  39
  40 /* For platform which support the ISO C amendment 1 functionality we
  41    support user defined character classes.  */
  42 #if defined _LIBC || WIDE_CHAR_SUPPORT
  43 # include <wctype.h>
  44 # include <wchar.h>
  45 #endif
  46
  47 /* We need some of the locale data (the collation sequence information)
  48    but there is no interface to get this information in general.  Therefore
  49    we support a correct implementation only in glibc.  */
  50 #ifdef _LIBC
  51 # include "../locale/localeinfo.h"
  52 # include "../locale/elem-hash.h"
  53 # include "../locale/coll-lookup.h"
  54 # include <shlib-compat.h>
  55
  56 # define CONCAT(a,b) __CONCAT(a,b)
  57 # define mbsrtowcs __mbsrtowcs
  58 # define fnmatch __fnmatch
  59 extern int fnmatch (const char *pattern, const char *string, int flags);
  60 #endif
  61
  62 #ifndef SIZE_MAX
  63 # define SIZE_MAX ((size_t) -1)
  64 #endif
  65
  66 #include "flexmember.h"
  67
  68 /* We often have to test for FNM_FILE_NAME and FNM_PERIOD being both set.  */
  69 #define NO_LEADING_PERIOD(flags) \
  70   ((flags & (FNM_FILE_NAME | FNM_PERIOD)) == (FNM_FILE_NAME | FNM_PERIOD))
  71
  72 /* Comment out all this code if we are using the GNU C Library, and are not
  73    actually compiling the library itself, and have not detected a bug
  74    in the library.  This code is part of the GNU C
  75    Library, but also included in many other GNU distributions.  Compiling
  76    and linking in this code is a waste when using the GNU C library
  77    (especially if it is a shared library).  Rather than having every GNU
  78    program understand 'configure --with-gnu-libc' and omit the object files,
  79    it is simpler to just do this in the source for each such file.  */
  80
  81 #if defined _LIBC || !defined __GNU_LIBRARY__ || !HAVE_FNMATCH_GNU
  82
  83
  84 # if ! (defined isblank || (HAVE_ISBLANK && HAVE_DECL_ISBLANK))
  85 #  define isblank(c) ((c) == ' ' || (c) == '\t')
  86 # endif
  87
  88 # define STREQ(s1, s2) (strcmp (s1, s2) == 0)
  89
  90 # if defined _LIBC || WIDE_CHAR_SUPPORT
  91 /* The GNU C library provides support for user-defined character classes
  92    and the functions from ISO C amendment 1.  */
  93 #  ifdef CHARCLASS_NAME_MAX
  94 #   define CHAR_CLASS_MAX_LENGTH CHARCLASS_NAME_MAX
  95 #  else
  96 /* This shouldn't happen but some implementation might still have this
  97    problem.  Use a reasonable default value.  */
  98 #   define CHAR_CLASS_MAX_LENGTH 256
  99 #  endif
 100
 101 #  ifdef _LIBC
 102 #   define IS_CHAR_CLASS(string) __wctype (string)
 103 #  else
 104 #   define IS_CHAR_CLASS(string) wctype (string)
 105 #  endif
 106
 107 #  ifdef _LIBC
 108 #   define ISWCTYPE(WC, WT)     __iswctype (WC, WT)
 109 #  else
 110 #   define ISWCTYPE(WC, WT)     iswctype (WC, WT)
 111 #  endif
 112
 113 #  if (HAVE_MBSTATE_T && HAVE_MBSRTOWCS) || _LIBC
 114 /* In this case we are implementing the multibyte character handling.  */
 115 #   define HANDLE_MULTIBYTE     1
 116 #  endif
 117
 118 # else
 119 #  define CHAR_CLASS_MAX_LENGTH  6 /* Namely, 'xdigit'.  */
 120
 121 #  define IS_CHAR_CLASS(string)                                               \
 122    (STREQ (string, "alpha") || STREQ (string, "upper")                        \
 123     || STREQ (string, "lower") || STREQ (string, "digit")                     \
 124     || STREQ (string, "alnum") || STREQ (string, "xdigit")                    \
 125     || STREQ (string, "space") || STREQ (string, "print")                     \
 126     || STREQ (string, "punct") || STREQ (string, "graph")                     \
 127     || STREQ (string, "cntrl") || STREQ (string, "blank"))
 128 # endif
 129
 130 /* Avoid depending on library functions or files
 131    whose names are inconsistent.  */
 132
 133 /* Global variable.  */
 134 static int posixly_correct;
 135
 136 # ifndef internal_function
