[thirdparty/git.git] / wildmatch.c

/*
**  Do shell-style pattern matching for ?, \, [], and * characters.
**  It is 8bit clean.
**
**  Written by Rich $alz, mirror!rs, Wed Nov 26 19:03:17 EST 1986.
**  Rich $alz is now <rsalz@bbn.com>.
**
**  Modified by Wayne Davison to special-case '/' matching, to make '**'
**  work differently than '*', and to fix the character-class code.
*/

#include "cache.h"
#include "wildmatch.h"

typedef unsigned char uchar;

/* What character marks an inverted character class? */
#define NEGATE_CLASS	'!'
#define NEGATE_CLASS2	'^'

#define CC_EQ(class, len, litmatch) ((len) == sizeof (litmatch)-1 \
				    && *(class) == *(litmatch) \
				    && strncmp((char*)class, litmatch, len) == 0)

#if defined STDC_HEADERS || !defined isascii
# define ISASCII(c) 1
#else
# define ISASCII(c) isascii(c)
#endif

#ifdef isblank
# define ISBLANK(c) (ISASCII(c) && isblank(c))
#else
# define ISBLANK(c) ((c) == ' ' || (c) == '\t')
#endif

#ifdef isgraph
# define ISGRAPH(c) (ISASCII(c) && isgraph(c))
#else
# define ISGRAPH(c) (ISASCII(c) && isprint(c) && !isspace(c))
#endif

#define ISPRINT(c) (ISASCII(c) && isprint(c))
#define ISDIGIT(c) (ISASCII(c) && isdigit(c))
#define ISALNUM(c) (ISASCII(c) && isalnum(c))
#define ISALPHA(c) (ISASCII(c) && isalpha(c))
#define ISCNTRL(c) (ISASCII(c) && iscntrl(c))
#define ISLOWER(c) (ISASCII(c) && islower(c))
#define ISPUNCT(c) (ISASCII(c) && ispunct(c))
#define ISSPACE(c) (ISASCII(c) && isspace(c))
#define ISUPPER(c) (ISASCII(c) && isupper(c))
#define ISXDIGIT(c) (ISASCII(c) && isxdigit(c))

/* Match pattern "p" against "text" */
static int dowild(const uchar *p, const uchar *text, unsigned int flags)
{
	uchar p_ch;
	const uchar *pattern = p;

