Show More
@@ -1,458 +1,490 b'' | |||
|
1 | 1 | # branchmap.py - logic to computes, maintain and stores branchmap for local repo |
|
2 | 2 | # |
|
3 | 3 | # Copyright 2005-2007 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 array |
|
11 | 11 | import struct |
|
12 | 12 | import time |
|
13 | 13 | |
|
14 | 14 | from .node import ( |
|
15 | 15 | bin, |
|
16 | 16 | hex, |
|
17 | 17 | nullid, |
|
18 | 18 | nullrev, |
|
19 | 19 | ) |
|
20 | 20 | from . import ( |
|
21 | 21 | encoding, |
|
22 | 22 | scmutil, |
|
23 | 23 | util, |
|
24 | 24 | ) |
|
25 | 25 | |
|
26 | 26 | array = array.array |
|
27 | 27 | calcsize = struct.calcsize |
|
28 | 28 | pack = struct.pack |
|
29 | 29 | unpack = struct.unpack |
|
30 | 30 | |
|
31 | 31 | def _filename(repo): |
|
32 | 32 | """name of a branchcache file for a given repo or repoview""" |
|
33 | 33 | filename = "cache/branch2" |
|
34 | 34 | if repo.filtername: |
|
35 | 35 | filename = '%s-%s' % (filename, repo.filtername) |
|
36 | 36 | return filename |
|
37 | 37 | |
|
38 | 38 | def read(repo): |
|
39 | 39 | try: |
|
40 | 40 | f = repo.vfs(_filename(repo)) |
|
41 | 41 | lines = f.read().split('\n') |
|
42 | 42 | f.close() |
|
43 | 43 | except (IOError, OSError): |
|
44 | 44 | return None |
|
45 | 45 | |
|
46 | 46 | try: |
|
47 | 47 | cachekey = lines.pop(0).split(" ", 2) |
|
48 | 48 | last, lrev = cachekey[:2] |
|
49 | 49 | last, lrev = bin(last), int(lrev) |
|
50 | 50 | filteredhash = None |
|
51 | 51 | if len(cachekey) > 2: |
|
52 | 52 | filteredhash = bin(cachekey[2]) |
|
53 | 53 | partial = branchcache(tipnode=last, tiprev=lrev, |
|
54 | 54 | filteredhash=filteredhash) |
|
55 | 55 | if not partial.validfor(repo): |
|
56 | 56 | # invalidate the cache |
|
57 | 57 | raise ValueError('tip differs') |
|
58 | 58 | for l in lines: |
|
59 | 59 | if not l: |
|
60 | 60 | continue |
|
61 | 61 | node, state, label = l.split(" ", 2) |
|
62 | 62 | if state not in 'oc': |
|
63 | 63 | raise ValueError('invalid branch state') |
|
64 | 64 | label = encoding.tolocal(label.strip()) |
|
65 | 65 | if not node in repo: |
|
66 | 66 | raise ValueError('node %s does not exist' % node) |
|
67 | 67 | node = bin(node) |
|
68 | 68 | partial.setdefault(label, []).append(node) |
|
69 | 69 | if state == 'c': |
|
70 | 70 | partial._closednodes.add(node) |
|
71 | 71 | except KeyboardInterrupt: |
|
72 | 72 | raise |
|
73 | 73 | except Exception as inst: |
|
74 | 74 | if repo.ui.debugflag: |
|
75 | 75 | msg = 'invalid branchheads cache' |
|
76 | 76 | if repo.filtername is not None: |
|
77 | 77 | msg += ' (%s)' % repo.filtername |
|
78 | 78 | msg += ': %s\n' |
|
79 | 79 | repo.ui.debug(msg % inst) |
|
80 | 80 | partial = None |
|
81 | 81 | return partial |
|
82 | 82 | |
|
83 | 83 | ### Nearest subset relation |
|
84 | 84 | # Nearest subset of filter X is a filter Y so that: |
|
85 | 85 | # * Y is included in X, |
|
86 | 86 | # * X - Y is as small as possible. |
|
87 | 87 | # This create and ordering used for branchmap purpose. |
|
88 | 88 | # the ordering may be partial |
|
89 | 89 | subsettable = {None: 'visible', |
|
90 | 90 | 'visible': 'served', |
|
91 | 91 | 'served': 'immutable', |
|
92 | 92 | 'immutable': 'base'} |
|
93 | 93 | |
|
94 | 94 | def updatecache(repo): |
|
95 | 95 | cl = repo.changelog |
|
96 | 96 | filtername = repo.filtername |
|
97 | 97 | partial = repo._branchcaches.get(filtername) |
|
98 | 98 | |
|
99 | 99 | revs = [] |
|
100 | 100 | if partial is None or not partial.validfor(repo): |
|
101 | 101 | partial = read(repo) |
|
102 | 102 | if partial is None: |
|
103 | 103 | subsetname = subsettable.get(filtername) |
|
104 | 104 | if subsetname is None: |
|
105 | 105 | partial = branchcache() |
|
106 | 106 | else: |
|
107 | 107 | subset = repo.filtered(subsetname) |
|
108 | 108 | partial = subset.branchmap().copy() |
|
109 | 109 | extrarevs = subset.changelog.filteredrevs - cl.filteredrevs |
|
110 | 110 | revs.extend(r for r in extrarevs if r <= partial.tiprev) |
|
111 | 111 | revs.extend(cl.revs(start=partial.tiprev + 1)) |
|
112 | 112 | if revs: |
|
113 | 113 | partial.update(repo, revs) |
|
114 | 114 | partial.write(repo) |
|
115 | 115 | |
|
116 | 116 | assert partial.validfor(repo), filtername |
|
117 | 117 | repo._branchcaches[repo.filtername] = partial |
