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>
47 #if !defined(HAVE_GLOB) || !defined(GLOB_HAS_ALTDIRFUNC) || \
48 !defined(GLOB_HAS_GL_MATCHC)
55 #elif defined(HAVE_SYSCONF) && defined(_SC_ARG_MAX)
56 return(sysconf(_SC_ARG_MAX));
58 return(256); /* XXX: arbitrary */
63 * glob(3) -- a superset of the one defined in POSIX 1003.2.
65 * The [!...] convention to negate a range is supported (SysV, Posix, ksh).
67 * Optional extra services, controlled by flags not defined by POSIX:
70 * Escaping convention: \ inhibits any special meaning the following
71 * character might have (except \ at end of string is retained).
73 * Set in gl_flags if pattern contained a globbing character.
75 * Same as GLOB_NOCHECK, but it will only append pattern if it did
76 * not contain any magic characters. [Used in csh style globbing]
78 * Use alternately specified directory access functions.
80 * expand ~user/foo to the /home/dir/of/user/foo
82 * expand {1,2}{a,b} to 1a 1b 2a 2b
84 * Number of matches in the current invocation of glob.
99 #undef TILDE /* Some platforms may already define it */
101 #define UNDERSCORE '_'
109 #define M_QUOTE 0x8000
110 #define M_PROTECT 0x4000
111 #define M_MASK 0xffff
112 #define M_ASCII 0x00ff
114 typedef u_short Char;
119 #define M_PROTECT 0x40
128 #define CHAR(c) ((Char)((c)&M_ASCII))
129 #define META(c) ((Char)((c)|M_QUOTE))
130 #define M_ALL META('*')
131 #define M_END META(']')
132 #define M_NOT META('!')
133 #define M_ONE META('?')
134 #define M_RNG META('-')
135 #define M_SET META('[')
136 #define ismeta(c) (((c)&M_QUOTE) != 0)
139 static int compare(const void *, const void *);
140 static int g_Ctoc(const Char *, char *, u_int);
141 static int g_lstat(Char *, struct stat *, glob_t *);
142 static DIR *g_opendir(Char *, glob_t *);
143 static Char *g_strchr(Char *, int);
144 static int g_stat(Char *, struct stat *, glob_t *);
145 static int glob0(const Char *, glob_t *);
146 static int glob1(Char *, Char *, glob_t *, size_t *);
147 static int glob2(Char *, Char *, Char *, Char *, Char *, Char *,
149 static int glob3(Char *, Char *, Char *, Char *, Char *, Char *,
150 Char *, Char *, glob_t *, size_t *);
151 static int globextend(const Char *, glob_t *, size_t *);
153 globtilde(const Char *, Char *, size_t, glob_t *);
154 static int globexp1(const Char *, glob_t *);
155 static int globexp2(const Char *, const Char *, glob_t *, int *);
156 static int match(Char *, Char *, Char *);
158 static void qprintf(const char *, Char *);
162 glob(const char *pattern, int flags, int (*errfunc)(const char *, int),
165 const u_char *patnext;
167 Char *bufnext, *bufend, patbuf[MAXPATHLEN];
169 patnext = (u_char *) pattern;
170 if (!(flags & GLOB_APPEND)) {
172 pglob->gl_pathv = NULL;
173 if (!(flags & GLOB_DOOFFS))
176 pglob->gl_flags = flags & ~GLOB_MAGCHAR;
177 pglob->gl_errfunc = errfunc;
178 pglob->gl_matchc = 0;
181 bufend = bufnext + MAXPATHLEN - 1;
182 if (flags & GLOB_NOESCAPE)
183 while (bufnext < bufend && (c = *patnext++) != EOS)
186 /* Protect the quoted characters. */
187 while (bufnext < bufend && (c = *patnext++) != EOS)
189 if ((c = *patnext++) == EOS) {
193 *bufnext++ = c | M_PROTECT;
199 if (flags & GLOB_BRACE)
200 return globexp1(patbuf, pglob);
202 return glob0(patbuf, pglob);
206 * Expand recursively a glob {} pattern. When there is no more expansion
207 * invoke the standard globbing routine to glob the rest of the magic
211 globexp1(const Char *pattern, glob_t *pglob)
213 const Char* ptr = pattern;
216 /* Protect a single {}, for find(1), like csh */
217 if (pattern[0] == LBRACE && pattern[1] == RBRACE && pattern[2] == EOS)
218 return glob0(pattern, pglob);
220 while ((ptr = (const Char *) g_strchr((Char *) ptr, LBRACE)) != NULL)
221 if (!globexp2(ptr, pattern, pglob, &rv))
224 return glob0(pattern, pglob);
229 * Recursive brace globbing helper. Tries to expand a single brace.
