[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 "rsync.h"

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

#define FALSE 0
#define TRUE 1
#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))

static int force_lower_case = 0;

/* Match pattern "p" against "text" */
static int dowild(const uchar *p, const uchar *text)
{
	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 FALSE;
			continue;
		case '?':
			/* Match anything but '/'. */
			if (t_ch == '/')
				return FALSE;
			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 FALSE;
				}
				return TRUE;
			}
			while (1) {
				if (t_ch == '\0')
					break;
				if ((matched = dowild(p, text)) != FALSE) {
					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 FALSE;
			continue;
		}
	}

	return *text ? FALSE : TRUE;
}

/* Match the "pattern" against the "text" string. */
int wildmatch(const char *pattern, const char *text)
{
	return dowild((const uchar*)pattern, (const uchar*)text) == TRUE;
}

/* Match the "pattern" against the forced-to-lower-case "text" string. */
int iwildmatch(const char *pattern, const char *text)
{
	int ret;
	force_lower_case = 1;
	ret = dowild((const uchar*)pattern, (const uchar*)text) == TRUE;
	force_lower_case = 0;
	return ret;
}
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
	12	#include "rsync.h"
	13
	14	/* What character marks an inverted character class? */
	15	#define NEGATE_CLASS '!'
	16	#define NEGATE_CLASS2 '^'
	17
	18	#define FALSE 0
	19	#define TRUE 1
	20	#define ABORT_ALL -1
	21	#define ABORT_TO_STARSTAR -2
	22
	23	#define CC_EQ(class, len, litmatch) ((len) == sizeof (litmatch)-1 \
	24	&& (class) == (litmatch) \
	25	&& strncmp((char*)class, litmatch, len) == 0)
	26
	27	#if defined STDC_HEADERS \|\| !defined isascii
	28	# define ISASCII(c) 1
	29	#else
	30	# define ISASCII(c) isascii(c)
	31	#endif
	32
	33	#ifdef isblank
	34	# define ISBLANK(c) (ISASCII(c) && isblank(c))
	35	#else
	36	# define ISBLANK(c) ((c) == ' ' \|\| (c) == '\t')
	37	#endif
	38
	39	#ifdef isgraph
	40	# define ISGRAPH(c) (ISASCII(c) && isgraph(c))
	41	#else
	42	# define ISGRAPH(c) (ISASCII(c) && isprint(c) && !isspace(c))
	43	#endif
	44
	45	#define ISPRINT(c) (ISASCII(c) && isprint(c))
	46	#define ISDIGIT(c) (ISASCII(c) && isdigit(c))
	47	#define ISALNUM(c) (ISASCII(c) && isalnum(c))
	48	#define ISALPHA(c) (ISASCII(c) && isalpha(c))
	49	#define ISCNTRL(c) (ISASCII(c) && iscntrl(c))
	50	#define ISLOWER(c) (ISASCII(c) && islower(c))
	51	#define ISPUNCT(c) (ISASCII(c) && ispunct(c))
	52	#define ISSPACE(c) (ISASCII(c) && isspace(c))
	53	#define ISUPPER(c) (ISASCII(c) && isupper(c))
	54	#define ISXDIGIT(c) (ISASCII(c) && isxdigit(c))
	55
5230f605 NTND	56	static int force_lower_case = 0;
5230f605 NTND	57
b0e02876 NTND	58	/* Match pattern "p" against "text" */
b0e02876 NTND	59	static int dowild(const uchar p, const uchar text)
5230f605	60	{
327f2f3e	61	uchar p_ch;
5230f605	62
327f2f3e NTND	63	for ( ; (p_ch = *p) != '\0'; text++, p++) {
	64	int matched, special;
	65	uchar t_ch, prev_ch;
	66	if ((t_ch = text) == '\0' && p_ch != '')
5230f605	67	return ABORT_ALL;
327f2f3e NTND	68	if (force_lower_case && ISUPPER(t_ch))
	69	t_ch = tolower(t_ch);
	70	switch (p_ch) {
	71	case '\\':
	72	/* Literal match with following character. Note that the test
	73	* in "default" handles the p[1] == '\0' failure case. */
5230f605	74	p_ch = *++p;
327f2f3e NTND	75	/* FALLTHROUGH */
	76	default:
	77	if (t_ch != p_ch)
	78	return FALSE;
	79	continue;
	80	case '?':
	81	/* Match anything but '/'. */
	82	if (t_ch == '/')
	83	return FALSE;
5230f605	84	continue;
327f2f3e NTND	85	case '*':
	86	if (++p == '') {
	87	while (++p == '') {}
	88	special = TRUE;
	89	} else
	90	special = FALSE;
	91	if (*p == '\0') {
