diff-lib.c

   1 /*
   2  * Copyright (C) 2005 Junio C Hamano
   3  */
   4 #include "cache.h"
   5 #include "quote.h"
   6 #include "commit.h"
   7 #include "diff.h"
   8 #include "diffcore.h"
   9 #include "revision.h"
  10 #include "cache-tree.h"
  11 #include "unpack-trees.h"
  12 #include "refs.h"
  13 #include "submodule.h"
  14 #include "dir.h"
  15 #include "fsmonitor.h"
  16 #include "commit-reach.h"
  17
  18 /*
  19  * diff-files
  20  */
  21
  22 /*
  23  * Has the work tree entity been removed?
  24  *
  25  * Return 1 if it was removed from the work tree, 0 if an entity to be
  26  * compared with the cache entry ce still exists (the latter includes
  27  * the case where a directory that is not a submodule repository
  28  * exists for ce that is a submodule -- it is a submodule that is not
  29  * checked out).  Return negative for an error.
  30  */
  31 static int check_removed(const struct cache_entry *ce, struct stat *st)
  32 {
  33         if (lstat(ce->name, st) < 0) {
  34                 if (!is_missing_file_error(errno))
  35                         return -1;
  36                 return 1;
  37         }
  38         if (has_symlink_leading_path(ce->name, ce_namelen(ce)))
  39                 return 1;
  40         if (S_ISDIR(st->st_mode)) {
  41                 struct object_id sub;
  42
  43                 /*
  44                  * If ce is already a gitlink, we can have a plain
  45                  * directory (i.e. the submodule is not checked out),
  46                  * or a checked out submodule.  Either case this is not
  47                  * a case where something was removed from the work tree,
  48                  * so we will return 0.
  49                  *
  50                  * Otherwise, if the directory is not a submodule
  51                  * repository, that means ce which was a blob turned into
  52                  * a directory --- the blob was removed!
  53                  */
  54                 if (!S_ISGITLINK(ce->ce_mode) &&
  55                     resolve_gitlink_ref(ce->name, "HEAD", &sub))
  56                         return 1;
  57         }
  58         return 0;
  59 }
  60
  61 /*
  62  * Has a file changed or has a submodule new commits or a dirty work tree?
  63  *
  64  * Return 1 when changes are detected, 0 otherwise. If the DIRTY_SUBMODULES
  65  * option is set, the caller does not only want to know if a submodule is
  66  * modified at all but wants to know all the conditions that are met (new
  67  * commits, untracked content and/or modified content).
  68  */
  69 static int match_stat_with_submodule(struct diff_options *diffopt,
  70                                      const struct cache_entry *ce,
  71                                      struct stat *st, unsigned ce_option,
  72                                      unsigned *dirty_submodule)
  73 {
  74         int changed = ie_match_stat(diffopt->repo->index, ce, st, ce_option);
  75         if (S_ISGITLINK(ce->ce_mode)) {
  76                 struct diff_flags orig_flags = diffopt->flags;
  77                 if (!diffopt->flags.override_submodule_config)
  78                         set_diffopt_flags_from_submodule_config(diffopt, ce->name);
  79                 if (diffopt->flags.ignore_submodules)
  80                         changed = 0;
  81                 else if (!diffopt->flags.ignore_dirty_submodules &&
  82                          (!changed || diffopt->flags.dirty_submodules))
  83                         *dirty_submodule = is_submodule_modified(ce->name,
  84                                                                  diffopt->flags.ignore_untracked_in_submodules);
  85                 diffopt->flags = orig_flags;
  86         }
  87         return changed;
  88 }
  89
  90 int run_diff_files(struct rev_info *revs, unsigned int option)
  91 {
  92         int entries, i;
  93         int diff_unmerged_stage = revs->max_count;
  94         unsigned ce_option = ((option & DIFF_RACY_IS_MODIFIED)
  95                               ? CE_MATCH_RACY_IS_DIRTY : 0);
  96         uint64_t start = getnanotime();
  97         struct index_state *istate = revs->diffopt.repo->index;
  98
  99         diff_set_mnemonic_prefix(&revs->diffopt, "i/", "w/");
 100
 101         if (diff_unmerged_stage < 0)
