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 */
44 #elif defined(HAVE_SYSCONF) && defined(_SC_ARG_MAX)
45 return(sysconf(_SC_ARG_MAX));
47 return(256); /* XXX: arbitrary */
51 #if !defined(HAVE_GLOB) || !defined(GLOB_HAS_ALTDIRFUNC) || \
52 !defined(GLOB_HAS_GL_MATCHC)
55 * glob(3) -- a superset of the one defined in POSIX 1003.2.
57 * The [!...] convention to negate a range is supported (SysV, Posix, ksh).
59 * Optional extra services, controlled by flags not defined by POSIX:
62 * Escaping convention: \ inhibits any special meaning the following
63 * character might have (except \ at end of string is retained).
65 * Set in gl_flags if pattern contained a globbing character.
67 * Same as GLOB_NOCHECK, but it will only append pattern if it did
68 * not contain any magic characters. [Used in csh style globbing]
70 * Use alternately specified directory access functions.
72 * expand ~user/foo to the /home/dir/of/user/foo
74 * expand {1,2}{a,b} to 1a 1b 2a 2b
76 * Number of matches in the current invocation of glob.
91 #undef TILDE /* Some platforms may already define it */
93 #define UNDERSCORE '_'
101 #define M_QUOTE 0x8000
102 #define M_PROTECT 0x4000
103 #define M_MASK 0xffff
104 #define M_ASCII 0x00ff
106 typedef u_short Char;
111 #define M_PROTECT 0x40
120 #define CHAR(c) ((Char)((c)&M_ASCII))
121 #define META(c) ((Char)((c)|M_QUOTE))
122 #define M_ALL META('*')
123 #define M_END META(']')
124 #define M_NOT META('!')
125 #define M_ONE META('?')
126 #define M_RNG META('-')
127 #define M_SET META('[')
128 #define ismeta(c) (((c)&M_QUOTE) != 0)
131 static int compare(const void *, const void *);
132 static int g_Ctoc(const Char *, char *, u_int);
133 static int g_lstat(Char *, struct stat *, glob_t *);
134 static DIR *g_opendir(Char *, glob_t *);
135 static Char *g_strchr(Char *, int);
136 static int g_stat(Char *, struct stat *, glob_t *);
137 static int glob0(const Char *, glob_t *);
138 static int glob1(Char *, Char *, glob_t *, size_t *);
139 static int glob2(Char *, Char *, Char *, Char *, Char *, Char *,
141 static int glob3(Char *, Char *, Char *, Char *, Char *, Char *,
142 Char *, Char *, glob_t *, size_t *);
143 static int globextend(const Char *, glob_t *, size_t *);
145 globtilde(const Char *, Char *, size_t, glob_t *);
146 static int globexp1(const Char *, glob_t *);
147 static int globexp2(const Char *, const Char *, glob_t *, int *);
148 static int match(Char *, Char *, Char *);
150 static void qprintf(const char *, Char *);
154 glob(const char *pattern, int flags, int (*errfunc)(const char *, int),
157 const u_char *patnext;
159 Char *bufnext, *bufend, patbuf[MAXPATHLEN];
161 patnext = (u_char *) pattern;
162 if (!(flags & GLOB_APPEND)) {
164 pglob->gl_pathv = NULL;
165 if (!(flags & GLOB_DOOFFS))
168 pglob->gl_flags = flags & ~GLOB_MAGCHAR;
169 pglob->gl_errfunc = errfunc;
170 pglob->gl_matchc = 0;
173 bufend = bufnext + MAXPATHLEN - 1;
174 if (flags & GLOB_NOESCAPE)
175 while (bufnext < bufend && (c = *patnext++) != EOS)
178 /* Protect the quoted characters. */
179 while (bufnext < bufend && (c = *patnext++) != EOS)
181 if ((c = *patnext++) == EOS) {
185 *bufnext++ = c | M_PROTECT;
191 if (flags & GLOB_BRACE)
192 return globexp1(patbuf, pglob);
194 return glob0(patbuf, pglob);
198 * Expand recursively a glob {} pattern. When there is no more expansion
199 * invoke the standard globbing routine to glob the rest of the magic
203 globexp1(const Char *pattern, glob_t *pglob)
205 const Char* ptr = pattern;
208 /* Protect a single {}, for find(1), like csh */
209 if (pattern[0] == LBRACE && pattern[1] == RBRACE && pattern[2] == EOS)
210 return glob0(pattern, pglob);
212 while ((ptr = (const Char *) g_strchr((Char *) ptr, LBRACE)) != NULL)
213 if (!globexp2(ptr, pattern, pglob, &rv))
216 return glob0(pattern, pglob);
221 * Recursive brace globbing helper. Tries to expand a single brace.
