tree-diff.c

   1 /*
   2  * Helper functions for tree diff generation
   3  */
   4 #include "cache.h"
   5 #include "diff.h"
   6 #include "diffcore.h"
   7 #include "tree.h"
   8
   9 static char *malloc_base(const char *base, int baselen, const char *path, int pathlen)
  10 {
  11         char *newbase = xmalloc(baselen + pathlen + 2);
  12         memcpy(newbase, base, baselen);
  13         memcpy(newbase + baselen, path, pathlen);
  14         memcpy(newbase + baselen + pathlen, "/", 2);
  15         return newbase;
  16 }
  17
  18 static char *malloc_fullname(const char *base, int baselen, const char *path, int pathlen)
  19 {
  20         char *fullname = xmalloc(baselen + pathlen + 1);
  21         memcpy(fullname, base, baselen);
  22         memcpy(fullname + baselen, path, pathlen);
  23         fullname[baselen + pathlen] = 0;
  24         return fullname;
  25 }
  26
  27 static void show_entry(struct diff_options *opt, const char *prefix, struct tree_desc *desc,
  28                        const char *base, int baselen);
  29
  30 static int compare_tree_entry(struct tree_desc *t1, struct tree_desc *t2, const char *base, int baselen, struct diff_options *opt)
  31 {
  32         unsigned mode1, mode2;
  33         const char *path1, *path2;
  34         const unsigned char *sha1, *sha2;
  35         int cmp, pathlen1, pathlen2;
  36         char *fullname;
  37
  38         sha1 = tree_entry_extract(t1, &path1, &mode1);
  39         sha2 = tree_entry_extract(t2, &path2, &mode2);
  40
  41         pathlen1 = tree_entry_len(path1, sha1);
  42         pathlen2 = tree_entry_len(path2, sha2);
  43         cmp = base_name_compare(path1, pathlen1, mode1, path2, pathlen2, mode2);
  44         if (cmp < 0) {
  45                 show_entry(opt, "-", t1, base, baselen);
  46                 return -1;
  47         }
  48         if (cmp > 0) {
  49                 show_entry(opt, "+", t2, base, baselen);
  50                 return 1;
  51         }
  52         if (!DIFF_OPT_TST(opt, FIND_COPIES_HARDER) && !hashcmp(sha1, sha2) && mode1 == mode2)
  53                 return 0;
  54
  55         /*
  56          * If the filemode has changed to/from a directory from/to a regular
  57          * file, we need to consider it a remove and an add.
  58          */
  59         if (S_ISDIR(mode1) != S_ISDIR(mode2)) {
  60                 show_entry(opt, "-", t1, base, baselen);
  61                 show_entry(opt, "+", t2, base, baselen);
  62                 return 0;
  63         }
  64
  65         if (DIFF_OPT_TST(opt, RECURSIVE) && S_ISDIR(mode1)) {
  66                 int retval;
  67                 char *newbase = malloc_base(base, baselen, path1, pathlen1);
  68                 if (DIFF_OPT_TST(opt, TREE_IN_RECURSIVE)) {
  69                         newbase[baselen + pathlen1] = 0;
  70                         opt->change(opt, mode1, mode2,
  71                                     sha1, sha2, newbase, 0, 0);
  72                         newbase[baselen + pathlen1] = '/';
  73                 }
  74                 retval = diff_tree_sha1(sha1, sha2, newbase, opt);
  75                 free(newbase);
  76                 return retval;
  77         }
  78
  79         fullname = malloc_fullname(base, baselen, path1, pathlen1);
  80         opt->change(opt, mode1, mode2, sha1, sha2, fullname, 0, 0);
  81         free(fullname);
  82         return 0;
  83 }
  84
  85 /*
  86  * Is a tree entry interesting given the pathspec we have?
  87  *
  88  * Return:
  89  *  - 2 for "yes, and all subsequent entries will be"
  90  *  - 1 for yes
  91  *  - zero for no
  92  *  - negative for "no, and no subsequent entries will be either"
  93  */
  94 static int tree_entry_interesting(struct tree_desc *desc, const char *base, int baselen, struct diff_options *opt)
  95 {
  96         const char *path;
  97         const unsigned char *sha1;
  98         unsigned mode;
  99         int i;
 100         int pathlen;
 101         int never_interesting = -1;
 102
 103         if (!opt->nr_paths)
 104                 return 1;
 105
 106         sha1 = tree_entry_extract(desc, &path, &mode);
 107
 108         pathlen = tree_entry_len(path, sha1);
 109
 110         for (i = 0; i < opt->nr_paths; i++) {
 111                 const char *match = opt->paths[i];
 112                 int matchlen = opt->pathlens[i];
 113                 int m = -1; /* signals that we haven't called strncmp() */
 114
 115                 if (baselen >= matchlen) {
 116                         /* If it doesn't match, move along... */
 117                         if (strncmp(base, match, matchlen))
 118                                 continue;
 119
 120                         /*
 121                          * If the base is a subdirectory of a path which
 122                          * was specified, all of them are interesting.
