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  * Pre-condition: baselen == 0 || base[baselen-1] == '/'
  89  *
  90  * Return:
  91  *  - 2 for "yes, and all subsequent entries will be"
  92  *  - 1 for yes
  93  *  - zero for no
  94  *  - negative for "no, and no subsequent entries will be either"
  95  */
  96 static int tree_entry_interesting(struct tree_desc *desc, const char *base, int baselen, struct diff_options *opt)
  97 {
  98         const char *path;
  99         const unsigned char *sha1;
 100         unsigned mode;
 101         int i;
 102         int pathlen;
 103         int never_interesting = -1;
 104
 105         if (!opt->pathspec.nr)
 106                 return 2;
 107
 108         sha1 = tree_entry_extract(desc, &path, &mode);
 109
 110         pathlen = tree_entry_len(path, sha1);
 111
 112         for (i = 0; i < opt->pathspec.nr; i++) {
 113                 const struct pathspec_item *item = opt->pathspec.items+i;
 114                 const char *match = item->match;
 115                 int matchlen = item->len;
 116                 int m = -1; /* signals that we haven't called strncmp() */
 117
 118                 if (baselen >= matchlen) {
 119                         /* If it doesn't match, move along... */
 120                         if (strncmp(base, match, matchlen))
 121                                 continue;
 122
 123                         /*
 124                          * If the base is a subdirectory of a path which
 125                          * was specified, all of them are interesting.
 126                          */
 127                         if (!matchlen ||
 128                             base[matchlen] == '/' ||
 129                             match[matchlen - 1] == '/')
 130                                 return 2;
 131
 132                         /* Just a random prefix match */
 133                         continue;
 134                 }
 135
 136                 /* Does the base match? */
 137                 if (strncmp(base, match, baselen))
 138                         continue;
 139
 140                 match += baselen;
 141                 matchlen -= baselen;
 142
 143                 if (never_interesting) {
 144                         /*
 145                          * We have not seen any match that sorts later
 146                          * than the current path.
 147                          */
 148
 149                         /*
 150                          * Does match sort strictly earlier than path
 151                          * with their common parts?
 152                          */
 153                         m = strncmp(match, path,
 154                                     (matchlen < pathlen) ? matchlen : pathlen);
 155                         if (m < 0)
 156                                 continue;
 157
 158                         /*
 159                          * If we come here even once, that means there is at
 160                          * least one pathspec that would sort equal to or
 161                          * later than the path we are currently looking at.
 162                          * In other words, if we have never reached this point
 163                          * after iterating all pathspecs, it means all
 164                          * pathspecs are either outside of base, or inside the
 165                          * base but sorts strictly earlier than the current
 166                          * one.  In either case, they will never match the
 167                          * subsequent entries.  In such a case, we initialized
 168                          * the variable to -1 and that is what will be
 169                          * returned, allowing the caller to terminate early.
 170                          */
 171                         never_interesting = 0;
 172                 }
 173
 174                 if (pathlen > matchlen)
 175                         continue;
 176
 177                 if (matchlen > pathlen) {
 178                         if (match[pathlen] != '/')
 179                                 continue;
 180                         if (!S_ISDIR(mode))
 181                                 continue;
 182                 }
 183
 184                 if (m == -1)
 185                         /*
 186                          * we cheated and did not do strncmp(), so we do
 187                          * that here.
 188                          */
 189                         m = strncmp(match, path, pathlen);
 190
 191                 /*
 192                  * If common part matched earlier then it is a hit,
 193                  * because we rejected the case where path is not a
 194                  * leading directory and is shorter than match.
