2 * Copyright (c) 1989 The Regents of the University of California.
5 * This code is derived from software contributed to Berkeley by
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
11 * 1. Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in the
15 * documentation and/or other materials provided with the distribution.
16 * 3. Neither the name of the University nor the names of its contributors
17 * may be used to endorse or promote products derived from this software
18 * without specific prior written permission.
20 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
21 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
22 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
23 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
24 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
25 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
26 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
27 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
28 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
29 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
32 #if defined(LIBC_SCCS) && !defined(lint)
33 static char sccsid[] = "@(#)glob.c 5.12 (Berkeley) 6/24/91";
34 #endif /* LIBC_SCCS and not lint */
36 * Glob: the interface is a superset of the one defined in POSIX 1003.2,
39 * The [!...] convention to negate a range is supported (SysV, Posix, ksh).
41 * Optional extra services, controlled by flags not defined by POSIX:
44 * Escaping convention: \ inhibits any special meaning the following
45 * character might have (except \ at end of string is retained).
47 * Set in gl_flags if pattern contained a globbing character.
49 * Use ^ instead of ! for "not".
51 * Number of matches in the current invocation of glob.
55 #include <sys/types.h>
56 #include <sys/param.h>
63 #pragma warning(disable:4244)
64 #endif /* WINNT_NATIVE */
79 #define S_ISDIR(a) (((a) & S_IFMT) == S_IFDIR)
82 #if !defined(S_ISLNK) && defined(S_IFLNK)
83 #define S_ISLNK(a) (((a) & S_IFMT) == S_IFLNK)
86 #if !defined(S_ISLNK) && !defined(lstat)
90 typedef unsigned short Char;
92 static int glob1 __P((Char *, glob_t *, int));
93 static int glob2 __P((Char *, Char *, Char *, glob_t *, int));
94 static int glob3 __P((Char *, Char *, Char *, Char *,
96 static int globextend __P((Char *, glob_t *));
97 static int match __P((Char *, Char *, Char *, int));
99 static int compare __P((const ptr_t, const ptr_t));
101 static DIR *Opendir __P((Char *));
103 static int Lstat __P((Char *, struct stat *));
105 static int Stat __P((Char *, struct stat *sb));
106 static Char *Strchr __P((Char *, int));
108 static void qprintf __P((Char *));
124 #define UNDERSCORE '_'
126 #define M_META 0x8000
127 #define M_PROTECT 0x4000
128 #define M_MASK 0xffff
129 #define M_ASCII 0x00ff
131 #define CHAR(c) ((c)&M_ASCII)
132 #define META(c) ((c)|M_META)
133 #define M_ALL META('*')
134 #define M_END META(']')
135 #define M_NOT META('!')
136 #define M_ALTNOT META('^')
137 #define M_ONE META('?')
138 #define M_RNG META('-')
139 #define M_SET META('[')
140 #define ismeta(c) (((c)&M_META) != 0)
143 #define GLOBBUFLEN MAXPATHLEN
145 #define GLOBBUFLEN BUFSIZE
152 #if defined(NLS) && defined(LC_COLLATE) && !defined(NOSTRCOLL)
158 * From kevin lyda <kevin@suberic.net>:
159 * strcoll does not guarantee case sorting, so we pre-process now:
161 if (islower(c1) && isupper(c2))
165 s1[1] = s2[1] = '\0';
166 return strcoll(s1, s2);
173 * Need to dodge two kernel bugs:
174 * opendir("") != opendir(".")
175 * NAMEI_BUG: on plain files trailing slashes are ignored in some kernels.
176 * POSIX specifies that they should be ignored in directories.
