1 /* $OpenBSD: glob.c,v 1.25 2005/08/08 08:05:34 espie Exp $ */
3 * Copyright (c) 1989, 1993
4 * The Regents of the University of California. All rights reserved.
6 * This code is derived from software contributed to Berkeley by
9 * Redistribution and use in source and binary forms, with or without
10 * modification, are permitted provided that the following conditions
12 * 1. Redistributions of source code must retain the above copyright
13 * notice, this list of conditions and the following disclaimer.
14 * 2. Redistributions in binary form must reproduce the above copyright
15 * notice, this list of conditions and the following disclaimer in the
16 * documentation and/or other materials provided with the distribution.
17 * 3. Neither the name of the University nor the names of its contributors
18 * may be used to endorse or promote products derived from this software
19 * without specific prior written permission.
21 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
22 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
23 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
24 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
25 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
26 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
27 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
28 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
29 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
30 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
34 /* OPENBSD ORIGINAL: lib/libc/gen/glob.c */
37 #include <sys/types.h>
48 #elif defined(HAVE_SYSCONF) && defined(_SC_ARG_MAX)
49 return(sysconf(_SC_ARG_MAX));
51 return(256); /* XXX: arbitrary */
55 #if !defined(HAVE_GLOB) || !defined(GLOB_HAS_ALTDIRFUNC) || \
56 !defined(GLOB_HAS_GL_MATCHC)
59 * glob(3) -- a superset of the one defined in POSIX 1003.2.
61 * The [!...] convention to negate a range is supported (SysV, Posix, ksh).
63 * Optional extra services, controlled by flags not defined by POSIX:
66 * Escaping convention: \ inhibits any special meaning the following
67 * character might have (except \ at end of string is retained).
69 * Set in gl_flags if pattern contained a globbing character.
71 * Same as GLOB_NOCHECK, but it will only append pattern if it did
72 * not contain any magic characters. [Used in csh style globbing]
74 * Use alternately specified directory access functions.
76 * expand ~user/foo to the /home/dir/of/user/foo
78 * expand {1,2}{a,b} to 1a 1b 2a 2b
80 * Number of matches in the current invocation of glob.
95 #undef TILDE /* Some platforms may already define it */
97 #define UNDERSCORE '_'
105 #define M_QUOTE 0x8000
106 #define M_PROTECT 0x4000
107 #define M_MASK 0xffff
108 #define M_ASCII 0x00ff
110 typedef u_short Char;
115 #define M_PROTECT 0x40
124 #define CHAR(c) ((Char)((c)&M_ASCII))
125 #define META(c) ((Char)((c)|M_QUOTE))
126 #define M_ALL META('*')
127 #define M_END META(']')
128 #define M_NOT META('!')
129 #define M_ONE META('?')
