Kiln » Unity3D Unity 3D's proposed fixes and extensions to Kiln BFiles
Clone URL:  

First public release of Kiln BFiles

Changeset e9febbf6d1d9

by Profile picture of User 12Benjamin Pollack <benjamin@fogcreek.com>

Changes to 65 files · Browse files at e9febbf6d1d9 Diff from another changeset...

Change 1 of 1 Show Entire File design.txt Stacked
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
@@ -0,0 +1,42 @@
+= bfiles - manage large binary files = +This extension is based off of Greg Ward's bfiles extension which can be found +at http://mercurial.selenic.com/wiki/BfilesExtension. + +== The Bfile Store == + +Bfile stores are simply directories where each file is a bfile. The filename +is the sha1 hash of the bfile. The path is not necessary because all interactions +with the store have one of these forms: + +-Download a bfile with this hash +-Upload a bfile with this hash +-Check if the store has a bfile with this hash + +== The Local Repository == + +The local repository has a bfile store in .hg/bfiles which holds a subset of the +bfiles needed. On a clone only the bfiles at tip are downloaded. When bfiles are +downloaded from the central store a copy is saved in this store. + +== The Global Cache == + +Bfiles in a local repository store are hard linked to files in the global cache. Before +a file is downloaded we check if it is in the global cache. + +== Implementation Details == + +Each bfile has a standin which is in .hgbfiles. The standin is tracked by Mercurial. +The contents of the standin is the SHA1 hash of the bfile. When a bfile is added/removed/ +copied/renamed/etc the same operation is applied to the standin. Thus the history of the +standin is the history of the bfile. + +For performance reasons the contents of a standin is only updated before a commit. +Standins are added/removed/copied/renamed from add/remove/copy/rename Mercurial +commands but their contents will not be updated. The contents of a standin will always +be the hash of the bfile as of the last commit. To support some commands (revert) some +standins are temporarily updated but will be changed back after the command is finished. + +A Mercurial dirstate object tracks the state of the bfiles. The dirstate uses the +last modified time and current size to detect if a file has changed (without reading +the entire contents of the file). +
Change 1 of 1 Show Entire File kbfiles/​__init__.py Stacked
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
@@ -0,0 +1,32 @@
+'''track large binary files + +Large binary files tend to be not very compressible, not very "diffable", +and not at all mergeable. Such files are not handled well by Mercurial\'s +storage format (revlog), which is based on compressed binary deltas. +bfiles solves this problem by adding a centralized client-server layer on +top of Mercurial: big files live in a *central store* out on the network +somewhere, and you only fetch the big files that you need when you need +them. + +bfiles works by maintaining a *standin* in .hgbfiles/ for each big file. +The standins are small (41 bytes: an SHA-1 hash plus newline) and are +tracked by Mercurial. Big file revisions are identified by the SHA-1 hash +of their contents, which is written to the standin. bfiles uses that +revision ID to get/put big file revisions from/to the central store. + +A complete tutorial for using bfiles is included in ``usage.txt`` in the +bfiles source distribution. See +http://vc.gerg.ca/hg/hg-bfiles/raw-file/tip/usage.txt for the latest +version. +''' + +from mercurial import commands +import bfsetup +import bfcommands + +reposetup = bfsetup.reposetup +uisetup = bfsetup.uisetup + +commands.norepo += " kbfconvert" + +cmdtable = bfcommands.cmdtable
Change 1 of 1 Show Entire File kbfiles/​basestore.py Stacked
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
@@ -0,0 +1,180 @@
+'''Base class for store implementations and store-related utility code.''' + +import os +import tempfile +import binascii +import bfutil + +from mercurial import util, node, error, url as url_, hg +from mercurial.i18n import _ + +class StoreError(Exception): + '''Raised when there is a problem getting files from or putting + files to a central store.''' + def __init__(self, filename, hash, url, detail): + self.filename = filename + self.hash = hash + self.url = url + self.detail = detail + + def longmessage(self): + return ("%s: %s\n" + "(failed URL: %s)\n" + % (self.filename, self.detail, self.url)) + + def __str__(self): + return "%s: %s" % (self.url, self.detail) + +class basestore(object): + def __init__(self, ui, repo, url): + self.ui = ui + self.repo = repo + self.url = url + + def put(self, source, hash): + '''Put source file into the store under <filename>/<hash>.''' + raise NotImplementedError('abstract method') + + def exists(self, hash): + '''Check to see if the store contains the given hash.''' + raise NotImplementedError('abstract method') + + def get(self, files): + '''Get the specified big files from the store and write to local + files under repo.root. files is a list of (filename, hash) + tuples. Return (success, missing), lists of files successfuly + downloaded and those not found in the store. success is a list + of (filename, hash) tuples; missing is a list of filenames that + we could not get. (The detailed error message will already have + been presented to the user, so missing is just supplied as a + summary.)''' + success = [] + missing = [] + ui = self.ui + + at = 0 + for filename, hash in files: + ui.progress(_('Getting kbfiles'), at, unit='kbfile', total=len(files)) + at += 1 + ui.note(_('getting %s\n') % filename) + outfilename = self.repo.wjoin(filename) + destdir = os.path.dirname(outfilename) + util.makedirs(destdir) + if not os.path.isdir(destdir): + self.abort(error.RepoError(_('cannot create dest directory %s') % destdir)) + + # No need to pass mode='wb' to fdopen(), since mkstemp() already + # opened the file in binary mode. + (tmpfd, tmpfilename) = tempfile.mkstemp( + dir=destdir, prefix=os.path.basename(filename)) + tmpfile = os.fdopen(tmpfd, 'w') + + try: + bhash = self._getfile(tmpfile, filename, hash) + except StoreError, err: + ui.warn(err.longmessage()) + os.remove(tmpfilename) + missing.append(filename) + continue + + hhash = binascii.hexlify(bhash) + if hhash != hash: + ui.warn(_('%s: data corruption (expected %s, got %s)\n') + % (filename, hash, hhash)) + os.remove(tmpfilename) + missing.append(filename) + else: + if os.path.exists(outfilename): # for windows + os.remove(outfilename) + os.rename(tmpfilename, outfilename) + bfutil.copy_to_cache(self.repo, self.repo['.'].node(), filename, True) + success.append((filename, hhash)) + + ui.progress(_('Getting bfiles'), None) + return (success, missing) + + def verify(self, revs, contents=False): + '''Verify the existence (and, optionally, contents) of every big + file revision referenced by every changeset in revs. + Return 0 if all is well, non-zero on any errors.''' + write = self.ui.write + failed = False + + write(_('searching %d changesets for big files\n') % len(revs)) + verified = set() # set of (filename, filenode) tuples + + for rev in revs: + cctx = self.repo[rev] + cset = "%d:%s" % (cctx.rev(), node.short(cctx.node())) + + for standin in cctx: + failed = (self._verifyfile(cctx, + cset, + contents, + standin, + verified) + or failed) + + num_revs = len(verified) + num_bfiles = len(set([fname for (fname, fnode) in verified])) + if contents: + write(_('verified contents of %d revisions of %d big files\n') + % (num_revs, num_bfiles)) + else: + write(_('verified existence of %d revisions of %d big files\n') + % (num_revs, num_bfiles)) + + return int(failed) + + def _getfile(self, tmpfile, filename, hash): + '''Fetch one revision of one file from the store and write it + to tmpfile. Compute the hash of the file on-the-fly as it + downloads and return the binary hash. Close tmpfile. Raise + StoreError if unable to download the file (e.g. it does not + exist in the store).''' + raise NotImplementedError('abstract method') + + def _verifyfile(self, cctx, cset, contents, standin, verified): + '''Perform the actual verification of a file in the store. + ''' + raise NotImplementedError('abstract method') + +import localstore, httpstore + +_store_provider = { + 'file': (localstore, 'localstore'), + 'http': (httpstore, 'httpstore'), + 'https': (httpstore, 'httpstore'), + } + +# During clone this function is passed the src's ui object +# but it needs the dest's ui object so it can read out of +# the config file. Use repo.ui instead. +def _open_store(repo, path=None, put=False): + ui = repo.ui + if not path: + path = ui.expandpath('default-push', 'default') + # If 'default-push' and 'default' can't be expanded + # they are just returned. In that case use the empty string which + # use the filescheme. + if path is 'default-push' or path is 'default': + path = '' + + # The path could be a scheme so use Mercurial's normal functionality + # to resolve the scheme to a repository and use its path + if path: + path = hg.repository(ui, path).path + + match = url_.scheme_re.match(path) + if not match: # regular filesystem path + scheme = 'file' + else: + scheme = match.group(1) + + try: + (mod, klass) = _store_provider[scheme] + except KeyError: + raise util.Abort(_('unsupported URL scheme %r') % scheme) + + klass = getattr(mod, klass) + return klass(ui, repo, path)
