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 */
38 #include <sys/types.h>
49 #if !defined(HAVE_GLOB) || !defined(GLOB_HAS_ALTDIRFUNC) || \
50 !defined(GLOB_HAS_GL_MATCHC)
57 #elif defined(HAVE_SYSCONF) && defined(_SC_ARG_MAX)
58 return(sysconf(_SC_ARG_MAX));
60 return(256); /* XXX: arbitrary */
65 * glob(3) -- a superset of the one defined in POSIX 1003.2.
67 * The [!...] convention to negate a range is supported (SysV, Posix, ksh).
69 * Optional extra services, controlled by flags not defined by POSIX:
72 * Escaping convention: \ inhibits any special meaning the following
73 * character might have (except \ at end of string is retained).
75 * Set in gl_flags if pattern contained a globbing character.
77 * Same as GLOB_NOCHECK, but it will only append pattern if it did
78 * not contain any magic characters. [Used in csh style globbing]
80 * Use alternately specified directory access functions.
82 * expand ~user/foo to the /home/dir/of/user/foo
84 * expand {1,2}{a,b} to 1a 1b 2a 2b
86 * Number of matches in the current invocation of glob.
101 #undef TILDE /* Some platforms may already define it */
103 #define UNDERSCORE '_'
111 #define M_QUOTE 0x8000
112 #define M_PROTECT 0x4000
113 #define M_MASK 0xffff
114 #define M_ASCII 0x00ff
116 typedef u_short Char;
121 #define M_PROTECT 0x40
130 #define CHAR(c) ((Char)((c)&M_ASCII))
131 #define META(c) ((Char)((c)|M_QUOTE))
132 #define M_ALL META('*')
133 #define M_END META(']')
134 #define M_NOT META('!')
135 #define M_ONE META('?')
136 #define M_RNG META('-')
137 #define M_SET META('[')
138 #define ismeta(c) (((c)&M_QUOTE) != 0)
141 static int compare(const void *, const void *);
142 static int g_Ctoc(const Char *, char *, u_int);
143 static int g_lstat(Char *, struct stat *, glob_t *);
144 static DIR *g_opendir(Char *, glob_t *);
145 static Char *g_strchr(Char *, int);
146 static int g_stat(Char *, struct stat *, glob_t *);
147 static int glob0(const Char *, glob_t *);
148 static int glob1(Char *, Char *, glob_t *, size_t *);
149 static int glob2(Char *, Char *, Char *, Char *, Char *, Char *,
151 static int glob3(Char *, Char *, Char *, Char *, Char *, Char *,
152 Char *, Char *, glob_t *, size_t *);
153 static int globextend(const Char *, glob_t *, size_t *);
155 globtilde(const Char *, Char *, size_t, glob_t *);
156 static int globexp1(const Char *, glob_t *);
157 static int globexp2(const Char *, const Char *, glob_t *, int *);
158 static int match(Char *, Char *, Char *);
160 static void qprintf(const char *, Char *);
164 glob(const char *pattern, int flags, int (*errfunc)(const char *, int),
167 const u_char *patnext;
169 Char *bufnext, *bufend, patbuf[MAXPATHLEN];
171 patnext = (u_char *) pattern;
172 if (!(flags & GLOB_APPEND)) {
174 pglob->gl_pathv = NULL;
175 if (!(flags & GLOB_DOOFFS))
178 pglob->gl_flags = flags & ~GLOB_MAGCHAR;
179 pglob->gl_errfunc = errfunc;
180 pglob->gl_matchc = 0;
183 bufend = bufnext + MAXPATHLEN - 1;
184 if (flags & GLOB_NOESCAPE)
185 while (bufnext < bufend && (c = *patnext++) != EOS)
188 /* Protect the quoted characters. */
189 while (bufnext < bufend && (c = *patnext++) != EOS)
191 if ((c = *patnext++) == EOS) {
195 *bufnext++ = c | M_PROTECT;
201 if (flags & GLOB_BRACE)
202 return globexp1(patbuf, pglob);
204 return glob0(patbuf, pglob);
208 * Expand recursively a glob {} pattern. When there is no more expansion
209 * invoke the standard globbing routine to glob the rest of the magic
213 globexp1(const Char *pattern, glob_t *pglob)
215 const Char* ptr = pattern;
218 /* Protect a single {}, for find(1), like csh */
219 if (pattern[0] == LBRACE && pattern[1] == RBRACE && pattern[2] == EOS)
220 return glob0(pattern, pglob);
222 while ((ptr = (const Char *) g_strchr((Char *) ptr, LBRACE)) != NULL)
223 if (!globexp2(ptr, pattern, pglob, &rv))
226 return glob0(pattern, pglob);
231 * Recursive brace globbing helper. Tries to expand a single brace.
