]>
Commit | Line | Data |
---|---|---|
f345b0a0 JH |
1 | /* |
2 | * Copyright (C) 2005 Junio C Hamano | |
3 | */ | |
08c46a49 | 4 | #include "git-compat-util.h" |
f345b0a0 JH |
5 | #include "diff.h" |
6 | #include "diffcore.h" | |
df6e8744 | 7 | #include "hash.h" |
08c46a49 | 8 | #include "object.h" |
95acf11a | 9 | #include "promisor-remote.h" |
f345b0a0 | 10 | |
b78ea5fc NTND |
11 | static int should_break(struct repository *r, |
12 | struct diff_filespec *src, | |
eeaa4603 JH |
13 | struct diff_filespec *dst, |
14 | int break_score, | |
15 | int *merge_score_p) | |
f345b0a0 JH |
16 | { |
17 | /* dst is recorded as a modification of src. Are they so | |
18 | * different that we are better off recording this as a pair | |
eeaa4603 | 19 | * of delete and create? |
f345b0a0 | 20 | * |
eeaa4603 JH |
21 | * There are two criteria used in this algorithm. For the |
22 | * purposes of helping later rename/copy, we take both delete | |
23 | * and insert into account and estimate the amount of "edit". | |
24 | * If the edit is very large, we break this pair so that | |
25 | * rename/copy can pick the pieces up to match with other | |
26 | * files. | |
27 | * | |
28 | * On the other hand, we would want to ignore inserts for the | |
29 | * pure "complete rewrite" detection. As long as most of the | |
30 | * existing contents were removed from the file, it is a | |
31 | * complete rewrite, and if sizable chunk from the original | |
32 | * still remains in the result, it is not a rewrite. It does | |
33 | * not matter how much or how little new material is added to | |
34 | * the file. | |
35 | * | |
36 | * The score we leave for such a broken filepair uses the | |
37 | * latter definition so that later clean-up stage can find the | |
38 | * pieces that should not have been broken according to the | |
39 | * latter definition after rename/copy runs, and merge the | |
40 | * broken pair that have a score lower than given criteria | |
41 | * back together. The break operation itself happens | |
42 | * according to the former definition. | |
43 | * | |
44 | * The minimum_edit parameter tells us when to break (the | |
45 | * amount of "edit" required for us to consider breaking the | |
46 | * pair). We leave the amount of deletion in *merge_score_p | |
47 | * when we return. | |
48 | * | |
49 | * The value we return is 1 if we want the pair to be broken, | |
50 | * or 0 if we do not. | |
f345b0a0 | 51 | */ |
eb3a9dd3 | 52 | unsigned long delta_size, max_size; |
6dd4b66f | 53 | unsigned long src_copied, literal_added, src_removed; |
eeaa4603 | 54 | |
95acf11a JT |
55 | struct diff_populate_filespec_options options = { 0 }; |
56 | ||
eeaa4603 JH |
57 | *merge_score_p = 0; /* assume no deletion --- "do not break" |
58 | * is the default. | |
59 | */ | |
f345b0a0 | 60 | |
b45563a2 JH |
61 | if (S_ISREG(src->mode) != S_ISREG(dst->mode)) { |
62 | *merge_score_p = (int)MAX_SCORE; | |
63 | return 1; /* even their types are different */ | |
64 | } | |
f345b0a0 | 65 | |
41c9560e | 66 | if (src->oid_valid && dst->oid_valid && |
4a7e27e9 | 67 | oideq(&src->oid, &dst->oid)) |
aeecd23a JH |
68 | return 0; /* they are the same */ |
69 | ||
a5183d76 | 70 | if (r == the_repository && repo_has_promisor_remote(the_repository)) { |
