Show More
@@ -1,754 +1,757 b'' | |||||
1 | # bundlerepo.py - repository class for viewing uncompressed bundles |
|
1 | # bundlerepo.py - repository class for viewing uncompressed bundles | |
2 | # |
|
2 | # | |
3 | # Copyright 2006, 2007 Benoit Boissinot <bboissin@gmail.com> |
|
3 | # Copyright 2006, 2007 Benoit Boissinot <bboissin@gmail.com> | |
4 | # |
|
4 | # | |
5 | # This software may be used and distributed according to the terms of the |
|
5 | # This software may be used and distributed according to the terms of the | |
6 | # GNU General Public License version 2 or any later version. |
|
6 | # GNU General Public License version 2 or any later version. | |
7 |
|
7 | |||
8 | """Repository class for viewing uncompressed bundles. |
|
8 | """Repository class for viewing uncompressed bundles. | |
9 |
|
9 | |||
10 | This provides a read-only repository interface to bundles as if they |
|
10 | This provides a read-only repository interface to bundles as if they | |
11 | were part of the actual repository. |
|
11 | were part of the actual repository. | |
12 | """ |
|
12 | """ | |
13 |
|
13 | |||
14 |
|
14 | |||
15 | import contextlib |
|
15 | import contextlib | |
16 | import os |
|
16 | import os | |
17 | import shutil |
|
17 | import shutil | |
18 |
|
18 | |||
19 | from .i18n import _ |
|
19 | from .i18n import _ | |
20 | from .node import ( |
|
20 | from .node import ( | |
21 | hex, |
|
21 | hex, | |
22 | nullrev, |
|
22 | nullrev, | |
23 | ) |
|
23 | ) | |
24 |
|
24 | |||
25 | from . import ( |
|
25 | from . import ( | |
26 | bundle2, |
|
26 | bundle2, | |
27 | changegroup, |
|
27 | changegroup, | |
28 | changelog, |
|
28 | changelog, | |
29 | cmdutil, |
|
29 | cmdutil, | |
30 | discovery, |
|
30 | discovery, | |
31 | encoding, |
|
31 | encoding, | |
32 | error, |
|
32 | error, | |
33 | exchange, |
|
33 | exchange, | |
34 | filelog, |
|
34 | filelog, | |
35 | localrepo, |
|
35 | localrepo, | |
36 | manifest, |
|
36 | manifest, | |
37 | mdiff, |
|
37 | mdiff, | |
38 | pathutil, |
|
38 | pathutil, | |
39 | phases, |
|
39 | phases, | |
40 | pycompat, |
|
40 | pycompat, | |
41 | revlog, |
|
41 | revlog, | |
42 | revlogutils, |
|
42 | revlogutils, | |
43 | util, |
|
43 | util, | |
44 | vfs as vfsmod, |
|
44 | vfs as vfsmod, | |
45 | ) |
|
45 | ) | |
46 | from .utils import ( |
|
46 | from .utils import ( | |
47 | urlutil, |
|
47 | urlutil, | |
48 | ) |
|
48 | ) | |
49 |
|
49 | |||
50 | from .revlogutils import ( |
|
50 | from .revlogutils import ( | |
51 | constants as revlog_constants, |
|
51 | constants as revlog_constants, | |
52 | ) |
|
52 | ) | |
53 |
|
53 | |||
54 |
|
54 | |||
55 | class bundlerevlog(revlog.revlog): |
|
55 | class bundlerevlog(revlog.revlog): | |
56 | def __init__(self, opener, target, radix, cgunpacker, linkmapper): |
|
56 | def __init__(self, opener, target, radix, cgunpacker, linkmapper): | |
57 | # How it works: |
|
57 | # How it works: | |
58 | # To retrieve a revision, we need to know the offset of the revision in |
|
58 | # To retrieve a revision, we need to know the offset of the revision in | |
59 | # the bundle (an unbundle object). We store this offset in the index |
|
59 | # the bundle (an unbundle object). We store this offset in the index | |
60 | # (start). The base of the delta is stored in the base field. |
|
60 | # (start). The base of the delta is stored in the base field. | |
61 | # |
|
61 | # | |
62 | # To differentiate a rev in the bundle from a rev in the revlog, we |
|
62 | # To differentiate a rev in the bundle from a rev in the revlog, we | |
63 | # check revision against repotiprev. |
|
63 | # check revision against repotiprev. | |
64 | opener = vfsmod.readonlyvfs(opener) |
|
64 | opener = vfsmod.readonlyvfs(opener) | |
65 | revlog.revlog.__init__(self, opener, target=target, radix=radix) |
|
65 | revlog.revlog.__init__(self, opener, target=target, radix=radix) | |
66 | self.bundle = cgunpacker |
|
66 | self.bundle = cgunpacker | |
67 | n = len(self) |
|
67 | n = len(self) | |
68 | self.repotiprev = n - 1 |
|
68 | self.repotiprev = n - 1 | |
69 | self.bundlerevs = set() # used by 'bundle()' revset expression |
|
69 | self.bundlerevs = set() # used by 'bundle()' revset expression | |
70 | for deltadata in cgunpacker.deltaiter(): |
|
70 | for deltadata in cgunpacker.deltaiter(): | |
71 | node, p1, p2, cs, deltabase, delta, flags, sidedata = deltadata |
|
71 | node, p1, p2, cs, deltabase, delta, flags, sidedata = deltadata | |
72 |
|
72 | |||
73 | size = len(delta) |
|
73 | size = len(delta) | |
74 | start = cgunpacker.tell() - size |
|
74 | start = cgunpacker.tell() - size | |
75 |
|
75 | |||
76 | if self.index.has_node(node): |
|
76 | if self.index.has_node(node): | |
77 | # this can happen if two branches make the same change |
|
77 | # this can happen if two branches make the same change | |
78 | self.bundlerevs.add(self.index.rev(node)) |
|
78 | self.bundlerevs.add(self.index.rev(node)) | |
79 | continue |
|
79 | continue | |
80 | if cs == node: |
|
80 | if cs == node: | |
81 | linkrev = nullrev |
|
81 | linkrev = nullrev | |
82 | else: |
|
82 | else: | |
83 | linkrev = linkmapper(cs) |
|
83 | linkrev = linkmapper(cs) | |
84 |
|
84 | |||
85 | for p in (p1, p2): |
|
85 | for p in (p1, p2): | |
86 | if not self.index.has_node(p): |
|
86 | if not self.index.has_node(p): | |
87 | raise error.LookupError( |
|
87 | raise error.LookupError( | |
88 | p, self.display_id, _(b"unknown parent") |
|
88 | p, self.display_id, _(b"unknown parent") | |
89 | ) |
|
89 | ) | |
90 |
|
90 | |||
91 | if not self.index.has_node(deltabase): |
|
91 | if not self.index.has_node(deltabase): | |
92 | raise error.LookupError( |
|
92 | raise error.LookupError( | |
93 | deltabase, self.display_id, _(b'unknown delta base') |
|
93 | deltabase, self.display_id, _(b'unknown delta base') | |
94 | ) |
|
94 | ) | |
95 |
|
95 | |||
96 | baserev = self.rev(deltabase) |
|
96 | baserev = self.rev(deltabase) | |
97 | # start, size, full unc. size, base (unused), link, p1, p2, node, sidedata_offset (unused), sidedata_size (unused) |
|
97 | # start, size, full unc. size, base (unused), link, p1, p2, node, sidedata_offset (unused), sidedata_size (unused) | |
98 | e = revlogutils.entry( |
|
98 | e = revlogutils.entry( | |
99 | flags=flags, |
|
99 | flags=flags, | |
100 | data_offset=start, |
|
100 | data_offset=start, | |
101 | data_compressed_length=size, |
|
101 | data_compressed_length=size, | |
102 | data_delta_base=baserev, |
|
102 | data_delta_base=baserev, | |
103 | link_rev=linkrev, |
|
103 | link_rev=linkrev, | |
104 | parent_rev_1=self.rev(p1), |
|
104 | parent_rev_1=self.rev(p1), | |
105 | parent_rev_2=self.rev(p2), |
|
105 | parent_rev_2=self.rev(p2), | |
106 | node_id=node, |
|
106 | node_id=node, | |
107 | ) |
|
107 | ) | |
