2 * This handles recursive filename detection with exclude
3 * files, index knowledge etc..
5 * Copyright (C) Linus Torvalds, 2005-2006
6 * Junio Hamano, 2005-2006
12 struct path_simplify {
17 static int read_directory_recursive(struct dir_struct *dir,
18 const char *path, const char *base, int baselen,
19 int check_only, const struct path_simplify *simplify);
20 static int get_dtype(struct dirent *de, const char *path);
22 static int common_prefix(const char **pathspec)
24 const char *path, *slash, *next;
31 slash = strrchr(path, '/');
35 prefix = slash - path + 1;
36 while ((next = *++pathspec) != NULL) {
37 int len = strlen(next);
38 if (len >= prefix && !memcmp(path, next, prefix))
44 if (next[--len] != '/')
46 if (memcmp(path, next, len+1))
55 int fill_directory(struct dir_struct *dir, const char **pathspec)
57 const char *path, *base;
61 * Calculate common prefix for the pathspec, and
62 * use that to optimize the directory walk
64 baselen = common_prefix(pathspec);
69 path = base = xmemdupz(*pathspec, baselen);
71 /* Read the directory and prune it */
72 read_directory(dir, path, base, baselen, pathspec);
77 * Does 'match' match the given name?
80 * (1) the 'match' string is leading directory of 'name', or
81 * (2) the 'match' string is a wildcard and matches 'name', or
82 * (3) the 'match' string is exactly the same as 'name'.
84 * and the return value tells which case it was.
86 * It returns 0 when there is no match.
88 static int match_one(const char *match, const char *name, int namelen)
92 /* If the match was just the prefix, we matched */
94 return MATCHED_RECURSIVELY;
97 unsigned char c1 = *match;
98 unsigned char c2 = *name;
99 if (c1 == '\0' || is_glob_special(c1))
110 * If we don't match the matchstring exactly,
111 * we need to match by fnmatch
113 matchlen = strlen(match);
114 if (strncmp(match, name, matchlen))
115 return !fnmatch(match, name, 0) ? MATCHED_FNMATCH : 0;
117 if (namelen == matchlen)
118 return MATCHED_EXACTLY;
119 if (match[matchlen-1] == '/' || name[matchlen] == '/')
120 return MATCHED_RECURSIVELY;
125 * Given a name and a list of pathspecs, see if the name matches
126 * any of the pathspecs. The caller is also interested in seeing
127 * all pathspec matches some names it calls this function with
128 * (otherwise the user could have mistyped the unmatched pathspec),
129 * and a mark is left in seen[] array for pathspec element that
130 * actually matched anything.
132 int match_pathspec(const char **pathspec, const char *name, int namelen,
133 int prefix, char *seen)
143 for (i = 0; pathspec[i] != NULL; i++) {
145 const char *match = pathspec[i] + prefix;
146 if (seen && seen[i] == MATCHED_EXACTLY)
148 how = match_one(match, name, namelen);
152 if (seen && seen[i] < how)
159 static int no_wildcard(const char *string)
161 return string[strcspn(string, "*?[{\\")] == '\0';
164 void add_exclude(const char *string, const char *base,
165 int baselen, struct exclude_list *which)
172 if (*string == '!') {
176 len = strlen(string);
177 if (len && string[len - 1] == '/') {
179 x = xmalloc(sizeof(*x) + len);
181 memcpy(s, string, len - 1);
185 flags = EXC_FLAG_MUSTBEDIR;
187 x = xmalloc(sizeof(*x));
190 x->to_exclude = to_exclude;
191 x->patternlen = strlen(string);
193 x->baselen = baselen;
195 if (!strchr(string, '/'))
196 x->flags |= EXC_FLAG_NODIR;
197 if (no_wildcard(string))
198 x->flags |= EXC_FLAG_NOWILDCARD;
199 if (*string == '*' && no_wildcard(string+1))
200 x->flags |= EXC_FLAG_ENDSWITH;
201 ALLOC_GROW(which->excludes, which->nr + 1, which->alloc);
202 which->excludes[which->nr++] = x;
205 static int add_excludes_from_file_1(const char *fname,
