Show More
@@ -1,1017 +1,1014 b'' | |||
|
1 | 1 | # copies.py - copy detection for Mercurial |
|
2 | 2 | # |
|
3 | 3 | # Copyright 2008 Matt Mackall <mpm@selenic.com> |
|
4 | 4 | # |
|
5 | 5 | # This software may be used and distributed according to the terms of the |
|
6 | 6 | # GNU General Public License version 2 or any later version. |
|
7 | 7 | |
|
8 | 8 | from __future__ import absolute_import |
|
9 | 9 | |
|
10 | 10 | import collections |
|
11 | 11 | import os |
|
12 | 12 | |
|
13 | 13 | from .i18n import _ |
|
14 | 14 | |
|
15 | 15 | |
|
16 | 16 | from .revlogutils.flagutil import REVIDX_SIDEDATA |
|
17 | 17 | |
|
18 | 18 | from . import ( |
|
19 | 19 | error, |
|
20 | 20 | match as matchmod, |
|
21 | 21 | node, |
|
22 | 22 | pathutil, |
|
23 | 23 | pycompat, |
|
24 | 24 | util, |
|
25 | 25 | ) |
|
26 | 26 | |
|
27 | 27 | from .revlogutils import sidedata as sidedatamod |
|
28 | 28 | |
|
29 | 29 | from .utils import stringutil |
|
30 | 30 | |
|
31 | 31 | |
|
32 | 32 | def _filter(src, dst, t): |
|
33 | 33 | """filters out invalid copies after chaining""" |
|
34 | 34 | |
|
35 | 35 | # When _chain()'ing copies in 'a' (from 'src' via some other commit 'mid') |
|
36 | 36 | # with copies in 'b' (from 'mid' to 'dst'), we can get the different cases |
|
37 | 37 | # in the following table (not including trivial cases). For example, case 2 |
|
38 | 38 | # is where a file existed in 'src' and remained under that name in 'mid' and |
|
39 | 39 | # then was renamed between 'mid' and 'dst'. |
|
40 | 40 | # |
|
41 | 41 | # case src mid dst result |
|
42 | 42 | # 1 x y - - |
|
43 | 43 | # 2 x y y x->y |
|
44 | 44 | # 3 x y x - |
|
45 | 45 | # 4 x y z x->z |
|
46 | 46 | # 5 - x y - |
|
47 | 47 | # 6 x x y x->y |
|
48 | 48 | # |
|
49 | 49 | # _chain() takes care of chaining the copies in 'a' and 'b', but it |
|
50 | 50 | # cannot tell the difference between cases 1 and 2, between 3 and 4, or |
|
51 | 51 | # between 5 and 6, so it includes all cases in its result. |
|
52 | 52 | # Cases 1, 3, and 5 are then removed by _filter(). |
|
53 | 53 | |
|
54 | 54 | for k, v in list(t.items()): |
|
55 | 55 | # remove copies from files that didn't exist |
|
56 | 56 | if v not in src: |
|
57 | 57 | del t[k] |
|
58 | 58 | # remove criss-crossed copies |
|
59 | 59 | elif k in src and v in dst: |
|
60 | 60 | del t[k] |
|
61 | 61 | # remove copies to files that were then removed |
|
62 | 62 | elif k not in dst: |
|
63 | 63 | del t[k] |
|
64 | 64 | |
|
65 | 65 | |
|
66 | 66 | def _chain(a, b): |
|
67 | 67 | """chain two sets of copies 'a' and 'b'""" |
|
68 | 68 | t = a.copy() |
|
69 | 69 | for k, v in pycompat.iteritems(b): |
|
70 | if v in t: | |
|
71 | t[k] = t[v] | |
|
72 | else: | |
|
73 | t[k] = v | |
|
70 | t[k] = t.get(v, v) | |
|
74 | 71 | return t |
|
75 | 72 | |
|
76 | 73 | |
|
77 | 74 | def _tracefile(fctx, am, basemf): |
|
78 | 75 | """return file context that is the ancestor of fctx present in ancestor |
|
79 | 76 | manifest am |
|
80 | 77 | |
|
81 | 78 | Note: we used to try and stop after a given limit, however checking if that |
|
82 | 79 | limit is reached turned out to be very expensive. we are better off |
|
83 | 80 | disabling that feature.""" |
|
84 | 81 | |
|
85 | 82 | for f in fctx.ancestors(): |
|
86 | 83 | path = f.path() |
|
87 | 84 | if am.get(path, None) == f.filenode(): |
|
88 | 85 | return path |
|
89 | 86 | if basemf and basemf.get(path, None) == f.filenode(): |
|
90 | 87 | return path |
|
91 | 88 | |
|
92 | 89 | |
|
93 | 90 | def _dirstatecopies(repo, match=None): |
|
94 | 91 | ds = repo.dirstate |
|
95 | 92 | c = ds.copies().copy() |
|
96 | 93 | for k in list(c): |
|
97 | 94 | if ds[k] not in b'anm' or (match and not match(k)): |
|
98 | 95 | del c[k] |
|
99 | 96 | return c |
|
100 | 97 | |
|
101 | 98 | |
|
102 | 99 | def _computeforwardmissing(a, b, match=None): |
|
103 | 100 | """Computes which files are in b but not a. |
|
104 | 101 | This is its own function so extensions can easily wrap this call to see what |
|
105 | 102 | files _forwardcopies is about to process. |
|
106 | 103 | """ |
|
107 | 104 | ma = a.manifest() |
|
108 | 105 | mb = b.manifest() |
|
109 | 106 | return mb.filesnotin(ma, match=match) |
|
110 | 107 | |
|
111 | 108 | |
|
112 | 109 | def usechangesetcentricalgo(repo): |
|
113 | 110 | """Checks if we should use changeset-centric copy algorithms""" |
|
114 | 111 | if repo.filecopiesmode == b'changeset-sidedata': |
|
115 | 112 | return True |
|
116 | 113 | readfrom = repo.ui.config(b'experimental', b'copies.read-from') |
|
117 | 114 | changesetsource = (b'changeset-only', b'compatibility') |
|
118 | 115 | return readfrom in changesetsource |
|
119 | 116 | |
|
120 | 117 | |
|
121 | 118 | def _committedforwardcopies(a, b, base, match): |
|
122 | 119 | """Like _forwardcopies(), but b.rev() cannot be None (working copy)""" |
|
123 | 120 | # files might have to be traced back to the fctx parent of the last |
|
124 | 121 | # one-side-only changeset, but not further back than that |
|
125 | 122 | repo = a._repo |
|
126 | 123 | |
|
127 | 124 | if usechangesetcentricalgo(repo): |
|
128 | 125 | return _changesetforwardcopies(a, b, match) |
|
129 | 126 | |
|
130 | 127 | debug = repo.ui.debugflag and repo.ui.configbool(b'devel', b'debug.copies') |
|
131 | 128 | dbg = repo.ui.debug |
|
132 | 129 | if debug: |
|
133 | 130 | dbg(b'debug.copies: looking into rename from %s to %s\n' % (a, b)) |
|
134 | 131 | am = a.manifest() |
|
135 | 132 | basemf = None if base is None else base.manifest() |
|
136 | 133 | |
|
137 | 134 | # find where new files came from |