108 | self.index.append(e) |
|
108 | self.index.append(e) | |
109 | self.bundlerevs.add(n) |
|
109 | self.bundlerevs.add(n) | |
110 | n += 1 |
|
110 | n += 1 | |
111 |
|
111 | |||
112 | @contextlib.contextmanager |
|
112 | @contextlib.contextmanager | |
113 | def reading(self): |
|
113 | def reading(self): | |
114 | if self.repotiprev < 0: |
|
114 | if self.repotiprev < 0: | |
115 | yield |
|
115 | yield | |
116 | else: |
|
116 | else: | |
117 | with super().reading() as x: |
|
117 | with super().reading() as x: | |
118 | yield x |
|
118 | yield x | |
119 |
|
119 | |||
120 | def _chunk(self, rev): |
|
120 | def _chunk(self, rev): | |
121 | # Warning: in case of bundle, the diff is against what we stored as |
|
121 | # Warning: in case of bundle, the diff is against what we stored as | |
122 | # delta base, not against rev - 1 |
|
122 | # delta base, not against rev - 1 | |
123 | # XXX: could use some caching |
|
123 | # XXX: could use some caching | |
124 | if rev <= self.repotiprev: |
|
124 | if rev <= self.repotiprev: | |
125 | return revlog.revlog._chunk(self, rev) |
|
125 | return revlog.revlog._chunk(self, rev) | |
126 | self.bundle.seek(self.start(rev)) |
|
126 | self.bundle.seek(self.start(rev)) | |
127 | return self.bundle.read(self.length(rev)) |
|
127 | return self.bundle.read(self.length(rev)) | |
128 |
|
128 | |||
129 | def revdiff(self, rev1, rev2): |
|
129 | def revdiff(self, rev1, rev2): | |
130 | """return or calculate a delta between two revisions""" |
|
130 | """return or calculate a delta between two revisions""" | |
131 | if rev1 > self.repotiprev and rev2 > self.repotiprev: |
|
131 | if rev1 > self.repotiprev and rev2 > self.repotiprev: | |
132 | # hot path for bundle |
|
132 | # hot path for bundle | |
133 | revb = self.index[rev2][3] |
|
133 | revb = self.index[rev2][3] | |
134 | if revb == rev1: |
|
134 | if revb == rev1: | |
135 | return self._chunk(rev2) |
|
135 | return self._chunk(rev2) | |
136 | elif rev1 <= self.repotiprev and rev2 <= self.repotiprev: |
|
136 | elif rev1 <= self.repotiprev and rev2 <= self.repotiprev: | |
137 | return revlog.revlog.revdiff(self, rev1, rev2) |
|
137 | return revlog.revlog.revdiff(self, rev1, rev2) | |
138 |
|
138 | |||
139 | return mdiff.textdiff(self.rawdata(rev1), self.rawdata(rev2)) |
|
139 | return mdiff.textdiff(self.rawdata(rev1), self.rawdata(rev2)) | |
140 |
|
140 | |||
141 | def _rawtext(self, node, rev): |
|
141 | def _rawtext(self, node, rev): | |
142 | if rev is None: |
|
142 | if rev is None: | |
143 | rev = self.rev(node) |
|
143 | rev = self.rev(node) | |
144 | validated = False |
|
144 | validated = False | |
145 | rawtext = None |
|
145 | rawtext = None | |
146 | chain = [] |
|
146 | chain = [] | |
147 | iterrev = rev |
|
147 | iterrev = rev | |
148 | # reconstruct the revision if it is from a changegroup |
|
148 | # reconstruct the revision if it is from a changegroup | |
149 | while iterrev > self.repotiprev: |
|
149 | while iterrev > self.repotiprev: | |
150 | if self._revisioncache and self._revisioncache[1] == iterrev: |
|
150 | if ( | |
151 |
|
|
151 | self._inner._revisioncache | |
|
152 | and self._inner._revisioncache[1] == iterrev | |||
|
153 | ): | |||
|
154 | rawtext = self._inner._revisioncache[2] | |||
152 | break |
|
155 | break | |
153 | chain.append(iterrev) |
|
156 | chain.append(iterrev) | |
154 | iterrev = self.index[iterrev][3] |
|
157 | iterrev = self.index[iterrev][3] | |
155 | if iterrev == nullrev: |
|
158 | if iterrev == nullrev: | |
156 | rawtext = b'' |
|
159 | rawtext = b'' | |
157 | elif rawtext is None: |
|
160 | elif rawtext is None: | |
158 | r = super(bundlerevlog, self)._rawtext( |
|
161 | r = super(bundlerevlog, self)._rawtext( | |
159 | self.node(iterrev), |
|
162 | self.node(iterrev), | |
160 | iterrev, |
|
163 | iterrev, | |
161 | ) |
|
164 | ) | |
162 | __, rawtext, validated = r |
|
165 | __, rawtext, validated = r | |
163 | if chain: |
|
166 | if chain: | |
164 | validated = False |
|
167 | validated = False | |
165 | while chain: |
|
168 | while chain: | |
166 | delta = self._chunk(chain.pop()) |
|
169 | delta = self._chunk(chain.pop()) | |
167 | rawtext = mdiff.patches(rawtext, [delta]) |
|
170 | rawtext = mdiff.patches(rawtext, [delta]) | |
168 | return rev, rawtext, validated |
|
171 | return rev, rawtext, validated | |
169 |
|
172 | |||
170 | def addrevision(self, *args, **kwargs): |
|
173 | def addrevision(self, *args, **kwargs): | |
171 | raise NotImplementedError |
|
174 | raise NotImplementedError | |
172 |
|
175 | |||
173 | def addgroup(self, *args, **kwargs): |
|
176 | def addgroup(self, *args, **kwargs): | |
174 | raise NotImplementedError |
|
177 | raise NotImplementedError | |
175 |
|
178 | |||
176 | def strip(self, *args, **kwargs): |
|
179 | def strip(self, *args, **kwargs): | |
177 | raise NotImplementedError |
|
180 | raise NotImplementedError | |
178 |
|
181 | |||
179 | def checksize(self): |
|
182 | def checksize(self): | |
180 | raise NotImplementedError |
|
183 | raise NotImplementedError | |
181 |
|
184 | |||
182 |
|
185 | |||
183 | class bundlechangelog(bundlerevlog, changelog.changelog): |
|
186 | class bundlechangelog(bundlerevlog, changelog.changelog): | |
184 | def __init__(self, opener, cgunpacker): |
|
187 | def __init__(self, opener, cgunpacker): | |
185 | changelog.changelog.__init__(self, opener) |
|
188 | changelog.changelog.__init__(self, opener) | |
186 | linkmapper = lambda x: x |
|
189 | linkmapper = lambda x: x | |
187 | bundlerevlog.__init__( |
|
190 | bundlerevlog.__init__( | |
188 | self, |
|
191 | self, | |
189 | opener, |
|
192 | opener, | |
190 | (revlog_constants.KIND_CHANGELOG, None), |
|
193 | (revlog_constants.KIND_CHANGELOG, None), | |
191 | self.radix, |
|
194 | self.radix, | |
192 | cgunpacker, |
|
195 | cgunpacker, | |
193 | linkmapper, |
|
196 | linkmapper, | |
194 | ) |
|
197 | ) | |
195 |
|
198 | |||
196 |
|
199 | |||
197 | class bundlemanifest(bundlerevlog, manifest.manifestrevlog): |
|
200 | class bundlemanifest(bundlerevlog, manifest.manifestrevlog): | |
198 | def __init__( |
|
201 | def __init__( | |
199 | self, |
|
202 | self, | |
200 | nodeconstants, |
|
203 | nodeconstants, | |
201 | opener, |
|
204 | opener, | |
202 | cgunpacker, |
|
205 | cgunpacker, | |
203 | linkmapper, |
|
206 | linkmapper, | |
204 | dirlogstarts=None, |
|
207 | dirlogstarts=None, | |
205 | dir=b'', |
|
208 | dir=b'', | |
206 | ): |
|
209 | ): | |
207 | # XXX manifestrevlog is not actually a revlog , so mixing it with |
|
210 | # XXX manifestrevlog is not actually a revlog , so mixing it with | |
208 | # bundlerevlog is not a good idea. |
|
211 | # bundlerevlog is not a good idea. | |
209 | manifest.manifestrevlog.__init__(self, nodeconstants, opener, tree=dir) |
|