	92	/* Trailing "*" matches everything. Trailing "" matches
	93	* only if there are no more slash characters. */
	94	if (!special) {
	95	if (strchr((char*)text, '/') != NULL)
	96	return FALSE;
	97	}
	98	return TRUE;
	99	}
	100	while (1) {
	101	if (t_ch == '\0')
	102	break;
	103	if ((matched = dowild(p, text)) != FALSE) {
	104	if (!special \|\| matched != ABORT_TO_STARSTAR)
	105	return matched;
	106	} else if (!special && t_ch == '/')
	107	return ABORT_TO_STARSTAR;
	108	t_ch = *++text;
	109	}
5230f605	110	return ABORT_ALL;
327f2f3e NTND	111	case '[':
	112	p_ch = *++p;
	113	#ifdef NEGATE_CLASS2
	114	if (p_ch == NEGATE_CLASS2)
	115	p_ch = NEGATE_CLASS;
	116	#endif
	117	/* Assign literal TRUE/FALSE because of "matched" comparison. */
	118	special = p_ch == NEGATE_CLASS? TRUE : FALSE;
	119	if (special) {
	120	/* Inverted character class. */
	121	p_ch = *++p;
	122	}
	123	prev_ch = 0;
	124	matched = FALSE;
	125	do {
	126	if (!p_ch)
	127	return ABORT_ALL;
	128	if (p_ch == '\\') {
	129	p_ch = *++p;
	130	if (!p_ch)
	131	return ABORT_ALL;
	132	if (t_ch == p_ch)
	133	matched = TRUE;
	134	} else if (p_ch == '-' && prev_ch && p[1] && p[1] != ']') {
	135	p_ch = *++p;
	136	if (p_ch == '\\') {
	137	p_ch = *++p;
	138	if (!p_ch)
	139	return ABORT_ALL;
	140	}
	141	if (t_ch <= p_ch && t_ch >= prev_ch)
	142	matched = TRUE;
	143	p_ch = 0; /* This makes "prev_ch" get set to 0. */
	144	} else if (p_ch == '[' && p[1] == ':') {
	145	const uchar *s;
	146	int i;
	147	for (s = p += 2; (p_ch = p) && p_ch != ']'; p++) {} /SHARED ITERATOR*/
	148	if (!p_ch)
	149	return ABORT_ALL;
	150	i = p - s - 1;
	151	if (i < 0 \|\| p[-1] != ':') {
	152	/* Didn't find ":]", so treat like a normal set. */
	153	p = s - 2;
	154	p_ch = '[';
	155	if (t_ch == p_ch)
	156	matched = TRUE;
	157	continue;
	158	}
	159	if (CC_EQ(s,i, "alnum")) {
	160	if (ISALNUM(t_ch))
	161	matched = TRUE;
	162	} else if (CC_EQ(s,i, "alpha")) {
	163	if (ISALPHA(t_ch))
	164	matched = TRUE;
	165	} else if (CC_EQ(s,i, "blank")) {
	166	if (ISBLANK(t_ch))
	167	matched = TRUE;
	168	} else if (CC_EQ(s,i, "cntrl")) {
	169	if (ISCNTRL(t_ch))
	170	matched = TRUE;
	171	} else if (CC_EQ(s,i, "digit")) {
	172	if (ISDIGIT(t_ch))
	173	matched = TRUE;
	174	} else if (CC_EQ(s,i, "graph")) {
175	if (ISGRAPH(t_ch))
176	matched = TRUE;
177	} else if (CC_EQ(s,i, "lower")) {
178	if (ISLOWER(t_ch))
179	matched = TRUE;
180	} else if (CC_EQ(s,i, "print")) {
181	if (ISPRINT(t_ch))
182	matched = TRUE;
183	} else if (CC_EQ(s,i, "punct")) {
184	if (ISPUNCT(t_ch))
185	matched = TRUE;
186	} else if (CC_EQ(s,i, "space")) {
187	if (ISSPACE(t_ch))
188	matched = TRUE;
189	} else if (CC_EQ(s,i, "upper")) {
190	if (ISUPPER(t_ch))
191	matched = TRUE;
192	} else if (CC_EQ(s,i, "xdigit")) {
193	if (ISXDIGIT(t_ch))
194	matched = TRUE;
195	} else /* malformed [:class:] string */
196	return ABORT_ALL;
197	p_ch = 0; /* This makes "prev_ch" get set to 0. */
198	} else if (t_ch == p_ch)
199	matched = TRUE;
200	} while (prev_ch = p_ch, (p_ch = *++p) != ']');
201	if (matched == special \|\| t_ch == '/')
202	return FALSE;
203	continue;
204	}
5230f605	205	}
5230f605	206
327f2f3e	207	return *text ? FALSE : TRUE;
5230f605 NTND	208	}
	209
	210	/* Match the "pattern" against the "text" string. */
	211	int wildmatch(const char pattern, const char text)
	212	{
327f2f3e	213	return dowild((const uchar)pattern, (const uchar)text) == TRUE;
5230f605 NTND	214	}
	215
	216	/* Match the "pattern" against the forced-to-lower-case "text" string. */
	217	int iwildmatch(const char pattern, const char text)
	218	{
327f2f3e NTND	219	int ret;
	220	force_lower_case = 1;
	221	ret = dowild((const uchar)pattern, (const uchar)text) == TRUE;
	222	force_lower_case = 0;
	223	return ret;
5230f605	224	}