95acf11a JT |
71 | options.missing_object_cb = diff_queued_diff_prefetch; |
72 | options.missing_object_data = r; | |
73 | } | |
74 | ||
75 | if (diff_populate_filespec(r, src, &options) || | |
76 | diff_populate_filespec(r, dst, &options)) | |
f345b0a0 JH |
77 | return 0; /* error but caught downstream */ |
78 | ||
6dd4b66f LT |
79 | max_size = ((src->size > dst->size) ? src->size : dst->size); |
80 | if (max_size < MINIMUM_BREAK_SIZE) | |
0532a5e4 | 81 | return 0; /* we do not break too small filepair */ |
f345b0a0 | 82 | |
e7b00c57 JK |
83 | if (!src->size) |
84 | return 0; /* we do not let empty files get renamed */ | |
85 | ||
b78ea5fc | 86 | if (diffcore_count_changes(r, src, dst, |
8282de94 | 87 | &src->cnt_data, &dst->cnt_data, |
65416758 JH |
88 | &src_copied, &literal_added)) |
89 | return 0; | |
355e76a4 | 90 | |
4d0f39ce JH |
91 | /* sanity */ |
92 | if (src->size < src_copied) | |
93 | src_copied = src->size; | |
94 | if (dst->size < literal_added + src_copied) { | |
95 | if (src_copied < dst->size) | |
96 | literal_added = dst->size - src_copied; | |
97 | else | |
98 | literal_added = 0; | |
99 | } | |
100 | src_removed = src->size - src_copied; | |
101 | ||
eeaa4603 JH |
102 | /* Compute merge-score, which is "how much is removed |
103 | * from the source material". The clean-up stage will | |
104 | * merge the surviving pair together if the score is | |
105 | * less than the minimum, after rename/copy runs. | |
106 | */ | |
dc49cd76 | 107 | *merge_score_p = (int)(src_removed * MAX_SCORE / src->size); |
6dd4b66f LT |
108 | if (*merge_score_p > break_score) |
109 | return 1; | |
4d0f39ce | 110 | |
eeaa4603 JH |
111 | /* Extent of damage, which counts both inserts and |
112 | * deletes. | |
113 | */ | |
4d0f39ce | 114 | delta_size = src_removed + literal_added; |
6dd4b66f | 115 | if (delta_size * MAX_SCORE / max_size < break_score) |
4d0f39ce JH |
116 | return 0; |
117 | ||
118 | /* If you removed a lot without adding new material, that is | |
119 | * not really a rewrite. | |
eeaa4603 | 120 | */ |
4d0f39ce JH |
121 | if ((src->size * break_score < src_removed * MAX_SCORE) && |
122 | (literal_added * 20 < src_removed) && | |
123 | (literal_added * 20 < src_copied)) | |
124 | return 0; | |
f345b0a0 | 125 | |
4d0f39ce | 126 | return 1; |
f345b0a0 JH |
127 | } |
128 | ||
b78ea5fc | 129 | void diffcore_break(struct repository *r, int break_score) |
f345b0a0 JH |
130 | { |
131 | struct diff_queue_struct *q = &diff_queued_diff; | |
132 | struct diff_queue_struct outq; | |
eeaa4603 JH |
133 | |
134 | /* When the filepair has this much edit (insert and delete), | |
135 | * it is first considered to be a rewrite and broken into a | |
136 | * create and delete filepair. This is to help breaking a | |
137 | * file that had too much new stuff added, possibly from | |
138 | * moving contents from another file, so that rename/copy can | |
139 | * match it with the other file. | |
140 | * | |
141 | * int break_score; we reuse incoming parameter for this. | |
142 | */ | |
143 | ||
144 | /* After a pair is broken according to break_score and | |
145 | * subjected to rename/copy, both of them may survive intact, | |