|
118 | 118 | |
|
119 | def replacecache(repo, bm): | |
|
120 | """Replace the branchmap cache for a repo with a branch mapping. | |
|
121 | ||
|
122 | This is likely only called during clone with a branch map from a remote. | |
|
123 | """ | |
|
124 | rbheads = [] | |
|
125 | closed = [] | |
|
126 | for bheads in bm.itervalues(): | |
|
127 | rbheads.extend(bheads) | |
|
128 | for h in bheads: | |
|
129 | r = repo.changelog.rev(h) | |
|
130 | b, c = repo.changelog.branchinfo(r) | |
|
131 | if c: | |
|
132 | closed.append(h) | |
|
133 | ||
|
134 | if rbheads: | |
|
135 | rtiprev = max((int(repo.changelog.rev(node)) | |
|
136 | for node in rbheads)) | |
|
137 | cache = branchcache(bm, | |
|
138 | repo[rtiprev].node(), | |
|
139 | rtiprev, | |
|
140 | closednodes=closed) | |
|
141 | ||
|
142 | # Try to stick it as low as possible | |
|
143 | # filter above served are unlikely to be fetch from a clone | |
|
144 | for candidate in ('base', 'immutable', 'served'): | |
|
145 | rview = repo.filtered(candidate) | |
|
146 | if cache.validfor(rview): | |
|
147 | repo._branchcaches[candidate] = cache | |
|
148 | cache.write(rview) | |
|
149 | break | |
|
150 | ||
|
119 | 151 | class branchcache(dict): |
|
120 | 152 | """A dict like object that hold branches heads cache. |
|
121 | 153 | |
|
122 | 154 | This cache is used to avoid costly computations to determine all the |
|
123 | 155 | branch heads of a repo. |
|
124 | 156 | |
|
125 | 157 | The cache is serialized on disk in the following format: |
|
126 | 158 | |
|
127 | 159 | <tip hex node> <tip rev number> [optional filtered repo hex hash] |
|
128 | 160 | <branch head hex node> <open/closed state> <branch name> |
|
129 | 161 | <branch head hex node> <open/closed state> <branch name> |
|
130 | 162 | ... |
|
131 | 163 | |
|
132 | 164 | The first line is used to check if the cache is still valid. If the |
|
133 | 165 | branch cache is for a filtered repo view, an optional third hash is |
|
134 | 166 | included that hashes the hashes of all filtered revisions. |
|
135 | 167 | |
|
136 | 168 | The open/closed state is represented by a single letter 'o' or 'c'. |
|
137 | 169 | This field can be used to avoid changelog reads when determining if a |
|
138 | 170 | branch head closes a branch or not. |
|
139 | 171 | """ |
|
140 | 172 | |
|
141 | 173 | def __init__(self, entries=(), tipnode=nullid, tiprev=nullrev, |
|
142 | 174 | filteredhash=None, closednodes=None): |
|
143 | 175 | super(branchcache, self).__init__(entries) |
|
144 | 176 | self.tipnode = tipnode |
|
145 | 177 | self.tiprev = tiprev |
|
146 | 178 | self.filteredhash = filteredhash |
|
147 | 179 | # closednodes is a set of nodes that close their branch. If the branch |
|
148 | 180 | # cache has been updated, it may contain nodes that are no longer |
|
149 | 181 | # heads. |
|
150 | 182 | if closednodes is None: |
|
151 | 183 | self._closednodes = set() |
|
152 | 184 | else: |
|
153 | 185 | self._closednodes = closednodes |
|
154 | 186 | |
|
155 | 187 | def validfor(self, repo): |
|
156 | 188 | """Is the cache content valid regarding a repo |
|
157 | 189 | |
|
158 | 190 | - False when cached tipnode is unknown or if we detect a strip. |
|
159 | 191 | - True when cache is up to date or a subset of current repo.""" |
|
160 | 192 | try: |
|
161 | 193 | return ((self.tipnode == repo.changelog.node(self.tiprev)) |
|
162 | 194 | and (self.filteredhash == \ |
|
163 | 195 | scmutil.filteredhash(repo, self.tiprev))) |
|
164 | 196 | except IndexError: |
|
165 | 197 | return False |
|
166 | 198 | |
|
167 | 199 | def _branchtip(self, heads): |
|
168 | 200 | '''Return tuple with last open head in heads and false, |
|
169 | 201 | otherwise return last closed head and true.''' |
|
170 | 202 | tip = heads[-1] |
|
171 | 203 | closed = True |
|
172 | 204 | for h in reversed(heads): |
|
173 | 205 | if h not in self._closednodes: |
|
174 | 206 | tip = h |
|
175 | 207 | closed = False |
|
176 | 208 | break |
|
177 | 209 | return tip, closed |
|
178 | 210 | |
|
179 | 211 | def branchtip(self, branch): |
|
180 | 212 | '''Return the tipmost open head on branch head, otherwise return the |
|
181 | 213 | tipmost closed head on branch. |
|
182 | 214 | Raise KeyError for unknown branch.''' |
|
