1 /* OPENBSD ORIGINAL: lib/libc/gen/glob.c */
4 * Copyright (c) 1989, 1993
5 * The Regents of the University of California. All rights reserved.
7 * This code is derived from software contributed to Berkeley by
10 * Redistribution and use in source and binary forms, with or without
11 * modification, are permitted provided that the following conditions
13 * 1. Redistributions of source code must retain the above copyright
14 * notice, this list of conditions and the following disclaimer.
15 * 2. Redistributions in binary form must reproduce the above copyright
16 * notice, this list of conditions and the following disclaimer in the
17 * documentation and/or other materials provided with the distribution.
18 * 3. Neither the name of the University nor the names of its contributors
19 * may be used to endorse or promote products derived from this software
20 * without specific prior written permission.
22 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
23 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
24 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
25 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
26 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
27 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
28 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
29 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
30 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
31 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
43 #elif defined(HAVE_SYSCONF) && defined(_SC_ARG_MAX)
44 return(sysconf(_SC_ARG_MAX));
46 return(256); /* XXX: arbitrary */
50 #if !defined(HAVE_GLOB) || !defined(GLOB_HAS_ALTDIRFUNC) || \
51 !defined(GLOB_HAS_GL_MATCHC)
53 #if defined(LIBC_SCCS) && !defined(lint)
55 static char sccsid[] = "@(#)glob.c 8.3 (Berkeley) 10/13/93";
57 static char rcsid[] = "$OpenBSD: glob.c,v 1.22 2003/06/25 21:16:47 deraadt Exp $";
59 #endif /* LIBC_SCCS and not lint */
62 * glob(3) -- a superset of the one defined in POSIX 1003.2.
64 * The [!...] convention to negate a range is supported (SysV, Posix, ksh).
66 * Optional extra services, controlled by flags not defined by POSIX:
69 * Escaping convention: \ inhibits any special meaning the following
70 * character might have (except \ at end of string is retained).
72 * Set in gl_flags if pattern contained a globbing character.
74 * Same as GLOB_NOCHECK, but it will only append pattern if it did
75 * not contain any magic characters. [Used in csh style globbing]
77 * Use alternately specified directory access functions.
79 * expand ~user/foo to the /home/dir/of/user/foo
81 * expand {1,2}{a,b} to 1a 1b 2a 2b
83 * Number of matches in the current invocation of glob.
98 #undef TILDE /* Some platforms may already define it */
100 #define UNDERSCORE '_'
108 #define M_QUOTE 0x8000
109 #define M_PROTECT 0x4000
110 #define M_MASK 0xffff
111 #define M_ASCII 0x00ff
113 typedef u_short Char;
118 #define M_PROTECT 0x40
127 #define CHAR(c) ((Char)((c)&M_ASCII))
128 #define META(c) ((Char)((c)|M_QUOTE))
129 #define M_ALL META('*')
130 #define M_END META(']')
131 #define M_NOT META('!')
132 #define M_ONE META('?')
133 #define M_RNG META('-')
134 #define M_SET META('[')
135 #define ismeta(c) (((c)&M_QUOTE) != 0)
138 static int compare(const void *, const void *);
139 static int g_Ctoc(const Char *, char *, u_int);
140 static int g_lstat(Char *, struct stat *, glob_t *);
141 static DIR *g_opendir(Char *, glob_t *);
142 static Char *g_strchr(Char *, int);
143 static int g_stat(Char *, struct stat *, glob_t *);
144 static int glob0(const Char *, glob_t *);
145 static int glob1(Char *, Char *, glob_t *, size_t *);
146 static int glob2(Char *, Char *, Char *, Char *, Char *, Char *,
148 static int glob3(Char *, Char *, Char *, Char *, Char *, Char *,
149 Char *, Char *, glob_t *, size_t *);
150 static int globextend(const Char *, glob_t *, size_t *);
152 globtilde(const Char *, Char *, size_t, glob_t *);
153 static int globexp1(const Char *, glob_t *);
154 static int globexp2(const Char *, const Char *, glob_t *, int *);
155 static int match(Char *, Char *, Char *);
157 static void qprintf(const char *, Char *);
161 glob(pattern, flags, errfunc, pglob)
163 int flags, (*errfunc)(const char *, int);
166 const u_char *patnext;
168 Char *bufnext, *bufend, patbuf[MAXPATHLEN];
170 patnext = (u_char *) pattern;
171 if (!(flags & GLOB_APPEND)) {
173 pglob->gl_pathv = NULL;
174 if (!(flags & GLOB_DOOFFS))
177 pglob->gl_flags = flags & ~GLOB_MAGCHAR;
178 pglob->gl_errfunc = errfunc;
179 pglob->gl_matchc = 0;
182 bufend = bufnext + MAXPATHLEN - 1;
183 if (flags & GLOB_NOESCAPE)
184 while (bufnext < bufend && (c = *patnext++) != EOS)
187 /* Protect the quoted characters. */
188 while (bufnext < bufend && (c = *patnext++) != EOS)
190 if ((c = *patnext++) == EOS) {
194 *bufnext++ = c | M_PROTECT;
200 if (flags & GLOB_BRACE)
201 return globexp1(patbuf, pglob);
203 return glob0(patbuf, pglob);
207 * Expand recursively a glob {} pattern. When there is no more expansion
208 * invoke the standard globbing routine to glob the rest of the magic
212 globexp1(pattern, 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(ptr, pattern, pglob, rv)
238 const Char *ptr, *pattern;
244 const Char *pe, *pm, *pl;
245 Char patbuf[MAXPATHLEN];
247 /* copy part up to the brace */
248 for (lm = patbuf, pm = pattern; pm != ptr; *lm++ = *pm++)
253 /* Find the balanced brace */
254 for (i = 0, pe = ++ptr; *pe; pe++)
255 if (*pe == LBRACKET) {
256 /* Ignore everything between [] */
257 for (pm = pe++; *pe != RBRACKET && *pe != EOS; pe++)
261 * We could not find a matching RBRACKET.