	for ( ; (p_ch = *p) != '\0'; text++, p++) {
		int matched, match_slash, negated;
		uchar t_ch, prev_ch;
		if ((t_ch = *text) == '\0' && p_ch != '*')
			return WM_ABORT_ALL;
		if ((flags & WM_CASEFOLD) && ISUPPER(t_ch))
			t_ch = tolower(t_ch);
		if ((flags & WM_CASEFOLD) && ISUPPER(p_ch))
			p_ch = tolower(p_ch);
		switch (p_ch) {
		case '\\':
			/* Literal match with following character.  Note that the test
			 * in "default" handles the p[1] == '\0' failure case. */
			p_ch = *++p;
			/* FALLTHROUGH */
		default:
			if (t_ch != p_ch)
				return WM_NOMATCH;
			continue;
		case '?':
			/* Match anything but '/'. */
			if (t_ch == '/')
				return WM_NOMATCH;
			continue;
		case '*':
			if (*++p == '*') {
				const uchar *prev_p = p - 2;
				while (*++p == '*') {}
				if ((prev_p < pattern || *prev_p == '/') &&
				    (*p == '\0' || *p == '/' ||
				     (p[0] == '\\' && p[1] == '/'))) {
					/*
					 * Assuming we already match 'foo/' and are at
					 * <star star slash>, just assume it matches
					 * nothing and go ahead match the rest of the
					 * pattern with the remaining string. This
					 * helps make foo/<*><*>/bar (<> because
					 * otherwise it breaks C comment syntax) match
					 * both foo/bar and foo/a/bar.
					 */
					if (p[0] == '/' &&
					    dowild(p + 1, text, flags) == WM_MATCH)
						return WM_MATCH;
					match_slash = 1;
				} else
					return WM_ABORT_MALFORMED;
			} else
				match_slash = 0;
			if (*p == '\0') {
				/* Trailing "**" matches everything.  Trailing "*" matches
				 * only if there are no more slash characters. */
				if (!match_slash) {
					if (strchr((char*)text, '/') != NULL)
						return WM_NOMATCH;
				}
				return WM_MATCH;
			}
			while (1) {
				if (t_ch == '\0')
					break;
				if ((matched = dowild(p, text, flags)) != WM_NOMATCH) {
					if (!match_slash || matched != WM_ABORT_TO_STARSTAR)
						return matched;
				} else if (!match_slash && t_ch == '/')
					return WM_ABORT_TO_STARSTAR;
				t_ch = *++text;
			}
			return WM_ABORT_ALL;
		case '[':
			p_ch = *++p;
#ifdef NEGATE_CLASS2
			if (p_ch == NEGATE_CLASS2)
				p_ch = NEGATE_CLASS;
#endif
			/* Assign literal 1/0 because of "matched" comparison. */
			negated = p_ch == NEGATE_CLASS ? 1 : 0;
			if (negated) {
				/* Inverted character class. */
				p_ch = *++p;
			}
			prev_ch = 0;
			matched = 0;
			do {
				if (!p_ch)
					return WM_ABORT_ALL;
				if (p_ch == '\\') {
					p_ch = *++p;
					if (!p_ch)
						return WM_ABORT_ALL;
					if (t_ch == p_ch)
						matched = 1;
				} else if (p_ch == '-' && prev_ch && p[1] && p[1] != ']') {
					p_ch = *++p;
					if (p_ch == '\\') {
						p_ch = *++p;
						if (!p_ch)
							return WM_ABORT_ALL;
					}
					if (t_ch <= p_ch && t_ch >= prev_ch)
						matched = 1;
					p_ch = 0; /* This makes "prev_ch" get set to 0. */
				} else if (p_ch == '[' && p[1] == ':') {
					const uchar *s;
					int i;
					for (s = p += 2; (p_ch = *p) && p_ch != ']'; p++) {} /*SHARED ITERATOR*/
					if (!p_ch)
						return WM_ABORT_ALL;
					i = p - s - 1;
					if (i < 0 || p[-1] != ':') {
						/* Didn't find ":]", so treat like a normal set. */
						p = s - 2;
						p_ch = '[';
						if (t_ch == p_ch)
							matched = 1;
						continue;
					}
					if (CC_EQ(s,i, "alnum")) {
						if (ISALNUM(t_ch))
							matched = 1;
					} else if (CC_EQ(s,i, "alpha")) {
						if (ISALPHA(t_ch))
							matched = 1;
					} else if (CC_EQ(s,i, "blank")) {
						if (ISBLANK(t_ch))
							matched = 1;
					} else if (CC_EQ(s,i, "cntrl")) {
						if (ISCNTRL(t_ch))
							matched = 1;
					} else if (CC_EQ(s,i, "digit")) {
						if (ISDIGIT(t_ch))
							matched = 1;
					} else if (CC_EQ(s,i, "graph")) {
						if (ISGRAPH(t_ch))
							matched = 1;
					} else if (CC_EQ(s,i, "lower")) {
						if (ISLOWER(t_ch))
							matched = 1;
					} else if (CC_EQ(s,i, "print")) {
						if (ISPRINT(t_ch))
							matched = 1;
					} else if (CC_EQ(s,i, "punct")) {
						if (ISPUNCT(t_ch))
							matched = 1;
					} else if (CC_EQ(s,i, "space")) {
						if (ISSPACE(t_ch))
							matched = 1;
					} else if (CC_EQ(s,i, "upper")) {
						if (ISUPPER(t_ch))
							matched = 1;
					} else if (CC_EQ(s,i, "xdigit")) {
						if (ISXDIGIT(t_ch))
							matched = 1;
					} else /* malformed [:class:] string */
						return WM_ABORT_ALL;
					p_ch = 0; /* This makes "prev_ch" get set to 0. */
				} else if (t_ch == p_ch)
					matched = 1;
			} while (prev_ch = p_ch, (p_ch = *++p) != ']');
			if (matched == negated || t_ch == '/')
				return WM_NOMATCH;
			continue;
		}
	}

