diffcore-break.c

   1 /*
   2  * Copyright (C) 2005 Junio C Hamano
   3  */
   4 #include "git-compat-util.h"
   5 #include "diff.h"
   6 #include "diffcore.h"
   7 #include "hash.h"
   8 #include "object.h"
   9 #include "promisor-remote.h"
  10
  11 static int should_break(struct repository *r,
  12                         struct diff_filespec *src,
  13                         struct diff_filespec *dst,
  14                         int break_score,
  15                         int *merge_score_p)
  16 {
  17         /* dst is recorded as a modification of src.  Are they so
  18          * different that we are better off recording this as a pair
  19          * of delete and create?
  20          *
  21          * There are two criteria used in this algorithm.  For the
  22          * purposes of helping later rename/copy, we take both delete
  23          * and insert into account and estimate the amount of "edit".
  24          * If the edit is very large, we break this pair so that
  25          * rename/copy can pick the pieces up to match with other
  26          * files.
  27          *
  28          * On the other hand, we would want to ignore inserts for the
  29          * pure "complete rewrite" detection.  As long as most of the
  30          * existing contents were removed from the file, it is a
  31          * complete rewrite, and if sizable chunk from the original
  32          * still remains in the result, it is not a rewrite.  It does
  33          * not matter how much or how little new material is added to
  34          * the file.
  35          *
  36          * The score we leave for such a broken filepair uses the
  37          * latter definition so that later clean-up stage can find the
  38          * pieces that should not have been broken according to the
  39          * latter definition after rename/copy runs, and merge the
  40          * broken pair that have a score lower than given criteria
  41          * back together.  The break operation itself happens
  42          * according to the former definition.
  43          *
  44          * The minimum_edit parameter tells us when to break (the
  45          * amount of "edit" required for us to consider breaking the
  46          * pair).  We leave the amount of deletion in *merge_score_p
  47          * when we return.
  48          *
  49          * The value we return is 1 if we want the pair to be broken,
  50          * or 0 if we do not.
  51          */
  52         unsigned long delta_size, max_size;
  53         unsigned long src_copied, literal_added, src_removed;
  54
  55         struct diff_populate_filespec_options options = { 0 };
  56
  57         *merge_score_p = 0; /* assume no deletion --- "do not break"
  58                              * is the default.
  59                              */
  60
  61         if (S_ISREG(src->mode) != S_ISREG(dst->mode)) {
  62                 *merge_score_p = (int)MAX_SCORE;
  63                 return 1; /* even their types are different */
  64         }
  65
  66         if (src->oid_valid && dst->oid_valid &&
  67             oideq(&src->oid, &dst->oid))
  68                 return 0; /* they are the same */
  69
  70         if (r == the_repository && repo_has_promisor_remote(the_repository)) {
  71                 options.missing_object_cb = diff_queued_diff_prefetch;
  72                 options.missing_object_data = r;
  73         }
  74
  75         if (diff_populate_filespec(r, src, &options) ||
  76             diff_populate_filespec(r, dst, &options))
  77                 return 0; /* error but caught downstream */
  78
  79         max_size = ((src->size > dst->size) ? src->size : dst->size);
  80         if (max_size < MINIMUM_BREAK_SIZE)
  81                 return 0; /* we do not break too small filepair */
  82
  83         if (!src->size)
  84                 return 0; /* we do not let empty files get renamed */
  85
  86         if (diffcore_count_changes(r, src, dst,
  87                                    &src->cnt_data, &dst->cnt_data,
  88                                    &src_copied, &literal_added))
  89                 return 0;
  90
  91         /* sanity */
  92         if (src->size < src_copied)
  93                 src_copied = src->size;
  94         if (dst->size < literal_added + src_copied) {
  95                 if (src_copied < dst->size)
  96                         literal_added = dst->size - src_copied;
  97                 else
  98                         literal_added = 0;
  99         }
 100         src_removed = src->size - src_copied;
 101
 102         /* Compute merge-score, which is "how much is removed
 103          * from the source material".  The clean-up stage will
 104          * merge the surviving pair together if the score is
 105          * less than the minimum, after rename/copy runs.