 102                 diff_unmerged_stage = 2;
 103         entries = istate->cache_nr;
 104         for (i = 0; i < entries; i++) {
 105                 unsigned int oldmode, newmode;
 106                 struct cache_entry *ce = istate->cache[i];
 107                 int changed;
 108                 unsigned dirty_submodule = 0;
 109                 const struct object_id *old_oid, *new_oid;
 110
 111                 if (diff_can_quit_early(&revs->diffopt))
 112                         break;
 113
 114                 if (!ce_path_match(istate, ce, &revs->prune_data, NULL))
 115                         continue;
 116
 117                 if (ce_stage(ce)) {
 118                         struct combine_diff_path *dpath;
 119                         struct diff_filepair *pair;
 120                         unsigned int wt_mode = 0;
 121                         int num_compare_stages = 0;
 122                         size_t path_len;
 123                         struct stat st;
 124
 125                         path_len = ce_namelen(ce);
 126
 127                         dpath = xmalloc(combine_diff_path_size(5, path_len));
 128                         dpath->path = (char *) &(dpath->parent[5]);
 129
 130                         dpath->next = NULL;
 131                         memcpy(dpath->path, ce->name, path_len);
 132                         dpath->path[path_len] = '\0';
 133                         oidclr(&dpath->oid);
 134                         memset(&(dpath->parent[0]), 0,
 135                                sizeof(struct combine_diff_parent)*5);
 136
 137                         changed = check_removed(ce, &st);
 138                         if (!changed)
 139                                 wt_mode = ce_mode_from_stat(ce, st.st_mode);
 140                         else {
 141                                 if (changed < 0) {
 142                                         perror(ce->name);
 143                                         continue;
 144                                 }
 145                                 wt_mode = 0;
 146                         }
 147                         dpath->mode = wt_mode;
 148
 149                         while (i < entries) {
 150                                 struct cache_entry *nce = istate->cache[i];
 151                                 int stage;
 152
 153                                 if (strcmp(ce->name, nce->name))
 154                                         break;
 155
 156                                 /* Stage #2 (ours) is the first parent,
 157                                  * stage #3 (theirs) is the second.
 158                                  */
 159                                 stage = ce_stage(nce);
 160                                 if (2 <= stage) {
 161                                         int mode = nce->ce_mode;
 162                                         num_compare_stages++;
 163                                         oidcpy(&dpath->parent[stage - 2].oid,
 164                                                &nce->oid);
 165                                         dpath->parent[stage-2].mode = ce_mode_from_stat(nce, mode);
 166                                         dpath->parent[stage-2].status =
 167                                                 DIFF_STATUS_MODIFIED;
 168                                 }
 169
 170                                 /* diff against the proper unmerged stage */
 171                                 if (stage == diff_unmerged_stage)
 172                                         ce = nce;
 173                                 i++;
 174                         }
 175                         /*
 176                          * Compensate for loop update
 177                          */
 178                         i--;
 179
 180                         if (revs->combine_merges && num_compare_stages == 2) {
 181                                 show_combined_diff(dpath, 2, revs);
 182                                 free(dpath);
 183                                 continue;
 184                         }
 185                         FREE_AND_NULL(dpath);
 186
 187                         /*
 188                          * Show the diff for the 'ce' if we found the one
 189                          * from the desired stage.