|
138 | 135 | # we currently don't try to find where old files went, too expensive |
|
139 | 136 | # this means we can miss a case like 'hg rm b; hg cp a b' |
|
140 | 137 | cm = {} |
|
141 | 138 | |
|
142 | 139 | # Computing the forward missing is quite expensive on large manifests, since |
|
143 | 140 | # it compares the entire manifests. We can optimize it in the common use |
|
144 | 141 | # case of computing what copies are in a commit versus its parent (like |
|
145 | 142 | # during a rebase or histedit). Note, we exclude merge commits from this |
|
146 | 143 | # optimization, since the ctx.files() for a merge commit is not correct for |
|
147 | 144 | # this comparison. |
|
148 | 145 | forwardmissingmatch = match |
|
149 | 146 | if b.p1() == a and b.p2().node() == node.nullid: |
|
150 | 147 | filesmatcher = matchmod.exact(b.files()) |
|
151 | 148 | forwardmissingmatch = matchmod.intersectmatchers(match, filesmatcher) |
|
152 | 149 | missing = _computeforwardmissing(a, b, match=forwardmissingmatch) |
|
153 | 150 | |
|
154 | 151 | ancestrycontext = a._repo.changelog.ancestors([b.rev()], inclusive=True) |
|
155 | 152 | |
|
156 | 153 | if debug: |
|
157 | 154 | dbg(b'debug.copies: missing files to search: %d\n' % len(missing)) |
|
158 | 155 | |
|
159 | 156 | for f in sorted(missing): |
|
160 | 157 | if debug: |
|
161 | 158 | dbg(b'debug.copies: tracing file: %s\n' % f) |
|
162 | 159 | fctx = b[f] |
|
163 | 160 | fctx._ancestrycontext = ancestrycontext |
|
164 | 161 | |
|
165 | 162 | if debug: |
|
166 | 163 | start = util.timer() |
|
167 | 164 | opath = _tracefile(fctx, am, basemf) |
|
168 | 165 | if opath: |
|
169 | 166 | if debug: |
|
170 | 167 | dbg(b'debug.copies: rename of: %s\n' % opath) |
|
171 | 168 | cm[f] = opath |
|
172 | 169 | if debug: |
|
173 | 170 | dbg( |
|
174 | 171 | b'debug.copies: time: %f seconds\n' |
|
175 | 172 | % (util.timer() - start) |
|
176 | 173 | ) |
|
177 | 174 | return cm |
|
178 | 175 | |
|
179 | 176 | |
|
180 | 177 | def _revinfogetter(repo): |
|
181 | 178 | """return a function that return multiple data given a <rev>"i |
|
182 | 179 | |
|
183 | 180 | * p1: revision number of first parent |
|
184 | 181 | * p2: revision number of first parent |
|
185 | 182 | * p1copies: mapping of copies from p1 |
|
186 | 183 | * p2copies: mapping of copies from p2 |
|
187 | 184 | * removed: a list of removed files |
|
188 | 185 | """ |
|
189 | 186 | cl = repo.changelog |
|
190 | 187 | parents = cl.parentrevs |
|
191 | 188 | |
|
192 | 189 | if repo.filecopiesmode == b'changeset-sidedata': |
|
193 | 190 | changelogrevision = cl.changelogrevision |
|
194 | 191 | flags = cl.flags |
|
195 | 192 | |
|
196 | 193 | # A small cache to avoid doing the work twice for merges |
|
197 | 194 | # |
|
198 | 195 | # In the vast majority of cases, if we ask information for a revision |
|
199 | 196 | # about 1 parent, we'll later ask it for the other. So it make sense to |
|
200 | 197 | # keep the information around when reaching the first parent of a merge |
|
201 | 198 | # and dropping it after it was provided for the second parents. |
|
202 | 199 | # |
|
203 | 200 | # It exists cases were only one parent of the merge will be walked. It |
|
204 | 201 | # happens when the "destination" the copy tracing is descendant from a |
|
205 | 202 | # new root, not common with the "source". In that case, we will only walk |
|
206 | 203 | # through merge parents that are descendant of changesets common |
|
207 | 204 | # between "source" and "destination". |
|
208 | 205 | # |
|
209 | 206 | # With the current case implementation if such changesets have a copy |
|
210 | 207 | # information, we'll keep them in memory until the end of |
|
211 | 208 | # _changesetforwardcopies. We don't expect the case to be frequent |
|
212 | 209 | # enough to matters. |
|
213 | 210 | # |
|
214 | 211 | # In addition, it would be possible to reach pathological case, were |
|
215 | 212 | # many first parent are met before any second parent is reached. In |
|
216 | 213 | # that case the cache could grow. If this even become an issue one can |
|
217 | 214 | # safely introduce a maximum cache size. This would trade extra CPU/IO |
|
218 | 215 | # time to save memory. |
|
219 | 216 | merge_caches = {} |
|
220 | 217 | |
|
221 | 218 | def revinfo(rev): |
|
222 | 219 | p1, p2 = parents(rev) |
|
223 | 220 | if flags(rev) & REVIDX_SIDEDATA: |
|
224 | 221 | e = merge_caches.pop(rev, None) |
|
225 | 222 | if e is not None: |
|
226 | 223 | return e |
|
227 | 224 | c = changelogrevision(rev) |
|
228 | 225 | p1copies = c.p1copies |
|
229 | 226 | p2copies = c.p2copies |
|
230 | 227 | removed = c.filesremoved |
|
231 | 228 | if p1 != node.nullrev and p2 != node.nullrev: |
|
232 | 229 | # XXX some case we over cache, IGNORE |
|
233 | 230 | merge_caches[rev] = (p1, p2, p1copies, p2copies, removed) |
|
234 | 231 | else: |
|
235 | 232 | p1copies = {} |
|
236 | 233 | p2copies = {} |
|
237 | 234 | removed = () |
|
238 | 235 | return p1, p2, p1copies, p2copies, removed |
|
239 | 236 | |
|
240 | 237 | else: |
|
241 | 238 | |
|
242 | 239 | def revinfo(rev): |
|
243 | 240 | p1, p2 = parents(rev) |
|
244 | 241 | ctx = repo[rev] |
|
245 | 242 | p1copies, p2copies = ctx._copies |
|
246 | 243 | removed = ctx.filesremoved() |
|
247 | 244 | return p1, p2, p1copies, p2copies, removed |
|
248 | 245 | |
|
249 | 246 | return revinfo |
|
250 | 247 | |
|
251 | 248 | |
|
252 | 249 | def _changesetforwardcopies(a, b, match): |
|
253 | 250 | if a.rev() in (node.nullrev, b.rev()): |
|
254 | 251 | return {} |