Change 1 of 1 Show Entire File kbfiles/​bfcommands.py Stacked
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
@@ -0,0 +1,523 @@
+'''High-level command functions: bfadd() et. al, plus the cmdtable.''' + +import os +import shutil + +from mercurial import util, match as match_, hg, node, context, error +from mercurial.i18n import _ + +import bfutil, basestore + +# -- Commands ---------------------------------------------------------- + +def bfconvert(ui, src, dest, *pats, **opts): + '''Convert a repository to a repository using bfiles + + Convert source repository creating an identical + repository, except that all files that match the + patterns given, or are over a given size will + be added as bfiles. The size of a file is the size of the + first version of the file. After running this command you + will need to set the store then run bfput on the new + repository to upload the bfiles to the central store. + ''' + + if opts['tonormal']: + tobfile = False + else: + tobfile = True + size = opts['size'] + if not size: + size = ui.config(bfutil.long_name, 'size', default=None) + try: + size = int(size) + except ValueError: + raise util.Abort(_('bfiles.size must be integer, was %s\n') % size) + except TypeError: + raise util.Abort(_('size must be specified')) + + try: + rsrc = hg.repository(ui, src) + if not rsrc.local(): + raise util.Abort(_('%s is not a local Mercurial repo') % src) + except error.RepoError, err: + ui.traceback() + raise util.Abort(err.args[0]) + if os.path.exists(dest): + if not os.path.isdir(dest): + raise util.Abort(_('destination %s already exists') % dest) + elif os.listdir(dest): + raise util.Abort(_('destination %s is not empty') % dest) + try: + ui.status(_('initializing destination %s\n') % dest) + rdst = hg.repository(ui, dest, create=True) + if not rdst.local(): + raise util.Abort(_('%s is not a local Mercurial repo') % dest) + except error.RepoError: + ui.traceback() + raise util.Abort(_('%s is not a repo') % dest) + + try: + # Lock destination to prevent modification while it is converted to. + # Don't need to lock src because we are just reading from its history + # which can't change. + dst_lock = rdst.lock() + + # Get a list of all changesets in the source. The easy way to do this + # is to simply walk the changelog, using changelog.nodesbewteen(). + # Take a look at mercurial/revlog.py:639 for more details. + # Use a generator instead of a list to decrease memory usage + ctxs = (rsrc[ctx] for ctx in rsrc.changelog.nodesbetween(None, rsrc.heads())[0]) + revmap = {node.nullid: node.nullid} + if tobfile: + bfiles = set() + normalfiles = set() + if not pats: + pats = ui.config(bfutil.long_name, 'patterns', default=()) + if pats: + pats = pats.split(' ') + if pats: + matcher = match_.match(rsrc.root, '', list(pats)) + else: + matcher = None + + bfiletohash = {} + for ctx in ctxs: + ui.progress(_('Converting revisions'), ctx.rev(), unit=_('revision'), total=rsrc['tip'].rev()) + _bfconvert_addchangeset(rsrc, rdst, ctx, revmap, + bfiles, normalfiles, matcher, size, bfiletohash) + ui.progress(_('Converting revisions'), None) + + if os.path.exists(rdst.wjoin(bfutil.short_name)): + shutil.rmtree(rdst.wjoin(bfutil.short_name)) + + for f in bfiletohash.keys(): + if os.path.isfile(rdst.wjoin(f)): + os.unlink(rdst.wjoin(f)) + try: + os.removedirs(os.path.dirname(rdst.wjoin(f))) + except: + pass + + else: + for ctx in ctxs: + ui.progress(_('Converting revisions'), ctx.rev(), unit=_('revision'), total=rsrc['tip'].rev()) + _addchangeset(ui, rsrc, rdst, ctx, revmap) + + ui.progress(_('Converting revisions'), None) + except: + # we failed, remove the new directory + shutil.rmtree(rdst.root) + raise + finally: + dst_lock.release() + +def _addchangeset(ui, rsrc, rdst, ctx, revmap): + # Convert src parents to dst parents + parents = [] + for p in ctx.parents(): + parents.append(revmap[p.node()]) + while len(parents) < 2: + parents.append(node.nullid) + + # Generate list of changed files + files = set(ctx.files()) + if node.nullid not in parents: + mc = ctx.manifest() + mp1 = ctx.parents()[0].manifest() + mp2 = ctx.parents()[1].manifest() + for f in mp1: + if f not in mc: + files.add(f) + for f in mp2: + if f not in mc: + files.add(f) + for f in mc: + if mc[f] != mp1.get(f, None) or mc[f] != mp2.get(f, None): + files.add(f) + + def getfilectx(repo, memctx, f): + if bfutil.standin(f) in files: + # if the file isn't in the manifest then it was removed + # or renamed, raise IOError to indicate this + try: + fctx = ctx.filectx(bfutil.standin(f)) + except error.LookupError: + raise IOError() + renamed = fctx.renamed() + if renamed: + renamed = bfutil.split_standin(renamed[0]) + + hash = fctx.data().strip() + path = bfutil.find_file(rsrc, hash) + ### TODO: What if the file is not cached? + data = '' + with open(path, 'rb') as fd: + data = fd.read() + return context.memfilectx(f, data, 'l' in fctx.flags(), + 'x' in fctx.flags(), renamed) + else: + try: + fctx = ctx.filectx(f) + except error.LookupError: + raise IOError() + renamed = fctx.renamed() + if renamed: + renamed = renamed[0] + data = fctx.data() + if f == '.hgtags': + newdata = [] + for line in data.splitlines(): + id, name = line.split(' ', 1) + newdata.append('%s %s\n' % (node.hex(revmap[node.bin(id)]), name)) + data = ''.join(newdata) + return context.memfilectx(f, data, 'l' in fctx.flags(), + 'x' in fctx.flags(), renamed) + + dstfiles = [] + for file in files: + if bfutil.is_standin(file): + dstfiles.append(bfutil.split_standin(file)) + else: + dstfiles.append(file) + # Commit + mctx = context.memctx(rdst, parents, ctx.description(), dstfiles, + getfilectx, ctx.user(), ctx.date(), ctx.extra()) + ret = rdst.commitctx(mctx) + rdst.dirstate.setparents(ret) + revmap[ctx.node()] = rdst.changelog.tip() + +def _bfconvert_addchangeset(rsrc, rdst, ctx, revmap, bfiles, normalfiles, matcher, size, bfiletohash): + # Convert src parents to dst parents + parents = [] + for p in ctx.parents(): + parents.append(revmap[p.node()]) + while len(parents) < 2: + parents.append(node.nullid) + + # Generate list of changed files + files = set(ctx.files()) + if node.nullid not in parents: + mc = ctx.manifest() + mp1 = ctx.parents()[0].manifest() + mp2 = ctx.parents()[1].manifest() + for f in mp1: + if f not in mc: + files.add(f) + for f in mp2: + if f not in mc: + files.add(f) + for f in mc: + if mc[f] != mp1.get(f, None) or mc[f] != mp2.get(f, None): + files.add(f) + + dstfiles = [] + for f in files: + if f not in bfiles and f not in normalfiles: + isbfile = _is_bfile(f, ctx, matcher, size) + # If this file was renamed or copied then copy + # the bfileness of its predecessor + if f in ctx.manifest(): + fctx = ctx.filectx(f) + renamed = fctx.renamed() + renamedbfile = renamed and renamed[0] in bfiles + isbfile |= renamedbfile + if 'l' in fctx.flags(): + if renamedbfile: + raise util.Abort(_('Renamed/copied bfile %s becomes symlink') % f) + isbfile = False + if isbfile: + bfiles.add(f) + else: + normalfiles.add(f) + + if f in bfiles: + dstfiles.append(bfutil.standin(f)) + # bfile in manifest if it has not been removed/renamed + if f in ctx.manifest(): + if 'l' in ctx.filectx(f).flags(): + if renamed and renamed[0] in bfiles: + raise util.Abort(_('bfile %s becomes symlink') % f) + + # bfile was modified, update standins + fullpath = rdst.wjoin(f) + bfutil.create_dir(os.path.dirname(fullpath)) + m = util.sha1('') + m.update(ctx[f].data()) + hash = m.hexdigest() + if f not in bfiletohash or bfiletohash[f] != hash: + with open(fullpath, 'wb') as fd: + fd.write(ctx[f].data()) + executable = 'x' in ctx[f].flags() + os.chmod(fullpath, bfutil.get_mode(executable)) + bfutil.write_standin(rdst, bfutil.standin(f), hash, executable) + bfiletohash[f] = hash + else: + # normal file + dstfiles.append(f) + + def getfilectx(repo, memctx, f): + if bfutil.is_standin(f): + # if the file isn't in the manifest then it was removed + # or renamed, raise IOError to indicate this + srcfname = bfutil.split_standin(f) + try: + fctx = ctx.filectx(srcfname) + except error.LookupError: + raise IOError() + renamed = fctx.renamed() + if renamed: + # standin is always a bfile because bfileness + # doesn't change after rename or copy + renamed = bfutil.standin(renamed[0]) + + return context.memfilectx(f, bfiletohash[srcfname], 'l' in fctx.flags(), + 'x' in fctx.flags(), renamed) + else: + try: + fctx = ctx.filectx(f) + except error.LookupError: + raise IOError() + renamed = fctx.renamed() + if renamed: + renamed = renamed[0] + + data = fctx.data() + if f == '.hgtags': + newdata = [] + for line in data.splitlines(): + id, name = line.split(' ', 1) + newdata.append('%s %s\n' % (node.hex(revmap[node.bin(id)]), name)) + data = ''.join(newdata) + return context.memfilectx(f, data, 'l' in fctx.flags(), + 'x' in fctx.flags(), renamed) + + # Commit + mctx = context.memctx(rdst, parents, ctx.description(), dstfiles, + getfilectx, ctx.user(), ctx.date(), ctx.extra()) + ret = rdst.commitctx(mctx) + rdst.dirstate.setparents(ret) + revmap[ctx.node()] = rdst.changelog.tip() + +def _is_bfile(file, ctx, matcher, size): + ''' + A file is a bfile if it matches a pattern or is over + the given size. + ''' + # Never store hgtags or hgignore as bfiles + if file == '.hgtags' or file == '.hgignore' or file == '.hgsigs': + return False + if matcher and matcher(file): + return True + try: + return ctx.filectx(file).size() >= size * 1024 * 1024 + except error.LookupError: + return False + +def upload_bfiles(ui, rsrc, rdst, files): + '''upload big files to the central store''' + + if not files: + return + + # Don't upload locally. All bfiles are in the system wide cache + # so the other repo can just get them from there. + if not rdst.path.startswith('http'): + return + + store = basestore._open_store(rsrc, rdst.path, put=True) + + at = 0 + for hash in files: + ui.progress(_('Uploading bfiles'), at, unit='bfile', total=len(files)) + if store.exists(hash): + at += 1 + continue + source = bfutil.find_file(rsrc, hash) + if not source: + raise util.Abort(_('Missing bfile %s needs to be uploaded') % hash) + # XXX check for errors here + store.put(source, hash) + at += 1 + ui.progress('Uploading bfiles', None) + +def verify_bfiles(ui, repo, all=False, contents=False): + '''Verify that every big file revision in the current changeset + exists in the central store. With --contents, also verify that + the contents of each big file revision are correct (SHA-1 hash + matches the revision ID). With --all, check every changeset in + this repository.''' + if all: + # Pass a list to the function rather than an iterator because we know a list will work. + revs = range(len(repo)) + else: + revs = ['.'] + + store = basestore._open_store(repo) + return store.verify(revs, contents=contents) + +def revert_bfiles(ui, repo): + wlock = repo.wlock() + try: + bfdirstate = bfutil.open_bfdirstate(ui, repo) + s = bfdirstate.status(match_.always(repo.root, repo.getcwd()), [], False, False, False) + (unsure, modified, added, removed, missing, unknown, ignored, clean) = s + + bfiles = bfutil.list_bfiles(repo) + toget = [] + at = 0 + updated = 0 + for bfile in bfiles: + if not os.path.exists(repo.wjoin(bfutil.standin(bfile))): + bfdirstate.remove(bfile) + continue + if os.path.exists(repo.wjoin(bfutil.standin(os.path.join(bfile + '.orig')))): + shutil.copyfile(repo.wjoin(bfile), repo.wjoin(bfile + '.orig')) + at += 1 + expectedhash = repo[None][bfutil.standin(bfile)].data().strip() + mode = os.stat(repo.wjoin(bfutil.standin(bfile))).st_mode + if not os.path.exists(repo.wjoin(bfile)) or expectedhash != bfutil.hashfile(repo.wjoin(bfile)): + path = bfutil.find_file(repo, expectedhash) + if path is None: + toget.append((bfile, expectedhash)) + else: + util.makedirs(os.path.dirname(repo.wjoin(bfile))) + shutil.copy(path, repo.wjoin(bfile)) + os.chmod(repo.wjoin(bfile), mode) + updated += 1 + if bfutil.standin(bfile) not in repo['.']: + bfdirstate.add(bfutil.unixpath(bfile)) + elif expectedhash == repo['.'][bfutil.standin(bfile)].data().strip(): + bfdirstate.normal(bfutil.unixpath(bfile)) + else: + bfutil.dirstate_normaldirty(bfdirstate, bfutil.unixpath(bfile)) + elif os.path.exists(repo.wjoin(bfile)) and mode != os.stat(repo.wjoin(bfile)).st_mode: + os.chmod(repo.wjoin(bfile), mode) + updated += 1 + if bfutil.standin(bfile) not in repo['.']: + bfdirstate.add(bfutil.unixpath(bfile)) + elif expectedhash == repo['.'][bfutil.standin(bfile)].data().strip(): + bfdirstate.normal(bfutil.unixpath(bfile)) + else: + bfutil.dirstate_normaldirty(bfdirstate, bfutil.unixpath(bfile)) + + if toget: + store = basestore._open_store(repo) + (success, missing) = store.get(toget) + else: + success, missing = [], [] + + for (filename, hash) in success: + mode = os.stat(repo.wjoin(bfutil.standin(filename))).st_mode + os.chmod(repo.wjoin(filename), mode) + updated += 1 + if bfutil.standin(filename) not in repo['.']: + bfdirstate.add(bfutil.unixpath(filename)) + elif hash == repo['.'][bfutil.standin(filename)].data().strip(): + bfdirstate.normal(bfutil.unixpath(filename)) + else: + bfutil.dirstate_normaldirty(bfdirstate, bfutil.unixpath(filename)) + + removed = 0 + for bfile in bfdirstate: + if not os.path.exists(repo.wjoin(bfutil.standin(bfile))): + if os.path.exists(repo.wjoin(bfile)): + os.unlink(repo.wjoin(bfile)) + removed += 1 + if bfutil.standin(bfile) in repo['.']: + bfdirstate.remove(bfutil.unixpath(bfile)) + else: + bfdirstate.forget(bfutil.unixpath(bfile)) + else: + state = repo.dirstate[bfutil.standin(bfile)] + if state == 'n': + bfdirstate.normal(bfile) + elif state == 'r': + bfdirstate.remove(bfile) + elif state == 'a': + bfdirstate.add(bfile) + elif state == '?': + bfdirstate.forget(bfile) + bfdirstate.write() + finally: + wlock.release() + +def update_bfiles(ui, repo): + wlock = repo.wlock() + try: + bfdirstate = bfutil.open_bfdirstate(ui, repo) + s = bfdirstate.status(match_.always(repo.root, repo.getcwd()), [], False, False, False) + (unsure, modified, added, removed, missing, unknown, ignored, clean) = s + + bfiles = bfutil.list_bfiles(repo) + toget = [] + at = 0 + updated = 0 + removed = 0 + printed = False + if bfiles: + ui.status(_('Getting changed bfiles\n')) + printed = True + + for bfile in bfiles: + at += 1 + if os.path.exists(repo.wjoin(bfile)) and not os.path.exists(repo.wjoin(bfutil.standin(bfile))): + os.unlink(repo.wjoin(bfile)) + removed += 1 + bfdirstate.forget(bfutil.unixpath(bfile)) + continue + expectedhash = repo[None][bfutil.standin(bfile)].data().strip() + mode = os.stat(repo.wjoin(bfutil.standin(bfile))).st_mode + if not os.path.exists(repo.wjoin(bfile)) or expectedhash != bfutil.hashfile(repo.wjoin(bfile)): + path = bfutil.find_file(repo, expectedhash) + if not path: + toget.append((bfile, expectedhash)) + else: + util.makedirs(os.path.dirname(repo.wjoin(bfile))) + shutil.copy(path, repo.wjoin(bfile)) + os.chmod(repo.wjoin(bfile), mode) + updated += 1 + bfdirstate.normal(bfutil.unixpath(bfile)) + elif os.path.exists(repo.wjoin(bfile)) and mode != os.stat(repo.wjoin(bfile)).st_mode: + os.chmod(repo.wjoin(bfile), mode) + updated += 1 + bfdirstate.normal(bfutil.unixpath(bfile)) + + if toget: + store = basestore._open_store(repo) + (success, missing) = store.get(toget) + else: + success, missing = [],[] + + for (filename, hash) in success: + mode = os.stat(repo.wjoin(bfutil.standin(filename))).st_mode + os.chmod(repo.wjoin(filename), mode) + updated += 1 + bfdirstate.normal(bfutil.unixpath(filename)) + + for bfile in bfdirstate: + if bfile not in bfiles: + if os.path.exists(repo.wjoin(bfile)): + if not printed: + ui.status(_('Getting changed bfiles\n')) + printed = True + os.unlink(repo.wjoin(bfile)) + removed += 1 + bfdirstate.forget(bfutil.unixpath(bfile)) + + bfdirstate.write() + if printed: + ui.status(_('%d big files updated, %d removed\n') % (updated, removed)) + finally: + wlock.release() + +# -- hg commands declarations ------------------------------------------------ + + +cmdtable = { + 'kbfconvert': (bfconvert, + [('s', 'size', 0, 'All files over this size ' + '(in megabytes) will be considered bfiles. This can also be specified in your hgrc as [bfiles].size.'), + ('','tonormal',False, 'Convert from a bfiles repo to a normal repo')], + _('hg kbfconvert SOURCE DEST [FILE ...]')), + }
Change 1 of 1 Show Entire File kbfiles/​bfsetup.py Stacked
