1/*
2 * Copyright (c) 1989, 1993
3 * The Regents of the University of California. All rights reserved.
4 *
5 * This code is derived from software contributed to Berkeley by
6 * Guido van Rossum.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
11 * 1. Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in the
15 * documentation and/or other materials provided with the distribution.
16 * 3. All advertising materials mentioning features or use of this software
17 * must display the following acknowledgement:
18 * This product includes software developed by the University of
19 * California, Berkeley and its contributors.
20 * 4. Neither the name of the University nor the names of its contributors
21 * may be used to endorse or promote products derived from this software
22 * without specific prior written permission.
23 *
24 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
25 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
26 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
27 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
28 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
29 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
30 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
31 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
32 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
33 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
34 * SUCH DAMAGE.
35 */
36
37#if defined(LIBC_SCCS) && !defined(lint)
38static char sccsid[] = "@(#)glob.c 8.3 (Berkeley) 10/13/93";
39#endif /* LIBC_SCCS and not lint */
40
41/*
42 * glob(3) -- a superset of the one defined in POSIX 1003.2.
43 *
44 * The [!...] convention to negate a range is supported (SysV, Posix, ksh).
45 *
46 * Optional extra services, controlled by flags not defined by POSIX:
47 *
48 * GLOB_QUOTE:
49 * Escaping convention: \ inhibits any special meaning the following
50 * character might have (except \ at end of string is retained).
51 * GLOB_MAGCHAR:
52 * Set in gl_flags if pattern contained a globbing character.
53 * GLOB_NOMAGIC:
54 * Same as GLOB_NOCHECK, but it will only append pattern if it did
55 * not contain any magic characters. [Used in csh style globbing]
56 * GLOB_ALTDIRFUNC:
57 * Use alternately specified directory access functions.
58 * GLOB_TILDE:
59 * expand ~user/foo to the /home/dir/of/user/foo
60 * GLOB_BRACE:
61 * expand {1,2}{a,b} to 1a 1b 2a 2b
62 * gl_matchc:
63 * Number of matches in the current invocation of glob.
64 */
65
66#include <sys/param.h>
67#include <sys/stat.h>
68
69#include <ctype.h>
70#include <dirent.h>
71#include <errno.h>
72#include <glob.h>
73#include <pwd.h>
74#include <stdio.h>
75#include <stdlib.h>
76#include <string.h>
77#include <unistd.h>
78
79#include "collate.h"
80
81#define DOLLAR '$'
82#define DOT '.'
83#define EOS '\0'
84#define LBRACKET '['
85#define NOT '!'
86#define QUESTION '?'
87#define QUOTE '\\'
88#define RANGE '-'
89#define RBRACKET ']'
90#define SEP '/'
91#define STAR '*'
92#define TILDE '~'
93#define UNDERSCORE '_'
94#define LBRACE '{'
95#define RBRACE '}'
96#define SLASH '/'
97#define COMMA ','
98
99#ifndef DEBUG
100
101#define M_QUOTE 0x8000
102#define M_PROTECT 0x4000
103#define M_MASK 0xffff
104#define M_ASCII 0x00ff
105
106typedef u_short Char;
107
108#else
109
110#define M_QUOTE 0x80
111#define M_PROTECT 0x40
112#define M_MASK 0xff
113#define M_ASCII 0x7f
114
115typedef char Char;
116
117#endif
118
119
120#define CHAR(c) ((Char)((c)&M_ASCII))
121#define META(c) ((Char)((c)|M_QUOTE))
122#define M_ALL META('*')
123#define M_END META(']')
124#define M_NOT META('!')
125#define M_ONE META('?')
126#define M_RNG META('-')
127#define M_SET META('[')
128#define ismeta(c) (((c)&M_QUOTE) != 0)
129
130
131static int compare __P((const void *, const void *));
132static void g_Ctoc __P((const Char *, char *));
133static int g_lstat __P((Char *, struct stat *, glob_t *));
134static DIR *g_opendir __P((Char *, glob_t *));
135static Char *g_strchr __P((Char *, int));
136#ifdef notdef
137static Char *g_strcat __P((Char *, const Char *));
138#endif
139static int g_stat __P((Char *, struct stat *, glob_t *));
140static int glob0 __P((const Char *, glob_t *));
141static int glob1 __P((Char *, glob_t *));
142static int glob2 __P((Char *, Char *, Char *, glob_t *));
143static int glob3 __P((Char *, Char *, Char *, Char *, glob_t *));
144static int globextend __P((const Char *, glob_t *));
145static const Char * globtilde __P((const Char *, Char *, size_t, glob_t *));
146static int globexp1 __P((const Char *, glob_t *));
147static int globexp2 __P((const Char *, const Char *, glob_t *, int *));
148static int match __P((Char *, Char *, Char *));
149#ifdef DEBUG
150static void qprintf __P((const char *, Char *));
151#endif
152
153int
154glob(pattern, flags, errfunc, pglob)
155 const char *pattern;
156 int flags, (*errfunc) __P((const char *, int));
157 glob_t *pglob;
158{
159 const u_char *patnext;
160 int c;
161 Char *bufnext, *bufend, patbuf[MAXPATHLEN+1];
162
163 patnext = (u_char *) pattern;
164 if (!(flags & GLOB_APPEND)) {
165 pglob->gl_pathc = 0;
166 pglob->gl_pathv = NULL;
167 if (!(flags & GLOB_DOOFFS))
168 pglob->gl_offs = 0;
169 }
170 pglob->gl_flags = flags & ~GLOB_MAGCHAR;
171 pglob->gl_errfunc = errfunc;
172 pglob->gl_matchc = 0;
173
174 bufnext = patbuf;
175 bufend = bufnext + MAXPATHLEN;
176 if (flags & GLOB_QUOTE) {
177 /* Protect the quoted characters. */
178 while (bufnext < bufend && (c = *patnext++) != EOS)
179 if (c == QUOTE) {
180 if ((c = *patnext++) == EOS) {
181 c = QUOTE;
182 --patnext;
183 }
184 *bufnext++ = c | M_PROTECT;
185 }
186 else
187 *bufnext++ = c;
188 }
189 else
190 while (bufnext < bufend && (c = *patnext++) != EOS)
191 *bufnext++ = c;
192 *bufnext = EOS;
193
194 if (flags & GLOB_BRACE)
195 return globexp1(patbuf, pglob);
196 else
197 return glob0(patbuf, pglob);
198}
199
200/*
201 * Expand recursively a glob {} pattern. When there is no more expansion
202 * invoke the standard globbing routine to glob the rest of the magic
203 * characters
204 */
205static int globexp1(pattern, pglob)
206 const Char *pattern;
207 glob_t *pglob;
208{
209 const Char* ptr = pattern;
210 int rv;
211
212 /* Protect a single {}, for find(1), like csh */
213 if (pattern[0] == LBRACE && pattern[1] == RBRACE && pattern[2] == EOS)
214 return glob0(pattern, pglob);
215
216 while ((ptr = (const Char *) g_strchr((Char *) ptr, LBRACE)) != NULL)
217 if (!globexp2(ptr, pattern, pglob, &rv))
218 return rv;
219
220 return glob0(pattern, pglob);
221}
222
223
224/*
225 * Recursive brace globbing helper. Tries to expand a single brace.