	return *text ? WM_NOMATCH : WM_MATCH;
}

/* Match the "pattern" against the "text" string. */
int wildmatch(const char *pattern, const char *text,
	      unsigned int flags, struct wildopts *wo)
{
	return dowild((const uchar*)pattern, (const uchar*)text, flags);
}
Commit	Line	Data
5230f605 NTND	1	/*
	2	** Do shell-style pattern matching for ?, \, [], and * characters.
	3	** It is 8bit clean.
	4	**
	5	** Written by Rich $alz, mirror!rs, Wed Nov 26 19:03:17 EST 1986.
	6	** Rich $alz is now <rsalz@bbn.com>.
	7	**
	8	Modified by Wayne Davison to special-case '/' matching, to make ''
	9	** work differently than '*', and to fix the character-class code.
	10	*/
	11
feabcc17 NTND	12	#include "cache.h"
	13	#include "wildmatch.h"
	14
	15	typedef unsigned char uchar;
5230f605 NTND	16
	17	/* What character marks an inverted character class? */
	18	#define NEGATE_CLASS '!'
	19	#define NEGATE_CLASS2 '^'
	20
5230f605 NTND	21	#define CC_EQ(class, len, litmatch) ((len) == sizeof (litmatch)-1 \
	22	&& (class) == (litmatch) \
	23	&& strncmp((char*)class, litmatch, len) == 0)
	24
	25	#if defined STDC_HEADERS \|\| !defined isascii
	26	# define ISASCII(c) 1
	27	#else
	28	# define ISASCII(c) isascii(c)
	29	#endif
	30
	31	#ifdef isblank
	32	# define ISBLANK(c) (ISASCII(c) && isblank(c))
	33	#else
	34	# define ISBLANK(c) ((c) == ' ' \|\| (c) == '\t')
	35	#endif
	36
	37	#ifdef isgraph
	38	# define ISGRAPH(c) (ISASCII(c) && isgraph(c))
	39	#else
	40	# define ISGRAPH(c) (ISASCII(c) && isprint(c) && !isspace(c))
	41	#endif
	42
	43	#define ISPRINT(c) (ISASCII(c) && isprint(c))
	44	#define ISDIGIT(c) (ISASCII(c) && isdigit(c))
	45	#define ISALNUM(c) (ISASCII(c) && isalnum(c))
	46	#define ISALPHA(c) (ISASCII(c) && isalpha(c))
	47	#define ISCNTRL(c) (ISASCII(c) && iscntrl(c))
	48	#define ISLOWER(c) (ISASCII(c) && islower(c))
	49	#define ISPUNCT(c) (ISASCII(c) && ispunct(c))
	50	#define ISSPACE(c) (ISASCII(c) && isspace(c))
	51	#define ISUPPER(c) (ISASCII(c) && isupper(c))
	52	#define ISXDIGIT(c) (ISASCII(c) && isxdigit(c))
	53
b0e02876	54	/* Match pattern "p" against "text" */
0c528168	55	static int dowild(const uchar p, const uchar text, unsigned int flags)
5230f605	56	{
327f2f3e	57	uchar p_ch;
3a078dec	58	const uchar *pattern = p;
5230f605	59
327f2f3e	60	for ( ; (p_ch = *p) != '\0'; text++, p++) {
b6a3d335	61	int matched, match_slash, negated;
327f2f3e NTND	62	uchar t_ch, prev_ch;
327f2f3e NTND	63	if ((t_ch = text) == '\0' && p_ch != '')
9b3497ca	64	return WM_ABORT_ALL;
0c528168	65	if ((flags & WM_CASEFOLD) && ISUPPER(t_ch))
327f2f3e	66	t_ch = tolower(t_ch);
0c528168	67	if ((flags & WM_CASEFOLD) && ISUPPER(p_ch))
164bf83a	68	p_ch = tolower(p_ch);
327f2f3e NTND	69	switch (p_ch) {
	70	case '\\':
	71	/* Literal match with following character. Note that the test