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
@@ -0,0 +1,1044 @@
+'''Setup code for bfiles extension: reposetup(), uisetup().''' + +import os +import types +import copy + +from mercurial import hg, extensions, commands, util, context, cmdutil, \ + match as match_, filemerge, node, archival, httprepo, error +from mercurial.i18n import _ +from mercurial.node import hex +import bfutil, bfcommands + +def hgversion(): + from mercurial.__version__ import version + return [int(n) for n in version.partition('+')[0].split('.')] +hgversion = hgversion() + +# -- Wrappers: modify existing commands -------------------------------- + +def reposetup(ui, repo): + # add a kbfiles-specific querystring argument to remote requests, so kiln can reject + # operations on a kbfiles-enabled remote repo from a non-kbfiles local repo. + if issubclass(repo.__class__, httprepo.httprepository): + class kbfilesrepo(repo.__class__): + # The function we want to override is do_cmd for Mercurial <= 1.6 + # and _callstream for Mercurial > 1.6. Wrap whichever one we can find. + if hasattr(repo.__class__, 'do_cmd'): + def do_cmd(self, cmd, **args): + args['kbfiles'] = 'true' + return super(repo.__class__, self).do_cmd(cmd, **args) + if hasattr(repo.__class__, '_callstream'): + def _callstream(self, cmd, **args): + args['kbfiles'] = 'true' + return super(repo.__class__, self)._callstream(cmd, **args) + repo.__class__ = kbfilesrepo + + # bfiles doesn't support non-local repositories -- get out quick in + # such a case + if not repo.local(): + return + + for name in ('status', 'commitctx', 'commit', 'push'): + method = getattr(repo, name) + #if not (isinstance(method, types.MethodType) and + # method.im_func is repo.__class__.commitctx.im_func): + if isinstance(method, types.FunctionType) and method.func_name == 'wrap': + ui.warn(_('kbfiles: repo method %r appears to have already been ' + 'wrapped by another extension: ' + 'kbfiles may behave incorrectly\n') + % name) + + class bfiles_repo(repo.__class__): + bfstatus = False + def status_nobfiles(self, *args, **kwargs): + return super(bfiles_repo, self).status(*args, **kwargs) + + # Figure out the status of big files and insert them into the + # appropriate list in the result. Also removes standin files from + # the listing. This function reverts to the original status if + # self.bfstatus is False + def status(self, node1='.', node2=None, match=None, ignored=False, clean=False, unknown=False, subrepos=None): + listignored, listclean, listunknown = ignored, clean, unknown + if not self.bfstatus: + try: + return super(bfiles_repo, self).status(node1, node2, match, listignored, listclean, listunknown, subrepos) + except TypeError: + return super(bfiles_repo, self).status(node1, node2, match, listignored, listclean, listunknown) + else: + # some calls in this function rely on the old version of status + self.bfstatus = False + if isinstance(node1, context.changectx): + ctx1 = node1 + else: + ctx1 = repo[node1] + if isinstance(node2, context.changectx): + ctx2 = node2 + else: + ctx2 = repo[node2] + working = ctx2.rev() is None + parentworking = working and ctx1 == self['.'] + + def inctx(file, ctx): + try: + if ctx.rev() is None: + return file in ctx.manifest() + ctx[file] + return True + except: + return False + + # create a copy of match that matches standins instead of bfiles + # if matcher not set then it is the always matcher so overwrite that + if match is None: + match = match_.always(self.root, self.getcwd()) + + def tostandin(file): + if inctx(bfutil.standin(file), ctx2): + return bfutil.standin(file) + return file + + m = copy.copy(match) + m._files = [tostandin(f) for f in m._files] + + # get ignored clean and unknown but remove them later if they were not asked for + try: + result = super(bfiles_repo, self).status(node1, node2, m, True, True, True, subrepos) + except TypeError: + result = super(bfiles_repo, self).status(node1, node2, m, True, True, True) + if working: + # Hold the wlock while we read bfiles and update the bfdirstate + wlock = repo.wlock() + try: + # Any non bfiles that were explicitly listed must be taken out or + # bfdirstate.status will report an error. The status of these files + # was already computed using super's status. + bfdirstate = bfutil.open_bfdirstate(ui, self) + match._files = [f for f in match._files if f in bfdirstate] + s = bfdirstate.status(match, [], listignored, listclean, listunknown) + (unsure, modified, added, removed, missing, unknown, ignored, clean) = s + if parentworking: + for bfile in unsure: + if ctx1[bfutil.standin(bfile)].data().strip() != bfutil.hashfile(self.wjoin(bfile)): + modified.append(bfile) + else: + clean.append(bfile) + bfdirstate.normal(bfutil.unixpath(bfile)) + bfdirstate.write() + else: + tocheck = unsure + modified + added + clean + modified, added, clean = [], [], [] + + for bfile in tocheck: + standin = bfutil.standin(bfile) + if inctx(standin, ctx1): + if ctx1[standin].data().strip() != bfutil.hashfile(self.wjoin(bfile)): + modified.append(bfile) + else: + clean.append(bfile) + else: + added.append(bfile) + finally: + wlock.release() + + for standin in ctx1.manifest(): + if not bfutil.is_standin(standin): + continue + bfile = bfutil.split_standin(standin) + if not match(bfile): + continue + if bfile not in bfdirstate: + removed.append(bfile) + # Handle unknown and ignored differently + bfiles = (modified, added, removed, missing, [], [], clean) + result = list(result) + # Unknown files + result[4] = [f for f in unknown if repo.dirstate[f] == '?' and not bfutil.is_standin(f)] + # Ignored files must be ignored by both the dirstate and bfdirstate + result[5] = set(ignored).intersection(set(result[5])) + # combine normal files and bfiles + normals = [[fn for fn in filelist if not bfutil.is_standin(fn)] for filelist in result] + result = [sorted(list1 + list2) for (list1, list2) in zip(normals, bfiles)] + else: + def toname(f): + if bfutil.is_standin(f): + return bfutil.split_standin(f) + return f + result = [[toname(f) for f in items] for items in result] + + if not listunknown: + result[4] = [] + if not listignored: + result[5] = [] + if not listclean: + result[6] = [] + self.bfstatus = True + return result + + # This call happens after a commit has occurred. Copy all of the bfiles + # into the cache + def commitctx(self, *args, **kwargs): + node = super(bfiles_repo, self).commitctx(*args, **kwargs) + ctx = self[node] + for filename in ctx.files(): + if bfutil.is_standin(filename) and filename in ctx.manifest(): + realfile = bfutil.split_standin(filename) + bfutil.copy_to_cache(self, ctx.node(), realfile) + + return node + + # This call happens before a commit has occurred. The bfile standins + # have not had their contents updated (to reflect the hash of their bfile). + # Do that here. + def commit(self, text="", user=None, date=None, match=None, force=False, + editor=False, extra={}): + orig = super(bfiles_repo, self).commit + + wlock = repo.wlock() + try: + # Case 1: user calls commit with no specific files or + # include/exclude patterns: refresh and commit everything. + if (match is None) or (not match.anypats() and not match.files()): + bfiles = bfutil.list_bfiles(self) + bfdirstate = bfutil.open_bfdirstate(ui, self) + # this only loops through bfiles that exist (not removed/renamed) + for bfile in bfiles: + if os.path.exists(self.wjoin(bfutil.standin(bfile))): + bfutil.update_standin(self, bfutil.standin(bfile)) + bfdirstate.normal(bfutil.unixpath(bfile)) + for bfile in bfdirstate: + if not os.path.exists(repo.wjoin(bfutil.standin(bfile))): + bfdirstate.forget(bfutil.unixpath(bfile)) + bfdirstate.write() + + return orig(text=text, user=user, date=date, match=match, + force=force, editor=editor, extra=extra) + + for file in match.files(): + if bfutil.is_standin(file): + raise util.Abort("Don't commit bfile standin. Commit bfile.") + + # Case 2: user calls commit with specified patterns: refresh any + # matching big files. + smatcher = bfutil.compose_standin_matcher(self, match) + standins = bfutil.dirstate_walk(self.dirstate, smatcher) + + # No matching big files: get out of the way and pass control to + # the usual commit() method. + if not standins: + return orig(text=text, user=user, date=date, match=match, + force=force, editor=editor, extra=extra) + + # Refresh all matching big files. It's possible that the commit + # will end up failing, in which case the big files will stay + # refreshed. No harm done: the user modified them and asked to + # commit them, so sooner or later we're going to refresh