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
45 #elif defined(HAVE_SYSCONF) && defined(_SC_ARG_MAX)
46 return(sysconf(_SC_ARG_MAX));
48 return(256); /* XXX: arbitrary */
52 #if !defined(HAVE_GLOB) || !defined(GLOB_HAS_ALTDIRFUNC) || \
53 !defined(GLOB_HAS_GL_MATCHC)
55 #if defined(LIBC_SCCS) && !defined(lint)
57 static char sccsid[] = "@(#)glob.c 8.3 (Berkeley) 10/13/93";
59 static char rcsid[] = "$OpenBSD: glob.c,v 1.20 2002/06/14 21:34:58 todd Exp $";
61 #endif /* LIBC_SCCS and not lint */
64 * glob(3) -- a superset of the one defined in POSIX 1003.2.
66 * The [!...] convention to negate a range is supported (SysV, Posix, ksh).
68 * Optional extra services, controlled by flags not defined by POSIX:
71 * Escaping convention: \ inhibits any special meaning the following
72 * character might have (except \ at end of string is retained).
74 * Set in gl_flags if pattern contained a globbing character.
76 * Same as GLOB_NOCHECK, but it will only append pattern if it did
77 * not contain any magic characters. [Used in csh style globbing]
79 * Use alternately specified directory access functions.
81 * expand ~user/foo to the /home/dir/of/user/foo
83 * expand {1,2}{a,b} to 1a 1b 2a 2b
85 * Number of matches in the current invocation of glob.
100 #undef TILDE /* Some platforms may already define it */
102 #define UNDERSCORE '_'
110 #define M_QUOTE 0x8000
111 #define M_PROTECT 0x4000
112 #define M_MASK 0xffff
113 #define M_ASCII 0x00ff
115 typedef u_short Char;
120 #define M_PROTECT 0x40
129 #define CHAR(c) ((Char)((c)&M_ASCII))
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 *, u_int);
142 static int g_lstat(Char *, struct stat *, glob_t *);
143 static DIR *g_opendir(Char *, glob_t *);
144 static Char *g_strchr(Char *, int);
145 static int g_stat(Char *, struct stat *, glob_t *);
146 static int glob0(const Char *, glob_t *);
147 static int glob1(Char *, Char *, glob_t *, size_t *);
148 static int glob2(Char *, Char *, Char *, Char *, Char *, Char *,
150 static int glob3(Char *, Char *, Char *, Char *, Char *, Char *,
151 Char *, Char *, glob_t *, size_t *);
152 static int globextend(const Char *, glob_t *, size_t *);
154 globtilde(const Char *, Char *, size_t, glob_t *);
155 static int globexp1(const Char *, glob_t *);
156 static int globexp2(const Char *, const Char *, glob_t *, int *);
157 static int match(Char *, Char *, Char *);
159 static void qprintf(const char *, Char *);
163 glob(pattern, flags, errfunc, pglob)
165 int flags, (*errfunc)(const char *, int);
168 const u_char *patnext;
170 Char *bufnext, *bufend, patbuf[MAXPATHLEN];
172 patnext = (u_char *) pattern;
173 if (!(flags & GLOB_APPEND)) {
175 pglob->gl_pathv = NULL;
176 if (!(flags & GLOB_DOOFFS))
179 pglob->gl_flags = flags & ~GLOB_MAGCHAR;
180 pglob->gl_errfunc = errfunc;
181 pglob->gl_matchc = 0;
184 bufend = bufnext + MAXPATHLEN - 1;
185 if (flags & GLOB_NOESCAPE)
186 while (bufnext < bufend && (c = *patnext++) != EOS)
189 /* Protect the quoted characters. */
190 while (bufnext < bufend && (c = *patnext++) != EOS)
192 if ((c = *patnext++) == EOS) {
196 *bufnext++ = c | M_PROTECT;
202 if (flags & GLOB_BRACE)
203 return globexp1(patbuf, pglob);
205 return glob0(patbuf, pglob);
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(pattern, pglob)
218 const Char* ptr = pattern;
221 /* Protect a single {}, for find(1), like csh */
222 if (pattern[0] == LBRACE && pattern[1] == RBRACE && pattern[2] == EOS)
223 return glob0(pattern, pglob);
225 while ((ptr = (const Char *) g_strchr((Char *) ptr, LBRACE)) != NULL)
226 if (!globexp2(ptr, pattern, pglob, &rv))
229 return glob0(pattern, pglob);
234 * Recursive brace globbing helper. Tries to expand a single brace.
235 * If it succeeds then it invokes globexp1 with the new pattern.
