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 * 4. Neither the name of the University nor the names of its contributors
17 * may be used to endorse or promote products derived from this software
18 * without specific prior written permission.
20 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
21 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
22 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
23 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
24 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
25 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
26 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
27 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
28 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
29 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
32 * @(#)glob.c 8.3 (Berkeley) 10/13/93
33 * $FreeBSD: src/lib/libc/gen/glob.c,v 1.27 2008/06/26 07:12:35 mtm Exp $
34 * $DragonFly: src/lib/libc/gen/glob.c,v 1.6 2005/12/07 02:28:15 corecode Exp $
38 * glob(3) -- a superset of the one defined in POSIX 1003.2.
40 * The [!...] convention to negate a range is supported (SysV, Posix, ksh).
42 * Optional extra services, controlled by flags not defined by POSIX:
45 * Escaping convention: \ inhibits any special meaning the following
46 * character might have (except \ at end of string is retained).
48 * Set in gl_flags if pattern contained a globbing character.
50 * Same as GLOB_NOCHECK, but it will only append pattern if it did
51 * not contain any magic characters. [Used in csh style globbing]
53 * Use alternately specified directory access functions.
55 * expand ~user/foo to the /home/dir/of/user/foo
57 * expand {1,2}{a,b} to 1a 1b 2a 2b
59 * Number of matches in the current invocation of glob.
63 * Some notes on multibyte character support:
64 * 1. Patterns with illegal byte sequences match nothing - even if
65 * GLOB_NOCHECK is specified.
66 * 2. Illegal byte sequences in filenames are handled by treating them as
67 * single-byte characters with a value of the first byte of the sequence
69 * 3. State-dependent encodings are not currently supported.
72 #include <sys/param.h>
102 #define UNDERSCORE '_'
110 #define M_QUOTE 0x8000000000ULL
111 #define M_PROTECT 0x4000000000ULL
112 #define M_MASK 0xffffffffffULL
113 #define M_CHAR 0x00ffffffffULL
115 typedef uint_fast64_t Char;
120 #define M_PROTECT 0x40
129 #define CHAR(c) ((Char)((c)&M_CHAR))
130 #define META(c) ((Char)((c)|M_QUOTE))
131 #define M_ALL META('*')
132 #define M_END META(']')
133 #define M_NOT META('!')
134 #define M_ONE META('?')
135 #define M_RNG META('-')
136 #define M_SET META('[')
137 #define ismeta(c) (((c)&M_QUOTE) != 0)
140 static int compare(const void *, const void *);
141 static int g_Ctoc(const Char *, char *, size_t);
142 static int g_lstat(const Char *, struct stat *, glob_t *);
143 static DIR *g_opendir(const Char *, glob_t *);
144 static const Char *g_strchr(const Char *, wchar_t);
146 static Char *g_strcat(Char *, const Char *);
148 static int g_stat(const Char *, struct stat *, glob_t *);
149 static int glob0(const Char *, glob_t *, size_t *);
150 static int glob1(const Char *, glob_t *, size_t *);
151 static int glob2(Char *, Char *, Char *, const Char *, glob_t *,
153 static int glob3(Char *, Char *, Char *, const Char *, const Char *,
155 static int globextend(const Char *, glob_t *, size_t *);
157 globtilde(const Char *, Char *, size_t, glob_t *);
158 static int globexp1(const Char *, glob_t *, size_t *);
