2 * Handle git attributes. See gitattributes(5) for a description of
3 * the file syntax, and Documentation/technical/api-gitattributes.txt
4 * for a description of the API.
6 * One basic design decision here is that we are not going to support
7 * an insanely large number of attributes.
10 #define NO_THE_INDEX_COMPATIBILITY_MACROS
15 const char git_attr__true
[] = "(builtin)true";
16 const char git_attr__false
[] = "\0(builtin)false";
17 static const char git_attr__unknown
[] = "(builtin)unknown";
18 #define ATTR__TRUE git_attr__true
19 #define ATTR__FALSE git_attr__false
20 #define ATTR__UNSET NULL
21 #define ATTR__UNKNOWN git_attr__unknown
23 static const char *attributes_file
;
25 /* This is a randomly chosen prime. */
33 struct git_attr
*next
;
36 char name
[FLEX_ARRAY
];
40 static struct git_attr_check
*check_all_attr
;
41 static struct git_attr
*(git_attr_hash
[HASHSIZE
]);
43 static unsigned hash_name(const char *name
, int namelen
)
49 val
= ((val
<< 7) | (val
>> 22)) ^ c
;
54 static int invalid_attr_name(const char *name
, int namelen
)
57 * Attribute name cannot begin with '-' and from
58 * [-A-Za-z0-9_.]. We'd specifically exclude '=' for now,
59 * as we might later want to allow non-binary value for
60 * attributes, e.g. "*.svg merge=special-merge-program-for-svg"
66 if (! (ch
== '-' || ch
== '.' || ch
== '_' ||
67 ('0' <= ch
&& ch
<= '9') ||
68 ('a' <= ch
&& ch
<= 'z') ||
69 ('A' <= ch
&& ch
<= 'Z')) )
75 static struct git_attr
*git_attr_internal(const char *name
, int len
)
77 unsigned hval
= hash_name(name
, len
);
78 unsigned pos
= hval
% HASHSIZE
;
81 for (a
= git_attr_hash
[pos
]; a
; a
= a
->next
) {
83 !memcmp(a
->name
, name
, len
) && !a
->name
[len
])
87 if (invalid_attr_name(name
, len
))
90 a
= xmalloc(sizeof(*a
) + len
+ 1);
91 memcpy(a
->name
, name
, len
);
94 a
->next
= git_attr_hash
[pos
];
95 a
->attr_nr
= attr_nr
++;
96 git_attr_hash
[pos
] = a
;
98 check_all_attr
= xrealloc(check_all_attr
,
99 sizeof(*check_all_attr
) * attr_nr
);
100 check_all_attr
[a
->attr_nr
].attr
= a
;
101 check_all_attr
[a
->attr_nr
].value
= ATTR__UNKNOWN
;
105 struct git_attr
*git_attr(const char *name
)
107 return git_attr_internal(name
, strlen(name
));
110 /* What does a matched pattern decide? */
112 struct git_attr
*attr
;
117 * One rule, as from a .gitattributes file.
119 * If is_macro is true, then u.attr is a pointer to the git_attr being
122 * If is_macro is false, then u.pattern points at the filename pattern
123 * to which the rule applies. (The memory pointed to is part of the
124 * memory block allocated for the match_attr instance.)
126 * In either case, num_attr is the number of attributes affected by
127 * this rule, and state is an array listing them. The attributes are
128 * listed as they appear in the file (macros unexpanded).
133 struct git_attr
*attr
;
137 struct attr_state state
[FLEX_ARRAY
];
140 static const char blank
[] = " \t\r\n";
143 * Parse a whitespace-delimited attribute state (i.e., "attr",
144 * "-attr", "!attr", or "attr=value") from the string starting at src.
145 * If e is not NULL, write the results to *e. Return a pointer to the
146 * remainder of the string (with leading whitespace removed), or NULL
147 * if there was an error.
