diffcore-break.c

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