1 /* $OpenBSD: glob.c,v 1.26 2005/11/28 17:50:12 deraadt Exp $ */
2 /**
3 * \file
4 * Copyright (c) 1989, 1993
5 * The Regents of the University of California. All rights reserved.
6 *
7 * This code is derived from software contributed to Berkeley by
8 * Guido van Rossum.
9 *
10 * Redistribution and use in source and binary forms, with or without
11 * modification, are permitted provided that the following conditions
12 * are met:
13 * 1. Redistributions of source code must retain the above copyright
14 * notice, this list of conditions and the following disclaimer.
15 * 2. Redistributions in binary form must reproduce the above copyright
16 * notice, this list of conditions and the following disclaimer in the
17 * documentation and/or other materials provided with the distribution.
18 * 3. Neither the name of the University nor the names of its contributors
19 * may be used to endorse or promote products derived from this software
20 * without specific prior written permission.
21 *
22 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
23 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
24 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
25 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
26 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
27 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
28 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
29 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
30 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
31 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
32 * SUCH DAMAGE.
33 */
34
35 /*
36 * mono_w32file_unix_glob(3) -- a subset of the one defined in POSIX 1003.2.
37 *
38 * Optional extra services, controlled by flags not defined by POSIX:
39 *
40 * GLOB_MAGCHAR:
41 * Set in gl_flags if pattern contained a globbing character.
42 */
43 #include <sys/types.h>
44 #include <sys/stat.h>
45
46 #include <glib.h>
47 #include <ctype.h>
48 #include <errno.h>
49 #include <stdio.h>
50 #include <stdlib.h>
51 #include <string.h>
52 #include <unistd.h>
53
54 #include "w32file-unix-glob.h"
55
56 #define EOS '\0'
57 #define NOT '!'
58 #define QUESTION '?'
59 #define QUOTE '\\'
60 #define STAR '*'
61
62 #ifndef DEBUG
63
64 #define M_QUOTE 0x8000
65 #define M_PROTECT 0x4000
66 #define M_MASK 0xffff
67 #define M_ASCII 0x00ff
68
69 typedef unsigned short Char;
70
71 #else
72
73 #define M_QUOTE 0x80
74 #define M_PROTECT 0x40
75 #define M_MASK 0xff
76 #define M_ASCII 0x7f
77
78 typedef char Char;
79
80 #endif
81
82
83 #define CHAR(c) ((gchar)((c)&M_ASCII))
84 #define META(c) ((gchar)((c)|M_QUOTE))
85 #define M_ALL META('*')
86 #define M_ONE META('?')
87 #define ismeta(c) (((c)&M_QUOTE) != 0)
88
89
90 static int
91 g_Ctoc(const gchar *, char *, unsigned int);
92
93 static int
94 glob0(GDir *dir, const gchar *, mono_w32file_unix_glob_t *, gboolean, gboolean);
95 static int
96 glob1(GDir *dir, gchar *, gchar *, mono_w32file_unix_glob_t *, size_t *, gboolean, gboolean);
97
98 static int
99 glob3(GDir *dir, gchar *, gchar *, mono_w32file_unix_glob_t *, size_t *, gboolean, gboolean);
100
101 static int
102 globextend(const gchar *, mono_w32file_unix_glob_t *, size_t *);
103
104 static int
105 match(const gchar *, gchar *, gchar *, gboolean);
106
107 #ifdef DEBUG_ENABLED
108 static void qprintf(const char *, Char *);
109 #endif
110
111 int
mono_w32file_unix_glob(GDir * dir,const char * pattern,int flags,mono_w32file_unix_glob_t * pglob)112 mono_w32file_unix_glob(GDir *dir, const char *pattern, int flags, mono_w32file_unix_glob_t *pglob)
113 {
114 const unsigned char *patnext;
115 int c;
116 gchar *bufnext, *bufend, patbuf[PATH_MAX];
117
118 patnext = (unsigned char *) pattern;
119 if (!(flags & W32FILE_UNIX_GLOB_APPEND)) {
120 pglob->gl_pathc = 0;
121 pglob->gl_pathv = NULL;
122 pglob->gl_offs = 0;
123 }
124 pglob->gl_flags = flags & ~W32FILE_UNIX_GLOB_MAGCHAR;
125
126 bufnext = patbuf;
127 bufend = bufnext + PATH_MAX - 1;
128
129 /* Protect the quoted characters. */
130 while (bufnext < bufend && (c = *patnext++) != EOS)
131 if (c == QUOTE) {
132 if ((c = *patnext++) == EOS) {
133 c = QUOTE;
134 --patnext;
135 }
136 *bufnext++ = c | M_PROTECT;
137 } else
138 *bufnext++ = c;
139
140 *bufnext = EOS;
141
142 return glob0(dir, patbuf, pglob, flags & W32FILE_UNIX_GLOB_IGNORECASE,
143 flags & W32FILE_UNIX_GLOB_UNIQUE);
144 }
145
146 /*
147 * The main glob() routine: compiles the pattern (optionally processing
148 * quotes), calls glob1() to do the real pattern matching, and finally
149 * sorts the list (unless unsorted operation is requested). Returns 0
150 * if things went well, nonzero if errors occurred. It is not an error
151 * to find no matches.
