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) || \
51 !defined(HAVE_DECL_GLOB_NOMATCH) || HAVE_DECL_GLOB_NOMATCH == 0
58 #elif defined(HAVE_SYSCONF) && defined(_SC_ARG_MAX)
59 return(sysconf(_SC_ARG_MAX));
61 return(256); /* XXX: arbitrary */
66 * glob(3) -- a superset of the one defined in POSIX 1003.2.
68 * The [!...] convention to negate a range is supported (SysV, Posix, ksh).
70 * Optional extra services, controlled by flags not defined by POSIX:
73 * Escaping convention: \ inhibits any special meaning the following
74 * character might have (except \ at end of string is retained).
76 * Set in gl_flags if pattern contained a globbing character.
78 * Same as GLOB_NOCHECK, but it will only append pattern if it did
79 * not contain any magic characters. [Used in csh style globbing]
81 * Use alternately specified directory access functions.
83 * expand ~user/foo to the /home/dir/of/user/foo
85 * expand {1,2}{a,b} to 1a 1b 2a 2b
87 * Number of matches in the current invocation of glob.
102 #undef TILDE /* Some platforms may already define it */
104 #define UNDERSCORE '_'
112 #define M_QUOTE 0x8000
113 #define M_PROTECT 0x4000
114 #define M_MASK 0xffff
115 #define M_ASCII 0x00ff
117 typedef u_short Char;
122 #define M_PROTECT 0x40
131 #define CHAR(c) ((Char)((c)&M_ASCII))
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 *, u_int);
144 static int g_lstat(Char *, struct stat *, glob_t *);
145 static DIR *g_opendir(Char *, glob_t *);
146 static Char *g_strchr(Char *, int);
147 static int g_stat(Char *, struct stat *, glob_t *);
148 static int glob0(const Char *, glob_t *);
149 static int glob1(Char *, Char *, glob_t *, size_t *);
150 static int glob2(Char *, Char *, Char *, Char *, Char *, Char *,
152 static int glob3(Char *, Char *, Char *, Char *, Char *, Char *,
153 Char *, Char *, glob_t *, size_t *);
154 static int globextend(const Char *, glob_t *, size_t *);
156 globtilde(const Char *, Char *, size_t, glob_t *);
157 static int globexp1(const Char *, glob_t *);
158 static int globexp2(const Char *, const Char *, glob_t *, int *);
159 static int match(Char *, Char *, Char *);
161 static void qprintf(const char *, Char *);
165 glob(const char *pattern, int flags, int (*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(const Char *pattern, glob_t *pglob)
216 const Char* ptr = pattern;
219 /* Protect a single {}, for find(1), like csh */
220 if (pattern[0] == LBRACE && pattern[1] == RBRACE && pattern[2] == EOS)
221 return glob0(pattern, pglob);
223 while ((ptr = (const Char *) g_strchr((Char *) ptr, LBRACE)) != NULL)
224 if (!globexp2(ptr, pattern, pglob, &rv))
227 return glob0(pattern, pglob);
232 * Recursive brace globbing helper. Tries to expand a single brace.
233 * If it succeeds then it invokes globexp1 with the new pattern.
234 * If it fails then it tries to glob the rest of the pattern and returns.
237 globexp2(const Char *ptr, const Char *pattern, glob_t *pglob, int *rv)
241 const Char *pe, *pm, *pl;
242 Char patbuf[MAXPATHLEN];
244 /* copy part up to the brace */
245 for (lm = patbuf, pm = pattern; pm != ptr; *lm++ = *pm++)
250 /* Find the balanced brace */
251 for (i = 0, pe = ++ptr; *pe; pe++)
252 if (*pe == LBRACKET) {
253 /* Ignore everything between [] */
254 for (pm = pe++; *pe != RBRACKET && *pe != EOS; pe++)
258 * We could not find a matching RBRACKET.
259 * Ignore and just look for RBRACE
263 } else if (*pe == LBRACE)
265 else if (*pe == RBRACE) {
271 /* Non matching braces; just glob the pattern */
272 if (i != 0 || *pe == EOS) {
273 *rv = glob0(patbuf, pglob);
277 for (i = 0, pl = pm = ptr; pm <= pe; pm++) {
280 /* Ignore everything between [] */
281 for (pl = pm++; *pm != RBRACKET && *pm != EOS; pm++)
285 * We could not find a matching RBRACKET.