209 struct exclude_list *which)
216 fd = open(fname, O_RDONLY);
217 if (fd < 0 || fstat(fd, &st) < 0)
219 size = xsize_t(st.st_size);
224 buf = xmalloc(size+1);
225 if (read_in_full(fd, buf, size) != size)
236 for (i = 0; i < size; i++) {
237 if (buf[i] == '\n') {
238 if (entry != buf + i && entry[0] != '#') {
239 buf[i - (i && buf[i-1] == '\r')] = 0;
240 add_exclude(entry, base, baselen, which);
253 void add_excludes_from_file(struct dir_struct *dir, const char *fname)
255 if (add_excludes_from_file_1(fname, "", 0, NULL,
256 &dir->exclude_list[EXC_FILE]) < 0)
257 die("cannot use %s as an exclude file", fname);
260 static void prep_exclude(struct dir_struct *dir, const char *base, int baselen)
262 struct exclude_list *el;
263 struct exclude_stack *stk = NULL;
266 if ((!dir->exclude_per_dir) ||
267 (baselen + strlen(dir->exclude_per_dir) >= PATH_MAX))
268 return; /* too long a path -- ignore */
270 /* Pop the ones that are not the prefix of the path being checked. */
271 el = &dir->exclude_list[EXC_DIRS];
272 while ((stk = dir->exclude_stack) != NULL) {
273 if (stk->baselen <= baselen &&
274 !strncmp(dir->basebuf, base, stk->baselen))
276 dir->exclude_stack = stk->prev;
277 while (stk->exclude_ix < el->nr)
278 free(el->excludes[--el->nr]);
283 /* Read from the parent directories and push them down. */
284 current = stk ? stk->baselen : -1;
285 while (current < baselen) {
286 struct exclude_stack *stk = xcalloc(1, sizeof(*stk));
294 cp = strchr(base + current + 1, '/');
296 die("oops in prep_exclude");
299 stk->prev = dir->exclude_stack;
300 stk->baselen = cp - base;
301 stk->exclude_ix = el->nr;
302 memcpy(dir->basebuf + current, base + current,
303 stk->baselen - current);
304 strcpy(dir->basebuf + stk->baselen, dir->exclude_per_dir);
305 add_excludes_from_file_1(dir->basebuf,
306 dir->basebuf, stk->baselen,
308 dir->exclude_stack = stk;
309 current = stk->baselen;
311 dir->basebuf[baselen] = '\0';
314 /* Scan the list and let the last match determine the fate.
315 * Return 1 for exclude, 0 for include and -1 for undecided.
317 static int excluded_1(const char *pathname,
318 int pathlen, const char *basename, int *dtype,
319 struct exclude_list *el)
324 for (i = el->nr - 1; 0 <= i; i--) {
325 struct exclude *x = el->excludes[i];
326 const char *exclude = x->pattern;
327 int to_exclude = x->to_exclude;
329 if (x->flags & EXC_FLAG_MUSTBEDIR) {
330 if (*dtype == DT_UNKNOWN)
331 *dtype = get_dtype(NULL, pathname);
332 if (*dtype != DT_DIR)
336 if (x->flags & EXC_FLAG_NODIR) {
338 if (x->flags & EXC_FLAG_NOWILDCARD) {
339 if (!strcmp(exclude, basename))
341 } else if (x->flags & EXC_FLAG_ENDSWITH) {
342 if (x->patternlen - 1 <= pathlen &&
343 !strcmp(exclude + 1, pathname + pathlen - x->patternlen + 1))
346 if (fnmatch(exclude, basename, 0) == 0)
351 /* match with FNM_PATHNAME:
352 * exclude has base (baselen long) implicitly
355 int baselen = x->baselen;
359 if (pathlen < baselen ||
360 (baselen && pathname[baselen-1] != '/') ||
361 strncmp(pathname, x->base, baselen))
364 if (x->flags & EXC_FLAG_NOWILDCARD) {
365 if (!strcmp(exclude, pathname + baselen))
368 if (fnmatch(exclude, pathname+baselen,
375 return -1; /* undecided */
378 int excluded(struct dir_struct *dir, const char *pathname, int *dtype_p)
380 int pathlen = strlen(pathname);
382 const char *basename = strrchr(pathname, '/');
383 basename = (basename) ? basename+1 : pathname;
385 prep_exclude(dir, pathname, basename-pathname);
386 for (st = EXC_CMDL; st <= EXC_FILE; st++) {
387 switch (excluded_1(pathname, pathlen, basename,
388 dtype_p, &dir->exclude_list[st])) {
398 static struct dir_entry *dir_entry_new(const char *pathname, int len)
