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 * Copyright (c) 2011 The FreeBSD Foundation
10 * Portions of this software were developed by David Chisnall
11 * under sponsorship from the FreeBSD Foundation.
13 * Redistribution and use in source and binary forms, with or without
14 * modification, are permitted provided that the following conditions
16 * 1. Redistributions of source code must retain the above copyright
17 * notice, this list of conditions and the following disclaimer.
18 * 2. Redistributions in binary form must reproduce the above copyright
19 * notice, this list of conditions and the following disclaimer in the
20 * documentation and/or other materials provided with the distribution.
21 * 4. Neither the name of the University nor the names of its contributors
22 * may be used to endorse or promote products derived from this software
23 * without specific prior written permission.
25 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
26 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
27 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
28 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
29 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
30 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
31 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
32 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
33 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
34 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
38 #if defined(LIBC_SCCS) && !defined(lint)
39 static char sccsid[] = "@(#)glob.c 8.3 (Berkeley) 10/13/93";
40 #endif /* LIBC_SCCS and not lint */
41 #include <sys/cdefs.h>
42 __FBSDID("$FreeBSD$");
45 * glob(3) -- a superset of the one defined in POSIX 1003.2.
47 * The [!...] convention to negate a range is supported (SysV, Posix, ksh).
49 * Optional extra services, controlled by flags not defined by POSIX:
52 * Escaping convention: \ inhibits any special meaning the following
53 * character might have (except \ at end of string is retained).
55 * Set in gl_flags if pattern contained a globbing character.
57 * Same as GLOB_NOCHECK, but it will only append pattern if it did
58 * not contain any magic characters. [Used in csh style globbing]
60 * Use alternately specified directory access functions.
62 * expand ~user/foo to the /home/dir/of/user/foo
64 * expand {1,2}{a,b} to 1a 1b 2a 2b
66 * Number of matches in the current invocation of glob.
70 * Some notes on multibyte character support:
71 * 1. Patterns with illegal byte sequences match nothing - even if
72 * GLOB_NOCHECK is specified.
73 * 2. Illegal byte sequences in filenames are handled by treating them as
74 * single-byte characters with a value of the first byte of the sequence
76 * 3. State-dependent encodings are not currently supported.
79 #include <sys/param.h>
109 #define UNDERSCORE '_'
117 #define M_QUOTE 0x8000000000ULL
118 #define M_PROTECT 0x4000000000ULL
119 #define M_MASK 0xffffffffffULL
120 #define M_CHAR 0x00ffffffffULL
122 typedef uint_fast64_t Char;
127 #define M_PROTECT 0x40
136 #define CHAR(c) ((Char)((c)&M_CHAR))
137 #define META(c) ((Char)((c)|M_QUOTE))
138 #define M_ALL META('*')
139 #define M_END META(']')
140 #define M_NOT META('!')
141 #define M_ONE META('?')
142 #define M_RNG META('-')
143 #define M_SET META('[')
144 #define ismeta(c) (((c)&M_QUOTE) != 0)
147 static int compare(const void *, const void *);
148 static int g_Ctoc(const Char *, char *, size_t);
149 static int g_lstat(Char *, struct stat *, glob_t *);
150 static DIR *g_opendir(Char *, glob_t *);
151 static const Char *g_strchr(const Char *, wchar_t);
153 static Char *g_strcat(Char *, const Char *);
155 static int g_stat(Char *, struct stat *, glob_t *);
156 static int glob0(const Char *, glob_t *, size_t *);
157 static int glob1(Char *, glob_t *, size_t *);