149 static const char *parse_attr(const char *src
, int lineno
, const char *cp
,
150 struct attr_state
*e
)
152 const char *ep
, *equals
;
155 ep
= cp
+ strcspn(cp
, blank
);
156 equals
= strchr(cp
, '=');
157 if (equals
&& ep
< equals
)
164 if (*cp
== '-' || *cp
== '!') {
168 if (invalid_attr_name(cp
, len
)) {
170 "%.*s is not a valid attribute name: %s:%d\n",
171 len
, cp
, src
, lineno
);
175 if (*cp
== '-' || *cp
== '!') {
176 e
->setto
= (*cp
== '-') ? ATTR__FALSE
: ATTR__UNSET
;
181 e
->setto
= ATTR__TRUE
;
183 e
->setto
= xmemdupz(equals
+ 1, ep
- equals
- 1);
185 e
->attr
= git_attr_internal(cp
, len
);
187 return ep
+ strspn(ep
, blank
);
190 static struct match_attr
*parse_attr_line(const char *line
, const char *src
,
191 int lineno
, int macro_ok
)
195 const char *cp
, *name
, *states
;
196 struct match_attr
*res
= NULL
;
200 cp
= line
+ strspn(line
, blank
);
201 if (!*cp
|| *cp
== '#')
204 namelen
= strcspn(name
, blank
);
205 if (strlen(ATTRIBUTE_MACRO_PREFIX
) < namelen
&&
206 !prefixcmp(name
, ATTRIBUTE_MACRO_PREFIX
)) {
208 fprintf(stderr
, "%s not allowed: %s:%d\n",
213 name
+= strlen(ATTRIBUTE_MACRO_PREFIX
);
214 name
+= strspn(name
, blank
);
215 namelen
= strcspn(name
, blank
);
216 if (invalid_attr_name(name
, namelen
)) {
218 "%.*s is not a valid attribute name: %s:%d\n",
219 namelen
, name
, src
, lineno
);
226 states
= name
+ namelen
;
227 states
+= strspn(states
, blank
);
229 for (pass
= 0; pass
< 2; pass
++) {
230 /* pass 0 counts and allocates, pass 1 fills */
234 cp
= parse_attr(src
, lineno
, cp
,
235 pass
? &(res
->state
[num_attr
]) : NULL
);
244 sizeof(struct attr_state
) * num_attr
+
245 (is_macro
? 0 : namelen
+ 1));
247 res
->u
.attr
= git_attr_internal(name
, namelen
);
249 res
->u
.pattern
= (char *)&(res
->state
[num_attr
]);
250 memcpy(res
->u
.pattern
, name
, namelen
);
251 res
->u
.pattern
[namelen
] = 0;
253 res
->is_macro
= is_macro
;
254 res
->num_attr
= num_attr
;
260 * Like info/exclude and .gitignore, the attribute information can
261 * come from many places.
263 * (1) .gitattribute file of the same directory;
264 * (2) .gitattribute file of the parent directory if (1) does not have
265 * any match; this goes recursively upwards, just like .gitignore.
266 * (3) $GIT_DIR/info/attributes, which overrides both of the above.
268 * In the same file, later entries override the earlier match, so in the
269 * global list, we would have entries from info/attributes the earliest
270 * (reading the file from top to bottom), .gitattribute of the root
271 * directory (again, reading the file from top to bottom) down to the
272 * current directory, and then scan the list backwards to find the first match.