 123                          */
 124                         if (!matchlen ||
 125                             base[matchlen] == '/' ||
 126                             match[matchlen - 1] == '/')
 127                                 return 2;
 128
 129                         /* Just a random prefix match */
 130                         continue;
 131                 }
 132
 133                 /* Does the base match? */
 134                 if (strncmp(base, match, baselen))
 135                         continue;
 136
 137                 match += baselen;
 138                 matchlen -= baselen;
 139
 140                 if (never_interesting) {
 141                         /*
 142                          * We have not seen any match that sorts later
 143                          * than the current path.
 144                          */
 145
 146                         /*
 147                          * Does match sort strictly earlier than path
 148                          * with their common parts?
 149                          */
 150                         m = strncmp(match, path,
 151                                     (matchlen < pathlen) ? matchlen : pathlen);
 152                         if (m < 0)
 153                                 continue;
 154
 155                         /*
 156                          * If we come here even once, that means there is at
 157                          * least one pathspec that would sort equal to or
 158                          * later than the path we are currently looking at.
 159                          * In other words, if we have never reached this point
 160                          * after iterating all pathspecs, it means all
 161                          * pathspecs are either outside of base, or inside the
 162                          * base but sorts strictly earlier than the current
 163                          * one.  In either case, they will never match the
 164                          * subsequent entries.  In such a case, we initialized
 165                          * the variable to -1 and that is what will be
 166                          * returned, allowing the caller to terminate early.
 167                          */
 168                         never_interesting = 0;
 169                 }
 170
 171                 if (pathlen > matchlen)
 172                         continue;
 173
 174                 if (matchlen > pathlen) {
 175                         if (match[pathlen] != '/')
 176                                 continue;
 177                         if (!S_ISDIR(mode))
 178                                 continue;
 179                 }
 180
 181                 if (m == -1)
 182                         /*
 183                          * we cheated and did not do strncmp(), so we do
 184                          * that here.
 185                          */
 186                         m = strncmp(match, path, pathlen);
 187
 188                 /*
 189                  * If common part matched earlier then it is a hit,
 190                  * because we rejected the case where path is not a
 191                  * leading directory and is shorter than match.
 192                  */
 193                 if (!m)
 194                         return 1;
 195         }
 196         return never_interesting; /* No matches */
 197 }
 198
 199 /* A whole sub-tree went away or appeared */
 200 static void show_tree(struct diff_options *opt, const char *prefix, struct tree_desc *desc, const char *base, int baselen)
 201 {
 202         int all_interesting = 0;
 203         while (desc->size) {
 204                 int show;
 205
 206                 if (all_interesting)
 207                         show = 1;
 208                 else {
 209                         show = tree_entry_interesting(desc, base, baselen,
 210                                                       opt);
 211                         if (show == 2)
 212                                 all_interesting = 1;
 213                 }
 214                 if (show < 0)
 215                         break;
 216                 if (show)
 217                         show_entry(opt, prefix, desc, base, baselen);
 218                 update_tree_entry(desc);
 219         }
 220 }
 221
 222 /* A file entry went away or appeared */
 223 static void show_entry(struct diff_options *opt, const char *prefix, struct tree_desc *desc,
 224                        const char *base, int baselen)
 225 {
 226         unsigned mode;
 227         const char *path;
 228         const unsigned char *sha1 = tree_entry_extract(desc, &path, &mode);
 229         int pathlen = tree_entry_len(path, sha1);
 230
 231         if (DIFF_OPT_TST(opt, RECURSIVE) && S_ISDIR(mode)) {
 232                 enum object_type type;
 233                 char *newbase = malloc_base(base, baselen, path, pathlen);
 234                 struct tree_desc inner;
 235                 void *tree;
 236                 unsigned long size;
 237
 238                 tree = read_sha1_file(sha1, &type, &size);
 239                 if (!tree || type != OBJ_TREE)
 240                         die("corrupt tree sha %s", sha1_to_hex(sha1));
 241