183 | 215 | return self._branchtip(self[branch])[0] |
|
184 | 216 | |
|
185 | 217 | def branchheads(self, branch, closed=False): |
|
186 | 218 | heads = self[branch] |
|
187 | 219 | if not closed: |
|
188 | 220 | heads = [h for h in heads if h not in self._closednodes] |
|
189 | 221 | return heads |
|
190 | 222 | |
|
191 | 223 | def iterbranches(self): |
|
192 | 224 | for bn, heads in self.iteritems(): |
|
193 | 225 | yield (bn, heads) + self._branchtip(heads) |
|
194 | 226 | |
|
195 | 227 | def copy(self): |
|
196 | 228 | """return an deep copy of the branchcache object""" |
|
197 | 229 | return branchcache(self, self.tipnode, self.tiprev, self.filteredhash, |
|
198 | 230 | self._closednodes) |
|
199 | 231 | |
|
200 | 232 | def write(self, repo): |
|
201 | 233 | try: |
|
202 | 234 | f = repo.vfs(_filename(repo), "w", atomictemp=True) |
|
203 | 235 | cachekey = [hex(self.tipnode), str(self.tiprev)] |
|
204 | 236 | if self.filteredhash is not None: |
|
205 | 237 | cachekey.append(hex(self.filteredhash)) |
|
206 | 238 | f.write(" ".join(cachekey) + '\n') |
|
207 | 239 | nodecount = 0 |
|
208 | 240 | for label, nodes in sorted(self.iteritems()): |
|
209 | 241 | for node in nodes: |
|
210 | 242 | nodecount += 1 |
|
211 | 243 | if node in self._closednodes: |
|
212 | 244 | state = 'c' |
|
213 | 245 | else: |
|
214 | 246 | state = 'o' |
|
215 | 247 | f.write("%s %s %s\n" % (hex(node), state, |
|
216 | 248 | encoding.fromlocal(label))) |
|
217 | 249 | f.close() |
|
218 | 250 | repo.ui.log('branchcache', |
|
219 | 251 | 'wrote %s branch cache with %d labels and %d nodes\n', |
|
220 | 252 | repo.filtername, len(self), nodecount) |
|
221 | 253 | except (IOError, OSError, util.Abort) as inst: |
|
222 | 254 | repo.ui.debug("couldn't write branch cache: %s\n" % inst) |
|
223 | 255 | # Abort may be raise by read only opener |
|
224 | 256 | pass |
|
225 | 257 | |
|
226 | 258 | def update(self, repo, revgen): |
|
227 | 259 | """Given a branchhead cache, self, that may have extra nodes or be |
|
228 | 260 | missing heads, and a generator of nodes that are strictly a superset of |
|
229 | 261 | heads missing, this function updates self to be correct. |
|
230 | 262 | """ |
|
231 | 263 | starttime = time.time() |
|
232 | 264 | cl = repo.changelog |
|
233 | 265 | # collect new branch entries |
|
234 | 266 | newbranches = {} |
|
235 | 267 | getbranchinfo = repo.revbranchcache().branchinfo |
|
236 | 268 | for r in revgen: |
|
237 | 269 | branch, closesbranch = getbranchinfo(r) |
|
238 | 270 | newbranches.setdefault(branch, []).append(r) |
|
239 | 271 | if closesbranch: |
|
240 | 272 | self._closednodes.add(cl.node(r)) |
|
241 | 273 | |
|
242 | 274 | # fetch current topological heads to speed up filtering |
|
243 | 275 | topoheads = set(cl.headrevs()) |
|
244 | 276 | |
|
245 | 277 | # if older branchheads are reachable from new ones, they aren't |
|
246 | 278 | # really branchheads. Note checking parents is insufficient: |
|
247 | 279 | # 1 (branch a) -> 2 (branch b) -> 3 (branch a) |
|
248 | 280 | for branch, newheadrevs in newbranches.iteritems(): |
|
249 | 281 | bheads = self.setdefault(branch, []) |
|
250 | 282 | bheadset = set(cl.rev(node) for node in bheads) |
|
251 | 283 | |
|
252 | 284 | # This have been tested True on all internal usage of this function. |
|
253 | 285 | # run it again in case of doubt |
|
254 | 286 | # assert not (set(bheadrevs) & set(newheadrevs)) |
|
255 | 287 | newheadrevs.sort() |
|
256 | 288 | bheadset.update(newheadrevs) |
|
257 | 289 | |
|
258 | 290 | # This prunes out two kinds of heads - heads that are superseded by |
|
259 | 291 | # a head in newheadrevs, and newheadrevs that are not heads because |
|
260 | 292 | # an existing head is their descendant. |
|
261 | 293 | uncertain = bheadset - topoheads |
|
262 | 294 | if uncertain: |
|
263 | 295 | floorrev = min(uncertain) |
|
264 | 296 | ancestors = set(cl.ancestors(newheadrevs, floorrev)) |
|
265 | 297 | bheadset -= ancestors |
|
266 | 298 | bheadrevs = sorted(bheadset) |
|
267 | 299 | self[branch] = [cl.node(rev) for rev in bheadrevs] |
|
268 | 300 | tiprev = bheadrevs[-1] |
|
269 | 301 | if tiprev > self.tiprev: |
|
270 | 302 | self.tipnode = cl.node(tiprev) |
|
271 | 303 | self.tiprev = tiprev |
|
272 | 304 | |
|
