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
33 #if defined(LIBC_SCCS) && !defined(lint)
34 static char sccsid[] = "@(#)glob.c 8.3 (Berkeley) 10/13/93";
35 #endif /* LIBC_SCCS and not lint */
36 #include <sys/cdefs.h>
37 __FBSDID("$FreeBSD$");
40 * glob(3) -- a superset of the one defined in POSIX 1003.2.
42 * The [!...] convention to negate a range is supported (SysV, Posix, ksh).
44 * Optional extra services, controlled by flags not defined by POSIX:
47 * Escaping convention: \ inhibits any special meaning the following
48 * character might have (except \ at end of string is retained).
50 * Set in gl_flags if pattern contained a globbing character.
52 * Same as GLOB_NOCHECK, but it will only append pattern if it did
53 * not contain any magic characters. [Used in csh style globbing]
55 * Use alternately specified directory access functions.
57 * expand ~user/foo to the /home/dir/of/user/foo
59 * expand {1,2}{a,b} to 1a 1b 2a 2b
61 * Number of matches in the current invocation of glob.
65 * Some notes on multibyte character support:
66 * 1. Patterns with illegal byte sequences match nothing - even if
67 * GLOB_NOCHECK is specified.
68 * 2. Illegal byte sequences in filenames are handled by treating them as
69 * single-byte characters with a value of the first byte of the sequence
71 * 3. State-dependent encodings are not currently supported.
74 #include <sys/param.h>
104 #define UNDERSCORE '_'
112 #define M_QUOTE 0x8000000000ULL
113 #define M_PROTECT 0x4000000000ULL
114 #define M_MASK 0xffffffffffULL
115 #define M_CHAR 0x00ffffffffULL
117 typedef uint_fast64_t Char;
122 #define M_PROTECT 0x40
131 #define CHAR(c) ((Char)((c)&M_CHAR))
132 #define META(c) ((Char)((c)|M_QUOTE))
133 #define M_ALL META('*')
134 #define M_END META(']')
135 #define M_NOT META('!')
136 #define M_ONE META('?')
137 #define M_RNG META('-')
138 #define M_SET META('[')
139 #define ismeta(c) (((c)&M_QUOTE) != 0)
142 static int compare(const void *, const void *);
143 static int g_Ctoc(const Char *, char *, size_t);
144 static int g_lstat(Char *, struct stat *, glob_t *);
145 static DIR *g_opendir(Char *, glob_t *);
146 static Char *g_strchr(Char *, wchar_t);
148 static Char *g_strcat(Char *, const Char *);
150 static int g_stat(Char *, struct stat *, glob_t *);
151 static int glob0(const Char *, glob_t *, size_t *);
152 static int glob1(Char *, glob_t *, size_t *);
153 static int glob2(Char *, Char *, Char *, Char *, glob_t *, size_t *);
154 static int glob3(Char *, Char *, Char *, Char *, Char *, glob_t *, size_t *);
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(Char *, Char *, Char *);
162 static void qprintf(const char *, Char *);
166 glob(const char *pattern, int flags, int (*errfunc)(const char *, int), glob_t *pglob)
170 Char *bufnext, *bufend, patbuf[MAXPATHLEN], prot;
176 if (!(flags & GLOB_APPEND)) {
178 pglob->gl_pathv = NULL;
179 if (!(flags & GLOB_DOOFFS))
182 if (flags & GLOB_LIMIT) {
183 limit = pglob->gl_matchc;