159 static int globexp2(const Char *, const Char *, glob_t *, int *, size_t *);
160 static int match(const Char *, const Char *, const Char *);
162 static void qprintf(const char *, const Char *);
166 glob(const char *pattern, int flags, int (*errfunc)(const char *, int),
171 Char *bufnext, *bufend, patbuf[MAXPATHLEN], prot;
177 if (!(flags & GLOB_APPEND)) {
179 pglob->gl_pathv = NULL;
180 if (!(flags & GLOB_DOOFFS))
183 if (flags & GLOB_LIMIT) {
184 limit = pglob->gl_matchc;
189 pglob->gl_flags = flags & ~GLOB_MAGCHAR;
190 pglob->gl_errfunc = errfunc;
191 pglob->gl_matchc = 0;
194 bufend = bufnext + MAXPATHLEN - 1;
195 if (flags & GLOB_NOESCAPE) {
196 memset(&mbs, 0, sizeof(mbs));
197 while (bufend - bufnext >= MB_CUR_MAX) {
198 clen = mbrtowc(&wc, patnext, MB_LEN_MAX, &mbs);
199 if (clen == (size_t)-1 || clen == (size_t)-2)
200 return (GLOB_NOMATCH);
207 /* Protect the quoted characters. */
208 memset(&mbs, 0, sizeof(mbs));
209 while (bufend - bufnext >= MB_CUR_MAX) {
210 if (*patnext == QUOTE) {
211 if (*++patnext == EOS) {
212 *bufnext++ = QUOTE | M_PROTECT;
218 clen = mbrtowc(&wc, patnext, MB_LEN_MAX, &mbs);
219 if (clen == (size_t)-1 || clen == (size_t)-2)
220 return (GLOB_NOMATCH);
223 *bufnext++ = wc | prot;
229 if (flags & GLOB_BRACE)
230 return globexp1(patbuf, pglob, &limit);
232 return glob0(patbuf, pglob, &limit);
236 * Expand recursively a glob {} pattern. When there is no more expansion
237 * invoke the standard globbing routine to glob the rest of the magic
241 globexp1(const Char *pattern, glob_t *pglob, size_t *limit)
243 const Char *ptr = pattern;
246 /* Protect a single {}, for find(1), like csh */
247 if (pattern[0] == LBRACE && pattern[1] == RBRACE && pattern[2] == EOS)
248 return glob0(pattern, pglob, limit);
250 while ((ptr = g_strchr(ptr, LBRACE)) != NULL)
251 if (!globexp2(ptr, pattern, pglob, &rv, limit))
254 return glob0(pattern, pglob, limit);
259 * Recursive brace globbing helper. Tries to expand a single brace.
260 * If it succeeds then it invokes globexp1 with the new pattern.
261 * If it fails then it tries to glob the rest of the pattern and returns.
264 globexp2(const Char *ptr, const Char *pattern, glob_t *pglob, int *rv,
269 const Char *pe, *pm, *pm1, *pl;
270 Char patbuf[MAXPATHLEN];
272 /* copy part up to the brace */
273 for (lm = patbuf, pm = pattern; pm != ptr; *lm++ = *pm++)
278 /* Find the balanced brace */
279 for (i = 0, pe = ++ptr; *pe; pe++)
280 if (*pe == LBRACKET) {
281 /* Ignore everything between [] */
282 for (pm = pe++; *pe != RBRACKET && *pe != EOS; pe++)
286 * We could not find a matching RBRACKET.
287 * Ignore and just look for RBRACE
292 else if (*pe == LBRACE)
294 else if (*pe == RBRACE) {
300 /* Non matching braces; just glob the pattern */
301 if (i != 0 || *pe == EOS) {
302 *rv = glob0(patbuf, pglob, limit);
306 for (i = 0, pl = pm = ptr; pm <= pe; pm++)
309 /* Ignore everything between [] */
310 for (pm1 = pm++; *pm != RBRACKET && *pm != EOS; pm++)
314 * We could not find a matching RBRACKET.
315 * Ignore and just look for RBRACE
332 if (i && *pm == COMMA)
335 /* Append the current string */
336 for (lm = ls; (pl < pm); *lm++ = *pl++)
339 * Append the rest of the pattern after the
342 for (pl = pe + 1; (*lm++ = *pl++) != EOS;)
345 /* Expand the current pattern */
347 qprintf("globexp2:", patbuf);
349 *rv = globexp1(patbuf, pglob, limit);
351 /* move after the comma, to the next string */
366 * expand tilde from the passwd file.