286 * Ignore and just look for RBRACE
303 if (i && *pm == COMMA)
306 /* Append the current string */
307 for (lm = ls; (pl < pm); *lm++ = *pl++)
311 * Append the rest of the pattern after the
314 for (pl = pe + 1; (*lm++ = *pl++) != EOS; )
317 /* Expand the current pattern */
319 qprintf("globexp2:", patbuf);
321 *rv = globexp1(patbuf, pglob);
323 /* move after the comma, to the next string */
339 * expand tilde from the passwd file.
342 globtilde(const Char *pattern, Char *patbuf, size_t patbuf_len, glob_t *pglob)
349 if (*pattern != TILDE || !(pglob->gl_flags & GLOB_TILDE))
352 /* Copy up to the end of the string or / */
353 eb = &patbuf[patbuf_len - 1];
354 for (p = pattern + 1, h = (char *) patbuf;
355 h < (char *)eb && *p && *p != SLASH; *h++ = *p++)
365 if (((char *) patbuf)[0] == EOS) {
367 * handle a plain ~ or ~/ by expanding $HOME
368 * first and then trying the password file
371 if (issetugid() != 0 || (h = getenv("HOME")) == NULL) {
373 if ((getuid() != geteuid()) || (h = getenv("HOME")) == NULL) {
374 if ((pwd = getpwuid(getuid())) == NULL)
383 if ((pwd = getpwnam((char*) patbuf)) == NULL)
389 /* Copy the home directory */
390 for (b = patbuf; b < eb && *h; *b++ = *h++)
393 /* Append the rest of the pattern */
394 while (b < eb && (*b++ = *p++) != EOS)
403 * The main glob() routine: compiles the pattern (optionally processing
404 * quotes), calls glob1() to do the real pattern matching, and finally
405 * sorts the list (unless unsorted operation is requested). Returns 0
406 * if things went well, nonzero if errors occurred. It is not an error
407 * to find no matches.
410 glob0(const Char *pattern, glob_t *pglob)
412 const Char *qpatnext;
413 int c, err, oldpathc;
414 Char *bufnext, patbuf[MAXPATHLEN];
417 qpatnext = globtilde(pattern, patbuf, MAXPATHLEN, pglob);
418 oldpathc = pglob->gl_pathc;
421 /* We don't need to check for buffer overflow any more. */
422 while ((c = *qpatnext++) != EOS) {
428 if (*qpatnext == EOS ||
429 g_strchr((Char *) qpatnext+1, RBRACKET) == NULL) {
430 *bufnext++ = LBRACKET;
440 *bufnext++ = CHAR(c);
441 if (*qpatnext == RANGE &&
442 (c = qpatnext[1]) != RBRACKET) {
444 *bufnext++ = CHAR(c);
447 } while ((c = *qpatnext++) != RBRACKET);
448 pglob->gl_flags |= GLOB_MAGCHAR;
452 pglob->gl_flags |= GLOB_MAGCHAR;
456 pglob->gl_flags |= GLOB_MAGCHAR;
457 /* collapse adjacent stars to one,
458 * to avoid exponential behavior
460 if (bufnext == patbuf || bufnext[-1] != M_ALL)
464 *bufnext++ = CHAR(c);
470 qprintf("glob0:", patbuf);
473 if ((err = glob1(patbuf, patbuf+MAXPATHLEN-1, pglob, &limit)) != 0)
477 * If there was no match we are going to append the pattern
478 * if GLOB_NOCHECK was specified or if GLOB_NOMAGIC was specified
479 * and the pattern did not contain any magic characters
480 * GLOB_NOMAGIC is there just for compatibility with csh.