230 * If it succeeds then it invokes globexp1 with the new pattern.
231 * If it fails then it tries to glob the rest of the pattern and returns.
234 globexp2(const Char *ptr, const Char *pattern, glob_t *pglob, int *rv)
238 const Char *pe, *pm, *pl;
239 Char patbuf[MAXPATHLEN];
241 /* copy part up to the brace */
242 for (lm = patbuf, pm = pattern; pm != ptr; *lm++ = *pm++)
247 /* Find the balanced brace */
248 for (i = 0, pe = ++ptr; *pe; pe++)
249 if (*pe == LBRACKET) {
250 /* Ignore everything between [] */
251 for (pm = pe++; *pe != RBRACKET && *pe != EOS; pe++)
255 * We could not find a matching RBRACKET.
256 * Ignore and just look for RBRACE
260 } else if (*pe == LBRACE)
262 else if (*pe == RBRACE) {
268 /* Non matching braces; just glob the pattern */
269 if (i != 0 || *pe == EOS) {
270 *rv = glob0(patbuf, pglob);
274 for (i = 0, pl = pm = ptr; pm <= pe; pm++) {
277 /* Ignore everything between [] */
278 for (pl = pm++; *pm != RBRACKET && *pm != EOS; pm++)
282 * We could not find a matching RBRACKET.
283 * Ignore and just look for RBRACE
300 if (i && *pm == COMMA)
303 /* Append the current string */
304 for (lm = ls; (pl < pm); *lm++ = *pl++)
308 * Append the rest of the pattern after the
311 for (pl = pe + 1; (*lm++ = *pl++) != EOS; )
314 /* Expand the current pattern */
316 qprintf("globexp2:", patbuf);
318 *rv = globexp1(patbuf, pglob);
320 /* move after the comma, to the next string */
336 * expand tilde from the passwd file.
339 globtilde(const Char *pattern, Char *patbuf, size_t patbuf_len, glob_t *pglob)
346 if (*pattern != TILDE || !(pglob->gl_flags & GLOB_TILDE))
349 /* Copy up to the end of the string or / */
350 eb = &patbuf[patbuf_len - 1];
351 for (p = pattern + 1, h = (char *) patbuf;
352 h < (char *)eb && *p && *p != SLASH; *h++ = *p++)
362 if (((char *) patbuf)[0] == EOS) {
364 * handle a plain ~ or ~/ by expanding $HOME
365 * first and then trying the password file
368 if (issetugid() != 0 || (h = getenv("HOME")) == NULL) {
370 if ((getuid() != geteuid()) || (h = getenv("HOME")) == NULL) {
371 if ((pwd = getpwuid(getuid())) == NULL)
380 if ((pwd = getpwnam((char*) patbuf)) == NULL)
386 /* Copy the home directory */
387 for (b = patbuf; b < eb && *h; *b++ = *h++)
390 /* Append the rest of the pattern */
391 while (b < eb && (*b++ = *p++) != EOS)
400 * The main glob() routine: compiles the pattern (optionally processing
401 * quotes), calls glob1() to do the real pattern matching, and finally
402 * sorts the list (unless unsorted operation is requested). Returns 0
403 * if things went well, nonzero if errors occurred. It is not an error
404 * to find no matches.
407 glob0(const Char *pattern, glob_t *pglob)
409 const Char *qpatnext;
410 int c, err, oldpathc;
411 Char *bufnext, patbuf[MAXPATHLEN];
414 qpatnext = globtilde(pattern, patbuf, MAXPATHLEN, pglob);
415 oldpathc = pglob->gl_pathc;
418 /* We don't need to check for buffer overflow any more. */
419 while ((c = *qpatnext++) != EOS) {
425 if (*qpatnext == EOS ||
426 g_strchr((Char *) qpatnext+1, RBRACKET) == NULL) {
427 *bufnext++ = LBRACKET;
437 *bufnext++ = CHAR(c);
438 if (*qpatnext == RANGE &&
439 (c = qpatnext[1]) != RBRACKET) {
441 *bufnext++ = CHAR(c);
444 } while ((c = *qpatnext++) != RBRACKET);
445 pglob->gl_flags |= GLOB_MAGCHAR;
449 pglob->gl_flags |= GLOB_MAGCHAR;
453 pglob->gl_flags |= GLOB_MAGCHAR;
454 /* collapse adjacent stars to one,
455 * to avoid exponential behavior
457 if (bufnext == patbuf || bufnext[-1] != M_ALL)
461 *bufnext++ = CHAR(c);
467 qprintf("glob0:", patbuf);
470 if ((err = glob1(patbuf, patbuf+MAXPATHLEN-1, pglob, &limit)) != 0)
474 * If there was no match we are going to append the pattern
475 * if GLOB_NOCHECK was specified or if GLOB_NOMAGIC was specified
476 * and the pattern did not contain any magic characters
477 * GLOB_NOMAGIC is there just for compatibility with csh.