183 char buf[GLOBBUFLEN];
184 register char *dc = buf;
185 #if defined(hpux) || defined(__hpux)
190 return (opendir("."));
191 while ((*dc++ = *str++) != '\0')
193 #if defined(hpux) || defined(__hpux)
195 * Opendir on some device files hangs, so avoid it
197 if (stat(buf, &st) == -1 || !S_ISDIR(st.st_mode))
200 return (opendir(buf));
209 char buf[GLOBBUFLEN];
210 register char *dc = buf;
212 while ((*dc++ = *fn++) != '\0')
221 return (*--dc == '/' && !S_ISDIR(sb->st_mode) ? -1 : st);
224 return (lstat(buf, sb));
225 # endif /* NAMEI_BUG */
236 char buf[GLOBBUFLEN];
237 register char *dc = buf;
239 while ((*dc++ = *fn++) != '\0')
248 return (*--dc == '/' && !S_ISDIR(sb->st_mode) ? -1 : st);
251 return (stat(buf, sb));
252 #endif /* NAMEI_BUG */
275 printf("%c", *p & 0xff);
278 printf("%c", *p & M_PROTECT ? '"' : ' ');
281 printf("%c", *p & M_META ? '_' : ' ');
290 #if defined(NLS) && !defined(NOSTRCOLL)
291 errno = 0; /* strcoll sets errno, another brain-damage */
293 return (strcoll(*(char **) p, *(char **) q));
295 return (strcmp(*(char **) p, *(char **) q));
296 #endif /* NLS && !NOSTRCOLL */
300 * The main glob() routine: compiles the pattern (optionally processing
301 * quotes), calls glob1() to do the real pattern matching, and finally
302 * sorts the list (unless unsorted operation is requested). Returns 0
303 * if things went well, nonzero if errors occurred. It is not an error
304 * to find no matches.
307 glob(pattern, flags, errfunc, pglob)
310 int (*errfunc) __P((const char *, int));
314 Char *bufnext, *bufend, *compilebuf, m_not;
315 const unsigned char *compilepat, *patnext;
317 Char patbuf[GLOBBUFLEN + 1], *qpatnext;
320 patnext = (unsigned char *) pattern;
321 if (!(flags & GLOB_APPEND)) {
323 pglob->gl_pathv = NULL;
324 if (!(flags & GLOB_DOOFFS))
327 pglob->gl_flags = flags & ~GLOB_MAGCHAR;
328 pglob->gl_errfunc = errfunc;
329 oldpathc = pglob->gl_pathc;
330 pglob->gl_matchc = 0;
332 if (pglob->gl_flags & GLOB_ALTNOT) {
342 bufend = bufnext + GLOBBUFLEN;
343 compilebuf = bufnext;
344 compilepat = patnext;
346 no_match = *patnext == not;
350 if (flags & GLOB_QUOTE) {
351 /* Protect the quoted characters */
352 while (bufnext < bufend && (c = *patnext++) != EOS)
354 if (Ismbyte1(c) && *patnext != EOS)
356 *bufnext++ = (Char) c;
357 *bufnext++ = (Char) *patnext++;
360 #endif /* DSPMBYTE */
362 if ((c = *patnext++) == EOS) {
366 *bufnext++ = (Char) (c | M_PROTECT);
369 *bufnext++ = (Char) c;
372 while (bufnext < bufend && (c = *patnext++) != EOS)
373 *bufnext++ = (Char) c;
378 /* we don't need to check for buffer overflow any more */
379 while ((c = *qpatnext++) != EOS) {
381 if (Ismbyte1(c) && *qpatnext != EOS)
383 *bufnext++ = CHAR(c);
384 *bufnext++ = CHAR(*qpatnext++);
387 #endif /* DSPMBYTE */
393 if (*qpatnext == EOS ||
394 Strchr(qpatnext + 1, RBRACKET) == NULL) {
395 *bufnext++ = LBRACKET;
400 pglob->gl_flags |= GLOB_MAGCHAR;
406 *bufnext++ = CHAR(c);
407 if (*qpatnext == RANGE &&
408 (c = qpatnext[1]) != RBRACKET) {
410 *bufnext++ = CHAR(c);
413 } while ((c = *qpatnext++) != RBRACKET);
417 pglob->gl_flags |= GLOB_MAGCHAR;
421 pglob->gl_flags |= GLOB_MAGCHAR;
422 /* collapse adjacent stars to one, to avoid
423 * exponential behavior
425 if (bufnext == patbuf || bufnext[-1] != M_ALL)
429 *bufnext++ = CHAR(c);
438 if ((err = glob1(patbuf, pglob, no_match)) != 0)
442 * If there was no match we are going to append the pattern
443 * if GLOB_NOCHECK was specified or if GLOB_NOMAGIC was specified
444 * and the pattern did not contain any magic characters
445 * GLOB_NOMAGIC is there just for compatibility with csh.
