2 * Copyright (c) 1989, 1993
3 * The Regents of the University of California. All rights reserved.
5 * This code is derived from software contributed to Berkeley by
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
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.
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
36 * $FreeBSD: src/lib/libc/gen/glob.c,v 1.11.6.6 2002/09/18 14:13:31 mikeh Exp $
37 * $DragonFly: src/lib/libc/gen/glob.c,v 1.4 2005/04/26 08:07:58 joerg Exp $
39 * @(#)glob.c 8.3 (Berkeley) 10/13/93
40 * $FreeBSD: src/lib/libc/gen/glob.c,v 1.11.6.6 2002/09/18 14:13:31 mikeh Exp $
44 * glob(3) -- a superset of the one defined in POSIX 1003.2.
46 * The [!...] convention to negate a range is supported (SysV, Posix, ksh).
48 * Optional extra services, controlled by flags not defined by POSIX:
51 * Escaping convention: \ inhibits any special meaning the following
52 * character might have (except \ at end of string is retained).
54 * Set in gl_flags if pattern contained a globbing character.
56 * Same as GLOB_NOCHECK, but it will only append pattern if it did
57 * not contain any magic characters. [Used in csh style globbing]
59 * Use alternately specified directory access functions.
61 * expand ~user/foo to the /home/dir/of/user/foo
63 * expand {1,2}{a,b} to 1a 1b 2a 2b
65 * Number of matches in the current invocation of glob.
68 #include <sys/param.h>
95 #define UNDERSCORE '_'
103 #define M_QUOTE 0x8000
104 #define M_PROTECT 0x4000
105 #define M_MASK 0xffff
106 #define M_ASCII 0x00ff
108 typedef u_short Char;
113 #define M_PROTECT 0x40
122 #define CHAR(c) ((Char)((c)&M_ASCII))
123 #define META(c) ((Char)((c)|M_QUOTE))
124 #define M_ALL META('*')
125 #define M_END META(']')
126 #define M_NOT META('!')
127 #define M_ONE META('?')
128 #define M_RNG META('-')
129 #define M_SET META('[')
130 #define ismeta(c) (((c)&M_QUOTE) != 0)
133 static int compare(const void *, const void *);
134 static int g_Ctoc(const Char *, char *, size_t);
135 static int g_lstat(const Char *, struct stat *, glob_t *);
136 static DIR *g_opendir(const Char *, glob_t *);
138 static Char *g_strcat(Char *, const Char *);
141 *g_strchr(const Char *, int ch);
142 static int g_stat(const Char *, struct stat *, glob_t *);
143 static int glob0(const Char *, glob_t *, int *);
144 static int glob1(const Char *, glob_t *, int *);
145 static int glob2(Char *, Char *, Char *, const Char *, glob_t *, int *);
146 static int glob3(Char *, Char *, Char *, const Char *, const Char *, glob_t *, int *);
147 static int globextend(const Char *, glob_t *, int *);
149 globtilde(const Char *, Char *, size_t, glob_t *);
150 static int globexp1(const Char *, glob_t *, int *);
151 static int globexp2(const Char *, const Char *, glob_t *, int *, int *);
152 static int match(const Char *, const Char *, const Char *);
154 static void qprintf(const char *, const Char *);
158 glob(const char *pattern, int flags, int (*errfunc)(const char *, int),
161 const u_char *patnext;
163 Char *bufnext, *bufend, patbuf[MAXPATHLEN];
165 patnext = (const u_char *) pattern;
166 if (!(flags & GLOB_APPEND)) {
168 pglob->gl_pathv = NULL;
169 if (!(flags & GLOB_DOOFFS))
172 if (flags & GLOB_LIMIT) {
173 limit = pglob->gl_matchc;
178 pglob->gl_flags = flags & ~GLOB_MAGCHAR;
179 pglob->gl_errfunc = errfunc;
180 pglob->gl_matchc = 0;
183 bufend = bufnext + MAXPATHLEN - 1;
184 if (flags & GLOB_NOESCAPE)
185 while (bufnext < bufend && (c = *patnext++) != EOS)
188 /* Protect the quoted characters. */
189 while (bufnext < bufend && (c = *patnext++) != EOS)
191 if ((c = *patnext++) == EOS) {
195 *bufnext++ = c | M_PROTECT;
202 if (flags & GLOB_BRACE)
203 return globexp1(patbuf, pglob, &limit);
205 return glob0(patbuf, pglob, &limit);
209 * Expand recursively a glob {} pattern. When there is no more expansion
210 * invoke the standard globbing routine to glob the rest of the magic
214 globexp1(const Char *pattern, glob_t *pglob, int *limit)
216 const Char *ptr = pattern;
219 /* Protect a single {}, for find(1), like csh */
220 if (pattern[0] == LBRACE && pattern[1] == RBRACE && pattern[2] == EOS)
221 return glob0(pattern, pglob, limit);
223 while ((ptr = g_strchr(ptr, LBRACE)) != NULL)
224 if (!globexp2(ptr, pattern, pglob, &rv, limit))
227 return glob0(pattern, pglob, limit);
232 * Recursive brace globbing helper. Tries to expand a single brace.