273 | 305 | if not self.validfor(repo): |
|
274 | 306 | # cache key are not valid anymore |
|
275 | 307 | self.tipnode = nullid |
|
276 | 308 | self.tiprev = nullrev |
|
277 | 309 | for heads in self.values(): |
|
278 | 310 | tiprev = max(cl.rev(node) for node in heads) |
|
279 | 311 | if tiprev > self.tiprev: |
|
280 | 312 | self.tipnode = cl.node(tiprev) |
|
281 | 313 | self.tiprev = tiprev |
|
282 | 314 | self.filteredhash = scmutil.filteredhash(repo, self.tiprev) |
|
283 | 315 | |
|
284 | 316 | duration = time.time() - starttime |
|
285 | 317 | repo.ui.log('branchcache', 'updated %s branch cache in %.4f seconds\n', |
|
286 | 318 | repo.filtername, duration) |
|
287 | 319 | |
|
288 | 320 | # Revision branch info cache |
|
289 | 321 | |
|
290 | 322 | _rbcversion = '-v1' |
|
291 | 323 | _rbcnames = 'cache/rbc-names' + _rbcversion |
|
292 | 324 | _rbcrevs = 'cache/rbc-revs' + _rbcversion |
|
293 | 325 | # [4 byte hash prefix][4 byte branch name number with sign bit indicating open] |
|
294 | 326 | _rbcrecfmt = '>4sI' |
|
295 | 327 | _rbcrecsize = calcsize(_rbcrecfmt) |
|
296 | 328 | _rbcnodelen = 4 |
|
297 | 329 | _rbcbranchidxmask = 0x7fffffff |
|
298 | 330 | _rbccloseflag = 0x80000000 |
|
299 | 331 | |
|
300 | 332 | class revbranchcache(object): |
|
301 | 333 | """Persistent cache, mapping from revision number to branch name and close. |
|
302 | 334 | This is a low level cache, independent of filtering. |
|
303 | 335 | |
|
304 | 336 | Branch names are stored in rbc-names in internal encoding separated by 0. |
|
305 | 337 | rbc-names is append-only, and each branch name is only stored once and will |
|
306 | 338 | thus have a unique index. |
|
307 | 339 | |
|
308 | 340 | The branch info for each revision is stored in rbc-revs as constant size |
|
309 | 341 | records. The whole file is read into memory, but it is only 'parsed' on |
|
310 | 342 | demand. The file is usually append-only but will be truncated if repo |
|
311 | 343 | modification is detected. |
|
312 | 344 | The record for each revision contains the first 4 bytes of the |
|
313 | 345 | corresponding node hash, and the record is only used if it still matches. |
|
314 | 346 | Even a completely trashed rbc-revs fill thus still give the right result |
|
315 | 347 | while converging towards full recovery ... assuming no incorrectly matching |
|
316 | 348 | node hashes. |
|
317 | 349 | The record also contains 4 bytes where 31 bits contains the index of the |
|
318 | 350 | branch and the last bit indicate that it is a branch close commit. |
|
319 | 351 | The usage pattern for rbc-revs is thus somewhat similar to 00changelog.i |
|
320 | 352 | and will grow with it but be 1/8th of its size. |
|
321 | 353 | """ |
|
322 | 354 | |
|
323 | 355 | def __init__(self, repo, readonly=True): |
|
324 | 356 | assert repo.filtername is None |
|
325 | 357 | self._repo = repo |
|
326 | 358 | self._names = [] # branch names in local encoding with static index |
|
327 | 359 | self._rbcrevs = array('c') # structs of type _rbcrecfmt |
|
328 | 360 | self._rbcsnameslen = 0 |
|
329 | 361 | try: |
|
330 | 362 | bndata = repo.vfs.read(_rbcnames) |
|
331 | 363 | self._rbcsnameslen = len(bndata) # for verification before writing |
|
332 | 364 | self._names = [encoding.tolocal(bn) for bn in bndata.split('\0')] |
|
333 | 365 | except (IOError, OSError) as inst: |
|
334 | 366 | if readonly: |
|
335 | 367 | # don't try to use cache - fall back to the slow path |
|
336 | 368 | self.branchinfo = self._branchinfo |
|
337 | 369 | |
|
338 | 370 | if self._names: |
|
339 | 371 | try: |
|
340 | 372 | data = repo.vfs.read(_rbcrevs) |
|
341 | 373 | self._rbcrevs.fromstring(data) |
|
342 | 374 | except (IOError, OSError) as inst: |
|
343 | 375 | repo.ui.debug("couldn't read revision branch cache: %s\n" % |
|
344 | 376 | inst) |
|
345 | 377 | # remember number of good records on disk |
|
346 | 378 | self._rbcrevslen = min(len(self._rbcrevs) // _rbcrecsize, |
|
347 | 379 | len(repo.changelog)) |
|
348 | 380 | if self._rbcrevslen == 0: |
|
349 | 381 | self._names = [] |
|
350 | 382 | self._rbcnamescount = len(self._names) # number of good names on disk |
|
351 | 383 | self._namesreverse = dict((b, r) for r, b in enumerate(self._names)) |
|
352 | 384 | |
|