130 #define M_RNG META('-')
131 #define M_SET META('[')
132 #define ismeta(c) (((c)&M_QUOTE) != 0)
135 static int compare(const void *, const void *);
136 static int g_Ctoc(const Char *, char *, u_int);
137 static int g_lstat(Char *, struct stat *, glob_t *);
138 static DIR *g_opendir(Char *, glob_t *);
139 static Char *g_strchr(Char *, int);
140 static int g_stat(Char *, struct stat *, glob_t *);
141 static int glob0(const Char *, glob_t *);
142 static int glob1(Char *, Char *, glob_t *, size_t *);
143 static int glob2(Char *, Char *, Char *, Char *, Char *, Char *,
145 static int glob3(Char *, Char *, Char *, Char *, Char *, Char *,
146 Char *, Char *, glob_t *, size_t *);
147 static int globextend(const Char *, glob_t *, size_t *);
149 globtilde(const Char *, Char *, size_t, glob_t *);
150 static int globexp1(const Char *, glob_t *);
151 static int globexp2(const Char *, const Char *, glob_t *, int *);
152 static int match(Char *, Char *, Char *);
154 static void qprintf(const char *, 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 = (u_char *) pattern;
166 if (!(flags & GLOB_APPEND)) {
168 pglob->gl_pathv = NULL;
169 if (!(flags & GLOB_DOOFFS))
172 pglob->gl_flags = flags & ~GLOB_MAGCHAR;
173 pglob->gl_errfunc = errfunc;
174 pglob->gl_matchc = 0;
177 bufend = bufnext + MAXPATHLEN - 1;
178 if (flags & GLOB_NOESCAPE)
179 while (bufnext < bufend && (c = *patnext++) != EOS)
182 /* Protect the quoted characters. */
183 while (bufnext < bufend && (c = *patnext++) != EOS)
185 if ((c = *patnext++) == EOS) {
189 *bufnext++ = c | M_PROTECT;
195 if (flags & GLOB_BRACE)
196 return globexp1(patbuf, pglob);
198 return glob0(patbuf, pglob);
202 * Expand recursively a glob {} pattern. When there is no more expansion
203 * invoke the standard globbing routine to glob the rest of the magic
207 globexp1(const Char *pattern, glob_t *pglob)
209 const Char* ptr = pattern;
212 /* Protect a single {}, for find(1), like csh */
213 if (pattern[0] == LBRACE && pattern[1] == RBRACE && pattern[2] == EOS)
214 return glob0(pattern, pglob);
216 while ((ptr = (const Char *) g_strchr((Char *) ptr, LBRACE)) != NULL)
217 if (!globexp2(ptr, pattern, pglob, &rv))
220 return glob0(pattern, pglob);
225 * Recursive brace globbing helper. Tries to expand a single brace.
226 * If it succeeds then it invokes globexp1 with the new pattern.
227 * If it fails then it tries to glob the rest of the pattern and returns.
230 globexp2(const Char *ptr, const Char *pattern, glob_t *pglob, int *rv)
234 const Char *pe, *pm, *pl;
235 Char patbuf[MAXPATHLEN];
237 /* copy part up to the brace */
238 for (lm = patbuf, pm = pattern; pm != ptr; *lm++ = *pm++)
243 /* Find the balanced brace */
244 for (i = 0, pe = ++ptr; *pe; pe++)
245 if (*pe == LBRACKET) {
246 /* Ignore everything between [] */
247 for (pm = pe++; *pe != RBRACKET && *pe != EOS; pe++)
251 * We could not find a matching RBRACKET.
252 * Ignore and just look for RBRACE
256 } else if (*pe == LBRACE)
258 else if (*pe == RBRACE) {
264 /* Non matching braces; just glob the pattern */
265 if (i != 0 || *pe == EOS) {
266 *rv = glob0(patbuf, pglob);
270 for (i = 0, pl = pm = ptr; pm <= pe; pm++) {
273 /* Ignore everything between [] */
274 for (pl = pm++; *pm != RBRACKET && *pm != EOS; pm++)
278 * We could not find a matching RBRACKET.
279 * Ignore and just look for RBRACE
296 if (i && *pm == COMMA)
299 /* Append the current string */
300 for (lm = ls; (pl < pm); *lm++ = *pl++)
304 * Append the rest of the pattern after the
307 for (pl = pe + 1; (*lm++ = *pl++) != EOS; )
310 /* Expand the current pattern */
312 qprintf("globexp2:", patbuf);
314 *rv = globexp1(patbuf, pglob);
316 /* move after the comma, to the next string */
332 * expand tilde from the passwd file.