|
255 | 252 | |
|
256 | 253 | repo = a.repo().unfiltered() |
|
257 | 254 | children = {} |
|
258 | 255 | revinfo = _revinfogetter(repo) |
|
259 | 256 | |
|
260 | 257 | cl = repo.changelog |
|
261 | 258 | missingrevs = cl.findmissingrevs(common=[a.rev()], heads=[b.rev()]) |
|
262 | 259 | mrset = set(missingrevs) |
|
263 | 260 | roots = set() |
|
264 | 261 | for r in missingrevs: |
|
265 | 262 | for p in cl.parentrevs(r): |
|
266 | 263 | if p == node.nullrev: |
|
267 | 264 | continue |
|
268 | 265 | if p not in children: |
|
269 | 266 | children[p] = [r] |
|
270 | 267 | else: |
|
271 | 268 | children[p].append(r) |
|
272 | 269 | if p not in mrset: |
|
273 | 270 | roots.add(p) |
|
274 | 271 | if not roots: |
|
275 | 272 | # no common revision to track copies from |
|
276 | 273 | return {} |
|
277 | 274 | min_root = min(roots) |
|
278 | 275 | |
|
279 | 276 | from_head = set( |
|
280 | 277 | cl.reachableroots(min_root, [b.rev()], list(roots), includepath=True) |
|
281 | 278 | ) |
|
282 | 279 | |
|
283 | 280 | iterrevs = set(from_head) |
|
284 | 281 | iterrevs &= mrset |
|
285 | 282 | iterrevs.update(roots) |
|
286 | 283 | iterrevs.remove(b.rev()) |
|
287 | 284 | all_copies = {r: {} for r in roots} |
|
288 | 285 | alwaysmatch = match.always() |
|
289 | 286 | for r in sorted(iterrevs): |
|
290 | 287 | copies = all_copies.pop(r) |
|
291 | 288 | for i, c in enumerate(children[r]): |
|
292 | 289 | p1, p2, p1copies, p2copies, removed = revinfo(c) |
|
293 | 290 | if r == p1: |
|
294 | 291 | parent = 1 |
|
295 | 292 | childcopies = p1copies |
|
296 | 293 | else: |
|
297 | 294 | assert r == p2 |
|
298 | 295 | parent = 2 |
|
299 | 296 | childcopies = p2copies |
|
300 | 297 | if not alwaysmatch: |
|
301 | 298 | childcopies = { |
|
302 | 299 | dst: src for dst, src in childcopies.items() if match(dst) |
|
303 | 300 | } |
|
304 | 301 | newcopies = copies |
|
305 | 302 | if childcopies: |
|
306 | 303 | newcopies = _chain(newcopies, childcopies) |
|
307 | 304 | # _chain makes a copies, we can avoid doing so in some |
|
308 | 305 | # simple/linear cases. |
|
309 | 306 | assert newcopies is not copies |
|
310 | 307 | for f in removed: |
|
311 | 308 | if f in newcopies: |
|
312 | 309 | if newcopies is copies: |
|
313 | 310 | # copy on write to avoid affecting potential other |
|
314 | 311 | # branches. when there are no other branches, this |
|
315 | 312 | # could be avoided. |
|
316 | 313 | newcopies = copies.copy() |
|
317 | 314 | del newcopies[f] |
|
318 | 315 | othercopies = all_copies.get(c) |
|
319 | 316 | if othercopies is None: |
|
320 | 317 | all_copies[c] = newcopies |
|
321 | 318 | else: |
|
322 | 319 | # we are the second parent to work on c, we need to merge our |
|
323 | 320 | # work with the other. |
|
324 | 321 | # |
|
325 | 322 | # Unlike when copies are stored in the filelog, we consider |
|
326 | 323 | # it a copy even if the destination already existed on the |
|
327 | 324 | # other branch. It's simply too expensive to check if the |
|
328 | 325 | # file existed in the manifest. |
|
329 | 326 | # |
|
330 | 327 | # In case of conflict, parent 1 take precedence over parent 2. |
|
331 | 328 | # This is an arbitrary choice made anew when implementing |
|
332 | 329 | # changeset based copies. It was made without regards with |
|
333 | 330 | # potential filelog related behavior. |
|
334 | 331 | if parent == 1: |
|
335 | 332 | othercopies.update(newcopies) |
|
336 | 333 | else: |
|
337 | 334 | newcopies.update(othercopies) |
|
338 | 335 | all_copies[c] = newcopies |
|
339 | 336 | return all_copies[b.rev()] |
|
340 | 337 | |
|
341 | 338 | |
|
342 | 339 | def _forwardcopies(a, b, base=None, match=None): |
|
343 | 340 | """find {dst@b: src@a} copy mapping where a is an ancestor of b""" |
|
344 | 341 | |
|
345 | 342 | if base is None: |
|
346 | 343 | base = a |
|
347 | 344 | match = a.repo().narrowmatch(match) |
|
348 | 345 | # check for working copy |
|
349 | 346 | if b.rev() is None: |
|
350 | 347 | cm = _committedforwardcopies(a, b.p1(), base, match) |
|
351 | 348 | # combine copies from dirstate if necessary |
|
352 | 349 | copies = _chain(cm, _dirstatecopies(b._repo, match)) |
|
353 | 350 | else: |
|
354 | 351 | copies = _committedforwardcopies(a, b, base, match) |
|
355 | 352 | return copies |
|
356 | 353 | |
|
357 | 354 | |
|
358 | 355 | def _backwardrenames(a, b, match): |
|
359 | 356 | if a._repo.ui.config(b'experimental', b'copytrace') == b'off': |
|
360 | 357 | return {} |
|
361 | 358 | |
|
362 | 359 | # Even though we're not taking copies into account, 1:n rename situations |
|
363 | 360 | # can still exist (e.g. hg cp a b; hg mv a c). In those cases we |
|
364 | 361 | # arbitrarily pick one of the renames. |
|
365 | 362 | # We don't want to pass in "match" here, since that would filter |
|
366 | 363 | # the destination by it. Since we're reversing the copies, we want |
|
367 | 364 | # to filter the source instead. |
|
368 | 365 | f = _forwardcopies(b, a) |
|
369 | 366 | r = {} |
|
370 | 367 | for k, v in sorted(pycompat.iteritems(f)): |
|
371 | 368 | if match and not match(v): |
|
372 | 369 | continue |
|
373 | 370 | # remove copies |
|
374 | 371 | if v in a: |
|
375 | 372 | continue |
|
376 | 373 | r[v] = k |
|
377 | 374 | return r |
|
378 | 375 | |
|
379 | 376 | |
|
380 | 377 | def pathcopies(x, y, match=None): |
|
381 | 378 | """find {dst@y: src@x} copy mapping for directed compare""" |
|
382 | 379 | repo = x._repo |
|
383 | 380 | debug = repo.ui.debugflag and repo.ui.configbool(b'devel', b'debug.copies') |
|
384 | 381 | if debug: |
|
