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, max_size;
  49         unsigned long src_copied, literal_added, src_removed;
  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                 *merge_score_p = (int)MAX_SCORE;
  57                 return 1; /* even their types are different */
  58         }
  59
  60         if (src->sha1_valid && dst->sha1_valid &&
  61             !hashcmp(src->sha1, dst->sha1))
  62                 return 0; /* they are the same */
  63
  64         if (diff_populate_filespec(src, 0) || diff_populate_filespec(dst, 0))
  65                 return 0; /* error but caught downstream */
  66
  67         base_size = ((src->size < dst->size) ? src->size : dst->size);
  68         max_size = ((src->size > dst->size) ? src->size : dst->size);
  69         if (max_size < MINIMUM_BREAK_SIZE)
  70                 return 0; /* we do not break too small filepair */
  71
  72         if (diffcore_count_changes(src, dst,
  73                                    NULL, NULL,
  74                                    0,
  75                                    &src_copied, &literal_added))
  76                 return 0;
  77
  78         /* sanity */
  79         if (src->size < src_copied)
  80                 src_copied = src->size;
  81         if (dst->size < literal_added + src_copied) {
  82                 if (src_copied < dst->size)
  83                         literal_added = dst->size - src_copied;
  84                 else
  85                         literal_added = 0;
  86         }
  87         src_removed = src->size - src_copied;
  88
  89         /* Compute merge-score, which is "how much is removed
  90          * from the source material".  The clean-up stage will
  91          * merge the surviving pair together if the score is
  92          * less than the minimum, after rename/copy runs.
  93          */
  94         *merge_score_p = (int)(src_removed * MAX_SCORE / src->size);
  95         if (*merge_score_p > break_score)
  96                 return 1;
  97
  98         /* Extent of damage, which counts both inserts and
  99          * deletes.
 100          */
 101         delta_size = src_removed + literal_added;
 102         if (delta_size * MAX_SCORE / max_size < break_score)
 103                 return 0;
 104
 105         /* If you removed a lot without adding new material, that is
 106          * not really a rewrite.
 107          */
 108         if ((src->size * break_score < src_removed * MAX_SCORE) &&
 109             (literal_added * 20 < src_removed) &&
 110             (literal_added * 20 < src_copied))
 111                 return 0;
 112
 113         return 1;
 114 }
 115
 116 void diffcore_break(int break_score)
 117 {
 118         struct diff_queue_struct *q = &diff_queued_diff;
 119         struct diff_queue_struct outq;
 120
 121         /* When the filepair has this much edit (insert and delete),
 122          * it is first considered to be a rewrite and broken into a
 123          * create and delete filepair.  This is to help breaking a
 124          * file that had too much new stuff added, possibly from
 125          * moving contents from another file, so that rename/copy can
 126          * match it with the other file.
 127          *
 128          * int break_score; we reuse incoming parameter for this.
 129          */
 130
 131         /* After a pair is broken according to break_score and
 132          * subjected to rename/copy, both of them may survive intact,
 133          * due to lack of suitable rename/copy peer.  Or, the caller
 134          * may be calling us without using rename/copy.  When that
 135          * happens, we merge the broken pieces back into one
 136          * modification together if the pair did not have more than
 137          * this much delete.  For this computation, we do not take
 138          * insert into account at all.  If you start from a 100-line
 139          * file and delete 97 lines of it, it does not matter if you
 140          * add 27 lines to it to make a new 30-line file or if you add
 141          * 997 lines to it to make a 1000-line file.  Either way what
 142          * you did was a rewrite of 97%.  On the other hand, if you
 143          * delete 3 lines, keeping 97 lines intact, it does not matter
 144          * if you add 3 lines to it to make a new 100-line file or if
 145          * you add 903 lines to it to make a new 1000-line file.
 146          * Either way you did a lot of additions and not a rewrite.
 147          * This merge happens to catch the latter case.  A merge_score
 148          * of 80% would be a good default value (a broken pair that
 149          * has score lower than merge_score will be merged back
 150          * together).
 151          */
 152         int merge_score;
 153         int i;
 154
 155         /* See comment on DEFAULT_BREAK_SCORE and
 156          * DEFAULT_MERGE_SCORE in diffcore.h
 157          */
 158         merge_score = (break_score >> 16) & 0xFFFF;
 159         break_score = (break_score & 0xFFFF);
 160
 161         if (!break_score)
 162                 break_score = DEFAULT_BREAK_SCORE;
 163         if (!merge_score)
 164                 merge_score = DEFAULT_MERGE_SCORE;
 165
 166         outq.nr = outq.alloc = 0;
 167         outq.queue = NULL;
 168
 169         for (i = 0; i < q->nr; i++) {
 170                 struct diff_filepair *p = q->queue[i];
 171                 int score;
 172
 173                 /*
 174                  * We deal only with in-place edit of blobs.
