Andrew's git - gitweb.git/blob

   1#include "cache.h"
   2#include "object.h"
   3#include "blob.h"
   4#include "tree.h"
   5#include "tree-walk.h"
   6#include "commit.h"
   7#include "tag.h"
   8#include "fsck.h"
   9#include "utf8.h"
  10
  11static int fsck_walk_tree(struct tree *tree, fsck_walk_func walk, void *data)
  12{
  13        struct tree_desc desc;
  14        struct name_entry entry;
  15        int res = 0;
  16
  17        if (parse_tree(tree))
  18                return -1;
  19
  20        init_tree_desc(&desc, tree->buffer, tree->size);
  21        while (tree_entry(&desc, &entry)) {
  22                int result;
  23
  24                if (S_ISGITLINK(entry.mode))
  25                        continue;
  26                if (S_ISDIR(entry.mode))
  27                        result = walk(&lookup_tree(entry.sha1)->object, OBJ_TREE, data);
  28                else if (S_ISREG(entry.mode) || S_ISLNK(entry.mode))
  29                        result = walk(&lookup_blob(entry.sha1)->object, OBJ_BLOB, data);
  30                else {
  31                        result = error("in tree %s: entry %s has bad mode %.6o",
  32                                        sha1_to_hex(tree->object.sha1), entry.path, entry.mode);
  33                }
  34                if (result < 0)
  35                        return result;
  36                if (!res)
  37                        res = result;
  38        }
  39        return res;
  40}
  41
  42static int fsck_walk_commit(struct commit *commit, fsck_walk_func walk, void *data)
  43{
  44        struct commit_list *parents;
  45        int res;
  46        int result;
  47
  48        if (parse_commit(commit))
  49                return -1;
  50
  51        result = walk((struct object *)commit->tree, OBJ_TREE, data);
  52        if (result < 0)
  53                return result;
  54        res = result;
  55
  56        parents = commit->parents;
  57        while (parents) {
  58                result = walk((struct object *)parents->item, OBJ_COMMIT, data);
  59                if (result < 0)
  60                        return result;
  61                if (!res)
  62                        res = result;
  63                parents = parents->next;
  64        }
  65        return res;
  66}
  67
  68static int fsck_walk_tag(struct tag *tag, fsck_walk_func walk, void *data)
  69{
  70        if (parse_tag(tag))
  71                return -1;
  72        return walk(tag->tagged, OBJ_ANY, data);
  73}
  74
  75int fsck_walk(struct object *obj, fsck_walk_func walk, void *data)
  76{
  77        if (!obj)
  78                return -1;
  79        switch (obj->type) {
  80        case OBJ_BLOB:
  81                return 0;
  82        case OBJ_TREE:
  83                return fsck_walk_tree((struct tree *)obj, walk, data);
  84        case OBJ_COMMIT:
  85                return fsck_walk_commit((struct commit *)obj, walk, data);
  86        case OBJ_TAG:
  87                return fsck_walk_tag((struct tag *)obj, walk, data);
  88        default:
  89                error("Unknown object type for %s", sha1_to_hex(obj->sha1));
  90                return -1;
  91        }
  92}
  93
  94/*
  95 * The entries in a tree are ordered in the _path_ order,
  96 * which means that a directory entry is ordered by adding
  97 * a slash to the end of it.
  98 *
  99 * So a directory called "a" is ordered _after_ a file
 100 * called "a.c", because "a/" sorts after "a.c".
 101 */
 102#define TREE_UNORDERED (-1)
 103#define TREE_HAS_DUPS  (-2)
 104
 105static int verify_ordered(unsigned mode1, const char *name1, unsigned mode2, const char *name2)
 106{
 107        int len1 = strlen(name1);
 108        int len2 = strlen(name2);
 109        int len = len1 < len2 ? len1 : len2;
 110        unsigned char c1, c2;
 111        int cmp;
 112
 113        cmp = memcmp(name1, name2, len);
 114        if (cmp < 0)
 115                return 0;
 116        if (cmp > 0)
 117                return TREE_UNORDERED;
 118
 119        /*
 120         * Ok, the first <len> characters are the same.
 121         * Now we need to order the next one, but turn
 122         * a '\0' into a '/' for a directory entry.
 123         */
 124        c1 = name1[len];
 125        c2 = name2[len];
 126        if (!c1 && !c2)
 127                /*
 128                 * git-write-tree used to write out a nonsense tree that has
 129                 * entries with the same name, one blob and one tree.  Make
 130                 * sure we do not have duplicate entries.