385 | 382 | repo.ui.debug( |
|
386 | 383 | b'debug.copies: searching copies from %s to %s\n' % (x, y) |
|
387 | 384 | ) |
|
388 | 385 | if x == y or not x or not y: |
|
389 | 386 | return {} |
|
390 | 387 | a = y.ancestor(x) |
|
391 | 388 | if a == x: |
|
392 | 389 | if debug: |
|
393 | 390 | repo.ui.debug(b'debug.copies: search mode: forward\n') |
|
394 | 391 | if y.rev() is None and x == y.p1(): |
|
395 | 392 | # short-circuit to avoid issues with merge states |
|
396 | 393 | return _dirstatecopies(repo, match) |
|
397 | 394 | copies = _forwardcopies(x, y, match=match) |
|
398 | 395 | elif a == y: |
|
399 | 396 | if debug: |
|
400 | 397 | repo.ui.debug(b'debug.copies: search mode: backward\n') |
|
401 | 398 | copies = _backwardrenames(x, y, match=match) |
|
402 | 399 | else: |
|
403 | 400 | if debug: |
|
404 | 401 | repo.ui.debug(b'debug.copies: search mode: combined\n') |
|
405 | 402 | base = None |
|
406 | 403 | if a.rev() != node.nullrev: |
|
407 | 404 | base = x |
|
408 | 405 | copies = _chain( |
|
409 | 406 | _backwardrenames(x, a, match=match), |
|
410 | 407 | _forwardcopies(a, y, base, match=match), |
|
411 | 408 | ) |
|
412 | 409 | _filter(x, y, copies) |
|
413 | 410 | return copies |
|
414 | 411 | |
|
415 | 412 | |
|
416 | 413 | def mergecopies(repo, c1, c2, base): |
|
417 | 414 | """ |
|
418 | 415 | Finds moves and copies between context c1 and c2 that are relevant for |
|
419 | 416 | merging. 'base' will be used as the merge base. |
|
420 | 417 | |
|
421 | 418 | Copytracing is used in commands like rebase, merge, unshelve, etc to merge |
|
422 | 419 | files that were moved/ copied in one merge parent and modified in another. |
|
423 | 420 | For example: |
|
424 | 421 | |
|
425 | 422 | o ---> 4 another commit |
|
426 | 423 | | |
|
427 | 424 | | o ---> 3 commit that modifies a.txt |
|
428 | 425 | | / |
|
429 | 426 | o / ---> 2 commit that moves a.txt to b.txt |
|
430 | 427 | |/ |
|
431 | 428 | o ---> 1 merge base |
|
432 | 429 | |
|
433 | 430 | If we try to rebase revision 3 on revision 4, since there is no a.txt in |
|
434 | 431 | revision 4, and if user have copytrace disabled, we prints the following |
|
435 | 432 | message: |
|
436 | 433 | |
|
437 | 434 | ```other changed <file> which local deleted``` |
|
438 | 435 | |
|
439 | 436 | Returns five dicts: "copy", "movewithdir", "diverge", "renamedelete" and |
|
440 | 437 | "dirmove". |
|
441 | 438 | |
|
442 | 439 | "copy" is a mapping from destination name -> source name, |
|
443 | 440 | where source is in c1 and destination is in c2 or vice-versa. |
|
444 | 441 | |
|
445 | 442 | "movewithdir" is a mapping from source name -> destination name, |
|
446 | 443 | where the file at source present in one context but not the other |
|
447 | 444 | needs to be moved to destination by the merge process, because the |
|
448 | 445 | other context moved the directory it is in. |
|
449 | 446 | |
|
450 | 447 | "diverge" is a mapping of source name -> list of destination names |
|
451 | 448 | for divergent renames. |
|
452 | 449 | |
|
453 | 450 | "renamedelete" is a mapping of source name -> list of destination |
|
454 | 451 | names for files deleted in c1 that were renamed in c2 or vice-versa. |
|
455 | 452 | |
|
456 | 453 | "dirmove" is a mapping of detected source dir -> destination dir renames. |
|
457 | 454 | This is needed for handling changes to new files previously grafted into |
|
458 | 455 | renamed directories. |
|
459 | 456 | |
|
460 | 457 | This function calls different copytracing algorithms based on config. |
|
461 | 458 | """ |
|
462 | 459 | # avoid silly behavior for update from empty dir |
|
463 | 460 | if not c1 or not c2 or c1 == c2: |
|
464 | 461 | return {}, {}, {}, {}, {} |
|
465 | 462 | |
|
466 | 463 | narrowmatch = c1.repo().narrowmatch() |
|
467 | 464 | |
|
468 | 465 | # avoid silly behavior for parent -> working dir |
|
469 | 466 | if c2.node() is None and c1.node() == repo.dirstate.p1(): |
|
470 | 467 | return _dirstatecopies(repo, narrowmatch), {}, {}, {}, {} |
|
471 | 468 | |
|
472 | 469 | copytracing = repo.ui.config(b'experimental', b'copytrace') |
|
473 | 470 | if stringutil.parsebool(copytracing) is False: |
|
474 | 471 | # stringutil.parsebool() returns None when it is unable to parse the |
|
475 | 472 | # value, so we should rely on making sure copytracing is on such cases |
|
476 | 473 | return {}, {}, {}, {}, {} |
|
477 | 474 | |
|
478 | 475 | if usechangesetcentricalgo(repo): |
|
479 | 476 | # The heuristics don't make sense when we need changeset-centric algos |
|
480 | 477 | return _fullcopytracing(repo, c1, c2, base) |
|
481 | 478 | |
|
482 | 479 | # Copy trace disabling is explicitly below the node == p1 logic above |
|
483 | 480 | # because the logic above is required for a simple copy to be kept across a |
|
484 | 481 | # rebase. |
|
485 | 482 | if copytracing == b'heuristics': |
|
486 | 483 | # Do full copytracing if only non-public revisions are involved as |
|
487 | 484 | # that will be fast enough and will also cover the copies which could |
|
488 | 485 | # be missed by heuristics |
|
489 | 486 | if _isfullcopytraceable(repo, c1, base): |
|
490 | 487 | return _fullcopytracing(repo, c1, c2, base) |
|
491 | 488 | return _heuristicscopytracing(repo, c1, c2, base) |
|
492 | 489 | else: |
|
493 | 490 | return _fullcopytracing(repo, c1, c2, base) |
|
494 | 491 | |
|
495 | 492 | |
|
496 | 493 | def _isfullcopytraceable(repo, c1, base): |
|
497 | 494 | """ Checks that if base, source and destination are all no-public branches, |
|