353 | 385 | def branchinfo(self, rev): |
|
354 | 386 | """Return branch name and close flag for rev, using and updating |
|
355 | 387 | persistent cache.""" |
|
356 | 388 | changelog = self._repo.changelog |
|
357 | 389 | rbcrevidx = rev * _rbcrecsize |
|
358 | 390 | |
|
359 | 391 | # avoid negative index, changelog.read(nullrev) is fast without cache |
|
360 | 392 | if rev == nullrev: |
|
361 | 393 | return changelog.branchinfo(rev) |
|
362 | 394 | |
|
363 | 395 | # if requested rev is missing, add and populate all missing revs |
|
364 | 396 | if len(self._rbcrevs) < rbcrevidx + _rbcrecsize: |
|
365 | 397 | self._rbcrevs.extend('\0' * (len(changelog) * _rbcrecsize - |
|
366 | 398 | len(self._rbcrevs))) |
|
367 | 399 | |
|
368 | 400 | # fast path: extract data from cache, use it if node is matching |
|
369 | 401 | reponode = changelog.node(rev)[:_rbcnodelen] |
|
370 | 402 | cachenode, branchidx = unpack( |
|
371 | 403 | _rbcrecfmt, buffer(self._rbcrevs, rbcrevidx, _rbcrecsize)) |
|
372 | 404 | close = bool(branchidx & _rbccloseflag) |
|
373 | 405 | if close: |
|
374 | 406 | branchidx &= _rbcbranchidxmask |
|
375 | 407 | if cachenode == '\0\0\0\0': |
|
376 | 408 | pass |
|
377 | 409 | elif cachenode == reponode: |
|
378 | 410 | return self._names[branchidx], close |
|
379 | 411 | else: |
|
380 | 412 | # rev/node map has changed, invalidate the cache from here up |
|
381 | 413 | truncate = rbcrevidx + _rbcrecsize |
|
382 | 414 | del self._rbcrevs[truncate:] |
|
383 | 415 | self._rbcrevslen = min(self._rbcrevslen, truncate) |
|
384 | 416 | |
|
385 | 417 | # fall back to slow path and make sure it will be written to disk |
|
386 | 418 | return self._branchinfo(rev) |
|
387 | 419 | |
|
388 | 420 | def _branchinfo(self, rev): |
|
389 | 421 | """Retrieve branch info from changelog and update _rbcrevs""" |
|
390 | 422 | changelog = self._repo.changelog |
|
391 | 423 | b, close = changelog.branchinfo(rev) |
|
392 | 424 | if b in self._namesreverse: |
|
393 | 425 | branchidx = self._namesreverse[b] |
|
394 | 426 | else: |
|
395 | 427 | branchidx = len(self._names) |
|
396 | 428 | self._names.append(b) |
|
397 | 429 | self._namesreverse[b] = branchidx |
|
398 | 430 | reponode = changelog.node(rev) |
|
399 | 431 | if close: |
|
400 | 432 | branchidx |= _rbccloseflag |
|
401 | 433 | self._setcachedata(rev, reponode, branchidx) |
|
402 | 434 | return b, close |
|
403 | 435 | |
|
404 | 436 | def _setcachedata(self, rev, node, branchidx): |
|
405 | 437 | """Writes the node's branch data to the in-memory cache data.""" |
|
406 | 438 | rbcrevidx = rev * _rbcrecsize |
|
407 | 439 | rec = array('c') |
|
408 | 440 | rec.fromstring(pack(_rbcrecfmt, node, branchidx)) |
|
409 | 441 | self._rbcrevs[rbcrevidx:rbcrevidx + _rbcrecsize] = rec |
|
410 | 442 | self._rbcrevslen = min(self._rbcrevslen, rev) |
|
411 | 443 | |
|
412 | 444 | tr = self._repo.currenttransaction() |
|
413 | 445 | if tr: |
|
414 | 446 | tr.addfinalize('write-revbranchcache', self.write) |
|
415 | 447 | |
|
416 | 448 | def write(self, tr=None): |
|
417 | 449 | """Save branch cache if it is dirty.""" |
|
418 | 450 | repo = self._repo |
|
419 | 451 | if self._rbcnamescount < len(self._names): |
|
420 | 452 | try: |
|
421 | 453 | if self._rbcnamescount != 0: |
|
422 | 454 | f = repo.vfs.open(_rbcnames, 'ab') |
|
423 | 455 | if f.tell() == self._rbcsnameslen: |
|
424 | 456 | f.write('\0') |
|
425 | 457 | else: |
|
426 | 458 | f.close() |
|
427 | 459 | repo.ui.debug("%s changed - rewriting it\n" % _rbcnames) |
|
428 | 460 | self._rbcnamescount = 0 |
|
429 | 461 | self._rbcrevslen = 0 |
|
430 | 462 | if self._rbcnamescount == 0: |
|
431 | 463 | f = repo.vfs.open(_rbcnames, 'wb') |
|
432 | 464 | f.write('\0'.join(encoding.fromlocal(b) |
|
433 | 465 | for b in self._names[self._rbcnamescount:])) |
|
434 | 466 | self._rbcsnameslen = f.tell() |
|
435 | 467 | f.close() |
|
436 | 468 | except (IOError, OSError, util.Abort) as inst: |
|
437 | 469 | repo.ui.debug("couldn't write revision branch cache names: " |
|
438 | 470 | "%s\n" % inst) |
|
439 | 471 | return |
|
440 | 472 | self._rbcnamescount = len(self._names) |
|
441 | 473 | |
|
442 | 474 | start = self._rbcrevslen * _rbcrecsize |
|
443 | 475 | if start != len(self._rbcrevs): |
|