233 * If it succeeds then it invokes globexp1 with the new pattern.
234 * If it fails then it tries to glob the rest of the pattern and returns.
237 globexp2(const Char *ptr, const Char *pattern, glob_t *pglob, int *rv,
242 const Char *pe, *pm, *pl;
243 Char patbuf[MAXPATHLEN];
245 /* copy part up to the brace */
246 for (lm = patbuf, pm = pattern; pm != ptr; *lm++ = *pm++)
251 /* Find the balanced brace */
252 for (i = 0, pe = ++ptr; *pe; pe++)
253 if (*pe == LBRACKET) {
254 /* Ignore everything between [] */
255 for (pm = pe++; *pe != RBRACKET && *pe != EOS; pe++)
259 * We could not find a matching RBRACKET.
260 * Ignore and just look for RBRACE
265 else if (*pe == LBRACE)
267 else if (*pe == RBRACE) {
273 /* Non matching braces; just glob the pattern */
274 if (i != 0 || *pe == EOS) {
275 *rv = glob0(patbuf, pglob, limit);
279 for (i = 0, pl = pm = ptr; pm <= pe; pm++)
282 /* Ignore everything between [] */
283 for (pl = pm++; *pm != RBRACKET && *pm != EOS; pm++)
287 * We could not find a matching RBRACKET.
288 * Ignore and just look for RBRACE
305 if (i && *pm == COMMA)
308 /* Append the current string */
309 for (lm = ls; (pl < pm); *lm++ = *pl++)
312 * Append the rest of the pattern after the
315 for (pl = pe + 1; (*lm++ = *pl++) != EOS;)
318 /* Expand the current pattern */
320 qprintf("globexp2:", patbuf);
322 *rv = globexp1(patbuf, pglob, limit);
324 /* move after the comma, to the next string */
339 * expand tilde from the passwd file.
342 globtilde(const Char *pattern, Char *patbuf, size_t patbuf_len, glob_t *pglob)
349 if (*pattern != TILDE || !(pglob->gl_flags & GLOB_TILDE))
353 * Copy up to the end of the string or /
355 eb = &patbuf[patbuf_len - 1];
356 for (p = pattern + 1, h = (char *) patbuf;
357 h < (char *)eb && *p && *p != SLASH; *h++ = *p++)
362 if (((char *) patbuf)[0] == EOS) {
364 * handle a plain ~ or ~/ by expanding $HOME first (iff
365 * we're not running setuid or setgid) and then trying
368 if (issetugid() != 0 ||
369 (h = getenv("HOME")) == NULL) {
370 if (((h = getlogin()) != NULL &&
371 (pwd = getpwnam(h)) != NULL) ||
372 (pwd = getpwuid(getuid())) != NULL)
382 if ((pwd = getpwnam((char*) patbuf)) == NULL)
388 /* Copy the home directory */
389 for (b = patbuf; b < eb && *h; *b++ = *h++)
392 /* Append the rest of the pattern */
393 while (b < eb && (*b++ = *p++) != EOS)
402 * The main glob() routine: compiles the pattern (optionally processing
403 * quotes), calls glob1() to do the real pattern matching, and finally
404 * sorts the list (unless unsorted operation is requested). Returns 0
405 * if things went well, nonzero if errors occurred.
408 glob0(const Char *pattern, glob_t *pglob, int *limit)
410 const Char *qpatnext;
411 int c, err, oldpathc;
412 Char *bufnext, patbuf[MAXPATHLEN];
414 qpatnext = globtilde(pattern, patbuf, MAXPATHLEN, pglob);
415 oldpathc = pglob->gl_pathc;
418 /* We don't need to check for buffer overflow any more. */
419 while ((c = *qpatnext++) != EOS) {
425 if (*qpatnext == EOS ||
426 g_strchr(qpatnext+1, RBRACKET) == NULL) {
427 *bufnext++ = LBRACKET;
437 *bufnext++ = CHAR(c);
438 if (*qpatnext == RANGE &&
439 (c = qpatnext[1]) != RBRACKET) {
441 *bufnext++ = CHAR(c);
444 } while ((c = *qpatnext++) != RBRACKET);
445 pglob->gl_flags |= GLOB_MAGCHAR;
449 pglob->gl_flags |= GLOB_MAGCHAR;
453 pglob->gl_flags |= GLOB_MAGCHAR;
454 /* collapse adjacent stars to one,
455 * to avoid exponential behavior
457 if (bufnext == patbuf || bufnext[-1] != M_ALL)
461 *bufnext++ = CHAR(c);
467 qprintf("glob0:", patbuf);
470 if ((err = glob1(patbuf, pglob, limit)) != 0)
474 * If there was no match we are going to append the pattern
475 * if GLOB_NOCHECK was specified or if GLOB_NOMAGIC was specified
476 * and the pattern did not contain any magic characters
477 * GLOB_NOMAGIC is there just for compatibility with csh.
