[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 FALSE 0
#define TRUE 1

#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, int force_lower_case)
{
	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 ABORT_ALL;
		if (force_lower_case && ISUPPER(t_ch))
			t_ch = tolower(t_ch);
		if (force_lower_case && 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 NOMATCH;
			continue;
		case '?':
			/* Match anything but '/'. */
			if (t_ch == '/')
				return 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, force_lower_case) == MATCH)
						return MATCH;
					match_slash = TRUE;
				} else
					return ABORT_MALFORMED;
			} else
				match_slash = FALSE;
			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 NOMATCH;
				}
				return MATCH;
			}
			while (1) {
				if (t_ch == '\0')
					break;
				if ((matched = dowild(p, text,  force_lower_case)) != NOMATCH) {
					if (!match_slash || matched != ABORT_TO_STARSTAR)
						return matched;
				} else if (!match_slash && t_ch == '/')
					return ABORT_TO_STARSTAR;
				t_ch = *++text;
			}
			return ABORT_ALL;
		case '[':
			p_ch = *++p;
#ifdef NEGATE_CLASS2
			if (p_ch == NEGATE_CLASS2)
				p_ch = NEGATE_CLASS;
#endif
			/* Assign literal TRUE/FALSE because of "matched" comparison. */
			negated = p_ch == NEGATE_CLASS? TRUE : FALSE;
			if (negated) {
				/* Inverted character class. */
				p_ch = *++p;
			}
			prev_ch = 0;
			matched = FALSE;
			do {
				if (!p_ch)
					return ABORT_ALL;
				if (p_ch == '\\') {
					p_ch = *++p;
					if (!p_ch)
						return ABORT_ALL;
					if (t_ch == p_ch)
						matched = TRUE;
				} else if (p_ch == '-' && prev_ch && p[1] && p[1] != ']') {
					p_ch = *++p;
					if (p_ch == '\\') {
						p_ch = *++p;
						if (!p_ch)
							return ABORT_ALL;
					}
					if (t_ch <= p_ch && t_ch >= prev_ch)
						matched = TRUE;
					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 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 = TRUE;
						continue;
					}
					if (CC_EQ(s,i, "alnum")) {
						if (ISALNUM(t_ch))
							matched = TRUE;
					} else if (CC_EQ(s,i, "alpha")) {
						if (ISALPHA(t_ch))
							matched = TRUE;
					} else if (CC_EQ(s,i, "blank")) {
						if (ISBLANK(t_ch))
							matched = TRUE;
					} else if (CC_EQ(s,i, "cntrl")) {
						if (ISCNTRL(t_ch))
							matched = TRUE;
					} else if (CC_EQ(s,i, "digit")) {
						if (ISDIGIT(t_ch))
							matched = TRUE;
					} else if (CC_EQ(s,i, "graph")) {
						if (ISGRAPH(t_ch))
							matched = TRUE;
					} else if (CC_EQ(s,i, "lower")) {
						if (ISLOWER(t_ch))
							matched = TRUE;
					} else if (CC_EQ(s,i, "print")) {
						if (ISPRINT(t_ch))
							matched = TRUE;
					} else if (CC_EQ(s,i, "punct")) {
						if (ISPUNCT(t_ch))
							matched = TRUE;
					} else if (CC_EQ(s,i, "space")) {
						if (ISSPACE(t_ch))
							matched = TRUE;
					} else if (CC_EQ(s,i, "upper")) {
						if (ISUPPER(t_ch))
							matched = TRUE;
					} else if (CC_EQ(s,i, "xdigit")) {
						if (ISXDIGIT(t_ch))
							matched = TRUE;
					} else /* malformed [:class:] string */
						return ABORT_ALL;
					p_ch = 0; /* This makes "prev_ch" get set to 0. */
				} else if (t_ch == p_ch)
					matched = TRUE;
			} while (prev_ch = p_ch, (p_ch = *++p) != ']');
			if (matched == negated || t_ch == '/')
				return NOMATCH;
			continue;
		}
	}

	return *text ? NOMATCH : MATCH;
}

/* Match the "pattern" against the "text" string. */
int wildmatch(const char *pattern, const char *text, int flags)
{
	return dowild((const uchar*)pattern, (const uchar*)text,
		      flags & FNM_CASEFOLD ? 1 :0);
}
Commit	Line	Data
	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
	12	#include "cache.h"
	13	#include "wildmatch.h"
	14
	15	typedef unsigned char uchar;
	16
	17	/* What character marks an inverted character class? */
	18	#define NEGATE_CLASS '!'
