[thirdparty/git.git] / attr.c

#include "cache.h"
#include "attr.h"

/*
 * The basic design decision here is that we are not going to have
 * insanely large number of attributes.
 *
 * This is a randomly chosen prime.
 */
#define HASHSIZE 257

#ifndef DEBUG_ATTR
#define DEBUG_ATTR 0
#endif

struct git_attr {
	struct git_attr *next;
	unsigned h;
	char name[FLEX_ARRAY];
};

static struct git_attr *(git_attr_hash[HASHSIZE]);

static unsigned hash_name(const char *name, int namelen)
{
	unsigned val = 0;
	unsigned char c;

	while (namelen--) {
		c = *name++;
		val = ((val << 7) | (val >> 22)) ^ c;
	}
	return val;
}

struct git_attr *git_attr(const char *name, int len)
{
	unsigned hval = hash_name(name, len);
	unsigned pos = hval % HASHSIZE;
	struct git_attr *a;

	for (a = git_attr_hash[pos]; a; a = a->next) {
		if (a->h == hval &&
		    !memcmp(a->name, name, len) && !a->name[len])
			return a;
	}

	a = xmalloc(sizeof(*a) + len + 1);
	memcpy(a->name, name, len);
	a->name[len] = 0;
	a->h = hval;
	a->next = git_attr_hash[pos];
	git_attr_hash[pos] = a;
	return a;
}

/*
 * .gitattributes file is one line per record, each of which is
 *
 * (1) glob pattern.
 * (2) whitespace
 * (3) whitespace separated list of attribute names, each of which
 *     could be prefixed with '!' to mean "not set".
 */

struct attr_state {
	int unset;
	struct git_attr *attr;
};

struct match_attr {
	char *pattern;
	unsigned num_attr;
	struct attr_state state[FLEX_ARRAY];
};

static const char blank[] = " \t\r\n";

static struct match_attr *parse_attr_line(const char *line)
{
	int namelen;
	int num_attr;
	const char *cp, *name;
	struct match_attr *res = res;
	int pass;

	cp = line + strspn(line, blank);
	if (!*cp || *cp == '#')
		return NULL;
	name = cp;
	namelen = strcspn(name, blank);

	for (pass = 0; pass < 2; pass++) {
		/* pass 0 counts and allocates, pass 1 fills */
		num_attr = 0;
		cp = name + namelen;
		cp = cp + strspn(cp, blank);
		while (*cp) {
			const char *ep;
			ep = cp + strcspn(cp, blank);
			if (pass) {
				struct attr_state *e;

				e = &(res->state[num_attr]);
				if (*cp == '!') {
					e->unset = 1;
					cp++;
				}
				e->attr = git_attr(cp, ep - cp);
			}
			num_attr++;
			cp = ep + strspn(ep, blank);
		}
		if (pass)
			break;
		res = xcalloc(1,
			      sizeof(*res) +
			      sizeof(struct attr_state) * num_attr +
			      namelen + 1);
		res->pattern = (char*)&(res->state[num_attr]);
		memcpy(res->pattern, name, namelen);
		res->pattern[namelen] = 0;
		res->num_attr = num_attr;
	}
	return res;
}

/*
 * Like info/exclude and .gitignore, the attribute information can
 * come from many places.
 *
 * (1) .gitattribute file of the same directory;
 * (2) .gitattribute file of the parent directory if (1) does not have any match;
 *     this goes recursively upwards, just like .gitignore
 * (3) perhaps $GIT_DIR/info/attributes, as the final fallback.
 *
 * In the same file, later entries override the earlier match, so in the
 * global list, we would have entries from info/attributes the earliest
 * (reading the file from top to bottom), .gitattribute of the root
 * directory (again, reading the file from top to bottom) down to the
 * current directory, and then scan the list backwards to find the first match.
 * This is exactly the same as what excluded() does in dir.c to deal with
 * .gitignore
 */

static struct attr_stack {
	struct attr_stack *prev;
	char *origin;
	unsigned num_matches;
	struct match_attr **attrs;
} *attr_stack;

static void free_attr_elem(struct attr_stack *e)
{
	int i;
	free(e->origin);
	for (i = 0; i < e->num_matches; i++)
		free(e->attrs[i]);
	free(e);
}

static const char *builtin_attr[] = {
	NULL,
};

static struct attr_stack *read_attr_from_array(const char **list)
{
	struct attr_stack *res;
	const char *line;