222 * If it succeeds then it invokes globexp1 with the new pattern.
223 * If it fails then it tries to glob the rest of the pattern and returns.
226 globexp2(const Char *ptr, const Char *pattern, glob_t *pglob, int *rv)
230 const Char *pe, *pm, *pl;
231 Char patbuf[MAXPATHLEN];
233 /* copy part up to the brace */
234 for (lm = patbuf, pm = pattern; pm != ptr; *lm++ = *pm++)
239 /* Find the balanced brace */
240 for (i = 0, pe = ++ptr; *pe; pe++)
241 if (*pe == LBRACKET) {
242 /* Ignore everything between [] */
243 for (pm = pe++; *pe != RBRACKET && *pe != EOS; pe++)
247 * We could not find a matching RBRACKET.
248 * Ignore and just look for RBRACE
252 } else if (*pe == LBRACE)
254 else if (*pe == RBRACE) {
260 /* Non matching braces; just glob the pattern */
261 if (i != 0 || *pe == EOS) {
262 *rv = glob0(patbuf, pglob);
266 for (i = 0, pl = pm = ptr; pm <= pe; pm++) {
269 /* Ignore everything between [] */
270 for (pl = pm++; *pm != RBRACKET && *pm != EOS; pm++)
274 * We could not find a matching RBRACKET.
275 * Ignore and just look for RBRACE
292 if (i && *pm == COMMA)
295 /* Append the current string */
296 for (lm = ls; (pl < pm); *lm++ = *pl++)
300 * Append the rest of the pattern after the
303 for (pl = pe + 1; (*lm++ = *pl++) != EOS; )
306 /* Expand the current pattern */
308 qprintf("globexp2:", patbuf);
310 *rv = globexp1(patbuf, pglob);
312 /* move after the comma, to the next string */
328 * expand tilde from the passwd file.
331 globtilde(const Char *pattern, Char *patbuf, size_t patbuf_len, glob_t *pglob)
338 if (*pattern != TILDE || !(pglob->gl_flags & GLOB_TILDE))
341 /* Copy up to the end of the string or / */
342 eb = &patbuf[patbuf_len - 1];
343 for (p = pattern + 1, h = (char *) patbuf;
344 h < (char *)eb && *p && *p != SLASH; *h++ = *p++)
354 if (((char *) patbuf)[0] == EOS) {
356 * handle a plain ~ or ~/ by expanding $HOME
357 * first and then trying the password file
360 if (issetugid() != 0 || (h = getenv("HOME")) == NULL) {
362 if ((getuid() != geteuid()) || (h = getenv("HOME")) == NULL) {
363 if ((pwd = getpwuid(getuid())) == NULL)
372 if ((pwd = getpwnam((char*) patbuf)) == NULL)
378 /* Copy the home directory */
379 for (b = patbuf; b < eb && *h; *b++ = *h++)
382 /* Append the rest of the pattern */
383 while (b < eb && (*b++ = *p++) != EOS)
392 * The main glob() routine: compiles the pattern (optionally processing
393 * quotes), calls glob1() to do the real pattern matching, and finally
394 * sorts the list (unless unsorted operation is requested). Returns 0
395 * if things went well, nonzero if errors occurred. It is not an error
396 * to find no matches.
399 glob0(const Char *pattern, glob_t *pglob)
401 const Char *qpatnext;
402 int c, err, oldpathc;
403 Char *bufnext, patbuf[MAXPATHLEN];
406 qpatnext = globtilde(pattern, patbuf, MAXPATHLEN, pglob);
407 oldpathc = pglob->gl_pathc;
410 /* We don't need to check for buffer overflow any more. */
411 while ((c = *qpatnext++) != EOS) {
417 if (*qpatnext == EOS ||
418 g_strchr((Char *) qpatnext+1, RBRACKET) == NULL) {
419 *bufnext++ = LBRACKET;
429 *bufnext++ = CHAR(c);
430 if (*qpatnext == RANGE &&
431 (c = qpatnext[1]) != RBRACKET) {
433 *bufnext++ = CHAR(c);
436 } while ((c = *qpatnext++) != RBRACKET);
437 pglob->gl_flags |= GLOB_MAGCHAR;
441 pglob->gl_flags |= GLOB_MAGCHAR;
445 pglob->gl_flags |= GLOB_MAGCHAR;
446 /* collapse adjacent stars to one,
447 * to avoid exponential behavior
449 if (bufnext == patbuf || bufnext[-1] != M_ALL)
453 *bufnext++ = CHAR(c);
459 qprintf("glob0:", patbuf);
462 if ((err = glob1(patbuf, patbuf+MAXPATHLEN-1, pglob, &limit)) != 0)
466 * If there was no match we are going to append the pattern
467 * if GLOB_NOCHECK was specified or if GLOB_NOMAGIC was specified
468 * and the pattern did not contain any magic characters
469 * GLOB_NOMAGIC is there just for compatibility with csh.