 175                  * We do not break anything else.
 176                  */
 177                 if (DIFF_FILE_VALID(p->one) && DIFF_FILE_VALID(p->two) &&
 178                     object_type(p->one->mode) == OBJ_BLOB &&
 179                     object_type(p->two->mode) == OBJ_BLOB &&
 180                     !strcmp(p->one->path, p->two->path)) {
 181                         if (should_break(p->one, p->two,
 182                                          break_score, &score)) {
 183                                 /* Split this into delete and create */
 184                                 struct diff_filespec *null_one, *null_two;
 185                                 struct diff_filepair *dp;
 186
 187                                 /* Set score to 0 for the pair that
 188                                  * needs to be merged back together
 189                                  * should they survive rename/copy.
 190                                  * Also we do not want to break very
 191                                  * small files.
 192                                  */
 193                                 if (score < merge_score)
 194                                         score = 0;
 195
 196                                 /* deletion of one */
 197                                 null_one = alloc_filespec(p->one->path);
 198                                 dp = diff_queue(&outq, p->one, null_one);
 199                                 dp->score = score;
 200                                 dp->broken_pair = 1;
 201
 202                                 /* creation of two */
 203                                 null_two = alloc_filespec(p->two->path);
 204                                 dp = diff_queue(&outq, null_two, p->two);
 205                                 dp->score = score;
 206                                 dp->broken_pair = 1;
 207
 208                                 free(p); /* not diff_free_filepair(), we are
 209                                           * reusing one and two here.
 210                                           */
 211                                 continue;
 212                         }
 213                 }
 214                 diff_q(&outq, p);
 215         }
 216         free(q->queue);
 217         *q = outq;
 218
 219         return;
 220 }
 221
 222 static void merge_broken(struct diff_filepair *p,
 223                          struct diff_filepair *pp,
 224                          struct diff_queue_struct *outq)
 225 {
 226         /* p and pp are broken pairs we want to merge */
 227         struct diff_filepair *c = p, *d = pp, *dp;
 228         if (DIFF_FILE_VALID(p->one)) {
 229                 /* this must be a delete half */
 230                 d = p; c = pp;
 231         }
 232         /* Sanity check */
 233         if (!DIFF_FILE_VALID(d->one))
 234                 die("internal error in merge #1");
 235         if (DIFF_FILE_VALID(d->two))
 236                 die("internal error in merge #2");
 237         if (DIFF_FILE_VALID(c->one))
 238                 die("internal error in merge #3");
 239         if (!DIFF_FILE_VALID(c->two))
 240                 die("internal error in merge #4");
 241
 242         dp = diff_queue(outq, d->one, c->two);
 243         dp->score = p->score;
 244         diff_free_filespec_data(d->two);
 245         diff_free_filespec_data(c->one);
 246         free(d);
 247         free(c);
 248 }
 249
 250 void diffcore_merge_broken(void)
 251 {
 252         struct diff_queue_struct *q = &diff_queued_diff;
 253         struct diff_queue_struct outq;
 254         int i, j;
 255
 256         outq.nr = outq.alloc = 0;
 257         outq.queue = NULL;
 258
 259         for (i = 0; i < q->nr; i++) {
 260                 struct diff_filepair *p = q->queue[i];
 261                 if (!p)
 262                         /* we already merged this with its peer */
 263                         continue;
 264                 else if (p->broken_pair &&
 265                          !strcmp(p->one->path, p->two->path)) {
 266                         /* If the peer also survived rename/copy, then
 267                          * we merge them back together.
 268                          */
 269                         for (j = i + 1; j < q->nr; j++) {
 270                                 struct diff_filepair *pp = q->queue[j];
 271                                 if (pp->broken_pair &&
 272                                     !strcmp(pp->one->path, pp->two->path) &&
 273                                     !strcmp(p->one->path, pp->two->path)) {
 274                                         /* Peer survived.  Merge them */
 275                                         merge_broken(p, pp, &outq);
 276                                         q->queue[j] = NULL;
 277                                         break;
 278                                 }
 279                         }
 280                         if (q->nr <= j)
 281                                 /* The peer did not survive, so we keep
 282                                  * it in the output.
 283                                  */
 284                                 diff_q(&outq, p);
 285                 }
 286                 else
 287                         diff_q(&outq, p);
 288         }
 289         free(q->queue);
 290         *q = outq;
 291
 292         return;
 293 }