479 if (pglob->gl_pathc == oldpathc) {
480 if ((pglob->gl_flags & GLOB_NOCHECK) ||
481 ((pglob->gl_flags & GLOB_NOMAGIC) &&
482 !(pglob->gl_flags & GLOB_MAGCHAR)))
483 return(globextend(pattern, pglob, &limit));
485 return(GLOB_NOMATCH);
487 if (!(pglob->gl_flags & GLOB_NOSORT))
488 qsort(pglob->gl_pathv + pglob->gl_offs + oldpathc,
489 pglob->gl_pathc - oldpathc, sizeof(char *), compare);
494 compare(const void *p, const void *q)
496 return(strcmp(*(char **)p, *(char **)q));
500 glob1(Char *pattern, Char *pattern_last, glob_t *pglob, size_t *limitp)
502 Char pathbuf[MAXPATHLEN];
504 /* A null pathname is invalid -- POSIX 1003.1 sect. 2.4. */
507 return(glob2(pathbuf, pathbuf+MAXPATHLEN-1,
508 pathbuf, pathbuf+MAXPATHLEN-1,
509 pattern, pattern_last, pglob, limitp));
513 * The functions glob2 and glob3 are mutually recursive; there is one level
514 * of recursion for each segment in the pattern that contains one or more
518 glob2(Char *pathbuf, Char *pathbuf_last, Char *pathend, Char *pathend_last,
519 Char *pattern, Char *pattern_last, glob_t *pglob, size_t *limitp)
526 * Loop over pattern segments until end of pattern or until
527 * segment with meta character found.
529 for (anymeta = 0;;) {
530 if (*pattern == EOS) { /* End of pattern? */
532 if (g_lstat(pathbuf, &sb, pglob))
535 if (((pglob->gl_flags & GLOB_MARK) &&
536 pathend[-1] != SEP) && (S_ISDIR(sb.st_mode) ||
537 (S_ISLNK(sb.st_mode) &&
538 (g_stat(pathbuf, &sb, pglob) == 0) &&
539 S_ISDIR(sb.st_mode)))) {
540 if (pathend+1 > pathend_last)
546 return(globextend(pathbuf, pglob, limitp));
549 /* Find end of next segment, copy tentatively to pathend. */
552 while (*p != EOS && *p != SEP) {
555 if (q+1 > pathend_last)
560 if (!anymeta) { /* No expansion, do next segment. */
563 while (*pattern == SEP) {
564 if (pathend+1 > pathend_last)
566 *pathend++ = *pattern++;
569 /* Need expansion, recurse. */
570 return(glob3(pathbuf, pathbuf_last, pathend,
571 pathend_last, pattern, pattern_last,
572 p, pattern_last, pglob, limitp));
578 glob3(Char *pathbuf, Char *pathbuf_last, Char *pathend, Char *pathend_last,
579 Char *pattern, Char *pattern_last, Char *restpattern,
580 Char *restpattern_last, glob_t *pglob, size_t *limitp)
585 char buf[MAXPATHLEN];
588 * The readdirfunc declaration can't be prototyped, because it is
589 * assigned, below, to two functions which are prototyped in glob.h
590 * and dirent.h as taking pointers to differently typed opaque
593 struct dirent *(*readdirfunc)(void *);
595 if (pathend > pathend_last)
600 if ((dirp = g_opendir(pathbuf, pglob)) == NULL) {
601 /* TODO: don't call for ENOENT or ENOTDIR? */
602 if (pglob->gl_errfunc) {
603 if (g_Ctoc(pathbuf, buf, sizeof(buf)))
604 return(GLOB_ABORTED);
605 if (pglob->gl_errfunc(buf, errno) ||
606 pglob->gl_flags & GLOB_ERR)
607 return(GLOB_ABORTED);
614 /* Search directory for matching names. */
615 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
616 readdirfunc = pglob->gl_readdir;
618 readdirfunc = (struct dirent *(*)(void *))readdir;
619 while ((dp = (*readdirfunc)(dirp))) {
623 /* Initial DOT must be matched literally. */
624 if (dp->d_name[0] == DOT && *pattern != DOT)
627 sc = (u_char *) dp->d_name;
628 while (dc < pathend_last && (*dc++ = *sc++) != EOS)
630 if (dc >= pathend_last) {
636 if (!match(pathend, pattern, restpattern)) {
640 err = glob2(pathbuf, pathbuf_last, --dc, pathend_last,
641 restpattern, restpattern_last, pglob, limitp);
646 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
647 (*pglob->gl_closedir)(dirp);
655 * Extend the gl_pathv member of a glob_t structure to accommodate a new item,
656 * add the new item, and update gl_pathc.