232 * If it succeeds then it invokes globexp1 with the new pattern.
233 * If it fails then it tries to glob the rest of the pattern and returns.
236 globexp2(const Char *ptr, const Char *pattern, glob_t *pglob, int *rv)
240 const Char *pe, *pm, *pl;
241 Char patbuf[MAXPATHLEN];
243 /* copy part up to the brace */
244 for (lm = patbuf, pm = pattern; pm != ptr; *lm++ = *pm++)
249 /* Find the balanced brace */
250 for (i = 0, pe = ++ptr; *pe; pe++)
251 if (*pe == LBRACKET) {
252 /* Ignore everything between [] */
253 for (pm = pe++; *pe != RBRACKET && *pe != EOS; pe++)
257 * We could not find a matching RBRACKET.
258 * Ignore and just look for RBRACE
262 } else if (*pe == LBRACE)
264 else if (*pe == RBRACE) {
270 /* Non matching braces; just glob the pattern */
271 if (i != 0 || *pe == EOS) {
272 *rv = glob0(patbuf, pglob);
276 for (i = 0, pl = pm = ptr; pm <= pe; pm++) {
279 /* Ignore everything between [] */
280 for (pl = pm++; *pm != RBRACKET && *pm != EOS; pm++)
284 * We could not find a matching RBRACKET.
285 * Ignore and just look for RBRACE
302 if (i && *pm == COMMA)
305 /* Append the current string */
306 for (lm = ls; (pl < pm); *lm++ = *pl++)
310 * Append the rest of the pattern after the
313 for (pl = pe + 1; (*lm++ = *pl++) != EOS; )
316 /* Expand the current pattern */
318 qprintf("globexp2:", patbuf);
320 *rv = globexp1(patbuf, pglob);
322 /* move after the comma, to the next string */
338 * expand tilde from the passwd file.
341 globtilde(const Char *pattern, Char *patbuf, size_t patbuf_len, glob_t *pglob)
348 if (*pattern != TILDE || !(pglob->gl_flags & GLOB_TILDE))
351 /* Copy up to the end of the string or / */
352 eb = &patbuf[patbuf_len - 1];
353 for (p = pattern + 1, h = (char *) patbuf;
354 h < (char *)eb && *p && *p != SLASH; *h++ = *p++)
364 if (((char *) patbuf)[0] == EOS) {
366 * handle a plain ~ or ~/ by expanding $HOME
367 * first and then trying the password file
370 if (issetugid() != 0 || (h = getenv("HOME")) == NULL) {
372 if ((getuid() != geteuid()) || (h = getenv("HOME")) == NULL) {
373 if ((pwd = getpwuid(getuid())) == NULL)
382 if ((pwd = getpwnam((char*) patbuf)) == NULL)
388 /* Copy the home directory */
389 for (b = patbuf; b < eb && *h; *b++ = *h++)
392 /* Append the rest of the pattern */
393 while (b < eb && (*b++ = *p++) != EOS)
402 * The main glob() routine: compiles the pattern (optionally processing
403 * quotes), calls glob1() to do the real pattern matching, and finally
404 * sorts the list (unless unsorted operation is requested). Returns 0
405 * if things went well, nonzero if errors occurred. It is not an error
406 * to find no matches.
409 glob0(const Char *pattern, glob_t *pglob)
411 const Char *qpatnext;
412 int c, err, oldpathc;
413 Char *bufnext, patbuf[MAXPATHLEN];
416 qpatnext = globtilde(pattern, patbuf, MAXPATHLEN, pglob);
417 oldpathc = pglob->gl_pathc;
420 /* We don't need to check for buffer overflow any more. */
421 while ((c = *qpatnext++) != EOS) {
427 if (*qpatnext == EOS ||
428 g_strchr((Char *) qpatnext+1, RBRACKET) == NULL) {
429 *bufnext++ = LBRACKET;
439 *bufnext++ = CHAR(c);
440 if (*qpatnext == RANGE &&
441 (c = qpatnext[1]) != RBRACKET) {
443 *bufnext++ = CHAR(c);
446 } while ((c = *qpatnext++) != RBRACKET);
447 pglob->gl_flags |= GLOB_MAGCHAR;
451 pglob->gl_flags |= GLOB_MAGCHAR;
455 pglob->gl_flags |= GLOB_MAGCHAR;
456 /* collapse adjacent stars to one,
457 * to avoid exponential behavior
459 if (bufnext == patbuf || bufnext[-1] != M_ALL)
463 *bufnext++ = CHAR(c);
469 qprintf("glob0:", patbuf);
472 if ((err = glob1(patbuf, patbuf+MAXPATHLEN-1, pglob, &limit)) != 0)
476 * If there was no match we are going to append the pattern
477 * if GLOB_NOCHECK was specified or if GLOB_NOMAGIC was specified
478 * and the pattern did not contain any magic characters
479 * GLOB_NOMAGIC is there just for compatibility with csh.