 190                          */
 191                         pair = diff_unmerge(&revs->diffopt, ce->name);
 192                         if (wt_mode)
 193                                 pair->two->mode = wt_mode;
 194                         if (ce_stage(ce) != diff_unmerged_stage)
 195                                 continue;
 196                 }
 197
 198                 if (ce_uptodate(ce) || ce_skip_worktree(ce))
 199                         continue;
 200
 201                 /* If CE_VALID is set, don't look at workdir for file removal */
 202                 if (ce->ce_flags & CE_VALID) {
 203                         changed = 0;
 204                         newmode = ce->ce_mode;
 205                 } else {
 206                         struct stat st;
 207
 208                         changed = check_removed(ce, &st);
 209                         if (changed) {
 210                                 if (changed < 0) {
 211                                         perror(ce->name);
 212                                         continue;
 213                                 }
 214                                 diff_addremove(&revs->diffopt, '-', ce->ce_mode,
 215                                                &ce->oid,
 216                                                !is_null_oid(&ce->oid),
 217                                                ce->name, 0);
 218                                 continue;
 219                         } else if (revs->diffopt.ita_invisible_in_index &&
 220                                    ce_intent_to_add(ce)) {
 221                                 newmode = ce_mode_from_stat(ce, st.st_mode);
 222                                 diff_addremove(&revs->diffopt, '+', newmode,
 223                                                &null_oid, 0, ce->name, 0);
 224                                 continue;
 225                         }
 226
 227                         changed = match_stat_with_submodule(&revs->diffopt, ce, &st,
 228                                                             ce_option, &dirty_submodule);
 229                         newmode = ce_mode_from_stat(ce, st.st_mode);
 230                 }
 231
 232                 if (!changed && !dirty_submodule) {
 233                         ce_mark_uptodate(ce);
 234                         mark_fsmonitor_valid(istate, ce);
 235                         if (!revs->diffopt.flags.find_copies_harder)
 236                                 continue;
 237                 }
 238                 oldmode = ce->ce_mode;
 239                 old_oid = &ce->oid;
 240                 new_oid = changed ? &null_oid : &ce->oid;
 241                 diff_change(&revs->diffopt, oldmode, newmode,
 242                             old_oid, new_oid,
 243                             !is_null_oid(old_oid),
 244                             !is_null_oid(new_oid),
 245                             ce->name, 0, dirty_submodule);
 246
 247         }
 248         diffcore_std(&revs->diffopt);
 249         diff_flush(&revs->diffopt);
 250         trace_performance_since(start, "diff-files");
 251         return 0;
 252 }
 253
 254 /*
 255  * diff-index
 256  */
 257
 258 /* A file entry went away or appeared */
 259 static void diff_index_show_file(struct rev_info *revs,
 260                                  const char *prefix,
 261                                  const struct cache_entry *ce,
 262                                  const struct object_id *oid, int oid_valid,
 263                                  unsigned int mode,
 264                                  unsigned dirty_submodule)
 265 {
 266         diff_addremove(&revs->diffopt, prefix[0], mode,
 267                        oid, oid_valid, ce->name, dirty_submodule);
 268 }
 269
 270 static int get_stat_data(const struct cache_entry *ce,
 271                          const struct object_id **oidp,
 272                          unsigned int *modep,
 273                          int cached, int match_missing,
 274                          unsigned *dirty_submodule, struct diff_options *diffopt)
 275 {
 276         const struct object_id *oid = &ce->oid;
 277         unsigned int mode = ce->ce_mode;
 278
 279         if (!cached && !ce_uptodate(ce)) {
 280                 int changed;
 281                 struct stat st;
 282                 changed = check_removed(ce, &st);
 283                 if (changed < 0)
 284                         return -1;
 285                 else if (changed) {
 286                         if (match_missing) {
 287                                 *oidp = oid;
 288                                 *modep = mode;
 289                                 return 0;
 290                         }
 291                         return -1;
 292                 }
 293                 changed = match_stat_with_submodule(diffopt, ce, &st,
 294                                                     0, dirty_submodule);
 295                 if (changed) {
 296                         mode = ce_mode_from_stat(ce, st.st_mode);
 297                         oid = &null_oid;
 298                 }
 299         }
 300
 301         *oidp = oid;
 302         *modep = mode;
 303         return 0;
 304 }
 305
 306 static void show_new_file(struct rev_info *revs,
 307                           const struct cache_entry *new_file,
 308                           int cached, int match_missing)
 309 {
 310         const struct object_id *oid;
 311         unsigned int mode;
 312         unsigned dirty_submodule = 0;
 313
 314         /*
 315          * New file in the index: it might actually be different in
 316          * the working tree.