226 * If it succeeds then it invokes globexp1 with the new pattern.
227 * If it fails then it tries to glob the rest of the pattern and returns.
228 */
229static int globexp2(ptr, pattern, pglob, rv)
230 const Char *ptr, *pattern;
231 glob_t *pglob;
232 int *rv;
233{
234 int i;
235 Char *lm, *ls;
236 const Char *pe, *pm, *pl;
237 Char patbuf[MAXPATHLEN + 1];
238
239 /* copy part up to the brace */
240 for (lm = patbuf, pm = pattern; pm != ptr; *lm++ = *pm++)
241 continue;
242 ls = lm;
243
244 /* Find the balanced brace */
245 for (i = 0, pe = ++ptr; *pe; pe++)
246 if (*pe == LBRACKET) {
247 /* Ignore everything between [] */
248 for (pm = pe++; *pe != RBRACKET && *pe != EOS; pe++)
249 continue;
250 if (*pe == EOS) {
251 /*
252 * We could not find a matching RBRACKET.
253 * Ignore and just look for RBRACE
254 */
255 pe = pm;
256 }
257 }
258 else if (*pe == LBRACE)
259 i++;
260 else if (*pe == RBRACE) {
261 if (i == 0)
262 break;
263 i--;
264 }
265
266 /* Non matching braces; just glob the pattern */
267 if (i != 0 || *pe == EOS) {
268 *rv = glob0(patbuf, pglob);
269 return 0;
270 }
271
272 for (i = 0, pl = pm = ptr; pm <= pe; pm++)
273 switch (*pm) {
274 case LBRACKET:
275 /* Ignore everything between [] */
276 for (pl = pm++; *pm != RBRACKET && *pm != EOS; pm++)
277 continue;
278 if (*pm == EOS) {
279 /*
280 * We could not find a matching RBRACKET.
281 * Ignore and just look for RBRACE
282 */
283 pm = pl;
284 }
285 break;
286
287 case LBRACE:
288 i++;
289 break;
290
291 case RBRACE:
292 if (i) {
293 i--;
294 break;
295 }
296 /* FALLTHROUGH */
297 case COMMA:
298 if (i && *pm == COMMA)
299 break;
300 else {
301 /* Append the current string */
302 for (lm = ls; (pl < pm); *lm++ = *pl++)
303 continue;
304 /*
305 * Append the rest of the pattern after the
306 * closing brace
307 */
308 for (pl = pe + 1; (*lm++ = *pl++) != EOS;)
309 continue;
310
311 /* Expand the current pattern */
312#ifdef DEBUG
313 qprintf("globexp2:", patbuf);
314#endif
315 *rv = globexp1(patbuf, pglob);
316
317 /* move after the comma, to the next string */
318 pl = pm + 1;
319 }
320 break;
321
322 default:
323 break;
324 }
325 *rv = 0;
326 return 0;
327}
328
329
330
331/*
332 * expand tilde from the passwd file.
333 */
334static const Char *
335globtilde(pattern, patbuf, patbuf_len, pglob)
336 const Char *pattern;
337 Char *patbuf;
338 size_t patbuf_len;
339 glob_t *pglob;
340{
341 struct passwd *pwd;
342 char *h;
343 const Char *p;
344 Char *b, *eb;
345
346 if (*pattern != TILDE || !(pglob->gl_flags & GLOB_TILDE))
347 return pattern;
348
349 /*
350 * Copy up to the end of the string or /
351 */
352 eb = &patbuf[patbuf_len - 1];
353 for (p = pattern + 1, h = (char *) patbuf;
354 h < (char *)eb && *p && *p != SLASH; *h++ = *p++)
355 continue;
356
357 *h = EOS;
358
359 if (((char *) patbuf)[0] == EOS) {
360 /*
361 * handle a plain ~ or ~/ by expanding $HOME
362 * first and then trying the password file
363 */
364 if ((h = getenv("HOME")) == NULL) {
365 if ((pwd = getpwuid(getuid())) == NULL)
366 return pattern;
367 else
368 h = pwd->pw_dir;
369 }
370 }
371 else {
372 /*
373 * Expand a ~user
374 */
375 if ((pwd = getpwnam((char*) patbuf)) == NULL)
376 return pattern;
377 else
378 h = pwd->pw_dir;
379 }
380
381 /* Copy the home directory */
382 for (b = patbuf; b < eb && *h; *b++ = *h++)
383 continue;
384
385 /* Append the rest of the pattern */
386 while (b < eb && (*b++ = *p++) != EOS)
387 continue;
388 *b = EOS;
389
390 return patbuf;
391}
392
393
394/*
395 * The main glob() routine: compiles the pattern (optionally processing
396 * quotes), calls glob1() to do the real pattern matching, and finally
397 * sorts the list (unless unsorted operation is requested). Returns 0
398 * if things went well, nonzero if errors occurred. It is not an error
399 * to find no matches.