146 | * due to lack of suitable rename/copy peer. Or, the caller | |
147 | * may be calling us without using rename/copy. When that | |
148 | * happens, we merge the broken pieces back into one | |
149 | * modification together if the pair did not have more than | |
150 | * this much delete. For this computation, we do not take | |
151 | * insert into account at all. If you start from a 100-line | |
152 | * file and delete 97 lines of it, it does not matter if you | |
153 | * add 27 lines to it to make a new 30-line file or if you add | |
154 | * 997 lines to it to make a 1000-line file. Either way what | |
155 | * you did was a rewrite of 97%. On the other hand, if you | |
156 | * delete 3 lines, keeping 97 lines intact, it does not matter | |
157 | * if you add 3 lines to it to make a new 100-line file or if | |
158 | * you add 903 lines to it to make a new 1000-line file. | |
159 | * Either way you did a lot of additions and not a rewrite. | |
160 | * This merge happens to catch the latter case. A merge_score | |
161 | * of 80% would be a good default value (a broken pair that | |
162 | * has score lower than merge_score will be merged back | |
163 | * together). | |
164 | */ | |
165 | int merge_score; | |
f345b0a0 JH |
166 | int i; |
167 | ||
eeaa4603 JH |
168 | /* See comment on DEFAULT_BREAK_SCORE and |
169 | * DEFAULT_MERGE_SCORE in diffcore.h | |
170 | */ | |
171 | merge_score = (break_score >> 16) & 0xFFFF; | |
172 | break_score = (break_score & 0xFFFF); | |
173 | ||
174 | if (!break_score) | |
175 | break_score = DEFAULT_BREAK_SCORE; | |
176 | if (!merge_score) | |
177 | merge_score = DEFAULT_MERGE_SCORE; | |
f345b0a0 | 178 | |
9ca5df90 | 179 | DIFF_QUEUE_CLEAR(&outq); |
f345b0a0 JH |
180 | |
181 | for (i = 0; i < q->nr; i++) { | |
182 | struct diff_filepair *p = q->queue[i]; | |
183 | int score; | |
184 | ||
b45563a2 JH |
185 | /* |
186 | * We deal only with in-place edit of blobs. | |
f345b0a0 JH |
187 | * We do not break anything else. |
188 | */ | |
189 | if (DIFF_FILE_VALID(p->one) && DIFF_FILE_VALID(p->two) && | |
b45563a2 JH |
190 | object_type(p->one->mode) == OBJ_BLOB && |
191 | object_type(p->two->mode) == OBJ_BLOB && | |
f345b0a0 | 192 | !strcmp(p->one->path, p->two->path)) { |
b78ea5fc | 193 | if (should_break(r, p->one, p->two, |
0532a5e4 | 194 | break_score, &score)) { |
f345b0a0 JH |
195 | /* Split this into delete and create */ |
196 | struct diff_filespec *null_one, *null_two; | |
197 | struct diff_filepair *dp; | |
198 | ||
eeaa4603 JH |
199 | /* Set score to 0 for the pair that |
200 | * needs to be merged back together | |
201 | * should they survive rename/copy. | |
202 | * Also we do not want to break very | |
203 | * small files. | |
204 | */ | |
f78c79c5 | 205 | if (score < merge_score) |
eeaa4603 JH |
206 | score = 0; |
207 | ||
f345b0a0 JH |
208 | /* deletion of one */ |
209 | null_one = alloc_filespec(p->one->path); | |
210 | dp = diff_queue(&outq, p->one, null_one); | |
211 | dp->score = score; | |
212 | dp->broken_pair = 1; | |
213 | ||
214 | /* creation of two */ | |
215 | null_two = alloc_filespec(p->two->path); | |
216 | dp = diff_queue(&outq, null_two, p->two); | |
217 | dp->score = score; | |
218 | dp->broken_pair = 1; | |
219 | ||
8282de94 JK |