 137 /* Inside GNU libc we mark some function in a special way.  In other
 138    environments simply ignore the marking.  */
 139 #  define internal_function
 140 # endif
 141
 142 /* Note that this evaluates C many times.  */
 143 # define FOLD(c) ((flags & FNM_CASEFOLD) ? tolower (c) : (c))
 144 # define CHAR   char
 145 # define UCHAR  unsigned char
 146 # define INT    int
 147 # define FCT    internal_fnmatch
 148 # define EXT    ext_match
 149 # define END    end_pattern
 150 # define L_(CS) CS
 151 # ifdef _LIBC
 152 #  define BTOWC(C)      __btowc (C)
 153 # else
 154 #  define BTOWC(C)      btowc (C)
 155 # endif
 156 # define STRLEN(S) strlen (S)
 157 # define STRCAT(D, S) strcat (D, S)
 158 # ifdef _LIBC
 159 #  define MEMPCPY(D, S, N) __mempcpy (D, S, N)
 160 # else
 161 #  if HAVE_MEMPCPY
 162 #   define MEMPCPY(D, S, N) mempcpy (D, S, N)
 163 #  else
 164 #   define MEMPCPY(D, S, N) ((void *) ((char *) memcpy (D, S, N) + (N)))
 165 #  endif
 166 # endif
 167 # define MEMCHR(S, C, N) memchr (S, C, N)
 168 # include "fnmatch_loop.c"
 169
 170
 171 # if HANDLE_MULTIBYTE
 172 #  define FOLD(c) ((flags & FNM_CASEFOLD) ? towlower (c) : (c))
 173 #  define CHAR  wchar_t
 174 #  define UCHAR wint_t
 175 #  define INT   wint_t
 176 #  define FCT   internal_fnwmatch
 177 #  define EXT   ext_wmatch
 178 #  define END   end_wpattern
 179 #  define L_(CS)        L##CS
 180 #  define BTOWC(C)      (C)
 181 #  ifdef _LIBC
 182 #   define STRLEN(S) __wcslen (S)
 183 #   define STRCAT(D, S) __wcscat (D, S)
 184 #   define MEMPCPY(D, S, N) __wmempcpy (D, S, N)
 185 #  else
 186 #   define STRLEN(S) wcslen (S)
 187 #   define STRCAT(D, S) wcscat (D, S)
 188 #   if HAVE_WMEMPCPY
 189 #    define MEMPCPY(D, S, N) wmempcpy (D, S, N)
 190 #   else
 191 #    define MEMPCPY(D, S, N) (wmemcpy (D, S, N) + (N))
 192 #   endif
 193 #  endif
 194 #  define MEMCHR(S, C, N) wmemchr (S, C, N)
 195 #  define WIDE_CHAR_VERSION 1
 196
 197 #  undef IS_CHAR_CLASS
 198 /* We have to convert the wide character string in a multibyte string.  But
 199    we know that the character class names consist of alphanumeric characters
 200    from the portable character set, and since the wide character encoding
 201    for a member of the portable character set is the same code point as
 202    its single-byte encoding, we can use a simplified method to convert the
 203    string to a multibyte character string.  */
 204 static wctype_t
 205 is_char_class (const wchar_t *wcs)
 206 {
 207   char s[CHAR_CLASS_MAX_LENGTH + 1];
 208   char *cp = s;
 209
 210   do
 211     {
 212       /* Test for a printable character from the portable character set.  */
 213 #  ifdef _LIBC
 214       if (*wcs < 0x20 || *wcs > 0x7e
 215           || *wcs == 0x24 || *wcs == 0x40 || *wcs == 0x60)
 216         return (wctype_t) 0;
 217 #  else
 218       switch (*wcs)
 219         {
 220         case L' ': case L'!': case L'"': case L'#': case L'%':
 221         case L'&': case L'\'': case L'(': case L')': case L'*':
 222         case L'+': case L',': case L'-': case L'.': case L'/':
 223         case L'0': case L'1': case L'2': case L'3': case L'4':
 224         case L'5': case L'6': case L'7': case L'8': case L'9':
 225         case L':': case L';': case L'<': case L'=': case L'>':
 226         case L'?':
 227         case L'A': case L'B': case L'C': case L'D': case L'E':
 228         case L'F': case L'G': case L'H': case L'I': case L'J':
 229         case L'K': case L'L': case L'M': case L'N': case L'O':
 230         case L'P': case L'Q': case L'R': case L'S': case L'T':
 231         case L'U': case L'V': case L'W': case L'X': case L'Y':
 232         case L'Z':
 233         case L'[': case L'\\': case L']': case L'^': case L'_':
 234         case L'a': case L'b': case L'c': case L'd': case L'e':
 235         case L'f': case L'g': case L'h': case L'i': case L'j':