 106          */
 107         *merge_score_p = (int)(src_removed * MAX_SCORE / src->size);
 108         if (*merge_score_p > break_score)
 109                 return 1;
 110
 111         /* Extent of damage, which counts both inserts and
 112          * deletes.
 113          */
 114         delta_size = src_removed + literal_added;
 115         if (delta_size * MAX_SCORE / max_size < break_score)
 116                 return 0;
 117
 118         /* If you removed a lot without adding new material, that is
 119          * not really a rewrite.
 120          */
 121         if ((src->size * break_score < src_removed * MAX_SCORE) &&
 122             (literal_added * 20 < src_removed) &&
 123             (literal_added * 20 < src_copied))
 124                 return 0;
 125
 126         return 1;
 127 }
 128
 129 void diffcore_break(struct repository *r, int break_score)
 130 {
 131         struct diff_queue_struct *q = &diff_queued_diff;
 132         struct diff_queue_struct outq;
 133
 134         /* When the filepair has this much edit (insert and delete),
 135          * it is first considered to be a rewrite and broken into a
 136          * create and delete filepair.  This is to help breaking a
 137          * file that had too much new stuff added, possibly from
 138          * moving contents from another file, so that rename/copy can
 139          * match it with the other file.
 140          *
 141          * int break_score; we reuse incoming parameter for this.
 142          */
 143
 144         /* After a pair is broken according to break_score and
 145          * subjected to rename/copy, both of them may survive intact,
 146          * due to lack of suitable rename/copy peer.  Or, the caller
 147          * may be calling us without using rename/copy.  When that
 148          * happens, we merge the broken pieces back into one
 149          * modification together if the pair did not have more than
 150          * this much delete.  For this computation, we do not take
 151          * insert into account at all.  If you start from a 100-line
 152          * file and delete 97 lines of it, it does not matter if you
 153          * add 27 lines to it to make a new 30-line file or if you add
 154          * 997 lines to it to make a 1000-line file.  Either way what
 155          * you did was a rewrite of 97%.  On the other hand, if you
 156          * delete 3 lines, keeping 97 lines intact, it does not matter
 157          * if you add 3 lines to it to make a new 100-line file or if
 158          * you add 903 lines to it to make a new 1000-line file.
 159          * Either way you did a lot of additions and not a rewrite.
 160          * This merge happens to catch the latter case.  A merge_score
 161          * of 80% would be a good default value (a broken pair that
 162          * has score lower than merge_score will be merged back
 163          * together).
 164          */
 165         int merge_score;
 166         int i;
 167
 168         /* See comment on DEFAULT_BREAK_SCORE and
 169          * DEFAULT_MERGE_SCORE in diffcore.h
 170          */
 171         merge_score = (break_score >> 16) & 0xFFFF;
 172         break_score = (break_score & 0xFFFF);
 173
 174         if (!break_score)
 175                 break_score = DEFAULT_BREAK_SCORE;
 176         if (!merge_score)
 177                 merge_score = DEFAULT_MERGE_SCORE;
 178
 179         DIFF_QUEUE_CLEAR(&outq);
 180
 181         for (i = 0; i < q->nr; i++) {
 182                 struct diff_filepair *p = q->queue[i];
 183                 int score;
 184
 185                 /*
 186                  * We deal only with in-place edit of blobs.
 187                  * We do not break anything else.
 188                  */
 189                 if (DIFF_FILE_VALID(p->one) && DIFF_FILE_VALID(p->two) &&
 190                     object_type(p->one->mode) == OBJ_BLOB &&
 191                     object_type(p->two->mode) == OBJ_BLOB &&
 192                     !strcmp(p->one->path, p->two->path)) {
 193                         if (should_break(r, p->one, p->two,
 194                                          break_score, &score)) {
 195                                 /* Split this into delete and create */
 196                                 struct diff_filespec *null_one, *null_two;
 197                                 struct diff_filepair *dp;
 198
 199                                 /* Set score to 0 for the pair that
 200                                  * needs to be merged back together
 201                                  * should they survive rename/copy.