220 | diff_free_filespec_blob(p->one); |
221 | diff_free_filespec_blob(p->two); | |
f345b0a0 JH |
222 | free(p); /* not diff_free_filepair(), we are |
223 | * reusing one and two here. | |
224 | */ | |
225 | continue; | |
226 | } | |
227 | } | |
f4f19fb6 JK |
228 | diff_free_filespec_data(p->one); |
229 | diff_free_filespec_data(p->two); | |
f345b0a0 JH |
230 | diff_q(&outq, p); |
231 | } | |
232 | free(q->queue); | |
233 | *q = outq; | |
234 | ||
235 | return; | |
236 | } | |
eeaa4603 JH |
237 | |
238 | static void merge_broken(struct diff_filepair *p, | |
239 | struct diff_filepair *pp, | |
240 | struct diff_queue_struct *outq) | |
241 | { | |
242 | /* p and pp are broken pairs we want to merge */ | |
366175ef | 243 | struct diff_filepair *c = p, *d = pp, *dp; |
eeaa4603 JH |
244 | if (DIFF_FILE_VALID(p->one)) { |
245 | /* this must be a delete half */ | |
246 | d = p; c = pp; | |
247 | } | |
248 | /* Sanity check */ | |
249 | if (!DIFF_FILE_VALID(d->one)) | |
250 | die("internal error in merge #1"); | |
251 | if (DIFF_FILE_VALID(d->two)) | |
252 | die("internal error in merge #2"); | |
253 | if (DIFF_FILE_VALID(c->one)) | |
254 | die("internal error in merge #3"); | |
255 | if (!DIFF_FILE_VALID(c->two)) | |
256 | die("internal error in merge #4"); | |
257 | ||
366175ef JH |
258 | dp = diff_queue(outq, d->one, c->two); |
259 | dp->score = p->score; | |
6936b585 JH |
260 | /* |
261 | * We will be one extra user of the same src side of the | |
262 | * broken pair, if it was used as the rename source for other | |
263 | * paths elsewhere. Increment to mark that the path stays | |
264 | * in the resulting tree. | |
265 | */ | |
266 | d->one->rename_used++; | |
19397b45 JH |
267 | diff_free_filespec_data(d->two); |
268 | diff_free_filespec_data(c->one); | |
eeaa4603 JH |
269 | free(d); |
270 | free(c); | |
271 | } | |
272 | ||
273 | void diffcore_merge_broken(void) | |
274 | { | |
275 | struct diff_queue_struct *q = &diff_queued_diff; | |
276 | struct diff_queue_struct outq; | |
277 | int i, j; | |
278 | ||
9ca5df90 | 279 | DIFF_QUEUE_CLEAR(&outq); |
eeaa4603 JH |
280 | |
281 | for (i = 0; i < q->nr; i++) { | |
282 | struct diff_filepair *p = q->queue[i]; | |
283 | if (!p) | |
284 | /* we already merged this with its peer */ | |
285 | continue; | |
286 | else if (p->broken_pair && | |
eeaa4603 JH |
287 | !strcmp(p->one->path, p->two->path)) { |
288 | /* If the peer also survived rename/copy, then | |
289 | * we merge them back together. | |
290 | */ | |
291 | for (j = i + 1; j < q->nr; j++) { | |
292 | struct diff_filepair *pp = q->queue[j]; | |
293 | if (pp->broken_pair && | |
eeaa4603 JH |
294 | !strcmp(pp->one->path, pp->two->path) && |
295 | !strcmp(p->one->path, pp->two->path)) { | |
296 | /* Peer survived. Merge them */ | |
297 | merge_broken(p, pp, &outq); | |
298 | q->queue[j] = NULL; | |
baed6bbb | 299 | goto next; |
eeaa4603 JH |
300 | } |
301 | } | |
baed6bbb AH |
302 | /* The peer did not survive, so we keep |
303 | * it in the output. | |
304 | */ | |
305 | diff_q(&outq, p); | |
eeaa4603 JH |
306 | } |
307 | else | |
308 | diff_q(&outq, p); | |
baed6bbb | 309 | next:; |
eeaa4603 JH |
310 | } |
311 | free(q->queue); | |
312 | *q = outq; | |
313 | ||
314 | return; | |
315 | } |