	72	* in "default" handles the p[1] == '\0' failure case. */
5230f605	73	p_ch = *++p;
327f2f3e NTND	74	/* FALLTHROUGH */
	75	default:
	76	if (t_ch != p_ch)
9b3497ca	77	return WM_NOMATCH;
327f2f3e NTND	78	continue;
	79	case '?':
	80	/* Match anything but '/'. */
	81	if (t_ch == '/')
9b3497ca	82	return WM_NOMATCH;
5230f605	83	continue;
327f2f3e NTND	84	case '*':
327f2f3e NTND	85	if (++p == '') {
40bbee0a	86	const uchar *prev_p = p - 2;
327f2f3e	87	while (++p == '') {}
3a078dec	88	if ((prev_p < pattern \|\| *prev_p == '/') &&
40bbee0a NTND	89	(p == '\0' \|\| p == '/' \|\|
40bbee0a NTND	90	(p[0] == '\\' && p[1] == '/'))) {
4c251e5c NTND	91	/*
	92	* Assuming we already match 'foo/' and are at
	93	* <star star slash>, just assume it matches
	94	* nothing and go ahead match the rest of the
	95	* pattern with the remaining string. This
	96	* helps make foo/<><>/bar (<> because
	97	* otherwise it breaks C comment syntax) match
	98	* both foo/bar and foo/a/bar.
	99	*/
	100	if (p[0] == '/' &&
0c528168	101	dowild(p + 1, text, flags) == WM_MATCH)
9b3497ca NTND	102	return WM_MATCH;
9b3497ca NTND	103	match_slash = 1;
40bbee0a	104	} else
9b3497ca	105	return WM_ABORT_MALFORMED;
327f2f3e	106	} else
9b3497ca	107	match_slash = 0;
327f2f3e NTND	108	if (*p == '\0') {
	109	/* Trailing "*" matches everything. Trailing "" matches
	110	* only if there are no more slash characters. */
b6a3d335	111	if (!match_slash) {
327f2f3e	112	if (strchr((char*)text, '/') != NULL)
9b3497ca	113	return WM_NOMATCH;
327f2f3e	114	}
9b3497ca	115	return WM_MATCH;
327f2f3e NTND	116	}
	117	while (1) {
	118	if (t_ch == '\0')
	119	break;
0c528168	120	if ((matched = dowild(p, text, flags)) != WM_NOMATCH) {
9b3497ca	121	if (!match_slash \|\| matched != WM_ABORT_TO_STARSTAR)
327f2f3e	122	return matched;
b6a3d335	123	} else if (!match_slash && t_ch == '/')
9b3497ca	124	return WM_ABORT_TO_STARSTAR;
327f2f3e NTND	125	t_ch = *++text;
327f2f3e NTND	126	}
9b3497ca	127	return WM_ABORT_ALL;
327f2f3e NTND	128	case '[':
	129	p_ch = *++p;
	130	#ifdef NEGATE_CLASS2
	131	if (p_ch == NEGATE_CLASS2)
	132	p_ch = NEGATE_CLASS;
	133	#endif
9b3497ca NTND	134	/* Assign literal 1/0 because of "matched" comparison. */
9b3497ca NTND	135	negated = p_ch == NEGATE_CLASS ? 1 : 0;
b6a3d335	136	if (negated) {
327f2f3e NTND	137	/* Inverted character class. */
	138	p_ch = *++p;
	139	}
	140	prev_ch = 0;
9b3497ca	141	matched = 0;
327f2f3e NTND	142	do {
327f2f3e NTND	143	if (!p_ch)
9b3497ca	144	return WM_ABORT_ALL;
327f2f3e NTND	145	if (p_ch == '\\') {
	146	p_ch = *++p;
	147	if (!p_ch)
9b3497ca	148	return WM_ABORT_ALL;
327f2f3e	149	if (t_ch == p_ch)
9b3497ca	150	matched = 1;
327f2f3e NTND	151	} else if (p_ch == '-' && prev_ch && p[1] && p[1] != ']') {
	152	p_ch = *++p;
	153	if (p_ch == '\\') {
	154	p_ch = *++p;
	155	if (!p_ch)
9b3497ca	156	return WM_ABORT_ALL;