400 */
401static int
402glob0(pattern, pglob)
403 const Char *pattern;
404 glob_t *pglob;
405{
406 const Char *qpatnext;
407 int c, err, oldpathc;
408 Char *bufnext, patbuf[MAXPATHLEN+1];
409
410 qpatnext = globtilde(pattern, patbuf, sizeof(patbuf) / sizeof(Char),
411 pglob);
412 oldpathc = pglob->gl_pathc;
413 bufnext = patbuf;
414
415 /* We don't need to check for buffer overflow any more. */
416 while ((c = *qpatnext++) != EOS) {
417 switch (c) {
418 case LBRACKET:
419 c = *qpatnext;
420 if (c == NOT)
421 ++qpatnext;
422 if (*qpatnext == EOS ||
423 g_strchr((Char *) qpatnext+1, RBRACKET) == NULL) {
424 *bufnext++ = LBRACKET;
425 if (c == NOT)
426 --qpatnext;
427 break;
428 }
429 *bufnext++ = M_SET;
430 if (c == NOT)
431 *bufnext++ = M_NOT;
432 c = *qpatnext++;
433 do {
434 *bufnext++ = CHAR(c);
435 if (*qpatnext == RANGE &&
436 (c = qpatnext[1]) != RBRACKET) {
437 *bufnext++ = M_RNG;
438 *bufnext++ = CHAR(c);
439 qpatnext += 2;
440 }
441 } while ((c = *qpatnext++) != RBRACKET);
442 pglob->gl_flags |= GLOB_MAGCHAR;
443 *bufnext++ = M_END;
444 break;
445 case QUESTION:
446 pglob->gl_flags |= GLOB_MAGCHAR;
447 *bufnext++ = M_ONE;
448 break;
449 case STAR:
450 pglob->gl_flags |= GLOB_MAGCHAR;
451 /* collapse adjacent stars to one,
452 * to avoid exponential behavior
453 */
454 if (bufnext == patbuf || bufnext[-1] != M_ALL)
455 *bufnext++ = M_ALL;
456 break;
457 default:
458 *bufnext++ = CHAR(c);
459 break;
460 }
461 }
462 *bufnext = EOS;
463#ifdef DEBUG
464 qprintf("glob0:", patbuf);
465#endif
466
467 if ((err = glob1(patbuf, pglob)) != 0)
468 return(err);
469
470 /*
471 * If there was no match we are going to append the pattern
472 * if GLOB_NOCHECK was specified or if GLOB_NOMAGIC was specified
473 * and the pattern did not contain any magic characters
474 * GLOB_NOMAGIC is there just for compatibility with csh.
475 */
476 if (pglob->gl_pathc == oldpathc &&
477 ((pglob->gl_flags & GLOB_NOCHECK) ||
478 ((pglob->gl_flags & GLOB_NOMAGIC) &&
479 !(pglob->gl_flags & GLOB_MAGCHAR))))
480 return(globextend(pattern, pglob));
481 else if (!(pglob->gl_flags & GLOB_NOSORT))
482 qsort(pglob->gl_pathv + pglob->gl_offs + oldpathc,
483 pglob->gl_pathc - oldpathc, sizeof(char *), compare);
484 return(0);
485}
486
487static int
488compare(p, q)
489 const void *p, *q;
490{
491 return(strcmp(*(char **)p, *(char **)q));
492}
493
494static int
495glob1(pattern, pglob)
496 Char *pattern;
497 glob_t *pglob;
498{
499 Char pathbuf[MAXPATHLEN+1];
500
501 /* A null pathname is invalid -- POSIX 1003.1 sect. 2.4. */
502 if (*pattern == EOS)
503 return(0);
504 return(glob2(pathbuf, pathbuf, pattern, pglob));
505}
506
507/*
508 * The functions glob2 and glob3 are mutually recursive; there is one level
509 * of recursion for each segment in the pattern that contains one or more
510 * meta characters.
511 */
512static int
513glob2(pathbuf, pathend, pattern, pglob)
514 Char *pathbuf, *pathend, *pattern;
515 glob_t *pglob;
516{
517 struct stat sb;
518 Char *p, *q;
519 int anymeta;
520
521 /*
522 * Loop over pattern segments until end of pattern or until
523 * segment with meta character found.
