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