188 pglob->gl_flags = flags & ~GLOB_MAGCHAR;
189 pglob->gl_errfunc = errfunc;
190 pglob->gl_matchc = 0;
193 bufend = bufnext + MAXPATHLEN - 1;
194 if (flags & GLOB_NOESCAPE) {
195 memset(&mbs, 0, sizeof(mbs));
196 while (bufend - bufnext >= MB_CUR_MAX) {
197 clen = mbrtowc(&wc, patnext, MB_LEN_MAX, &mbs);
198 if (clen == (size_t)-1 || clen == (size_t)-2)
199 return (GLOB_NOMATCH);
206 /* Protect the quoted characters. */
207 memset(&mbs, 0, sizeof(mbs));
208 while (bufend - bufnext >= MB_CUR_MAX) {
209 if (*patnext == QUOTE) {
210 if (*++patnext == EOS) {
211 *bufnext++ = QUOTE | M_PROTECT;
217 clen = mbrtowc(&wc, patnext, MB_LEN_MAX, &mbs);
218 if (clen == (size_t)-1 || clen == (size_t)-2)
219 return (GLOB_NOMATCH);
222 *bufnext++ = wc | prot;
228 if (flags & GLOB_BRACE)
229 return globexp1(patbuf, pglob, &limit);
231 return glob0(patbuf, pglob, &limit);
235 * Expand recursively a glob {} pattern. When there is no more expansion
236 * invoke the standard globbing routine to glob the rest of the magic
240 globexp1(const Char *pattern, glob_t *pglob, size_t *limit)
242 const Char* ptr = pattern;
245 /* Protect a single {}, for find(1), like csh */
246 if (pattern[0] == LBRACE && pattern[1] == RBRACE && pattern[2] == EOS)
247 return glob0(pattern, pglob, limit);
249 while ((ptr = (const Char *) g_strchr((Char *) ptr, LBRACE)) != NULL)
250 if (!globexp2(ptr, pattern, pglob, &rv, limit))
253 return glob0(pattern, pglob, limit);
258 * Recursive brace globbing helper. Tries to expand a single brace.
259 * If it succeeds then it invokes globexp1 with the new pattern.
260 * If it fails then it tries to glob the rest of the pattern and returns.
263 globexp2(const Char *ptr, const Char *pattern, glob_t *pglob, int *rv, size_t *limit)
267 const Char *pe, *pm, *pm1, *pl;
268 Char patbuf[MAXPATHLEN];
270 /* copy part up to the brace */
271 for (lm = patbuf, pm = pattern; pm != ptr; *lm++ = *pm++)
276 /* Find the balanced brace */
277 for (i = 0, pe = ++ptr; *pe; pe++)
278 if (*pe == LBRACKET) {
279 /* Ignore everything between [] */
280 for (pm = pe++; *pe != RBRACKET && *pe != EOS; pe++)
284 * We could not find a matching RBRACKET.
285 * Ignore and just look for RBRACE
290 else if (*pe == LBRACE)
292 else if (*pe == RBRACE) {
298 /* Non matching braces; just glob the pattern */
299 if (i != 0 || *pe == EOS) {
300 *rv = glob0(patbuf, pglob, limit);
304 for (i = 0, pl = pm = ptr; pm <= pe; pm++)
307 /* Ignore everything between [] */
308 for (pm1 = pm++; *pm != RBRACKET && *pm != EOS; pm++)
312 * We could not find a matching RBRACKET.
313 * Ignore and just look for RBRACE
330 if (i && *pm == COMMA)
333 /* Append the current string */
334 for (lm = ls; (pl < pm); *lm++ = *pl++)
337 * Append the rest of the pattern after the
340 for (pl = pe + 1; (*lm++ = *pl++) != EOS;)
343 /* Expand the current pattern */
345 qprintf("globexp2:", patbuf);
347 *rv = globexp1(patbuf, pglob, limit);
349 /* move after the comma, to the next string */
364 * expand tilde from the passwd file.