524 */
525 for (anymeta = 0;;) {
526 if (*pattern == EOS) { /* End of pattern? */
527 *pathend = EOS;
528 if (g_lstat(pathbuf, &sb, pglob))
529 return(0);
530
531 if (((pglob->gl_flags & GLOB_MARK) &&
532 pathend[-1] != SEP) && (S_ISDIR(sb.st_mode)
533 || (S_ISLNK(sb.st_mode) &&
534 (g_stat(pathbuf, &sb, pglob) == 0) &&
535 S_ISDIR(sb.st_mode)))) {
536 *pathend++ = SEP;
537 *pathend = EOS;
538 }
539 ++pglob->gl_matchc;
540 return(globextend(pathbuf, pglob));
541 }
542
543 /* Find end of next segment, copy tentatively to pathend. */
544 q = pathend;
545 p = pattern;
546 while (*p != EOS && *p != SEP) {
547 if (ismeta(*p))
548 anymeta = 1;
549 *q++ = *p++;
550 }
551
552 if (!anymeta) { /* No expansion, do next segment. */
553 pathend = q;
554 pattern = p;
555 while (*pattern == SEP)
556 *pathend++ = *pattern++;
557 } else /* Need expansion, recurse. */
558 return(glob3(pathbuf, pathend, pattern, p, pglob));
559 }
560 /* NOTREACHED */
561}
562
563static int
564glob3(pathbuf, pathend, pattern, restpattern, pglob)
565 Char *pathbuf, *pathend, *pattern, *restpattern;
566 glob_t *pglob;
567{
568 register struct dirent *dp;
569 DIR *dirp;
570 int err;
571 char buf[MAXPATHLEN];
572
573 /*
574 * The readdirfunc declaration can't be prototyped, because it is
575 * assigned, below, to two functions which are prototyped in glob.h
576 * and dirent.h as taking pointers to differently typed opaque
577 * structures.
578 */
579 struct dirent *(*readdirfunc)();
580
581 *pathend = EOS;
582 errno = 0;
583
584 if ((dirp = g_opendir(pathbuf, pglob)) == NULL) {
585 /* TODO: don't call for ENOENT or ENOTDIR? */
586 if (pglob->gl_errfunc) {
587 g_Ctoc(pathbuf, buf);
588 if (pglob->gl_errfunc(buf, errno) ||
589 pglob->gl_flags & GLOB_ERR)
590 return (GLOB_ABEND);
591 }
592 return(0);
593 }
594
595 err = 0;
596
597 /* Search directory for matching names. */
598 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
599 readdirfunc = pglob->gl_readdir;
600 else
601 readdirfunc = readdir;
602 while ((dp = (*readdirfunc)(dirp))) {
603 register u_char *sc;
604 register Char *dc;
605
606 /* Initial DOT must be matched literally. */
607 if (dp->d_name[0] == DOT && *pattern != DOT)
608 continue;
609 for (sc = (u_char *) dp->d_name, dc = pathend;
610 (*dc++ = *sc++) != EOS;)
611 continue;
612 if (!match(pathend, pattern, restpattern)) {
613 *pathend = EOS;
614 continue;
615 }
616 err = glob2(pathbuf, --dc, restpattern, pglob);
617 if (err)
618 break;
619 }
620
621 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
622 (*pglob->gl_closedir)(dirp);
623 else
624 closedir(dirp);
625 return(err);
626}
627
628
629/*
630 * Extend the gl_pathv member of a glob_t structure to accomodate a new item,
631 * add the new item, and update gl_pathc.
632 *
633 * This assumes the BSD realloc, which only copies the block when its size
634 * crosses a power-of-two boundary; for v7 realloc, this would cause quadratic
635 * behavior.
636 *
637 * Return 0 if new item added, error code if memory couldn't be allocated.
638 *
639 * Invariant of the glob_t structure:
640 * Either gl_pathc is zero and gl_pathv is NULL; or gl_pathc > 0 and
641 * gl_pathv points to (gl_offs + gl_pathc + 1) items.
642 */
643static int
644globextend(path, pglob)
645 const Char *path;
646 glob_t *pglob;
647{
648 register char **pathv;
649 register int i;
650 u_int newsize;
651 char *copy;
652 const Char *p;
653
654 newsize = sizeof(*pathv) * (2 + pglob->gl_pathc + pglob->gl_offs);
655 pathv = pglob->gl_pathv ?
656 realloc((char *)pglob->gl_pathv, newsize) :
657 malloc(newsize);
658 if (pathv == NULL)
659 return(GLOB_NOSPACE);
660
661 if (pglob->gl_pathv == NULL && pglob->gl_offs > 0) {
662 /* first time around -- clear initial gl_offs items */
663 pathv += pglob->gl_offs;
664 for (i = pglob->gl_offs; --i >= 0; )
665 *--pathv = NULL;
666 }
667 pglob->gl_pathv = pathv;
668
669 for (p = path; *p++;)
670 continue;
671 if ((copy = malloc(p - path)) != NULL) {
672 g_Ctoc(path, copy);
673 pathv[pglob->gl_offs + pglob->gl_pathc++] = copy;
674 }
675 pathv[pglob->gl_offs + pglob->gl_pathc] = NULL;
676 return(copy == NULL ? GLOB_NOSPACE : 0);
677}
678
679/*
680 * pattern matching function for filenames. Each occurrence of the *
681 * pattern causes a recursion level.
682 */
683static int
684match(name, pat, patend)
685 register Char *name, *pat, *patend;
686{
687 int ok, negate_range;
688 Char c, k;
689
690 while (pat < patend) {
691 c = *pat++;
692 switch (c & M_MASK) {
693 case M_ALL:
694 if (pat == patend)
695 return(1);
696 do
697 if (match(name, pat, patend))
698 return(1);
699 while (*name++ != EOS);
700 return(0);
701 case M_ONE:
702 if (*name++ == EOS)
703 return(0);
704 break;
705 case M_SET:
706 ok = 0;
707 if ((k = *name++) == EOS)
708 return(0);
709 if ((negate_range = ((*pat & M_MASK) == M_NOT)) != EOS)
710 ++pat;
711 while (((c = *pat++) & M_MASK) != M_END)
712 if ((*pat & M_MASK) == M_RNG) {
| 1/*
2 * Copyright (c) 1989, 1993
3 * The Regents of the University of California. All rights reserved.