335 globtilde(const Char *pattern, Char *patbuf, size_t patbuf_len, glob_t *pglob)
342 if (*pattern != TILDE || !(pglob->gl_flags & GLOB_TILDE))
345 /* Copy up to the end of the string or / */
346 eb = &patbuf[patbuf_len - 1];
347 for (p = pattern + 1, h = (char *) patbuf;
348 h < (char *)eb && *p && *p != SLASH; *h++ = *p++)
358 if (((char *) patbuf)[0] == EOS) {
360 * handle a plain ~ or ~/ by expanding $HOME
361 * first and then trying the password file
364 if (issetugid() != 0 || (h = getenv("HOME")) == NULL) {
366 if ((getuid() != geteuid()) || (h = getenv("HOME")) == NULL) {
367 if ((pwd = getpwuid(getuid())) == NULL)
376 if ((pwd = getpwnam((char*) patbuf)) == NULL)
382 /* Copy the home directory */
383 for (b = patbuf; b < eb && *h; *b++ = *h++)
386 /* Append the rest of the pattern */
387 while (b < eb && (*b++ = *p++) != EOS)
396 * The main glob() routine: compiles the pattern (optionally processing
397 * quotes), calls glob1() to do the real pattern matching, and finally
398 * sorts the list (unless unsorted operation is requested). Returns 0
399 * if things went well, nonzero if errors occurred. It is not an error
400 * to find no matches.
403 glob0(const Char *pattern, glob_t *pglob)
405 const Char *qpatnext;
406 int c, err, oldpathc;
407 Char *bufnext, patbuf[MAXPATHLEN];
410 qpatnext = globtilde(pattern, patbuf, MAXPATHLEN, pglob);
411 oldpathc = pglob->gl_pathc;
414 /* We don't need to check for buffer overflow any more. */
415 while ((c = *qpatnext++) != EOS) {
421 if (*qpatnext == EOS ||
422 g_strchr((Char *) qpatnext+1, RBRACKET) == NULL) {
423 *bufnext++ = LBRACKET;
433 *bufnext++ = CHAR(c);
434 if (*qpatnext == RANGE &&
435 (c = qpatnext[1]) != RBRACKET) {
437 *bufnext++ = CHAR(c);
440 } while ((c = *qpatnext++) != RBRACKET);
441 pglob->gl_flags |= GLOB_MAGCHAR;
445 pglob->gl_flags |= GLOB_MAGCHAR;
449 pglob->gl_flags |= GLOB_MAGCHAR;
450 /* collapse adjacent stars to one,
451 * to avoid exponential behavior
453 if (bufnext == patbuf || bufnext[-1] != M_ALL)
457 *bufnext++ = CHAR(c);
463 qprintf("glob0:", patbuf);
466 if ((err = glob1(patbuf, patbuf+MAXPATHLEN-1, pglob, &limit)) != 0)
470 * If there was no match we are going to append the pattern
471 * if GLOB_NOCHECK was specified or if GLOB_NOMAGIC was specified
472 * and the pattern did not contain any magic characters
473 * GLOB_NOMAGIC is there just for compatibility with csh.
475 if (pglob->gl_pathc == oldpathc) {
476 if ((pglob->gl_flags & GLOB_NOCHECK) ||
477 ((pglob->gl_flags & GLOB_NOMAGIC) &&
478 !(pglob->gl_flags & GLOB_MAGCHAR)))
479 return(globextend(pattern, pglob, &limit));
481 return(GLOB_NOMATCH);
483 if (!(pglob->gl_flags & GLOB_NOSORT))
484 qsort(pglob->gl_pathv + pglob->gl_offs + oldpathc,
485 pglob->gl_pathc - oldpathc, sizeof(char *), compare);
490 compare(const void *p, const void *q)
492 return(strcmp(*(char **)p, *(char **)q));
496 glob1(Char *pattern, Char *pattern_last, glob_t *pglob, size_t *limitp)
498 Char pathbuf[MAXPATHLEN];
500 /* A null pathname is invalid -- POSIX 1003.1 sect. 2.4. */
503 return(glob2(pathbuf, pathbuf+MAXPATHLEN-1,
504 pathbuf, pathbuf+MAXPATHLEN-1,
505 pattern, pattern_last, pglob, limitp));
509 * The functions glob2 and glob3 are mutually recursive; there is one level
510 * of recursion for each segment in the pattern that contains one or more
514 glob2(Char *pathbuf, Char *pathbuf_last, Char *pathend, Char *pathend_last,
515 Char *pattern, Char *pattern_last, glob_t *pglob, size_t *limitp)
522 * Loop over pattern segments until end of pattern or until
523 * segment with meta character found.