447 if (pglob->gl_pathc == oldpathc &&
448 ((flags & GLOB_NOCHECK) ||
449 ((flags & GLOB_NOMAGIC) && !(pglob->gl_flags & GLOB_MAGCHAR)))) {
450 if (!(flags & GLOB_QUOTE)) {
451 Char *dp = compilebuf;
452 const unsigned char *sp = compilepat;
454 while ((*dp++ = *sp++) != '\0')
459 * copy pattern, interpreting quotes; this is slightly different
460 * than the interpretation of quotes above -- which should prevail?
462 while (*compilepat != EOS) {
463 if (*compilepat == QUOTE) {
464 if (*++compilepat == EOS)
467 *compilebuf++ = (unsigned char) *compilepat++;
471 return (globextend(patbuf, pglob));
473 else if (!(flags & GLOB_NOSORT) && (pglob->gl_pathc != oldpathc))
474 qsort((char *) (pglob->gl_pathv + pglob->gl_offs + oldpathc),
475 pglob->gl_pathc - oldpathc, sizeof(char *),
476 (int (*) __P((const void *, const void *))) compare);
481 glob1(pattern, pglob, no_match)
486 Char pathbuf[GLOBBUFLEN + 1];
489 * a null pathname is invalid -- POSIX 1003.1 sect. 2.4.
493 return (glob2(pathbuf, pathbuf, pattern, pglob, no_match));
497 * functions glob2 and glob3 are mutually recursive; there is one level
498 * of recursion for each segment in the pattern that contains one or
499 * more meta characters.
502 glob2(pathbuf, pathend, pattern, pglob, no_match)
503 Char *pathbuf, *pathend, *pattern;
512 * loop over pattern segments until end of pattern or until segment with
513 * meta character found.
517 if (*pattern == EOS) { /* end of pattern? */
520 if (Lstat(pathbuf, &sbuf))
523 if (((pglob->gl_flags & GLOB_MARK) &&
524 pathend[-1] != SEP) &&
525 (S_ISDIR(sbuf.st_mode)
527 || (S_ISLNK(sbuf.st_mode) &&
528 (Stat(pathbuf, &sbuf) == 0) &&
529 S_ISDIR(sbuf.st_mode))
536 return (globextend(pathbuf, pglob));
539 /* find end of next segment, copy tentatively to pathend */
542 while (*p != EOS && *p != SEP) {
548 if (!anymeta) { /* no expansion, do next segment */
551 while (*pattern == SEP)
552 *pathend++ = *pattern++;
554 else /* need expansion, recurse */
555 return (glob3(pathbuf, pathend, pattern, p, pglob, no_match));
562 glob3(pathbuf, pathend, pattern, restpattern, pglob, no_match)
563 Char *pathbuf, *pathend, *pattern, *restpattern;
570 Char m_not = (pglob->gl_flags & GLOB_ALTNOT) ? M_ALTNOT : M_NOT;
571 char cpathbuf[GLOBBUFLEN], *ptr;;
576 if (!(dirp = Opendir(pathbuf))) {
577 /* todo: don't call for ENOENT or ENOTDIR? */
578 for (ptr = cpathbuf; (*ptr++ = (char) *pathbuf++) != EOS;)
580 if ((pglob->gl_errfunc && (*pglob->gl_errfunc) (cpathbuf, errno)) ||
581 (pglob->gl_flags & GLOB_ERR))
589 /* search directory for matching names */
590 while ((dp = readdir(dirp)) != NULL) {
591 register unsigned char *sc;
594 /* initial DOT must be matched literally */
595 if (dp->d_name[0] == DOT && *pattern != DOT)
597 for (sc = (unsigned char *) dp->d_name, dc = pathend;
598 (*dc++ = *sc++) != '\0';)
600 if (match(pathend, pattern, restpattern, (int) m_not) == no_match) {
604 err = glob2(pathbuf, --dc, restpattern, pglob, no_match);
608 /* todo: check error from readdir? */
609 (void) closedir(dirp);
615 * Extend the gl_pathv member of a glob_t structure to accomodate a new item,
616 * add the new item, and update gl_pathc.