481 if (pglob->gl_pathc == oldpathc) {
482 if ((pglob->gl_flags & GLOB_NOCHECK) ||
483 ((pglob->gl_flags & GLOB_NOMAGIC) &&
484 !(pglob->gl_flags & GLOB_MAGCHAR)))
485 return(globextend(pattern, pglob, &limit));
487 return(GLOB_NOMATCH);
489 if (!(pglob->gl_flags & GLOB_NOSORT))
490 qsort(pglob->gl_pathv + pglob->gl_offs + oldpathc,
491 pglob->gl_pathc - oldpathc, sizeof(char *), compare);
496 compare(const void *p, const void *q)
498 return(strcmp(*(char **)p, *(char **)q));
502 glob1(Char *pattern, Char *pattern_last, glob_t *pglob, size_t *limitp)
504 Char pathbuf[MAXPATHLEN];
506 /* A null pathname is invalid -- POSIX 1003.1 sect. 2.4. */
509 return(glob2(pathbuf, pathbuf+MAXPATHLEN-1,
510 pathbuf, pathbuf+MAXPATHLEN-1,
511 pattern, pattern_last, pglob, limitp));
515 * The functions glob2 and glob3 are mutually recursive; there is one level
516 * of recursion for each segment in the pattern that contains one or more
520 glob2(Char *pathbuf, Char *pathbuf_last, Char *pathend, Char *pathend_last,
521 Char *pattern, Char *pattern_last, glob_t *pglob, size_t *limitp)
528 * Loop over pattern segments until end of pattern or until
529 * segment with meta character found.
531 for (anymeta = 0;;) {
532 if (*pattern == EOS) { /* End of pattern? */
534 if (g_lstat(pathbuf, &sb, pglob))
537 if (((pglob->gl_flags & GLOB_MARK) &&
538 pathend[-1] != SEP) && (S_ISDIR(sb.st_mode) ||
539 (S_ISLNK(sb.st_mode) &&
540 (g_stat(pathbuf, &sb, pglob) == 0) &&
541 S_ISDIR(sb.st_mode)))) {
542 if (pathend+1 > pathend_last)
548 return(globextend(pathbuf, pglob, limitp));
551 /* Find end of next segment, copy tentatively to pathend. */
554 while (*p != EOS && *p != SEP) {
557 if (q+1 > pathend_last)
562 if (!anymeta) { /* No expansion, do next segment. */
565 while (*pattern == SEP) {
566 if (pathend+1 > pathend_last)
568 *pathend++ = *pattern++;
571 /* Need expansion, recurse. */
572 return(glob3(pathbuf, pathbuf_last, pathend,
573 pathend_last, pattern, pattern_last,
574 p, pattern_last, pglob, limitp));
580 glob3(Char *pathbuf, Char *pathbuf_last, Char *pathend, Char *pathend_last,
581 Char *pattern, Char *pattern_last, Char *restpattern,
582 Char *restpattern_last, glob_t *pglob, size_t *limitp)
587 char buf[MAXPATHLEN];
590 * The readdirfunc declaration can't be prototyped, because it is
591 * assigned, below, to two functions which are prototyped in glob.h
592 * and dirent.h as taking pointers to differently typed opaque
595 struct dirent *(*readdirfunc)(void *);
597 if (pathend > pathend_last)
602 if ((dirp = g_opendir(pathbuf, pglob)) == NULL) {
603 /* TODO: don't call for ENOENT or ENOTDIR? */
604 if (pglob->gl_errfunc) {
605 if (g_Ctoc(pathbuf, buf, sizeof(buf)))
606 return(GLOB_ABORTED);
607 if (pglob->gl_errfunc(buf, errno) ||
608 pglob->gl_flags & GLOB_ERR)
609 return(GLOB_ABORTED);
616 /* Search directory for matching names. */
617 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
618 readdirfunc = pglob->gl_readdir;
620 readdirfunc = (struct dirent *(*)(void *))readdir;
621 while ((dp = (*readdirfunc)(dirp))) {
625 /* Initial DOT must be matched literally. */
626 if (dp->d_name[0] == DOT && *pattern != DOT)
629 sc = (u_char *) dp->d_name;
630 while (dc < pathend_last && (*dc++ = *sc++) != EOS)
632 if (dc >= pathend_last) {
638 if (!match(pathend, pattern, restpattern)) {
642 err = glob2(pathbuf, pathbuf_last, --dc, pathend_last,
643 restpattern, restpattern_last, pglob, limitp);
648 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
649 (*pglob->gl_closedir)(dirp);
657 * Extend the gl_pathv member of a glob_t structure to accommodate a new item,
658 * add the new item, and update gl_pathc.