444 | 476 | revs = min(len(repo.changelog), len(self._rbcrevs) // _rbcrecsize) |
|
445 | 477 | try: |
|
446 | 478 | f = repo.vfs.open(_rbcrevs, 'ab') |
|
447 | 479 | if f.tell() != start: |
|
448 | 480 | repo.ui.debug("truncating %s to %s\n" % (_rbcrevs, start)) |
|
449 | 481 | f.seek(start) |
|
450 | 482 | f.truncate() |
|
451 | 483 | end = revs * _rbcrecsize |
|
452 | 484 | f.write(self._rbcrevs[start:end]) |
|
453 | 485 | f.close() |
|
454 | 486 | except (IOError, OSError, util.Abort) as inst: |
|
455 | 487 | repo.ui.debug("couldn't write revision branch cache: %s\n" % |
|
456 | 488 | inst) |
|
457 | 489 | return |
|
458 | 490 | self._rbcrevslen = revs |
@@ -1,283 +1,260 b'' | |||
|
1 | 1 | # streamclone.py - producing and consuming streaming repository data |
|
2 | 2 | # |
|
3 | 3 | # Copyright 2015 Gregory Szorc <gregory.szorc@gmail.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 time |
|
11 | 11 | |
|
12 | 12 | from .i18n import _ |
|
13 | 13 | from . import ( |
|
14 | 14 | branchmap, |
|
15 | 15 | error, |
|
16 | 16 | store, |
|
17 | 17 | util, |
|
18 | 18 | ) |
|
19 | 19 | |
|
20 | 20 | def canperformstreamclone(repo, remote, heads, streamrequested=None): |
|
21 | 21 | """Whether it is possible to perform a streaming clone as part of pull. |
|
22 | 22 | |
|
23 | 23 | Returns a tuple of (supported, requirements). ``supported`` is True if |
|
24 | 24 | streaming clone is supported and False otherwise. ``requirements`` is |
|
25 | 25 | a set of repo requirements from the remote, or ``None`` if stream clone |
|
26 | 26 | isn't supported. |
|
27 | 27 | """ |
|
28 | 28 | # Streaming clone only works on empty repositories. |
|
29 | 29 | if len(repo): |
|
30 | 30 | return False, None |
|
31 | 31 | |
|
32 | 32 | # Streaming clone only works if all data is being requested. |
|
33 | 33 | if heads: |
|
34 | 34 | return False, None |
|
35 | 35 | |
|
36 | 36 | # If we don't have a preference, let the server decide for us. This |
|
37 | 37 | # likely only comes into play in LANs. |
|
38 | 38 | if streamrequested is None: |
|
39 | 39 | # The server can advertise whether to prefer streaming clone. |
|
40 | 40 | streamrequested = remote.capable('stream-preferred') |
|
41 | 41 | |
|
42 | 42 | if not streamrequested: |
|
43 | 43 | return False, None |
|
44 | 44 | |
|
45 | 45 | # In order for stream clone to work, the client has to support all the |
|
46 | 46 | # requirements advertised by the server. |
|
47 | 47 | # |
|
48 | 48 | # The server advertises its requirements via the "stream" and "streamreqs" |
|
49 | 49 | # capability. "stream" (a value-less capability) is advertised if and only |
|
50 | 50 | # if the only requirement is "revlogv1." Else, the "streamreqs" capability |
|
51 | 51 | # is advertised and contains a comma-delimited list of requirements. |
|
52 | 52 | requirements = set() |
|
53 | 53 | if remote.capable('stream'): |
|
54 | 54 | requirements.add('revlogv1') |
|
55 | 55 | else: |
|
56 | 56 | streamreqs = remote.capable('streamreqs') |
|
57 | 57 | # This is weird and shouldn't happen with modern servers. |
|
58 | 58 | if not streamreqs: |
|
59 | 59 | return False, None |
|
60 | 60 | |
|
61 | 61 | streamreqs = set(streamreqs.split(',')) |
|
62 | 62 | # Server requires something we don't support. Bail. |
|
63 | 63 | if streamreqs - repo.supportedformats: |
|
64 | 64 | return False, None |
|
65 | 65 | requirements = streamreqs |
|
66 | 66 | |
|
67 | 67 | return True, requirements |
|
68 | 68 | |
|
69 | 69 | def maybeperformstreamclone(pullop): |
|
70 | 70 | repo = pullop.repo |
|
71 | 71 | remote = pullop.remote |
|
72 | 72 | |
|
73 | 73 | r = canperformstreamclone(repo, remote, pullop.heads, |
|
74 | 74 | streamrequested=pullop.streamclonerequested) |
|
75 | 75 | supported, requirements = r |
|
76 | 76 | |
|
77 | 77 | if not supported: |
|
78 | 78 | return |
|
79 | 79 | |
|
80 | 80 | # Save remote branchmap. We will use it later to speed up branchcache |
|
81 | 81 | # creation. |
|
82 | 82 | rbranchmap = None |
|
83 | 83 | if remote.capable('branchmap'): |
|
84 | 84 | rbranchmap = remote.branchmap() |
|
85 | 85 | |
|
86 | 86 | fp = remote.stream_out() |
|
87 | 87 | l = fp.readline() |
|
88 | 88 | try: |
|
89 | 89 | resp = int(l) |
|
90 | 90 | except ValueError: |
|