	res = xcalloc(1, sizeof(*res));
	while ((line = *(list++)) != NULL) {
		struct match_attr *a = parse_attr_line(line);
		if (!a)
			continue;
		res->attrs = xrealloc(res->attrs, res->num_matches + 1);
		res->attrs[res->num_matches++] = a;
	}
	return res;
}

static struct attr_stack *read_attr_from_file(const char *path)
{
	FILE *fp;
	struct attr_stack *res;
	char buf[2048];

	res = xcalloc(1, sizeof(*res));
	fp = fopen(path, "r");
	if (!fp)
		return res;

	while (fgets(buf, sizeof(buf), fp)) {
		struct match_attr *a = parse_attr_line(buf);
		if (!a)
			continue;
		res->attrs = xrealloc(res->attrs, res->num_matches + 1);
		res->attrs[res->num_matches++] = a;
	}
	fclose(fp);
	return res;
}

#if DEBUG_ATTR
static void debug_info(const char *what, struct attr_stack *elem)
{
	fprintf(stderr, "%s: %s\n", what, elem->origin ? elem->origin : "()");
}
#define debug_push(a) debug_info("push", (a))
#define debug_pop(a) debug_info("pop", (a))
#else
#define debug_push(a) do { ; } while (0)
#define debug_pop(a) do { ; } while (0)
#endif

static void prepare_attr_stack(const char *path, int dirlen)
{
	struct attr_stack *elem, *info;
	int len;
	char pathbuf[PATH_MAX];

	/*
	 * At the bottom of the attribute stack is the built-in
	 * set of attribute definitions.  Then, contents from
	 * .gitattribute files from directories closer to the
	 * root to the ones in deeper directories are pushed
	 * to the stack.  Finally, at the very top of the stack
	 * we always keep the contents of $GIT_DIR/info/attributes.
	 *
	 * When checking, we use entries from near the top of the
	 * stack, preferring $GIT_DIR/info/attributes, then
	 * .gitattributes in deeper directories to shallower ones,
	 * and finally use the built-in set as the default.
	 */
	if (!attr_stack) {
		elem = read_attr_from_array(builtin_attr);
		elem->origin = NULL;
		elem->prev = attr_stack;
		attr_stack = elem;

		elem = read_attr_from_file(GITATTRIBUTES_FILE);
		elem->origin = strdup("");
		elem->prev = attr_stack;
		attr_stack = elem;
		debug_push(elem);

		elem = read_attr_from_file(git_path(INFOATTRIBUTES_FILE));
		elem->origin = NULL;
		elem->prev = attr_stack;
		attr_stack = elem;
	}

	/*
	 * Pop the "info" one that is always at the top of the stack.
	 */
	info = attr_stack;
	attr_stack = info->prev;

	/*
	 * Pop the ones from directories that are not the prefix of
	 * the path we are checking.
	 */
	while (attr_stack && attr_stack->origin) {
		int namelen = strlen(attr_stack->origin);

		elem = attr_stack;
		if (namelen <= dirlen &&
		    !strncmp(elem->origin, path, namelen))
			break;

		debug_pop(elem);
		attr_stack = elem->prev;
		free_attr_elem(elem);
	}

	/*
	 * Read from parent directories and push them down
	 */
	while (1) {
		char *cp;

		len = strlen(attr_stack->origin);
		if (dirlen <= len)
			break;
		memcpy(pathbuf, path, dirlen);
		memcpy(pathbuf + dirlen, "/", 2);
		cp = strchr(pathbuf + len + 1, '/');
		strcpy(cp + 1, GITATTRIBUTES_FILE);
		elem = read_attr_from_file(pathbuf);
		*cp = '\0';
		elem->origin = strdup(pathbuf);
		elem->prev = attr_stack;
		attr_stack = elem;
		debug_push(elem);
	}

	/*
	 * Finally push the "info" one at the top of the stack.
	 */
	info->prev = attr_stack;
	attr_stack = info;
}

static int path_matches(const char *pathname, int pathlen,
			const char *pattern,
			const char *base, int baselen)
{
	if (!strchr(pattern, '/')) {
		/* match basename */
		const char *basename = strrchr(pathname, '/');
		basename = basename ? basename + 1 : pathname;
		return (fnmatch(pattern, basename, 0) == 0);
	}
	/*
	 * match with FNM_PATHNAME; the pattern has base implicitly
	 * in front of it.
	 */
	if (*pattern == '/')
		pattern++;
	if (pathlen < baselen ||
	    (baselen && pathname[baselen - 1] != '/') ||
	    strncmp(pathname, base, baselen))
		return 0;
	return fnmatch(pattern, pathname + baselen, FNM_PATHNAME) == 0;
}