212 | manifest.manifestrevlog.__init__(self, nodeconstants, opener, tree=dir) | |
210 | bundlerevlog.__init__( |
|
213 | bundlerevlog.__init__( | |
211 | self, |
|
214 | self, | |
212 | opener, |
|
215 | opener, | |
213 | (revlog_constants.KIND_MANIFESTLOG, dir), |
|
216 | (revlog_constants.KIND_MANIFESTLOG, dir), | |
214 | self._revlog.radix, |
|
217 | self._revlog.radix, | |
215 | cgunpacker, |
|
218 | cgunpacker, | |
216 | linkmapper, |
|
219 | linkmapper, | |
217 | ) |
|
220 | ) | |
218 | if dirlogstarts is None: |
|
221 | if dirlogstarts is None: | |
219 | dirlogstarts = {} |
|
222 | dirlogstarts = {} | |
220 | if self.bundle.version == b"03": |
|
223 | if self.bundle.version == b"03": | |
221 | dirlogstarts = _getfilestarts(self.bundle) |
|
224 | dirlogstarts = _getfilestarts(self.bundle) | |
222 | self._dirlogstarts = dirlogstarts |
|
225 | self._dirlogstarts = dirlogstarts | |
223 | self._linkmapper = linkmapper |
|
226 | self._linkmapper = linkmapper | |
224 |
|
227 | |||
225 | def dirlog(self, d): |
|
228 | def dirlog(self, d): | |
226 | if d in self._dirlogstarts: |
|
229 | if d in self._dirlogstarts: | |
227 | self.bundle.seek(self._dirlogstarts[d]) |
|
230 | self.bundle.seek(self._dirlogstarts[d]) | |
228 | return bundlemanifest( |
|
231 | return bundlemanifest( | |
229 | self.nodeconstants, |
|
232 | self.nodeconstants, | |
230 | self.opener, |
|
233 | self.opener, | |
231 | self.bundle, |
|
234 | self.bundle, | |
232 | self._linkmapper, |
|
235 | self._linkmapper, | |
233 | self._dirlogstarts, |
|
236 | self._dirlogstarts, | |
234 | dir=d, |
|
237 | dir=d, | |
235 | ) |
|
238 | ) | |
236 | return super(bundlemanifest, self).dirlog(d) |
|
239 | return super(bundlemanifest, self).dirlog(d) | |
237 |
|
240 | |||
238 |
|
241 | |||
239 | class bundlefilelog(filelog.filelog): |
|
242 | class bundlefilelog(filelog.filelog): | |
240 | def __init__(self, opener, path, cgunpacker, linkmapper): |
|
243 | def __init__(self, opener, path, cgunpacker, linkmapper): | |
241 | filelog.filelog.__init__(self, opener, path) |
|
244 | filelog.filelog.__init__(self, opener, path) | |
242 | self._revlog = bundlerevlog( |
|
245 | self._revlog = bundlerevlog( | |
243 | opener, |
|
246 | opener, | |
244 | # XXX should use the unencoded path |
|
247 | # XXX should use the unencoded path | |
245 | target=(revlog_constants.KIND_FILELOG, path), |
|
248 | target=(revlog_constants.KIND_FILELOG, path), | |
246 | radix=self._revlog.radix, |
|
249 | radix=self._revlog.radix, | |
247 | cgunpacker=cgunpacker, |
|
250 | cgunpacker=cgunpacker, | |
248 | linkmapper=linkmapper, |
|
251 | linkmapper=linkmapper, | |
249 | ) |
|
252 | ) | |
250 |
|
253 | |||
251 |
|
254 | |||
252 | class bundlepeer(localrepo.localpeer): |
|
255 | class bundlepeer(localrepo.localpeer): | |
253 | def canpush(self): |
|
256 | def canpush(self): | |
254 | return False |
|
257 | return False | |
255 |
|
258 | |||
256 |
|
259 | |||
257 | class bundlephasecache(phases.phasecache): |
|
260 | class bundlephasecache(phases.phasecache): | |
258 | def __init__(self, *args, **kwargs): |
|
261 | def __init__(self, *args, **kwargs): | |
259 | super(bundlephasecache, self).__init__(*args, **kwargs) |
|
262 | super(bundlephasecache, self).__init__(*args, **kwargs) | |
260 | if hasattr(self, 'opener'): |
|
263 | if hasattr(self, 'opener'): | |
261 | self.opener = vfsmod.readonlyvfs(self.opener) |
|
264 | self.opener = vfsmod.readonlyvfs(self.opener) | |
262 |
|
265 | |||
263 | def write(self): |
|
266 | def write(self): | |
264 | raise NotImplementedError |
|
267 | raise NotImplementedError | |
265 |
|
268 | |||
266 | def _write(self, fp): |
|
269 | def _write(self, fp): | |
267 | raise NotImplementedError |
|
270 | raise NotImplementedError | |
268 |
|
271 | |||
269 | def _updateroots(self, phase, newroots, tr): |
|
272 | def _updateroots(self, phase, newroots, tr): | |
270 | self.phaseroots[phase] = newroots |
|
273 | self.phaseroots[phase] = newroots | |
271 | self.invalidate() |
|
274 | self.invalidate() | |
272 | self.dirty = True |
|
275 | self.dirty = True | |
273 |
|
276 | |||
274 |
|
277 | |||
275 | def _getfilestarts(cgunpacker): |
|
278 | def _getfilestarts(cgunpacker): | |
276 | filespos = {} |
|
279 | filespos = {} | |
277 | for chunkdata in iter(cgunpacker.filelogheader, {}): |
|
280 | for chunkdata in iter(cgunpacker.filelogheader, {}): | |
278 | fname = chunkdata[b'filename'] |
|
281 | fname = chunkdata[b'filename'] | |
279 | filespos[fname] = cgunpacker.tell() |
|
282 | filespos[fname] = cgunpacker.tell() | |
280 | for chunk in iter(lambda: cgunpacker.deltachunk(None), {}): |
|
283 | for chunk in iter(lambda: cgunpacker.deltachunk(None), {}): | |
281 | pass |
|
284 | pass | |
282 | return filespos |
|
285 | return filespos | |
283 |
|
286 | |||
284 |
|
287 | |||
285 | class bundlerepository: |
|
288 | class bundlerepository: | |
286 | """A repository instance that is a union of a local repo and a bundle. |
|
289 | """A repository instance that is a union of a local repo and a bundle. | |
287 |
|
290 | |||
288 | Instances represent a read-only repository composed of a local repository |
|
291 | Instances represent a read-only repository composed of a local repository | |
289 | with the contents of a bundle file applied. The repository instance is |
|
292 | with the contents of a bundle file applied. The repository instance is | |
290 | conceptually similar to the state of a repository after an |
|
293 | conceptually similar to the state of a repository after an | |
291 | ``hg unbundle`` operation. However, the contents of the bundle are never |
|
294 | ``hg unbundle`` operation. However, the contents of the bundle are never | |
292 | applied to the actual base repository. |
|
295 | applied to the actual base repository. | |
293 |
|
296 | |||
294 | Instances constructed directly are not usable as repository objects. |
|
297 | Instances constructed directly are not usable as repository objects. | |
295 | Use instance() or makebundlerepository() to create instances. |
|
298 | Use instance() or makebundlerepository() to create instances. | |
296 | """ |
|
299 | """ | |
297 |
|
300 | |||
298 | def __init__(self, bundlepath, url, tempparent): |
|
301 | def __init__(self, bundlepath, url, tempparent): | |
299 | self._tempparent = tempparent |
|
302 | self._tempparent = tempparent | |
300 | self._url = url |
|
303 | self._url = url | |
301 |
|
304 | |||
302 | self.ui.setconfig(b'phases', b'publish', False, b'bundlerepo') |
|
305 | self.ui.setconfig(b'phases', b'publish', False, b'bundlerepo') | |
303 |
|
306 | |||
304 | # dict with the mapping 'filename' -> position in the changegroup. |
|
307 | # dict with the mapping 'filename' -> position in the changegroup. | |
305 | self._cgfilespos = {} |
|