158 static int glob2(Char *, Char *, Char *, Char *, glob_t *, size_t *);
159 static int glob3(Char *, Char *, Char *, Char *, Char *, glob_t *, size_t *);
160 static int globextend(const Char *, glob_t *, size_t *);
162 globtilde(const Char *, Char *, size_t, glob_t *);
163 static int globexp1(const Char *, glob_t *, size_t *);
164 static int globexp2(const Char *, const Char *, glob_t *, int *, size_t *);
165 static int match(Char *, Char *, Char *);
167 static void qprintf(const char *, Char *);
171 glob(const char * __restrict pattern, int flags,
172 int (*errfunc)(const char *, int), glob_t * __restrict pglob)
176 Char *bufnext, *bufend, patbuf[MAXPATHLEN], prot;
182 if (!(flags & GLOB_APPEND)) {
184 pglob->gl_pathv = NULL;
185 if (!(flags & GLOB_DOOFFS))
188 if (flags & GLOB_LIMIT) {
189 limit = pglob->gl_matchc;
194 pglob->gl_flags = flags & ~GLOB_MAGCHAR;
195 pglob->gl_errfunc = errfunc;
196 pglob->gl_matchc = 0;
199 bufend = bufnext + MAXPATHLEN - 1;
200 if (flags & GLOB_NOESCAPE) {
201 memset(&mbs, 0, sizeof(mbs));
202 while (bufend - bufnext >= MB_CUR_MAX) {
203 clen = mbrtowc(&wc, patnext, MB_LEN_MAX, &mbs);
204 if (clen == (size_t)-1 || clen == (size_t)-2)
205 return (GLOB_NOMATCH);
212 /* Protect the quoted characters. */
213 memset(&mbs, 0, sizeof(mbs));
214 while (bufend - bufnext >= MB_CUR_MAX) {
215 if (*patnext == QUOTE) {
216 if (*++patnext == EOS) {
217 *bufnext++ = QUOTE | M_PROTECT;
223 clen = mbrtowc(&wc, patnext, MB_LEN_MAX, &mbs);
224 if (clen == (size_t)-1 || clen == (size_t)-2)
225 return (GLOB_NOMATCH);
228 *bufnext++ = wc | prot;
234 if (flags & GLOB_BRACE)
235 return globexp1(patbuf, pglob, &limit);
237 return glob0(patbuf, pglob, &limit);
241 * Expand recursively a glob {} pattern. When there is no more expansion
242 * invoke the standard globbing routine to glob the rest of the magic
246 globexp1(const Char *pattern, glob_t *pglob, size_t *limit)
248 const Char* ptr = pattern;
251 /* Protect a single {}, for find(1), like csh */
252 if (pattern[0] == LBRACE && pattern[1] == RBRACE && pattern[2] == EOS)
253 return glob0(pattern, pglob, limit);
255 while ((ptr = g_strchr(ptr, LBRACE)) != NULL)
256 if (!globexp2(ptr, pattern, pglob, &rv, limit))
259 return glob0(pattern, pglob, limit);
264 * Recursive brace globbing helper. Tries to expand a single brace.
265 * If it succeeds then it invokes globexp1 with the new pattern.
266 * If it fails then it tries to glob the rest of the pattern and returns.
269 globexp2(const Char *ptr, const Char *pattern, glob_t *pglob, int *rv, size_t *limit)
273 const Char *pe, *pm, *pm1, *pl;
274 Char patbuf[MAXPATHLEN];
276 /* copy part up to the brace */
277 for (lm = patbuf, pm = pattern; pm != ptr; *lm++ = *pm++)
282 /* Find the balanced brace */
283 for (i = 0, pe = ++ptr; *pe; pe++)
284 if (*pe == LBRACKET) {
285 /* Ignore everything between [] */
286 for (pm = pe++; *pe != RBRACKET && *pe != EOS; pe++)
290 * We could not find a matching RBRACKET.
291 * Ignore and just look for RBRACE
296 else if (*pe == LBRACE)
298 else if (*pe == RBRACE) {
304 /* Non matching braces; just glob the pattern */
305 if (i != 0 || *pe == EOS) {
306 *rv = glob0(patbuf, pglob, limit);
310 for (i = 0, pl = pm = ptr; pm <= pe; pm++)
313 /* Ignore everything between [] */
314 for (pm1 = pm++; *pm != RBRACKET && *pm != EOS; pm++)
318 * We could not find a matching RBRACKET.
319 * Ignore and just look for RBRACE
336 if (i && *pm == COMMA)
339 /* Append the current string */
340 for (lm = ls; (pl < pm); *lm++ = *pl++)
343 * Append the rest of the pattern after the
346 for (pl = pe + 1; (*lm++ = *pl++) != EOS;)
349 /* Expand the current pattern */
351 qprintf("globexp2:", patbuf);
353 *rv = globexp1(patbuf, pglob, limit);
355 /* move after the comma, to the next string */
370 * expand tilde from the passwd file.