471 if (pglob->gl_pathc == oldpathc) {
472 if ((pglob->gl_flags & GLOB_NOCHECK) ||
473 ((pglob->gl_flags & GLOB_NOMAGIC) &&
474 !(pglob->gl_flags & GLOB_MAGCHAR)))
475 return(globextend(pattern, pglob, &limit));
477 return(GLOB_NOMATCH);
479 if (!(pglob->gl_flags & GLOB_NOSORT))
480 qsort(pglob->gl_pathv + pglob->gl_offs + oldpathc,
481 pglob->gl_pathc - oldpathc, sizeof(char *), compare);
486 compare(const void *p, const void *q)
488 return(strcmp(*(char **)p, *(char **)q));
492 glob1(Char *pattern, Char *pattern_last, glob_t *pglob, size_t *limitp)
494 Char pathbuf[MAXPATHLEN];
496 /* A null pathname is invalid -- POSIX 1003.1 sect. 2.4. */
499 return(glob2(pathbuf, pathbuf+MAXPATHLEN-1,
500 pathbuf, pathbuf+MAXPATHLEN-1,
501 pattern, pattern_last, pglob, limitp));
505 * The functions glob2 and glob3 are mutually recursive; there is one level
506 * of recursion for each segment in the pattern that contains one or more
510 glob2(Char *pathbuf, Char *pathbuf_last, Char *pathend, Char *pathend_last,
511 Char *pattern, Char *pattern_last, glob_t *pglob, size_t *limitp)
518 * Loop over pattern segments until end of pattern or until
519 * segment with meta character found.
521 for (anymeta = 0;;) {
522 if (*pattern == EOS) { /* End of pattern? */
524 if (g_lstat(pathbuf, &sb, pglob))
527 if (((pglob->gl_flags & GLOB_MARK) &&
528 pathend[-1] != SEP) && (S_ISDIR(sb.st_mode) ||
529 (S_ISLNK(sb.st_mode) &&
530 (g_stat(pathbuf, &sb, pglob) == 0) &&
531 S_ISDIR(sb.st_mode)))) {
532 if (pathend+1 > pathend_last)
538 return(globextend(pathbuf, pglob, limitp));
541 /* Find end of next segment, copy tentatively to pathend. */
544 while (*p != EOS && *p != SEP) {
547 if (q+1 > pathend_last)
552 if (!anymeta) { /* No expansion, do next segment. */
555 while (*pattern == SEP) {
556 if (pathend+1 > pathend_last)
558 *pathend++ = *pattern++;
561 /* Need expansion, recurse. */
562 return(glob3(pathbuf, pathbuf_last, pathend,
563 pathend_last, pattern, pattern_last,
564 p, pattern_last, pglob, limitp));
570 glob3(Char *pathbuf, Char *pathbuf_last, Char *pathend, Char *pathend_last,
571 Char *pattern, Char *pattern_last, Char *restpattern,
572 Char *restpattern_last, glob_t *pglob, size_t *limitp)
577 char buf[MAXPATHLEN];
580 * The readdirfunc declaration can't be prototyped, because it is
581 * assigned, below, to two functions which are prototyped in glob.h
582 * and dirent.h as taking pointers to differently typed opaque
585 struct dirent *(*readdirfunc)(void *);
587 if (pathend > pathend_last)
592 if ((dirp = g_opendir(pathbuf, pglob)) == NULL) {
593 /* TODO: don't call for ENOENT or ENOTDIR? */
594 if (pglob->gl_errfunc) {
595 if (g_Ctoc(pathbuf, buf, sizeof(buf)))
596 return(GLOB_ABORTED);
597 if (pglob->gl_errfunc(buf, errno) ||
598 pglob->gl_flags & GLOB_ERR)
599 return(GLOB_ABORTED);
606 /* Search directory for matching names. */
607 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
608 readdirfunc = pglob->gl_readdir;
610 readdirfunc = (struct dirent *(*)(void *))readdir;
611 while ((dp = (*readdirfunc)(dirp))) {
615 /* Initial DOT must be matched literally. */
616 if (dp->d_name[0] == DOT && *pattern != DOT)
619 sc = (u_char *) dp->d_name;
620 while (dc < pathend_last && (*dc++ = *sc++) != EOS)
622 if (dc >= pathend_last) {
628 if (!match(pathend, pattern, restpattern)) {
632 err = glob2(pathbuf, pathbuf_last, --dc, pathend_last,
633 restpattern, restpattern_last, pglob, limitp);
638 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
639 (*pglob->gl_closedir)(dirp);
647 * Extend the gl_pathv member of a glob_t structure to accommodate a new item,
648 * add the new item, and update gl_pathc.
