diffcore-break.c

   1 /*
   2  * Copyright (C) 2005 Junio C Hamano
   3  */
   4 #include "cache.h"
   5 #include "diff.h"
   6 #include "diffcore.h"
   7
   8 static int should_break(struct diff_filespec *src,
   9                         struct diff_filespec *dst,
  10                         int break_score,
  11                         int *merge_score_p)
  12 {
  13         /* dst is recorded as a modification of src.  Are they so
  14          * different that we are better off recording this as a pair
  15          * of delete and create?
  16          *
  17          * There are two criteria used in this algorithm.  For the
  18          * purposes of helping later rename/copy, we take both delete
  19          * and insert into account and estimate the amount of "edit".
  20          * If the edit is very large, we break this pair so that
  21          * rename/copy can pick the pieces up to match with other
  22          * files.
  23          *
  24          * On the other hand, we would want to ignore inserts for the
  25          * pure "complete rewrite" detection.  As long as most of the
  26          * existing contents were removed from the file, it is a
  27          * complete rewrite, and if sizable chunk from the original
  28          * still remains in the result, it is not a rewrite.  It does
  29          * not matter how much or how little new material is added to
  30          * the file.
  31          *
  32          * The score we leave for such a broken filepair uses the
  33          * latter definition so that later clean-up stage can find the
  34          * pieces that should not have been broken according to the
  35          * latter definition after rename/copy runs, and merge the
  36          * broken pair that have a score lower than given criteria
  37          * back together.  The break operation itself happens
  38          * according to the former definition.
  39          *
  40          * The minimum_edit parameter tells us when to break (the
  41          * amount of "edit" required for us to consider breaking the
  42          * pair).  We leave the amount of deletion in *merge_score_p
  43          * when we return.
  44          *
  45          * The value we return is 1 if we want the pair to be broken,
  46          * or 0 if we do not.
  47          */
  48         unsigned long delta_size, base_size, src_copied, literal_added;
  49         int to_break = 0;
  50
  51         *merge_score_p = 0; /* assume no deletion --- "do not break"
  52                              * is the default.
  53                              */
  54
  55         if (!S_ISREG(src->mode) || !S_ISREG(dst->mode))
  56                 return 0; /* leave symlink rename alone */
  57
  58         if (src->sha1_valid && dst->sha1_valid &&
  59             !memcmp(src->sha1, dst->sha1, 20))
  60                 return 0; /* they are the same */
  61
  62         if (diff_populate_filespec(src, 0) || diff_populate_filespec(dst, 0))
  63                 return 0; /* error but caught downstream */
  64
  65         base_size = ((src->size < dst->size) ? src->size : dst->size);
  66         if (base_size < MINIMUM_BREAK_SIZE)
  67                 return 0; /* we do not break too small filepair */
  68
  69         if (diffcore_count_changes(src->data, src->size,
  70                                    dst->data, dst->size,
  71                                    0,
  72                                    &src_copied, &literal_added))
  73                 return 0;
  74
  75         /* Compute merge-score, which is "how much is removed
  76          * from the source material".  The clean-up stage will
  77          * merge the surviving pair together if the score is
  78          * less than the minimum, after rename/copy runs.
  79          */
  80         if (src->size <= src_copied)
  81                 ; /* all copied, nothing removed */
  82         else {
  83                 delta_size = src->size - src_copied;
  84                 *merge_score_p = delta_size * MAX_SCORE / src->size;
  85         }
  86
  87         /* Extent of damage, which counts both inserts and
  88          * deletes.
  89          */
  90         if (src->size + literal_added <= src_copied)
  91                 delta_size = 0; /* avoid wrapping around */
  92         else
  93                 delta_size = (src->size - src_copied) + literal_added;
  94
  95         /* We break if the edit exceeds the minimum.
  96          * i.e. (break_score / MAX_SCORE < delta_size / base_size)
  97          */
  98         if (break_score * base_size < delta_size * MAX_SCORE)
  99                 to_break = 1;
 100
 101         return to_break;
 102 }
 103
 104 void diffcore_break(int break_score)
 105 {
 106         struct diff_queue_struct *q = &diff_queued_diff;
 107         struct diff_queue_struct outq;
 108
 109         /* When the filepair has this much edit (insert and delete),
 110          * it is first considered to be a rewrite and broken into a
 111          * create and delete filepair.  This is to help breaking a
 112          * file that had too much new stuff added, possibly from
 113          * moving contents from another file, so that rename/copy can
 114          * match it with the other file.
 115          *
 116          * int break_score; we reuse incoming parameter for this.
 117          */
 118
 119         /* After a pair is broken according to break_score and
 120          * subjected to rename/copy, both of them may survive intact,
 121          * due to lack of suitable rename/copy peer.  Or, the caller
 122          * may be calling us without using rename/copy.  When that
 123          * happens, we merge the broken pieces back into one
 124          * modification together if the pair did not have more than
 125          * this much delete.  For this computation, we do not take
 126          * insert into account at all.  If you start from a 100-line
 127          * file and delete 97 lines of it, it does not matter if you
 128          * add 27 lines to it to make a new 30-line file or if you add
 129          * 997 lines to it to make a 1000-line file.  Either way what
 130          * you did was a rewrite of 97%.  On the other hand, if you
 131          * delete 3 lines, keeping 97 lines intact, it does not matter
 132          * if you add 3 lines to it to make a new 100-line file or if
 133          * you add 903 lines to it to make a new 1000-line file.