498 | 495 | if yes let's use the full copytrace algorithm for increased capabilities |
|
499 | 496 | since it will be fast enough. |
|
500 | 497 | |
|
501 | 498 | `experimental.copytrace.sourcecommitlimit` can be used to set a limit for |
|
502 | 499 | number of changesets from c1 to base such that if number of changesets are |
|
503 | 500 | more than the limit, full copytracing algorithm won't be used. |
|
504 | 501 | """ |
|
505 | 502 | if c1.rev() is None: |
|
506 | 503 | c1 = c1.p1() |
|
507 | 504 | if c1.mutable() and base.mutable(): |
|
508 | 505 | sourcecommitlimit = repo.ui.configint( |
|
509 | 506 | b'experimental', b'copytrace.sourcecommitlimit' |
|
510 | 507 | ) |
|
511 | 508 | commits = len(repo.revs(b'%d::%d', base.rev(), c1.rev())) |
|
512 | 509 | return commits < sourcecommitlimit |
|
513 | 510 | return False |
|
514 | 511 | |
|
515 | 512 | |
|
516 | 513 | def _checksinglesidecopies( |
|
517 | 514 | src, dsts1, m1, m2, mb, c2, base, copy, renamedelete |
|
518 | 515 | ): |
|
519 | 516 | if src not in m2: |
|
520 | 517 | # deleted on side 2 |
|
521 | 518 | if src not in m1: |
|
522 | 519 | # renamed on side 1, deleted on side 2 |
|
523 | 520 | renamedelete[src] = dsts1 |
|
524 | 521 | elif m2[src] != mb[src]: |
|
525 | 522 | if not _related(c2[src], base[src]): |
|
526 | 523 | return |
|
527 | 524 | # modified on side 2 |
|
528 | 525 | for dst in dsts1: |
|
529 | 526 | if dst not in m2: |
|
530 | 527 | # dst not added on side 2 (handle as regular |
|
531 | 528 | # "both created" case in manifestmerge otherwise) |
|
532 | 529 | copy[dst] = src |
|
533 | 530 | |
|
534 | 531 | |
|
535 | 532 | def _fullcopytracing(repo, c1, c2, base): |
|
536 | 533 | """ The full copytracing algorithm which finds all the new files that were |
|
537 | 534 | added from merge base up to the top commit and for each file it checks if |
|
538 | 535 | this file was copied from another file. |
|
539 | 536 | |
|
540 | 537 | This is pretty slow when a lot of changesets are involved but will track all |
|
541 | 538 | the copies. |
|
542 | 539 | """ |
|
543 | 540 | m1 = c1.manifest() |
|
544 | 541 | m2 = c2.manifest() |
|
545 | 542 | mb = base.manifest() |
|
546 | 543 | |
|
547 | 544 | copies1 = pathcopies(base, c1) |
|
548 | 545 | copies2 = pathcopies(base, c2) |
|
549 | 546 | |
|
550 | 547 | inversecopies1 = {} |
|
551 | 548 | inversecopies2 = {} |
|
552 | 549 | for dst, src in copies1.items(): |
|
553 | 550 | inversecopies1.setdefault(src, []).append(dst) |
|
554 | 551 | for dst, src in copies2.items(): |
|
555 | 552 | inversecopies2.setdefault(src, []).append(dst) |
|
556 | 553 | |
|
557 | 554 | copy = {} |
|
558 | 555 | diverge = {} |
|
559 | 556 | renamedelete = {} |
|
560 | 557 | allsources = set(inversecopies1) | set(inversecopies2) |
|
561 | 558 | for src in allsources: |
|
562 | 559 | dsts1 = inversecopies1.get(src) |
|
563 | 560 | dsts2 = inversecopies2.get(src) |
|
564 | 561 | if dsts1 and dsts2: |
|
565 | 562 | # copied/renamed on both sides |
|
566 | 563 | if src not in m1 and src not in m2: |
|
567 | 564 | # renamed on both sides |
|
568 | 565 | dsts1 = set(dsts1) |
|
569 | 566 | dsts2 = set(dsts2) |
|
570 | 567 | # If there's some overlap in the rename destinations, we |
|
571 | 568 | # consider it not divergent. For example, if side 1 copies 'a' |
|
572 | 569 | # to 'b' and 'c' and deletes 'a', and side 2 copies 'a' to 'c' |
|
573 | 570 | # and 'd' and deletes 'a'. |
|
574 | 571 | if dsts1 & dsts2: |
|
575 | 572 | for dst in dsts1 & dsts2: |
|
576 | 573 | copy[dst] = src |
|
577 | 574 | else: |
|
578 | 575 | diverge[src] = sorted(dsts1 | dsts2) |
|
579 | 576 | elif src in m1 and src in m2: |
|
580 | 577 | # copied on both sides |
|
581 | 578 | dsts1 = set(dsts1) |
|
582 | 579 | dsts2 = set(dsts2) |
|
583 | 580 | for dst in dsts1 & dsts2: |
|
584 | 581 | copy[dst] = src |
|
585 | 582 | # TODO: Handle cases where it was renamed on one side and copied |
|
586 | 583 | # on the other side |
|
587 | 584 | elif dsts1: |
|
588 | 585 | # copied/renamed only on side 1 |
|
589 | 586 | _checksinglesidecopies( |
|
590 | 587 | src, dsts1, m1, m2, mb, c2, base, copy, renamedelete |
|
591 | 588 | ) |
|
592 | 589 | elif dsts2: |
|
593 | 590 | # copied/renamed only on side 2 |
|
594 | 591 | _checksinglesidecopies( |
|
595 | 592 | src, dsts2, m2, m1, mb, c1, base, copy, renamedelete |
|
596 | 593 | ) |
|
597 | 594 | |
|
598 | 595 | renamedeleteset = set() |
|
599 | 596 | divergeset = set() |
|
600 | 597 | for dsts in diverge.values(): |
|
601 | 598 | divergeset.update(dsts) |
|
602 | 599 | for dsts in renamedelete.values(): |
|
603 | 600 | renamedeleteset.update(dsts) |
|
604 | 601 | |
|
605 | 602 | # find interesting file sets from manifests |
|
606 | 603 | addedinm1 = m1.filesnotin(mb, repo.narrowmatch()) |
|
607 | 604 | addedinm2 = m2.filesnotin(mb, repo.narrowmatch()) |
|
608 | 605 | u1 = sorted(addedinm1 - addedinm2) |
|
609 | 606 | u2 = sorted(addedinm2 - addedinm1) |
|
610 | 607 | |
|
611 | 608 | header = b" unmatched files in %s" |
|
612 | 609 | if u1: |
|
613 | 610 | repo.ui.debug(b"%s:\n %s\n" % (header % b'local', b"\n ".join(u1))) |
|
614 | 611 | if u2: |
|
615 | 612 | repo.ui.debug(b"%s:\n %s\n" % (header % b'other', b"\n ".join(u2))) |
|
616 | 613 | |
|
617 | 614 | fullcopy = copies1.copy() |
|
618 | 615 | fullcopy.update(copies2) |
|
619 | 616 | if not fullcopy: |
|
620 | 617 | return copy, {}, diverge, renamedelete, {} |
|
621 | 618 | |
|
622 | 619 | if repo.ui.debugflag: |
|
623 | 620 | repo.ui.debug( |
|
624 | 621 | b" all copies found (* = to merge, ! = divergent, " |
|