/*
 * I do not like this at all.  Only because we allow individual
 * attribute to be set or unset incrementally by individual
 * lines in .gitattribute files, we need to do this triple
 * loop which looks quite wasteful.
 */
static int fill(const char *path, int pathlen,
		struct attr_stack *stk, struct git_attr_check *check,
		int num, int rem)
{
	int i, j, k;
	const char *base = stk->origin ? stk->origin : "";

	for (i = stk->num_matches - 1; 0 < rem && 0 <= i; i--) {
		struct match_attr *a = stk->attrs[i];
		if (path_matches(path, pathlen,
				 a->pattern, base, strlen(base))) {
			for (j = 0; j < a->num_attr; j++) {
				struct git_attr *attr = a->state[j].attr;
				int set = !a->state[j].unset;
				for (k = 0; k < num; k++) {
					if (0 <= check[k].isset ||
					    check[k].attr != attr)
						continue;
					check[k].isset = set;
					rem--;
				}
			}
		}
	}
	return rem;
}

int git_checkattr(const char *path, int num, struct git_attr_check *check)
{
	struct attr_stack *stk;
	const char *cp;
	int dirlen, pathlen, i, rem;

	for (i = 0; i < num; i++)
		check[i].isset = -1;

	pathlen = strlen(path);
	cp = strrchr(path, '/');
	if (!cp)
		dirlen = 0;
	else
		dirlen = cp - path;
	prepare_attr_stack(path, dirlen);
	rem = num;
	for (stk = attr_stack; 0 < rem && stk; stk = stk->prev)
		rem = fill(path, pathlen, stk, check, num, rem);
	return 0;
}
Commit	Line	Data
d0bfd026 JH	1	#include "cache.h"
	2	#include "attr.h"
	3
	4	/*
	5	* The basic design decision here is that we are not going to have
	6	* insanely large number of attributes.
	7	*
	8	* This is a randomly chosen prime.
	9	*/
	10	#define HASHSIZE 257
	11
	12	#ifndef DEBUG_ATTR
	13	#define DEBUG_ATTR 0
	14	#endif
	15
	16	struct git_attr {
	17	struct git_attr *next;
	18	unsigned h;
	19	char name[FLEX_ARRAY];
	20	};
	21
	22	static struct git_attr *(git_attr_hash[HASHSIZE]);
	23
	24	static unsigned hash_name(const char *name, int namelen)
	25	{
	26	unsigned val = 0;
	27	unsigned char c;
	28
	29	while (namelen--) {
	30	c = *name++;
	31	val = ((val << 7) \| (val >> 22)) ^ c;
	32	}
	33	return val;
	34	}
	35
	36	struct git_attr git_attr(const char name, int len)
	37	{
	38	unsigned hval = hash_name(name, len);
	39	unsigned pos = hval % HASHSIZE;
	40	struct git_attr *a;
	41
	42	for (a = git_attr_hash[pos]; a; a = a->next) {
	43	if (a->h == hval &&
	44	!memcmp(a->name, name, len) && !a->name[len])
	45	return a;
	46	}
	47
	48	a = xmalloc(sizeof(*a) + len + 1);
	49	memcpy(a->name, name, len);
	50	a->name[len] = 0;
	51	a->h = hval;
	52	a->next = git_attr_hash[pos];
	53	git_attr_hash[pos] = a;
	54	return a;
	55	}
	56
	57	/*
	58	* .gitattributes file is one line per record, each of which is
	59	*
	60	* (1) glob pattern.
	61	* (2) whitespace
	62	* (3) whitespace separated list of attribute names, each of which
	63	* could be prefixed with '!' to mean "not set".
	64	*/
65
66	struct attr_state {
67	int unset;
68	struct git_attr *attr;
69	};
70
71	struct match_attr {
72	char *pattern;
73	unsigned num_attr;
74	struct attr_state state[FLEX_ARRAY];
75	};
76
77	static const char blank[] = " \t\r\n";
78
79	static struct match_attr parse_attr_line(const char line)
80	{
81	int namelen;
82	int num_attr;
83	const char cp, name;
84	struct match_attr *res = res;
85	int pass;
86
87	cp = line + strspn(line, blank);
88	if (!cp \|\| cp == '#')
89	return NULL;
90	name = cp;
91	namelen = strcspn(name, blank);
92
93	for (pass = 0; pass < 2; pass++) {
94	/* pass 0 counts and allocates, pass 1 fills */
95	num_attr = 0;
96	cp = name + namelen;
97	cp = cp + strspn(cp, blank);
98	while (*cp) {
99	const char *ep;
100	ep = cp + strcspn(cp, blank);
101	if (pass) {
102	struct attr_state *e;
103
104	e = &(res->state[num_attr]);
105	if (*cp == '!') {
106	e->unset = 1;
107	cp++;
108	}