the + # standins. Might as well leave them refreshed. + bfdirstate = bfutil.open_bfdirstate(ui, self) + for standin in standins: + bfile = bfutil.split_standin(standin) + if bfdirstate[bfile] is not 'r': + bfutil.update_standin(self, standin) + bfdirstate.normal(bfutil.unixpath(bfile)) + else: + bfdirstate.forget(bfutil.unixpath(bfile)) + bfdirstate.write() + + # Cook up a new matcher that only matches regular files or + # standins corresponding to the big files requested by the user. + # Have to modify _files to prevent commit() from complaining + # "not tracked" for big files. + bfiles = bfutil.list_bfiles(repo) + match = copy.copy(match) + orig_matchfn = match.matchfn + + # Check both the list of bfiles and the list of standins because if a bfile was removed, it + # won't be in the list of bfiles at this point + match._files += sorted(standins) + + actualfiles = [] + for f in match._files: + fstandin = bfutil.standin(f) + + # Ignore known bfiles and standins + if f in bfiles or fstandin in standins: + continue + + # Append directory separator to avoid collisions + if not fstandin.endswith('/'): + fstandin += '/' + + # Prevalidate matching standin directories + if any(st for st in match._files if st.startswith(fstandin)): + continue + actualfiles.append(f) + match._files = actualfiles + + def matchfn(f): + if orig_matchfn(f): + return f not in bfiles + else: + return f in standins + + match.matchfn = matchfn + return orig(text=text, user=user, date=date, match=match, + force=force, editor=editor, extra=extra) + finally: + wlock.release() + + def push(self, remote, force=False, revs=None, newbranch=False): + o = bfutil.findoutgoing(repo, remote, force) + if o: + toupload = set() + o = repo.changelog.nodesbetween(o, revs)[0] + for n in o: + parents = [p for p in repo.changelog.parents(n) if p != node.nullid] + ctx = repo[n] + files = set(ctx.files()) + if len(parents) == 2: + mc = ctx.manifest() + mp1 = ctx.parents()[0].manifest() + mp2 = ctx.parents()[1].manifest() + for f in mp1: + if f not in mc: + files.add(f) + for f in mp2: + if f not in mc: + files.add(f) + for f in mc: + if mc[f] != mp1.get(f, None) or mc[f] != mp2.get(f, None): + files.add(f) + + toupload = toupload.union(set([ctx[f].data().strip() for f in files if bfutil.is_standin(f) and f in ctx])) + bfcommands.upload_bfiles(ui, self, remote, toupload) + # Mercurial >= 1.6 takes the newbranch argument, try that first. + try: + return super(bfiles_repo, self).push(remote, force, revs, newbranch) + except TypeError: + return super(bfiles_repo, self).push(remote, force, revs) + + repo.__class__ = bfiles_repo + +# Add works by going through the files that the user wanted to add +# and checking if they should be added as bfiles. Then making a new +# matcher which matches only the normal files and running the original +# version of add. +def override_add(orig, ui, repo, *pats, **opts): + bf = opts.pop('bf', None) + + bfsize = opts.pop('bfsize', None) + if bfsize: + try: + bfsize = int(bfsize) + except ValueError: + raise util.Abort(_('size must be an integer, was %s\n') % bfsize) + else: + if os.path.exists(repo.wjoin(bfutil.short_name)): + bfsize = ui.config(bfutil.long_name, 'size', default='10') + if bfsize: + try: + bfsize = int(bfsize) + except ValueError: + raise util.Abort(_('bfiles.size must be integer, was %s\n') % bfsize) + + bfmatcher = None + if os.path.exists(repo.wjoin(bfutil.short_name)): + bfpats = ui.config(bfutil.long_name, 'patterns', default=()) + if bfpats: + bfpats = bfpats.split(' ') + bfmatcher = match_.match(repo.root, '', list(bfpats)) + + bfnames = [] + m = cmdutil.match(repo, pats, opts) + m.bad = lambda x,y: None + wctx = repo[None] + for f in repo.walk(m): + exact = m.exact(f) + bfile = bfutil.standin(f) in wctx + nfile = f in wctx + + if exact and bfile: + ui.warn(_('%s already a bfile\n') % f) + continue + # Don't warn the user when they attempt to add a normal tracked file. The normal add code + # will do that for us. + if exact and nfile: + continue + if exact or (not bfile and not nfile): + if bf or (bfsize and os.path.getsize(repo.wjoin(f)) >= bfsize*1024*1024) \ + or (bfmatcher and bfmatcher(f)): + bfnames.append(f) + if ui.verbose or not exact: + ui.status(_('adding %s as bfile\n') % m.rel(f)) + + bad = [] + standins = [] + + # Need to lock otherwise there could be a race condition inbetween when standins are created + # and added to the repo + wlock = repo.wlock() + try: + if not opts.get('dry_run'): + bfdirstate = bfutil.open_bfdirstate(ui, repo) + for f in bfnames: + standinname = bfutil.standin(f) + bfutil.write_standin(repo, standinname, hash='', executable=bfutil.get_executable(repo.wjoin(f))) + standins.append(standinname) + if bfdirstate[bfutil.unixpath(f)] == 'r': + bfdirstate.normallookup(bfutil.unixpath(f)) + else: + bfdirstate.add(bfutil.unixpath(f)) + bfdirstate.write() + bad += [bfutil.split_standin(f) for f in bfutil.repo_add(repo, standins) if f in m.files()] + finally: + wlock.release() + + oldmatch = cmdutil.match + manifest = repo[None].manifest() + def override_match(repo, pats=[], opts={}, globbed=False, default='relpath'): + match = oldmatch(repo, pats, opts, globbed, default) + m = copy.copy(match) + notbfile = lambda f: not bfutil.is_standin(f) and bfutil.standin(f) not in manifest + m._files = [f for f in m._files if notbfile(f)] + m._fmap = set(m._files) + orig_matchfn = m.matchfn + m.matchfn = lambda f: notbfile(f) and orig_matchfn(f) or None + return m + cmdutil.match = override_match + result = orig(ui, repo, *pats, **opts) + cmdutil.match = oldmatch + + return (result is 1 or bad) and 1 or 0 + +def override_remove(orig, ui, repo, *pats, **opts): + wctx = repo[None].manifest() + oldmatch = cmdutil.match + def override_match(repo, pats=[], opts={}, globbed=False, default='relpath'): + match = oldmatch(repo, pats, opts, globbed, default) + m = copy.copy(match) + notbfile = lambda f: not bfutil.is_standin(f) and bfutil.standin(f) not in wctx + m._files = [f for f in m._files if notbfile(f)] + m._fmap = set(m._files) + orig_matchfn = m.matchfn + m.matchfn = lambda f: orig_matchfn(f) and notbfile(f) + return m + cmdutil.match = override_match + orig(ui, repo, *pats, **opts) + cmdutil.match = oldmatch + + after, force = opts.get('after'), opts.get('force') + if not pats and not after: + raise util.Abort(_('no files specified')) + m = cmdutil.match(repo, pats, opts) + try: + repo.bfstatus = True + s = repo.status(match=m, clean=True) + finally: + repo.bfstatus = False + modified, added, deleted, clean = [[f for f in list if bfutil.standin(f) in wctx] for list in [s[0], s[1], s[3], s[6]]] + + def warn(files, reason): + for f in files: + ui.warn(_('not removing %s: file %s (use -f to force removal)\n') + % (m.rel(f), reason)) + + if force: + remove, forget = modified + deleted + clean, added + elif after: + remove, forget = deleted, [] + warn(modified + added + clean, _('still exists')) + else: + remove, forget = deleted + clean, [] + warn(modified, _('is modified')) + warn(added, _('has been marked for add')) + + for f in sorted(remove + forget): + if ui.verbose or not m.exact(f): + ui.status(_('removing %s\n') % m.rel(f)) + + # Need to lock because standin files are deleted then removed from the repository + # and we could race inbetween. + wlock = repo.wlock() + try: + bfdirstate = bfutil.open_bfdirstate(ui, repo) + for f in remove: + if not after: + os.unlink(repo.wjoin(f)) + currentdir = os.path.split(f)[0] + while currentdir and not os.listdir(repo.wjoin(currentdir)): + os.rmdir(repo.wjoin(currentdir)) + currentdir = os.path.split(currentdir)[0] + bfdirstate.remove(bfutil.unixpath(f)) + bfdirstate.write() + + forget = [bfutil.standin(f) for f in forget] + remove = [bfutil.standin(f) for f in remove] + bfutil.repo_forget(repo, forget) + bfutil.repo_remove(repo, remove, unlink=True) + finally: + wlock.release() + +def override_status(orig, ui, repo, *pats, **opts): + try: + repo.bfstatus = True + return orig(ui, repo, *pats, **opts) + finally: + repo.bfstatus = False + +def override_verify(orig, ui, repo, *pats, **opts): + bf = opts.pop('bf', False) + all = opts.pop('bfa', False) + contents = opts.pop('bfc', False) + + result = orig(ui, repo, *pats, **opts) + if bf: + result = result or bfcommands.verify_bfiles(ui, repo, all, contents) + return result + +# Override needs to refresh standins so that update's normal merge +# will go through properly. Then the other update hook (overriding repo.update) +# will get the new files. Filemerge is also overriden so that the merge +# will merge standins