625 | 622 | b"% = renamed and deleted):\n" |
|
626 | 623 | ) |
|
627 | 624 | for f in sorted(fullcopy): |
|
628 | 625 | note = b"" |
|
629 | 626 | if f in copy: |
|
630 | 627 | note += b"*" |
|
631 | 628 | if f in divergeset: |
|
632 | 629 | note += b"!" |
|
633 | 630 | if f in renamedeleteset: |
|
634 | 631 | note += b"%" |
|
635 | 632 | repo.ui.debug( |
|
636 | 633 | b" src: '%s' -> dst: '%s' %s\n" % (fullcopy[f], f, note) |
|
637 | 634 | ) |
|
638 | 635 | del divergeset |
|
639 | 636 | |
|
640 | 637 | repo.ui.debug(b" checking for directory renames\n") |
|
641 | 638 | |
|
642 | 639 | # generate a directory move map |
|
643 | 640 | d1, d2 = c1.dirs(), c2.dirs() |
|
644 | 641 | invalid = set() |
|
645 | 642 | dirmove = {} |
|
646 | 643 | |
|
647 | 644 | # examine each file copy for a potential directory move, which is |
|
648 | 645 | # when all the files in a directory are moved to a new directory |
|
649 | 646 | for dst, src in pycompat.iteritems(fullcopy): |
|
650 | 647 | dsrc, ddst = pathutil.dirname(src), pathutil.dirname(dst) |
|
651 | 648 | if dsrc in invalid: |
|
652 | 649 | # already seen to be uninteresting |
|
653 | 650 | continue |
|
654 | 651 | elif dsrc in d1 and ddst in d1: |
|
655 | 652 | # directory wasn't entirely moved locally |
|
656 | 653 | invalid.add(dsrc) |
|
657 | 654 | elif dsrc in d2 and ddst in d2: |
|
658 | 655 | # directory wasn't entirely moved remotely |
|
659 | 656 | invalid.add(dsrc) |
|
660 | 657 | elif dsrc in dirmove and dirmove[dsrc] != ddst: |
|
661 | 658 | # files from the same directory moved to two different places |
|
662 | 659 | invalid.add(dsrc) |
|
663 | 660 | else: |
|
664 | 661 | # looks good so far |
|
665 | 662 | dirmove[dsrc] = ddst |
|
666 | 663 | |
|
667 | 664 | for i in invalid: |
|
668 | 665 | if i in dirmove: |
|
669 | 666 | del dirmove[i] |
|
670 | 667 | del d1, d2, invalid |
|
671 | 668 | |
|
672 | 669 | if not dirmove: |
|
673 | 670 | return copy, {}, diverge, renamedelete, {} |
|
674 | 671 | |
|
675 | 672 | dirmove = {k + b"/": v + b"/" for k, v in pycompat.iteritems(dirmove)} |
|
676 | 673 | |
|
677 | 674 | for d in dirmove: |
|
678 | 675 | repo.ui.debug( |
|
679 | 676 | b" discovered dir src: '%s' -> dst: '%s'\n" % (d, dirmove[d]) |
|
680 | 677 | ) |
|
681 | 678 | |
|
682 | 679 | movewithdir = {} |
|
683 | 680 | # check unaccounted nonoverlapping files against directory moves |
|
684 | 681 | for f in u1 + u2: |
|
685 | 682 | if f not in fullcopy: |
|
686 | 683 | for d in dirmove: |
|
687 | 684 | if f.startswith(d): |
|
688 | 685 | # new file added in a directory that was moved, move it |
|
689 | 686 | df = dirmove[d] + f[len(d) :] |
|
690 | 687 | if df not in copy: |
|
691 | 688 | movewithdir[f] = df |
|
692 | 689 | repo.ui.debug( |
|
693 | 690 | b" pending file src: '%s' -> dst: '%s'\n" |
|
694 | 691 | % (f, df) |
|
695 | 692 | ) |
|
696 | 693 | break |
|
697 | 694 | |
|
698 | 695 | return copy, movewithdir, diverge, renamedelete, dirmove |
|
699 | 696 | |
|
700 | 697 | |
|
701 | 698 | def _heuristicscopytracing(repo, c1, c2, base): |
|
702 | 699 | """ Fast copytracing using filename heuristics |
|
703 | 700 | |
|
704 | 701 | Assumes that moves or renames are of following two types: |
|
705 | 702 | |
|
706 | 703 | 1) Inside a directory only (same directory name but different filenames) |
|
707 | 704 | 2) Move from one directory to another |
|
708 | 705 | (same filenames but different directory names) |
|
709 | 706 | |
|
710 | 707 | Works only when there are no merge commits in the "source branch". |
|
711 | 708 | Source branch is commits from base up to c2 not including base. |
|
712 | 709 | |
|
713 | 710 | If merge is involved it fallbacks to _fullcopytracing(). |
|
714 | 711 | |
|
715 | 712 | Can be used by setting the following config: |
|
716 | 713 | |
|
717 | 714 | [experimental] |
|
718 | 715 | copytrace = heuristics |
|
719 | 716 | |
|
720 | 717 | In some cases the copy/move candidates found by heuristics can be very large |
|
721 | 718 | in number and that will make the algorithm slow. The number of possible |
|
722 | 719 | candidates to check can be limited by using the config |
|
723 | 720 | `experimental.copytrace.movecandidateslimit` which defaults to 100. |
|
724 | 721 | """ |
|
725 | 722 | |
|
726 | 723 | if c1.rev() is None: |
|
727 | 724 | c1 = c1.p1() |
|
728 | 725 | if c2.rev() is None: |
|
729 | 726 | c2 = c2.p1() |
|
730 | 727 | |
|
731 | 728 | copies = {} |
|
732 | 729 | |
|
733 | 730 | changedfiles = set() |
|
734 | 731 | m1 = c1.manifest() |
|
735 | 732 | if not repo.revs(b'%d::%d', base.rev(), c2.rev()): |
|
736 | 733 | # If base is not in c2 branch, we switch to fullcopytracing |
|
737 | 734 | repo.ui.debug( |
|
738 | 735 | b"switching to full copytracing as base is not " |
|
739 | 736 | b"an ancestor of c2\n" |
|
740 | 737 | ) |
|
741 | 738 | return _fullcopytracing(repo, c1, c2, base) |
|
742 | 739 | |
|
743 | 740 | ctx = c2 |
|
744 | 741 | while ctx != base: |
|
745 | 742 | if len(ctx.parents()) == 2: |
|
746 | 743 | # To keep things simple let's not handle merges |
|
747 | 744 | repo.ui.debug(b"switching to full copytracing because of merges\n") |
|
748 | 745 | return _fullcopytracing(repo, c1, c2, base) |
|
749 | 746 | changedfiles.update(ctx.files()) |
|
750 | 747 | ctx = ctx.p1() |
|
751 | 748 | |
|
752 | 749 | cp = _forwardcopies(base, c2) |
|
753 | 750 | for dst, src in pycompat.iteritems(cp): |
|
754 | 751 | if src in m1: |
|
755 | 752 | copies[dst] = src |
|
756 | 753 | |
|
757 | 754 | # file is missing if it isn't present in the destination, but is present in |