525 for (anymeta = 0;;) {
526 if (*pattern == EOS) { /* End of pattern? */
528 if (g_lstat(pathbuf, &sb, pglob))
531 if (((pglob->gl_flags & GLOB_MARK) &&
532 pathend[-1] != SEP) && (S_ISDIR(sb.st_mode) ||
533 (S_ISLNK(sb.st_mode) &&
534 (g_stat(pathbuf, &sb, pglob) == 0) &&
535 S_ISDIR(sb.st_mode)))) {
536 if (pathend+1 > pathend_last)
542 return(globextend(pathbuf, pglob, limitp));
545 /* Find end of next segment, copy tentatively to pathend. */
548 while (*p != EOS && *p != SEP) {
551 if (q+1 > pathend_last)
556 if (!anymeta) { /* No expansion, do next segment. */
559 while (*pattern == SEP) {
560 if (pathend+1 > pathend_last)
562 *pathend++ = *pattern++;
565 /* Need expansion, recurse. */
566 return(glob3(pathbuf, pathbuf_last, pathend,
567 pathend_last, pattern, pattern_last,
568 p, pattern_last, pglob, limitp));
574 glob3(Char *pathbuf, Char *pathbuf_last, Char *pathend, Char *pathend_last,
575 Char *pattern, Char *pattern_last, Char *restpattern,
576 Char *restpattern_last, glob_t *pglob, size_t *limitp)
581 char buf[MAXPATHLEN];
584 * The readdirfunc declaration can't be prototyped, because it is
585 * assigned, below, to two functions which are prototyped in glob.h
586 * and dirent.h as taking pointers to differently typed opaque
589 struct dirent *(*readdirfunc)(void *);
591 if (pathend > pathend_last)
596 if ((dirp = g_opendir(pathbuf, pglob)) == NULL) {
597 /* TODO: don't call for ENOENT or ENOTDIR? */
598 if (pglob->gl_errfunc) {
599 if (g_Ctoc(pathbuf, buf, sizeof(buf)))
600 return(GLOB_ABORTED);
601 if (pglob->gl_errfunc(buf, errno) ||
602 pglob->gl_flags & GLOB_ERR)
603 return(GLOB_ABORTED);
610 /* Search directory for matching names. */
611 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
612 readdirfunc = pglob->gl_readdir;
614 readdirfunc = (struct dirent *(*)(void *))readdir;
615 while ((dp = (*readdirfunc)(dirp))) {
619 /* Initial DOT must be matched literally. */
620 if (dp->d_name[0] == DOT && *pattern != DOT)
623 sc = (u_char *) dp->d_name;
624 while (dc < pathend_last && (*dc++ = *sc++) != EOS)
626 if (dc >= pathend_last) {
632 if (!match(pathend, pattern, restpattern)) {
636 err = glob2(pathbuf, pathbuf_last, --dc, pathend_last,
637 restpattern, restpattern_last, pglob, limitp);
642 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
643 (*pglob->gl_closedir)(dirp);
651 * Extend the gl_pathv member of a glob_t structure to accommodate a new item,
652 * add the new item, and update gl_pathc.
654 * This assumes the BSD realloc, which only copies the block when its size
655 * crosses a power-of-two boundary; for v7 realloc, this would cause quadratic
658 * Return 0 if new item added, error code if memory couldn't be allocated.
660 * Invariant of the glob_t structure:
661 * Either gl_pathc is zero and gl_pathv is NULL; or gl_pathc > 0 and
662 * gl_pathv points to (gl_offs + gl_pathc + 1) items.