4 *
5 * This code is derived from software contributed to Berkeley by
6 * Guido van Rossum.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
11 * 1. Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in the
15 * documentation and/or other materials provided with the distribution.
16 * 3. All advertising materials mentioning features or use of this software
17 * must display the following acknowledgement:
18 * This product includes software developed by the University of
19 * California, Berkeley and its contributors.
20 * 4. Neither the name of the University nor the names of its contributors
21 * may be used to endorse or promote products derived from this software
22 * without specific prior written permission.
23 *
24 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
25 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
26 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
27 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
28 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
29 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
30 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
31 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
32 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
33 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
34 * SUCH DAMAGE.
35 */
36
37#if defined(LIBC_SCCS) && !defined(lint)
38static char sccsid[] = "@(#)glob.c 8.3 (Berkeley) 10/13/93";
39#endif /* LIBC_SCCS and not lint */
40
41/*
42 * glob(3) -- a superset of the one defined in POSIX 1003.2.
43 *
44 * The [!...] convention to negate a range is supported (SysV, Posix, ksh).
45 *
46 * Optional extra services, controlled by flags not defined by POSIX:
47 *
48 * GLOB_QUOTE:
49 * Escaping convention: \ inhibits any special meaning the following
50 * character might have (except \ at end of string is retained).
51 * GLOB_MAGCHAR:
52 * Set in gl_flags if pattern contained a globbing character.
53 * GLOB_NOMAGIC:
54 * Same as GLOB_NOCHECK, but it will only append pattern if it did
55 * not contain any magic characters. [Used in csh style globbing]
56 * GLOB_ALTDIRFUNC:
57 * Use alternately specified directory access functions.
58 * GLOB_TILDE:
59 * expand ~user/foo to the /home/dir/of/user/foo
60 * GLOB_BRACE:
61 * expand {1,2}{a,b} to 1a 1b 2a 2b
62 * gl_matchc:
63 * Number of matches in the current invocation of glob.
64 */
65
66#include <sys/param.h>
67#include <sys/stat.h>
68
69#include <ctype.h>
70#include <dirent.h>
71#include <errno.h>
72#include <glob.h>
73#include <pwd.h>
74#include <stdio.h>
75#include <stdlib.h>
76#include <string.h>
77#include <unistd.h>
78
79#include "collate.h"
80
81#define DOLLAR '$'
82#define DOT '.'
83#define EOS '\0'
84#define LBRACKET '['
85#define NOT '!'
86#define QUESTION '?'
87#define QUOTE '\\'
88#define RANGE '-'
89#define RBRACKET ']'
90#define SEP '/'
91#define STAR '*'
92#define TILDE '~'
93#define UNDERSCORE '_'
94#define LBRACE '{'
95#define RBRACE '}'
96#define SLASH '/'
97#define COMMA ','
98
99#ifndef DEBUG
100
101#define M_QUOTE 0x8000
102#define M_PROTECT 0x4000
103#define M_MASK 0xffff
104#define M_ASCII 0x00ff
105
106typedef u_short Char;
107
108#else
109
110#define M_QUOTE 0x80
111#define M_PROTECT 0x40
112#define M_MASK 0xff
113#define M_ASCII 0x7f
114
115typedef char Char;
116
117#endif
118
119
120#define CHAR(c) ((Char)((c)&M_ASCII))
121#define META(c) ((Char)((c)|M_QUOTE))
122#define M_ALL META('*')
123#define M_END META(']')
124#define M_NOT META('!')
125#define M_ONE META('?')
126#define M_RNG META('-')
127#define M_SET META('[')
128#define ismeta(c) (((c)&M_QUOTE) != 0)
129
130
131static int compare __P((const void *, const void *));
132static void g_Ctoc __P((const Char *, char *));
133static int g_lstat __P((Char *, struct stat *, glob_t *));
134static DIR *g_opendir __P((Char *, glob_t *));
135static Char *g_strchr __P((Char *, int));
136#ifdef notdef
137static Char *g_strcat __P((Char *, const Char *));
138#endif
139static int g_stat __P((Char *, struct stat *, glob_t *));
140static int glob0 __P((const Char *, glob_t *));
141static int glob1 __P((Char *, glob_t *));
142static int glob2 __P((Char *, Char *, Char *, glob_t *));
143static int glob3 __P((Char *, Char *, Char *, Char *, glob_t *));
144static int globextend __P((const Char *, glob_t *));
145static const Char * globtilde __P((const Char *, Char *, size_t, glob_t *));
146static int globexp1 __P((const Char *, glob_t *));
147static int globexp2 __P((const Char *, const Char *, glob_t *, int *));
148static int match __P((Char *, Char *, Char *));
149#ifdef DEBUG
150static void qprintf __P((const char *, Char *));
151#endif
152
153int
154glob(pattern, flags, errfunc, pglob)
155 const char *pattern;
156 int flags, (*errfunc) __P((const char *, int));
157 glob_t *pglob;
158{
159 const u_char *patnext;
160 int c;
161 Char *bufnext, *bufend, patbuf[MAXPATHLEN+1];
162
163 patnext = (u_char *) pattern;
164 if (!(flags & GLOB_APPEND)) {
165 pglob->gl_pathc = 0;
166 pglob->gl_pathv = NULL;
167 if (!(flags & GLOB_DOOFFS))
168 pglob->gl_offs = 0;
169 }
170 pglob->gl_flags = flags & ~GLOB_MAGCHAR;
171 pglob->gl_errfunc = errfunc;
172 pglob->gl_matchc = 0;
173
174 bufnext = patbuf;
175 bufend = bufnext + MAXPATHLEN;
176 if (flags & GLOB_QUOTE) {
177 /* Protect the quoted characters. */
178 while (bufnext < bufend && (c = *patnext++) != EOS)
179 if (c == QUOTE) {
180 if ((c = *patnext++) == EOS) {
181 c = QUOTE;
182 --patnext;
183 }
184 *bufnext++ = c | M_PROTECT;
185 }
186 else
187 *bufnext++ = c;
188 }
189 else
190 while (bufnext < bufend && (c = *patnext++) != EOS)
191 *bufnext++ = c;
192 *bufnext = EOS;
193
194 if (flags & GLOB_BRACE)
195 return globexp1(patbuf, pglob);
196 else
197 return glob0(patbuf, pglob);
198}
199
200/*
201 * Expand recursively a glob {} pattern. When there is no more expansion
202 * invoke the standard globbing routine to glob the rest of the magic
203 * characters
204 */
205static int globexp1(pattern, pglob)
206 const Char *pattern;
207 glob_t *pglob;
208{
209 const Char* ptr = pattern;
210 int rv;
211
212 /* Protect a single {}, for find(1), like csh */
213 if (pattern[0] == LBRACE && pattern[1] == RBRACE && pattern[2] == EOS)
214 return glob0(pattern, pglob);
215
216 while ((ptr = (const Char *) g_strchr((Char *) ptr, LBRACE)) != NULL)
217 if (!globexp2(ptr, pattern, pglob, &rv))
218 return rv;
219
220 return glob0(pattern, pglob);
221}
222
223
224/*
225 * Recursive brace globbing helper. Tries to expand a single brace.