369 globtilde(const Char *pattern, Char *patbuf, size_t patbuf_len, glob_t *pglob)
376 if (*pattern != TILDE || !(pglob->gl_flags & GLOB_TILDE))
380 * Copy up to the end of the string or /
382 eb = &patbuf[patbuf_len - 1];
383 for (p = pattern + 1, h = (char *) patbuf;
384 h < (char *)eb && *p && *p != SLASH; *h++ = *p++)
389 if (((char *) patbuf)[0] == EOS) {
391 * handle a plain ~ or ~/ by expanding $HOME first (iff
392 * we're not running setuid or setgid) and then trying
395 if (issetugid() != 0 ||
396 (h = getenv("HOME")) == NULL) {
397 if (((h = getlogin()) != NULL &&
398 (pwd = getpwnam(h)) != NULL) ||
399 (pwd = getpwuid(getuid())) != NULL)
409 if ((pwd = getpwnam((char*) patbuf)) == NULL)
415 /* Copy the home directory */
416 for (b = patbuf; b < eb && *h; *b++ = *h++)
419 /* Append the rest of the pattern */
420 while (b < eb && (*b++ = *p++) != EOS)
429 * The main glob() routine: compiles the pattern (optionally processing
430 * quotes), calls glob1() to do the real pattern matching, and finally
431 * sorts the list (unless unsorted operation is requested). Returns 0
432 * if things went well, nonzero if errors occurred.
435 glob0(const Char *pattern, glob_t *pglob, size_t *limit)
437 const Char *qpatnext;
440 Char *bufnext, patbuf[MAXPATHLEN];
442 qpatnext = globtilde(pattern, patbuf, MAXPATHLEN, pglob);
443 oldpathc = pglob->gl_pathc;
446 /* We don't need to check for buffer overflow any more. */
447 while ((c = *qpatnext++) != EOS) {
453 if (*qpatnext == EOS ||
454 g_strchr(qpatnext+1, RBRACKET) == NULL) {
455 *bufnext++ = LBRACKET;
465 *bufnext++ = CHAR(c);
466 if (*qpatnext == RANGE &&
467 (c = qpatnext[1]) != RBRACKET) {
469 *bufnext++ = CHAR(c);
472 } while ((c = *qpatnext++) != RBRACKET);
473 pglob->gl_flags |= GLOB_MAGCHAR;
477 pglob->gl_flags |= GLOB_MAGCHAR;
481 pglob->gl_flags |= GLOB_MAGCHAR;
482 /* collapse adjacent stars to one,
483 * to avoid exponential behavior
485 if (bufnext == patbuf || bufnext[-1] != M_ALL)
489 *bufnext++ = CHAR(c);
495 qprintf("glob0:", patbuf);
498 if ((err = glob1(patbuf, pglob, limit)) != 0)
502 * If there was no match we are going to append the pattern
503 * if GLOB_NOCHECK was specified or if GLOB_NOMAGIC was specified
504 * and the pattern did not contain any magic characters
505 * GLOB_NOMAGIC is there just for compatibility with csh.
507 if (pglob->gl_pathc == oldpathc) {
508 if (((pglob->gl_flags & GLOB_NOCHECK) ||
509 ((pglob->gl_flags & GLOB_NOMAGIC) &&
510 !(pglob->gl_flags & GLOB_MAGCHAR))))
511 return(globextend(pattern, pglob, limit));
513 return(GLOB_NOMATCH);
515 if (!(pglob->gl_flags & GLOB_NOSORT))
516 qsort(pglob->gl_pathv + pglob->gl_offs + oldpathc,
517 pglob->gl_pathc - oldpathc, sizeof(char *), compare);
522 compare(const void *p, const void *q)
524 return(strcmp(*(const char * const *)p, *(const char * const *)q));
528 glob1(const Char *pattern, glob_t *pglob, size_t *limit)
530 Char pathbuf[MAXPATHLEN];
532 /* A null pathname is invalid -- POSIX 1003.1 sect. 2.4. */
535 return(glob2(pathbuf, pathbuf, pathbuf + MAXPATHLEN - 1,
536 pattern, pglob, limit));
540 * The functions glob2 and glob3 are mutually recursive; there is one level
541 * of recursion for each segment in the pattern that contains one or more
545 glob2(Char *pathbuf, Char *pathend, Char *pathend_last, const Char *pattern,
546 glob_t *pglob, size_t *limit)
554 * Loop over pattern segments until end of pattern or until
555 * segment with meta character found.