91 | 91 | raise error.ResponseError( |
|
92 | 92 | _('unexpected response from remote server:'), l) |
|
93 | 93 | if resp == 1: |
|
94 | 94 | raise util.Abort(_('operation forbidden by server')) |
|
95 | 95 | elif resp == 2: |
|
96 | 96 | raise util.Abort(_('locking the remote repository failed')) |
|
97 | 97 | elif resp != 0: |
|
98 | 98 | raise util.Abort(_('the server sent an unknown error code')) |
|
99 | 99 | |
|
100 | 100 | applyremotedata(repo, requirements, rbranchmap, fp) |
|
101 | 101 | |
|
102 | 102 | def allowservergeneration(ui): |
|
103 | 103 | """Whether streaming clones are allowed from the server.""" |
|
104 | 104 | return ui.configbool('server', 'uncompressed', True, untrusted=True) |
|
105 | 105 | |
|
106 | 106 | # This is it's own function so extensions can override it. |
|
107 | 107 | def _walkstreamfiles(repo): |
|
108 | 108 | return repo.store.walk() |
|
109 | 109 | |
|
110 | 110 | def generatev1(repo): |
|
111 | 111 | """Emit content for version 1 of a streaming clone. |
|
112 | 112 | |
|
113 | 113 | This is a generator of raw chunks that constitute a streaming clone. |
|
114 | 114 | |
|
115 | 115 | The stream begins with a line of 2 space-delimited integers containing the |
|
116 | 116 | number of entries and total bytes size. |
|
117 | 117 | |
|
118 | 118 | Next, are N entries for each file being transferred. Each file entry starts |
|
119 | 119 | as a line with the file name and integer size delimited by a null byte. |
|
120 | 120 | The raw file data follows. Following the raw file data is the next file |
|
121 | 121 | entry, or EOF. |
|
122 | 122 | |
|
123 | 123 | When used on the wire protocol, an additional line indicating protocol |
|
124 | 124 | success will be prepended to the stream. This function is not responsible |
|
125 | 125 | for adding it. |
|
126 | 126 | |
|
127 | 127 | This function will obtain a repository lock to ensure a consistent view of |
|
128 | 128 | the store is captured. It therefore may raise LockError. |
|
129 | 129 | """ |
|
130 | 130 | entries = [] |
|
131 | 131 | total_bytes = 0 |
|
132 | 132 | # Get consistent snapshot of repo, lock during scan. |
|
133 | 133 | lock = repo.lock() |
|
134 | 134 | try: |
|
135 | 135 | repo.ui.debug('scanning\n') |
|
136 | 136 | for name, ename, size in _walkstreamfiles(repo): |
|
137 | 137 | if size: |
|
138 | 138 | entries.append((name, size)) |
|
139 | 139 | total_bytes += size |
|
140 | 140 | finally: |
|
141 | 141 | lock.release() |
|
142 | 142 | |
|
143 | 143 | repo.ui.debug('%d files, %d bytes to transfer\n' % |
|
144 | 144 | (len(entries), total_bytes)) |
|
145 | 145 | yield '%d %d\n' % (len(entries), total_bytes) |
|
146 | 146 | |
|
147 | 147 | svfs = repo.svfs |
|
148 | 148 | oldaudit = svfs.mustaudit |
|
149 | 149 | debugflag = repo.ui.debugflag |
|
150 | 150 | svfs.mustaudit = False |
|
151 | 151 | |
|
152 | 152 | try: |
|
153 | 153 | for name, size in entries: |
|
154 | 154 | if debugflag: |
|
155 | 155 | repo.ui.debug('sending %s (%d bytes)\n' % (name, size)) |
|
156 | 156 | # partially encode name over the wire for backwards compat |
|
157 | 157 | yield '%s\0%d\n' % (store.encodedir(name), size) |
|
158 | 158 | if size <= 65536: |
|
159 | 159 | fp = svfs(name) |
|
160 | 160 | try: |
|
161 | 161 | data = fp.read(size) |
|
162 | 162 | finally: |
|
163 | 163 | fp.close() |
|
164 | 164 | yield data |
|
165 | 165 | else: |
|
166 | 166 | for chunk in util.filechunkiter(svfs(name), limit=size): |
|
167 | 167 | yield chunk |
|
168 | 168 | finally: |
|
169 | 169 | svfs.mustaudit = oldaudit |
|
170 | 170 | |
|
171 | 171 | def consumev1(repo, fp): |
|
172 | 172 | """Apply the contents from version 1 of a streaming clone file handle. |
|
173 | 173 | |
|
174 | 174 | This takes the output from "streamout" and applies it to the specified |
|
175 | 175 | repository. |
|
176 | 176 | |
|
177 | 177 | Like "streamout," the status line added by the wire protocol is not handled |
|
178 | 178 | by this function. |
|
179 | 179 | """ |
|
180 | 180 | lock = repo.lock() |
|
181 | 181 | try: |
|
182 | 182 | repo.ui.status(_('streaming all changes\n')) |
|
183 | 183 | l = fp.readline() |
|
184 | 184 | try: |
|
185 | 185 | total_files, total_bytes = map(int, l.split(' ', 1)) |
|
186 | 186 | except (ValueError, TypeError): |