 131                 */
 132                return TREE_HAS_DUPS;
 133        if (!c1 && S_ISDIR(mode1))
 134                c1 = '/';
 135        if (!c2 && S_ISDIR(mode2))
 136                c2 = '/';
 137        return c1 < c2 ? 0 : TREE_UNORDERED;
 138}
 139
 140static int fsck_tree(struct tree *item, int strict, fsck_error error_func)
 141{
 142        int retval;
 143        int has_null_sha1 = 0;
 144        int has_full_path = 0;
 145        int has_empty_name = 0;
 146        int has_dot = 0;
 147        int has_dotdot = 0;
 148        int has_dotgit = 0;
 149        int has_zero_pad = 0;
 150        int has_bad_modes = 0;
 151        int has_dup_entries = 0;
 152        int not_properly_sorted = 0;
 153        struct tree_desc desc;
 154        unsigned o_mode;
 155        const char *o_name;
 156
 157        init_tree_desc(&desc, item->buffer, item->size);
 158
 159        o_mode = 0;
 160        o_name = NULL;
 161
 162        while (desc.size) {
 163                unsigned mode;
 164                const char *name;
 165                const unsigned char *sha1;
 166
 167                sha1 = tree_entry_extract(&desc, &name, &mode);
 168
 169                if (is_null_sha1(sha1))
 170                        has_null_sha1 = 1;
 171                if (strchr(name, '/'))
 172                        has_full_path = 1;
 173                if (!*name)
 174                        has_empty_name = 1;
 175                if (!strcmp(name, "."))
 176                        has_dot = 1;
 177                if (!strcmp(name, ".."))
 178                        has_dotdot = 1;
 179                if (!strcasecmp(name, ".git") || is_hfs_dotgit(name) ||
 180                                is_ntfs_dotgit(name))
 181                        has_dotgit = 1;
 182                has_zero_pad |= *(char *)desc.buffer == '0';
 183                update_tree_entry(&desc);
 184
 185                switch (mode) {
 186                /*
 187                 * Standard modes..
 188                 */
 189                case S_IFREG | 0755:
 190                case S_IFREG | 0644:
 191                case S_IFLNK:
 192                case S_IFDIR:
 193                case S_IFGITLINK:
 194                        break;
 195                /*
 196                 * This is nonstandard, but we had a few of these
 197                 * early on when we honored the full set of mode
 198                 * bits..
 199                 */
 200                case S_IFREG | 0664:
 201                        if (!strict)
 202                                break;
 203                default:
 204                        has_bad_modes = 1;
 205                }
 206
 207                if (o_name) {
 208                        switch (verify_ordered(o_mode, o_name, mode, name)) {
 209                        case TREE_UNORDERED:
 210                                not_properly_sorted = 1;
 211                                break;
 212                        case TREE_HAS_DUPS:
 213                                has_dup_entries = 1;
 214                                break;
 215                        default:
 216                                break;
 217                        }
 218                }
 219
 220                o_mode = mode;
 221                o_name = name;
 222        }
 223
 224        retval = 0;
 225        if (has_null_sha1)
 226                retval += error_func(&item->object, FSCK_WARN, "contains entries pointing to null sha1");
 227        if (has_full_path)
 228                retval += error_func(&item->object, FSCK_WARN, "contains full pathnames");
 229        if (has_empty_name)
 230                retval += error_func(&item->object, FSCK_WARN, "contains empty pathname");
 231        if (has_dot)
 232                retval += error_func(&item->object, FSCK_WARN, "contains '.'");
 233        if (has_dotdot)
 234                retval += error_func(&item->object, FSCK_WARN, "contains '..'");
 235        if (has_dotgit)
 236                retval += error_func(&item->object, FSCK_WARN, "contains '.git'");
 237        if (has_zero_pad)
 238                retval += error_func(&item->object, FSCK_WARN, "contains zero-padded file modes");
 239        if (has_bad_modes)
 240                retval += error_func(&item->object, FSCK_WARN, "contains bad file modes");
 241        if (has_dup_entries)
 242                retval += error_func(&item->object, FSCK_ERROR, "contains duplicate file entries");
 243        if (not_properly_sorted)
 244                retval += error_func(&item->object, FSCK_ERROR, "not properly sorted");
 245        return retval;
 246}
 247
 248static int fsck_ident(char **ident, struct object *obj, fsck_error error_func)
 249{
 250        if (**ident == '<')
 251                return error_func(obj, FSCK_ERROR, "invalid author/committer line - missing space before email");
 252        *ident += strcspn(*ident, "<>\n");
 253        if (**ident == '>')
 254                return error_func(obj, FSCK_ERROR, "invalid author/committer line - bad name");
 255        if (**ident != '<')
 256                return error_func(obj, FSCK_ERROR, "invalid author/committer line - missing email");
 257        if ((*ident)[-1] != ' ')
 258                return error_func(obj, FSCK_ERROR, "invalid author/committer line - missing space before email");
 259        (*ident)++;
 260        *ident += strcspn(*ident, "<>\n");
 261        if (**ident != '>')