373 globtilde(const Char *pattern, Char *patbuf, size_t patbuf_len, glob_t *pglob)
380 if (*pattern != TILDE || !(pglob->gl_flags & GLOB_TILDE))
384 * Copy up to the end of the string or /
386 eb = &patbuf[patbuf_len - 1];
387 for (p = pattern + 1, h = (char *) patbuf;
388 h < (char *)eb && *p && *p != SLASH; *h++ = *p++)
393 if (((char *) patbuf)[0] == EOS) {
395 * handle a plain ~ or ~/ by expanding $HOME first (iff
396 * we're not running setuid or setgid) and then trying
399 if (issetugid() != 0 ||
400 (h = getenv("HOME")) == NULL) {
401 if (((h = getlogin()) != NULL &&
402 (pwd = getpwnam(h)) != NULL) ||
403 (pwd = getpwuid(getuid())) != NULL)
413 if ((pwd = getpwnam((char*) patbuf)) == NULL)
419 /* Copy the home directory */
420 for (b = patbuf; b < eb && *h; *b++ = *h++)
423 /* Append the rest of the pattern */
424 while (b < eb && (*b++ = *p++) != EOS)
433 * The main glob() routine: compiles the pattern (optionally processing
434 * quotes), calls glob1() to do the real pattern matching, and finally
435 * sorts the list (unless unsorted operation is requested). Returns 0
436 * if things went well, nonzero if errors occurred.
439 glob0(const Char *pattern, glob_t *pglob, size_t *limit)
441 const Char *qpatnext;
444 Char *bufnext, c, patbuf[MAXPATHLEN];
446 qpatnext = globtilde(pattern, patbuf, MAXPATHLEN, pglob);
447 oldpathc = pglob->gl_pathc;
450 /* We don't need to check for buffer overflow any more. */
451 while ((c = *qpatnext++) != EOS) {
457 if (*qpatnext == EOS ||
458 g_strchr(qpatnext+1, RBRACKET) == NULL) {
459 *bufnext++ = LBRACKET;
469 *bufnext++ = CHAR(c);
470 if (*qpatnext == RANGE &&
471 (c = qpatnext[1]) != RBRACKET) {
473 *bufnext++ = CHAR(c);
476 } while ((c = *qpatnext++) != RBRACKET);
477 pglob->gl_flags |= GLOB_MAGCHAR;
481 pglob->gl_flags |= GLOB_MAGCHAR;
485 pglob->gl_flags |= GLOB_MAGCHAR;
486 /* collapse adjacent stars to one,
487 * to avoid exponential behavior
489 if (bufnext == patbuf || bufnext[-1] != M_ALL)
493 *bufnext++ = CHAR(c);
499 qprintf("glob0:", patbuf);
502 if ((err = glob1(patbuf, pglob, limit)) != 0)
506 * If there was no match we are going to append the pattern
507 * if GLOB_NOCHECK was specified or if GLOB_NOMAGIC was specified
508 * and the pattern did not contain any magic characters
509 * GLOB_NOMAGIC is there just for compatibility with csh.
511 if (pglob->gl_pathc == oldpathc) {
512 if (((pglob->gl_flags & GLOB_NOCHECK) ||
513 ((pglob->gl_flags & GLOB_NOMAGIC) &&
514 !(pglob->gl_flags & GLOB_MAGCHAR))))
515 return(globextend(pattern, pglob, limit));
517 return(GLOB_NOMATCH);
519 if (!(pglob->gl_flags & GLOB_NOSORT))
520 qsort(pglob->gl_pathv + pglob->gl_offs + oldpathc,
521 pglob->gl_pathc - oldpathc, sizeof(char *), compare);
526 compare(const void *p, const void *q)
528 return(strcmp(*(char **)p, *(char **)q));
532 glob1(Char *pattern, glob_t *pglob, size_t *limit)
534 Char pathbuf[MAXPATHLEN];
536 /* A null pathname is invalid -- POSIX 1003.1 sect. 2.4. */
539 return(glob2(pathbuf, pathbuf, pathbuf + MAXPATHLEN - 1,
540 pattern, pglob, limit));
544 * The functions glob2 and glob3 are mutually recursive; there is one level
545 * of recursion for each segment in the pattern that contains one or more
549 glob2(Char *pathbuf, Char *pathend, Char *pathend_last, Char *pattern,
550 glob_t *pglob, size_t *limit)
557 * Loop over pattern segments until end of pattern or until
558 * segment with meta character found.
