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