 134          * Either way you did a lot of additions and not a rewrite.
 135          * This merge happens to catch the latter case.  A merge_score
 136          * of 80% would be a good default value (a broken pair that
 137          * has score lower than merge_score will be merged back
 138          * together).
 139          */
 140         int merge_score;
 141         int i;
 142
 143         /* See comment on DEFAULT_BREAK_SCORE and
 144          * DEFAULT_MERGE_SCORE in diffcore.h
 145          */
 146         merge_score = (break_score >> 16) & 0xFFFF;
 147         break_score = (break_score & 0xFFFF);
 148
 149         if (!break_score)
 150                 break_score = DEFAULT_BREAK_SCORE;
 151         if (!merge_score)
 152                 merge_score = DEFAULT_MERGE_SCORE;
 153
 154         outq.nr = outq.alloc = 0;
 155         outq.queue = NULL;
 156
 157         for (i = 0; i < q->nr; i++) {
 158                 struct diff_filepair *p = q->queue[i];
 159                 int score;
 160
 161                 /* We deal only with in-place edit of non directory.
 162                  * We do not break anything else.
 163                  */
 164                 if (DIFF_FILE_VALID(p->one) && DIFF_FILE_VALID(p->two) &&
 165                     !S_ISDIR(p->one->mode) && !S_ISDIR(p->two->mode) &&
 166                     !strcmp(p->one->path, p->two->path)) {
 167                         if (should_break(p->one, p->two,
 168                                          break_score, &score)) {
 169                                 /* Split this into delete and create */
 170                                 struct diff_filespec *null_one, *null_two;
 171                                 struct diff_filepair *dp;
 172
 173                                 /* Set score to 0 for the pair that
 174                                  * needs to be merged back together
 175                                  * should they survive rename/copy.
 176                                  * Also we do not want to break very
 177                                  * small files.
 178                                  */
 179                                 if (score < merge_score)
 180                                         score = 0;
 181
 182                                 /* deletion of one */
 183                                 null_one = alloc_filespec(p->one->path);
 184                                 dp = diff_queue(&outq, p->one, null_one);
 185                                 dp->score = score;
 186                                 dp->broken_pair = 1;
 187
 188                                 /* creation of two */
 189                                 null_two = alloc_filespec(p->two->path);
 190                                 dp = diff_queue(&outq, null_two, p->two);
 191                                 dp->score = score;
 192                                 dp->broken_pair = 1;
 193
 194                                 free(p); /* not diff_free_filepair(), we are
 195                                           * reusing one and two here.
 196                                           */
 197                                 continue;
 198                         }
 199                 }
 200                 diff_q(&outq, p);
 201         }
 202         free(q->queue);
 203         *q = outq;
 204
 205         return;
 206 }
 207
 208 static void merge_broken(struct diff_filepair *p,
 209                          struct diff_filepair *pp,
 210                          struct diff_queue_struct *outq)
 211 {
 212         /* p and pp are broken pairs we want to merge */
 213         struct diff_filepair *c = p, *d = pp, *dp;
 214         if (DIFF_FILE_VALID(p->one)) {
 215                 /* this must be a delete half */
 216                 d = p; c = pp;
 217         }
 218         /* Sanity check */
 219         if (!DIFF_FILE_VALID(d->one))
 220                 die("internal error in merge #1");
 221         if (DIFF_FILE_VALID(d->two))
 222                 die("internal error in merge #2");
 223         if (DIFF_FILE_VALID(c->one))
 224                 die("internal error in merge #3");
 225         if (!DIFF_FILE_VALID(c->two))
 226                 die("internal error in merge #4");
 227
 228         dp = diff_queue(outq, d->one, c->two);
 229         dp->score = p->score;
 230         diff_free_filespec_data(d->two);
 231         diff_free_filespec_data(c->one);
 232         free(d);
 233         free(c);
 234 }
 235
 236 void diffcore_merge_broken(void)
 237 {
 238         struct diff_queue_struct *q = &diff_queued_diff;
 239         struct diff_queue_struct outq;
 240         int i, j;
 241
 242         outq.nr = outq.alloc = 0;
 243         outq.queue = NULL;
 244
 245         for (i = 0; i < q->nr; i++) {
 246                 struct diff_filepair *p = q->queue[i];
 247                 if (!p)
 248                         /* we already merged this with its peer */
 249                         continue;
 250                 else if (p->broken_pair &&
 251                          !strcmp(p->one->path, p->two->path)) {
 252                         /* If the peer also survived rename/copy, then
 253                          * we merge them back together.
 254                          */
 255                         for (j = i + 1; j < q->nr; j++) {
 256                                 struct diff_filepair *pp = q->queue[j];
 257                                 if (pp->broken_pair &&
 258                                     !strcmp(pp->one->path, pp->two->path) &&
 259                                     !strcmp(p->one->path, pp->two->path)) {
 260                                         /* Peer survived.  Merge them */
 261                                         merge_broken(p, pp, &outq);
 262                                         q->queue[j] = NULL;
 263                                         break;
 264                                 }
 265                         }
 266                         if (q->nr <= j)
 267                                 /* The peer did not survive, so we keep
 268                                  * it in the output.
 269                                  */
 270                                 diff_q(&outq, p);
 271                 }
 272                 else
 273                         diff_q(&outq, p);
 274         }
 275         free(q->queue);
 276         *q = outq;
 277
 278         return;
 279 }