|
758 | 755 | # the base and present in the source. |
|
759 | 756 | # Presence in the base is important to exclude added files, presence in the |
|
760 | 757 | # source is important to exclude removed files. |
|
761 | 758 | filt = lambda f: f not in m1 and f in base and f in c2 |
|
762 | 759 | missingfiles = [f for f in changedfiles if filt(f)] |
|
763 | 760 | |
|
764 | 761 | if missingfiles: |
|
765 | 762 | basenametofilename = collections.defaultdict(list) |
|
766 | 763 | dirnametofilename = collections.defaultdict(list) |
|
767 | 764 | |
|
768 | 765 | for f in m1.filesnotin(base.manifest()): |
|
769 | 766 | basename = os.path.basename(f) |
|
770 | 767 | dirname = os.path.dirname(f) |
|
771 | 768 | basenametofilename[basename].append(f) |
|
772 | 769 | dirnametofilename[dirname].append(f) |
|
773 | 770 | |
|
774 | 771 | for f in missingfiles: |
|
775 | 772 | basename = os.path.basename(f) |
|
776 | 773 | dirname = os.path.dirname(f) |
|
777 | 774 | samebasename = basenametofilename[basename] |
|
778 | 775 | samedirname = dirnametofilename[dirname] |
|
779 | 776 | movecandidates = samebasename + samedirname |
|
780 | 777 | # f is guaranteed to be present in c2, that's why |
|
781 | 778 | # c2.filectx(f) won't fail |
|
782 | 779 | f2 = c2.filectx(f) |
|
783 | 780 | # we can have a lot of candidates which can slow down the heuristics |
|
784 | 781 | # config value to limit the number of candidates moves to check |
|
785 | 782 | maxcandidates = repo.ui.configint( |
|
786 | 783 | b'experimental', b'copytrace.movecandidateslimit' |
|
787 | 784 | ) |
|
788 | 785 | |
|
789 | 786 | if len(movecandidates) > maxcandidates: |
|
790 | 787 | repo.ui.status( |
|
791 | 788 | _( |
|
792 | 789 | b"skipping copytracing for '%s', more " |
|
793 | 790 | b"candidates than the limit: %d\n" |
|
794 | 791 | ) |
|
795 | 792 | % (f, len(movecandidates)) |
|
796 | 793 | ) |
|
797 | 794 | continue |
|
798 | 795 | |
|
799 | 796 | for candidate in movecandidates: |
|
800 | 797 | f1 = c1.filectx(candidate) |
|
801 | 798 | if _related(f1, f2): |
|
802 | 799 | # if there are a few related copies then we'll merge |
|
803 | 800 | # changes into all of them. This matches the behaviour |
|
804 | 801 | # of upstream copytracing |
|
805 | 802 | copies[candidate] = f |
|
806 | 803 | |
|
807 | 804 | return copies, {}, {}, {}, {} |
|
808 | 805 | |
|
809 | 806 | |
|
810 | 807 | def _related(f1, f2): |
|
811 | 808 | """return True if f1 and f2 filectx have a common ancestor |
|
812 | 809 | |
|
813 | 810 | Walk back to common ancestor to see if the two files originate |
|
814 | 811 | from the same file. Since workingfilectx's rev() is None it messes |
|
815 | 812 | up the integer comparison logic, hence the pre-step check for |
|
816 | 813 | None (f1 and f2 can only be workingfilectx's initially). |
|
817 | 814 | """ |
|
818 | 815 | |
|
819 | 816 | if f1 == f2: |
|
820 | 817 | return True # a match |
|
821 | 818 | |
|
822 | 819 | g1, g2 = f1.ancestors(), f2.ancestors() |
|
823 | 820 | try: |
|
824 | 821 | f1r, f2r = f1.linkrev(), f2.linkrev() |
|
825 | 822 | |
|
826 | 823 | if f1r is None: |
|
827 | 824 | f1 = next(g1) |
|
828 | 825 | if f2r is None: |
|
829 | 826 | f2 = next(g2) |
|
830 | 827 | |
|
831 | 828 | while True: |
|
832 | 829 | f1r, f2r = f1.linkrev(), f2.linkrev() |
|
833 | 830 | if f1r > f2r: |
|
834 | 831 | f1 = next(g1) |
|
835 | 832 | elif f2r > f1r: |
|
836 | 833 | f2 = next(g2) |
|
837 | 834 | else: # f1 and f2 point to files in the same linkrev |
|
838 | 835 | return f1 == f2 # true if they point to the same file |
|
839 | 836 | except StopIteration: |
|
840 | 837 | return False |
|
841 | 838 | |
|
842 | 839 | |
|
843 | 840 | def duplicatecopies(repo, wctx, rev, fromrev, skiprev=None): |
|
844 | 841 | """reproduce copies from fromrev to rev in the dirstate |
|
845 | 842 | |
|
846 | 843 | If skiprev is specified, it's a revision that should be used to |
|
847 | 844 | filter copy records. Any copies that occur between fromrev and |
|
848 | 845 | skiprev will not be duplicated, even if they appear in the set of |
|
849 | 846 | copies between fromrev and rev. |
|
850 | 847 | """ |
|
851 | 848 | exclude = {} |
|
852 | 849 | ctraceconfig = repo.ui.config(b'experimental', b'copytrace') |
|
853 | 850 | bctrace = stringutil.parsebool(ctraceconfig) |
|
854 | 851 | if skiprev is not None and ( |
|
855 | 852 | ctraceconfig == b'heuristics' or bctrace or bctrace is None |
|
856 | 853 | ): |
|
857 | 854 | # copytrace='off' skips this line, but not the entire function because |
|
858 | 855 | # the line below is O(size of the repo) during a rebase, while the rest |
|
859 | 856 | # of the function is much faster (and is required for carrying copy |
|
860 | 857 | # metadata across the rebase anyway). |
|
861 | 858 | exclude = pathcopies(repo[fromrev], repo[skiprev]) |
|
862 | 859 | for dst, src in pycompat.iteritems(pathcopies(repo[fromrev], repo[rev])): |
|
863 | 860 | if dst in exclude: |
|
864 | 861 | continue |
|
865 | 862 | if dst in wctx: |
|
866 | 863 | wctx[dst].markcopied(src) |
|
867 | 864 | |
|
868 | 865 | |
|
869 | 866 | def computechangesetfilesadded(ctx): |
|
870 | 867 | """return the list of files added in a changeset |
|
871 | 868 | """ |
|
872 | 869 | added = [] |
|
873 | 870 | for f in ctx.files(): |
|
874 | 871 | if not any(f in p for p in ctx.parents()): |
|
875 | 872 | added.append(f) |
|
876 | 873 | return added |
|
877 | 874 | |
|
878 | 875 | |
|
879 | 876 | def computechangesetfilesremoved(ctx): |
|
880 | 877 | """return the list of files removed in a changeset |