660 * This assumes the BSD realloc, which only copies the block when its size
661 * crosses a power-of-two boundary; for v7 realloc, this would cause quadratic
664 * Return 0 if new item added, error code if memory couldn't be allocated.
666 * Invariant of the glob_t structure:
667 * Either gl_pathc is zero and gl_pathv is NULL; or gl_pathc > 0 and
668 * gl_pathv points to (gl_offs + gl_pathc + 1) items.
671 globextend(const Char *path, glob_t *pglob, size_t *limitp)
679 newsize = sizeof(*pathv) * (2 + pglob->gl_pathc + pglob->gl_offs);
680 pathv = pglob->gl_pathv ? realloc((char *)pglob->gl_pathv, newsize) :
683 if (pglob->gl_pathv) {
684 free(pglob->gl_pathv);
685 pglob->gl_pathv = NULL;
687 return(GLOB_NOSPACE);
690 if (pglob->gl_pathv == NULL && pglob->gl_offs > 0) {
691 /* first time around -- clear initial gl_offs items */
692 pathv += pglob->gl_offs;
693 for (i = pglob->gl_offs; --i >= 0; )
696 pglob->gl_pathv = pathv;
698 for (p = path; *p++;)
700 len = (size_t)(p - path);
702 if ((copy = malloc(len)) != NULL) {
703 if (g_Ctoc(path, copy, len)) {
705 return(GLOB_NOSPACE);
707 pathv[pglob->gl_offs + pglob->gl_pathc++] = copy;
709 pathv[pglob->gl_offs + pglob->gl_pathc] = NULL;
711 if ((pglob->gl_flags & GLOB_LIMIT) &&
712 newsize + *limitp >= (u_int) get_arg_max()) {
714 return(GLOB_NOSPACE);
717 return(copy == NULL ? GLOB_NOSPACE : 0);
722 * pattern matching function for filenames. Each occurrence of the *
723 * pattern causes a recursion level.
726 match(Char *name, Char *pat, Char *patend)
728 int ok, negate_range;
731 while (pat < patend) {
733 switch (c & M_MASK) {
738 if (match(name, pat, patend))
740 } while (*name++ != EOS);
748 if ((k = *name++) == EOS)
750 if ((negate_range = ((*pat & M_MASK) == M_NOT)) != EOS)
752 while (((c = *pat++) & M_MASK) != M_END)
753 if ((*pat & M_MASK) == M_RNG) {
754 if (c <= k && k <= pat[1])
759 if (ok == negate_range)
768 return(*name == EOS);
771 /* Free allocated data belonging to a glob_t structure. */
773 globfree(glob_t *pglob)
778 if (pglob->gl_pathv != NULL) {
779 pp = pglob->gl_pathv + pglob->gl_offs;
780 for (i = pglob->gl_pathc; i--; ++pp)
783 free(pglob->gl_pathv);
784 pglob->gl_pathv = NULL;
789 g_opendir(Char *str, glob_t *pglob)
791 char buf[MAXPATHLEN];
794 strlcpy(buf, ".", sizeof buf);
796 if (g_Ctoc(str, buf, sizeof(buf)))
800 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
801 return((*pglob->gl_opendir)(buf));
803 return(opendir(buf));
807 g_lstat(Char *fn, struct stat *sb, glob_t *pglob)
809 char buf[MAXPATHLEN];
811 if (g_Ctoc(fn, buf, sizeof(buf)))
813 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
814 return((*pglob->gl_lstat)(buf, sb));
815 return(lstat(buf, sb));
819 g_stat(Char *fn, struct stat *sb, glob_t *pglob)
821 char buf[MAXPATHLEN];
823 if (g_Ctoc(fn, buf, sizeof(buf)))
825 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
826 return((*pglob->gl_stat)(buf, sb));
827 return(stat(buf, sb));
831 g_strchr(Char *str, int ch)
841 g_Ctoc(const Char *str, char *buf, u_int len)
845 if ((*buf++ = *str++) == EOS)
853 qprintf(const char *str, Char *s)
857 (void)printf("%s:\n", str);
859 (void)printf("%c", CHAR(*p));
862 (void)printf("%c", *p & M_PROTECT ? '"' : ' ');
865 (void)printf("%c", ismeta(*p) ? '_' : ' ');
870 #endif /* !defined(HAVE_GLOB) || !defined(GLOB_HAS_ALTDIRFUNC) ||
871 !defined(GLOB_HAS_GL_MATCHC) */