658 * This assumes the BSD realloc, which only copies the block when its size
659 * crosses a power-of-two boundary; for v7 realloc, this would cause quadratic
662 * Return 0 if new item added, error code if memory couldn't be allocated.
664 * Invariant of the glob_t structure:
665 * Either gl_pathc is zero and gl_pathv is NULL; or gl_pathc > 0 and
666 * gl_pathv points to (gl_offs + gl_pathc + 1) items.
669 globextend(const Char *path, glob_t *pglob, size_t *limitp)
677 newsize = sizeof(*pathv) * (2 + pglob->gl_pathc + pglob->gl_offs);
678 pathv = pglob->gl_pathv ? realloc((char *)pglob->gl_pathv, newsize) :
681 if (pglob->gl_pathv) {
682 free(pglob->gl_pathv);
683 pglob->gl_pathv = NULL;
685 return(GLOB_NOSPACE);
688 if (pglob->gl_pathv == NULL && pglob->gl_offs > 0) {
689 /* first time around -- clear initial gl_offs items */
690 pathv += pglob->gl_offs;
691 for (i = pglob->gl_offs; --i >= 0; )
694 pglob->gl_pathv = pathv;
696 for (p = path; *p++;)
698 len = (size_t)(p - path);
700 if ((copy = malloc(len)) != NULL) {
701 if (g_Ctoc(path, copy, len)) {
703 return(GLOB_NOSPACE);
705 pathv[pglob->gl_offs + pglob->gl_pathc++] = copy;
707 pathv[pglob->gl_offs + pglob->gl_pathc] = NULL;
709 if ((pglob->gl_flags & GLOB_LIMIT) &&
710 newsize + *limitp >= (u_int) get_arg_max()) {
712 return(GLOB_NOSPACE);
715 return(copy == NULL ? GLOB_NOSPACE : 0);
720 * pattern matching function for filenames. Each occurrence of the *
721 * pattern causes a recursion level.
724 match(Char *name, Char *pat, Char *patend)
726 int ok, negate_range;
729 while (pat < patend) {
731 switch (c & M_MASK) {
736 if (match(name, pat, patend))
738 } while (*name++ != EOS);
746 if ((k = *name++) == EOS)
748 if ((negate_range = ((*pat & M_MASK) == M_NOT)) != EOS)
750 while (((c = *pat++) & M_MASK) != M_END)
751 if ((*pat & M_MASK) == M_RNG) {
752 if (c <= k && k <= pat[1])
757 if (ok == negate_range)
766 return(*name == EOS);
769 /* Free allocated data belonging to a glob_t structure. */
771 globfree(glob_t *pglob)
776 if (pglob->gl_pathv != NULL) {
777 pp = pglob->gl_pathv + pglob->gl_offs;
778 for (i = pglob->gl_pathc; i--; ++pp)
781 free(pglob->gl_pathv);
782 pglob->gl_pathv = NULL;
787 g_opendir(Char *str, glob_t *pglob)
789 char buf[MAXPATHLEN];
792 strlcpy(buf, ".", sizeof buf);
794 if (g_Ctoc(str, buf, sizeof(buf)))
798 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
799 return((*pglob->gl_opendir)(buf));
801 return(opendir(buf));
805 g_lstat(Char *fn, struct stat *sb, glob_t *pglob)
807 char buf[MAXPATHLEN];
809 if (g_Ctoc(fn, buf, sizeof(buf)))
811 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
812 return((*pglob->gl_lstat)(buf, sb));
813 return(lstat(buf, sb));
817 g_stat(Char *fn, struct stat *sb, glob_t *pglob)
819 char buf[MAXPATHLEN];
821 if (g_Ctoc(fn, buf, sizeof(buf)))
823 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
824 return((*pglob->gl_stat)(buf, sb));
825 return(stat(buf, sb));
829 g_strchr(Char *str, int ch)
839 g_Ctoc(const Char *str, char *buf, u_int len)
843 if ((*buf++ = *str++) == EOS)
851 qprintf(const char *str, Char *s)
855 (void)printf("%s:\n", str);
857 (void)printf("%c", CHAR(*p));
860 (void)printf("%c", *p & M_PROTECT ? '"' : ' ');
863 (void)printf("%c", ismeta(*p) ? '_' : ' ');
868 #endif /* !defined(HAVE_GLOB) || !defined(GLOB_HAS_ALTDIRFUNC) ||
869 !defined(GLOB_HAS_GL_MATCHC) */