226 * If it succeeds then it invokes globexp1 with the new pattern.
227 * If it fails then it tries to glob the rest of the pattern and returns.
228 */
229static int globexp2(ptr, pattern, pglob, rv)
230 const Char *ptr, *pattern;
231 glob_t *pglob;
232 int *rv;
233{
234 int i;
235 Char *lm, *ls;
236 const Char *pe, *pm, *pl;
237 Char patbuf[MAXPATHLEN + 1];
238
239 /* copy part up to the brace */
240 for (lm = patbuf, pm = pattern; pm != ptr; *lm++ = *pm++)
241 continue;
242 ls = lm;
243
244 /* Find the balanced brace */
245 for (i = 0, pe = ++ptr; *pe; pe++)
246 if (*pe == LBRACKET) {
247 /* Ignore everything between [] */
248 for (pm = pe++; *pe != RBRACKET && *pe != EOS; pe++)
249 continue;
250 if (*pe == EOS) {
251 /*
252 * We could not find a matching RBRACKET.
253 * Ignore and just look for RBRACE
254 */
255 pe = pm;
256 }
257 }
258 else if (*pe == LBRACE)
259 i++;
260 else if (*pe == RBRACE) {
261 if (i == 0)
262 break;
263 i--;
264 }
265
266 /* Non matching braces; just glob the pattern */
267 if (i != 0 || *pe == EOS) {
268 *rv = glob0(patbuf, pglob);
269 return 0;
270 }
271
272 for (i = 0, pl = pm = ptr; pm <= pe; pm++)
273 switch (*pm) {
274 case LBRACKET:
275 /* Ignore everything between [] */
276 for (pl = pm++; *pm != RBRACKET && *pm != EOS; pm++)
277 continue;
278 if (*pm == EOS) {
279 /*
280 * We could not find a matching RBRACKET.
281 * Ignore and just look for RBRACE
282 */
283 pm = pl;
284 }
285 break;
286
287 case LBRACE:
288 i++;
289 break;
290
291 case RBRACE:
292 if (i) {
293 i--;
294 break;
295 }
296 /* FALLTHROUGH */
297 case COMMA:
298 if (i && *pm == COMMA)
299 break;
300 else {
301 /* Append the current string */
302 for (lm = ls; (pl < pm); *lm++ = *pl++)
303 continue;
304 /*
305 * Append the rest of the pattern after the
306 * closing brace
307 */
308 for (pl = pe + 1; (*lm++ = *pl++) != EOS;)
309 continue;
310
311 /* Expand the current pattern */
312#ifdef DEBUG
313 qprintf("globexp2:", patbuf);
314#endif
315 *rv = globexp1(patbuf, pglob);
316
317 /* move after the comma, to the next string */
318 pl = pm + 1;
319 }
320 break;
321
322 default:
323 break;
324 }
325 *rv = 0;
326 return 0;
327}
328
329
330
331/*
332 * expand tilde from the passwd file.
333 */
334static const Char *
335globtilde(pattern, patbuf, patbuf_len, pglob)
336 const Char *pattern;
337 Char *patbuf;
338 size_t patbuf_len;
339 glob_t *pglob;
340{
341 struct passwd *pwd;
342 char *h;
343 const Char *p;
344 Char *b, *eb;
345
346 if (*pattern != TILDE || !(pglob->gl_flags & GLOB_TILDE))
347 return pattern;
348
349 /*
350 * Copy up to the end of the string or /
351 */
352 eb = &patbuf[patbuf_len - 1];
353 for (p = pattern + 1, h = (char *) patbuf;
354 h < (char *)eb && *p && *p != SLASH; *h++ = *p++)
355 continue;
356
357 *h = EOS;
358
359 if (((char *) patbuf)[0] == EOS) {
360 /*
361 * handle a plain ~ or ~/ by expanding $HOME
362 * first and then trying the password file
363 */
364 if ((h = getenv("HOME")) == NULL) {
365 if ((pwd = getpwuid(getuid())) == NULL)
366 return pattern;
367 else
368 h = pwd->pw_dir;
369 }
370 }
371 else {
372 /*
373 * Expand a ~user
374 */
375 if ((pwd = getpwnam((char*) patbuf)) == NULL)
376 return pattern;
377 else
378 h = pwd->pw_dir;
379 }
380
381 /* Copy the home directory */
382 for (b = patbuf; b < eb && *h; *b++ = *h++)
383 continue;
384
385 /* Append the rest of the pattern */
386 while (b < eb && (*b++ = *p++) != EOS)
387 continue;
388 *b = EOS;
389
390 return patbuf;
391}
392
393
394/*
395 * The main glob() routine: compiles the pattern (optionally processing
396 * quotes), calls glob1() to do the real pattern matching, and finally
397 * sorts the list (unless unsorted operation is requested). Returns 0
398 * if things went well, nonzero if errors occurred. It is not an error
399 * to find no matches.