308 | self._cgfilespos = {} | |
306 | self._bundlefile = None |
|
309 | self._bundlefile = None | |
307 | self._cgunpacker = None |
|
310 | self._cgunpacker = None | |
308 | self.tempfile = None |
|
311 | self.tempfile = None | |
309 | f = util.posixfile(bundlepath, b"rb") |
|
312 | f = util.posixfile(bundlepath, b"rb") | |
310 | bundle = exchange.readbundle(self.ui, f, bundlepath) |
|
313 | bundle = exchange.readbundle(self.ui, f, bundlepath) | |
311 |
|
314 | |||
312 | if isinstance(bundle, bundle2.unbundle20): |
|
315 | if isinstance(bundle, bundle2.unbundle20): | |
313 | self._bundlefile = bundle |
|
316 | self._bundlefile = bundle | |
314 |
|
317 | |||
315 | cgpart = None |
|
318 | cgpart = None | |
316 | for part in bundle.iterparts(seekable=True): |
|
319 | for part in bundle.iterparts(seekable=True): | |
317 | if part.type == b'phase-heads': |
|
320 | if part.type == b'phase-heads': | |
318 | self._handle_bundle2_phase_part(bundle, part) |
|
321 | self._handle_bundle2_phase_part(bundle, part) | |
319 | elif part.type == b'changegroup': |
|
322 | elif part.type == b'changegroup': | |
320 | if cgpart: |
|
323 | if cgpart: | |
321 | raise NotImplementedError( |
|
324 | raise NotImplementedError( | |
322 | b"can't process multiple changegroups" |
|
325 | b"can't process multiple changegroups" | |
323 | ) |
|
326 | ) | |
324 | cgpart = part |
|
327 | cgpart = part | |
325 | self._handle_bundle2_cg_part(bundle, part) |
|
328 | self._handle_bundle2_cg_part(bundle, part) | |
326 |
|
329 | |||
327 | if not cgpart: |
|
330 | if not cgpart: | |
328 | raise error.Abort(_(b"No changegroups found")) |
|
331 | raise error.Abort(_(b"No changegroups found")) | |
329 |
|
332 | |||
330 | # This is required to placate a later consumer, which expects |
|
333 | # This is required to placate a later consumer, which expects | |
331 | # the payload offset to be at the beginning of the changegroup. |
|
334 | # the payload offset to be at the beginning of the changegroup. | |
332 | # We need to do this after the iterparts() generator advances |
|
335 | # We need to do this after the iterparts() generator advances | |
333 | # because iterparts() will seek to end of payload after the |
|
336 | # because iterparts() will seek to end of payload after the | |
334 | # generator returns control to iterparts(). |
|
337 | # generator returns control to iterparts(). | |
335 | cgpart.seek(0, os.SEEK_SET) |
|
338 | cgpart.seek(0, os.SEEK_SET) | |
336 |
|
339 | |||
337 | elif isinstance(bundle, changegroup.cg1unpacker): |
|
340 | elif isinstance(bundle, changegroup.cg1unpacker): | |
338 | self._handle_bundle1(bundle, bundlepath) |
|
341 | self._handle_bundle1(bundle, bundlepath) | |
339 | else: |
|
342 | else: | |
340 | raise error.Abort( |
|
343 | raise error.Abort( | |
341 | _(b'bundle type %r cannot be read') % type(bundle) |
|
344 | _(b'bundle type %r cannot be read') % type(bundle) | |
342 | ) |
|
345 | ) | |
343 |
|
346 | |||
344 | def _handle_bundle1(self, bundle, bundlepath): |
|
347 | def _handle_bundle1(self, bundle, bundlepath): | |
345 | if bundle.compressed(): |
|
348 | if bundle.compressed(): | |
346 | f = self._writetempbundle(bundle.read, b'.hg10un', header=b'HG10UN') |
|
349 | f = self._writetempbundle(bundle.read, b'.hg10un', header=b'HG10UN') | |
347 | bundle = exchange.readbundle(self.ui, f, bundlepath, self.vfs) |
|
350 | bundle = exchange.readbundle(self.ui, f, bundlepath, self.vfs) | |
348 |
|
351 | |||
349 | self._bundlefile = bundle |
|
352 | self._bundlefile = bundle | |
350 | self._cgunpacker = bundle |
|
353 | self._cgunpacker = bundle | |
351 |
|
354 | |||
352 | self.firstnewrev = self.changelog.repotiprev + 1 |
|
355 | self.firstnewrev = self.changelog.repotiprev + 1 | |
353 | phases.retractboundary( |
|
356 | phases.retractboundary( | |
354 | self, |
|
357 | self, | |
355 | None, |
|
358 | None, | |
356 | phases.draft, |
|
359 | phases.draft, | |
357 | [ctx.node() for ctx in self[self.firstnewrev :]], |
|
360 | [ctx.node() for ctx in self[self.firstnewrev :]], | |
358 | ) |
|
361 | ) | |
359 |
|
362 | |||
360 | def _handle_bundle2_cg_part(self, bundle, part): |
|
363 | def _handle_bundle2_cg_part(self, bundle, part): | |
361 | assert part.type == b'changegroup' |
|
364 | assert part.type == b'changegroup' | |
362 | cgstream = part |
|
365 | cgstream = part | |
363 | targetphase = part.params.get(b'targetphase') |
|
366 | targetphase = part.params.get(b'targetphase') | |
364 | try: |
|
367 | try: | |
365 | targetphase = int(targetphase) |
|
368 | targetphase = int(targetphase) | |
366 | except TypeError: |
|
369 | except TypeError: | |
367 | pass |
|
370 | pass | |
368 | if targetphase is None: |
|
371 | if targetphase is None: | |
369 | targetphase = phases.draft |
|
372 | targetphase = phases.draft | |
370 | if targetphase not in phases.allphases: |
|
373 | if targetphase not in phases.allphases: | |
371 | m = _(b'unsupported targetphase: %d') |
|
374 | m = _(b'unsupported targetphase: %d') | |
372 | m %= targetphase |
|
375 | m %= targetphase | |
373 | raise error.Abort(m) |
|
376 | raise error.Abort(m) | |
374 | version = part.params.get(b'version', b'01') |
|
377 | version = part.params.get(b'version', b'01') | |
375 | legalcgvers = changegroup.supportedincomingversions(self) |
|
378 | legalcgvers = changegroup.supportedincomingversions(self) | |
376 | if version not in legalcgvers: |
|
379 | if version not in legalcgvers: | |
377 | msg = _(b'Unsupported changegroup version: %s') |
|
380 | msg = _(b'Unsupported changegroup version: %s') | |
378 | raise error.Abort(msg % version) |
|
381 | raise error.Abort(msg % version) | |
379 | if bundle.compressed(): |
|
382 | if bundle.compressed(): | |
380 | cgstream = self._writetempbundle(part.read, b'.cg%sun' % version) |
|
383 | cgstream = self._writetempbundle(part.read, b'.cg%sun' % version) | |
381 |
|
384 | |||
382 | self._cgunpacker = changegroup.getunbundler(version, cgstream, b'UN') |
|
385 | self._cgunpacker = changegroup.getunbundler(version, cgstream, b'UN') | |
383 |
|
386 | |||
384 | self.firstnewrev = self.changelog.repotiprev + 1 |
|
387 | self.firstnewrev = self.changelog.repotiprev + 1 | |
385 | phases.retractboundary( |
|
388 | phases.retractboundary( | |
386 | self, |
|
389 | self, | |
387 | None, |
|
390 | None, | |
388 | targetphase, |
|
391 | targetphase, | |
389 | [ctx.node() for ctx in self[self.firstnewrev :]], |
|
392 | [ctx.node() for ctx in self[self.firstnewrev :]], | |
390 | ) |
|
393 | ) | |
391 |
|
394 | |||
392 | def _handle_bundle2_phase_part(self, bundle, part): |
|
395 | def _handle_bundle2_phase_part(self, bundle, part): | |
393 | assert part.type == b'phase-heads' |
|
396 | assert part.type == b'phase-heads' | |
394 |
|
397 | |||
395 | unfi = self.unfiltered() |