 236         case L'k': case L'l': case L'm': case L'n': case L'o':
 237         case L'p': case L'q': case L'r': case L's': case L't':
 238         case L'u': case L'v': case L'w': case L'x': case L'y':
 239         case L'z': case L'{': case L'|': case L'}': case L'~':
 240           break;
 241         default:
 242           return (wctype_t) 0;
 243         }
 244 #  endif
 245
 246       /* Avoid overrunning the buffer.  */
 247       if (cp == s + CHAR_CLASS_MAX_LENGTH)
 248         return (wctype_t) 0;
 249
 250       *cp++ = (char) *wcs++;
 251     }
 252   while (*wcs != L'\0');
 253
 254   *cp = '\0';
 255
 256 #  ifdef _LIBC
 257   return __wctype (s);
 258 #  else
 259   return wctype (s);
 260 #  endif
 261 }
 262 #  define IS_CHAR_CLASS(string) is_char_class (string)
 263
 264 #  include "fnmatch_loop.c"
 265 # endif
 266
 267
 268 int
 269 fnmatch (const char *pattern, const char *string, int flags)
 270 {
 271 # if HANDLE_MULTIBYTE
 272 #  define ALLOCA_LIMIT 2000
 273   if (__builtin_expect (MB_CUR_MAX, 1) != 1)
 274     {
 275       mbstate_t ps;
 276       size_t patsize;
 277       size_t strsize;
 278       size_t totsize;
 279       wchar_t *wpattern;
 280       wchar_t *wstring;
 281       int res;
 282
 283       /* Calculate the size needed to convert the strings to
 284          wide characters.  */
 285       memset (&ps, '\0', sizeof (ps));
 286       patsize = mbsrtowcs (NULL, &pattern, 0, &ps) + 1;
 287       if (__builtin_expect (patsize != 0, 1))
 288         {
 289           assert (mbsinit (&ps));
 290           strsize = mbsrtowcs (NULL, &string, 0, &ps) + 1;
 291           if (__builtin_expect (strsize != 0, 1))
 292             {
 293               assert (mbsinit (&ps));
 294               totsize = patsize + strsize;
 295               if (__builtin_expect (! (patsize <= totsize
 296                                        && totsize <= SIZE_MAX / sizeof (wchar_t)),
 297                                     0))
 298                 {
 299                   errno = ENOMEM;
 300                   return -1;
 301                 }
 302
 303               /* Allocate room for the wide characters.  */
 304               if (__builtin_expect (totsize < ALLOCA_LIMIT, 1))
 305                 wpattern = (wchar_t *) alloca (totsize * sizeof (wchar_t));
 306               else
 307                 {
 308                   wpattern = malloc (totsize * sizeof (wchar_t));
 309                   if (__builtin_expect (! wpattern, 0))
 310                     {
 311                       errno = ENOMEM;
 312                       return -1;
 313                     }
 314                 }
 315               wstring = wpattern + patsize;
 316
 317               /* Convert the strings into wide characters.  */
 318               mbsrtowcs (wpattern, &pattern, patsize, &ps);
 319               assert (mbsinit (&ps));
 320               mbsrtowcs (wstring, &string, strsize, &ps);
 321
 322               res = internal_fnwmatch (wpattern, wstring, wstring + strsize - 1,
 323                                        flags & FNM_PERIOD, flags);
 324
 325               if (__builtin_expect (! (totsize < ALLOCA_LIMIT), 0))
 326                 free (wpattern);
 327               return res;
 328             }
 329         }
 330     }
 331
 332 # endif /* HANDLE_MULTIBYTE */
 333
 334   return internal_fnmatch (pattern, string, string + strlen (string),
 335                            flags & FNM_PERIOD, flags);
 336 }
 337
 338 # ifdef _LIBC
 339 #  undef fnmatch
 340 versioned_symbol (libc, __fnmatch, fnmatch, GLIBC_2_2_3);
 341 #  if SHLIB_COMPAT(libc, GLIBC_2_0, GLIBC_2_2_3)
 342 strong_alias (__fnmatch, __fnmatch_old)
 343 compat_symbol (libc, __fnmatch_old, fnmatch, GLIBC_2_0);
 344 #  endif
 345 libc_hidden_ver (__fnmatch, fnmatch)
 346 # endif
 347
 348 #endif  /* _LIBC or not __GNU_LIBRARY__.  */