400 */
401static int
402glob0(pattern, pglob)
403 const Char *pattern;
404 glob_t *pglob;
405{
406 const Char *qpatnext;
407 int c, err, oldpathc;
408 Char *bufnext, patbuf[MAXPATHLEN+1];
409
410 qpatnext = globtilde(pattern, patbuf, sizeof(patbuf) / sizeof(Char),
411 pglob);
412 oldpathc = pglob->gl_pathc;
413 bufnext = patbuf;
414
415 /* We don't need to check for buffer overflow any more. */
416 while ((c = *qpatnext++) != EOS) {
417 switch (c) {
418 case LBRACKET:
419 c = *qpatnext;
420 if (c == NOT)
421 ++qpatnext;
422 if (*qpatnext == EOS ||
423 g_strchr((Char *) qpatnext+1, RBRACKET) == NULL) {
424 *bufnext++ = LBRACKET;
425 if (c == NOT)
426 --qpatnext;
427 break;
428 }
429 *bufnext++ = M_SET;
430 if (c == NOT)
431 *bufnext++ = M_NOT;
432 c = *qpatnext++;
433 do {
434 *bufnext++ = CHAR(c);
435 if (*qpatnext == RANGE &&
436 (c = qpatnext[1]) != RBRACKET) {
437 *bufnext++ = M_RNG;
438 *bufnext++ = CHAR(c);
439 qpatnext += 2;
440 }
441 } while ((c = *qpatnext++) != RBRACKET);
442 pglob->gl_flags |= GLOB_MAGCHAR;
443 *bufnext++ = M_END;
444 break;
445 case QUESTION:
446 pglob->gl_flags |= GLOB_MAGCHAR;
447 *bufnext++ = M_ONE;
448 break;
449 case STAR:
450 pglob->gl_flags |= GLOB_MAGCHAR;
451 /* collapse adjacent stars to one,
452 * to avoid exponential behavior
453 */
454 if (bufnext == patbuf || bufnext[-1] != M_ALL)
455 *bufnext++ = M_ALL;
456 break;
457 default:
458 *bufnext++ = CHAR(c);
459 break;
460 }
461 }
462 *bufnext = EOS;
463#ifdef DEBUG
464 qprintf("glob0:", patbuf);
465#endif
466
467 if ((err = glob1(patbuf, pglob)) != 0)
468 return(err);
469
470 /*
471 * If there was no match we are going to append the pattern
472 * if GLOB_NOCHECK was specified or if GLOB_NOMAGIC was specified
473 * and the pattern did not contain any magic characters
474 * GLOB_NOMAGIC is there just for compatibility with csh.
475 */
476 if (pglob->gl_pathc == oldpathc &&
477 ((pglob->gl_flags & GLOB_NOCHECK) ||
478 ((pglob->gl_flags & GLOB_NOMAGIC) &&
479 !(pglob->gl_flags & GLOB_MAGCHAR))))
480 return(globextend(pattern, pglob));
481 else if (!(pglob->gl_flags & GLOB_NOSORT))
482 qsort(pglob->gl_pathv + pglob->gl_offs + oldpathc,
483 pglob->gl_pathc - oldpathc, sizeof(char *), compare);
484 return(0);
485}
486
487static int
488compare(p, q)
489 const void *p, *q;
490{
491 return(strcmp(*(char **)p, *(char **)q));
492}
493
494static int
495glob1(pattern, pglob)
496 Char *pattern;
497 glob_t *pglob;
498{
499 Char pathbuf[MAXPATHLEN+1];
500
501 /* A null pathname is invalid -- POSIX 1003.1 sect. 2.4. */
502 if (*pattern == EOS)
503 return(0);
504 return(glob2(pathbuf, pathbuf, pattern, pglob));
505}
506
507/*
508 * The functions glob2 and glob3 are mutually recursive; there is one level
509 * of recursion for each segment in the pattern that contains one or more
510 * meta characters.
511 */
512static int
513glob2(pathbuf, pathend, pattern, pglob)
514 Char *pathbuf, *pathend, *pattern;
515 glob_t *pglob;
516{
517 struct stat sb;
518 Char *p, *q;
519 int anymeta;
520
521 /*
522 * Loop over pattern segments until end of pattern or until
523 * segment with meta character found.