327f2f3e NTND	157	}
327f2f3e NTND	158	if (t_ch <= p_ch && t_ch >= prev_ch)
9b3497ca	159	matched = 1;
327f2f3e NTND	160	p_ch = 0; /* This makes "prev_ch" get set to 0. */
	161	} else if (p_ch == '[' && p[1] == ':') {
	162	const uchar *s;
	163	int i;
	164	for (s = p += 2; (p_ch = p) && p_ch != ']'; p++) {} /SHARED ITERATOR*/
	165	if (!p_ch)
9b3497ca	166	return WM_ABORT_ALL;
327f2f3e NTND	167	i = p - s - 1;
	168	if (i < 0 \|\| p[-1] != ':') {
	169	/* Didn't find ":]", so treat like a normal set. */
	170	p = s - 2;
	171	p_ch = '[';
	172	if (t_ch == p_ch)
9b3497ca	173	matched = 1;
327f2f3e NTND	174	continue;
	175	}
	176	if (CC_EQ(s,i, "alnum")) {
	177	if (ISALNUM(t_ch))
9b3497ca	178	matched = 1;
327f2f3e NTND	179	} else if (CC_EQ(s,i, "alpha")) {
327f2f3e NTND	180	if (ISALPHA(t_ch))
9b3497ca	181	matched = 1;
327f2f3e NTND	182	} else if (CC_EQ(s,i, "blank")) {
327f2f3e NTND	183	if (ISBLANK(t_ch))
9b3497ca	184	matched = 1;
327f2f3e NTND	185	} else if (CC_EQ(s,i, "cntrl")) {
327f2f3e NTND	186	if (ISCNTRL(t_ch))
9b3497ca	187	matched = 1;
327f2f3e NTND	188	} else if (CC_EQ(s,i, "digit")) {
327f2f3e NTND	189	if (ISDIGIT(t_ch))
9b3497ca	190	matched = 1;
327f2f3e NTND	191	} else if (CC_EQ(s,i, "graph")) {
327f2f3e NTND	192	if (ISGRAPH(t_ch))
9b3497ca	193	matched = 1;
327f2f3e NTND	194	} else if (CC_EQ(s,i, "lower")) {
327f2f3e NTND	195	if (ISLOWER(t_ch))
9b3497ca	196	matched = 1;
327f2f3e NTND	197	} else if (CC_EQ(s,i, "print")) {
327f2f3e NTND	198	if (ISPRINT(t_ch))
9b3497ca	199	matched = 1;
327f2f3e NTND	200	} else if (CC_EQ(s,i, "punct")) {
327f2f3e NTND	201	if (ISPUNCT(t_ch))
9b3497ca	202	matched = 1;
327f2f3e NTND	203	} else if (CC_EQ(s,i, "space")) {
327f2f3e NTND	204	if (ISSPACE(t_ch))
9b3497ca	205	matched = 1;
327f2f3e NTND	206	} else if (CC_EQ(s,i, "upper")) {
327f2f3e NTND	207	if (ISUPPER(t_ch))
9b3497ca	208	matched = 1;
327f2f3e NTND	209	} else if (CC_EQ(s,i, "xdigit")) {
327f2f3e NTND	210	if (ISXDIGIT(t_ch))
9b3497ca	211	matched = 1;
327f2f3e	212	} else /* malformed [:class:] string */
9b3497ca	213	return WM_ABORT_ALL;
327f2f3e NTND	214	p_ch = 0; /* This makes "prev_ch" get set to 0. */
327f2f3e NTND	215	} else if (t_ch == p_ch)
9b3497ca	216	matched = 1;
327f2f3e	217	} while (prev_ch = p_ch, (p_ch = *++p) != ']');
b6a3d335	218	if (matched == negated \|\| t_ch == '/')
9b3497ca	219	return WM_NOMATCH;
327f2f3e NTND	220	continue;
327f2f3e NTND	221	}
5230f605	222	}
5230f605	223
9b3497ca	224	return *text ? WM_NOMATCH : WM_MATCH;
5230f605 NTND	225	}
	226
	227	/* Match the "pattern" against the "text" string. */
9b3497ca NTND	228	int wildmatch(const char pattern, const char text,
9b3497ca NTND	229	unsigned int flags, struct wildopts *wo)
5230f605	230	{
0c528168	231	return dowild((const uchar)pattern, (const uchar)text, flags);
5230f605	232	}