560 for (anymeta = 0;;) {
561 if (*pattern == EOS) { /* End of pattern? */
563 if (g_lstat(pathbuf, &sb, pglob))
566 if (((pglob->gl_flags & GLOB_MARK) &&
567 pathend[-1] != SEP) && (S_ISDIR(sb.st_mode)
568 || (S_ISLNK(sb.st_mode) &&
569 (g_stat(pathbuf, &sb, pglob) == 0) &&
570 S_ISDIR(sb.st_mode)))) {
571 if (pathend + 1 > pathend_last)
572 return (GLOB_ABORTED);
577 return(globextend(pathbuf, pglob, limit));
580 /* Find end of next segment, copy tentatively to pathend. */
583 while (*p != EOS && *p != SEP) {
586 if (q + 1 > pathend_last)
587 return (GLOB_ABORTED);
591 if (!anymeta) { /* No expansion, do next segment. */
594 while (*pattern == SEP) {
595 if (pathend + 1 > pathend_last)
596 return (GLOB_ABORTED);
597 *pathend++ = *pattern++;
599 } else /* Need expansion, recurse. */
600 return(glob3(pathbuf, pathend, pathend_last, pattern, p,
607 glob3(Char *pathbuf, Char *pathend, Char *pathend_last,
608 Char *pattern, Char *restpattern,
609 glob_t *pglob, size_t *limit)
614 char buf[MAXPATHLEN];
617 * The readdirfunc declaration can't be prototyped, because it is
618 * assigned, below, to two functions which are prototyped in glob.h
619 * and dirent.h as taking pointers to differently typed opaque
622 struct dirent *(*readdirfunc)();
624 if (pathend > pathend_last)
625 return (GLOB_ABORTED);
629 if ((dirp = g_opendir(pathbuf, pglob)) == NULL) {
630 /* TODO: don't call for ENOENT or ENOTDIR? */
631 if (pglob->gl_errfunc) {
632 if (g_Ctoc(pathbuf, buf, sizeof(buf)))
633 return (GLOB_ABORTED);
634 if (pglob->gl_errfunc(buf, errno) ||
635 pglob->gl_flags & GLOB_ERR)
636 return (GLOB_ABORTED);
643 /* Search directory for matching names. */
644 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
645 readdirfunc = pglob->gl_readdir;
647 readdirfunc = readdir;
648 while ((dp = (*readdirfunc)(dirp))) {
655 /* Initial DOT must be matched literally. */
656 if (dp->d_name[0] == DOT && *pattern != DOT)
658 memset(&mbs, 0, sizeof(mbs));
661 while (dc < pathend_last) {
662 clen = mbrtowc(&wc, sc, MB_LEN_MAX, &mbs);
663 if (clen == (size_t)-1 || clen == (size_t)-2) {
666 memset(&mbs, 0, sizeof(mbs));
668 if ((*dc++ = wc) == EOS)
672 if (!match(pathend, pattern, restpattern)) {
676 err = glob2(pathbuf, --dc, pathend_last, restpattern,
682 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
683 (*pglob->gl_closedir)(dirp);
691 * Extend the gl_pathv member of a glob_t structure to accomodate a new item,
692 * add the new item, and update gl_pathc.
694 * This assumes the BSD realloc, which only copies the block when its size
695 * crosses a power-of-two boundary; for v7 realloc, this would cause quadratic
698 * Return 0 if new item added, error code if memory couldn't be allocated.
700 * Invariant of the glob_t structure:
701 * Either gl_pathc is zero and gl_pathv is NULL; or gl_pathc > 0 and
702 * gl_pathv points to (gl_offs + gl_pathc + 1) items.
705 globextend(const Char *path, glob_t *pglob, size_t *limit)
708 size_t i, newsize, len;
712 if (*limit && pglob->gl_pathc > *limit) {
714 return (GLOB_NOSPACE);
717 newsize = sizeof(*pathv) * (2 + pglob->gl_pathc + pglob->gl_offs);
718 pathv = pglob->gl_pathv ?