|
187 | 187 | raise error.ResponseError( |
|
188 | 188 | _('unexpected response from remote server:'), l) |
|
189 | 189 | repo.ui.status(_('%d files to transfer, %s of data\n') % |
|
190 | 190 | (total_files, util.bytecount(total_bytes))) |
|
191 | 191 | handled_bytes = 0 |
|
192 | 192 | repo.ui.progress(_('clone'), 0, total=total_bytes) |
|
193 | 193 | start = time.time() |
|
194 | 194 | |
|
195 | 195 | tr = repo.transaction(_('clone')) |
|
196 | 196 | try: |
|
197 | 197 | for i in xrange(total_files): |
|
198 | 198 | # XXX doesn't support '\n' or '\r' in filenames |
|
199 | 199 | l = fp.readline() |
|
200 | 200 | try: |
|
201 | 201 | name, size = l.split('\0', 1) |
|
202 | 202 | size = int(size) |
|
203 | 203 | except (ValueError, TypeError): |
|
204 | 204 | raise error.ResponseError( |
|
205 | 205 | _('unexpected response from remote server:'), l) |
|
206 | 206 | if repo.ui.debugflag: |
|
207 | 207 | repo.ui.debug('adding %s (%s)\n' % |
|
208 | 208 | (name, util.bytecount(size))) |
|
209 | 209 | # for backwards compat, name was partially encoded |
|
210 | 210 | ofp = repo.svfs(store.decodedir(name), 'w') |
|
211 | 211 | for chunk in util.filechunkiter(fp, limit=size): |
|
212 | 212 | handled_bytes += len(chunk) |
|
213 | 213 | repo.ui.progress(_('clone'), handled_bytes, |
|
214 | 214 | total=total_bytes) |
|
215 | 215 | ofp.write(chunk) |
|
216 | 216 | ofp.close() |
|
217 | 217 | tr.close() |
|
218 | 218 | finally: |
|
219 | 219 | tr.release() |
|
220 | 220 | |
|
221 | 221 | # Writing straight to files circumvented the inmemory caches |
|
222 | 222 | repo.invalidate() |
|
223 | 223 | |
|
224 | 224 | elapsed = time.time() - start |
|
225 | 225 | if elapsed <= 0: |
|
226 | 226 | elapsed = 0.001 |
|
227 | 227 | repo.ui.progress(_('clone'), None) |
|
228 | 228 | repo.ui.status(_('transferred %s in %.1f seconds (%s/sec)\n') % |
|
229 | 229 | (util.bytecount(total_bytes), elapsed, |
|
230 | 230 | util.bytecount(total_bytes / elapsed))) |
|
231 | 231 | finally: |
|
232 | 232 | lock.release() |
|
233 | 233 | |
|
234 | 234 | def applyremotedata(repo, remotereqs, remotebranchmap, fp): |
|
235 | 235 | """Apply stream clone data to a repository. |
|
236 | 236 | |
|
237 | 237 | "remotereqs" is a set of requirements to handle the incoming data. |
|
238 | 238 | "remotebranchmap" is the result of a branchmap lookup on the remote. It |
|
239 | 239 | can be None. |
|
240 | 240 | "fp" is a file object containing the raw stream data, suitable for |
|
241 | 241 | feeding into consumev1(). |
|
242 | 242 | """ |
|
243 | 243 | lock = repo.lock() |
|
244 | 244 | try: |
|
245 | 245 | consumev1(repo, fp) |
|
246 | 246 | |
|
247 | 247 | # new requirements = old non-format requirements + |
|
248 | 248 | # new format-related remote requirements |
|
249 | 249 | # requirements from the streamed-in repository |
|
250 | 250 | repo.requirements = remotereqs | ( |
|
251 | 251 | repo.requirements - repo.supportedformats) |
|
252 | 252 | repo._applyopenerreqs() |
|
253 | 253 | repo._writerequirements() |
|
254 | 254 | |
|
255 | 255 | if remotebranchmap: |
|
256 | rbheads = [] | |
|
257 | closed = [] | |
|
258 | for bheads in remotebranchmap.itervalues(): | |
|
259 | rbheads.extend(bheads) | |
|
260 | for h in bheads: | |
|
261 | r = repo.changelog.rev(h) | |
|
262 | b, c = repo.changelog.branchinfo(r) | |
|
263 | if c: | |
|
264 | closed.append(h) | |
|
256 | branchmap.replacecache(repo, remotebranchmap) | |
|
265 | 257 | |
|
266 | if rbheads: | |
|
267 | rtiprev = max((int(repo.changelog.rev(node)) | |
|
268 | for node in rbheads)) | |
|
269 | cache = branchmap.branchcache(remotebranchmap, | |
|
270 | repo[rtiprev].node(), | |
|
271 | rtiprev, | |
|
272 | closednodes=closed) | |
|
273 | # Try to stick it as low as possible | |
|
274 | # filter above served are unlikely to be fetch from a clone | |
|
275 | for candidate in ('base', 'immutable', 'served'): | |
|
276 | rview = repo.filtered(candidate) | |
|
277 | if cache.validfor(rview): | |
|
278 | repo._branchcaches[candidate] = cache | |
|
279 | cache.write(rview) | |
|
280 | break | |
|
281 | 258 | repo.invalidate() |
|
282 | 259 | finally: |
|
283 | 260 | lock.release() |
General Comments 0
You need to be logged in to leave comments.
Login now