 195                  */
 196                 if (!m)
 197                         return 1;
 198         }
 199         return never_interesting; /* No matches */
 200 }
 201
 202 /* A whole sub-tree went away or appeared */
 203 static void show_tree(struct diff_options *opt, const char *prefix, struct tree_desc *desc, const char *base, int baselen)
 204 {
 205         int all_interesting = 0;
 206         while (desc->size) {
 207                 int show;
 208
 209                 if (all_interesting)
 210                         show = 1;
 211                 else {
 212                         show = tree_entry_interesting(desc, base, baselen,
 213                                                       opt);
 214                         if (show == 2)
 215                                 all_interesting = 1;
 216                 }
 217                 if (show < 0)
 218                         break;
 219                 if (show)
 220                         show_entry(opt, prefix, desc, base, baselen);
 221                 update_tree_entry(desc);
 222         }
 223 }
 224
 225 /* A file entry went away or appeared */
 226 static void show_entry(struct diff_options *opt, const char *prefix, struct tree_desc *desc,
 227                        const char *base, int baselen)
 228 {
 229         unsigned mode;
 230         const char *path;
 231         const unsigned char *sha1 = tree_entry_extract(desc, &path, &mode);
 232         int pathlen = tree_entry_len(path, sha1);
 233
 234         if (DIFF_OPT_TST(opt, RECURSIVE) && S_ISDIR(mode)) {
 235                 enum object_type type;
 236                 char *newbase = malloc_base(base, baselen, path, pathlen);
 237                 struct tree_desc inner;
 238                 void *tree;
 239                 unsigned long size;
 240
 241                 tree = read_sha1_file(sha1, &type, &size);
 242                 if (!tree || type != OBJ_TREE)
 243                         die("corrupt tree sha %s", sha1_to_hex(sha1));
 244
 245                 if (DIFF_OPT_TST(opt, TREE_IN_RECURSIVE)) {
 246                         newbase[baselen + pathlen] = 0;
 247                         opt->add_remove(opt, *prefix, mode, sha1, newbase, 0);
 248                         newbase[baselen + pathlen] = '/';
 249                 }
 250
 251                 init_tree_desc(&inner, tree, size);
 252                 show_tree(opt, prefix, &inner, newbase, baselen + 1 + pathlen);
 253
 254                 free(tree);
 255                 free(newbase);
 256         } else {
 257                 char *fullname = malloc_fullname(base, baselen, path, pathlen);
 258                 opt->add_remove(opt, prefix[0], mode, sha1, fullname, 0);
 259                 free(fullname);
 260         }
 261 }
 262
 263 static void skip_uninteresting(struct tree_desc *t, const char *base, int baselen, struct diff_options *opt, int *all_interesting)
 264 {
 265         while (t->size) {
 266                 int show = tree_entry_interesting(t, base, baselen, opt);
 267                 if (show == 2)
 268                         *all_interesting = 1;
 269                 if (!show) {
 270                         update_tree_entry(t);
 271                         continue;
 272                 }
 273                 /* Skip it all? */
 274                 if (show < 0)
 275                         t->size = 0;
 276                 return;
 277         }
 278 }
 279
 280 int diff_tree(struct tree_desc *t1, struct tree_desc *t2, const char *base, struct diff_options *opt)
 281 {
 282         int baselen = strlen(base);
 283         int all_t1_interesting = 0;
 284         int all_t2_interesting = 0;
 285
 286         for (;;) {
 287                 if (DIFF_OPT_TST(opt, QUICK) &&
 288                     DIFF_OPT_TST(opt, HAS_CHANGES))
 289                         break;
 290                 if (opt->pathspec.nr) {
 291                         if (!all_t1_interesting)
 292                                 skip_uninteresting(t1, base, baselen, opt,
 293                                                    &all_t1_interesting);
 294                         if (!all_t2_interesting)
 295                                 skip_uninteresting(t2, base, baselen, opt,
 296                                                    &all_t2_interesting);
 297                 }
 298                 if (!t1->size) {
 299                         if (!t2->size)
 300                                 break;
 301                         show_entry(opt, "+", t2, base, baselen);
 302                         update_tree_entry(t2);
 303                         continue;
 304                 }
 305                 if (!t2->size) {
 306                         show_entry(opt, "-", t1, base, baselen);
 307                         update_tree_entry(t1);
 308                         continue;
 309                 }
 310                 switch (compare_tree_entry(t1, t2, base, baselen, opt)) {
 311                 case -1:
 312                         update_tree_entry(t1);
 313                         continue;
 314                 case 0:
 315                         update_tree_entry(t1);
 316                         /* Fallthrough */
 317                 case 1:
 318                         update_tree_entry(t2);
 319                         continue;
 320                 }
 321                 die("git diff-tree: internal error");
 322         }
 323         return 0;
 324 }
 325
 326 /*
 327  * Does it look like the resulting diff might be due to a rename?