correctly. +def override_update(orig, ui, repo, *pats, **opts): + bfdirstate = bfutil.open_bfdirstate(ui, repo) + s = bfdirstate.status(match_.always(repo.root, repo.getcwd()), [], False, False, False) + (unsure, modified, added, removed, missing, unknown, ignored, clean) = s + + # Need to lock between the standins getting updated and their bfiles getting updated + wlock = repo.wlock() + try: + if opts['check']: + mod = len(modified) > 0 + for bfile in unsure: + standin = bfutil.standin(bfile) + if repo['.'][standin].data().strip() != bfutil.hashfile(repo.wjoin(bfile)): + mod = True + else: + bfdirstate.normal(bfutil.unixpath(bfile)) + bfdirstate.write() + if mod: + raise util.Abort(_('uncommitted local changes')) + # XXX handle removed differently + if not opts['clean']: + for bfile in unsure + modified + added: + bfutil.update_standin(repo, bfutil.standin(bfile)) + finally: + wlock.release() + return orig(ui, repo, *pats, **opts) + +# Override filemerge to prompt the user about how they wish to merge bfiles. +# This will handle identical edits, and copy/rename + edit without prompting the user. +def override_filemerge(origfn, repo, mynode, orig, fcd, fco, fca): + # Use better variable names here. Because this is a wrapper we cannot change + # the variable names in the function declaration. + fcdest, fcother, fcancestor = fcd, fco, fca + if not bfutil.is_standin(orig): + return origfn(repo, mynode, orig, fcdest, fcother, fcancestor) + else: + if not fcother.cmp(fcdest): # files identical? + return None + + if fcancestor == fcother: # backwards, use working dir parent as ancestor + fcancestor = fcdest.parents()[0] + + if orig != fcother.path(): + repo.ui.status(_('merging %s and %s to %s\n') + % (bfutil.split_standin(orig), bfutil.split_standin(fcother.path()), bfutil.split_standin(fcdest.path()))) + else: + repo.ui.status(_('merging %s\n') % bfutil.split_standin(fcdest.path())) + + if fcancestor.path() != fcother.path() and fcother.data() == fcancestor.data(): + return 0 + if fcancestor.path() != fcdest.path() and fcdest.data() == fcancestor.data(): + repo.wwrite(fcdest.path(), fcother.data(), fcother.flags()) + return 0 + + if repo.ui.promptchoice(_('bfile %s has a merge conflict\n' + 'keep (l)ocal or take (o)ther?') % bfutil.split_standin(orig), + (_('&Local'), _('&Other')), 0) == 0: + return 0 + else: + repo.wwrite(fcdest.path(), fcother.data(), fcother.flags()) + return 0 + +# Copy first changes the matchers to match standins instead of bfiles. +# Then it overrides util.copyfile in that function it checks if the destination +# bfile already exists. It also keeps a list of copied files so that the bfiles +# can be copied and the dirstate updated. +def override_copy(orig, ui, repo, pats, opts, rename=False): + # doesn't remove bfile on rename + if len(pats) < 2: + # this isn't legal, let the original function deal with it + return orig(ui, repo, pats, opts, rename) + + def makestandin(relpath): + return os.path.join(os.path.relpath('.', repo.getcwd()), bfutil.standin(util.canonpath(repo.root, repo.getcwd(), relpath))) + + fullpats = cmdutil.expandpats(pats) + dest = fullpats[-1] + + if os.path.isdir(dest): + if not os.path.isdir(makestandin(dest)): + os.makedirs(makestandin(dest)) + # This could copy both bfiles and normal files in one command, but we don't want + # to do that first replace their matcher to only match normal files and run it + # then replace it to just match bfiles and run it again + nonormalfiles = False + nobfiles = False + oldmatch = cmdutil.match + try: + manifest = repo[None].manifest() + def override_match(repo, pats=[], opts={}, globbed=False, default='relpath'): + match = oldmatch(repo, pats, opts, globbed, default) + m = copy.copy(match) + notbfile = lambda f: not bfutil.is_standin(f) and bfutil.standin(f) not in manifest + m._files = [f for f in m._files if notbfile(f)] + m._fmap = set(m._files) + orig_matchfn = m.matchfn + m.matchfn = lambda f: notbfile(f) and orig_matchfn(f) or None + return m + cmdutil.match = override_match + result = orig(ui, repo, pats, opts, rename) + except util.Abort as e: + if str(e) != 'no files to copy': + raise e + else: + nonormalfiles = True + result = 0 + finally: + cmdutil.match = oldmatch + + # The first rename can cause our current working directory to be removed. In that case + # there is nothing left to copy/rename so just quit. + try: + repo.getcwd() + except OSError: + return result + + try: + # When we call orig below it creates the standins but we don't add them to the dir state + # until later so lock during that time. + wlock = repo.wlock() + + manifest = repo[None].manifest() + def override_match(repo, pats=[], opts={}, globbed=False, default='relpath'): + newpats = [] + # The patterns were previously mangled to add .hgbfiles, we need to remove that now + for pat in pats: + if match_.patkind(pat) == None and bfutil.short_name in pat: + newpats.append(pat.replace( bfutil.short_name, '')) + else: + newpats.append(pat) + match = oldmatch(repo, newpats, opts, globbed, default) + m = copy.copy(match) + bfile = lambda f: bfutil.standin(f) in manifest + m._files = [bfutil.standin(f) for f in m._files if bfile(f)] + m._fmap = set(m._files) + orig_matchfn = m.matchfn + m.matchfn = lambda f: bfutil.is_standin(f) and bfile(bfutil.split_standin(f)) and orig_matchfn(bfutil.split_standin(f)) or None + return m + cmdutil.match = override_match + listpats = [] + for pat in pats: + if match_.patkind(pat) != None: + listpats.append(pat) + else: + listpats.append(makestandin(pat)) + + try: + origcopyfile = util.copyfile + copiedfiles = [] + def override_copyfile(src, dest): + if bfutil.short_name in src and bfutil.short_name in dest: + destbfile = dest.replace(bfutil.short_name, '') + if not opts['force'] and os.path.exists(destbfile): + raise IOError('', _('destination bfile already exists')) + copiedfiles.append((src, dest)) + origcopyfile(src, dest) + + util.copyfile = override_copyfile + result += orig(ui, repo, listpats, opts, rename) + finally: + util.copyfile = origcopyfile + + bfdirstate = bfutil.open_bfdirstate(ui, repo) + for (src, dest) in copiedfiles: + if bfutil.short_name in src and bfutil.short_name in dest: + srcbfile = src.replace(bfutil.short_name, '') + destbfile = dest.replace(bfutil.short_name, '') + destbfiledir = os.path.dirname(destbfile) or '.' + if not os.path.isdir(destbfiledir): + os.makedirs(destbfiledir) + if rename: + os.rename(srcbfile, destbfile) + bfdirstate.remove(bfutil.unixpath(os.path.relpath(srcbfile, repo.root))) + else: + util.copyfile(srcbfile, destbfile) + bfdirstate.add(bfutil.unixpath(os.path.relpath(destbfile, repo.root))) + bfdirstate.write() + except util.Abort as e: + if str(e) != 'no files to copy': + raise e + else: + nobfiles = True + finally: + cmdutil.match = oldmatch + wlock.release() + + if nobfiles and nonormalfiles: + raise util.Abort(_('no files to copy')) + + return result + +# Standins are only updated (to match the hash of bfiles) before commits. +# Update the standins then run the original revert (changing the matcher to hit standins +# instead of bfiles). Based on the resulting standins update the bfiles. Then return the +# standins to their proper state +def override_revert(orig, ui, repo, *pats, **opts): + # Because we put the standins in a bad state (by updating them) and then return them + # to a correct state we need to lock to prevent others from changing them in their + # incorrect state. + wlock = repo.wlock() + try: + bfdirstate = bfutil.open_bfdirstate(ui, repo) + (modified, added, removed, missing, unknown, ignored, clean) = bfutil.bfdirstate_status(bfdirstate, repo, repo['.'].rev()) + for bfile in modified: + bfutil.update_standin(repo, bfutil.standin(bfile)) + + oldmatch = cmdutil.match + try: + ctx = repo[opts.get('rev')] + def override_match(repo, pats=[], opts={}, globbed=False, default='relpath'): + match = oldmatch(repo, pats, opts, globbed, default) + m = copy.copy(match) + def tostandin(f): + if bfutil.standin(f) in repo[None] or bfutil.standin(f) in ctx: + return bfutil.standin(f) + return f + m._files = [tostandin(f) for f in m._files] + m._fmap = set(m._files) + orig_matchfn = m.matchfn + def matchfn(f): + if bfutil.is_standin(f): + return orig_matchfn(bfutil.split_standin(f)) and (f in repo[None] or f in ctx) + return orig_matchfn(f) + m.matchfn = matchfn + return m + cmdutil.match = override_match + orig(ui, repo, *pats, **opts) + finally: + cmdutil.match = oldmatch + bfcommands.revert_bfiles(ui, repo) + for bfile in modified: + if os.path.exists(repo.wjoin(bfutil.standin(bfile))) and