 262                return error_func(obj, FSCK_ERROR, "invalid author/committer line - bad email");
 263        (*ident)++;
 264        if (**ident != ' ')
 265                return error_func(obj, FSCK_ERROR, "invalid author/committer line - missing space before date");
 266        (*ident)++;
 267        if (**ident == '0' && (*ident)[1] != ' ')
 268                return error_func(obj, FSCK_ERROR, "invalid author/committer line - zero-padded date");
 269        *ident += strspn(*ident, "0123456789");
 270        if (**ident != ' ')
 271                return error_func(obj, FSCK_ERROR, "invalid author/committer line - bad date");
 272        (*ident)++;
 273        if ((**ident != '+' && **ident != '-') ||
 274            !isdigit((*ident)[1]) ||
 275            !isdigit((*ident)[2]) ||
 276            !isdigit((*ident)[3]) ||
 277            !isdigit((*ident)[4]) ||
 278            ((*ident)[5] != '\n'))
 279                return error_func(obj, FSCK_ERROR, "invalid author/committer line - bad time zone");
 280        (*ident) += 6;
 281        return 0;
 282}
 283
 284static int fsck_commit(struct commit *commit, fsck_error error_func)
 285{
 286        char *buffer = commit->buffer;
 287        unsigned char tree_sha1[20], sha1[20];
 288        struct commit_graft *graft;
 289        int parents = 0;
 290        int err;
 291
 292        if (commit->date == ULONG_MAX)
 293                return error_func(&commit->object, FSCK_ERROR, "invalid author/committer line");
 294
 295        if (memcmp(buffer, "tree ", 5))
 296                return error_func(&commit->object, FSCK_ERROR, "invalid format - expected 'tree' line");
 297        if (get_sha1_hex(buffer+5, tree_sha1) || buffer[45] != '\n')
 298                return error_func(&commit->object, FSCK_ERROR, "invalid 'tree' line format - bad sha1");
 299        buffer += 46;
 300        while (!memcmp(buffer, "parent ", 7)) {
 301                if (get_sha1_hex(buffer+7, sha1) || buffer[47] != '\n')
 302                        return error_func(&commit->object, FSCK_ERROR, "invalid 'parent' line format - bad sha1");
 303                buffer += 48;
 304                parents++;
 305        }
 306        graft = lookup_commit_graft(commit->object.sha1);
 307        if (graft) {
 308                struct commit_list *p = commit->parents;
 309                parents = 0;
 310                while (p) {
 311                        p = p->next;
 312                        parents++;
 313                }
 314                if (graft->nr_parent == -1 && !parents)
 315                        ; /* shallow commit */
 316                else if (graft->nr_parent != parents)
 317                        return error_func(&commit->object, FSCK_ERROR, "graft objects missing");
 318        } else {
 319                struct commit_list *p = commit->parents;
 320                while (p && parents) {
 321                        p = p->next;
 322                        parents--;
 323                }
 324                if (p || parents)
 325                        return error_func(&commit->object, FSCK_ERROR, "parent objects missing");
 326        }
 327        if (memcmp(buffer, "author ", 7))
 328                return error_func(&commit->object, FSCK_ERROR, "invalid format - expected 'author' line");
 329        buffer += 7;
 330        err = fsck_ident(&buffer, &commit->object, error_func);
 331        if (err)
 332                return err;
 333        if (memcmp(buffer, "committer ", strlen("committer ")))
 334                return error_func(&commit->object, FSCK_ERROR, "invalid format - expected 'committer' line");
 335        buffer += strlen("committer ");
 336        err = fsck_ident(&buffer, &commit->object, error_func);
 337        if (err)
 338                return err;
 339        if (!commit->tree)
 340                return error_func(&commit->object, FSCK_ERROR, "could not load commit's tree %s", sha1_to_hex(tree_sha1));
 341
 342        return 0;
 343}
 344
 345static int fsck_tag(struct tag *tag, fsck_error error_func)
 346{
 347        struct object *tagged = tag->tagged;
 348
 349        if (!tagged)
 350                return error_func(&tag->object, FSCK_ERROR, "could not load tagged object");
 351        return 0;
 352}
 353
 354int fsck_object(struct object *obj, int strict, fsck_error error_func)
 355{
 356        if (!obj)
 357                return error_func(obj, FSCK_ERROR, "no valid object to fsck");
 358
 359        if (obj->type == OBJ_BLOB)
 360                return 0;
 361        if (obj->type == OBJ_TREE)
 362                return fsck_tree((struct tree *) obj, strict, error_func);
 363        if (obj->type == OBJ_COMMIT)
 364                return fsck_commit((struct commit *) obj, error_func);
 365        if (obj->type == OBJ_TAG)
 366                return fsck_tag((struct tag *) obj, error_func);
 367
 368        return error_func(obj, FSCK_ERROR, "unknown type '%d' (internal fsck error)",
 369                          obj->type);
 370}
 371
 372int fsck_error_function(struct object *obj, int type, const char *fmt, ...)
 373{
 374        va_list ap;
 375        struct strbuf sb = STRBUF_INIT;
 376
 377        strbuf_addf(&sb, "object %s:", sha1_to_hex(obj->sha1));
 378
 379        va_start(ap, fmt);
 380        strbuf_vaddf(&sb, fmt, ap);
 381        va_end(ap);
 382
 383        error("%s", sb.buf);
 384        strbuf_release(&sb);
 385        return 1;
 386}