236 * If it fails then it tries to glob the rest of the pattern and returns.
239 globexp2(ptr, pattern, pglob, rv)
240 const Char *ptr, *pattern;
246 const Char *pe, *pm, *pl;
247 Char patbuf[MAXPATHLEN];
249 /* copy part up to the brace */
250 for (lm = patbuf, pm = pattern; pm != ptr; *lm++ = *pm++)
255 /* Find the balanced brace */
256 for (i = 0, pe = ++ptr; *pe; pe++)
257 if (*pe == LBRACKET) {
258 /* Ignore everything between [] */
259 for (pm = pe++; *pe != RBRACKET && *pe != EOS; pe++)
263 * We could not find a matching RBRACKET.
264 * Ignore and just look for RBRACE
268 } else if (*pe == LBRACE)
270 else if (*pe == RBRACE) {
276 /* Non matching braces; just glob the pattern */
277 if (i != 0 || *pe == EOS) {
278 *rv = glob0(patbuf, pglob);
282 for (i = 0, pl = pm = ptr; pm <= pe; pm++) {
285 /* Ignore everything between [] */
286 for (pl = pm++; *pm != RBRACKET && *pm != EOS; pm++)
290 * We could not find a matching RBRACKET.
291 * Ignore and just look for RBRACE
308 if (i && *pm == COMMA)
311 /* Append the current string */
312 for (lm = ls; (pl < pm); *lm++ = *pl++)
316 * Append the rest of the pattern after the
319 for (pl = pe + 1; (*lm++ = *pl++) != EOS; )
322 /* Expand the current pattern */
324 qprintf("globexp2:", patbuf);
326 *rv = globexp1(patbuf, pglob);
328 /* move after the comma, to the next string */
344 * expand tilde from the passwd file.
347 globtilde(pattern, patbuf, patbuf_len, pglob)
358 if (*pattern != TILDE || !(pglob->gl_flags & GLOB_TILDE))
361 /* Copy up to the end of the string or / */
362 eb = &patbuf[patbuf_len - 1];
363 for (p = pattern + 1, h = (char *) patbuf;
364 h < (char *)eb && *p && *p != SLASH; *h++ = *p++)
374 if (((char *) patbuf)[0] == EOS) {
376 * handle a plain ~ or ~/ by expanding $HOME
377 * first and then trying the password file
380 if (issetugid() != 0 || (h = getenv("HOME")) == NULL) {
382 if ((getuid() != geteuid()) || (h = getenv("HOME")) == NULL) {
383 if ((pwd = getpwuid(getuid())) == NULL)
392 if ((pwd = getpwnam((char*) patbuf)) == NULL)
398 /* Copy the home directory */
399 for (b = patbuf; b < eb && *h; *b++ = *h++)
402 /* Append the rest of the pattern */
403 while (b < eb && (*b++ = *p++) != EOS)
412 * The main glob() routine: compiles the pattern (optionally processing
413 * quotes), calls glob1() to do the real pattern matching, and finally
414 * sorts the list (unless unsorted operation is requested). Returns 0
415 * if things went well, nonzero if errors occurred. It is not an error
416 * to find no matches.
419 glob0(pattern, pglob)
423 const Char *qpatnext;
424 int c, err, oldpathc;
425 Char *bufnext, patbuf[MAXPATHLEN];
428 qpatnext = globtilde(pattern, patbuf, MAXPATHLEN, pglob);
429 oldpathc = pglob->gl_pathc;
432 /* We don't need to check for buffer overflow any more. */
433 while ((c = *qpatnext++) != EOS) {
439 if (*qpatnext == EOS ||
440 g_strchr((Char *) qpatnext+1, RBRACKET) == NULL) {
441 *bufnext++ = LBRACKET;
451 *bufnext++ = CHAR(c);
452 if (*qpatnext == RANGE &&
453 (c = qpatnext[1]) != RBRACKET) {
455 *bufnext++ = CHAR(c);
458 } while ((c = *qpatnext++) != RBRACKET);
459 pglob->gl_flags |= GLOB_MAGCHAR;
463 pglob->gl_flags |= GLOB_MAGCHAR;
467 pglob->gl_flags |= GLOB_MAGCHAR;
468 /* collapse adjacent stars to one,
469 * to avoid exponential behavior
471 if (bufnext == patbuf || bufnext[-1] != M_ALL)
475 *bufnext++ = CHAR(c);
481 qprintf("glob0:", patbuf);
484 if ((err = glob1(patbuf, patbuf+MAXPATHLEN-1, pglob, &limit)) != 0)
488 * If there was no match we are going to append the pattern
489 * if GLOB_NOCHECK was specified or if GLOB_NOMAGIC was specified
490 * and the pattern did not contain any magic characters
491 * GLOB_NOMAGIC is there just for compatibility with csh.
