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 #pragma warning(disable:4244)
56 #endif /* WINNT_NATIVE */
70 #define S_ISDIR(a) (((a) & S_IFMT) == S_IFDIR)
73 #if !defined(S_ISLNK) && defined(S_IFLNK)
74 #define S_ISLNK(a) (((a) & S_IFMT) == S_IFLNK)
77 #if !defined(S_ISLNK) && !defined(lstat)
81 typedef unsigned short Char
;
83 static int glob1 (Char
*, glob_t
*, int);
84 static int glob2 (struct strbuf
*, const Char
*, glob_t
*, int);
85 static int glob3 (struct strbuf
*, const Char
*, const Char
*,
87 static void globextend (const char *, glob_t
*);
88 static int match (const char *, const Char
*, const Char
*,
90 static int compare (const void *, const void *);
91 static DIR *Opendir (const char *);
93 static int Lstat (const char *, struct stat
*);
95 static int Stat (const char *, struct stat
*sb
);
96 static Char
*Strchr (Char
*, int);
98 static void qprintf (const Char
*);
114 #define UNDERSCORE '_'
116 #define M_META 0x8000
117 #define M_PROTECT 0x4000
118 #define M_MASK 0xffff
119 #define M_ASCII 0x00ff
121 #define LCHAR(c) ((c)&M_ASCII)
122 #define META(c) ((c)|M_META)
123 #define M_ALL META('*')
124 #define M_END META(']')
125 #define M_NOT META('!')
126 #define M_ALTNOT META('^')
127 #define M_ONE META('?')
128 #define M_RNG META('-')
129 #define M_SET META('[')
130 #define ismeta(c) (((c)&M_META) != 0)
133 globcharcoll(__Char c1
, __Char c2
, int cs
)
135 #if defined(NLS) && defined(LC_COLLATE) && defined(HAVE_STRCOLL)
136 # if defined(WIDE_STRINGS)
137 wchar_t s1
[2], s2
[2];
145 /* This should not be here, but I'll rather leave it in than engage in
146 a LC_COLLATE flamewar about a shell I don't use... */
147 if (iswlower(c1
) && iswupper(c2
))
149 if (iswupper(c1
) && iswlower(c2
))
154 s1
[1] = s2
[1] = '\0';
155 return wcscoll(s1
, s2
);
156 # else /* not WIDE_STRINGS */
162 * From kevin lyda <kevin@suberic.net>:
163 * strcoll does not guarantee case sorting, so we pre-process now:
166 c1
= islower(c1
) ? c1
: tolower(c1
);
167 c2
= islower(c2
) ? c2
: tolower(c2
);
169 if (islower(c1
) && isupper(c2
))
171 if (isupper(c1
) && islower(c2
))
176 s1
[1] = s2
[1] = '\0';
177 return strcoll(s1
, s2
);
185 * Need to dodge two kernel bugs:
186 * opendir("") != opendir(".")
187 * NAMEI_BUG: on plain files trailing slashes are ignored in some kernels.
188 * POSIX specifies that they should be ignored in directories.
192 Opendir(const char *str
)
194 #if defined(hpux) || defined(__hpux)
199 return (opendir("."));
200 #if defined(hpux) || defined(__hpux)
202 * Opendir on some device files hangs, so avoid it
204 if (stat(str
, &st
) == -1 || !S_ISDIR(st
.st_mode
))
212 Lstat(const char *fn
, struct stat
*sb
)
218 if (*fn
!= 0 && strend(fn
)[-1] == '/' && !S_ISDIR(sb
->st_mode
))
220 # endif /* NAMEI_BUG */
228 Stat(const char *fn
, struct stat
*sb
)
234 if (*fn
!= 0 && strend(fn
)[-1] == '/' && !S_ISDIR(sb
->st_mode
))
236 #endif /* NAMEI_BUG */
241 Strchr(Char
*str
, int ch
)
252 qprintf(const Char
*s
)
257 printf("%c", *p
& 0xff);
260 printf("%c", *p
& M_PROTECT
? '"' : ' ');
263 printf("%c", *p
& M_META
? '_' : ' ');
269 compare(const void *p
, const void *q
)
271 #if defined(NLS) && defined(HAVE_STRCOLL)
272 return (strcoll(*(char *const *) p
, *(char *const *) q
));
274 return (strcmp(*(char *const *) p
, *(char *const *) q
));
275 #endif /* NLS && HAVE_STRCOLL */
279 * The main glob() routine: compiles the pattern (optionally processing
280 * quotes), calls glob1() to do the real pattern matching, and finally
281 * sorts the list (unless unsorted operation is requested). Returns 0
282 * if things went well, nonzero if errors occurred. It is not an error
283 * to find no matches.