618 * This assumes the BSD realloc, which only copies the block when its size
619 * crosses a power-of-two boundary; for v7 realloc, this would cause quadratic
622 * Return 0 if new item added, error code if memory couldn't be allocated.
624 * Invariant of the glob_t structure:
625 * Either gl_pathc is zero and gl_pathv is NULL; or gl_pathc > 0 and
626 * gl_pathv points to (gl_offs + gl_pathc + 1) items.
629 globextend(path, pglob)
633 register char **pathv;
635 unsigned int newsize;
639 newsize = sizeof(*pathv) * (2 + pglob->gl_pathc + pglob->gl_offs);
640 pathv = (char **) (pglob->gl_pathv ?
641 xrealloc((ptr_t) pglob->gl_pathv, (size_t) newsize) :
642 xmalloc((size_t) newsize));
644 return (GLOB_NOSPACE);
646 if (pglob->gl_pathv == NULL && pglob->gl_offs > 0) {
647 /* first time around -- clear initial gl_offs items */
648 pathv += pglob->gl_offs;
649 for (i = pglob->gl_offs; --i >= 0;)
652 pglob->gl_pathv = pathv;
654 for (p = path; *p++;)
656 if ((copy = (char *) xmalloc((size_t) (p - path))) != NULL) {
657 register char *dc = copy;
658 register Char *sc = path;
660 while ((*dc++ = *sc++) != '\0')
662 pathv[pglob->gl_offs + pglob->gl_pathc++] = copy;
664 pathv[pglob->gl_offs + pglob->gl_pathc] = NULL;
665 return ((copy == NULL) ? GLOB_NOSPACE : 0);
670 * pattern matching function for filenames. Each occurrence of the *
671 * pattern causes a recursion level.
674 match(name, pat, patend, m_not)
675 register Char *name, *pat, *patend;
678 int ok, negate_range;
681 while (pat < patend) {
683 switch (c & M_MASK) {
688 if (match(name, pat, patend, m_not))
690 while (*name++ != EOS);
698 if ((k = *name++) == EOS)
700 if ((negate_range = ((*pat & M_MASK) == m_not)) != 0)
702 while (((c = *pat++) & M_MASK) != M_END) {
703 if ((*pat & M_MASK) == M_RNG) {
704 if (globcharcoll(CHAR(c), CHAR(k)) <= 0 &&
705 globcharcoll(CHAR(k), CHAR(pat[1])) <= 0)
712 if (ok == negate_range)
717 if (samecase(k) != samecase(c))
722 return (*name == EOS);
725 /* free allocated data belonging to a glob_t structure */
733 if (pglob->gl_pathv != NULL) {
734 pp = pglob->gl_pathv + pglob->gl_offs;
735 for (i = pglob->gl_pathc; i--; ++pp)
737 xfree((ptr_t) *pp), *pp = NULL;
738 xfree((ptr_t) pglob->gl_pathv), pglob->gl_pathv = NULL;