493 if (pglob->gl_pathc == oldpathc) {
494 if ((pglob->gl_flags & GLOB_NOCHECK) ||
495 ((pglob->gl_flags & GLOB_NOMAGIC) &&
496 !(pglob->gl_flags & GLOB_MAGCHAR)))
497 return(globextend(pattern, pglob, &limit));
499 return(GLOB_NOMATCH);
501 if (!(pglob->gl_flags & GLOB_NOSORT))
502 qsort(pglob->gl_pathv + pglob->gl_offs + oldpathc,
503 pglob->gl_pathc - oldpathc, sizeof(char *), compare);
511 return(strcmp(*(char **)p, *(char **)q));
515 glob1(pattern, pattern_last, pglob, limitp)
516 Char *pattern, *pattern_last;
520 Char pathbuf[MAXPATHLEN];
522 /* A null pathname is invalid -- POSIX 1003.1 sect. 2.4. */
525 return(glob2(pathbuf, pathbuf+MAXPATHLEN-1,
526 pathbuf, pathbuf+MAXPATHLEN-1,
527 pattern, pattern_last, pglob, limitp));
531 * The functions glob2 and glob3 are mutually recursive; there is one level
532 * of recursion for each segment in the pattern that contains one or more
536 glob2(pathbuf, pathbuf_last, pathend, pathend_last, pattern,
537 pattern_last, pglob, limitp)
538 Char *pathbuf, *pathbuf_last, *pathend, *pathend_last;
539 Char *pattern, *pattern_last;
548 * Loop over pattern segments until end of pattern or until
549 * segment with meta character found.
551 for (anymeta = 0;;) {
552 if (*pattern == EOS) { /* End of pattern? */
554 if (g_lstat(pathbuf, &sb, pglob))
557 if (((pglob->gl_flags & GLOB_MARK) &&
558 pathend[-1] != SEP) && (S_ISDIR(sb.st_mode) ||
559 (S_ISLNK(sb.st_mode) &&
560 (g_stat(pathbuf, &sb, pglob) == 0) &&
561 S_ISDIR(sb.st_mode)))) {
562 if (pathend+1 > pathend_last)
568 return(globextend(pathbuf, pglob, limitp));
571 /* Find end of next segment, copy tentatively to pathend. */
574 while (*p != EOS && *p != SEP) {
577 if (q+1 > pathend_last)
582 if (!anymeta) { /* No expansion, do next segment. */
585 while (*pattern == SEP) {
586 if (pathend+1 > pathend_last)
588 *pathend++ = *pattern++;
591 /* Need expansion, recurse. */
592 return(glob3(pathbuf, pathbuf_last, pathend,
593 pathend_last, pattern, pattern_last,
594 p, pattern_last, pglob, limitp));
600 glob3(pathbuf, pathbuf_last, pathend, pathend_last, pattern, pattern_last,
601 restpattern, restpattern_last, pglob, limitp)
602 Char *pathbuf, *pathbuf_last, *pathend, *pathend_last;
603 Char *pattern, *pattern_last, *restpattern, *restpattern_last;
607 register struct dirent *dp;
610 char buf[MAXPATHLEN];
613 * The readdirfunc declaration can't be prototyped, because it is
614 * assigned, below, to two functions which are prototyped in glob.h
615 * and dirent.h as taking pointers to differently typed opaque
618 struct dirent *(*readdirfunc)();
620 if (pathend > pathend_last)
625 if ((dirp = g_opendir(pathbuf, pglob)) == NULL) {
626 /* TODO: don't call for ENOENT or ENOTDIR? */
627 if (pglob->gl_errfunc) {
628 if (g_Ctoc(pathbuf, buf, sizeof(buf)))
629 return(GLOB_ABORTED);
630 if (pglob->gl_errfunc(buf, errno) ||
631 pglob->gl_flags & GLOB_ERR)
632 return(GLOB_ABORTED);
639 /* Search directory for matching names. */
640 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
641 readdirfunc = pglob->gl_readdir;
643 readdirfunc = readdir;
644 while ((dp = (*readdirfunc)(dirp))) {
648 /* Initial DOT must be matched literally. */
649 if (dp->d_name[0] == DOT && *pattern != DOT)
652 sc = (u_char *) dp->d_name;
653 while (dc < pathend_last && (*dc++ = *sc++) != EOS)
655 if (dc >= pathend_last) {
661 if (!match(pathend, pattern, restpattern)) {
665 err = glob2(pathbuf, pathbuf_last, --dc, pathend_last,
666 restpattern, restpattern_last, pglob, limitp);
671 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
672 (*pglob->gl_closedir)(dirp);
680 * Extend the gl_pathv member of a glob_t structure to accommodate a new item,
681 * add the new item, and update gl_pathc.