367 globtilde(const Char *pattern, Char *patbuf, size_t patbuf_len, glob_t *pglob)
374 if (*pattern != TILDE || !(pglob->gl_flags & GLOB_TILDE))
378 * Copy up to the end of the string or /
380 eb = &patbuf[patbuf_len - 1];
381 for (p = pattern + 1, h = (char *) patbuf;
382 h < (char *)eb && *p && *p != SLASH; *h++ = *p++)
387 if (((char *) patbuf)[0] == EOS) {
389 * handle a plain ~ or ~/ by expanding $HOME first (iff
390 * we're not running setuid or setgid) and then trying
393 if (issetugid() != 0 ||
394 (h = getenv("HOME")) == NULL) {
395 if (((h = getlogin()) != NULL &&
396 (pwd = getpwnam(h)) != NULL) ||
397 (pwd = getpwuid(getuid())) != NULL)
407 if ((pwd = getpwnam((char*) patbuf)) == NULL)
413 /* Copy the home directory */
414 for (b = patbuf; b < eb && *h; *b++ = *h++)
417 /* Append the rest of the pattern */
418 while (b < eb && (*b++ = *p++) != EOS)
427 * The main glob() routine: compiles the pattern (optionally processing
428 * quotes), calls glob1() to do the real pattern matching, and finally
429 * sorts the list (unless unsorted operation is requested). Returns 0
430 * if things went well, nonzero if errors occurred.
433 glob0(const Char *pattern, glob_t *pglob, size_t *limit)
435 const Char *qpatnext;
438 Char *bufnext, patbuf[MAXPATHLEN];
440 qpatnext = globtilde(pattern, patbuf, MAXPATHLEN, pglob);
441 oldpathc = pglob->gl_pathc;
444 /* We don't need to check for buffer overflow any more. */
445 while ((c = *qpatnext++) != EOS) {
451 if (*qpatnext == EOS ||
452 g_strchr((Char *) qpatnext+1, RBRACKET) == NULL) {
453 *bufnext++ = LBRACKET;
463 *bufnext++ = CHAR(c);
464 if (*qpatnext == RANGE &&
465 (c = qpatnext[1]) != RBRACKET) {
467 *bufnext++ = CHAR(c);
470 } while ((c = *qpatnext++) != RBRACKET);
471 pglob->gl_flags |= GLOB_MAGCHAR;
475 pglob->gl_flags |= GLOB_MAGCHAR;
479 pglob->gl_flags |= GLOB_MAGCHAR;
480 /* collapse adjacent stars to one,
481 * to avoid exponential behavior
483 if (bufnext == patbuf || bufnext[-1] != M_ALL)
487 *bufnext++ = CHAR(c);
493 qprintf("glob0:", patbuf);
496 if ((err = glob1(patbuf, pglob, limit)) != 0)
500 * If there was no match we are going to append the pattern
501 * if GLOB_NOCHECK was specified or if GLOB_NOMAGIC was specified
502 * and the pattern did not contain any magic characters
503 * GLOB_NOMAGIC is there just for compatibility with csh.
505 if (pglob->gl_pathc == oldpathc) {
506 if (((pglob->gl_flags & GLOB_NOCHECK) ||
507 ((pglob->gl_flags & GLOB_NOMAGIC) &&
508 !(pglob->gl_flags & GLOB_MAGCHAR))))
509 return(globextend(pattern, pglob, limit));
511 return(GLOB_NOMATCH);
513 if (!(pglob->gl_flags & GLOB_NOSORT))
514 qsort(pglob->gl_pathv + pglob->gl_offs + oldpathc,
515 pglob->gl_pathc - oldpathc, sizeof(char *), compare);
520 compare(const void *p, const void *q)
522 return(strcmp(*(char **)p, *(char **)q));
526 glob1(Char *pattern, glob_t *pglob, size_t *limit)
528 Char pathbuf[MAXPATHLEN];
530 /* A null pathname is invalid -- POSIX 1003.1 sect. 2.4. */
533 return(glob2(pathbuf, pathbuf, pathbuf + MAXPATHLEN - 1,
534 pattern, pglob, limit));
538 * The functions glob2 and glob3 are mutually recursive; there is one level
539 * of recursion for each segment in the pattern that contains one or more
543 glob2(Char *pathbuf, Char *pathend, Char *pathend_last, Char *pattern,
544 glob_t *pglob, size_t *limit)
551 * Loop over pattern segments until end of pattern or until
552 * segment with meta character found.
