[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 NOMATCH 1
#define MATCH 0
#define ABORT_ALL -1
#define ABORT_TO_STARSTAR -2

#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;

	for ( ; (p_ch = *p) != '\0'; text++, p++) {
		int matched, special;
		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);
		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 == '*') {
				while (*++p == '*') {}
				special = TRUE;
			} else
				special = FALSE;
			if (*p == '\0') {
				/* Trailing "**" matches everything.  Trailing "*" matches
				 * only if there are no more slash characters. */
				if (!special) {
					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 (!special || matched != ABORT_TO_STARSTAR)
						return matched;
				} else if (!special && 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. */
			special = p_ch == NEGATE_CLASS? TRUE : FALSE;
			if (special) {
				/* 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 == special || 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
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
	21	#define FALSE 0
	22	#define TRUE 1
3ae5396c NTND	23
	24	#define NOMATCH 1
	25	#define MATCH 0
5230f605 NTND	26	#define ABORT_ALL -1
	27	#define ABORT_TO_STARSTAR -2
	28
	29	#define CC_EQ(class, len, litmatch) ((len) == sizeof (litmatch)-1 \
	30	&& (class) == (litmatch) \
	31	&& strncmp((char*)class, litmatch, len) == 0)
	32
	33	#if defined STDC_HEADERS \|\| !defined isascii
	34	# define ISASCII(c) 1
	35	#else
	36	# define ISASCII(c) isascii(c)
	37	#endif
	38
	39	#ifdef isblank
	40	# define ISBLANK(c) (ISASCII(c) && isblank(c))
	41	#else
	42	# define ISBLANK(c) ((c) == ' ' \|\| (c) == '\t')
	43	#endif
	44
	45	#ifdef isgraph
	46	# define ISGRAPH(c) (ISASCII(c) && isgraph(c))
	47	#else
	48	# define ISGRAPH(c) (ISASCII(c) && isprint(c) && !isspace(c))
	49	#endif
	50
	51	#define ISPRINT(c) (ISASCII(c) && isprint(c))
	52	#define ISDIGIT(c) (ISASCII(c) && isdigit(c))
	53	#define ISALNUM(c) (ISASCII(c) && isalnum(c))
	54	#define ISALPHA(c) (ISASCII(c) && isalpha(c))
	55	#define ISCNTRL(c) (ISASCII(c) && iscntrl(c))
	56	#define ISLOWER(c) (ISASCII(c) && islower(c))
	57	#define ISPUNCT(c) (ISASCII(c) && ispunct(c))
	58	#define ISSPACE(c) (ISASCII(c) && isspace(c))
	59	#define ISUPPER(c) (ISASCII(c) && isupper(c))
	60	#define ISXDIGIT(c) (ISASCII(c) && isxdigit(c))
	61
b0e02876	62	/* Match pattern "p" against "text" */
9b4edc0a	63	static int dowild(const uchar p, const uchar text, int force_lower_case)
5230f605	64	{
327f2f3e	65	uchar p_ch;
5230f605	66
327f2f3e NTND	67	for ( ; (p_ch = *p) != '\0'; text++, p++) {
	68	int matched, special;
	69	uchar t_ch, prev_ch;
	70	if ((t_ch = text) == '\0' && p_ch != '')
5230f605	71	return ABORT_ALL;
327f2f3e NTND	72	if (force_lower_case && ISUPPER(t_ch))
	73	t_ch = tolower(t_ch);
	74	switch (p_ch) {
	75	case '\\':
	76	/* Literal match with following character. Note that the test
	77	* in "default" handles the p[1] == '\0' failure case. */
5230f605	78	p_ch = *++p;
327f2f3e NTND	79	/* FALLTHROUGH */
	80	default:
	81	if (t_ch != p_ch)
3ae5396c	82	return NOMATCH;
327f2f3e NTND	83	continue;
	84	case '?':
	85	/* Match anything but '/'. */
	86	if (t_ch == '/')
3ae5396c	87	return NOMATCH;
5230f605	88	continue;
327f2f3e NTND	89	case '*':
	90	if (++p == '') {
	91	while (++p == '') {}
	92	special = TRUE;
	93	} else
	94	special = FALSE;
	95	if (*p == '\0') {
	96	/* Trailing "*" matches everything. Trailing "" matches
	97	* only if there are no more slash characters. */