262 * Ignore and just look for RBRACE
266 } else if (*pe == LBRACE)
268 else if (*pe == RBRACE) {
274 /* Non matching braces; just glob the pattern */
275 if (i != 0 || *pe == EOS) {
276 *rv = glob0(patbuf, pglob);
280 for (i = 0, pl = pm = ptr; pm <= pe; pm++) {
283 /* Ignore everything between [] */
284 for (pl = pm++; *pm != RBRACKET && *pm != EOS; pm++)
288 * We could not find a matching RBRACKET.
289 * Ignore and just look for RBRACE
306 if (i && *pm == COMMA)
309 /* Append the current string */
310 for (lm = ls; (pl < pm); *lm++ = *pl++)
314 * Append the rest of the pattern after the
317 for (pl = pe + 1; (*lm++ = *pl++) != EOS; )
320 /* Expand the current pattern */
322 qprintf("globexp2:", patbuf);
324 *rv = globexp1(patbuf, pglob);
326 /* move after the comma, to the next string */
342 * expand tilde from the passwd file.
345 globtilde(pattern, patbuf, patbuf_len, pglob)
356 if (*pattern != TILDE || !(pglob->gl_flags & GLOB_TILDE))
359 /* Copy up to the end of the string or / */
360 eb = &patbuf[patbuf_len - 1];
361 for (p = pattern + 1, h = (char *) patbuf;
362 h < (char *)eb && *p && *p != SLASH; *h++ = *p++)
372 if (((char *) patbuf)[0] == EOS) {
374 * handle a plain ~ or ~/ by expanding $HOME
375 * first and then trying the password file
378 if (issetugid() != 0 || (h = getenv("HOME")) == NULL) {
380 if ((getuid() != geteuid()) || (h = getenv("HOME")) == NULL) {
381 if ((pwd = getpwuid(getuid())) == NULL)
390 if ((pwd = getpwnam((char*) patbuf)) == NULL)
396 /* Copy the home directory */
397 for (b = patbuf; b < eb && *h; *b++ = *h++)
400 /* Append the rest of the pattern */
401 while (b < eb && (*b++ = *p++) != EOS)
410 * The main glob() routine: compiles the pattern (optionally processing
411 * quotes), calls glob1() to do the real pattern matching, and finally
412 * sorts the list (unless unsorted operation is requested). Returns 0
413 * if things went well, nonzero if errors occurred. It is not an error
414 * to find no matches.
417 glob0(pattern, pglob)
421 const Char *qpatnext;
422 int c, err, oldpathc;
423 Char *bufnext, patbuf[MAXPATHLEN];
426 qpatnext = globtilde(pattern, patbuf, MAXPATHLEN, pglob);
427 oldpathc = pglob->gl_pathc;
430 /* We don't need to check for buffer overflow any more. */
431 while ((c = *qpatnext++) != EOS) {
437 if (*qpatnext == EOS ||
438 g_strchr((Char *) qpatnext+1, RBRACKET) == NULL) {
439 *bufnext++ = LBRACKET;
449 *bufnext++ = CHAR(c);
450 if (*qpatnext == RANGE &&
451 (c = qpatnext[1]) != RBRACKET) {
453 *bufnext++ = CHAR(c);
456 } while ((c = *qpatnext++) != RBRACKET);
457 pglob->gl_flags |= GLOB_MAGCHAR;
461 pglob->gl_flags |= GLOB_MAGCHAR;
465 pglob->gl_flags |= GLOB_MAGCHAR;
466 /* collapse adjacent stars to one,
467 * to avoid exponential behavior
469 if (bufnext == patbuf || bufnext[-1] != M_ALL)
473 *bufnext++ = CHAR(c);
479 qprintf("glob0:", patbuf);
482 if ((err = glob1(patbuf, patbuf+MAXPATHLEN-1, pglob, &limit)) != 0)
486 * If there was no match we are going to append the pattern
487 * if GLOB_NOCHECK was specified or if GLOB_NOMAGIC was specified
488 * and the pattern did not contain any magic characters
489 * GLOB_NOMAGIC is there just for compatibility with csh.
