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