 242                 if (DIFF_OPT_TST(opt, TREE_IN_RECURSIVE)) {
 243                         newbase[baselen + pathlen] = 0;
 244                         opt->add_remove(opt, *prefix, mode, sha1, newbase, 0);
 245                         newbase[baselen + pathlen] = '/';
 246                 }
 247
 248                 init_tree_desc(&inner, tree, size);
 249                 show_tree(opt, prefix, &inner, newbase, baselen + 1 + pathlen);
 250
 251                 free(tree);
 252                 free(newbase);
 253         } else {
 254                 char *fullname = malloc_fullname(base, baselen, path, pathlen);
 255                 opt->add_remove(opt, prefix[0], mode, sha1, fullname, 0);
 256                 free(fullname);
 257         }
 258 }
 259
 260 static void skip_uninteresting(struct tree_desc *t, const char *base, int baselen, struct diff_options *opt)
 261 {
 262         int all_interesting = 0;
 263         while (t->size) {
 264                 int show;
 265
 266                 if (all_interesting)
 267                         show = 1;
 268                 else {
 269                         show = tree_entry_interesting(t, base, baselen, opt);
 270                         if (show == 2)
 271                                 all_interesting = 1;
 272                 }
 273                 if (!show) {
 274                         update_tree_entry(t);
 275                         continue;
 276                 }
 277                 /* Skip it all? */
 278                 if (show < 0)
 279                         t->size = 0;
 280                 return;
 281         }
 282 }
 283
 284 int diff_tree(struct tree_desc *t1, struct tree_desc *t2, const char *base, struct diff_options *opt)
 285 {
 286         int baselen = strlen(base);
 287
 288         for (;;) {
 289                 if (DIFF_OPT_TST(opt, QUICK) &&
 290                     DIFF_OPT_TST(opt, HAS_CHANGES))
 291                         break;
 292                 if (opt->nr_paths) {
 293                         skip_uninteresting(t1, base, baselen, opt);
 294                         skip_uninteresting(t2, base, baselen, opt);
 295                 }
 296                 if (!t1->size) {
 297                         if (!t2->size)
 298                                 break;
 299                         show_entry(opt, "+", t2, base, baselen);
 300                         update_tree_entry(t2);
 301                         continue;
 302                 }
 303                 if (!t2->size) {
 304                         show_entry(opt, "-", t1, base, baselen);
 305                         update_tree_entry(t1);
 306                         continue;
 307                 }
 308                 switch (compare_tree_entry(t1, t2, base, baselen, opt)) {
 309                 case -1:
 310                         update_tree_entry(t1);
 311                         continue;
 312                 case 0:
 313                         update_tree_entry(t1);
 314                         /* Fallthrough */
 315                 case 1:
 316                         update_tree_entry(t2);
 317                         continue;
 318                 }
 319                 die("git diff-tree: internal error");
 320         }
 321         return 0;
 322 }
 323
 324 /*
 325  * Does it look like the resulting diff might be due to a rename?
 326  *  - single entry
 327  *  - not a valid previous file
 328  */
 329 static inline int diff_might_be_rename(void)
 330 {
 331         return diff_queued_diff.nr == 1 &&
 332                 !DIFF_FILE_VALID(diff_queued_diff.queue[0]->one);
 333 }
 334
 335 static void try_to_follow_renames(struct tree_desc *t1, struct tree_desc *t2, const char *base, struct diff_options *opt)
 336 {
 337         struct diff_options diff_opts;
 338         struct diff_queue_struct *q = &diff_queued_diff;
 339         struct diff_filepair *choice;
 340         const char *paths[1];
 341         int i;
 342
 343         /* Remove the file creation entry from the diff queue, and remember it */
 344         choice = q->queue[0];
 345         q->nr = 0;
 346
 347         diff_setup(&diff_opts);
 348         DIFF_OPT_SET(&diff_opts, RECURSIVE);
 349         DIFF_OPT_SET(&diff_opts, FIND_COPIES_HARDER);
 350         diff_opts.output_format = DIFF_FORMAT_NO_OUTPUT;
 351         diff_opts.single_follow = opt->paths[0];
 352         diff_opts.break_opt = opt->break_opt;
 353         paths[0] = NULL;
 354         diff_tree_setup_paths(paths, &diff_opts);
 355         if (diff_setup_done(&diff_opts) < 0)
 356                 die("unable to set up diff options to follow renames");
 357         diff_tree(t1, t2, base, &diff_opts);
 358         diffcore_std(&diff_opts);
 359         diff_tree_release_paths(&diff_opts);
 360
 361         /* Go through the new set of filepairing, and see if we find a more interesting one */
 362         opt->found_follow = 0;
 363         for (i = 0; i < q->nr; i++) {
 364                 struct diff_filepair *p = q->queue[i];
 365
 366                 /*
 367                  * Found a source? Not only do we use that for the new
 368                  * diff_queued_diff, we will also use that as the path in
 369                  * the future!