479 if (pglob->gl_pathc == oldpathc) {
480 if (((pglob->gl_flags & GLOB_NOCHECK) ||
481 ((pglob->gl_flags & GLOB_NOMAGIC) &&
482 !(pglob->gl_flags & GLOB_MAGCHAR))))
483 return(globextend(pattern, pglob, limit));
485 return(GLOB_NOMATCH);
487 if (!(pglob->gl_flags & GLOB_NOSORT))
488 qsort(pglob->gl_pathv + pglob->gl_offs + oldpathc,
489 pglob->gl_pathc - oldpathc, sizeof(char *), compare);
494 compare(const void *p, const void *q)
496 return(strcmp(*(const char * const *)p, *(const char * const *)q));
500 glob1(const Char *pattern, glob_t *pglob, int *limit)
502 Char pathbuf[MAXPATHLEN];
504 /* A null pathname is invalid -- POSIX 1003.1 sect. 2.4. */
507 return(glob2(pathbuf, pathbuf, pathbuf + MAXPATHLEN - 1,
508 pattern, pglob, limit));
512 * The functions glob2 and glob3 are mutually recursive; there is one level
513 * of recursion for each segment in the pattern that contains one or more
517 glob2(Char *pathbuf, Char *pathend, Char *pathend_last, const Char *pattern,
518 glob_t *pglob, int *limit)
526 * Loop over pattern segments until end of pattern or until
527 * segment with meta character found.
529 for (anymeta = 0;;) {
530 if (*pattern == EOS) { /* End of pattern? */
532 if (g_lstat(pathbuf, &sb, pglob))
535 if (((pglob->gl_flags & GLOB_MARK) &&
536 pathend[-1] != SEP) && (S_ISDIR(sb.st_mode)
537 || (S_ISLNK(sb.st_mode) &&
538 (g_stat(pathbuf, &sb, pglob) == 0) &&
539 S_ISDIR(sb.st_mode)))) {
540 if (pathend + 1 > pathend_last)
541 return (GLOB_ABORTED);
546 return(globextend(pathbuf, pglob, limit));
549 /* Find end of next segment, copy tentatively to pathend. */
552 while (*p != EOS && *p != SEP) {
555 if (q + 1 > pathend_last)
556 return (GLOB_ABORTED);
560 if (!anymeta) { /* No expansion, do next segment. */
563 while (*pattern == SEP) {
564 if (pathend + 1 > pathend_last)
565 return (GLOB_ABORTED);
566 *pathend++ = *pattern++;
568 } else /* Need expansion, recurse. */
569 return(glob3(pathbuf, pathend, pathend_last, pattern, p,
576 glob3(Char *pathbuf, Char *pathend, Char *pathend_last, const Char *pattern,
577 const Char *restpattern, glob_t *pglob, int *limit)
582 char buf[MAXPATHLEN];
583 struct dirent *(*readdirfunc)(DIR *);
585 if (pathend > pathend_last)
586 return (GLOB_ABORTED);
590 if ((dirp = g_opendir(pathbuf, pglob)) == NULL) {
591 /* TODO: don't call for ENOENT or ENOTDIR? */
592 if (pglob->gl_errfunc) {
593 if (g_Ctoc(pathbuf, buf, sizeof(buf)))
594 return (GLOB_ABORTED);
595 if (pglob->gl_errfunc(buf, errno) ||
596 pglob->gl_flags & GLOB_ERR)
597 return (GLOB_ABORTED);
604 /* pglob->gl_readdir takes a void *, fix this manually. */
605 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
606 readdirfunc = (struct dirent *(*)(DIR *))pglob->gl_readdir;
608 readdirfunc = readdir;
610 /* Search directory for matching names. */
611 while ((dp = (*readdirfunc)(dirp)) != NULL) {
615 /* Initial DOT must be matched literally. */
616 if (dp->d_name[0] == DOT && *pattern != DOT)
619 sc = (u_char *) dp->d_name;
620 while (dc < pathend_last && (*dc++ = *sc++) != EOS)
622 if (!match(pathend, pattern, restpattern)) {
626 err = glob2(pathbuf, --dc, pathend_last, restpattern,
632 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
633 (*pglob->gl_closedir)(dirp);
641 * Extend the gl_pathv member of a glob_t structure to accomodate a new item,
642 * add the new item, and update gl_pathc.
