git-cvsimport-script: use private index.
[git.git] / fsck-cache.c
index cb01095..9167924 100644 (file)
@@ -12,6 +12,7 @@
 static int show_root = 0;
 static int show_tags = 0;
 static int show_unreachable = 0;
+static int keep_cache_objects = 0; 
 static unsigned char head_sha1[20];
 
 static void check_connectivity(void)
@@ -24,7 +25,8 @@ static void check_connectivity(void)
                struct object_list *refs;
 
                if (!obj->parsed) {
-                       printf("missing %s %s\n", obj->type, sha1_to_hex(obj->sha1));
+                       printf("missing %s %s\n",
+                              obj->type, sha1_to_hex(obj->sha1));
                        continue;
                }
 
@@ -34,15 +36,12 @@ static void check_connectivity(void)
                        printf("broken link from %7s %s\n",
                               obj->type, sha1_to_hex(obj->sha1));
                        printf("              to %7s %s\n",
-                              obj->type, sha1_to_hex(refs->item->sha1));
+                              refs->item->type, sha1_to_hex(refs->item->sha1));
                }
 
-               /* Don't bother with tag reachability. */
-               if (obj->type == tag_type)
-                       continue;
-
                if (show_unreachable && !(obj->flags & REACHABLE)) {
-                       printf("unreachable %s %s\n", obj->type, sha1_to_hex(obj->sha1));
+                       printf("unreachable %s %s\n",
+                              obj->type, sha1_to_hex(obj->sha1));
                        continue;
                }
 
@@ -53,18 +52,116 @@ static void check_connectivity(void)
        }
 }
 
+/*
+ * The entries in a tree are ordered in the _path_ order,
+ * which means that a directory entry is ordered by adding
+ * a slash to the end of it.
+ *
+ * So a directory called "a" is ordered _after_ a file
+ * called "a.c", because "a/" sorts after "a.c".
+ */
+#define TREE_UNORDERED (-1)
+#define TREE_HAS_DUPS  (-2)
+
+static int verify_ordered(struct tree_entry_list *a, struct tree_entry_list *b)
+{
+       int len1 = strlen(a->name);
+       int len2 = strlen(b->name);
+       int len = len1 < len2 ? len1 : len2;
+       unsigned char c1, c2;
+       int cmp;
+
+       cmp = memcmp(a->name, b->name, len);
+       if (cmp < 0)
+               return 0;
+       if (cmp > 0)
+               return TREE_UNORDERED;
+
+       /*
+        * Ok, the first <len> characters are the same.
+        * Now we need to order the next one, but turn
+        * a '\0' into a '/' for a directory entry.
+        */
+       c1 = a->name[len];
+       c2 = b->name[len];
+       if (!c1 && !c2)
+               /*
+                * git-write-tree used to write out a nonsense tree that has
+                * entries with the same name, one blob and one tree.  Make
+                * sure we do not have duplicate entries.
+                */
+               return TREE_HAS_DUPS;
+       if (!c1 && a->directory)
+               c1 = '/';
+       if (!c2 && b->directory)
+               c2 = '/';
+       return c1 < c2 ? 0 : TREE_UNORDERED;
+}
+
 static int fsck_tree(struct tree *item)
 {
-       if (item->has_full_path) {
-               fprintf(stderr, "warning: fsck-cache: tree %s "
+       int has_full_path = 0;
+       struct tree_entry_list *entry, *last;
+
+       last = NULL;
+       for (entry = item->entries; entry; entry = entry->next) {
+               if (strchr(entry->name, '/'))
+                       has_full_path = 1;
+
+               switch (entry->mode) {
+               /*
+                * Standard modes.. 
+                */
+               case S_IFREG | 0755:
+               case S_IFREG | 0644:
+               case S_IFLNK:
+               case S_IFDIR:
+                       break;
+               /*
+                * This is nonstandard, but we had a few of these
+                * early on when we honored the full set of mode
+                * bits..
+                */
+               case S_IFREG | 0664:
+                       break;
+               default:
+                       printf("tree %s has entry %o %s\n",
+                               sha1_to_hex(item->object.sha1),
+                               entry->mode, entry->name);
+               }
+
+               if (last) {
+                       switch (verify_ordered(last, entry)) {
+                       case TREE_UNORDERED:
+                               fprintf(stderr, "tree %s not ordered\n",
+                                       sha1_to_hex(item->object.sha1));
+                               return -1;
+                       case TREE_HAS_DUPS:
+                               fprintf(stderr, "tree %s has duplicate entries for '%s'\n",
+                                       sha1_to_hex(item->object.sha1),
+                                       entry->name);
+                               return -1;
+                       default:
+                               break;
+                       }
+               }
+
+               last = entry;
+       }
+
+       if (has_full_path) {
+               fprintf(stderr, "warning: git-fsck-cache: tree %s "
                        "has full pathnames in it\n", 
                        sha1_to_hex(item->object.sha1));
        }
+
        return 0;
 }
 
 static int fsck_commit(struct commit *commit)
 {
+       free(commit->buffer);
+       commit->buffer = NULL;
        if (!commit->tree)
                return -1;
        if (!commit->parents && show_root)
@@ -77,14 +174,17 @@ static int fsck_commit(struct commit *commit)
 
 static int fsck_tag(struct tag *tag)
 {
+       struct object *tagged = tag->tagged;
+
+       if (!tagged) {
+               printf("bad object in tag %s\n", sha1_to_hex(tag->object.sha1));
+               return -1;
+       }
        if (!show_tags)
                return 0;
 