|
398 | unfi = self.unfiltered() | |
396 | headsbyphase = phases.binarydecode(part) |
|
399 | headsbyphase = phases.binarydecode(part) | |
397 | phases.updatephases(unfi, lambda: None, headsbyphase) |
|
400 | phases.updatephases(unfi, lambda: None, headsbyphase) | |
398 |
|
401 | |||
399 | def _writetempbundle(self, readfn, suffix, header=b''): |
|
402 | def _writetempbundle(self, readfn, suffix, header=b''): | |
400 | """Write a temporary file to disk""" |
|
403 | """Write a temporary file to disk""" | |
401 | fdtemp, temp = self.vfs.mkstemp(prefix=b"hg-bundle-", suffix=suffix) |
|
404 | fdtemp, temp = self.vfs.mkstemp(prefix=b"hg-bundle-", suffix=suffix) | |
402 | self.tempfile = temp |
|
405 | self.tempfile = temp | |
403 |
|
406 | |||
404 | with os.fdopen(fdtemp, 'wb') as fptemp: |
|
407 | with os.fdopen(fdtemp, 'wb') as fptemp: | |
405 | fptemp.write(header) |
|
408 | fptemp.write(header) | |
406 | while True: |
|
409 | while True: | |
407 | chunk = readfn(2 ** 18) |
|
410 | chunk = readfn(2 ** 18) | |
408 | if not chunk: |
|
411 | if not chunk: | |
409 | break |
|
412 | break | |
410 | fptemp.write(chunk) |
|
413 | fptemp.write(chunk) | |
411 |
|
414 | |||
412 | return self.vfs.open(self.tempfile, mode=b"rb") |
|
415 | return self.vfs.open(self.tempfile, mode=b"rb") | |
413 |
|
416 | |||
414 | @localrepo.unfilteredpropertycache |
|
417 | @localrepo.unfilteredpropertycache | |
415 | def _phasecache(self): |
|
418 | def _phasecache(self): | |
416 | return bundlephasecache(self, self._phasedefaults) |
|
419 | return bundlephasecache(self, self._phasedefaults) | |
417 |
|
420 | |||
418 | @localrepo.unfilteredpropertycache |
|
421 | @localrepo.unfilteredpropertycache | |
419 | def changelog(self): |
|
422 | def changelog(self): | |
420 | # consume the header if it exists |
|
423 | # consume the header if it exists | |
421 | self._cgunpacker.changelogheader() |
|
424 | self._cgunpacker.changelogheader() | |
422 | c = bundlechangelog(self.svfs, self._cgunpacker) |
|
425 | c = bundlechangelog(self.svfs, self._cgunpacker) | |
423 | self.manstart = self._cgunpacker.tell() |
|
426 | self.manstart = self._cgunpacker.tell() | |
424 | return c |
|
427 | return c | |
425 |
|
428 | |||
426 | def _refreshchangelog(self): |
|
429 | def _refreshchangelog(self): | |
427 | # changelog for bundle repo are not filecache, this method is not |
|
430 | # changelog for bundle repo are not filecache, this method is not | |
428 | # applicable. |
|
431 | # applicable. | |
429 | pass |
|
432 | pass | |
430 |
|
433 | |||
431 | @localrepo.unfilteredpropertycache |
|
434 | @localrepo.unfilteredpropertycache | |
432 | def manifestlog(self): |
|
435 | def manifestlog(self): | |
433 | self._cgunpacker.seek(self.manstart) |
|
436 | self._cgunpacker.seek(self.manstart) | |
434 | # consume the header if it exists |
|
437 | # consume the header if it exists | |
435 | self._cgunpacker.manifestheader() |
|
438 | self._cgunpacker.manifestheader() | |
436 | linkmapper = self.unfiltered().changelog.rev |
|
439 | linkmapper = self.unfiltered().changelog.rev | |
437 | rootstore = bundlemanifest( |
|
440 | rootstore = bundlemanifest( | |
438 | self.nodeconstants, self.svfs, self._cgunpacker, linkmapper |
|
441 | self.nodeconstants, self.svfs, self._cgunpacker, linkmapper | |
439 | ) |
|
442 | ) | |
440 | self.filestart = self._cgunpacker.tell() |
|
443 | self.filestart = self._cgunpacker.tell() | |
441 |
|
444 | |||
442 | return manifest.manifestlog( |
|
445 | return manifest.manifestlog( | |
443 | self.svfs, self, rootstore, self.narrowmatch() |
|
446 | self.svfs, self, rootstore, self.narrowmatch() | |
444 | ) |
|
447 | ) | |
445 |
|
448 | |||
446 | def _consumemanifest(self): |
|
449 | def _consumemanifest(self): | |
447 | """Consumes the manifest portion of the bundle, setting filestart so the |
|
450 | """Consumes the manifest portion of the bundle, setting filestart so the | |
448 | file portion can be read.""" |
|
451 | file portion can be read.""" | |
449 | self._cgunpacker.seek(self.manstart) |
|
452 | self._cgunpacker.seek(self.manstart) | |
450 | self._cgunpacker.manifestheader() |
|
453 | self._cgunpacker.manifestheader() | |
451 | for delta in self._cgunpacker.deltaiter(): |
|
454 | for delta in self._cgunpacker.deltaiter(): | |
452 | pass |
|
455 | pass | |
453 | self.filestart = self._cgunpacker.tell() |
|
456 | self.filestart = self._cgunpacker.tell() | |
454 |
|
457 | |||
455 | @localrepo.unfilteredpropertycache |
|
458 | @localrepo.unfilteredpropertycache | |
456 | def manstart(self): |
|
459 | def manstart(self): | |
457 | self.changelog |
|
460 | self.changelog | |
458 | return self.manstart |
|
461 | return self.manstart | |
459 |
|
462 | |||
460 | @localrepo.unfilteredpropertycache |
|
463 | @localrepo.unfilteredpropertycache | |
461 | def filestart(self): |
|
464 | def filestart(self): | |
462 | self.manifestlog |
|
465 | self.manifestlog | |
463 |
|
466 | |||
464 | # If filestart was not set by self.manifestlog, that means the |
|
467 | # If filestart was not set by self.manifestlog, that means the | |
465 | # manifestlog implementation did not consume the manifests from the |
|
468 | # manifestlog implementation did not consume the manifests from the | |
466 | # changegroup (ex: it might be consuming trees from a separate bundle2 |
|
469 | # changegroup (ex: it might be consuming trees from a separate bundle2 | |
467 | # part instead). So we need to manually consume it. |
|
470 | # part instead). So we need to manually consume it. | |
468 | if 'filestart' not in self.__dict__: |
|
471 | if 'filestart' not in self.__dict__: | |
469 | self._consumemanifest() |
|
472 | self._consumemanifest() | |
470 |
|
473 | |||
471 | return self.filestart |
|
474 | return self.filestart | |
472 |
|
475 | |||
473 | def url(self): |
|
476 | def url(self): | |
474 | return self._url |
|
477 | return self._url | |
475 |
|
478 | |||
476 | def file(self, f): |
|
479 | def file(self, f): | |
477 | if not self._cgfilespos: |
|
480 | if not self._cgfilespos: | |
478 | self._cgunpacker.seek(self.filestart) |
|
481 | self._cgunpacker.seek(self.filestart) | |
479 | self._cgfilespos = _getfilestarts(self._cgunpacker) |
|
482 | self._cgfilespos = _getfilestarts(self._cgunpacker) | |
480 |
|
483 | |||
481 | if f in self._cgfilespos: |
|
484 | if f in self._cgfilespos: | |
482 | self._cgunpacker.seek(self._cgfilespos[f]) |
|
485 | self._cgunpacker.seek(self._cgfilespos[f]) | |
483 | linkmapper = self.unfiltered().changelog.rev |
|
486 | linkmapper = self.unfiltered().changelog.rev | |
484 | return bundlefilelog(self.svfs, f, self._cgunpacker, linkmapper) |
|
487 | return bundlefilelog(self.svfs, f, self._cgunpacker, linkmapper) | |
485 | else: |
|
488 | else: | |