557 for (anymeta = 0;;) {
558 if (*pattern == EOS) { /* End of pattern? */
560 if (g_lstat(pathbuf, &sb, pglob))
563 if (((pglob->gl_flags & GLOB_MARK) &&
564 pathend[-1] != SEP) && (S_ISDIR(sb.st_mode)
565 || (S_ISLNK(sb.st_mode) &&
566 (g_stat(pathbuf, &sb, pglob) == 0) &&
567 S_ISDIR(sb.st_mode)))) {
568 if (pathend + 1 > pathend_last)
569 return (GLOB_ABORTED);
574 return(globextend(pathbuf, pglob, limit));
577 /* Find end of next segment, copy tentatively to pathend. */
580 while (*p != EOS && *p != SEP) {
583 if (q + 1 > pathend_last)
584 return (GLOB_ABORTED);
588 if (!anymeta) { /* No expansion, do next segment. */
591 while (*pattern == SEP) {
592 if (pathend + 1 > pathend_last)
593 return (GLOB_ABORTED);
594 *pathend++ = *pattern++;
596 } else /* Need expansion, recurse. */
597 return(glob3(pathbuf, pathend, pathend_last, pattern, p,
604 glob3(Char *pathbuf, Char *pathend, Char *pathend_last, const Char *pattern,
605 const Char *restpattern, glob_t *pglob, size_t *limit)
610 char buf[MAXPATHLEN];
611 struct dirent *(*readdirfunc)(DIR *);
613 if (pathend > pathend_last)
614 return (GLOB_ABORTED);
618 if ((dirp = g_opendir(pathbuf, pglob)) == NULL) {
619 /* TODO: don't call for ENOENT or ENOTDIR? */
620 if (pglob->gl_errfunc) {
621 if (g_Ctoc(pathbuf, buf, sizeof(buf)))
622 return (GLOB_ABORTED);
623 if (pglob->gl_errfunc(buf, errno) ||
624 pglob->gl_flags & GLOB_ERR)
625 return (GLOB_ABORTED);
632 /* pglob->gl_readdir takes a void *, fix this manually. */
633 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
634 readdirfunc = (struct dirent *(*)(DIR *))pglob->gl_readdir;
636 readdirfunc = readdir;
638 /* Search directory for matching names. */
639 while ((dp = (*readdirfunc)(dirp)) != NULL) {
646 /* Initial DOT must be matched literally. */
647 if (dp->d_name[0] == DOT && *pattern != DOT &&
648 !(pglob->gl_flags & GLOB_PERIOD))
650 memset(&mbs, 0, sizeof(mbs));
653 while (dc < pathend_last) {
654 clen = mbrtowc(&wc, sc, MB_LEN_MAX, &mbs);
655 if (clen == (size_t)-1 || clen == (size_t)-2) {
658 memset(&mbs, 0, sizeof(mbs));
660 if ((*dc++ = wc) == EOS)
664 if (!match(pathend, pattern, restpattern)) {
668 err = glob2(pathbuf, --dc, pathend_last, restpattern,
674 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
675 (*pglob->gl_closedir)(dirp);
683 * Extend the gl_pathv member of a glob_t structure to accomodate a new item,
684 * add the new item, and update gl_pathc.
686 * This assumes the BSD realloc, which only copies the block when its size
687 * crosses a power-of-two boundary; for v7 realloc, this would cause quadratic
690 * Return 0 if new item added, error code if memory couldn't be allocated.
692 * Invariant of the glob_t structure:
693 * Either gl_pathc is zero and gl_pathv is NULL; or gl_pathc > 0 and
694 * gl_pathv points to (gl_offs + gl_pathc + 1) items.
697 globextend(const Char *path, glob_t *pglob, size_t *limit)
700 size_t i, newsize, len;
704 if (*limit && pglob->gl_pathc > *limit) {
706 return (GLOB_NOSPACE);
709 newsize = sizeof(*pathv) * (2 + pglob->gl_pathc + pglob->gl_offs);
710 pathv = pglob->gl_pathv ?