554 for (anymeta = 0;;) {
555 if (*pattern == EOS) { /* End of pattern? */
557 if (g_lstat(pathbuf, &sb, pglob))
560 if (((pglob->gl_flags & GLOB_MARK) &&
561 pathend[-1] != SEP) && (S_ISDIR(sb.st_mode)
562 || (S_ISLNK(sb.st_mode) &&
563 (g_stat(pathbuf, &sb, pglob) == 0) &&
564 S_ISDIR(sb.st_mode)))) {
565 if (pathend + 1 > pathend_last)
566 return (GLOB_ABORTED);
571 return(globextend(pathbuf, pglob, limit));
574 /* Find end of next segment, copy tentatively to pathend. */
577 while (*p != EOS && *p != SEP) {
580 if (q + 1 > pathend_last)
581 return (GLOB_ABORTED);
585 if (!anymeta) { /* No expansion, do next segment. */
588 while (*pattern == SEP) {
589 if (pathend + 1 > pathend_last)
590 return (GLOB_ABORTED);
591 *pathend++ = *pattern++;
593 } else /* Need expansion, recurse. */
594 return(glob3(pathbuf, pathend, pathend_last, pattern, p,
601 glob3(Char *pathbuf, Char *pathend, Char *pathend_last,
602 Char *pattern, Char *restpattern,
603 glob_t *pglob, size_t *limit)
608 char buf[MAXPATHLEN];
611 * The readdirfunc declaration can't be prototyped, because it is
612 * assigned, below, to two functions which are prototyped in glob.h
613 * and dirent.h as taking pointers to differently typed opaque
616 struct dirent *(*readdirfunc)();
618 if (pathend > pathend_last)
619 return (GLOB_ABORTED);
623 if ((dirp = g_opendir(pathbuf, pglob)) == NULL) {
624 /* TODO: don't call for ENOENT or ENOTDIR? */
625 if (pglob->gl_errfunc) {
626 if (g_Ctoc(pathbuf, buf, sizeof(buf)))
627 return (GLOB_ABORTED);
628 if (pglob->gl_errfunc(buf, errno) ||
629 pglob->gl_flags & GLOB_ERR)
630 return (GLOB_ABORTED);
637 /* Search directory for matching names. */
638 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
639 readdirfunc = pglob->gl_readdir;
641 readdirfunc = readdir;
642 while ((dp = (*readdirfunc)(dirp))) {
649 /* Initial DOT must be matched literally. */
650 if (dp->d_name[0] == DOT && *pattern != DOT)
652 memset(&mbs, 0, sizeof(mbs));
655 while (dc < pathend_last) {
656 clen = mbrtowc(&wc, sc, MB_LEN_MAX, &mbs);
657 if (clen == (size_t)-1 || clen == (size_t)-2) {
660 memset(&mbs, 0, sizeof(mbs));
662 if ((*dc++ = wc) == EOS)
666 if (!match(pathend, pattern, restpattern)) {
670 err = glob2(pathbuf, --dc, pathend_last, restpattern,
676 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
677 (*pglob->gl_closedir)(dirp);
685 * Extend the gl_pathv member of a glob_t structure to accomodate a new item,
686 * add the new item, and update gl_pathc.
688 * This assumes the BSD realloc, which only copies the block when its size
689 * crosses a power-of-two boundary; for v7 realloc, this would cause quadratic
692 * Return 0 if new item added, error code if memory couldn't be allocated.
694 * Invariant of the glob_t structure:
695 * Either gl_pathc is zero and gl_pathv is NULL; or gl_pathc > 0 and
696 * gl_pathv points to (gl_offs + gl_pathc + 1) items.
699 globextend(const Char *path, glob_t *pglob, size_t *limit)
702 size_t i, newsize, len;
706 if (*limit && pglob->gl_pathc > *limit) {
708 return (GLOB_NOSPACE);
711 newsize = sizeof(*pathv) * (2 + pglob->gl_pathc + pglob->gl_offs);
712 pathv = pglob->gl_pathv ?