400 struct dir_entry *ent;
402 ent = xmalloc(sizeof(*ent) + len + 1);
404 memcpy(ent->name, pathname, len);
409 static struct dir_entry *dir_add_name(struct dir_struct *dir, const char *pathname, int len)
411 if (cache_name_exists(pathname, len, ignore_case))
414 ALLOC_GROW(dir->entries, dir->nr+1, dir->alloc);
415 return dir->entries[dir->nr++] = dir_entry_new(pathname, len);
418 static struct dir_entry *dir_add_ignored(struct dir_struct *dir, const char *pathname, int len)
420 if (!cache_name_is_other(pathname, len))
423 ALLOC_GROW(dir->ignored, dir->ignored_nr+1, dir->ignored_alloc);
424 return dir->ignored[dir->ignored_nr++] = dir_entry_new(pathname, len);
428 index_nonexistent = 0,
434 * The index sorts alphabetically by entry name, which
435 * means that a gitlink sorts as '\0' at the end, while
436 * a directory (which is defined not as an entry, but as
437 * the files it contains) will sort with the '/' at the
440 static enum exist_status directory_exists_in_index(const char *dirname, int len)
442 int pos = cache_name_pos(dirname, len);
445 while (pos < active_nr) {
446 struct cache_entry *ce = active_cache[pos++];
447 unsigned char endchar;
449 if (strncmp(ce->name, dirname, len))
451 endchar = ce->name[len];
455 return index_directory;
456 if (!endchar && S_ISGITLINK(ce->ce_mode))
459 return index_nonexistent;
463 * When we find a directory when traversing the filesystem, we
464 * have three distinct cases:
467 * - see it as a directory
470 * and which one we choose depends on a combination of existing
471 * git index contents and the flags passed into the directory
474 * Case 1: If we *already* have entries in the index under that
475 * directory name, we always recurse into the directory to see
478 * Case 2: If we *already* have that directory name as a gitlink,
479 * we always continue to see it as a gitlink, regardless of whether
480 * there is an actual git directory there or not (it might not
481 * be checked out as a subproject!)
483 * Case 3: if we didn't have it in the index previously, we
484 * have a few sub-cases:
486 * (a) if "show_other_directories" is true, we show it as
487 * just a directory, unless "hide_empty_directories" is
488 * also true and the directory is empty, in which case
489 * we just ignore it entirely.
490 * (b) if it looks like a git directory, and we don't have
491 * 'no_gitlinks' set we treat it as a gitlink, and show it
493 * (c) otherwise, we recurse into it.
495 enum directory_treatment {
498 recurse_into_directory,
501 static enum directory_treatment treat_directory(struct dir_struct *dir,
502 const char *dirname, int len,
503 const struct path_simplify *simplify)
505 /* The "len-1" is to strip the final '/' */
506 switch (directory_exists_in_index(dirname, len-1)) {
507 case index_directory:
508 return recurse_into_directory;
511 if (dir->flags & DIR_SHOW_OTHER_DIRECTORIES)
512 return ignore_directory;
513 return show_directory;
515 case index_nonexistent:
516 if (dir->flags & DIR_SHOW_OTHER_DIRECTORIES)
518 if (!(dir->flags & DIR_NO_GITLINKS)) {
519 unsigned char sha1[20];
520 if (resolve_gitlink_ref(dirname, "HEAD", sha1) == 0)
521 return show_directory;
523 return recurse_into_directory;
526 /* This is the "show_other_directories" case */
527 if (!(dir->flags & DIR_HIDE_EMPTY_DIRECTORIES))
528 return show_directory;
529 if (!read_directory_recursive(dir, dirname, dirname, len, 1, simplify))
530 return ignore_directory;
531 return show_directory;
535 * This is an inexact early pruning of any recursive directory
536 * reading - if the path cannot possibly be in the pathspec,
537 * return true, and we'll skip it early.
