1 /* Copyright (C) 1991-1993, 1996-1999, 2000 Free Software Foundation, Inc.
2 This file is part of the GNU C Library.
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.
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.
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. */
23 /* Enable GNU extensions in fnmatch.h. */
25 # define _GNU_SOURCE 1
33 #if HAVE_STRING_H || defined _LIBC
39 #if defined STDC_HEADERS || defined _LIBC
43 /* For platform which support the ISO C amendement 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>. */
51 /* We need some of the locale data (the collation sequence information)
52 but there is no interface to get this information in general. Therefore
53 we support a correct implementation only in glibc. */
55 # include "../locale/localeinfo.h"
56 # include "../locale/elem-hash.h"
57 # include "../locale/coll-lookup.h"
59 # define CONCAT(a,b) __CONCAT(a,b)
60 # define mbsinit __mbsinit
61 # define mbsrtowcs __mbsrtowcs
64 /* Comment out all this code if we are using the GNU C Library, and are not
65 actually compiling the library itself. This code is part of the GNU C
66 Library, but also included in many other GNU distributions. Compiling
67 and linking in this code is a waste when using the GNU C library
68 (especially if it is a shared library). Rather than having every GNU
69 program understand `configure --with-gnu-libc' and omit the object files,
70 it is simpler to just do this in the source for each such file. */
72 #if defined _LIBC || !defined __GNU_LIBRARY__
75 # if defined STDC_HEADERS || !defined isascii
78 # define ISASCII(c) isascii(c)
82 # define ISBLANK(c) (ISASCII (c) && isblank (c))
84 # define ISBLANK(c) ((c) == ' ' || (c) == '\t')
87 # define ISGRAPH(c) (ISASCII (c) && isgraph (c))
89 # define ISGRAPH(c) (ISASCII (c) && isprint (c) && !isspace (c))
92 # define ISPRINT(c) (ISASCII (c) && isprint (c))
93 # define ISDIGIT(c) (ISASCII (c) && isdigit (c))
94 # define ISALNUM(c) (ISASCII (c) && isalnum (c))
95 # define ISALPHA(c) (ISASCII (c) && isalpha (c))
96 # define ISCNTRL(c) (ISASCII (c) && iscntrl (c))
97 # define ISLOWER(c) (ISASCII (c) && islower (c))
98 # define ISPUNCT(c) (ISASCII (c) && ispunct (c))
99 # define ISSPACE(c) (ISASCII (c) && isspace (c))
100 # define ISUPPER(c) (ISASCII (c) && isupper (c))
101 # define ISXDIGIT(c) (ISASCII (c) && isxdigit (c))
103 # define STREQ(s1, s2) ((strcmp (s1, s2) == 0))
105 # if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H)
106 /* The GNU C library provides support for user-defined character classes
107 and the functions from ISO C amendement 1. */
108 # ifdef CHARCLASS_NAME_MAX
109 # define CHAR_CLASS_MAX_LENGTH CHARCLASS_NAME_MAX
111 /* This shouldn't happen but some implementation might still have this
112 problem. Use a reasonable default value. */
113 # define CHAR_CLASS_MAX_LENGTH 256
117 # define IS_CHAR_CLASS(string) __wctype (string)
119 # define IS_CHAR_CLASS(string) wctype (string)
123 # define ISWCTYPE(WC, WT) __iswctype (WC, WT)
125 # define ISWCTYPE(WC, WT) iswctype (WC, WT)
128 # if (HAVE_MBSTATE_T && HAVE_MBSRTOWCS) || _LIBC
129 /* In this case we are implementing the multibyte character handling. */
130 # define HANDLE_MULTIBYTE 1
134 # define CHAR_CLASS_MAX_LENGTH 6 /* Namely, `xdigit'. */
136 # define IS_CHAR_CLASS(string) \
137 (STREQ (string, "alpha") || STREQ (string, "upper") \
138 || STREQ (string, "lower") || STREQ (string, "digit") \
139 || STREQ (string, "alnum") || STREQ (string, "xdigit") \
140 || STREQ (string, "space") || STREQ (string, "print") \
141 || STREQ (string, "punct") || STREQ (string, "graph") \
142 || STREQ (string, "cntrl") || STREQ (string, "blank"))
145 /* Avoid depending on library functions or files
146 whose names are inconsistent. */
148 # if !defined _LIBC && !defined getenv
149 extern char *getenv ();
156 /* This function doesn't exist on most systems. */
158 # if !defined HAVE___STRCHRNUL && !defined _LIBC
164 char *result
= strchr (s
, c
);
166 result
= strchr (s
, '\0');
171 # if HANDLE_MULTIBYTE && !defined HAVE___STRCHRNUL && !defined _LIBC
177 wchar_t *result
= wcschr (s
, c
);
179 result
= wcschr (s
, '\0');
184 # ifndef internal_function
185 /* Inside GNU libc we mark some function in a special way. In other
186 environments simply ignore the marking. */