286 glob(const char *pattern
, int flags
, int (*errfunc
) (const char *, int),
290 Char
*bufnext
, m_not
;
291 const unsigned char *patnext
;
293 Char
*qpatnext
, *patbuf
;
296 patnext
= (const unsigned char *) pattern
;
297 if (!(flags
& GLOB_APPEND
)) {
299 pglob
->gl_pathv
= NULL
;
300 if (!(flags
& GLOB_DOOFFS
))
303 pglob
->gl_flags
= flags
& ~GLOB_MAGCHAR
;
304 pglob
->gl_errfunc
= errfunc
;
305 oldpathc
= pglob
->gl_pathc
;
306 pglob
->gl_matchc
= 0;
308 if (pglob
->gl_flags
& GLOB_ALTNOT
) {
317 patbuf
= xmalloc((strlen(pattern
) + 1) * sizeof(*patbuf
));
320 no_match
= *patnext
== not;
324 if (flags
& GLOB_QUOTE
) {
325 /* Protect the quoted characters */
326 while ((c
= *patnext
++) != EOS
) {
330 len
= mblen((const char *)(patnext
- 1), MB_LEN_MAX
);
334 *bufnext
++ = (Char
) c
;
336 *bufnext
++ = (Char
) (*patnext
++ | M_PROTECT
);
338 #endif /* WIDE_STRINGS */
340 if ((c
= *patnext
++) == EOS
) {
344 *bufnext
++ = (Char
) (c
| M_PROTECT
);
347 *bufnext
++ = (Char
) c
;
351 while ((c
= *patnext
++) != EOS
)
352 *bufnext
++ = (Char
) c
;
357 while ((c
= *qpatnext
++) != EOS
) {
363 if (*qpatnext
== EOS
||
364 Strchr(qpatnext
+ 1, RBRACKET
) == NULL
) {
365 *bufnext
++ = LBRACKET
;
370 pglob
->gl_flags
|= GLOB_MAGCHAR
;
376 *bufnext
++ = LCHAR(c
);
377 if (*qpatnext
== RANGE
&&
378 (c
= qpatnext
[1]) != RBRACKET
) {
380 *bufnext
++ = LCHAR(c
);
383 } while ((c
= *qpatnext
++) != RBRACKET
);
387 pglob
->gl_flags
|= GLOB_MAGCHAR
;
391 pglob
->gl_flags
|= GLOB_MAGCHAR
;
392 /* collapse adjacent stars to one, to avoid
393 * exponential behavior
395 if (bufnext
== patbuf
|| bufnext
[-1] != M_ALL
)
399 *bufnext
++ = LCHAR(c
);
408 if ((err
= glob1(patbuf
, pglob
, no_match
)) != 0) {
414 * If there was no match we are going to append the pattern
415 * if GLOB_NOCHECK was specified or if GLOB_NOMAGIC was specified
416 * and the pattern did not contain any magic characters
417 * GLOB_NOMAGIC is there just for compatibility with csh.