539 static int simplify_away(const char *path, int pathlen, const struct path_simplify *simplify)
543 const char *match = simplify->path;
544 int len = simplify->len;
550 if (!memcmp(path, match, len))
559 static int in_pathspec(const char *path, int len, const struct path_simplify *simplify)
562 for (; simplify->path; simplify++) {
563 if (len == simplify->len
564 && !memcmp(path, simplify->path, len))
571 static int get_dtype(struct dirent *de, const char *path)
573 int dtype = de ? DTYPE(de) : DT_UNKNOWN;
576 if (dtype != DT_UNKNOWN)
578 if (lstat(path, &st))
580 if (S_ISREG(st.st_mode))
582 if (S_ISDIR(st.st_mode))
584 if (S_ISLNK(st.st_mode))
590 * Read a directory tree. We currently ignore anything but
591 * directories, regular files and symlinks. That's because git
592 * doesn't handle them at all yet. Maybe that will change some
595 * Also, we ignore the name ".git" (even if it is not a directory).
596 * That likely will not change.
598 static int read_directory_recursive(struct dir_struct *dir, const char *path, const char *base, int baselen, int check_only, const struct path_simplify *simplify)
600 DIR *fdir = opendir(*path ? path : ".");
605 char fullname[PATH_MAX + 1];
606 memcpy(fullname, base, baselen);
608 while ((de = readdir(fdir)) != NULL) {
612 if (is_dot_or_dotdot(de->d_name) ||
613 !strcmp(de->d_name, ".git"))
615 len = strlen(de->d_name);
616 /* Ignore overly long pathnames! */
617 if (len + baselen + 8 > sizeof(fullname))
619 memcpy(fullname + baselen, de->d_name, len+1);
620 if (simplify_away(fullname, baselen + len, simplify))
624 exclude = excluded(dir, fullname, &dtype);
625 if (exclude && (dir->flags & DIR_COLLECT_IGNORED)
626 && in_pathspec(fullname, baselen + len, simplify))
627 dir_add_ignored(dir, fullname, baselen + len);
630 * Excluded? If we don't explicitly want to show
631 * ignored files, ignore it
633 if (exclude && !(dir->flags & DIR_SHOW_IGNORED))
636 if (dtype == DT_UNKNOWN)
637 dtype = get_dtype(de, fullname);
640 * Do we want to see just the ignored files?
641 * We still need to recurse into directories,
642 * even if we don't ignore them, since the
643 * directory may contain files that we do..
645 if (!exclude && (dir->flags & DIR_SHOW_IGNORED)) {
654 memcpy(fullname + baselen + len, "/", 2);
656 switch (treat_directory(dir, fullname, baselen + len, simplify)) {
658 if (exclude != !!(dir->flags
662 case recurse_into_directory:
663 contents += read_directory_recursive(dir,
664 fullname, fullname, baselen + len, 0, simplify);
666 case ignore_directory:
678 dir_add_name(dir, fullname, baselen + len);
687 static int cmp_name(const void *p1, const void *p2)
689 const struct dir_entry *e1 = *(const struct dir_entry **)p1;
690 const struct dir_entry *e2 = *(const struct dir_entry **)p2;
692 return cache_name_compare(e1->name, e1->len,
697 * Return the length of the "simple" part of a path match limiter.