644 * This assumes the BSD realloc, which only copies the block when its size
645 * crosses a power-of-two boundary; for v7 realloc, this would cause quadratic
648 * Return 0 if new item added, error code if memory couldn't be allocated.
650 * Invariant of the glob_t structure:
651 * Either gl_pathc is zero and gl_pathv is NULL; or gl_pathc > 0 and
652 * gl_pathv points to (gl_offs + gl_pathc + 1) items.
655 globextend(const Char *path, glob_t *pglob, int *limit)
663 if (*limit && pglob->gl_pathc > *limit) {
665 return (GLOB_NOSPACE);
668 newsize = sizeof(*pathv) * (2 + pglob->gl_pathc + pglob->gl_offs);
669 pathv = pglob->gl_pathv ?
670 realloc((char *)pglob->gl_pathv, newsize) :
673 if (pglob->gl_pathv) {
674 free(pglob->gl_pathv);
675 pglob->gl_pathv = NULL;
677 return(GLOB_NOSPACE);
680 if (pglob->gl_pathv == NULL && pglob->gl_offs > 0) {
681 /* first time around -- clear initial gl_offs items */
682 pathv += pglob->gl_offs;
683 for (i = pglob->gl_offs; --i >= 0; )
686 pglob->gl_pathv = pathv;
688 for (p = path; *p++;)
690 len = (size_t)(p - path);
691 if ((copy = malloc(len)) != NULL) {
692 if (g_Ctoc(path, copy, len)) {
694 return (GLOB_NOSPACE);
696 pathv[pglob->gl_offs + pglob->gl_pathc++] = copy;
698 pathv[pglob->gl_offs + pglob->gl_pathc] = NULL;
699 return(copy == NULL ? GLOB_NOSPACE : 0);
703 * pattern matching function for filenames. Each occurrence of the *
704 * pattern causes a recursion level.
707 match(const Char * name, const Char *pat, const Char *patend)
709 int ok, negate_range;
712 while (pat < patend) {
714 switch (c & M_MASK) {
719 if (match(name, pat, patend))
721 while (*name++ != EOS);
729 if ((k = *name++) == EOS)
731 if ((negate_range = ((*pat & M_MASK) == M_NOT)) != EOS)
733 while (((c = *pat++) & M_MASK) != M_END)
734 if ((*pat & M_MASK) == M_RNG) {
735 if (__collate_load_error ?
736 CHAR(c) <= CHAR(k) && CHAR(k) <= CHAR(pat[1]) :
737 __collate_range_cmp(CHAR(c), CHAR(k)) <= 0
738 && __collate_range_cmp(CHAR(k), CHAR(pat[1])) <= 0
744 if (ok == negate_range)
753 return(*name == EOS);
756 /* Free allocated data belonging to a glob_t structure. */
758 globfree(glob_t *pglob)
763 if (pglob->gl_pathv != NULL) {
764 pp = pglob->gl_pathv + pglob->gl_offs;
765 for (i = pglob->gl_pathc; i--; ++pp)
768 free(pglob->gl_pathv);
769 pglob->gl_pathv = NULL;
774 g_opendir(const Char *str, glob_t *pglob)
776 char buf[MAXPATHLEN];
780 else if (g_Ctoc(str, buf, sizeof(buf)))
783 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
784 return((*pglob->gl_opendir)(buf));
786 return(opendir(buf));
790 g_lstat(const Char *fn, struct stat *sb, glob_t *pglob)
792 char buf[MAXPATHLEN];
794 if (g_Ctoc(fn, buf, sizeof(buf))) {
795 errno = ENAMETOOLONG;
798 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
799 return((*pglob->gl_lstat)(buf, sb));
800 return(lstat(buf, sb));
804 g_stat(const Char *fn, struct stat *sb, glob_t *pglob)
806 char buf[MAXPATHLEN];
808 if (g_Ctoc(fn, buf, sizeof(buf))) {
809 errno = ENAMETOOLONG;
812 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
813 return((*pglob->gl_stat)(buf, sb));
814 return(stat(buf, sb));
818 g_strchr(const Char *str, int ch)
829 g_Ctoc(const Char *str, char *buf, size_t len)
833 if ((*buf++ = *str++) == '\0')
841 qprintf(const char *str, const Char *s)
845 (void)printf("%s:\n", str);
847 (void)printf("%c", CHAR(*p));
850 (void)printf("%c", *p & M_PROTECT ? '"' : ' ');
853 (void)printf("%c", ismeta(*p) ? '_' : ' ');