bfile in repo['.']: + bfutil.write_standin(repo, bfutil.standin(bfile), repo['.'][bfile].data().strip(), 'x' in repo['.'][bfile].flags()) + finally: + wlock.release() + +def hg_update(orig, repo, node): + result = orig(repo, node) + # XXX check if it worked first + bfcommands.update_bfiles(repo.ui, repo) + return result + +def hg_clean(orig, repo, node, show_stats=True): + result = orig(repo, node, show_stats) + bfcommands.update_bfiles(repo.ui, repo) + return result + +def hg_merge(orig, repo, node, force=None, remind=True): + result = orig(repo, node, force, remind) + bfcommands.update_bfiles(repo.ui, repo) + return result + +# When we rebase a repository with remotely changed bfiles, we need +# to explicitly do a clean update so that the entries in .kbf are +# udpated and the new bfiles are pulled +def override_pull(orig, ui, repo, source="default", **opts): + result = orig(ui, repo, source, **opts) + if opts.get('rebase', False): + commands.update(repo.ui, repo, clean = True) + return result + +def override_archive(orig, repo, dest, node, kind, decode=True, matchfn=None, + prefix=None, mtime=None, subrepos=None): + # No need to lock because we are only reading history and bfile caches + # neither of which are modified + + if kind not in archival.archivers: + raise util.Abort(_("unknown archive type '%s'") % kind) + + ctx = repo[node] + + # In Mercurial <= 1.5 the prefix is passed to the archiver so try that + # if that doesn't work we are probably in Mercurial >= 1.6 where the + # prefix is not handled by the archiver + try: + archiver = archival.archivers[kind](dest, prefix, mtime or ctx.date()[0]) + + def write(name, mode, islink, getdata): + if matchfn and not matchfn(name): + return + data = getdata() + if decode: + data = repo.wwritedata(name, data) + archiver.addfile(name, mode, islink, data) + except TypeError: + if kind == 'files': + if prefix: + raise util.Abort(_('cannot give prefix when archiving to files')) + else: + prefix = archival.tidyprefix(dest, kind, prefix) + + def write(name, mode, islink, getdata): + if matchfn and not matchfn(name): + return + data = getdata() + if decode: + data = repo.wwritedata(name, data) + archiver.addfile(prefix + name, mode, islink, data) + + archiver = archival.archivers[kind](dest, mtime or ctx.date()[0]) + + if repo.ui.configbool("ui", "archivemeta", True): + def metadata(): + base = 'repo: %s\nnode: %s\nbranch: %s\n' % ( + hex(repo.changelog.node(0)), hex(node), ctx.branch()) + + tags = ''.join('tag: %s\n' % t for t in ctx.tags() + if repo.tagtype(t) == 'global') + if not tags: + repo.ui.pushbuffer() + opts = {'template': '{latesttag}\n{latesttagdistance}', + 'style': '', 'patch': None, 'git': None} + cmdutil.show_changeset(repo.ui, repo, opts).show(ctx) + ltags, dist = repo.ui.popbuffer().split('\n') + tags = ''.join('latesttag: %s\n' % t for t in ltags.split(':')) + tags += 'latesttagdistance: %s\n' % dist + + return base + tags + + write('.hg_archival.txt', 0644, False, metadata) + + for f in ctx: + ff = ctx.flags(f) + getdata = ctx[f].data + if bfutil.is_standin(f): + path = bfutil.find_file(repo, getdata().strip()) + ### TODO: What if the file is not cached? + f = bfutil.split_standin(f) + + def getdatafn(): + with open(path, 'rb') as fd: + return fd.read() + + getdata = getdatafn + write(f, 'x' in ff and 0755 or 0644, 'l' in ff, getdata) + archiver.done() + +# If a bfile is modified the change is not reflected in its standin until a commit. +# cmdutil.bail_if_changed raises an exception if the repo has uncommitted changes. +# Wrap it to also check if bfiles were changed. This is used by bisect and backout. +def override_bail_if_changed(orig, repo): + orig(repo) + repo.bfstatus = True + modified, added, removed, deleted = repo.status()[:4] + repo.bfstatus = False + if modified or added or removed or deleted: + raise util.Abort(_('outstanding uncommitted changes')) + +# Fetch doesn't use cmdutil.bail_if_changed so override it to add the check +def override_fetch(orig, ui, repo, *pats, **opts): + repo.bfstatus = True + modified, added, removed, deleted = repo.status()[:4] + repo.bfstatus = False + if modified or added or removed or deleted: + raise util.Abort(_('outstanding uncommitted changes')) + return orig(ui, repo, *pats, **opts) + +def override_forget(orig, ui, repo, *pats, **opts): + wctx = repo[None].manifest() + oldmatch = cmdutil.match + def override_match(repo, pats=[], opts={}, globbed=False, default='relpath'): + match = oldmatch(repo, pats, opts, globbed, default) + m = copy.copy(match) + notbfile = lambda f: not bfutil.is_standin(f) and bfutil.standin(f) not in wctx + m._files = [f for f in m._files if notbfile(f)] + m._fmap = set(m._files) + orig_matchfn = m.matchfn + m.matchfn = lambda f: orig_matchfn(f) and notbfile(f) + return m + cmdutil.match = override_match + orig(ui, repo, *pats, **opts) + cmdutil.match = oldmatch + + m = cmdutil.match(repo, pats, opts) + try: + repo.bfstatus = True + s = repo.status(match=m, clean=True) + finally: + repo.bfstatus = False + forget = sorted(s[0] + s[1] + s[3] + s[6]) + forget = [f for f in forget if bfutil.standin(f) in wctx] + + for f in forget: + if bfutil.standin(f) not in repo.dirstate and not os.path.isdir(m.rel(bfutil.standin(f))): + ui.warn(_('not removing %s: file is already untracked\n') + % m.rel(f)) + + for f in forget: + if ui.verbose or not m.exact(f): + ui.status(_('removing %s\n') % m.rel(f)) + + # Need to lock because standin files are deleted then removed from the repository + # and we could race inbetween. + wlock = repo.wlock() + try: + bfdirstate = bfutil.open_bfdirstate(ui, repo) + for f in forget: + bfdirstate.remove(bfutil.unixpath(f)) + bfdirstate.write() + bfutil.repo_remove(repo, [bfutil.standin(f) for f in forget], unlink=True) + finally: + wlock.release() + +def get_outgoing_bfiles(ui, repo, dest=None, **opts): + dest = ui.expandpath(dest or 'default-push', dest or 'default') + dest, branches = hg.parseurl(dest, opts.get('branch')) + revs, checkout = hg.addbranchrevs(repo, repo, branches, opts.get('rev')) + if revs: + revs = [repo.lookup(rev) for rev in revs] + + # Mercurial <= 1.5 had remoteui in cmdutil, then it moved to hg + try: + remoteui = cmdutil.remoteui + except AttributeError: + remoteui = hg.remoteui + + try: + remote = hg.repository(remoteui(repo, opts), dest) + except error.RepoError: + return None + o = bfutil.findoutgoing(repo, remote, False) + if not o: + return None + o = repo.changelog.nodesbetween(o, revs)[0] + if opts.get('newest_first'): + o.reverse() + + toupload = set() + for n in o: + parents = [p for p in repo.changelog.parents(n) if p != node.nullid] + ctx = repo[n] + files = set(ctx.files()) + if len(parents) == 2: + mc = ctx.manifest() + mp1 = ctx.parents()[0].manifest() + mp2 = ctx.parents()[1].manifest() + for f in mp1: + if f not in mc: + files.add(f) + for f in mp2: + if f not in mc: + files.add(f) + for f in mc: + if mc[f] != mp1.get(f, None) or mc[f] != mp2.get(f, None): + files.add(f) + toupload = toupload.union(set([f for f in files if bfutil.is_standin(f) and f in ctx])) + return toupload + +def override_outgoing(orig, ui, repo, dest=None, **opts): + orig(ui, repo, dest, **opts) + + if opts.pop('bf', None): + toupload = get_outgoing_bfiles(ui, repo, dest, **opts) + if toupload is None: + ui.status(_('kbfiles: No remote repo\n')) + else: + ui.status(_('kbfiles to upload:\n')) + for file in toupload: + ui.status(bfutil.split_standin(file) + '\n') + ui.status('\n') + +def override_summary(orig, ui, repo, *pats, **opts): + orig(ui, repo, *pats, **opts) + + if opts.pop('bf', None): + toupload = get_outgoing_bfiles(ui, repo, None, **opts) + if toupload is None: + ui.status(_('kbfiles: No remote repo\n')) + else: + ui.status(_('kbfiles: %d to upload\n') % len(toupload)) + +def override_addremove(orig, ui, repo, *pats, **opts): + # Check if the parent or child has bfiles if they do don't allow it. + # If there is a symlink in the manifest then getting the manifest throws an exception + # catch it and let addremove deal with it. This happens in Mercurial's test + # test-addremove-symlink + try: + manifesttip = set(repo['tip'].manifest()) + except util.Abort: + manifesttip = set() + try: + manifestworking = set(repo[None].manifest()) + except util.Abort: + manifestworking = set() + + # Manifests are only iterable so turn them into sets then union + for file in manifesttip.union(manifestworking): + if file.startswith(bfutil.short_name): + raise util.Abort(_('addremove cannot be run on a repo with bfiles')) + + return