650 * This assumes the BSD realloc, which only copies the block when its size
651 * crosses a power-of-two boundary; for v7 realloc, this would cause quadratic
654 * Return 0 if new item added, error code if memory couldn't be allocated.
656 * Invariant of the glob_t structure:
657 * Either gl_pathc is zero and gl_pathv is NULL; or gl_pathc > 0 and
658 * gl_pathv points to (gl_offs + gl_pathc + 1) items.
661 globextend(const Char *path, glob_t *pglob, size_t *limitp)
669 newsize = sizeof(*pathv) * (2 + pglob->gl_pathc + pglob->gl_offs);
670 pathv = pglob->gl_pathv ? realloc((char *)pglob->gl_pathv, newsize) :
673 if (pglob->gl_pathv) {
674 free(pglob->gl_pathv);
675 pglob->gl_pathv = NULL;
677 return(GLOB_NOSPACE);
680 if (pglob->gl_pathv == NULL && pglob->gl_offs > 0) {
681 /* first time around -- clear initial gl_offs items */
682 pathv += pglob->gl_offs;
683 for (i = pglob->gl_offs; --i >= 0; )
686 pglob->gl_pathv = pathv;
688 for (p = path; *p++;)
690 len = (size_t)(p - path);
692 if ((copy = malloc(len)) != NULL) {
693 if (g_Ctoc(path, copy, len)) {
695 return(GLOB_NOSPACE);
697 pathv[pglob->gl_offs + pglob->gl_pathc++] = copy;
699 pathv[pglob->gl_offs + pglob->gl_pathc] = NULL;
701 if ((pglob->gl_flags & GLOB_LIMIT) &&
702 newsize + *limitp >= (u_int) get_arg_max()) {
704 return(GLOB_NOSPACE);
707 return(copy == NULL ? GLOB_NOSPACE : 0);
712 * pattern matching function for filenames. Each occurrence of the *
713 * pattern causes a recursion level.
716 match(Char *name, Char *pat, Char *patend)
718 int ok, negate_range;
721 while (pat < patend) {
723 switch (c & M_MASK) {
728 if (match(name, pat, patend))
730 } while (*name++ != EOS);
738 if ((k = *name++) == EOS)
740 if ((negate_range = ((*pat & M_MASK) == M_NOT)) != EOS)
742 while (((c = *pat++) & M_MASK) != M_END)
743 if ((*pat & M_MASK) == M_RNG) {
744 if (c <= k && k <= pat[1])
749 if (ok == negate_range)
758 return(*name == EOS);
761 /* Free allocated data belonging to a glob_t structure. */
763 globfree(glob_t *pglob)
768 if (pglob->gl_pathv != NULL) {
769 pp = pglob->gl_pathv + pglob->gl_offs;
770 for (i = pglob->gl_pathc; i--; ++pp)
773 free(pglob->gl_pathv);
774 pglob->gl_pathv = NULL;
779 g_opendir(Char *str, glob_t *pglob)
781 char buf[MAXPATHLEN];
784 strlcpy(buf, ".", sizeof buf);
786 if (g_Ctoc(str, buf, sizeof(buf)))
790 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
791 return((*pglob->gl_opendir)(buf));
793 return(opendir(buf));
797 g_lstat(Char *fn, struct stat *sb, glob_t *pglob)
799 char buf[MAXPATHLEN];
801 if (g_Ctoc(fn, buf, sizeof(buf)))
803 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
804 return((*pglob->gl_lstat)(buf, sb));
805 return(lstat(buf, sb));
809 g_stat(Char *fn, struct stat *sb, glob_t *pglob)
811 char buf[MAXPATHLEN];
813 if (g_Ctoc(fn, buf, sizeof(buf)))
815 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
816 return((*pglob->gl_stat)(buf, sb));
817 return(stat(buf, sb));
821 g_strchr(Char *str, int ch)
831 g_Ctoc(const Char *str, char *buf, u_int len)
835 if ((*buf++ = *str++) == EOS)
843 qprintf(const char *str, Char *s)
847 (void)printf("%s:\n", str);
849 (void)printf("%c", CHAR(*p));
852 (void)printf("%c", *p & M_PROTECT ? '"' : ' ');
855 (void)printf("%c", ismeta(*p) ? '_' : ' ');
860 #endif /* !defined(HAVE_GLOB) || !defined(GLOB_HAS_ALTDIRFUNC) ||
861 !defined(GLOB_HAS_GL_MATCHC) */