273 * This is exactly the same as what excluded() does in dir.c to deal with
277 static struct attr_stack
{
278 struct attr_stack
*prev
;
280 unsigned num_matches
;
282 struct match_attr
**attrs
;
285 static void free_attr_elem(struct attr_stack
*e
)
289 for (i
= 0; i
< e
->num_matches
; i
++) {
290 struct match_attr
*a
= e
->attrs
[i
];
292 for (j
= 0; j
< a
->num_attr
; j
++) {
293 const char *setto
= a
->state
[j
].setto
;
294 if (setto
== ATTR__TRUE
||
295 setto
== ATTR__FALSE
||
296 setto
== ATTR__UNSET
||
297 setto
== ATTR__UNKNOWN
)
300 free((char *) setto
);
307 static const char *builtin_attr
[] = {
308 "[attr]binary -diff -text",
312 static void handle_attr_line(struct attr_stack
*res
,
318 struct match_attr
*a
;
320 a
= parse_attr_line(line
, src
, lineno
, macro_ok
);
323 if (res
->alloc
<= res
->num_matches
) {
324 res
->alloc
= alloc_nr(res
->num_matches
);
325 res
->attrs
= xrealloc(res
->attrs
,
326 sizeof(struct match_attr
*) *
329 res
->attrs
[res
->num_matches
++] = a
;
332 static struct attr_stack
*read_attr_from_array(const char **list
)
334 struct attr_stack
*res
;
338 res
= xcalloc(1, sizeof(*res
));
339 while ((line
= *(list
++)) != NULL
)
340 handle_attr_line(res
, line
, "[builtin]", ++lineno
, 1);
344 static enum git_attr_direction direction
;
345 static struct index_state
*use_index
;
347 static struct attr_stack
*read_attr_from_file(const char *path
, int macro_ok
)
349 FILE *fp
= fopen(path
, "r");
350 struct attr_stack
*res
;
356 res
= xcalloc(1, sizeof(*res
));
357 while (fgets(buf
, sizeof(buf
), fp
))
358 handle_attr_line(res
, buf
, path
, ++lineno
, macro_ok
);
363 static void *read_index_data(const char *path
)
367 enum object_type type
;
369 struct index_state
*istate
= use_index
? use_index
: &the_index
;
372 pos
= index_name_pos(istate
, path
, len
);
375 * We might be in the middle of a merge, in which
376 * case we would read stage #2 (ours).
380 (pos
< 0 && i
< istate
->cache_nr
&&
381 !strcmp(istate
->cache
[i
]->name
, path
));
383 if (ce_stage(istate
->cache
[i
]) == 2)
388 data
= read_sha1_file(istate
->cache
[pos
]->sha1
, &type
, &sz
);
389 if (!data
|| type
!= OBJ_BLOB
) {
396 static struct attr_stack
*read_attr_from_index(const char *path
, int macro_ok
)
398 struct attr_stack
*res
;
402 buf
= read_index_data(path
);
406 res
= xcalloc(1, sizeof(*res
));
407 for (sp
= buf
; *sp
; ) {
410 for (ep
= sp
; *ep
&& *ep
!= '\n'; ep
++)
412 more
= (*ep
== '\n');
414 handle_attr_line(res
, sp
, path
, ++lineno
, macro_ok
);
421 static struct attr_stack
*read_attr(const char *path
, int macro_ok
)
423 struct attr_stack
*res
;
425 if (direction
== GIT_ATTR_CHECKOUT
) {
426 res
= read_attr_from_index(path
, macro_ok
);
428 res
= read_attr_from_file(path
, macro_ok
);
430 else if (direction
== GIT_ATTR_CHECKIN
) {
431 res
= read_attr_from_file(path
, macro_ok
);
434 * There is no checked out .gitattributes file there, but
435 * we might have it in the index. We allow operation in a
436 * sparsely checked out work tree, so read from it.