491 if (pglob->gl_pathc == oldpathc) {
492 if ((pglob->gl_flags & GLOB_NOCHECK) ||
493 ((pglob->gl_flags & GLOB_NOMAGIC) &&
494 !(pglob->gl_flags & GLOB_MAGCHAR)))
495 return(globextend(pattern, pglob, &limit));
497 return(GLOB_NOMATCH);
499 if (!(pglob->gl_flags & GLOB_NOSORT))
500 qsort(pglob->gl_pathv + pglob->gl_offs + oldpathc,
501 pglob->gl_pathc - oldpathc, sizeof(char *), compare);
509 return(strcmp(*(char **)p, *(char **)q));
513 glob1(pattern, pattern_last, pglob, limitp)
514 Char *pattern, *pattern_last;
518 Char pathbuf[MAXPATHLEN];
520 /* A null pathname is invalid -- POSIX 1003.1 sect. 2.4. */
523 return(glob2(pathbuf, pathbuf+MAXPATHLEN-1,
524 pathbuf, pathbuf+MAXPATHLEN-1,
525 pattern, pattern_last, pglob, limitp));
529 * The functions glob2 and glob3 are mutually recursive; there is one level
530 * of recursion for each segment in the pattern that contains one or more
534 glob2(pathbuf, pathbuf_last, pathend, pathend_last, pattern,
535 pattern_last, pglob, limitp)
536 Char *pathbuf, *pathbuf_last, *pathend, *pathend_last;
537 Char *pattern, *pattern_last;
546 * Loop over pattern segments until end of pattern or until
547 * segment with meta character found.
549 for (anymeta = 0;;) {
550 if (*pattern == EOS) { /* End of pattern? */
552 if (g_lstat(pathbuf, &sb, pglob))
555 if (((pglob->gl_flags & GLOB_MARK) &&
556 pathend[-1] != SEP) && (S_ISDIR(sb.st_mode) ||
557 (S_ISLNK(sb.st_mode) &&
558 (g_stat(pathbuf, &sb, pglob) == 0) &&
559 S_ISDIR(sb.st_mode)))) {
560 if (pathend+1 > pathend_last)
566 return(globextend(pathbuf, pglob, limitp));
569 /* Find end of next segment, copy tentatively to pathend. */
572 while (*p != EOS && *p != SEP) {
575 if (q+1 > pathend_last)
580 if (!anymeta) { /* No expansion, do next segment. */
583 while (*pattern == SEP) {
584 if (pathend+1 > pathend_last)
586 *pathend++ = *pattern++;
589 /* Need expansion, recurse. */
590 return(glob3(pathbuf, pathbuf_last, pathend,
591 pathend_last, pattern, pattern_last,
592 p, pattern_last, pglob, limitp));
598 glob3(pathbuf, pathbuf_last, pathend, pathend_last, pattern, pattern_last,
599 restpattern, restpattern_last, pglob, limitp)
600 Char *pathbuf, *pathbuf_last, *pathend, *pathend_last;
601 Char *pattern, *pattern_last, *restpattern, *restpattern_last;
605 register struct dirent *dp;
608 char buf[MAXPATHLEN];
611 * The readdirfunc declaration can't be prototyped, because it is
612 * assigned, below, to two functions which are prototyped in glob.h
613 * and dirent.h as taking pointers to differently typed opaque
616 struct dirent *(*readdirfunc)(void *);
618 if (pathend > pathend_last)
623 if ((dirp = g_opendir(pathbuf, pglob)) == NULL) {
624 /* TODO: don't call for ENOENT or ENOTDIR? */
625 if (pglob->gl_errfunc) {
626 if (g_Ctoc(pathbuf, buf, sizeof(buf)))
627 return(GLOB_ABORTED);
628 if (pglob->gl_errfunc(buf, errno) ||
629 pglob->gl_flags & GLOB_ERR)
630 return(GLOB_ABORTED);
637 /* Search directory for matching names. */
638 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
639 readdirfunc = pglob->gl_readdir;
641 readdirfunc = (struct dirent *(*)(void *))readdir;
642 while ((dp = (*readdirfunc)(dirp))) {
646 /* Initial DOT must be matched literally. */
647 if (dp->d_name[0] == DOT && *pattern != DOT)
650 sc = (u_char *) dp->d_name;
651 while (dc < pathend_last && (*dc++ = *sc++) != EOS)
653 if (dc >= pathend_last) {
659 if (!match(pathend, pattern, restpattern)) {
663 err = glob2(pathbuf, pathbuf_last, --dc, pathend_last,
664 restpattern, restpattern_last, pglob, limitp);
669 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
670 (*pglob->gl_closedir)(dirp);
678 * Extend the gl_pathv member of a glob_t structure to accommodate a new item,
679 * add the new item, and update gl_pathc.