	19	#define NEGATE_CLASS2 '^'
	20
	21	#define FALSE 0
	22	#define TRUE 1
	23
	24	#define CC_EQ(class, len, litmatch) ((len) == sizeof (litmatch)-1 \
	25	&& (class) == (litmatch) \
	26	&& strncmp((char*)class, litmatch, len) == 0)
	27
	28	#if defined STDC_HEADERS \|\| !defined isascii
	29	# define ISASCII(c) 1
	30	#else
	31	# define ISASCII(c) isascii(c)
	32	#endif
	33
	34	#ifdef isblank
	35	# define ISBLANK(c) (ISASCII(c) && isblank(c))
	36	#else
	37	# define ISBLANK(c) ((c) == ' ' \|\| (c) == '\t')
	38	#endif
	39
	40	#ifdef isgraph
	41	# define ISGRAPH(c) (ISASCII(c) && isgraph(c))
	42	#else
	43	# define ISGRAPH(c) (ISASCII(c) && isprint(c) && !isspace(c))
	44	#endif
	45
	46	#define ISPRINT(c) (ISASCII(c) && isprint(c))
	47	#define ISDIGIT(c) (ISASCII(c) && isdigit(c))
	48	#define ISALNUM(c) (ISASCII(c) && isalnum(c))
	49	#define ISALPHA(c) (ISASCII(c) && isalpha(c))
	50	#define ISCNTRL(c) (ISASCII(c) && iscntrl(c))
	51	#define ISLOWER(c) (ISASCII(c) && islower(c))
	52	#define ISPUNCT(c) (ISASCII(c) && ispunct(c))
	53	#define ISSPACE(c) (ISASCII(c) && isspace(c))
	54	#define ISUPPER(c) (ISASCII(c) && isupper(c))
	55	#define ISXDIGIT(c) (ISASCII(c) && isxdigit(c))
	56
	57	/* Match pattern "p" against "text" */
	58	static int dowild(const uchar p, const uchar text, int force_lower_case)
	59	{
	60	uchar p_ch;
	61	const uchar *pattern = p;
	62
	63	for ( ; (p_ch = *p) != '\0'; text++, p++) {
	64	int matched, match_slash, negated;
	65	uchar t_ch, prev_ch;
	66	if ((t_ch = text) == '\0' && p_ch != '')
	67	return ABORT_ALL;
	68	if (force_lower_case && ISUPPER(t_ch))
	69	t_ch = tolower(t_ch);
	70	if (force_lower_case && ISUPPER(p_ch))
	71	p_ch = tolower(p_ch);
	72	switch (p_ch) {
	73	case '\\':
	74	/* Literal match with following character. Note that the test
	75	* in "default" handles the p[1] == '\0' failure case. */
	76	p_ch = *++p;
	77	/* FALLTHROUGH */
	78	default:
	79	if (t_ch != p_ch)
	80	return NOMATCH;
	81	continue;
	82	case '?':
	83	/* Match anything but '/'. */
	84	if (t_ch == '/')
	85	return NOMATCH;
	86	continue;
	87	case '*':
	88	if (++p == '') {
	89	const uchar *prev_p = p - 2;
	90	while (++p == '') {}
	91	if ((prev_p < pattern \|\| *prev_p == '/') &&
	92	(p == '\0' \|\| p == '/' \|\|
	93	(p[0] == '\\' && p[1] == '/'))) {
	94	/*
	95	* Assuming we already match 'foo/' and are at
	96	* <star star slash>, just assume it matches
	97	* nothing and go ahead match the rest of the
	98	* pattern with the remaining string. This
	99	* helps make foo/<><>/bar (<> because
	100	* otherwise it breaks C comment syntax) match
	101	* both foo/bar and foo/a/bar.