683 * This assumes the BSD realloc, which only copies the block when its size
684 * crosses a power-of-two boundary; for v7 realloc, this would cause quadratic
687 * Return 0 if new item added, error code if memory couldn't be allocated.
689 * Invariant of the glob_t structure:
690 * Either gl_pathc is zero and gl_pathv is NULL; or gl_pathc > 0 and
691 * gl_pathv points to (gl_offs + gl_pathc + 1) items.
694 globextend(path, pglob, limitp)
699 register char **pathv;
705 newsize = sizeof(*pathv) * (2 + pglob->gl_pathc + pglob->gl_offs);
706 pathv = pglob->gl_pathv ? realloc((char *)pglob->gl_pathv, newsize) :
709 if (pglob->gl_pathv) {
710 free(pglob->gl_pathv);
711 pglob->gl_pathv = NULL;
713 return(GLOB_NOSPACE);
716 if (pglob->gl_pathv == NULL && pglob->gl_offs > 0) {
717 /* first time around -- clear initial gl_offs items */
718 pathv += pglob->gl_offs;
719 for (i = pglob->gl_offs; --i >= 0; )
722 pglob->gl_pathv = pathv;
724 for (p = path; *p++;)
726 len = (size_t)(p - path);
728 if ((copy = malloc(len)) != NULL) {
729 if (g_Ctoc(path, copy, len)) {
731 return(GLOB_NOSPACE);
733 pathv[pglob->gl_offs + pglob->gl_pathc++] = copy;
735 pathv[pglob->gl_offs + pglob->gl_pathc] = NULL;
737 if ((pglob->gl_flags & GLOB_LIMIT) &&
738 newsize + *limitp >= (u_int) get_arg_max()) {
740 return(GLOB_NOSPACE);
743 return(copy == NULL ? GLOB_NOSPACE : 0);
748 * pattern matching function for filenames. Each occurrence of the *
749 * pattern causes a recursion level.
752 match(name, pat, patend)
753 register Char *name, *pat, *patend;
755 int ok, negate_range;
758 while (pat < patend) {
760 switch (c & M_MASK) {
765 if (match(name, pat, patend))
767 while (*name++ != EOS)
776 if ((k = *name++) == EOS)
778 if ((negate_range = ((*pat & M_MASK) == M_NOT)) != EOS)
780 while (((c = *pat++) & M_MASK) != M_END)
781 if ((*pat & M_MASK) == M_RNG) {
782 if (c <= k && k <= pat[1])
787 if (ok == negate_range)
796 return(*name == EOS);
799 /* Free allocated data belonging to a glob_t structure. */
807 if (pglob->gl_pathv != NULL) {
808 pp = pglob->gl_pathv + pglob->gl_offs;
809 for (i = pglob->gl_pathc; i--; ++pp)
812 free(pglob->gl_pathv);
813 pglob->gl_pathv = NULL;
818 g_opendir(str, pglob)
822 char buf[MAXPATHLEN];
825 strlcpy(buf, ".", sizeof buf);
827 if (g_Ctoc(str, buf, sizeof(buf)))
831 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
832 return((*pglob->gl_opendir)(buf));
834 return(opendir(buf));
838 g_lstat(fn, sb, pglob)
843 char buf[MAXPATHLEN];
845 if (g_Ctoc(fn, buf, sizeof(buf)))
847 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
848 return((*pglob->gl_lstat)(buf, sb));
849 return(lstat(buf, sb));
853 g_stat(fn, sb, pglob)
858 char buf[MAXPATHLEN];
860 if (g_Ctoc(fn, buf, sizeof(buf)))
862 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
863 return((*pglob->gl_stat)(buf, sb));
864 return(stat(buf, sb));
880 g_Ctoc(str, buf, len)
881 register const Char *str;
887 if ((*buf++ = *str++) == EOS)
901 (void)printf("%s:\n", str);
903 (void)printf("%c", CHAR(*p));
906 (void)printf("%c", *p & M_PROTECT ? '"' : ' ');
909 (void)printf("%c", ismeta(*p) ? '_' : ' ');
914 #endif /* !defined(HAVE_GLOB) || !defined(GLOB_HAS_ALTDIRFUNC) ||
915 !defined(GLOB_HAS_GL_MATCHC) */