-       printf("tagged %s %s",
-              tag->tagged->type,
-              sha1_to_hex(tag->tagged->sha1));
-       printf(" (%s) in %s\n",
-              tag->tag, sha1_to_hex(tag->object.sha1));
+       printf("tagged %s %s", tagged->type, sha1_to_hex(tagged->sha1));
+       printf(" (%s) in %s\n", tag->tag, sha1_to_hex(tag->object.sha1));
        return 0;
 }
 
@@ -196,6 +296,76 @@ static int fsck_dir(int i, char *path)
        return 0;
 }
 
+static int read_sha1_reference(const char *path)
+{
+       char hexname[60];
+       unsigned char sha1[20];
+       int fd = open(path, O_RDONLY), len;
+       struct object *obj;
+
+       if (fd < 0)
+               return -1;
+
+       len = read(fd, hexname, sizeof(hexname));
+       close(fd);
+       if (len < 40)
+               return -1;
+
+       if (get_sha1_hex(hexname, sha1) < 0)
+               return -1;
+
+       obj = lookup_object(sha1);
+       if (!obj)
+               return error("%s: invalid sha1 pointer %.40s", path, hexname);
+
+       obj->used = 1;
+       mark_reachable(obj, REACHABLE);
+       return 0;
+}
+
+static int find_file_objects(const char *base, const char *name)
+{
+       int baselen = strlen(base);
+       int namelen = strlen(name);
+       char *path = xmalloc(baselen + namelen + 2);
+       struct stat st;
+
+       memcpy(path, base, baselen);
+       path[baselen] = '/';
+       memcpy(path + baselen + 1, name, namelen+1);
+       if (stat(path, &st) < 0)
+               return 0;
+
+       /*
+        * Recurse into directories
+        */
+       if (S_ISDIR(st.st_mode)) {
+               int count = 0;
+               DIR *dir = opendir(path);
+               if (dir) {
+                       struct dirent *de;
+                       while ((de = readdir(dir)) != NULL) {
+                               if (de->d_name[0] == '.')
+                                       continue;
+                               count += find_file_objects(path, de->d_name);
+                       }
+                       closedir(dir);
+               }
+               return count;
+       }
+       if (S_ISREG(st.st_mode))
+               return read_sha1_reference(path) == 0;
+       return 0;
+}
+
+static void get_default_heads(void)
+{
+       char *git_dir = gitenv(GIT_DIR_ENVIRONMENT) ? : DEFAULT_GIT_DIR_ENVIRONMENT;
+       int count = find_file_objects(git_dir, "refs");
+       if (!count)
+               die("No default references");
+}
+
 int main(int argc, char **argv)
 {
        int i, heads;
@@ -216,11 +386,15 @@ int main(int argc, char **argv)
                        show_root = 1;
                        continue;
                }
+               if (!strcmp(arg, "--cache")) {
+                       keep_cache_objects = 1;
+                       continue;
+               }
                if (*arg == '-')
-                       usage("fsck-cache [--tags] [[--unreachable] <head-sha1>*]");
+                       usage("git-fsck-cache [--tags] [[--unreachable] [--cache] <head-sha1>*]");
        }
 
-       sha1_dir = getenv(DB_ENVIRONMENT) ? : DEFAULT_DB_ENVIRONMENT;
+       sha1_dir = get_object_directory();
        for (i = 0; i < 256; i++) {
                static char dir[4096];
                sprintf(dir, "%s/%02x", sha1_dir, i);
@@ -236,14 +410,12 @@ int main(int argc, char **argv)
                        continue;
 
                if (!get_sha1(arg, head_sha1)) {
-                       struct commit *commit = lookup_commit(head_sha1);
-                       struct object *obj;
+                       struct object *obj = lookup_object(head_sha1);
 
-                       /* Error is printed by lookup_commit(). */
-                       if (!commit)
+                       /* Error is printed by lookup_object(). */
+                       if (!obj)
                                continue;
 
-                       obj = &commit->object;
                        obj->used = 1;
                        mark_reachable(obj, REACHABLE);
                        heads++;
@@ -252,12 +424,28 @@ int main(int argc, char **argv)
                error("expected sha1, got %s", arg);
        }
 
+       /*
+        * If we've not been given any explicit head information, do the
+        * default ones from .git/refs. We also consider the index file
+        * in this case (ie this implies --cache).
+        */
        if (!heads) {
-               if (show_unreachable) {
-                       fprintf(stderr, "unable to do reachability without a head\n");
-                       show_unreachable = 0; 
+               get_default_heads();
+               keep_cache_objects = 1;
+       }
+
+       if (keep_cache_objects) {
+               int i;
+               read_cache();
+               for (i = 0; i < active_nr; i++) {
+                       struct blob *blob = lookup_blob(active_cache[i]->sha1);
+                       struct object *obj;
+                       if (!blob)
+                               continue;
+                       obj = &blob->object;
+                       obj->used = 1;
+                       mark_reachable(obj, REACHABLE);
                }
-               fprintf(stderr, "expect dangling commits - potential heads - due to lack of head information\n");
        }
 
        check_connectivity();