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                 return 0; /* leave symlink rename alone */
  57
  58         if (src->sha1_valid && dst->sha1_valid &&
  59             !hashcmp(src->sha1, dst->sha1))
  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         max_size = ((src->size > dst->size) ? src->size : dst->size);
  67         if (max_size < MINIMUM_BREAK_SIZE)
  68                 return 0; /* we do not break too small filepair */
  69
  70         if (diffcore_count_changes(src, dst,
  71                                    NULL, NULL,
  72                                    0,
  73                                    &src_copied, &literal_added))
  74                 return 0;
  75
  76         /* sanity */
  77         if (src->size < src_copied)
  78                 src_copied = src->size;
  79         if (dst->size < literal_added + src_copied) {
  80                 if (src_copied < dst->size)
  81                         literal_added = dst->size - src_copied;
  82                 else
  83                         literal_added = 0;
  84         }
  85         src_removed = src->size - src_copied;
  86
  87         /* Compute merge-score, which is "how much is removed
  88          * from the source material".  The clean-up stage will
  89          * merge the surviving pair together if the score is
  90          * less than the minimum, after rename/copy runs.
  91          */
  92         *merge_score_p = (int)(src_removed * MAX_SCORE / src->size);
  93         if (*merge_score_p > break_score)
  94                 return 1;
  95
  96         /* Extent of damage, which counts both inserts and
  97          * deletes.
  98          */
  99         delta_size = src_removed + literal_added;
 100         if (delta_size * MAX_SCORE / max_size < break_score)
 101                 return 0;
 102
 103         /* If you removed a lot without adding new material, that is
 104          * not really a rewrite.
 105          */
 106         if ((src->size * break_score < src_removed * MAX_SCORE) &&
 107             (literal_added * 20 < src_removed) &&
 108             (literal_added * 20 < src_copied))
 109                 return 0;
 110
 111         return 1;
 112 }
 113
 114 void diffcore_break(int break_score)
 115 {
 116         struct diff_queue_struct *q = &diff_queued_diff;
 117         struct diff_queue_struct outq;
 118
 119         /* When the filepair has this much edit (insert and delete),
 120          * it is first considered to be a rewrite and broken into a
 121          * create and delete filepair.  This is to help breaking a
 122          * file that had too much new stuff added, possibly from
 123          * moving contents from another file, so that rename/copy can
 124          * match it with the other file.
 125          *
 126          * int break_score; we reuse incoming parameter for this.
 127          */
 128
 129         /* After a pair is broken according to break_score and
 130          * subjected to rename/copy, both of them may survive intact,
 131          * due to lack of suitable rename/copy peer.  Or, the caller
 132          * may be calling us without using rename/copy.  When that
 133          * happens, we merge the broken pieces back into one
 134          * modification together if the pair did not have more than
 135          * this much delete.  For this computation, we do not take
 136          * insert into account at all.  If you start from a 100-line
 137          * file and delete 97 lines of it, it does not matter if you
 138          * add 27 lines to it to make a new 30-line file or if you add
 139          * 997 lines to it to make a 1000-line file.  Either way what
 140          * you did was a rewrite of 97%.  On the other hand, if you
 141          * delete 3 lines, keeping 97 lines intact, it does not matter
 142          * if you add 3 lines to it to make a new 100-line file or if
 143          * you add 903 lines to it to make a new 1000-line file.
 144          * Either way you did a lot of additions and not a rewrite.
 145          * This merge happens to catch the latter case.  A merge_score
 146          * of 80% would be a good default value (a broken pair that
 147          * has score lower than merge_score will be merged back
 148          * together).
 149          */
 150         int merge_score;
 151         int i;
 152
 153         /* See comment on DEFAULT_BREAK_SCORE and
 154          * DEFAULT_MERGE_SCORE in diffcore.h
 155          */
 156         merge_score = (break_score >> 16) & 0xFFFF;
 157         break_score = (break_score & 0xFFFF);
 158
 159         if (!break_score)
 160                 break_score = DEFAULT_BREAK_SCORE;
 161         if (!merge_score)
 162                 merge_score = DEFAULT_MERGE_SCORE;
 163
 164         outq.nr = outq.alloc = 0;
 165         outq.queue = NULL;
 166
 167         for (i = 0; i < q->nr; i++) {
 168                 struct diff_filepair *p = q->queue[i];
 169                 int score;
 170
 171                 /* We deal only with in-place edit of non directory.
 172                  * We do not break anything else.
 173                  */
 174                 if (DIFF_FILE_VALID(p->one) && DIFF_FILE_VALID(p->two) &&
 175                     !S_ISDIR(p->one->mode) && !S_ISDIR(p->two->mode) &&
 176                     !strcmp(p->one->path, p->two->path)) {
 177                         if (should_break(p->one, p->two,
 178                                          break_score, &score)) {
 179                                 /* Split this into delete and create */
 180                                 struct diff_filespec *null_one, *null_two;
 181                                 struct diff_filepair *dp;
 182
 183                                 /* Set score to 0 for the pair that
 184                                  * needs to be merged back together
 185                                  * should they survive rename/copy.
 186                                  * Also we do not want to break very
 187                                  * small files.
 188                                  */
 189                                 if (score < merge_score)
 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, *dp;
 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         dp = diff_queue(outq, d->one, c->two);
 239         dp->score = p->score;
 240         diff_free_filespec_data(d->two);
 241         diff_free_filespec_data(c->one);
 242         free(d);
 243         free(c);
 244 }
 245
 246 void diffcore_merge_broken(void)
 247 {
 248         struct diff_queue_struct *q = &diff_queued_diff;
 249         struct diff_queue_struct outq;
 250         int i, j;
 251
 252         outq.nr = outq.alloc = 0;
 253         outq.queue = NULL;
 254
 255         for (i = 0; i < q->nr; i++) {
 256                 struct diff_filepair *p = q->queue[i];
 257                 if (!p)
 258                         /* we already merged this with its peer */
 259                         continue;
 260                 else if (p->broken_pair &&
 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                                     !strcmp(pp->one->path, pp->two->path) &&
 269                                     !strcmp(p->one->path, pp->two->path)) {
 270                                         /* Peer survived.  Merge them */
 271                                         merge_broken(p, pp, &outq);
 272                                         q->queue[j] = NULL;
 273                                         break;
 274                                 }
 275                         }
 276                         if (q->nr <= j)
 277                                 /* The peer did not survive, so we keep
 278                                  * it in the output.
 279                                  */
 280                                 diff_q(&outq, p);
 281                 }
 282                 else
 283                         diff_q(&outq, p);
 284         }
 285         free(q->queue);
 286         *q = outq;
 287
 288         return;
 289 }