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