109	e->attr = git_attr(cp, ep - cp);
110	}
111	num_attr++;
112	cp = ep + strspn(ep, blank);
113	}
114	if (pass)
115	break;
116	res = xcalloc(1,
117	sizeof(*res) +
118	sizeof(struct attr_state) * num_attr +
119	namelen + 1);
120	res->pattern = (char*)&(res->state[num_attr]);
121	memcpy(res->pattern, name, namelen);
122	res->pattern[namelen] = 0;
123	res->num_attr = num_attr;
124	}
125	return res;
126	}
127
128	/*
129	* Like info/exclude and .gitignore, the attribute information can
130	* come from many places.
131	*
132	* (1) .gitattribute file of the same directory;
133	* (2) .gitattribute file of the parent directory if (1) does not have any match;
134	* this goes recursively upwards, just like .gitignore
135	* (3) perhaps $GIT_DIR/info/attributes, as the final fallback.
136	*
137	* In the same file, later entries override the earlier match, so in the
138	* global list, we would have entries from info/attributes the earliest
139	* (reading the file from top to bottom), .gitattribute of the root
140	* directory (again, reading the file from top to bottom) down to the
141	* current directory, and then scan the list backwards to find the first match.
142	* This is exactly the same as what excluded() does in dir.c to deal with
143	* .gitignore
144	*/
145
146	static struct attr_stack {
147	struct attr_stack *prev;
148	char *origin;
149	unsigned num_matches;
150	struct match_attr **attrs;
151	} *attr_stack;
152
153	static void free_attr_elem(struct attr_stack *e)
154	{
155	int i;
156	free(e->origin);
157	for (i = 0; i < e->num_matches; i++)
158	free(e->attrs[i]);
159	free(e);
160	}
161
162	static const char *builtin_attr[] = {
163	NULL,
164	};
165
166	static struct attr_stack read_attr_from_array(const char *list)
167	{
168	struct attr_stack *res;
169	const char *line;
170
171	res = xcalloc(1, sizeof(*res));
172	while ((line = *(list++)) != NULL) {
173	struct match_attr *a = parse_attr_line(line);
174	if (!a)
175	continue;
176	res->attrs = xrealloc(res->attrs, res->num_matches + 1);
177	res->attrs[res->num_matches++] = a;
178	}
179	return res;
180	}
181
182	static struct attr_stack read_attr_from_file(const char path)
183	{
184	FILE *fp;
185	struct attr_stack *res;
186	char buf[2048];
187
188	res = xcalloc(1, sizeof(*res));
189	fp = fopen(path, "r");
190	if (!fp)
191	return res;
192
193	while (fgets(buf, sizeof(buf), fp)) {
194	struct match_attr *a = parse_attr_line(buf);
195	if (!a)
196	continue;
197	res->attrs = xrealloc(res->attrs, res->num_matches + 1);
198	res->attrs[res->num_matches++] = a;
199	}
200	fclose(fp);
201	return res;
202	}
203
204	#if DEBUG_ATTR
205	static void debug_info(const char what, struct attr_stack elem)
206	{
207	fprintf(stderr, "%s: %s\n", what, elem->origin ? elem->origin : "()");
208	}
209	#define debug_push(a) debug_info("push", (a))
210	#define debug_pop(a) debug_info("pop", (a))
211	#else
212	#define debug_push(a) do { ; } while (0)
213	#define debug_pop(a) do { ; } while (0)
214	#endif
215
216	static void prepare_attr_stack(const char *path, int dirlen)
217	{
218	struct attr_stack elem, info;
219	int len;
220	char pathbuf[PATH_MAX];
221
222	/*
223	* At the bottom of the attribute stack is the built-in
224	* set of attribute definitions. Then, contents from
225	* .gitattribute files from directories closer to the
226	* root to the ones in deeper directories are pushed
227	* to the stack. Finally, at the very top of the stack
228	* we always keep the contents of $GIT_DIR/info/attributes.
229	*
230	* When checking, we use entries from near the top of the
231	* stack, preferring $GIT_DIR/info/attributes, then
232	* .gitattributes in deeper directories to shallower ones,
233	* and finally use the built-in set as the default.
234	*/
235	if (!attr_stack) {
236	elem = read_attr_from_array(builtin_attr);