482 if (pglob->gl_pathc == oldpathc) {
483 if ((pglob->gl_flags & GLOB_NOCHECK) ||
484 ((pglob->gl_flags & GLOB_NOMAGIC) &&
485 !(pglob->gl_flags & GLOB_MAGCHAR)))
486 return(globextend(pattern, pglob, &limit));
488 return(GLOB_NOMATCH);
490 if (!(pglob->gl_flags & GLOB_NOSORT))
491 qsort(pglob->gl_pathv + pglob->gl_offs + oldpathc,
492 pglob->gl_pathc - oldpathc, sizeof(char *), compare);
497 compare(const void *p, const void *q)
499 return(strcmp(*(char **)p, *(char **)q));
503 glob1(Char *pattern, Char *pattern_last, glob_t *pglob, size_t *limitp)
505 Char pathbuf[MAXPATHLEN];
507 /* A null pathname is invalid -- POSIX 1003.1 sect. 2.4. */
510 return(glob2(pathbuf, pathbuf+MAXPATHLEN-1,
511 pathbuf, pathbuf+MAXPATHLEN-1,
512 pattern, pattern_last, pglob, limitp));
516 * The functions glob2 and glob3 are mutually recursive; there is one level
517 * of recursion for each segment in the pattern that contains one or more
521 glob2(Char *pathbuf, Char *pathbuf_last, Char *pathend, Char *pathend_last,
522 Char *pattern, Char *pattern_last, glob_t *pglob, size_t *limitp)
529 * Loop over pattern segments until end of pattern or until
530 * segment with meta character found.
532 for (anymeta = 0;;) {
533 if (*pattern == EOS) { /* End of pattern? */
535 if (g_lstat(pathbuf, &sb, pglob))
538 if (((pglob->gl_flags & GLOB_MARK) &&
539 pathend[-1] != SEP) && (S_ISDIR(sb.st_mode) ||
540 (S_ISLNK(sb.st_mode) &&
541 (g_stat(pathbuf, &sb, pglob) == 0) &&
542 S_ISDIR(sb.st_mode)))) {
543 if (pathend+1 > pathend_last)
549 return(globextend(pathbuf, pglob, limitp));
552 /* Find end of next segment, copy tentatively to pathend. */
555 while (*p != EOS && *p != SEP) {
558 if (q+1 > pathend_last)
563 if (!anymeta) { /* No expansion, do next segment. */
566 while (*pattern == SEP) {
567 if (pathend+1 > pathend_last)
569 *pathend++ = *pattern++;
572 /* Need expansion, recurse. */
573 return(glob3(pathbuf, pathbuf_last, pathend,
574 pathend_last, pattern, pattern_last,
575 p, pattern_last, pglob, limitp));
581 glob3(Char *pathbuf, Char *pathbuf_last, Char *pathend, Char *pathend_last,
582 Char *pattern, Char *pattern_last, Char *restpattern,
583 Char *restpattern_last, glob_t *pglob, size_t *limitp)
588 char buf[MAXPATHLEN];
591 * The readdirfunc declaration can't be prototyped, because it is
592 * assigned, below, to two functions which are prototyped in glob.h
593 * and dirent.h as taking pointers to differently typed opaque
596 struct dirent *(*readdirfunc)(void *);
598 if (pathend > pathend_last)
603 if ((dirp = g_opendir(pathbuf, pglob)) == NULL) {
604 /* TODO: don't call for ENOENT or ENOTDIR? */
605 if (pglob->gl_errfunc) {
606 if (g_Ctoc(pathbuf, buf, sizeof(buf)))
607 return(GLOB_ABORTED);
608 if (pglob->gl_errfunc(buf, errno) ||
609 pglob->gl_flags & GLOB_ERR)
610 return(GLOB_ABORTED);
617 /* Search directory for matching names. */
618 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
619 readdirfunc = pglob->gl_readdir;
621 readdirfunc = (struct dirent *(*)(void *))readdir;
622 while ((dp = (*readdirfunc)(dirp))) {
626 /* Initial DOT must be matched literally. */
627 if (dp->d_name[0] == DOT && *pattern != DOT)
630 sc = (u_char *) dp->d_name;
631 while (dc < pathend_last && (*dc++ = *sc++) != EOS)
633 if (dc >= pathend_last) {
639 if (!match(pathend, pattern, restpattern)) {
643 err = glob2(pathbuf, pathbuf_last, --dc, pathend_last,
644 restpattern, restpattern_last, pglob, limitp);
649 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
650 (*pglob->gl_closedir)(dirp);
658 * Extend the gl_pathv member of a glob_t structure to accommodate a new item,
659 * add the new item, and update gl_pathc.
661 * This assumes the BSD realloc, which only copies the block when its size
662 * crosses a power-of-two boundary; for v7 realloc, this would cause quadratic
665 * Return 0 if new item added, error code if memory couldn't be allocated.