152 */
153 static int
glob0(GDir * dir,const gchar * pattern,mono_w32file_unix_glob_t * pglob,gboolean ignorecase,gboolean unique)154 glob0(GDir *dir, const gchar *pattern, mono_w32file_unix_glob_t *pglob, gboolean ignorecase,
155 gboolean unique)
156 {
157 const gchar *qpatnext;
158 int c, err, oldpathc;
159 gchar *bufnext, patbuf[PATH_MAX];
160 size_t limit = 0;
161
162 qpatnext = pattern;
163 oldpathc = pglob->gl_pathc;
164 bufnext = patbuf;
165
166 /* We don't need to check for buffer overflow any more. */
167 while ((c = *qpatnext++) != EOS) {
168 switch (c) {
169 case QUESTION:
170 pglob->gl_flags |= W32FILE_UNIX_GLOB_MAGCHAR;
171 *bufnext++ = M_ONE;
172 break;
173 case STAR:
174 pglob->gl_flags |= W32FILE_UNIX_GLOB_MAGCHAR;
175 /* collapse adjacent stars to one,
176 * to avoid exponential behavior
177 */
178 if (bufnext == patbuf || bufnext[-1] != M_ALL)
179 *bufnext++ = M_ALL;
180 break;
181 default:
182 *bufnext++ = CHAR(c);
183 break;
184 }
185 }
186 *bufnext = EOS;
187 #ifdef DEBUG_ENABLED
188 qprintf("glob0:", patbuf);
189 #endif
190
191 if ((err = glob1(dir, patbuf, patbuf+PATH_MAX-1, pglob, &limit,
192 ignorecase, unique)) != 0)
193 return(err);
194
195 if (pglob->gl_pathc == oldpathc) {
196 return(W32FILE_UNIX_GLOB_NOMATCH);
197 }
198
199 return(0);
200 }
201
202 static int
glob1(GDir * dir,gchar * pattern,gchar * pattern_last,mono_w32file_unix_glob_t * pglob,size_t * limitp,gboolean ignorecase,gboolean unique)203 glob1(GDir *dir, gchar *pattern, gchar *pattern_last, mono_w32file_unix_glob_t *pglob,
204 size_t *limitp, gboolean ignorecase, gboolean unique)
205 {
206 /* A null pathname is invalid -- POSIX 1003.1 sect. 2.4. */
207 if (*pattern == EOS)
208 return(0);
209 return(glob3(dir, pattern, pattern_last, pglob, limitp, ignorecase,
210 unique));
211 }
212
contains(mono_w32file_unix_glob_t * pglob,const gchar * name)213 static gboolean contains (mono_w32file_unix_glob_t *pglob, const gchar *name)
214 {
215 int i;
216 char **pp;
217
218 if (pglob->gl_pathv != NULL) {
219 pp = pglob->gl_pathv + pglob->gl_offs;
220 for (i = pglob->gl_pathc; i--; ++pp) {
221 if (*pp) {
222 if (!strcmp (*pp, name)) {
223 return(TRUE);
224 }
225 }
226 }
227 }
228
229 return(FALSE);
230 }
231
232 static int
glob3(GDir * dir,gchar * pattern,gchar * pattern_last,mono_w32file_unix_glob_t * pglob,size_t * limitp,gboolean ignorecase,gboolean unique)233 glob3(GDir *dir, gchar *pattern, gchar *pattern_last, mono_w32file_unix_glob_t *pglob,
234 size_t *limitp, gboolean ignorecase, gboolean unique)
235 {
236 const gchar *name;
237
238 /* Search directory for matching names. */
239 while ((name = g_dir_read_name(dir))) {
240 if (!match(name, pattern, pattern + strlen (pattern),
241 ignorecase)) {
242 continue;
243 }
244 if (!unique ||
245 !contains (pglob, name)) {
246 globextend (name, pglob, limitp);
247 }
248 }
249
250 return(0);
251 }
252
253
254 /*
255 * Extend the gl_pathv member of a mono_w32file_unix_glob_t structure to accommodate a new item,
256 * add the new item, and update gl_pathc.
257 *
258 * This assumes the BSD realloc, which only copies the block when its size
259 * crosses a power-of-two boundary; for v7 realloc, this would cause quadratic
260 * behavior.
261 *
262 * Return 0 if new item added, error code if memory couldn't be allocated.
263 *
264 * Invariant of the mono_w32file_unix_glob_t structure:
265 * Either gl_pathc is zero and gl_pathv is NULL; or gl_pathc > 0 and
266 * gl_pathv points to (gl_offs + gl_pathc + 1) items.