419 if (pglob
->gl_pathc
== oldpathc
&&
420 ((flags
& GLOB_NOCHECK
) ||
421 ((flags
& GLOB_NOMAGIC
) && !(pglob
->gl_flags
& GLOB_MAGCHAR
)))) {
422 if (!(flags
& GLOB_QUOTE
))
423 globextend(pattern
, pglob
);
428 /* copy pattern, interpreting quotes */
429 copy
= xmalloc(strlen(pattern
) + 1);
432 while (*src
!= EOS
) {
440 globextend(copy
, pglob
);
446 else if (!(flags
& GLOB_NOSORT
) && (pglob
->gl_pathc
!= oldpathc
))
447 qsort(pglob
->gl_pathv
+ pglob
->gl_offs
+ oldpathc
,
448 pglob
->gl_pathc
- oldpathc
, sizeof(char *), compare
);
454 glob1(Char
*pattern
, glob_t
*pglob
, int no_match
)
456 struct strbuf pathbuf
= strbuf_INIT
;
460 * a null pathname is invalid -- POSIX 1003.1 sect. 2.4.
464 err
= glob2(&pathbuf
, pattern
, pglob
, no_match
);
470 * functions glob2 and glob3 are mutually recursive; there is one level
471 * of recursion for each segment in the pattern that contains one or
472 * more meta characters.
475 glob2(struct strbuf
*pathbuf
, const Char
*pattern
, glob_t
*pglob
, int no_match
)
483 * loop over pattern segments until end of pattern or until segment with
484 * meta character found.
488 if (*pattern
== EOS
) { /* end of pattern? */
489 strbuf_terminate(pathbuf
);
491 if (Lstat(pathbuf
->s
, &sbuf
))
494 if (((pglob
->gl_flags
& GLOB_MARK
) &&
495 pathbuf
->s
[pathbuf
->len
- 1] != SEP
) &&
496 (S_ISDIR(sbuf
.st_mode
)
498 || (S_ISLNK(sbuf
.st_mode
) &&
499 (Stat(pathbuf
->s
, &sbuf
) == 0) &&
500 S_ISDIR(sbuf
.st_mode
))
503 strbuf_append1(pathbuf
, SEP
);
504 strbuf_terminate(pathbuf
);
507 globextend(pathbuf
->s
, pglob
);
511 /* find end of next segment, tentatively copy to pathbuf */
513 orig_len
= pathbuf
->len
;
514 while (*p
!= EOS
&& *p
!= SEP
) {
517 strbuf_append1(pathbuf
, *p
++);
520 if (!anymeta
) { /* no expansion, do next segment */
522 while (*pattern
== SEP
)
523 strbuf_append1(pathbuf
, *pattern
++);
525 else { /* need expansion, recurse */
526 pathbuf
->len
= orig_len
;
527 return (glob3(pathbuf
, pattern
, p
, pglob
, no_match
));
535 glob3(struct strbuf
*pathbuf
, const Char
*pattern
, const Char
*restpattern
,
536 glob_t
*pglob
, int no_match
)
541 Char m_not
= (pglob
->gl_flags
& GLOB_ALTNOT
) ? M_ALTNOT
: M_NOT
;
544 strbuf_terminate(pathbuf
);
547 if (!(dirp
= Opendir(pathbuf
->s
))) {
548 /* todo: don't call for ENOENT or ENOTDIR? */
549 if ((pglob
->gl_errfunc
&& (*pglob
->gl_errfunc
) (pathbuf
->s
, errno
)) ||
550 (pglob
->gl_flags
& GLOB_ERR
))
558 orig_len
= pathbuf
->len
;
559 /* search directory for matching names */
560 while ((dp
= readdir(dirp
)) != NULL
) {
561 /* initial DOT must be matched literally */
562 if (dp
->d_name
[0] == DOT
&& *pattern
!= DOT
)
564 pathbuf
->len
= orig_len
;
565 strbuf_append(pathbuf
, dp
->d_name
);
566 strbuf_terminate(pathbuf
);
567 if (match(pathbuf
->s
+ orig_len
, pattern
, restpattern
, (int) m_not
)
570 err
= glob2(pathbuf
, restpattern
, pglob
, no_match
);
574 /* todo: check error from readdir? */
581 * Extend the gl_pathv member of a glob_t structure to accomodate a new item,
582 * add the new item, and update gl_pathc.