711 realloc((char *)pglob->gl_pathv, newsize) :
714 if (pglob->gl_pathv) {
715 free(pglob->gl_pathv);
716 pglob->gl_pathv = NULL;
718 return(GLOB_NOSPACE);
721 if (pglob->gl_pathv == NULL && pglob->gl_offs > 0) {
722 /* first time around -- clear initial gl_offs items */
723 pathv += pglob->gl_offs;
724 for (i = pglob->gl_offs + 1; --i > 0; )
727 pglob->gl_pathv = pathv;
729 for (p = path; *p++;)
731 len = MB_CUR_MAX * (size_t)(p - path); /* XXX overallocation */
732 if ((copy = malloc(len)) != NULL) {
733 if (g_Ctoc(path, copy, len)) {
735 return (GLOB_NOSPACE);
737 pathv[pglob->gl_offs + pglob->gl_pathc++] = copy;
739 pathv[pglob->gl_offs + pglob->gl_pathc] = NULL;
740 return(copy == NULL ? GLOB_NOSPACE : 0);
744 * pattern matching function for filenames. Each occurrence of the *
745 * pattern causes a recursion level.
748 match(const Char * name, const Char *pat, const Char *patend)
750 int ok, negate_range;
753 while (pat < patend) {
755 switch (c & M_MASK) {
760 if (match(name, pat, patend))
762 while (*name++ != EOS);
770 if ((k = *name++) == EOS)
772 if ((negate_range = ((*pat & M_MASK) == M_NOT)) != EOS)
774 while (((c = *pat++) & M_MASK) != M_END)
775 if ((*pat & M_MASK) == M_RNG) {
776 if (__collate_load_error ?
777 CHAR(c) <= CHAR(k) && CHAR(k) <= CHAR(pat[1]) :
778 __collate_range_cmp(CHAR(c), CHAR(k)) <= 0
779 && __collate_range_cmp(CHAR(k), CHAR(pat[1])) <= 0
785 if (ok == negate_range)
794 return(*name == EOS);
797 /* Free allocated data belonging to a glob_t structure. */
799 globfree(glob_t *pglob)
804 if (pglob->gl_pathv != NULL) {
805 pp = pglob->gl_pathv + pglob->gl_offs;
806 for (i = pglob->gl_pathc; i--; ++pp)
809 free(pglob->gl_pathv);
810 pglob->gl_pathv = NULL;
815 g_opendir(const Char *str, glob_t *pglob)
817 char buf[MAXPATHLEN];
821 else if (g_Ctoc(str, buf, sizeof(buf)))
824 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
825 return((*pglob->gl_opendir)(buf));
827 return(opendir(buf));
831 g_lstat(const Char *fn, struct stat *sb, glob_t *pglob)
833 char buf[MAXPATHLEN];
835 if (g_Ctoc(fn, buf, sizeof(buf))) {
836 errno = ENAMETOOLONG;
839 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
840 return((*pglob->gl_lstat)(buf, sb));
841 return(lstat(buf, sb));
845 g_stat(const Char *fn, struct stat *sb, glob_t *pglob)
847 char buf[MAXPATHLEN];
849 if (g_Ctoc(fn, buf, sizeof(buf))) {
850 errno = ENAMETOOLONG;
853 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
854 return((*pglob->gl_stat)(buf, sb));
855 return(stat(buf, sb));
859 g_strchr(const Char *str, wchar_t ch)
870 g_Ctoc(const Char *str, char *buf, size_t len)
875 memset(&mbs, 0, sizeof(mbs));
876 while (len >= MB_CUR_MAX) {
877 clen = wcrtomb(buf, *str, &mbs);
878 if (clen == (size_t)-1)
891 qprintf(const char *str, const Char *s)
895 printf("%s:\n", str);
897 printf("%c", CHAR(*p));
900 printf("%c", *p & M_PROTECT ? '"' : ' ');
903 printf("%c", ismeta(*p) ? '_' : ' ');