681 * This assumes the BSD realloc, which only copies the block when its size
682 * crosses a power-of-two boundary; for v7 realloc, this would cause quadratic
685 * Return 0 if new item added, error code if memory couldn't be allocated.
687 * Invariant of the glob_t structure:
688 * Either gl_pathc is zero and gl_pathv is NULL; or gl_pathc > 0 and
689 * gl_pathv points to (gl_offs + gl_pathc + 1) items.
692 globextend(path, pglob, limitp)
697 register char **pathv;
703 newsize = sizeof(*pathv) * (2 + pglob->gl_pathc + pglob->gl_offs);
704 pathv = pglob->gl_pathv ? realloc((char *)pglob->gl_pathv, newsize) :
707 if (pglob->gl_pathv) {
708 free(pglob->gl_pathv);
709 pglob->gl_pathv = NULL;
711 return(GLOB_NOSPACE);
714 if (pglob->gl_pathv == NULL && pglob->gl_offs > 0) {
715 /* first time around -- clear initial gl_offs items */
716 pathv += pglob->gl_offs;
717 for (i = pglob->gl_offs; --i >= 0; )
720 pglob->gl_pathv = pathv;
722 for (p = path; *p++;)
724 len = (size_t)(p - path);
726 if ((copy = malloc(len)) != NULL) {
727 if (g_Ctoc(path, copy, len)) {
729 return(GLOB_NOSPACE);
731 pathv[pglob->gl_offs + pglob->gl_pathc++] = copy;
733 pathv[pglob->gl_offs + pglob->gl_pathc] = NULL;
735 if ((pglob->gl_flags & GLOB_LIMIT) &&
736 newsize + *limitp >= (u_int) get_arg_max()) {
738 return(GLOB_NOSPACE);
741 return(copy == NULL ? GLOB_NOSPACE : 0);
746 * pattern matching function for filenames. Each occurrence of the *
747 * pattern causes a recursion level.
750 match(name, pat, patend)
751 register Char *name, *pat, *patend;
753 int ok, negate_range;
756 while (pat < patend) {
758 switch (c & M_MASK) {
763 if (match(name, pat, patend))
765 while (*name++ != EOS)
774 if ((k = *name++) == EOS)
776 if ((negate_range = ((*pat & M_MASK) == M_NOT)) != EOS)
778 while (((c = *pat++) & M_MASK) != M_END)
779 if ((*pat & M_MASK) == M_RNG) {
780 if (c <= k && k <= pat[1])
785 if (ok == negate_range)
794 return(*name == EOS);
797 /* Free allocated data belonging to a glob_t structure. */
805 if (pglob->gl_pathv != NULL) {
806 pp = pglob->gl_pathv + pglob->gl_offs;
807 for (i = pglob->gl_pathc; i--; ++pp)
810 free(pglob->gl_pathv);
811 pglob->gl_pathv = NULL;
816 g_opendir(str, pglob)
820 char buf[MAXPATHLEN];
823 strlcpy(buf, ".", sizeof buf);
825 if (g_Ctoc(str, buf, sizeof(buf)))
829 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
830 return((*pglob->gl_opendir)(buf));
832 return(opendir(buf));
836 g_lstat(fn, sb, pglob)
841 char buf[MAXPATHLEN];
843 if (g_Ctoc(fn, buf, sizeof(buf)))
845 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
846 return((*pglob->gl_lstat)(buf, sb));
847 return(lstat(buf, sb));
851 g_stat(fn, sb, pglob)
856 char buf[MAXPATHLEN];
858 if (g_Ctoc(fn, buf, sizeof(buf)))
860 if (pglob->gl_flags & GLOB_ALTDIRFUNC)
861 return((*pglob->gl_stat)(buf, sb));
862 return(stat(buf, sb));
878 g_Ctoc(str, buf, len)
879 register const Char *str;
885 if ((*buf++ = *str++) == EOS)
899 (void)printf("%s:\n", str);
901 (void)printf("%c", CHAR(*p));
904 (void)printf("%c", *p & M_PROTECT ? '"' : ' ');
907 (void)printf("%c", ismeta(*p) ? '_' : ' ');
912 #endif /* !defined(HAVE_GLOB) || !defined(GLOB_HAS_ALTDIRFUNC) ||
913 !defined(GLOB_HAS_GL_MATCHC) */