orig(ui, repo, *pats, **opts) + +def uisetup(ui): + # Disable auto-status for some commands which assume that all + # files in the result are under Mercurial's control + + entry = extensions.wrapcommand(commands.table, 'add', override_add) + addopt = [('', 'bf', None, _('add as bfile')), + ('', 'bfsize', '', _('add all files above this size (in megabytes) as bfiles (default: 10)'))] + entry[1].extend(addopt) + + entry = extensions.wrapcommand(commands.table, 'addremove', override_addremove) + entry = extensions.wrapcommand(commands.table, 'remove', override_remove) + entry = extensions.wrapcommand(commands.table, 'forget', override_forget) + entry = extensions.wrapcommand(commands.table, 'status', override_status) + + entry = extensions.wrapcommand(commands.table, 'verify', override_verify) + verifyopt = [('', 'bf', None, _('verify bfiles')), + ('', 'bfa', None, _('verify all revisions of bfiles not just current')), + ('', 'bfc', None, _('verify bfile contents not just existence'))] + entry[1].extend(verifyopt) + + entry = extensions.wrapcommand(commands.table, 'outgoing', override_outgoing) + outgoingopt = [('', 'bf', None, _('display outgoing bfiles'))] + entry[1].extend(outgoingopt) + entry = extensions.wrapcommand(commands.table, 'summary', override_summary) + summaryopt = [('', 'bf', None, _('display outgoing bfiles'))] + entry[1].extend(summaryopt) + + entry = extensions.wrapcommand(commands.table, 'update', override_update) + entry = extensions.wrapcommand(commands.table, 'pull', override_pull) + entry = extensions.wrapfunction(filemerge, 'filemerge', override_filemerge) + entry = extensions.wrapfunction(cmdutil, 'copy', override_copy) + + # Backout calls revert so we need to override both the command and the function + entry = extensions.wrapcommand(commands.table, 'revert', override_revert) + entry = extensions.wrapfunction(commands, 'revert', override_revert) + + # clone uses hg._update instead of hg.update even though they are the + # same function... so wrap both of them) + extensions.wrapfunction(hg, 'update', hg_update) + extensions.wrapfunction(hg, '_update', hg_update) + extensions.wrapfunction(hg, 'clean', hg_clean) + extensions.wrapfunction(hg, 'merge', hg_merge) + + extensions.wrapfunction(archival, 'archive', override_archive) + extensions.wrapfunction(cmdutil, 'bail_if_changed', override_bail_if_changed) + + for name, module in extensions.extensions(): + if name == 'fetch': + extensions.wrapcommand(getattr(module, 'cmdtable'), 'fetch', override_fetch) + +
Show Entire File kbfiles/​bfutil.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File kbfiles/​httpstore.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Change 1 of 1 Show Entire File kbfiles/​localstore.py Stacked
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
@@ -0,0 +1,59 @@
+'''Store class for local filesystem.''' + +import os + +from mercurial import util +from mercurial.i18n import _ +import bfutil, basestore + +class localstore(basestore.basestore): + '''Because there is a system wide cache, the local store always uses that cache. + Since the cache is updated elsewhere, we can just read from it here as if it were the store.''' + + def __init__(self, ui, repo, url): + url = os.path.join(url, '.hg', bfutil.long_name) + super(localstore, self).__init__(ui, repo, util.expandpath(url)) + + def put(self, source, filename, hash): + '''Any file that is put must already be in the system wide cache so do nothing.''' + return + + def exists(self, hash): + return bfutil.in_system_cache(self.repo.ui, hash) + + def _getfile(self, tmpfile, filename, hash): + if bfutil.in_system_cache(self.ui, hash): + return bfutil.system_cache_path(self.ui, hash) + raise basestore.StoreError(filename, hash, '', _("Can't get file locally")) + + def _verifyfile(self, cctx, cset, contents, standin, verified): + filename = bfutil.split_standin(standin) + if not filename: + return False + fctx = cctx[standin] + key = (filename, fctx.filenode()) + if key in verified: + return False + + expect_hash = fctx.data()[0:40] + verified.add(key) + if not bfutil.in_system_cache(self.ui, expect_hash): + self.ui.warn( + _('changeset %s: %s missing\n' + ' (%s: %s)\n') + % (cset, filename, expect_hash, err.strerror)) + return True # failed + + if contents: + store_path = bfutil.system_cache_path(self.ui, expect_hash) + actual_hash = bfutil.hashfile(store_path) + if actual_hash != expect_hash: + self.ui.warn( + _('changeset %s: %s: contents differ\n' + ' (%s:\n' + ' expected hash %s,\n' + ' but got %s)\n') + % (cset, filename, + store_path, expect_hash, actual_hash)) + return True # failed + return False
Show Entire File tests/​README Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​common.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​hgtest.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​kilntest.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-add.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-add.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-addremove.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Change 1 of 1 Show Entire File tests/​test-addremove.py.out Stacked
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
@@ -0,0 +1,13 @@
+% test with bfiles +hg init +hg add --bf foo +hg addremove +hg commit -m 'added foo' +hg addremove + +% test without bfiles +hg init +hg add foo +hg addremove +hg commit -m 'added foo' +hg addremove
Show Entire File tests/​test-archive.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-archive.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-backout.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-backout.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-bisect.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-bisect.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-clone.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-clone.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-commit.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-commit.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-convert.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-convert.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-converttags.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-converttags.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-copyrename.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-copyrename.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-edit.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-edit.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-fetch.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-fetch.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-forget.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-forget.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-kiln.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-kiln.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-kilnnobfiles.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-kilnnobfiles.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-outgoing.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-outgoing.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-permissions.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-permissions.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-pushpull.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-pushpull.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-remove.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-remove.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
 
 
Show Entire File tests/​test-reposetup.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-reposetup.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-revert.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-revert.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-schemes.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-schemes.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-status.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-status.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-summary.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-summary.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-update.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-update.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-verify.py Stacked
This file's diff was not loaded because this changeset is very large. Load changes
Show Entire File tests/​test-verify.py.out Stacked
This file's diff was not loaded because this changeset is very large. Load changes