Commit | Line | Data |
---|---|---|
c0c3707f | 1 | /* Copyright (C) 1991-1993, 1996-2007, 2009-2019 Free Software Foundation, Inc. |
8690e634 JK |
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 3, 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 | |
c0c3707f | 14 | along with this program; if not, see <https://www.gnu.org/licenses/>. */ |
8690e634 JK |
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 | ||
8690e634 JK |
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 | ||
49e4877c PA |
66 | #include "flexmember.h" |
67 | ||
c0c3707f CB |
68 | #ifndef FALLTHROUGH |
69 | # if __GNUC__ < 7 | |
70 | # define FALLTHROUGH ((void) 0) | |
71 | # else | |
72 | # define FALLTHROUGH __attribute__ ((__fallthrough__)) | |
73 | # endif | |
74 | #endif | |
75 | ||
8690e634 JK |
76 | /* We often have to test for FNM_FILE_NAME and FNM_PERIOD being both set. */ |
77 | #define NO_LEADING_PERIOD(flags) \ | |
78 | ((flags & (FNM_FILE_NAME | FNM_PERIOD)) == (FNM_FILE_NAME | FNM_PERIOD)) | |
79 | ||
80 | /* Comment out all this code if we are using the GNU C Library, and are not | |
81 | actually compiling the library itself, and have not detected a bug | |
82 | in the library. This code is part of the GNU C | |
83 | Library, but also included in many other GNU distributions. Compiling | |
84 | and linking in this code is a waste when using the GNU C library | |
85 | (especially if it is a shared library). Rather than having every GNU | |
86 | program understand 'configure --with-gnu-libc' and omit the object files, | |
87 | it is simpler to just do this in the source for each such file. */ | |
88 | ||
89 | #if defined _LIBC || !defined __GNU_LIBRARY__ || !HAVE_FNMATCH_GNU | |
90 | ||
91 | ||
92 | # if ! (defined isblank || (HAVE_ISBLANK && HAVE_DECL_ISBLANK)) | |
93 | # define isblank(c) ((c) == ' ' || (c) == '\t') | |
94 | # endif | |
95 | ||
96 | # define STREQ(s1, s2) (strcmp (s1, s2) == 0) | |
97 | ||
98 | # if defined _LIBC || WIDE_CHAR_SUPPORT | |
99 | /* The GNU C library provides support for user-defined character classes | |
100 | and the functions from ISO C amendment 1. */ | |
101 | # ifdef CHARCLASS_NAME_MAX | |
102 | # define CHAR_CLASS_MAX_LENGTH CHARCLASS_NAME_MAX | |
103 | # else | |
104 | /* This shouldn't happen but some implementation might still have this | |
105 | problem. Use a reasonable default value. */ | |
106 | # define CHAR_CLASS_MAX_LENGTH 256 | |
107 | # endif | |
108 | ||
109 | # ifdef _LIBC | |
110 | # define IS_CHAR_CLASS(string) __wctype (string) | |
111 | # else | |
112 | # define IS_CHAR_CLASS(string) wctype (string) | |
113 | # endif | |
114 | ||
115 | # ifdef _LIBC | |
116 | # define ISWCTYPE(WC, WT) __iswctype (WC, WT) | |
117 | # else | |
118 | # define ISWCTYPE(WC, WT) iswctype (WC, WT) | |
119 | # endif | |
120 | ||
121 | # if (HAVE_MBSTATE_T && HAVE_MBSRTOWCS) || _LIBC | |
122 | /* In this case we are implementing the multibyte character handling. */ | |
123 | # define HANDLE_MULTIBYTE 1 | |
124 | # endif | |
125 | ||
126 | # else | |
127 | # define CHAR_CLASS_MAX_LENGTH 6 /* Namely, 'xdigit'. */ | |
128 | ||
129 | # define IS_CHAR_CLASS(string) \ | |
130 | (STREQ (string, "alpha") || STREQ (string, "upper") \ | |
131 | || STREQ (string, "lower") || STREQ (string, "digit") \ | |
132 | || STREQ (string, "alnum") || STREQ (string, "xdigit") \ | |
133 | || STREQ (string, "space") || STREQ (string, "print") \ | |
134 | || STREQ (string, "punct") || STREQ (string, "graph") \ | |
135 | || STREQ (string, "cntrl") || STREQ (string, "blank")) | |
136 | # endif | |
137 | ||
138 | /* Avoid depending on library functions or files | |
139 | whose names are inconsistent. */ | |
140 | ||
141 | /* Global variable. */ | |
142 | static int posixly_correct; | |
143 | ||
144 | # ifndef internal_function | |