486 | return super(bundlerepository, self).file(f) |
|
489 | return super(bundlerepository, self).file(f) | |
487 |
|
490 | |||
488 | def close(self): |
|
491 | def close(self): | |
489 | """Close assigned bundle file immediately.""" |
|
492 | """Close assigned bundle file immediately.""" | |
490 | self._bundlefile.close() |
|
493 | self._bundlefile.close() | |
491 | if self.tempfile is not None: |
|
494 | if self.tempfile is not None: | |
492 | self.vfs.unlink(self.tempfile) |
|
495 | self.vfs.unlink(self.tempfile) | |
493 | if self._tempparent: |
|
496 | if self._tempparent: | |
494 | shutil.rmtree(self._tempparent, True) |
|
497 | shutil.rmtree(self._tempparent, True) | |
495 |
|
498 | |||
496 | def cancopy(self): |
|
499 | def cancopy(self): | |
497 | return False |
|
500 | return False | |
498 |
|
501 | |||
499 | def peer(self, path=None, remotehidden=False): |
|
502 | def peer(self, path=None, remotehidden=False): | |
500 | return bundlepeer(self, path=path, remotehidden=remotehidden) |
|
503 | return bundlepeer(self, path=path, remotehidden=remotehidden) | |
501 |
|
504 | |||
502 | def getcwd(self): |
|
505 | def getcwd(self): | |
503 | return encoding.getcwd() # always outside the repo |
|
506 | return encoding.getcwd() # always outside the repo | |
504 |
|
507 | |||
505 | # Check if parents exist in localrepo before setting |
|
508 | # Check if parents exist in localrepo before setting | |
506 | def setparents(self, p1, p2=None): |
|
509 | def setparents(self, p1, p2=None): | |
507 | if p2 is None: |
|
510 | if p2 is None: | |
508 | p2 = self.nullid |
|
511 | p2 = self.nullid | |
509 | p1rev = self.changelog.rev(p1) |
|
512 | p1rev = self.changelog.rev(p1) | |
510 | p2rev = self.changelog.rev(p2) |
|
513 | p2rev = self.changelog.rev(p2) | |
511 | msg = _(b"setting parent to node %s that only exists in the bundle\n") |
|
514 | msg = _(b"setting parent to node %s that only exists in the bundle\n") | |
512 | if self.changelog.repotiprev < p1rev: |
|
515 | if self.changelog.repotiprev < p1rev: | |
513 | self.ui.warn(msg % hex(p1)) |
|
516 | self.ui.warn(msg % hex(p1)) | |
514 | if self.changelog.repotiprev < p2rev: |
|
517 | if self.changelog.repotiprev < p2rev: | |
515 | self.ui.warn(msg % hex(p2)) |
|
518 | self.ui.warn(msg % hex(p2)) | |
516 | return super(bundlerepository, self).setparents(p1, p2) |
|
519 | return super(bundlerepository, self).setparents(p1, p2) | |
517 |
|
520 | |||
518 |
|
521 | |||
519 | def instance(ui, path, create, intents=None, createopts=None): |
|
522 | def instance(ui, path, create, intents=None, createopts=None): | |
520 | if create: |
|
523 | if create: | |
521 | raise error.Abort(_(b'cannot create new bundle repository')) |
|
524 | raise error.Abort(_(b'cannot create new bundle repository')) | |
522 | # internal config: bundle.mainreporoot |
|
525 | # internal config: bundle.mainreporoot | |
523 | parentpath = ui.config(b"bundle", b"mainreporoot") |
|
526 | parentpath = ui.config(b"bundle", b"mainreporoot") | |
524 | if not parentpath: |
|
527 | if not parentpath: | |
525 | # try to find the correct path to the working directory repo |
|
528 | # try to find the correct path to the working directory repo | |
526 | parentpath = cmdutil.findrepo(encoding.getcwd()) |
|
529 | parentpath = cmdutil.findrepo(encoding.getcwd()) | |
527 | if parentpath is None: |
|
530 | if parentpath is None: | |
528 | parentpath = b'' |
|
531 | parentpath = b'' | |
529 | if parentpath: |
|
532 | if parentpath: | |
530 | # Try to make the full path relative so we get a nice, short URL. |
|
533 | # Try to make the full path relative so we get a nice, short URL. | |
531 | # In particular, we don't want temp dir names in test outputs. |
|
534 | # In particular, we don't want temp dir names in test outputs. | |
532 | cwd = encoding.getcwd() |
|
535 | cwd = encoding.getcwd() | |
533 | if parentpath == cwd: |
|
536 | if parentpath == cwd: | |
534 | parentpath = b'' |
|
537 | parentpath = b'' | |
535 | else: |
|
538 | else: | |
536 | cwd = pathutil.normasprefix(cwd) |
|
539 | cwd = pathutil.normasprefix(cwd) | |
537 | if parentpath.startswith(cwd): |
|
540 | if parentpath.startswith(cwd): | |
538 | parentpath = parentpath[len(cwd) :] |
|
541 | parentpath = parentpath[len(cwd) :] | |
539 | u = urlutil.url(path) |
|
542 | u = urlutil.url(path) | |
540 | path = u.localpath() |
|
543 | path = u.localpath() | |
541 | if u.scheme == b'bundle': |
|
544 | if u.scheme == b'bundle': | |
542 | s = path.split(b"+", 1) |
|
545 | s = path.split(b"+", 1) | |
543 | if len(s) == 1: |
|
546 | if len(s) == 1: | |
544 | repopath, bundlename = parentpath, s[0] |
|
547 | repopath, bundlename = parentpath, s[0] | |
545 | else: |
|
548 | else: | |
546 | repopath, bundlename = s |
|
549 | repopath, bundlename = s | |
547 | else: |
|
550 | else: | |
548 | repopath, bundlename = parentpath, path |
|
551 | repopath, bundlename = parentpath, path | |
549 |
|
552 | |||
550 | return makebundlerepository(ui, repopath, bundlename) |
|
553 | return makebundlerepository(ui, repopath, bundlename) | |
551 |
|
554 | |||
552 |
|
555 | |||
553 | def makebundlerepository(ui, repopath, bundlepath): |
|
556 | def makebundlerepository(ui, repopath, bundlepath): | |
554 | """Make a bundle repository object based on repo and bundle paths.""" |
|
557 | """Make a bundle repository object based on repo and bundle paths.""" | |
555 | if repopath: |
|
558 | if repopath: | |
556 | url = b'bundle:%s+%s' % (util.expandpath(repopath), bundlepath) |
|
559 | url = b'bundle:%s+%s' % (util.expandpath(repopath), bundlepath) | |
557 | else: |
|
560 | else: | |
558 | url = b'bundle:%s' % bundlepath |
|
561 | url = b'bundle:%s' % bundlepath | |
559 |
|
562 | |||
560 | # Because we can't make any guarantees about the type of the base |
|
563 | # Because we can't make any guarantees about the type of the base | |
561 | # repository, we can't have a static class representing the bundle |
|
564 | # repository, we can't have a static class representing the bundle | |
562 | # repository. We also can't make any guarantees about how to even |
|
565 | # repository. We also can't make any guarantees about how to even | |
563 | # call the base repository's constructor! |
|
566 | # call the base repository's constructor! | |
564 | # |
|
567 | # | |
565 | # So, our strategy is to go through ``localrepo.instance()`` to construct |
|
568 | # So, our strategy is to go through ``localrepo.instance()`` to construct | |
566 | # a repo instance. Then, we dynamically create a new type derived from |
|
569 | # a repo instance. Then, we dynamically create a new type derived from | |
567 | # both it and our ``bundlerepository`` class which overrides some |
|
570 | # both it and our ``bundlerepository`` class which overrides some | |