719 realloc((char *)pglob->gl_pathv, newsize) :
722 if (pglob->gl_pathv) {
723 free(pglob->gl_pathv);
724 pglob->gl_pathv = NULL;
726 return(GLOB_NOSPACE);
729 if (pglob->gl_pathv == NULL && pglob->gl_offs > 0) {
730 /* first time around -- clear initial gl_offs items */
731 pathv += pglob->gl_offs;
732 for (i = pglob->gl_offs + 1; --i > 0; )
735 pglob->gl_pathv = pathv;
737 for (p = path; *p++;)
739 len = MB_CUR_MAX * (size_t)(p - path); /* XXX overallocation */
740 if ((copy = malloc(len)) != NULL) {
741 if (g_Ctoc(path, copy, len)) {
743 return (GLOB_NOSPACE);
745 pathv[pglob->gl_offs + pglob->gl_pathc++] = copy;
747 pathv[pglob->gl_offs + pglob->gl_pathc] = NULL;
748 return(copy == NULL ? GLOB_NOSPACE : 0);
752 * pattern matching function for filenames. Each occurrence of the *
753 * pattern causes a recursion level.
756 match(Char *name, Char *pat, Char *patend)
758 int ok, negate_range;
760 struct xlocale_collate *table =
761 (struct xlocale_collate*)__get_locale()->components[XLC_COLLATE];
763 while (pat < patend) {
765 switch (c & M_MASK) {
770 if (match(name, pat, patend))
772 while (*name++ != EOS);
780 if ((k = *name++) == EOS)
782 if ((negate_range = ((*pat & M_MASK) == M_NOT)) != EOS)
784 while (((c = *pat++) & M_MASK) != M_END)
785 if ((*pat & M_MASK) == M_RNG) {
786 if (table->__collate_load_error ?
787 CHAR(c) <= CHAR(k) && CHAR(k) <= CHAR(pat[1]) :
788 __collate_range_cmp(table, CHAR(c), CHAR(k)) <= 0
789 && __collate_range_cmp(table, CHAR(k), CHAR(pat[1])) <= 0
795 if (ok == negate_range)
804 return(*name == EOS);
807 /* Free allocated data belonging to a glob_t structure. */
809 globfree(glob_t *pglob)
814 if (pglob->gl_pathv != NULL) {
815 pp = pglob->gl_pathv + pglob->gl_offs;
816 for (i = pglob->gl_pathc; i--; ++pp)
819 free(pglob->gl_pathv);
820 pglob->gl_pathv = NULL;
825 g_opendir(Char *str, glob_t *pglob)
827 char buf[MAXPATHLEN];
832 if (g_Ctoc(str, buf, sizeof(buf)))
836 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
837 return((*pglob->gl_opendir)(buf));
839 return(opendir(buf));
843 g_lstat(Char *fn, struct stat *sb, glob_t *pglob)
845 char buf[MAXPATHLEN];
847 if (g_Ctoc(fn, buf, sizeof(buf))) {
848 errno = ENAMETOOLONG;
851 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
852 return((*pglob->gl_lstat)(buf, sb));
853 return(lstat(buf, sb));
857 g_stat(Char *fn, struct stat *sb, glob_t *pglob)
859 char buf[MAXPATHLEN];
861 if (g_Ctoc(fn, buf, sizeof(buf))) {
862 errno = ENAMETOOLONG;
865 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
866 return((*pglob->gl_stat)(buf, sb));
867 return(stat(buf, sb));
871 g_strchr(const Char *str, wchar_t ch)
882 g_Ctoc(const Char *str, char *buf, size_t len)
887 memset(&mbs, 0, sizeof(mbs));
888 while (len >= MB_CUR_MAX) {
889 clen = wcrtomb(buf, *str, &mbs);
890 if (clen == (size_t)-1)
903 qprintf(const char *str, Char *s)
907 (void)printf("%s:\n", str);
909 (void)printf("%c", CHAR(*p));
912 (void)printf("%c", *p & M_PROTECT ? '"' : ' ');
915 (void)printf("%c", ismeta(*p) ? '_' : ' ');