 317          */
 318         if (get_stat_data(new_file, &oid, &mode, cached, match_missing,
 319             &dirty_submodule, &revs->diffopt) < 0)
 320                 return;
 321
 322         diff_index_show_file(revs, "+", new_file, oid, !is_null_oid(oid), mode, dirty_submodule);
 323 }
 324
 325 static int show_modified(struct rev_info *revs,
 326                          const struct cache_entry *old_entry,
 327                          const struct cache_entry *new_entry,
 328                          int report_missing,
 329                          int cached, int match_missing)
 330 {
 331         unsigned int mode, oldmode;
 332         const struct object_id *oid;
 333         unsigned dirty_submodule = 0;
 334
 335         if (get_stat_data(new_entry, &oid, &mode, cached, match_missing,
 336                           &dirty_submodule, &revs->diffopt) < 0) {
 337                 if (report_missing)
 338                         diff_index_show_file(revs, "-", old_entry,
 339                                              &old_entry->oid, 1, old_entry->ce_mode,
 340                                              0);
 341                 return -1;
 342         }
 343
 344         if (revs->combine_merges && !cached &&
 345             (!oideq(oid, &old_entry->oid) || !oideq(&old_entry->oid, &new_entry->oid))) {
 346                 struct combine_diff_path *p;
 347                 int pathlen = ce_namelen(new_entry);
 348
 349                 p = xmalloc(combine_diff_path_size(2, pathlen));
 350                 p->path = (char *) &p->parent[2];
 351                 p->next = NULL;
 352                 memcpy(p->path, new_entry->name, pathlen);
 353                 p->path[pathlen] = 0;
 354                 p->mode = mode;
 355                 oidclr(&p->oid);
 356                 memset(p->parent, 0, 2 * sizeof(struct combine_diff_parent));
 357                 p->parent[0].status = DIFF_STATUS_MODIFIED;
 358                 p->parent[0].mode = new_entry->ce_mode;
 359                 oidcpy(&p->parent[0].oid, &new_entry->oid);
 360                 p->parent[1].status = DIFF_STATUS_MODIFIED;
 361                 p->parent[1].mode = old_entry->ce_mode;
 362                 oidcpy(&p->parent[1].oid, &old_entry->oid);
 363                 show_combined_diff(p, 2, revs);
 364                 free(p);
 365                 return 0;
 366         }
 367
 368         oldmode = old_entry->ce_mode;
 369         if (mode == oldmode && oideq(oid, &old_entry->oid) && !dirty_submodule &&
 370             !revs->diffopt.flags.find_copies_harder)
 371                 return 0;
 372
 373         diff_change(&revs->diffopt, oldmode, mode,
 374                     &old_entry->oid, oid, 1, !is_null_oid(oid),
 375                     old_entry->name, 0, dirty_submodule);
 376         return 0;
 377 }
 378
 379 /*
 380  * This gets a mix of an existing index and a tree, one pathname entry
 381  * at a time. The index entry may be a single stage-0 one, but it could
 382  * also be multiple unmerged entries (in which case idx_pos/idx_nr will
 383  * give you the position and number of entries in the index).
 384  */
 385 static void do_oneway_diff(struct unpack_trees_options *o,
 386                            const struct cache_entry *idx,
 387                            const struct cache_entry *tree)
 388 {
 389         struct rev_info *revs = o->unpack_data;
 390         int match_missing, cached;
 391
 392         /*
 393          * i-t-a entries do not actually exist in the index (if we're
 394          * looking at its content)
 395          */
 396         if (o->index_only &&
 397             revs->diffopt.ita_invisible_in_index &&
 398             idx && ce_intent_to_add(idx)) {
 399                 idx = NULL;
 400                 if (!tree)
 401                         return; /* nothing to diff.. */
 402         }
 403
 404         /* if the entry is not checked out, don't examine work tree */
 405         cached = o->index_only ||
 406                 (idx && ((idx->ce_flags & CE_VALID) || ce_skip_worktree(idx)));
 407
 408         match_missing = revs->match_missing;
 409
 410         if (cached && idx && ce_stage(idx)) {
 411                 struct diff_filepair *pair;
 412                 pair = diff_unmerge(&revs->diffopt, idx->name);
 413                 if (tree)
 414                         fill_filespec(pair->one, &tree->oid, 1,
 415                                       tree->ce_mode);
 416                 return;
 417         }
 418
 419         /*
 420          * Something added to the tree?
 421          */
 422         if (!tree) {
 423                 show_new_file(revs, idx, cached, match_missing);
 424                 return;
 425         }
 426
 427         /*
 428          * Something removed from the tree?