667 * Invariant of the glob_t structure:
668 * Either gl_pathc is zero and gl_pathv is NULL; or gl_pathc > 0 and
669 * gl_pathv points to (gl_offs + gl_pathc + 1) items.
672 globextend(const Char *path, glob_t *pglob, size_t *limitp)
680 newsize = sizeof(*pathv) * (2 + pglob->gl_pathc + pglob->gl_offs);
681 pathv = pglob->gl_pathv ? realloc((char *)pglob->gl_pathv, newsize) :
684 if (pglob->gl_pathv) {
685 free(pglob->gl_pathv);
686 pglob->gl_pathv = NULL;
688 return(GLOB_NOSPACE);
691 if (pglob->gl_pathv == NULL && pglob->gl_offs > 0) {
692 /* first time around -- clear initial gl_offs items */
693 pathv += pglob->gl_offs;
694 for (i = pglob->gl_offs; --i >= 0; )
697 pglob->gl_pathv = pathv;
699 for (p = path; *p++;)
701 len = (size_t)(p - path);
703 if ((copy = malloc(len)) != NULL) {
704 if (g_Ctoc(path, copy, len)) {
706 return(GLOB_NOSPACE);
708 pathv[pglob->gl_offs + pglob->gl_pathc++] = copy;
710 pathv[pglob->gl_offs + pglob->gl_pathc] = NULL;
712 if ((pglob->gl_flags & GLOB_LIMIT) &&
713 newsize + *limitp >= (u_int) get_arg_max()) {
715 return(GLOB_NOSPACE);
718 return(copy == NULL ? GLOB_NOSPACE : 0);
723 * pattern matching function for filenames. Each occurrence of the *
724 * pattern causes a recursion level.
727 match(Char *name, Char *pat, Char *patend)
729 int ok, negate_range;
732 while (pat < patend) {
734 switch (c & M_MASK) {
739 if (match(name, pat, patend))
741 } while (*name++ != EOS);
749 if ((k = *name++) == EOS)
751 if ((negate_range = ((*pat & M_MASK) == M_NOT)) != EOS)
753 while (((c = *pat++) & M_MASK) != M_END)
754 if ((*pat & M_MASK) == M_RNG) {
755 if (c <= k && k <= pat[1])
760 if (ok == negate_range)
769 return(*name == EOS);
772 /* Free allocated data belonging to a glob_t structure. */
774 globfree(glob_t *pglob)
779 if (pglob->gl_pathv != NULL) {
780 pp = pglob->gl_pathv + pglob->gl_offs;
781 for (i = pglob->gl_pathc; i--; ++pp)
784 free(pglob->gl_pathv);
785 pglob->gl_pathv = NULL;
790 g_opendir(Char *str, glob_t *pglob)
792 char buf[MAXPATHLEN];
795 strlcpy(buf, ".", sizeof buf);
797 if (g_Ctoc(str, buf, sizeof(buf)))
801 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
802 return((*pglob->gl_opendir)(buf));
804 return(opendir(buf));
808 g_lstat(Char *fn, struct stat *sb, glob_t *pglob)
810 char buf[MAXPATHLEN];
812 if (g_Ctoc(fn, buf, sizeof(buf)))
814 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
815 return((*pglob->gl_lstat)(buf, sb));
816 return(lstat(buf, sb));
820 g_stat(Char *fn, struct stat *sb, glob_t *pglob)
822 char buf[MAXPATHLEN];
824 if (g_Ctoc(fn, buf, sizeof(buf)))
826 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
827 return((*pglob->gl_stat)(buf, sb));
828 return(stat(buf, sb));
832 g_strchr(Char *str, int ch)
842 g_Ctoc(const Char *str, char *buf, u_int len)
846 if ((*buf++ = *str++) == EOS)
854 qprintf(const char *str, Char *s)
858 (void)printf("%s:\n", str);
860 (void)printf("%c", CHAR(*p));
863 (void)printf("%c", *p & M_PROTECT ? '"' : ' ');
866 (void)printf("%c", ismeta(*p) ? '_' : ' ');
871 #endif /* !defined(HAVE_GLOB) || !defined(GLOB_HAS_ALTDIRFUNC) ||
872 !defined(GLOB_HAS_GL_MATCHC) */