438 res
= read_attr_from_index(path
, macro_ok
);
441 res
= read_attr_from_index(path
, macro_ok
);
443 res
= xcalloc(1, sizeof(*res
));
448 static void debug_info(const char *what
, struct attr_stack
*elem
)
450 fprintf(stderr
, "%s: %s\n", what
, elem
->origin
? elem
->origin
: "()");
452 static void debug_set(const char *what
, const char *match
, struct git_attr
*attr
, const void *v
)
454 const char *value
= v
;
456 if (ATTR_TRUE(value
))
458 else if (ATTR_FALSE(value
))
460 else if (ATTR_UNSET(value
))
461 value
= "unspecified";
463 fprintf(stderr
, "%s: %s => %s (%s)\n",
464 what
, attr
->name
, (char *) value
, match
);
466 #define debug_push(a) debug_info("push", (a))
467 #define debug_pop(a) debug_info("pop", (a))
469 #define debug_push(a) do { ; } while (0)
470 #define debug_pop(a) do { ; } while (0)
471 #define debug_set(a,b,c,d) do { ; } while (0)
474 static void drop_attr_stack(void)
477 struct attr_stack
*elem
= attr_stack
;
478 attr_stack
= elem
->prev
;
479 free_attr_elem(elem
);
483 static const char *git_etc_gitattributes(void)
485 static const char *system_wide
;
487 system_wide
= system_path(ETC_GITATTRIBUTES
);
491 static int git_attr_system(void)
493 return !git_env_bool("GIT_ATTR_NOSYSTEM", 0);
496 static int git_attr_config(const char *var
, const char *value
, void *dummy
)
498 if (!strcmp(var
, "core.attributesfile"))
499 return git_config_pathname(&attributes_file
, var
, value
);
504 static void bootstrap_attr_stack(void)
507 struct attr_stack
*elem
;
509 elem
= read_attr_from_array(builtin_attr
);
511 elem
->prev
= attr_stack
;
514 if (git_attr_system()) {
515 elem
= read_attr_from_file(git_etc_gitattributes(), 1);
518 elem
->prev
= attr_stack
;
523 git_config(git_attr_config
, NULL
);
524 if (attributes_file
) {
525 elem
= read_attr_from_file(attributes_file
, 1);
528 elem
->prev
= attr_stack
;
533 if (!is_bare_repository() || direction
== GIT_ATTR_INDEX
) {
534 elem
= read_attr(GITATTRIBUTES_FILE
, 1);
535 elem
->origin
= strdup("");
536 elem
->prev
= attr_stack
;
541 elem
= read_attr_from_file(git_path(INFOATTRIBUTES_FILE
), 1);
543 elem
= xcalloc(1, sizeof(*elem
));
545 elem
->prev
= attr_stack
;
550 static void prepare_attr_stack(const char *path
, int dirlen
)
552 struct attr_stack
*elem
, *info
;
554 struct strbuf pathbuf
;
556 strbuf_init(&pathbuf
, dirlen
+2+strlen(GITATTRIBUTES_FILE
));
559 * At the bottom of the attribute stack is the built-in
560 * set of attribute definitions, followed by the contents
561 * of $(prefix)/etc/gitattributes and a file specified by
562 * core.attributesfile. Then, contents from
563 * .gitattribute files from directories closer to the
564 * root to the ones in deeper directories are pushed
565 * to the stack. Finally, at the very top of the stack
566 * we always keep the contents of $GIT_DIR/info/attributes.
568 * When checking, we use entries from near the top of the
569 * stack, preferring $GIT_DIR/info/attributes, then
570 * .gitattributes in deeper directories to shallower ones,
571 * and finally use the built-in set as the default.
574 bootstrap_attr_stack();
577 * Pop the "info" one that is always at the top of the stack.
580 attr_stack
= info
->prev
;
583 * Pop the ones from directories that are not the prefix of
584 * the path we are checking.
586 while (attr_stack
&& attr_stack
->origin
) {
587 int namelen
= strlen(attr_stack
->origin
);
590 if (namelen
<= dirlen
&&
591 !strncmp(elem
->origin
, path
, namelen
))
595 attr_stack
= elem
->prev
;
596 free_attr_elem(elem
);
600 * Read from parent directories and push them down
602 if (!is_bare_repository() || direction
== GIT_ATTR_INDEX
) {
606 len
= strlen(attr_stack
->origin
);
609 strbuf_reset(&pathbuf
);
610 strbuf_add(&pathbuf
, path
, dirlen
);
611 strbuf_addch(&pathbuf
, '/');
612 cp
= strchr(pathbuf
.buf
+ len
+ 1, '/');
613 strcpy(cp
+ 1, GITATTRIBUTES_FILE
);
614 elem
= read_attr(pathbuf
.buf
, 0);
616 elem
->origin
= strdup(pathbuf
.buf
);
617 elem
->prev
= attr_stack
;
623 strbuf_release(&pathbuf
);
626 * Finally push the "info" one at the top of the stack.