145 | /* Inside GNU libc we mark some function in a special way. In other | |
146 | environments simply ignore the marking. */ | |
147 | # define internal_function | |
148 | # endif | |
149 | ||
150 | /* Note that this evaluates C many times. */ | |
151 | # define FOLD(c) ((flags & FNM_CASEFOLD) ? tolower (c) : (c)) | |
152 | # define CHAR char | |
153 | # define UCHAR unsigned char | |
154 | # define INT int | |
155 | # define FCT internal_fnmatch | |
156 | # define EXT ext_match | |
157 | # define END end_pattern | |
158 | # define L_(CS) CS | |
159 | # ifdef _LIBC | |
160 | # define BTOWC(C) __btowc (C) | |
161 | # else | |
162 | # define BTOWC(C) btowc (C) | |
163 | # endif | |
164 | # define STRLEN(S) strlen (S) | |
165 | # define STRCAT(D, S) strcat (D, S) | |
166 | # ifdef _LIBC | |
167 | # define MEMPCPY(D, S, N) __mempcpy (D, S, N) | |
168 | # else | |
169 | # if HAVE_MEMPCPY | |
170 | # define MEMPCPY(D, S, N) mempcpy (D, S, N) | |
171 | # else | |
172 | # define MEMPCPY(D, S, N) ((void *) ((char *) memcpy (D, S, N) + (N))) | |
173 | # endif | |
174 | # endif | |
175 | # define MEMCHR(S, C, N) memchr (S, C, N) | |
176 | # include "fnmatch_loop.c" | |
177 | ||
178 | ||
179 | # if HANDLE_MULTIBYTE | |
180 | # define FOLD(c) ((flags & FNM_CASEFOLD) ? towlower (c) : (c)) | |
181 | # define CHAR wchar_t | |
182 | # define UCHAR wint_t | |
183 | # define INT wint_t | |
184 | # define FCT internal_fnwmatch | |
185 | # define EXT ext_wmatch | |
186 | # define END end_wpattern | |
187 | # define L_(CS) L##CS | |
188 | # define BTOWC(C) (C) | |
189 | # ifdef _LIBC | |
190 | # define STRLEN(S) __wcslen (S) | |
191 | # define STRCAT(D, S) __wcscat (D, S) | |
192 | # define MEMPCPY(D, S, N) __wmempcpy (D, S, N) | |
193 | # else | |
194 | # define STRLEN(S) wcslen (S) | |
195 | # define STRCAT(D, S) wcscat (D, S) | |
196 | # if HAVE_WMEMPCPY | |
197 | # define MEMPCPY(D, S, N) wmempcpy (D, S, N) | |
198 | # else | |
199 | # define MEMPCPY(D, S, N) (wmemcpy (D, S, N) + (N)) | |
200 | # endif | |
201 | # endif | |
202 | # define MEMCHR(S, C, N) wmemchr (S, C, N) | |
203 | # define WIDE_CHAR_VERSION 1 | |
204 | ||
205 | # undef IS_CHAR_CLASS | |
206 | /* We have to convert the wide character string in a multibyte string. But | |
207 | we know that the character class names consist of alphanumeric characters | |
208 | from the portable character set, and since the wide character encoding | |
209 | for a member of the portable character set is the same code point as | |
210 | its single-byte encoding, we can use a simplified method to convert the | |
211 | string to a multibyte character string. */ | |
212 | static wctype_t | |
213 | is_char_class (const wchar_t *wcs) | |
214 | { | |
215 | char s[CHAR_CLASS_MAX_LENGTH + 1]; | |
216 | char *cp = s; | |
217 | ||
218 | do | |
219 | { | |
220 | /* Test for a printable character from the portable character set. */ | |
221 | # ifdef _LIBC | |
222 | if (*wcs < 0x20 || *wcs > 0x7e | |
223 | || *wcs == 0x24 || *wcs == 0x40 || *wcs == 0x60) | |
224 | return (wctype_t) 0; | |
225 | # else | |
226 | switch (*wcs) | |
227 | { | |
228 | case L' ': case L'!': case L'"': case L'#': case L'%': | |
229 | case L'&': case L'\'': case L'(': case L')': case L'*': | |
230 | case L'+': case L',': case L'-': case L'.': case L'/': | |
231 | case L'0': case L'1': case L'2': case L'3': case L'4': | |
232 | case L'5': case L'6': case L'7': case L'8': case L'9': | |
233 | case L':': case L';': case L'<': case L'=': case L'>': | |
234 | case L'?': | |
235 | case L'A': case L'B': case L'C': case L'D': case L'E': | |
236 | case L'F': case L'G': case L'H': case L'I': case L'J': | |
237 | case L'K': case L'L': case L'M': case L'N': case L'O': | |
238 | case L'P': case L'Q': case L'R': case L'S': case L'T': | |
239 | case L'U': case L'V': case L'W': case L'X': case L'Y': | |