584 * This assumes the BSD realloc, which only copies the block when its size
585 * crosses a power-of-two boundary; for v7 realloc, this would cause quadratic
588 * Return 0 if new item added, error code if memory couldn't be allocated.
590 * Invariant of the glob_t structure:
591 * Either gl_pathc is zero and gl_pathv is NULL; or gl_pathc > 0 and
592 * gl_pathv points to (gl_offs + gl_pathc + 1) items.
595 globextend(const char *path
, glob_t
*pglob
)
601 newsize
= sizeof(*pathv
) * (2 + pglob
->gl_pathc
+ pglob
->gl_offs
);
602 pathv
= xrealloc(pglob
->gl_pathv
, newsize
);
604 if (pglob
->gl_pathv
== NULL
&& pglob
->gl_offs
> 0) {
605 /* first time around -- clear initial gl_offs items */
606 pathv
+= pglob
->gl_offs
;
607 for (i
= pglob
->gl_offs
; --i
>= 0;)
610 pglob
->gl_pathv
= pathv
;
612 pathv
[pglob
->gl_offs
+ pglob
->gl_pathc
++] = strsave(path
);
613 pathv
[pglob
->gl_offs
+ pglob
->gl_pathc
] = NULL
;
617 One_Char_mbtowc(__Char
*pwc
, const Char
*s
, size_t n
)
620 char buf
[MB_LEN_MAX
], *p
;
625 while (p
< buf
+ n
&& (*p
++ = LCHAR(*s
++)) != 0)
627 return one_mbtowc(pwc
, buf
, n
);
635 * pattern matching function for filenames. Each occurrence of the *
636 * pattern causes a recursion level.
639 match(const char *name
, const Char
*pat
, const Char
*patend
, int m_not
)
641 int ok
, negate_range
;
644 while (pat
< patend
) {
648 c
= *pat
; /* Only for M_MASK bits */
649 pat
+= One_Char_mbtowc(&wc
, pat
, MB_LEN_MAX
);
650 lwk
= one_mbtowc(&wk
, name
, MB_LEN_MAX
);
651 switch (c
& M_MASK
) {
656 if (match(name
, pat
, patend
, m_not
))
661 lwk
= one_mbtowc(&wk
, name
, MB_LEN_MAX
);
674 if ((negate_range
= ((*pat
& M_MASK
) == m_not
)) != 0)
676 while ((*pat
& M_MASK
) != M_END
) {
677 pat
+= One_Char_mbtowc(&wc
, pat
, MB_LEN_MAX
);
678 if ((*pat
& M_MASK
) == M_RNG
) {
682 pat
+= One_Char_mbtowc(&wc2
, pat
, MB_LEN_MAX
);
683 if (globcharcoll(wc
, wk
, 0) <= 0 &&
684 globcharcoll(wk
, wc2
, 0) <= 0)
689 pat
+= One_Char_mbtowc(&wc
, pat
, MB_LEN_MAX
);
690 if (ok
== negate_range
)
695 if (samecase(wk
) != samecase(wc
))
700 return (*name
== EOS
);
703 /* free allocated data belonging to a glob_t structure */
705 globfree(glob_t
*pglob
)
710 if (pglob
->gl_pathv
!= NULL
) {
711 pp
= pglob
->gl_pathv
+ pglob
->gl_offs
;
712 for (i
= pglob
->gl_pathc
; i
--; ++pp
)
714 xfree(*pp
), *pp
= NULL
;
715 xfree(pglob
->gl_pathv
), pglob
->gl_pathv
= NULL
;