628 info
->prev
= attr_stack
;
632 static int path_matches(const char *pathname
, int pathlen
,
634 const char *base
, int baselen
)
636 if (!strchr(pattern
, '/')) {
638 const char *basename
= strrchr(pathname
, '/');
639 basename
= basename
? basename
+ 1 : pathname
;
640 return (fnmatch(pattern
, basename
, 0) == 0);
643 * match with FNM_PATHNAME; the pattern has base implicitly
648 if (pathlen
< baselen
||
649 (baselen
&& pathname
[baselen
] != '/') ||
650 strncmp(pathname
, base
, baselen
))
654 return fnmatch(pattern
, pathname
+ baselen
, FNM_PATHNAME
) == 0;
657 static int macroexpand_one(int attr_nr
, int rem
);
659 static int fill_one(const char *what
, struct match_attr
*a
, int rem
)
661 struct git_attr_check
*check
= check_all_attr
;
664 for (i
= a
->num_attr
- 1; 0 < rem
&& 0 <= i
; i
--) {
665 struct git_attr
*attr
= a
->state
[i
].attr
;
666 const char **n
= &(check
[attr
->attr_nr
].value
);
667 const char *v
= a
->state
[i
].setto
;
669 if (*n
== ATTR__UNKNOWN
) {
671 a
->is_macro
? a
->u
.attr
->name
: a
->u
.pattern
,
675 rem
= macroexpand_one(attr
->attr_nr
, rem
);
681 static int fill(const char *path
, int pathlen
, struct attr_stack
*stk
, int rem
)
684 const char *base
= stk
->origin
? stk
->origin
: "";
686 for (i
= stk
->num_matches
- 1; 0 < rem
&& 0 <= i
; i
--) {
687 struct match_attr
*a
= stk
->attrs
[i
];
690 if (path_matches(path
, pathlen
,
691 a
->u
.pattern
, base
, strlen(base
)))
692 rem
= fill_one("fill", a
, rem
);
697 static int macroexpand_one(int attr_nr
, int rem
)
699 struct attr_stack
*stk
;
700 struct match_attr
*a
= NULL
;
703 if (check_all_attr
[attr_nr
].value
!= ATTR__TRUE
)
706 for (stk
= attr_stack
; !a
&& stk
; stk
= stk
->prev
)
707 for (i
= stk
->num_matches
- 1; !a
&& 0 <= i
; i
--) {
708 struct match_attr
*ma
= stk
->attrs
[i
];
711 if (ma
->u
.attr
->attr_nr
== attr_nr
)
716 rem
= fill_one("expand", a
, rem
);
721 int git_checkattr(const char *path
, int num
, struct git_attr_check
*check
)
723 struct attr_stack
*stk
;
725 int dirlen
, pathlen
, i
, rem
;
727 bootstrap_attr_stack();
728 for (i
= 0; i
< attr_nr
; i
++)
729 check_all_attr
[i
].value
= ATTR__UNKNOWN
;
731 pathlen
= strlen(path
);
732 cp
= strrchr(path
, '/');
737 prepare_attr_stack(path
, dirlen
);
739 for (stk
= attr_stack
; 0 < rem
&& stk
; stk
= stk
->prev
)
740 rem
= fill(path
, pathlen
, stk
, rem
);
742 for (i
= 0; i
< num
; i
++) {
743 const char *value
= check_all_attr
[check
[i
].attr
->attr_nr
].value
;
744 if (value
== ATTR__UNKNOWN
)
746 check
[i
].value
= value
;
752 void git_attr_set_direction(enum git_attr_direction
new, struct index_state
*istate
)
754 enum git_attr_direction old
= direction
;
756 if (is_bare_repository() && new != GIT_ATTR_INDEX
)
757 die("BUG: non-INDEX attr direction in a bare repo");