187 # define internal_function
190 /* Note that this evaluates C many times. */
192 # define FOLD(c) ((flags & FNM_CASEFOLD) ? tolower (c) : (c))
194 # define FOLD(c) ((flags & FNM_CASEFOLD) && ISUPPER (c) ? tolower (c) : (c))
197 # define UCHAR unsigned char
198 # define FCT internal_fnmatch
201 # define BTOWC(C) __btowc (C)
203 # define BTOWC(C) btowc (C)
205 # define STRCHR(S, C) strchr (S, C)
206 # define STRCHRNUL(S, C) __strchrnul (S, C)
207 # define STRCOLL(S1, S2) strcoll (S1, S2)
208 # include "fnmatch_loop.c"
211 # if HANDLE_MULTIBYTE
212 /* Note that this evaluates C many times. */
214 # define FOLD(c) ((flags & FNM_CASEFOLD) ? towlower (c) : (c))
216 # define FOLD(c) ((flags & FNM_CASEFOLD) && ISUPPER (c) ? towlower (c) : (c))
218 # define CHAR wchar_t
219 # define UCHAR wint_t
220 # define FCT internal_fnwmatch
222 # define BTOWC(C) (C)
223 # define STRCHR(S, C) wcschr (S, C)
224 # define STRCHRNUL(S, C) __wcschrnul (S, C)
225 # define STRCOLL(S1, S2) wcscoll (S1, S2)
226 # define WIDE_CHAR_VERSION 1
228 # undef IS_CHAR_CLASS
229 /* We have to convert the wide character string in a multibyte string. But
230 we know that the character class names consist of alphanumeric characters
231 from the portable character set, and since the wide character encoding
232 for a member of the portable character set is the same code point as
233 its single-byte encoding, we can use a simplified method to convert the
234 string to a multibyte character string. */
236 is_char_class (const wchar_t *wcs
)
238 char s
[CHAR_CLASS_MAX_LENGTH
+ 1];
243 /* Test for a printable character from the portable character set. */
245 if (*wcs
< 0x20 || *wcs
> 0x7e
246 || *wcs
== 0x24 || *wcs
== 0x40 || *wcs
== 0x60)
251 case L
' ': case L
'!': case L
'"': case L
'#': case L
'%':
252 case L
'&': case L
'\'': case L
'(': case L
')': case L
'*':
253 case L
'+': case L
',': case L
'-': case L
'.': case L
'/':
254 case L
'0': case L
'1': case L
'2': case L
'3': case L
'4':
255 case L
'5': case L
'6': case L
'7': case L
'8': case L
'9':
256 case L
':': case L
';': case L
'<': case L
'=': case L
'>':
258 case L
'A': case L
'B': case L
'C': case L
'D': case L
'E':
259 case L
'F': case L
'G': case L
'H': case L
'I': case L
'J':
260 case L
'K': case L
'L': case L
'M': case L
'N': case L
'O':
261 case L
'P': case L
'Q': case L
'R': case L
'S': case L
'T':
262 case L
'U': case L
'V': case L
'W': case L
'X': case L
'Y':
264 case L
'[': case L
'\\': case L
']': case L
'^': case L
'_':
265 case L
'a': case L
'b': case L
'c': case L
'd': case L
'e':
266 case L
'f': case L
'g': case L
'h': case L
'i': case L
'j':
267 case L
'k': case L
'l': case L
'm': case L
'n': case L
'o':
268 case L
'p': case L
'q': case L
'r': case L
's': case L
't':
269 case L
'u': case L
'v': case L
'w': case L
'x': case L
'y':
270 case L
'z': case L
'{': case L
'|': case L
'}': case L
'~':
277 /* Avoid overrunning the buffer. */
278 if (cp
== s
+ CHAR_CLASS_MAX_LENGTH
)
281 *cp
++ = (char) *wcs
++;
283 while (*wcs
!= L
'\0');
293 # define IS_CHAR_CLASS(string) is_char_class (string)
295 # include "fnmatch_loop.c"
300 fnmatch (pattern
, string
, flags
)
305 # if HANDLE_MULTIBYTE
312 /* This is an optimization for 8-bit character set. */
313 return internal_fnmatch (pattern
, string
, flags
& FNM_PERIOD
, flags
);
315 /* Convert the strings into wide characters. */
316 memset (&ps
, '\0', sizeof (ps
));
317 n
= mbsrtowcs (NULL
, &pattern
, 0, &ps
);
318 if (n
== (size_t) -1)
320 XXX Do we have to set `errno' to something which mbsrtows hasn't
323 wpattern
= (wchar_t *) alloca ((n
+ 1) * sizeof (wchar_t));
324 assert (mbsinit (&ps
));
325 (void) mbsrtowcs (wpattern
, &pattern
, n
+ 1, &ps
);
327 assert (mbsinit (&ps
));
328 n
= mbsrtowcs (NULL
, &string
, 0, &ps
);
329 if (n
== (size_t) -1)
331 XXX Do we have to set `errno' to something which mbsrtows hasn't
334 wstring
= (wchar_t *) alloca ((n
+ 1) * sizeof (wchar_t));
335 assert (mbsinit (&ps
));
336 (void) mbsrtowcs (wstring
, &string
, n
+ 1, &ps
);
338 return internal_fnwmatch (wpattern
, wstring
, flags
& FNM_PERIOD
, flags
);
340 return internal_fnmatch (pattern
, string
, flags
& FNM_PERIOD
, flags
);
341 # endif /* mbstate_t and mbsrtowcs or _LIBC. */
344 #endif /* _LIBC or not __GNU_LIBRARY__. */