699 static int simple_length(const char *match)
704 unsigned char c = *match++;
706 if (c == '\0' || is_glob_special(c))
711 static struct path_simplify *create_simplify(const char **pathspec)
714 struct path_simplify *simplify = NULL;
719 for (nr = 0 ; ; nr++) {
722 alloc = alloc_nr(alloc);
723 simplify = xrealloc(simplify, alloc * sizeof(*simplify));
728 simplify[nr].path = match;
729 simplify[nr].len = simple_length(match);
731 simplify[nr].path = NULL;
732 simplify[nr].len = 0;
736 static void free_simplify(struct path_simplify *simplify)
741 int read_directory(struct dir_struct *dir, const char *path, const char *base, int baselen, const char **pathspec)
743 struct path_simplify *simplify;
745 if (has_symlink_leading_path(path, strlen(path)))
748 simplify = create_simplify(pathspec);
749 read_directory_recursive(dir, path, base, baselen, 0, simplify);
750 free_simplify(simplify);
751 qsort(dir->entries, dir->nr, sizeof(struct dir_entry *), cmp_name);
752 qsort(dir->ignored, dir->ignored_nr, sizeof(struct dir_entry *), cmp_name);
756 int file_exists(const char *f)
759 return lstat(f, &sb) == 0;
763 * get_relative_cwd() gets the prefix of the current working directory
764 * relative to 'dir'. If we are not inside 'dir', it returns NULL.
766 * As a convenience, it also returns NULL if 'dir' is already NULL. The
767 * reason for this behaviour is that it is natural for functions returning
768 * directory names to return NULL to say "this directory does not exist"
769 * or "this directory is invalid". These cases are usually handled the
770 * same as if the cwd is not inside 'dir' at all, so get_relative_cwd()
771 * returns NULL for both of them.
773 * Most notably, get_relative_cwd(buffer, size, get_git_work_tree())
774 * unifies the handling of "outside work tree" with "no work tree at all".
776 char *get_relative_cwd(char *buffer, int size, const char *dir)
782 if (!getcwd(buffer, size))
783 die_errno("can't find the current directory");
785 if (!is_absolute_path(dir))
786 dir = make_absolute_path(dir);
788 while (*dir && *dir == *cwd) {
799 int is_inside_dir(const char *dir)
801 char buffer[PATH_MAX];
802 return get_relative_cwd(buffer, sizeof(buffer), dir) != NULL;
805 int is_empty_dir(const char *path)
807 DIR *dir = opendir(path);
814 while ((e = readdir(dir)) != NULL)
815 if (!is_dot_or_dotdot(e->d_name)) {
824 int remove_dir_recursively(struct strbuf *path, int only_empty)
826 DIR *dir = opendir(path->buf);
828 int ret = 0, original_len = path->len, len;
832 if (path->buf[original_len - 1] != '/')
833 strbuf_addch(path, '/');
836 while ((e = readdir(dir)) != NULL) {
838 if (is_dot_or_dotdot(e->d_name))
841 strbuf_setlen(path, len);
842 strbuf_addstr(path, e->d_name);
843 if (lstat(path->buf, &st))
845 else if (S_ISDIR(st.st_mode)) {
846 if (!remove_dir_recursively(path, only_empty))
847 continue; /* happy */
848 } else if (!only_empty && !unlink(path->buf))
849 continue; /* happy, too */
851 /* path too long, stat fails, or non-directory still exists */
857 strbuf_setlen(path, original_len);
859 ret = rmdir(path->buf);
863 void setup_standard_excludes(struct dir_struct *dir)
867 dir->exclude_per_dir = ".gitignore";
868 path = git_path("info/exclude");
869 if (!access(path, R_OK))
870 add_excludes_from_file(dir, path);
871 if (excludes_file && !access(excludes_file, R_OK))
872 add_excludes_from_file(dir, excludes_file);
875 int remove_path(const char *name)
879 if (unlink(name) && errno != ENOENT)
882 slash = strrchr(name, '/');
884 char *dirs = xstrdup(name);
885 slash = dirs + (slash - name);
888 } while (rmdir(dirs) && (slash = strrchr(dirs, '/')));