 429          */
 430         if (!idx) {
 431                 diff_index_show_file(revs, "-", tree, &tree->oid, 1,
 432                                      tree->ce_mode, 0);
 433                 return;
 434         }
 435
 436         /* Show difference between old and new */
 437         show_modified(revs, tree, idx, 1, cached, match_missing);
 438 }
 439
 440 /*
 441  * The unpack_trees() interface is designed for merging, so
 442  * the different source entries are designed primarily for
 443  * the source trees, with the old index being really mainly
 444  * used for being replaced by the result.
 445  *
 446  * For diffing, the index is more important, and we only have a
 447  * single tree.
 448  *
 449  * We're supposed to advance o->pos to skip what we have already processed.
 450  *
 451  * This wrapper makes it all more readable, and takes care of all
 452  * the fairly complex unpack_trees() semantic requirements, including
 453  * the skipping, the path matching, the type conflict cases etc.
 454  */
 455 static int oneway_diff(const struct cache_entry * const *src,
 456                        struct unpack_trees_options *o)
 457 {
 458         const struct cache_entry *idx = src[0];
 459         const struct cache_entry *tree = src[1];
 460         struct rev_info *revs = o->unpack_data;
 461
 462         /*
 463          * Unpack-trees generates a DF/conflict entry if
 464          * there was a directory in the index and a tree
 465          * in the tree. From a diff standpoint, that's a
 466          * delete of the tree and a create of the file.
 467          */
 468         if (tree == o->df_conflict_entry)
 469                 tree = NULL;
 470
 471         if (ce_path_match(revs->diffopt.repo->index,
 472                           idx ? idx : tree,
 473                           &revs->prune_data, NULL)) {
 474                 do_oneway_diff(o, idx, tree);
 475                 if (diff_can_quit_early(&revs->diffopt)) {
 476                         o->exiting_early = 1;
 477                         return -1;
 478                 }
 479         }
 480
 481         return 0;
 482 }
 483
 484 static int diff_cache(struct rev_info *revs,
 485                       const struct object_id *tree_oid,
 486                       const char *tree_name,
 487                       int cached)
 488 {
 489         struct tree *tree;
 490         struct tree_desc t;
 491         struct unpack_trees_options opts;
 492
 493         tree = parse_tree_indirect(tree_oid);
 494         if (!tree)
 495                 return error("bad tree object %s",
 496                              tree_name ? tree_name : oid_to_hex(tree_oid));
 497         memset(&opts, 0, sizeof(opts));
 498         opts.head_idx = 1;
 499         opts.index_only = cached;
 500         opts.diff_index_cached = (cached &&
 501                                   !revs->diffopt.flags.find_copies_harder);
 502         opts.merge = 1;
 503         opts.fn = oneway_diff;
 504         opts.unpack_data = revs;
 505         opts.src_index = revs->diffopt.repo->index;
 506         opts.dst_index = NULL;
 507         opts.pathspec = &revs->diffopt.pathspec;
 508         opts.pathspec->recursive = 1;
 509
 510         init_tree_desc(&t, tree->buffer, tree->size);
 511         return unpack_trees(1, &t, &opts);
 512 }
 513
 514 void diff_get_merge_base(const struct rev_info *revs, struct object_id *mb)
 515 {
 516         int i;
 517         struct commit *mb_child[2] = {0};
 518         struct commit_list *merge_bases;
 519
 520         for (i = 0; i < revs->pending.nr; i++) {
 521                 struct object *obj = revs->pending.objects[i].item;
 522                 if (obj->flags)
 523                         die(_("--merge-base does not work with ranges"));
 524                 if (obj->type != OBJ_COMMIT)
 525                         die(_("--merge-base only works with commits"));
 526         }
 527
 528         /*
 529          * This check must go after the for loop above because A...B
 530          * ranges produce three pending commits, resulting in a
 531          * misleading error message.