237	elem->origin = NULL;
238	elem->prev = attr_stack;
239	attr_stack = elem;
240
241	elem = read_attr_from_file(GITATTRIBUTES_FILE);
242	elem->origin = strdup("");
243	elem->prev = attr_stack;
244	attr_stack = elem;
245	debug_push(elem);
246
247	elem = read_attr_from_file(git_path(INFOATTRIBUTES_FILE));
248	elem->origin = NULL;
249	elem->prev = attr_stack;
250	attr_stack = elem;
251	}
252
253	/*
254	* Pop the "info" one that is always at the top of the stack.
255	*/
256	info = attr_stack;
257	attr_stack = info->prev;
258
259	/*
260	* Pop the ones from directories that are not the prefix of
261	* the path we are checking.
262	*/
263	while (attr_stack && attr_stack->origin) {
264	int namelen = strlen(attr_stack->origin);
265
266	elem = attr_stack;
267	if (namelen <= dirlen &&
268	!strncmp(elem->origin, path, namelen))
269	break;
270
271	debug_pop(elem);
272	attr_stack = elem->prev;
273	free_attr_elem(elem);
274	}
275
276	/*
277	* Read from parent directories and push them down
278	*/
279	while (1) {
280	char *cp;
281
282	len = strlen(attr_stack->origin);
283	if (dirlen <= len)
284	break;
285	memcpy(pathbuf, path, dirlen);
286	memcpy(pathbuf + dirlen, "/", 2);
287	cp = strchr(pathbuf + len + 1, '/');
288	strcpy(cp + 1, GITATTRIBUTES_FILE);
289	elem = read_attr_from_file(pathbuf);
290	*cp = '\0';
291	elem->origin = strdup(pathbuf);
292	elem->prev = attr_stack;
293	attr_stack = elem;
294	debug_push(elem);
295	}
296
297	/*
298	* Finally push the "info" one at the top of the stack.
299	*/
300	info->prev = attr_stack;
301	attr_stack = info;
302	}
303
304	static int path_matches(const char *pathname, int pathlen,
305	const char *pattern,
306	const char *base, int baselen)
307	{
308	if (!strchr(pattern, '/')) {
309	/* match basename */
310	const char *basename = strrchr(pathname, '/');
311	basename = basename ? basename + 1 : pathname;
312	return (fnmatch(pattern, basename, 0) == 0);
313	}
314	/*
315	* match with FNM_PATHNAME; the pattern has base implicitly
316	* in front of it.
317	*/
318	if (*pattern == '/')
319	pattern++;
320	if (pathlen < baselen \|\|
321	(baselen && pathname[baselen - 1] != '/') \|\|
322	strncmp(pathname, base, baselen))
323	return 0;
324	return fnmatch(pattern, pathname + baselen, FNM_PATHNAME) == 0;
325	}
326
327	/*
328	* I do not like this at all. Only because we allow individual
329	* attribute to be set or unset incrementally by individual
330	* lines in .gitattribute files, we need to do this triple
331	* loop which looks quite wasteful.
332	*/
333	static int fill(const char *path, int pathlen,
334	struct attr_stack stk, struct git_attr_check check,
335	int num, int rem)
336	{
337	int i, j, k;
338	const char *base = stk->origin ? stk->origin : "";
339
340	for (i = stk->num_matches - 1; 0 < rem && 0 <= i; i--) {
341	struct match_attr *a = stk->attrs[i];
342	if (path_matches(path, pathlen,
343	a->pattern, base, strlen(base))) {
344	for (j = 0; j < a->num_attr; j++) {
345	struct git_attr *attr = a->state[j].attr;
346	int set = !a->state[j].unset;
347	for (k = 0; k < num; k++) {
348	if (0 <= check[k].isset \|\|
349	check[k].attr != attr)
350	continue;
351	check[k].isset = set;
352	rem--;
353	}
354	}
355	}
356	}
357	return rem;
358	}
359
360	int git_checkattr(const char path, int num, struct git_attr_check check)
361	{
362	struct attr_stack *stk;
363	const char *cp;
364	int dirlen, pathlen, i, rem;
365
366	for (i = 0; i < num; i++)
367	check[i].isset = -1;
368
369	pathlen = strlen(path);
370	cp = strrchr(path, '/');
371	if (!cp)
372	dirlen = 0;
373	else
374	dirlen = cp - path;
375	prepare_attr_stack(path, dirlen);
376	rem = num;
377	for (stk = attr_stack; 0 < rem && stk; stk = stk->prev)
378	rem = fill(path, pathlen, stk, check, num, rem);
379	return 0;
380	}