|
881 | 878 | """ |
|
882 | 879 | removed = [] |
|
883 | 880 | for f in ctx.files(): |
|
884 | 881 | if f not in ctx: |
|
885 | 882 | removed.append(f) |
|
886 | 883 | return removed |
|
887 | 884 | |
|
888 | 885 | |
|
889 | 886 | def computechangesetcopies(ctx): |
|
890 | 887 | """return the copies data for a changeset |
|
891 | 888 | |
|
892 | 889 | The copies data are returned as a pair of dictionnary (p1copies, p2copies). |
|
893 | 890 | |
|
894 | 891 | Each dictionnary are in the form: `{newname: oldname}` |
|
895 | 892 | """ |
|
896 | 893 | p1copies = {} |
|
897 | 894 | p2copies = {} |
|
898 | 895 | p1 = ctx.p1() |
|
899 | 896 | p2 = ctx.p2() |
|
900 | 897 | narrowmatch = ctx._repo.narrowmatch() |
|
901 | 898 | for dst in ctx.files(): |
|
902 | 899 | if not narrowmatch(dst) or dst not in ctx: |
|
903 | 900 | continue |
|
904 | 901 | copied = ctx[dst].renamed() |
|
905 | 902 | if not copied: |
|
906 | 903 | continue |
|
907 | 904 | src, srcnode = copied |
|
908 | 905 | if src in p1 and p1[src].filenode() == srcnode: |
|
909 | 906 | p1copies[dst] = src |
|
910 | 907 | elif src in p2 and p2[src].filenode() == srcnode: |
|
911 | 908 | p2copies[dst] = src |
|
912 | 909 | return p1copies, p2copies |
|
913 | 910 | |
|
914 | 911 | |
|
915 | 912 | def encodecopies(files, copies): |
|
916 | 913 | items = [] |
|
917 | 914 | for i, dst in enumerate(files): |
|
918 | 915 | if dst in copies: |
|
919 | 916 | items.append(b'%d\0%s' % (i, copies[dst])) |
|
920 | 917 | if len(items) != len(copies): |
|
921 | 918 | raise error.ProgrammingError( |
|
922 | 919 | b'some copy targets missing from file list' |
|
923 | 920 | ) |
|
924 | 921 | return b"\n".join(items) |
|
925 | 922 | |
|
926 | 923 | |
|
927 | 924 | def decodecopies(files, data): |
|
928 | 925 | try: |
|
929 | 926 | copies = {} |
|
930 | 927 | if not data: |
|
931 | 928 | return copies |
|
932 | 929 | for l in data.split(b'\n'): |
|
933 | 930 | strindex, src = l.split(b'\0') |
|
934 | 931 | i = int(strindex) |
|
935 | 932 | dst = files[i] |
|
936 | 933 | copies[dst] = src |
|
937 | 934 | return copies |
|
938 | 935 | except (ValueError, IndexError): |
|
939 | 936 | # Perhaps someone had chosen the same key name (e.g. "p1copies") and |
|
940 | 937 | # used different syntax for the value. |
|
941 | 938 | return None |
|
942 | 939 | |
|
943 | 940 | |
|
944 | 941 | def encodefileindices(files, subset): |
|
945 | 942 | subset = set(subset) |
|
946 | 943 | indices = [] |
|
947 | 944 | for i, f in enumerate(files): |
|
948 | 945 | if f in subset: |
|
949 | 946 | indices.append(b'%d' % i) |
|
950 | 947 | return b'\n'.join(indices) |
|
951 | 948 | |
|
952 | 949 | |
|
953 | 950 | def decodefileindices(files, data): |
|
954 | 951 | try: |
|
955 | 952 | subset = [] |
|
956 | 953 | if not data: |
|
957 | 954 | return subset |
|
958 | 955 | for strindex in data.split(b'\n'): |
|
959 | 956 | i = int(strindex) |
|
960 | 957 | if i < 0 or i >= len(files): |
|
961 | 958 | return None |
|
962 | 959 | subset.append(files[i]) |
|
963 | 960 | return subset |
|
964 | 961 | except (ValueError, IndexError): |
|
965 | 962 | # Perhaps someone had chosen the same key name (e.g. "added") and |
|
966 | 963 | # used different syntax for the value. |
|
967 | 964 | return None |
|
968 | 965 | |
|
969 | 966 | |
|
970 | 967 | def _getsidedata(srcrepo, rev): |
|
971 | 968 | ctx = srcrepo[rev] |
|
972 | 969 | filescopies = computechangesetcopies(ctx) |
|
973 | 970 | filesadded = computechangesetfilesadded(ctx) |
|
974 | 971 | filesremoved = computechangesetfilesremoved(ctx) |
|
975 | 972 | sidedata = {} |
|
976 | 973 | if any([filescopies, filesadded, filesremoved]): |
|
977 | 974 | sortedfiles = sorted(ctx.files()) |
|
978 | 975 | p1copies, p2copies = filescopies |
|
979 | 976 | p1copies = encodecopies(sortedfiles, p1copies) |
|
980 | 977 | p2copies = encodecopies(sortedfiles, p2copies) |
|
981 | 978 | filesadded = encodefileindices(sortedfiles, filesadded) |
|
982 | 979 | filesremoved = encodefileindices(sortedfiles, filesremoved) |
|
983 | 980 | if p1copies: |
|
984 | 981 | sidedata[sidedatamod.SD_P1COPIES] = p1copies |
|
985 | 982 | if p2copies: |
|
986 | 983 | sidedata[sidedatamod.SD_P2COPIES] = p2copies |
|
987 | 984 | if filesadded: |
|
988 | 985 | sidedata[sidedatamod.SD_FILESADDED] = filesadded |
|
989 | 986 | if filesremoved: |
|
990 | 987 | sidedata[sidedatamod.SD_FILESREMOVED] = filesremoved |
|
991 | 988 | return sidedata |
|
992 | 989 | |
|
993 | 990 | |
|
994 | 991 | def getsidedataadder(srcrepo, destrepo): |
|
995 | 992 | def sidedatacompanion(revlog, rev): |
|
996 | 993 | sidedata = {} |
|
997 | 994 | if util.safehasattr(revlog, 'filteredrevs'): # this is a changelog |
|
998 | 995 | sidedata = _getsidedata(srcrepo, rev) |
|
999 | 996 | return False, (), sidedata |
|
1000 | 997 | |
|
1001 | 998 | return sidedatacompanion |
|
1002 | 999 | |
|
1003 | 1000 | |
|
1004 | 1001 | def getsidedataremover(srcrepo, destrepo): |
|
1005 | 1002 | def sidedatacompanion(revlog, rev): |
|
1006 | 1003 | f = () |
|
1007 | 1004 | if util.safehasattr(revlog, 'filteredrevs'): # this is a changelog |
|
1008 | 1005 | if revlog.flags(rev) & REVIDX_SIDEDATA: |
|
1009 | 1006 | f = ( |
|
1010 | 1007 | sidedatamod.SD_P1COPIES, |
|
1011 | 1008 | sidedatamod.SD_P2COPIES, |
|
1012 | 1009 | sidedatamod.SD_FILESADDED, |
|
1013 | 1010 | sidedatamod.SD_FILESREMOVED, |
|
1014 | 1011 | ) |
|
1015 | 1012 | return False, f, {} |
|
1016 | 1013 | |
|
1017 | 1014 | return sidedatacompanion |
General Comments 0
You need to be logged in to leave comments.
Login now