524 */
525 for (anymeta = 0;;) {
526 if (*pattern == EOS) { /* End of pattern? */
527 *pathend = EOS;
528 if (g_lstat(pathbuf, &sb, pglob))
529 return(0);
530
531 if (((pglob->gl_flags & GLOB_MARK) &&
532 pathend[-1] != SEP) && (S_ISDIR(sb.st_mode)
533 || (S_ISLNK(sb.st_mode) &&
534 (g_stat(pathbuf, &sb, pglob) == 0) &&
535 S_ISDIR(sb.st_mode)))) {
536 *pathend++ = SEP;
537 *pathend = EOS;
538 }
539 ++pglob->gl_matchc;
540 return(globextend(pathbuf, pglob));
541 }
542
543 /* Find end of next segment, copy tentatively to pathend. */
544 q = pathend;
545 p = pattern;
546 while (*p != EOS && *p != SEP) {
547 if (ismeta(*p))
548 anymeta = 1;
549 *q++ = *p++;
550 }
551
552 if (!anymeta) { /* No expansion, do next segment. */
553 pathend = q;
554 pattern = p;
555 while (*pattern == SEP)
556 *pathend++ = *pattern++;
557 } else /* Need expansion, recurse. */
558 return(glob3(pathbuf, pathend, pattern, p, pglob));
559 }
560 /* NOTREACHED */
561}
562
563static int
564glob3(pathbuf, pathend, pattern, restpattern, pglob)
565 Char *pathbuf, *pathend, *pattern, *restpattern;
566 glob_t *pglob;
567{
568 register struct dirent *dp;
569 DIR *dirp;
570 int err;
571 char buf[MAXPATHLEN];
572
573 /*
574 * The readdirfunc declaration can't be prototyped, because it is
575 * assigned, below, to two functions which are prototyped in glob.h
576 * and dirent.h as taking pointers to differently typed opaque
577 * structures.
578 */
579 struct dirent *(*readdirfunc)();
580
581 *pathend = EOS;
582 errno = 0;
583
584 if ((dirp = g_opendir(pathbuf, pglob)) == NULL) {
585 /* TODO: don't call for ENOENT or ENOTDIR? */
586 if (pglob->gl_errfunc) {
587 g_Ctoc(pathbuf, buf);
588 if (pglob->gl_errfunc(buf, errno) ||
589 pglob->gl_flags & GLOB_ERR)
590 return (GLOB_ABEND);
591 }
592 return(0);
593 }
594
595 err = 0;
596
597 /* Search directory for matching names. */
598 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
599 readdirfunc = pglob->gl_readdir;
600 else
601 readdirfunc = readdir;
602 while ((dp = (*readdirfunc)(dirp))) {
603 register u_char *sc;
604 register Char *dc;
605
606 /* Initial DOT must be matched literally. */
607 if (dp->d_name[0] == DOT && *pattern != DOT)
608 continue;
609 for (sc = (u_char *) dp->d_name, dc = pathend;
610 (*dc++ = *sc++) != EOS;)
611 continue;
612 if (!match(pathend, pattern, restpattern)) {
613 *pathend = EOS;
614 continue;
615 }
616 err = glob2(pathbuf, --dc, restpattern, pglob);
617 if (err)
618 break;
619 }
620
621 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
622 (*pglob->gl_closedir)(dirp);
623 else
624 closedir(dirp);
625 return(err);
626}
627
628
629/*
630 * Extend the gl_pathv member of a glob_t structure to accomodate a new item,
631 * add the new item, and update gl_pathc.
632 *
633 * This assumes the BSD realloc, which only copies the block when its size
634 * crosses a power-of-two boundary; for v7 realloc, this would cause quadratic
635 * behavior.
636 *
637 * Return 0 if new item added, error code if memory couldn't be allocated.
638 *
639 * Invariant of the glob_t structure:
640 * Either gl_pathc is zero and gl_pathv is NULL; or gl_pathc > 0 and
641 * gl_pathv points to (gl_offs + gl_pathc + 1) items.
642 */
643static int
644globextend(path, pglob)
645 const Char *path;
646 glob_t *pglob;
647{
648 register char **pathv;
649 register int i;
650 u_int newsize;
651 char *copy;
652 const Char *p;
653
654 newsize = sizeof(*pathv) * (2 + pglob->gl_pathc + pglob->gl_offs);
655 pathv = pglob->gl_pathv ?
656 realloc((char *)pglob->gl_pathv, newsize) :
657 malloc(newsize);
658 if (pathv == NULL)
659 return(GLOB_NOSPACE);
660
661 if (pglob->gl_pathv == NULL && pglob->gl_offs > 0) {
662 /* first time around -- clear initial gl_offs items */
663 pathv += pglob->gl_offs;
664 for (i = pglob->gl_offs; --i >= 0; )
665 *--pathv = NULL;
666 }
667 pglob->gl_pathv = pathv;
668
669 for (p = path; *p++;)
670 continue;
671 if ((copy = malloc(p - path)) != NULL) {
672 g_Ctoc(path, copy);
673 pathv[pglob->gl_offs + pglob->gl_pathc++] = copy;
674 }
675 pathv[pglob->gl_offs + pglob->gl_pathc] = NULL;
676 return(copy == NULL ? GLOB_NOSPACE : 0);
677}
678
679/*
680 * pattern matching function for filenames. Each occurrence of the *
681 * pattern causes a recursion level.
682 */
683static int
684match(name, pat, patend)
685 register Char *name, *pat, *patend;
686{
687 int ok, negate_range;
688 Char c, k;
689
690 while (pat < patend) {
691 c = *pat++;
692 switch (c & M_MASK) {
693 case M_ALL:
694 if (pat == patend)
695 return(1);
696 do
697 if (match(name, pat, patend))
698 return(1);
699 while (*name++ != EOS);
700 return(0);
701 case M_ONE:
702 if (*name++ == EOS)
703 return(0);
704 break;
705 case M_SET:
706 ok = 0;
707 if ((k = *name++) == EOS)
708 return(0);
709 if ((negate_range = ((*pat & M_MASK) == M_NOT)) != EOS)
710 ++pat;
711 while (((c = *pat++) & M_MASK) != M_END)
712 if ((*pat & M_MASK) == M_RNG) {
|