 370                  */
 371                 if ((p->status == 'R' || p->status == 'C') && !strcmp(p->two->path, opt->paths[0])) {
 372                         /* Switch the file-pairs around */
 373                         q->queue[i] = choice;
 374                         choice = p;
 375
 376                         /* Update the path we use from now on.. */
 377                         diff_tree_release_paths(opt);
 378                         opt->paths[0] = xstrdup(p->one->path);
 379                         diff_tree_setup_paths(opt->paths, opt);
 380
 381                         /*
 382                          * The caller expects us to return a set of vanilla
 383                          * filepairs to let a later call to diffcore_std()
 384                          * it makes to sort the renames out (among other
 385                          * things), but we already have found renames
 386                          * ourselves; signal diffcore_std() not to muck with
 387                          * rename information.
 388                          */
 389                         opt->found_follow = 1;
 390                         break;
 391                 }
 392         }
 393
 394         /*
 395          * Then, discard all the non-relevant file pairs...
 396          */
 397         for (i = 0; i < q->nr; i++) {
 398                 struct diff_filepair *p = q->queue[i];
 399                 diff_free_filepair(p);
 400         }
 401
 402         /*
 403          * .. and re-instate the one we want (which might be either the
 404          * original one, or the rename/copy we found)
 405          */
 406         q->queue[0] = choice;
 407         q->nr = 1;
 408 }
 409
 410 int diff_tree_sha1(const unsigned char *old, const unsigned char *new, const char *base, struct diff_options *opt)
 411 {
 412         void *tree1, *tree2;
 413         struct tree_desc t1, t2;
 414         unsigned long size1, size2;
 415         int retval;
 416
 417         tree1 = read_object_with_reference(old, tree_type, &size1, NULL);
 418         if (!tree1)
 419                 die("unable to read source tree (%s)", sha1_to_hex(old));
 420         tree2 = read_object_with_reference(new, tree_type, &size2, NULL);
 421         if (!tree2)
 422                 die("unable to read destination tree (%s)", sha1_to_hex(new));
 423         init_tree_desc(&t1, tree1, size1);
 424         init_tree_desc(&t2, tree2, size2);
 425         retval = diff_tree(&t1, &t2, base, opt);
 426         if (!*base && DIFF_OPT_TST(opt, FOLLOW_RENAMES) && diff_might_be_rename()) {
 427                 init_tree_desc(&t1, tree1, size1);
 428                 init_tree_desc(&t2, tree2, size2);
 429                 try_to_follow_renames(&t1, &t2, base, opt);
 430         }
 431         free(tree1);
 432         free(tree2);
 433         return retval;
 434 }
 435
 436 int diff_root_tree_sha1(const unsigned char *new, const char *base, struct diff_options *opt)
 437 {
 438         int retval;
 439         void *tree;
 440         unsigned long size;
 441         struct tree_desc empty, real;
 442
 443         tree = read_object_with_reference(new, tree_type, &size, NULL);
 444         if (!tree)
 445                 die("unable to read root tree (%s)", sha1_to_hex(new));
 446         init_tree_desc(&real, tree, size);
 447
 448         init_tree_desc(&empty, "", 0);
 449         retval = diff_tree(&empty, &real, base, opt);
 450         free(tree);
 451         return retval;
 452 }
 453
 454 static int count_paths(const char **paths)
 455 {
 456         int i = 0;
 457         while (*paths++)
 458                 i++;
 459         return i;
 460 }
 461
 462 void diff_tree_release_paths(struct diff_options *opt)
 463 {
 464         free(opt->pathlens);
 465 }
 466
 467 void diff_tree_setup_paths(const char **p, struct diff_options *opt)
 468 {
 469         opt->nr_paths = 0;
 470         opt->pathlens = NULL;
 471         opt->paths = NULL;
 472
 473         if (p) {
 474                 int i;
 475
 476                 opt->paths = p;
 477                 opt->nr_paths = count_paths(p);
 478                 if (opt->nr_paths == 0) {
 479                         opt->pathlens = NULL;
 480                         return;
 481                 }
 482                 opt->pathlens = xmalloc(opt->nr_paths * sizeof(int));
 483                 for (i=0; i < opt->nr_paths; i++)
 484                         opt->pathlens[i] = strlen(p[i]);
 485         }
 486 }