713 realloc((char *)pglob->gl_pathv, newsize) :
716 if (pglob->gl_pathv) {
717 free(pglob->gl_pathv);
718 pglob->gl_pathv = NULL;
720 return(GLOB_NOSPACE);
723 if (pglob->gl_pathv == NULL && pglob->gl_offs > 0) {
724 /* first time around -- clear initial gl_offs items */
725 pathv += pglob->gl_offs;
726 for (i = pglob->gl_offs + 1; --i > 0; )
729 pglob->gl_pathv = pathv;
731 for (p = path; *p++;)
733 len = MB_CUR_MAX * (size_t)(p - path); /* XXX overallocation */
734 if ((copy = malloc(len)) != NULL) {
735 if (g_Ctoc(path, copy, len)) {
737 return (GLOB_NOSPACE);
739 pathv[pglob->gl_offs + pglob->gl_pathc++] = copy;
741 pathv[pglob->gl_offs + pglob->gl_pathc] = NULL;
742 return(copy == NULL ? GLOB_NOSPACE : 0);
746 * pattern matching function for filenames. Each occurrence of the *
747 * pattern causes a recursion level.
750 match(Char *name, Char *pat, Char *patend)
752 int ok, negate_range;
755 while (pat < patend) {
757 switch (c & M_MASK) {
762 if (match(name, pat, patend))
764 while (*name++ != EOS);
772 if ((k = *name++) == EOS)
774 if ((negate_range = ((*pat & M_MASK) == M_NOT)) != EOS)
776 while (((c = *pat++) & M_MASK) != M_END)
777 if ((*pat & M_MASK) == M_RNG) {
778 if (__collate_load_error ?
779 CHAR(c) <= CHAR(k) && CHAR(k) <= CHAR(pat[1]) :
780 __collate_range_cmp(CHAR(c), CHAR(k)) <= 0
781 && __collate_range_cmp(CHAR(k), CHAR(pat[1])) <= 0
787 if (ok == negate_range)
796 return(*name == EOS);
799 /* Free allocated data belonging to a glob_t structure. */
801 globfree(glob_t *pglob)
806 if (pglob->gl_pathv != NULL) {
807 pp = pglob->gl_pathv + pglob->gl_offs;
808 for (i = pglob->gl_pathc; i--; ++pp)
811 free(pglob->gl_pathv);
812 pglob->gl_pathv = NULL;
817 g_opendir(Char *str, glob_t *pglob)
819 char buf[MAXPATHLEN];
824 if (g_Ctoc(str, buf, sizeof(buf)))
828 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
829 return((*pglob->gl_opendir)(buf));
831 return(opendir(buf));
835 g_lstat(Char *fn, struct stat *sb, glob_t *pglob)
837 char buf[MAXPATHLEN];
839 if (g_Ctoc(fn, buf, sizeof(buf))) {
840 errno = ENAMETOOLONG;
843 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
844 return((*pglob->gl_lstat)(buf, sb));
845 return(lstat(buf, sb));
849 g_stat(Char *fn, struct stat *sb, glob_t *pglob)
851 char buf[MAXPATHLEN];
853 if (g_Ctoc(fn, buf, sizeof(buf))) {
854 errno = ENAMETOOLONG;
857 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
858 return((*pglob->gl_stat)(buf, sb));
859 return(stat(buf, sb));
863 g_strchr(Char *str, wchar_t ch)
874 g_Ctoc(const Char *str, char *buf, size_t len)
879 memset(&mbs, 0, sizeof(mbs));
880 while (len >= MB_CUR_MAX) {
881 clen = wcrtomb(buf, *str, &mbs);
882 if (clen == (size_t)-1)
895 qprintf(const char *str, Char *s)
899 (void)printf("%s:\n", str);
901 (void)printf("%c", CHAR(*p));
904 (void)printf("%c", *p & M_PROTECT ? '"' : ' ');
907 (void)printf("%c", ismeta(*p) ? '_' : ' ');