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