665 globextend(const Char *path, glob_t *pglob, size_t *limitp)
673 newsize = sizeof(*pathv) * (2 + pglob->gl_pathc + pglob->gl_offs);
674 pathv = pglob->gl_pathv ? realloc((char *)pglob->gl_pathv, newsize) :
677 if (pglob->gl_pathv) {
678 free(pglob->gl_pathv);
679 pglob->gl_pathv = NULL;
681 return(GLOB_NOSPACE);
684 if (pglob->gl_pathv == NULL && pglob->gl_offs > 0) {
685 /* first time around -- clear initial gl_offs items */
686 pathv += pglob->gl_offs;
687 for (i = pglob->gl_offs; --i >= 0; )
690 pglob->gl_pathv = pathv;
692 for (p = path; *p++;)
694 len = (size_t)(p - path);
696 if ((copy = malloc(len)) != NULL) {
697 if (g_Ctoc(path, copy, len)) {
699 return(GLOB_NOSPACE);
701 pathv[pglob->gl_offs + pglob->gl_pathc++] = copy;
703 pathv[pglob->gl_offs + pglob->gl_pathc] = NULL;
705 if ((pglob->gl_flags & GLOB_LIMIT) &&
706 newsize + *limitp >= (u_int) get_arg_max()) {
708 return(GLOB_NOSPACE);
711 return(copy == NULL ? GLOB_NOSPACE : 0);
716 * pattern matching function for filenames. Each occurrence of the *
717 * pattern causes a recursion level.
720 match(Char *name, Char *pat, Char *patend)
722 int ok, negate_range;
725 while (pat < patend) {
727 switch (c & M_MASK) {
732 if (match(name, pat, patend))
734 } while (*name++ != EOS);
742 if ((k = *name++) == EOS)
744 if ((negate_range = ((*pat & M_MASK) == M_NOT)) != EOS)
746 while (((c = *pat++) & M_MASK) != M_END)
747 if ((*pat & M_MASK) == M_RNG) {
748 if (c <= k && k <= pat[1])
753 if (ok == negate_range)
762 return(*name == EOS);
765 /* Free allocated data belonging to a glob_t structure. */
767 globfree(glob_t *pglob)
772 if (pglob->gl_pathv != NULL) {
773 pp = pglob->gl_pathv + pglob->gl_offs;
774 for (i = pglob->gl_pathc; i--; ++pp)
777 free(pglob->gl_pathv);
778 pglob->gl_pathv = NULL;
783 g_opendir(Char *str, glob_t *pglob)
785 char buf[MAXPATHLEN];
788 strlcpy(buf, ".", sizeof buf);
790 if (g_Ctoc(str, buf, sizeof(buf)))
794 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
795 return((*pglob->gl_opendir)(buf));
797 return(opendir(buf));
801 g_lstat(Char *fn, struct stat *sb, glob_t *pglob)
803 char buf[MAXPATHLEN];
805 if (g_Ctoc(fn, buf, sizeof(buf)))
807 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
808 return((*pglob->gl_lstat)(buf, sb));
809 return(lstat(buf, sb));
813 g_stat(Char *fn, struct stat *sb, glob_t *pglob)
815 char buf[MAXPATHLEN];
817 if (g_Ctoc(fn, buf, sizeof(buf)))
819 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
820 return((*pglob->gl_stat)(buf, sb));
821 return(stat(buf, sb));
825 g_strchr(Char *str, int ch)
835 g_Ctoc(const Char *str, char *buf, u_int len)
839 if ((*buf++ = *str++) == EOS)
847 qprintf(const char *str, Char *s)
851 (void)printf("%s:\n", str);
853 (void)printf("%c", CHAR(*p));
856 (void)printf("%c", *p & M_PROTECT ? '"' : ' ');
859 (void)printf("%c", ismeta(*p) ? '_' : ' ');
864 #endif /* !defined(HAVE_GLOB) || !defined(GLOB_HAS_ALTDIRFUNC) ||
865 !defined(GLOB_HAS_GL_MATCHC) */