 202                                  * Also we do not want to break very
 203                                  * small files.
 204                                  */
 205                                 if (score < merge_score)
 206                                         score = 0;
 207
 208                                 /* deletion of one */
 209                                 null_one = alloc_filespec(p->one->path);
 210                                 dp = diff_queue(&outq, p->one, null_one);
 211                                 dp->score = score;
 212                                 dp->broken_pair = 1;
 213
 214                                 /* creation of two */
 215                                 null_two = alloc_filespec(p->two->path);
 216                                 dp = diff_queue(&outq, null_two, p->two);
 217                                 dp->score = score;
 218                                 dp->broken_pair = 1;
 219
 220                                 diff_free_filespec_blob(p->one);
 221                                 diff_free_filespec_blob(p->two);
 222                                 free(p); /* not diff_free_filepair(), we are
 223                                           * reusing one and two here.
 224                                           */
 225                                 continue;
 226                         }
 227                 }
 228                 diff_free_filespec_data(p->one);
 229                 diff_free_filespec_data(p->two);
 230                 diff_q(&outq, p);
 231         }
 232         free(q->queue);
 233         *q = outq;
 234
 235         return;
 236 }
 237
 238 static void merge_broken(struct diff_filepair *p,
 239                          struct diff_filepair *pp,
 240                          struct diff_queue_struct *outq)
 241 {
 242         /* p and pp are broken pairs we want to merge */
 243         struct diff_filepair *c = p, *d = pp, *dp;
 244         if (DIFF_FILE_VALID(p->one)) {
 245                 /* this must be a delete half */
 246                 d = p; c = pp;
 247         }
 248         /* Sanity check */
 249         if (!DIFF_FILE_VALID(d->one))
 250                 die("internal error in merge #1");
 251         if (DIFF_FILE_VALID(d->two))
 252                 die("internal error in merge #2");
 253         if (DIFF_FILE_VALID(c->one))
 254                 die("internal error in merge #3");
 255         if (!DIFF_FILE_VALID(c->two))
 256                 die("internal error in merge #4");
 257
 258         dp = diff_queue(outq, d->one, c->two);
 259         dp->score = p->score;
 260         /*
 261          * We will be one extra user of the same src side of the
 262          * broken pair, if it was used as the rename source for other
 263          * paths elsewhere.  Increment to mark that the path stays
 264          * in the resulting tree.
 265          */
 266         d->one->rename_used++;
 267         diff_free_filespec_data(d->two);
 268         diff_free_filespec_data(c->one);
 269         free(d);
 270         free(c);
 271 }
 272
 273 void diffcore_merge_broken(void)
 274 {
 275         struct diff_queue_struct *q = &diff_queued_diff;
 276         struct diff_queue_struct outq;
 277         int i, j;
 278
 279         DIFF_QUEUE_CLEAR(&outq);
 280
 281         for (i = 0; i < q->nr; i++) {
 282                 struct diff_filepair *p = q->queue[i];
 283                 if (!p)
 284                         /* we already merged this with its peer */
 285                         continue;
 286                 else if (p->broken_pair &&
 287                          !strcmp(p->one->path, p->two->path)) {
 288                         /* If the peer also survived rename/copy, then
 289                          * we merge them back together.
 290                          */
 291                         for (j = i + 1; j < q->nr; j++) {
 292                                 struct diff_filepair *pp = q->queue[j];
 293                                 if (pp->broken_pair &&
 294                                     !strcmp(pp->one->path, pp->two->path) &&
 295                                     !strcmp(p->one->path, pp->two->path)) {
 296                                         /* Peer survived.  Merge them */
 297                                         merge_broken(p, pp, &outq);
 298                                         q->queue[j] = NULL;
 299                                         goto next;
 300                                 }
 301                         }
 302                         /* The peer did not survive, so we keep
 303                          * it in the output.
 304                          */
 305                         diff_q(&outq, p);
 306                 }
 307                 else
 308                         diff_q(&outq, p);
 309 next:;
 310         }
 311         free(q->queue);
 312         *q = outq;
 313
 314         return;
 315 }