240 | case L'Z': | |
241 | case L'[': case L'\\': case L']': case L'^': case L'_': | |
242 | case L'a': case L'b': case L'c': case L'd': case L'e': | |
243 | case L'f': case L'g': case L'h': case L'i': case L'j': | |
244 | case L'k': case L'l': case L'm': case L'n': case L'o': | |
245 | case L'p': case L'q': case L'r': case L's': case L't': | |
246 | case L'u': case L'v': case L'w': case L'x': case L'y': | |
247 | case L'z': case L'{': case L'|': case L'}': case L'~': | |
248 | break; | |
249 | default: | |
250 | return (wctype_t) 0; | |
251 | } | |
252 | # endif | |
253 | ||
254 | /* Avoid overrunning the buffer. */ | |
255 | if (cp == s + CHAR_CLASS_MAX_LENGTH) | |
256 | return (wctype_t) 0; | |
257 | ||
258 | *cp++ = (char) *wcs++; | |
259 | } | |
260 | while (*wcs != L'\0'); | |
261 | ||
262 | *cp = '\0'; | |
263 | ||
264 | # ifdef _LIBC | |
265 | return __wctype (s); | |
266 | # else | |
267 | return wctype (s); | |
268 | # endif | |
269 | } | |
270 | # define IS_CHAR_CLASS(string) is_char_class (string) | |
271 | ||
272 | # include "fnmatch_loop.c" | |
273 | # endif | |
274 | ||
275 | ||
276 | int | |
277 | fnmatch (const char *pattern, const char *string, int flags) | |
278 | { | |
279 | # if HANDLE_MULTIBYTE | |
280 | # define ALLOCA_LIMIT 2000 | |
281 | if (__builtin_expect (MB_CUR_MAX, 1) != 1) | |
282 | { | |
283 | mbstate_t ps; | |
284 | size_t patsize; | |
285 | size_t strsize; | |
286 | size_t totsize; | |
287 | wchar_t *wpattern; | |
288 | wchar_t *wstring; | |
289 | int res; | |
290 | ||
291 | /* Calculate the size needed to convert the strings to | |
292 | wide characters. */ | |
293 | memset (&ps, '\0', sizeof (ps)); | |
294 | patsize = mbsrtowcs (NULL, &pattern, 0, &ps) + 1; | |
295 | if (__builtin_expect (patsize != 0, 1)) | |
296 | { | |
297 | assert (mbsinit (&ps)); | |
298 | strsize = mbsrtowcs (NULL, &string, 0, &ps) + 1; | |
299 | if (__builtin_expect (strsize != 0, 1)) | |
300 | { | |
301 | assert (mbsinit (&ps)); | |
302 | totsize = patsize + strsize; | |
303 | if (__builtin_expect (! (patsize <= totsize | |
304 | && totsize <= SIZE_MAX / sizeof (wchar_t)), | |
305 | 0)) | |
306 | { | |
307 | errno = ENOMEM; | |
308 | return -1; | |
309 | } | |
310 | ||
311 | /* Allocate room for the wide characters. */ | |
312 | if (__builtin_expect (totsize < ALLOCA_LIMIT, 1)) | |
313 | wpattern = (wchar_t *) alloca (totsize * sizeof (wchar_t)); | |
314 | else | |
315 | { | |
316 | wpattern = malloc (totsize * sizeof (wchar_t)); | |
317 | if (__builtin_expect (! wpattern, 0)) | |
318 | { | |
319 | errno = ENOMEM; | |
320 | return -1; | |
321 | } | |
322 | } | |
323 | wstring = wpattern + patsize; | |
324 | ||
325 | /* Convert the strings into wide characters. */ | |
326 | mbsrtowcs (wpattern, &pattern, patsize, &ps); | |
327 | assert (mbsinit (&ps)); | |
328 | mbsrtowcs (wstring, &string, strsize, &ps); | |
329 | ||
330 | res = internal_fnwmatch (wpattern, wstring, wstring + strsize - 1, | |
331 | flags & FNM_PERIOD, flags); | |
332 | ||
333 | if (__builtin_expect (! (totsize < ALLOCA_LIMIT), 0)) | |
334 | free (wpattern); | |
335 | return res; | |
336 | } | |
337 | } | |
338 | } | |
339 | ||
340 | # endif /* HANDLE_MULTIBYTE */ | |
341 | ||
342 | return internal_fnmatch (pattern, string, string + strlen (string), | |
343 | flags & FNM_PERIOD, flags); | |
344 | } | |
345 | ||
346 | # ifdef _LIBC | |
347 | # undef fnmatch | |
348 | versioned_symbol (libc, __fnmatch, fnmatch, GLIBC_2_2_3); | |
349 | # if SHLIB_COMPAT(libc, GLIBC_2_0, GLIBC_2_2_3) | |
350 | strong_alias (__fnmatch, __fnmatch_old) | |
351 | compat_symbol (libc, __fnmatch_old, fnmatch, GLIBC_2_0); | |
352 | # endif | |
353 | libc_hidden_ver (__fnmatch, fnmatch) | |
354 | # endif | |
355 | ||
356 | #endif /* _LIBC or not __GNU_LIBRARY__. */ |