	98	if (!special) {
	99	if (strchr((char*)text, '/') != NULL)
3ae5396c	100	return NOMATCH;
327f2f3e	101	}
3ae5396c	102	return MATCH;
327f2f3e NTND	103	}
	104	while (1) {
	105	if (t_ch == '\0')
	106	break;
9b4edc0a	107	if ((matched = dowild(p, text, force_lower_case)) != NOMATCH) {
327f2f3e NTND	108	if (!special \|\| matched != ABORT_TO_STARSTAR)
	109	return matched;
	110	} else if (!special && t_ch == '/')
	111	return ABORT_TO_STARSTAR;
	112	t_ch = *++text;
	113	}
5230f605	114	return ABORT_ALL;
327f2f3e NTND	115	case '[':
	116	p_ch = *++p;
	117	#ifdef NEGATE_CLASS2
	118	if (p_ch == NEGATE_CLASS2)
	119	p_ch = NEGATE_CLASS;
	120	#endif
	121	/* Assign literal TRUE/FALSE because of "matched" comparison. */
	122	special = p_ch == NEGATE_CLASS? TRUE : FALSE;
	123	if (special) {
	124	/* Inverted character class. */
	125	p_ch = *++p;
	126	}
	127	prev_ch = 0;
	128	matched = FALSE;
	129	do {
	130	if (!p_ch)
	131	return ABORT_ALL;
	132	if (p_ch == '\\') {
	133	p_ch = *++p;
	134	if (!p_ch)
	135	return ABORT_ALL;
	136	if (t_ch == p_ch)
	137	matched = TRUE;
	138	} else if (p_ch == '-' && prev_ch && p[1] && p[1] != ']') {
	139	p_ch = *++p;
	140	if (p_ch == '\\') {
	141	p_ch = *++p;
	142	if (!p_ch)
	143	return ABORT_ALL;
	144	}
	145	if (t_ch <= p_ch && t_ch >= prev_ch)
	146	matched = TRUE;
	147	p_ch = 0; /* This makes "prev_ch" get set to 0. */
	148	} else if (p_ch == '[' && p[1] == ':') {
	149	const uchar *s;
	150	int i;
	151	for (s = p += 2; (p_ch = p) && p_ch != ']'; p++) {} /SHARED ITERATOR*/
	152	if (!p_ch)
	153	return ABORT_ALL;
	154	i = p - s - 1;
	155	if (i < 0 \|\| p[-1] != ':') {
	156	/* Didn't find ":]", so treat like a normal set. */
	157	p = s - 2;
	158	p_ch = '[';
	159	if (t_ch == p_ch)
	160	matched = TRUE;
	161	continue;
	162	}
	163	if (CC_EQ(s,i, "alnum")) {
	164	if (ISALNUM(t_ch))
	165	matched = TRUE;
	166	} else if (CC_EQ(s,i, "alpha")) {
	167	if (ISALPHA(t_ch))
	168	matched = TRUE;
	169	} else if (CC_EQ(s,i, "blank")) {
	170	if (ISBLANK(t_ch))
	171	matched = TRUE;
	172	} else if (CC_EQ(s,i, "cntrl")) {
	173	if (ISCNTRL(t_ch))
	174	matched = TRUE;
	175	} else if (CC_EQ(s,i, "digit")) {
	176	if (ISDIGIT(t_ch))
	177	matched = TRUE;
	178	} else if (CC_EQ(s,i, "graph")) {
179	if (ISGRAPH(t_ch))
180	matched = TRUE;
181	} else if (CC_EQ(s,i, "lower")) {
182	if (ISLOWER(t_ch))
183	matched = TRUE;
184	} else if (CC_EQ(s,i, "print")) {
185	if (ISPRINT(t_ch))
186	matched = TRUE;
187	} else if (CC_EQ(s,i, "punct")) {
188	if (ISPUNCT(t_ch))
189	matched = TRUE;
190	} else if (CC_EQ(s,i, "space")) {
191	if (ISSPACE(t_ch))
192	matched = TRUE;
193	} else if (CC_EQ(s,i, "upper")) {
194	if (ISUPPER(t_ch))
195	matched = TRUE;
196	} else if (CC_EQ(s,i, "xdigit")) {
197	if (ISXDIGIT(t_ch))
198	matched = TRUE;
199	} else /* malformed [:class:] string */
200	return ABORT_ALL;
201	p_ch = 0; /* This makes "prev_ch" get set to 0. */
202	} else if (t_ch == p_ch)
203	matched = TRUE;
204	} while (prev_ch = p_ch, (p_ch = *++p) != ']');
205	if (matched == special \|\| t_ch == '/')
3ae5396c	206	return NOMATCH;
327f2f3e NTND	207	continue;
327f2f3e NTND	208	}
5230f605	209	}
5230f605	210
3ae5396c	211	return *text ? NOMATCH : MATCH;
5230f605 NTND	212	}
	213
	214	/* Match the "pattern" against the "text" string. */
9b4edc0a	215	int wildmatch(const char pattern, const char text, int flags)
5230f605	216	{
9b4edc0a NTND	217	return dowild((const uchar)pattern, (const uchar)text,
9b4edc0a NTND	218	flags & FNM_CASEFOLD ? 1 :0);
5230f605	219	}