 328  *  - single entry
 329  *  - not a valid previous file
 330  */
 331 static inline int diff_might_be_rename(void)
 332 {
 333         return diff_queued_diff.nr == 1 &&
 334                 !DIFF_FILE_VALID(diff_queued_diff.queue[0]->one);
 335 }
 336
 337 static void try_to_follow_renames(struct tree_desc *t1, struct tree_desc *t2, const char *base, struct diff_options *opt)
 338 {
 339         struct diff_options diff_opts;
 340         struct diff_queue_struct *q = &diff_queued_diff;
 341         struct diff_filepair *choice;
 342         const char *paths[1];
 343         int i;
 344
 345         /* Remove the file creation entry from the diff queue, and remember it */
 346         choice = q->queue[0];
 347         q->nr = 0;
 348
 349         diff_setup(&diff_opts);
 350         DIFF_OPT_SET(&diff_opts, RECURSIVE);
 351         DIFF_OPT_SET(&diff_opts, FIND_COPIES_HARDER);
 352         diff_opts.output_format = DIFF_FORMAT_NO_OUTPUT;
 353         diff_opts.single_follow = opt->pathspec.raw[0];
 354         diff_opts.break_opt = opt->break_opt;
 355         paths[0] = NULL;
 356         diff_tree_setup_paths(paths, &diff_opts);
 357         if (diff_setup_done(&diff_opts) < 0)
 358                 die("unable to set up diff options to follow renames");
 359         diff_tree(t1, t2, base, &diff_opts);
 360         diffcore_std(&diff_opts);
 361         diff_tree_release_paths(&diff_opts);
 362
 363         /* Go through the new set of filepairing, and see if we find a more interesting one */
 364         opt->found_follow = 0;
 365         for (i = 0; i < q->nr; i++) {
 366                 struct diff_filepair *p = q->queue[i];
 367
 368                 /*
 369                  * Found a source? Not only do we use that for the new
 370                  * diff_queued_diff, we will also use that as the path in
 371                  * the future!
 372                  */
 373                 if ((p->status == 'R' || p->status == 'C') &&
 374                     !strcmp(p->two->path, opt->pathspec.raw[0])) {
 375                         /* Switch the file-pairs around */
 376                         q->queue[i] = choice;
 377                         choice = p;
 378
 379                         /* Update the path we use from now on.. */
 380                         diff_tree_release_paths(opt);
 381                         opt->pathspec.raw[0] = xstrdup(p->one->path);
 382                         diff_tree_setup_paths(opt->pathspec.raw, opt);
 383
 384                         /*
 385                          * The caller expects us to return a set of vanilla
 386                          * filepairs to let a later call to diffcore_std()
 387                          * it makes to sort the renames out (among other
 388                          * things), but we already have found renames
 389                          * ourselves; signal diffcore_std() not to muck with
 390                          * rename information.
 391                          */
 392                         opt->found_follow = 1;
 393                         break;
 394                 }
 395         }
 396
 397         /*
 398          * Then, discard all the non-relevant file pairs...
 399          */
 400         for (i = 0; i < q->nr; i++) {
 401                 struct diff_filepair *p = q->queue[i];
 402                 diff_free_filepair(p);
 403         }
 404
 405         /*
 406          * .. and re-instate the one we want (which might be either the
 407          * original one, or the rename/copy we found)
 408          */
 409         q->queue[0] = choice;
 410         q->nr = 1;
 411 }
 412
 413 int diff_tree_sha1(const unsigned char *old, const unsigned char *new, const char *base, struct diff_options *opt)
 414 {
 415         void *tree1, *tree2;
 416         struct tree_desc t1, t2;
 417         unsigned long size1, size2;
 418         int retval;
 419
 420         tree1 = read_object_with_reference(old, tree_type, &size1, NULL);
 421         if (!tree1)
 422                 die("unable to read source tree (%s)", sha1_to_hex(old));
 423         tree2 = read_object_with_reference(new, tree_type, &size2, NULL);
 424         if (!tree2)
 425                 die("unable to read destination tree (%s)", sha1_to_hex(new));
 426         init_tree_desc(&t1, tree1, size1);
 427         init_tree_desc(&t2, tree2, size2);
 428         retval = diff_tree(&t1, &t2, base, opt);
 429         if (!*base && DIFF_OPT_TST(opt, FOLLOW_RENAMES) && diff_might_be_rename()) {
 430                 init_tree_desc(&t1, tree1, size1);
 431                 init_tree_desc(&t2, tree2, size2);
 432                 try_to_follow_renames(&t1, &t2, base, opt);
 433         }
 434         free(tree1);
 435         free(tree2);
 436         return retval;
 437 }
 438
 439 int diff_root_tree_sha1(const unsigned char *new, const char *base, struct diff_options *opt)
 440 {
 441         int retval;
 442         void *tree;
 443         unsigned long size;
 444         struct tree_desc empty, real;
 445
 446         tree = read_object_with_reference(new, tree_type, &size, NULL);
 447         if (!tree)
 448                 die("unable to read root tree (%s)", sha1_to_hex(new));
 449         init_tree_desc(&real, tree, size);
 450
 451         init_tree_desc(&empty, "", 0);
 452         retval = diff_tree(&empty, &real, base, opt);
 453         free(tree);
 454         return retval;
 455 }
 456
 457 void diff_tree_release_paths(struct diff_options *opt)
 458 {
 459         free_pathspec(&opt->pathspec);
 460 }
 461
 462 void diff_tree_setup_paths(const char **p, struct diff_options *opt)
 463 {
 464         init_pathspec(&opt->pathspec, p);
 465 }