 532          */
 533         if (revs->pending.nr < 1 || revs->pending.nr > 2)
 534                 BUG("unexpected revs->pending.nr: %d", revs->pending.nr);
 535
 536         for (i = 0; i < revs->pending.nr; i++)
 537                 mb_child[i] = lookup_commit_reference(the_repository, &revs->pending.objects[i].item->oid);
 538         if (revs->pending.nr == 1) {
 539                 struct object_id oid;
 540
 541                 if (get_oid("HEAD", &oid))
 542                         die(_("unable to get HEAD"));
 543
 544                 mb_child[1] = lookup_commit_reference(the_repository, &oid);
 545         }
 546
 547         merge_bases = repo_get_merge_bases(the_repository, mb_child[0], mb_child[1]);
 548         if (!merge_bases)
 549                 die(_("no merge base found"));
 550         if (merge_bases->next)
 551                 die(_("multiple merge bases found"));
 552
 553         oidcpy(mb, &merge_bases->item->object.oid);
 554
 555         free_commit_list(merge_bases);
 556 }
 557
 558 int run_diff_index(struct rev_info *revs, unsigned int option)
 559 {
 560         struct object_array_entry *ent;
 561         int cached = !!(option & DIFF_INDEX_CACHED);
 562         int merge_base = !!(option & DIFF_INDEX_MERGE_BASE);
 563         struct object_id oid;
 564         const char *name;
 565         char merge_base_hex[GIT_MAX_HEXSZ + 1];
 566
 567         if (revs->pending.nr != 1)
 568                 BUG("run_diff_index must be passed exactly one tree");
 569
 570         trace_performance_enter();
 571         ent = revs->pending.objects;
 572
 573         if (merge_base) {
 574                 diff_get_merge_base(revs, &oid);
 575                 name = oid_to_hex_r(merge_base_hex, &oid);
 576         } else {
 577                 oidcpy(&oid, &ent->item->oid);
 578                 name = ent->name;
 579         }
 580
 581         if (diff_cache(revs, &oid, name, cached))
 582                 exit(128);
 583
 584         diff_set_mnemonic_prefix(&revs->diffopt, "c/", cached ? "i/" : "w/");
 585         diffcore_fix_diff_index();
 586         diffcore_std(&revs->diffopt);
 587         diff_flush(&revs->diffopt);
 588         trace_performance_leave("diff-index");
 589         return 0;
 590 }
 591
 592 int do_diff_cache(const struct object_id *tree_oid, struct diff_options *opt)
 593 {
 594         struct rev_info revs;
 595
 596         repo_init_revisions(opt->repo, &revs, NULL);
 597         copy_pathspec(&revs.prune_data, &opt->pathspec);
 598         revs.diffopt = *opt;
 599
 600         if (diff_cache(&revs, tree_oid, NULL, 1))
 601                 exit(128);
 602         return 0;
 603 }
 604
 605 int index_differs_from(struct repository *r,
 606                        const char *def, const struct diff_flags *flags,
 607                        int ita_invisible_in_index)
 608 {
 609         struct rev_info rev;
 610         struct setup_revision_opt opt;
 611
 612         repo_init_revisions(r, &rev, NULL);
 613         memset(&opt, 0, sizeof(opt));
 614         opt.def = def;
 615         setup_revisions(0, NULL, &rev, &opt);
 616         rev.diffopt.flags.quick = 1;
 617         rev.diffopt.flags.exit_with_status = 1;
 618         if (flags)
 619                 diff_flags_or(&rev.diffopt.flags, flags);
 620         rev.diffopt.ita_invisible_in_index = ita_invisible_in_index;
 621         run_diff_index(&rev, 1);
 622         object_array_clear(&rev.pending);
 623         return (rev.diffopt.flags.has_changes != 0);
 624 }
 625
 626 static struct strbuf *idiff_prefix_cb(struct diff_options *opt, void *data)
 627 {
 628         return data;
 629 }
 630
 631 void show_interdiff(const struct object_id *oid1, const struct object_id *oid2,
 632                     int indent, struct diff_options *diffopt)
 633 {
 634         struct diff_options opts;
 635         struct strbuf prefix = STRBUF_INIT;
 636
 637         memcpy(&opts, diffopt, sizeof(opts));
 638         opts.output_format = DIFF_FORMAT_PATCH;
 639         opts.output_prefix = idiff_prefix_cb;
 640         strbuf_addchars(&prefix, ' ', indent);
 641         opts.output_prefix_data = &prefix;
 642         diff_setup_done(&opts);
 643
 644         diff_tree_oid(oid1, oid2, "", &opts);
 645         diffcore_std(&opts);
 646         diff_flush(&opts);
 647
 648         strbuf_release(&prefix);
 649 }