568 | # functionality. We then change the type of the constructed repository |
|
571 | # functionality. We then change the type of the constructed repository | |
569 | # to this new type and initialize the bundle-specific bits of it. |
|
572 | # to this new type and initialize the bundle-specific bits of it. | |
570 |
|
573 | |||
571 | try: |
|
574 | try: | |
572 | repo = localrepo.instance(ui, repopath, create=False) |
|
575 | repo = localrepo.instance(ui, repopath, create=False) | |
573 | tempparent = None |
|
576 | tempparent = None | |
574 | except error.RequirementError: |
|
577 | except error.RequirementError: | |
575 | raise # no fallback if the backing repo is unsupported |
|
578 | raise # no fallback if the backing repo is unsupported | |
576 | except error.RepoError: |
|
579 | except error.RepoError: | |
577 | tempparent = pycompat.mkdtemp() |
|
580 | tempparent = pycompat.mkdtemp() | |
578 | try: |
|
581 | try: | |
579 | repo = localrepo.instance(ui, tempparent, create=True) |
|
582 | repo = localrepo.instance(ui, tempparent, create=True) | |
580 | except Exception: |
|
583 | except Exception: | |
581 | shutil.rmtree(tempparent) |
|
584 | shutil.rmtree(tempparent) | |
582 | raise |
|
585 | raise | |
583 |
|
586 | |||
584 | class derivedbundlerepository(bundlerepository, repo.__class__): |
|
587 | class derivedbundlerepository(bundlerepository, repo.__class__): | |
585 | pass |
|
588 | pass | |
586 |
|
589 | |||
587 | repo.__class__ = derivedbundlerepository |
|
590 | repo.__class__ = derivedbundlerepository | |
588 | bundlerepository.__init__(repo, bundlepath, url, tempparent) |
|
591 | bundlerepository.__init__(repo, bundlepath, url, tempparent) | |
589 |
|
592 | |||
590 | return repo |
|
593 | return repo | |
591 |
|
594 | |||
592 |
|
595 | |||
593 | class bundletransactionmanager: |
|
596 | class bundletransactionmanager: | |
594 | def transaction(self): |
|
597 | def transaction(self): | |
595 | return None |
|
598 | return None | |
596 |
|
599 | |||
597 | def close(self): |
|
600 | def close(self): | |
598 | raise NotImplementedError |
|
601 | raise NotImplementedError | |
599 |
|
602 | |||
600 | def release(self): |
|
603 | def release(self): | |
601 | raise NotImplementedError |
|
604 | raise NotImplementedError | |
602 |
|
605 | |||
603 |
|
606 | |||
604 | def getremotechanges( |
|
607 | def getremotechanges( | |
605 | ui, repo, peer, onlyheads=None, bundlename=None, force=False |
|
608 | ui, repo, peer, onlyheads=None, bundlename=None, force=False | |
606 | ): |
|
609 | ): | |
607 | """obtains a bundle of changes incoming from peer |
|
610 | """obtains a bundle of changes incoming from peer | |
608 |
|
611 | |||
609 | "onlyheads" restricts the returned changes to those reachable from the |
|
612 | "onlyheads" restricts the returned changes to those reachable from the | |
610 | specified heads. |
|
613 | specified heads. | |
611 | "bundlename", if given, stores the bundle to this file path permanently; |
|
614 | "bundlename", if given, stores the bundle to this file path permanently; | |
612 | otherwise it's stored to a temp file and gets deleted again when you call |
|
615 | otherwise it's stored to a temp file and gets deleted again when you call | |
613 | the returned "cleanupfn". |
|
616 | the returned "cleanupfn". | |
614 | "force" indicates whether to proceed on unrelated repos. |
|
617 | "force" indicates whether to proceed on unrelated repos. | |
615 |
|
618 | |||
616 | Returns a tuple (local, csets, cleanupfn): |
|
619 | Returns a tuple (local, csets, cleanupfn): | |
617 |
|
620 | |||
618 | "local" is a local repo from which to obtain the actual incoming |
|
621 | "local" is a local repo from which to obtain the actual incoming | |
619 | changesets; it is a bundlerepo for the obtained bundle when the |
|
622 | changesets; it is a bundlerepo for the obtained bundle when the | |
620 | original "peer" is remote. |
|
623 | original "peer" is remote. | |
621 | "csets" lists the incoming changeset node ids. |
|
624 | "csets" lists the incoming changeset node ids. | |
622 | "cleanupfn" must be called without arguments when you're done processing |
|
625 | "cleanupfn" must be called without arguments when you're done processing | |
623 | the changes; it closes both the original "peer" and the one returned |
|
626 | the changes; it closes both the original "peer" and the one returned | |
624 | here. |
|
627 | here. | |
625 | """ |
|
628 | """ | |
626 | tmp = discovery.findcommonincoming(repo, peer, heads=onlyheads, force=force) |
|
629 | tmp = discovery.findcommonincoming(repo, peer, heads=onlyheads, force=force) | |
627 | common, incoming, rheads = tmp |
|
630 | common, incoming, rheads = tmp | |
628 | if not incoming: |
|
631 | if not incoming: | |
629 | try: |
|
632 | try: | |
630 | if bundlename: |
|
633 | if bundlename: | |
631 | os.unlink(bundlename) |
|
634 | os.unlink(bundlename) | |
632 | except OSError: |
|
635 | except OSError: | |
633 | pass |
|
636 | pass | |
634 | return repo, [], peer.close |
|
637 | return repo, [], peer.close | |
635 |
|
638 | |||
636 | commonset = set(common) |
|
639 | commonset = set(common) | |
637 | rheads = [x for x in rheads if x not in commonset] |
|
640 | rheads = [x for x in rheads if x not in commonset] | |
638 |
|
641 | |||
639 | bundle = None |
|
642 | bundle = None | |
640 | bundlerepo = None |
|
643 | bundlerepo = None | |
641 | localrepo = peer.local() |
|
644 | localrepo = peer.local() | |
642 | if bundlename or not localrepo: |
|
645 | if bundlename or not localrepo: | |
643 | # create a bundle (uncompressed if peer repo is not local) |
|
646 | # create a bundle (uncompressed if peer repo is not local) | |
644 |
|
647 | |||
645 | # developer config: devel.legacy.exchange |
|
648 | # developer config: devel.legacy.exchange | |
646 | legexc = ui.configlist(b'devel', b'legacy.exchange') |
|
649 | legexc = ui.configlist(b'devel', b'legacy.exchange') | |
647 | forcebundle1 = b'bundle2' not in legexc and b'bundle1' in legexc |
|
650 | forcebundle1 = b'bundle2' not in legexc and b'bundle1' in legexc | |
648 | canbundle2 = ( |
|
651 | canbundle2 = ( | |
649 | not forcebundle1 |
|
652 | not forcebundle1 | |
650 | and peer.capable(b'getbundle') |
|
653 | and peer.capable(b'getbundle') | |
651 | and peer.capable(b'bundle2') |
|
654 | and peer.capable(b'bundle2') | |
652 | ) |
|
655 | ) | |
653 | if canbundle2: |
|
656 | if canbundle2: | |
654 | with peer.commandexecutor() as e: |
|
657 | with peer.commandexecutor() as e: | |
655 | b2 = e.callcommand( |
|
658 | b2 = e.callcommand( | |
656 | b'getbundle', |
|
659 | b'getbundle', | |
657 | { |
|
660 | { | |
658 | b'source': b'incoming', |
|
661 | b'source': b'incoming', | |
659 | b'common': common, |
|
662 | b'common': common, | |