	102	*/
	103	if (p[0] == '/' &&
	104	dowild(p + 1, text, force_lower_case) == MATCH)
	105	return MATCH;
	106	match_slash = TRUE;
	107	} else
	108	return ABORT_MALFORMED;
	109	} else
	110	match_slash = FALSE;
	111	if (*p == '\0') {
	112	/* Trailing "*" matches everything. Trailing "" matches
	113	* only if there are no more slash characters. */
	114	if (!match_slash) {
	115	if (strchr((char*)text, '/') != NULL)
	116	return NOMATCH;
	117	}
	118	return MATCH;
	119	}
	120	while (1) {
	121	if (t_ch == '\0')
	122	break;
	123	if ((matched = dowild(p, text, force_lower_case)) != NOMATCH) {
	124	if (!match_slash \|\| matched != ABORT_TO_STARSTAR)
	125	return matched;
	126	} else if (!match_slash && t_ch == '/')
	127	return ABORT_TO_STARSTAR;
	128	t_ch = *++text;
	129	}
	130	return ABORT_ALL;
	131	case '[':
	132	p_ch = *++p;
	133	#ifdef NEGATE_CLASS2
	134	if (p_ch == NEGATE_CLASS2)
	135	p_ch = NEGATE_CLASS;
	136	#endif
	137	/* Assign literal TRUE/FALSE because of "matched" comparison. */
	138	negated = p_ch == NEGATE_CLASS? TRUE : FALSE;
	139	if (negated) {
	140	/* Inverted character class. */
	141	p_ch = *++p;
	142	}
	143	prev_ch = 0;
	144	matched = FALSE;
	145	do {
	146	if (!p_ch)
	147	return ABORT_ALL;
	148	if (p_ch == '\\') {
	149	p_ch = *++p;
	150	if (!p_ch)
	151	return ABORT_ALL;
	152	if (t_ch == p_ch)
	153	matched = TRUE;
	154	} else if (p_ch == '-' && prev_ch && p[1] && p[1] != ']') {
	155	p_ch = *++p;
	156	if (p_ch == '\\') {
	157	p_ch = *++p;
	158	if (!p_ch)
	159	return ABORT_ALL;
	160	}
	161	if (t_ch <= p_ch && t_ch >= prev_ch)
	162	matched = TRUE;
	163	p_ch = 0; /* This makes "prev_ch" get set to 0. */
	164	} else if (p_ch == '[' && p[1] == ':') {
	165	const uchar *s;
	166	int i;
	167	for (s = p += 2; (p_ch = p) && p_ch != ']'; p++) {} /SHARED ITERATOR*/
	168	if (!p_ch)
	169	return ABORT_ALL;
	170	i = p - s - 1;
	171	if (i < 0 \|\| p[-1] != ':') {
	172	/* Didn't find ":]", so treat like a normal set. */
	173	p = s - 2;
	174	p_ch = '[';
	175	if (t_ch == p_ch)
	176	matched = TRUE;
	177	continue;
	178	}
	179	if (CC_EQ(s,i, "alnum")) {
	180	if (ISALNUM(t_ch))
	181	matched = TRUE;
	182	} else if (CC_EQ(s,i, "alpha")) {
	183	if (ISALPHA(t_ch))
	184	matched = TRUE;
	185	} else if (CC_EQ(s,i, "blank")) {
	186	if (ISBLANK(t_ch))
	187	matched = TRUE;
	188	} else if (CC_EQ(s,i, "cntrl")) {
	189	if (ISCNTRL(t_ch))
	190	matched = TRUE;
	191	} else if (CC_EQ(s,i, "digit")) {
	192	if (ISDIGIT(t_ch))
	193	matched = TRUE;
	194	} else if (CC_EQ(s,i, "graph")) {
	195	if (ISGRAPH(t_ch))
	196	matched = TRUE;
	197	} else if (CC_EQ(s,i, "lower")) {
	198	if (ISLOWER(t_ch))
	199	matched = TRUE;
	200	} else if (CC_EQ(s,i, "print")) {
	201	if (ISPRINT(t_ch))
	202	matched = TRUE;
	203	} else if (CC_EQ(s,i, "punct")) {
	204	if (ISPUNCT(t_ch))
	205	matched = TRUE;
	206	} else if (CC_EQ(s,i, "space")) {
	207	if (ISSPACE(t_ch))
	208	matched = TRUE;
	209	} else if (CC_EQ(s,i, "upper")) {
	210	if (ISUPPER(t_ch))
	211	matched = TRUE;
	212	} else if (CC_EQ(s,i, "xdigit")) {
	213	if (ISXDIGIT(t_ch))
	214	matched = TRUE;
	215	} else /* malformed [:class:] string */
	216	return ABORT_ALL;
	217	p_ch = 0; /* This makes "prev_ch" get set to 0. */
	218	} else if (t_ch == p_ch)
	219	matched = TRUE;
	220	} while (prev_ch = p_ch, (p_ch = *++p) != ']');
	221	if (matched == negated \|\| t_ch == '/')
	222	return NOMATCH;
	223	continue;
	224	}
	225	}
	226
	227	return *text ? NOMATCH : MATCH;
	228	}
	229
	230	/* Match the "pattern" against the "text" string. */
	231	int wildmatch(const char pattern, const char text, int flags)
	232	{
	233	return dowild((const uchar)pattern, (const uchar)text,
	234	flags & FNM_CASEFOLD ? 1 :0);
	235	}