267 */
268 static int
globextend(const gchar * path,mono_w32file_unix_glob_t * pglob,size_t * limitp)269 globextend(const gchar *path, mono_w32file_unix_glob_t *pglob, size_t *limitp)
270 {
271 char **pathv;
272 int i;
273 unsigned int newsize, len;
274 char *copy;
275 const gchar *p;
276
277 newsize = sizeof(*pathv) * (2 + pglob->gl_pathc + pglob->gl_offs);
278 /* FIXME: Can just use realloc(). */
279 pathv = (char **)(pglob->gl_pathv ? g_realloc ((char *)pglob->gl_pathv, newsize) :
280 g_malloc (newsize));
281 if (pathv == NULL) {
282 if (pglob->gl_pathv) {
283 g_free (pglob->gl_pathv);
284 pglob->gl_pathv = NULL;
285 }
286 return(W32FILE_UNIX_GLOB_NOSPACE);
287 }
288
289 if (pglob->gl_pathv == NULL && pglob->gl_offs > 0) {
290 /* first time around -- clear initial gl_offs items */
291 pathv += pglob->gl_offs;
292 for (i = pglob->gl_offs; --i >= 0; )
293 *--pathv = NULL;
294 }
295 pglob->gl_pathv = pathv;
296
297 for (p = path; *p++;)
298 ;
299 len = (size_t)(p - path);
300 *limitp += len;
301 if ((copy = (char *)malloc(len)) != NULL) {
302 if (g_Ctoc(path, copy, len)) {
303 g_free (copy);
304 return(W32FILE_UNIX_GLOB_NOSPACE);
305 }
306 pathv[pglob->gl_offs + pglob->gl_pathc++] = copy;
307 }
308 pathv[pglob->gl_offs + pglob->gl_pathc] = NULL;
309
310 #if 0
311 /* Broken on opensuse 11 */
312 if ((pglob->gl_flags & W32FILE_UNIX_GLOB_LIMIT) &&
313 newsize + *limitp >= ARG_MAX) {
314 errno = 0;
315 return(W32FILE_UNIX_GLOB_NOSPACE);
316 }
317 #endif
318
319 return(copy == NULL ? W32FILE_UNIX_GLOB_NOSPACE : 0);
320 }
321
322
323 /*
324 * pattern matching function for filenames. Each occurrence of the *
325 * pattern causes a recursion level.
326 */
327 static int
match(const gchar * name,gchar * pat,gchar * patend,gboolean ignorecase)328 match(const gchar *name, gchar *pat, gchar *patend, gboolean ignorecase)
329 {
330 gchar c;
331
332 while (pat < patend) {
333 c = *pat++;
334 switch (c & M_MASK) {
335 case M_ALL:
336 if (pat == patend)
337 return(1);
338 do {
339 if (match(name, pat, patend, ignorecase))
340 return(1);
341 } while (*name++ != EOS);
342 return(0);
343 case M_ONE:
344 if (*name++ == EOS)
345 return(0);
346 break;
347 default:
348 if (ignorecase) {
349 if (g_ascii_tolower (*name++) != g_ascii_tolower (c))
350 return(0);
351 } else {
352 if (*name++ != c)
353 return(0);
354 }
355
356 break;
357 }
358 }
359 return(*name == EOS);
360 }
361
362 /* Free allocated data belonging to a mono_w32file_unix_glob_t structure. */
363 void
mono_w32file_unix_globfree(mono_w32file_unix_glob_t * pglob)364 mono_w32file_unix_globfree(mono_w32file_unix_glob_t *pglob)
365 {
366 int i;
367 char **pp;
368
369 if (pglob->gl_pathv != NULL) {
370 pp = pglob->gl_pathv + pglob->gl_offs;
371 for (i = pglob->gl_pathc; i--; ++pp)
372 if (*pp)
373 g_free (*pp);
374 g_free (pglob->gl_pathv);
375 pglob->gl_pathv = NULL;
376 }
377 }
378
379 static int
g_Ctoc(const gchar * str,char * buf,unsigned int len)380 g_Ctoc(const gchar *str, char *buf, unsigned int len)
381 {
382
383 while (len--) {
384 if ((*buf++ = *str++) == EOS)
385 return (0);
386 }
387 return (1);
388 }
389
390 #ifdef DEBUG_ENABLED
391 static void
qprintf(const char * str,Char * s)392 qprintf(const char *str, Char *s)
393 {
394 Char *p;
395
396 (void)printf("%s:\n", str);
397 for (p = s; *p; p++)
398 (void)printf("%c", CHAR(*p));
399 (void)printf("\n");
400 for (p = s; *p; p++)
401 (void)printf("%c", *p & M_PROTECT ? '"' : ' ');
402 (void)printf("\n");
403 for (p = s; *p; p++)
404 (void)printf("%c", ismeta(*p) ? '_' : ' ');
405 (void)printf("\n");
406 }
407 #endif
408