660 | b'heads': rheads, |
|
663 | b'heads': rheads, | |
661 | b'bundlecaps': exchange.caps20to10( |
|
664 | b'bundlecaps': exchange.caps20to10( | |
662 | repo, role=b'client' |
|
665 | repo, role=b'client' | |
663 | ), |
|
666 | ), | |
664 | b'cg': True, |
|
667 | b'cg': True, | |
665 | }, |
|
668 | }, | |
666 | ).result() |
|
669 | ).result() | |
667 |
|
670 | |||
668 | fname = bundle = changegroup.writechunks( |
|
671 | fname = bundle = changegroup.writechunks( | |
669 | ui, b2._forwardchunks(), bundlename |
|
672 | ui, b2._forwardchunks(), bundlename | |
670 | ) |
|
673 | ) | |
671 | else: |
|
674 | else: | |
672 | if peer.capable(b'getbundle'): |
|
675 | if peer.capable(b'getbundle'): | |
673 | with peer.commandexecutor() as e: |
|
676 | with peer.commandexecutor() as e: | |
674 | cg = e.callcommand( |
|
677 | cg = e.callcommand( | |
675 | b'getbundle', |
|
678 | b'getbundle', | |
676 | { |
|
679 | { | |
677 | b'source': b'incoming', |
|
680 | b'source': b'incoming', | |
678 | b'common': common, |
|
681 | b'common': common, | |
679 | b'heads': rheads, |
|
682 | b'heads': rheads, | |
680 | }, |
|
683 | }, | |
681 | ).result() |
|
684 | ).result() | |
682 | elif onlyheads is None and not peer.capable(b'changegroupsubset'): |
|
685 | elif onlyheads is None and not peer.capable(b'changegroupsubset'): | |
683 | # compat with older servers when pulling all remote heads |
|
686 | # compat with older servers when pulling all remote heads | |
684 |
|
687 | |||
685 | with peer.commandexecutor() as e: |
|
688 | with peer.commandexecutor() as e: | |
686 | cg = e.callcommand( |
|
689 | cg = e.callcommand( | |
687 | b'changegroup', |
|
690 | b'changegroup', | |
688 | { |
|
691 | { | |
689 | b'nodes': incoming, |
|
692 | b'nodes': incoming, | |
690 | b'source': b'incoming', |
|
693 | b'source': b'incoming', | |
691 | }, |
|
694 | }, | |
692 | ).result() |
|
695 | ).result() | |
693 |
|
696 | |||
694 | rheads = None |
|
697 | rheads = None | |
695 | else: |
|
698 | else: | |
696 | with peer.commandexecutor() as e: |
|
699 | with peer.commandexecutor() as e: | |
697 | cg = e.callcommand( |
|
700 | cg = e.callcommand( | |
698 | b'changegroupsubset', |
|
701 | b'changegroupsubset', | |
699 | { |
|
702 | { | |
700 | b'bases': incoming, |
|
703 | b'bases': incoming, | |
701 | b'heads': rheads, |
|
704 | b'heads': rheads, | |
702 | b'source': b'incoming', |
|
705 | b'source': b'incoming', | |
703 | }, |
|
706 | }, | |
704 | ).result() |
|
707 | ).result() | |
705 |
|
708 | |||
706 | if localrepo: |
|
709 | if localrepo: | |
707 | bundletype = b"HG10BZ" |
|
710 | bundletype = b"HG10BZ" | |
708 | else: |
|
711 | else: | |
709 | bundletype = b"HG10UN" |
|
712 | bundletype = b"HG10UN" | |
710 | fname = bundle = bundle2.writebundle(ui, cg, bundlename, bundletype) |
|
713 | fname = bundle = bundle2.writebundle(ui, cg, bundlename, bundletype) | |
711 | # keep written bundle? |
|
714 | # keep written bundle? | |
712 | if bundlename: |
|
715 | if bundlename: | |
713 | bundle = None |
|
716 | bundle = None | |
714 | if not localrepo: |
|
717 | if not localrepo: | |
715 | # use the created uncompressed bundlerepo |
|
718 | # use the created uncompressed bundlerepo | |
716 | localrepo = bundlerepo = makebundlerepository( |
|
719 | localrepo = bundlerepo = makebundlerepository( | |
717 | repo.baseui, repo.root, fname |
|
720 | repo.baseui, repo.root, fname | |
718 | ) |
|
721 | ) | |
719 |
|
722 | |||
720 | # this repo contains local and peer now, so filter out local again |
|
723 | # this repo contains local and peer now, so filter out local again | |
721 | common = repo.heads() |
|
724 | common = repo.heads() | |
722 | if localrepo: |
|
725 | if localrepo: | |
723 | # Part of common may be remotely filtered |
|
726 | # Part of common may be remotely filtered | |
724 | # So use an unfiltered version |
|
727 | # So use an unfiltered version | |
725 | # The discovery process probably need cleanup to avoid that |
|
728 | # The discovery process probably need cleanup to avoid that | |
726 | localrepo = localrepo.unfiltered() |
|
729 | localrepo = localrepo.unfiltered() | |
727 |
|
730 | |||
728 | csets = localrepo.changelog.findmissing(common, rheads) |
|
731 | csets = localrepo.changelog.findmissing(common, rheads) | |
729 |
|
732 | |||
730 | if bundlerepo: |
|
733 | if bundlerepo: | |
731 | reponodes = [ctx.node() for ctx in bundlerepo[bundlerepo.firstnewrev :]] |
|
734 | reponodes = [ctx.node() for ctx in bundlerepo[bundlerepo.firstnewrev :]] | |
732 |
|
735 | |||
733 | with peer.commandexecutor() as e: |
|
736 | with peer.commandexecutor() as e: | |
734 | remotephases = e.callcommand( |
|
737 | remotephases = e.callcommand( | |
735 | b'listkeys', |
|
738 | b'listkeys', | |
736 | { |
|
739 | { | |
737 | b'namespace': b'phases', |
|
740 | b'namespace': b'phases', | |
738 | }, |
|
741 | }, | |
739 | ).result() |
|
742 | ).result() | |
740 |
|
743 | |||
741 | pullop = exchange.pulloperation( |
|
744 | pullop = exchange.pulloperation( | |
742 | bundlerepo, peer, path=None, heads=reponodes |
|
745 | bundlerepo, peer, path=None, heads=reponodes | |
743 | ) |
|
746 | ) | |
744 | pullop.trmanager = bundletransactionmanager() |
|
747 | pullop.trmanager = bundletransactionmanager() | |
745 | exchange._pullapplyphases(pullop, remotephases) |
|
748 | exchange._pullapplyphases(pullop, remotephases) | |
746 |
|
749 | |||
747 | def cleanup(): |
|
750 | def cleanup(): | |
748 | if bundlerepo: |
|
751 | if bundlerepo: | |
749 | bundlerepo.close() |
|
752 | bundlerepo.close() | |
750 | if bundle: |
|
753 | if bundle: | |
751 | os.unlink(bundle) |
|
754 | os.unlink(bundle) | |
752 | peer.close() |
|
755 | peer.close() | |
753 |
|
756 | |||
754 | return (localrepo, csets, cleanup) |
|
757 | return (localrepo, csets, cleanup) |
@@ -1,3971 +1,3974 b'' | |||||
1 | # revlog.py - storage back-end for mercurial |
|
1 | # revlog.py - storage back-end for mercurial | |
2 | # coding: utf8 |
|
2 | # coding: utf8 | |
3 | # |
|
3 | # | |
4 | # Copyright 2005-2007 Olivia Mackall <olivia@selenic.com> |
|
4 | # Copyright 2005-2007 Olivia Mackall <olivia@selenic.com> | |
5 | # |
|
5 | # | |
6 | # This software may be used and distributed according to the terms of the |
|
6 | # This software may be used and distributed according to the terms of the | |
7 | # GNU General Public License version 2 or any later version. |
|
7 | # GNU General Public License version 2 or any later version. | |
8 |
|
8 | |||
9 | """Storage back-end for Mercurial. |
|
9 | """Storage back-end for Mercurial. | |
10 |
|
10 | |||
11 | This provides efficient delta storage with O(1) retrieve and append |
|
11 | This provides efficient delta storage with O(1) retrieve and append | |
12 | and O(changes) merge between branches. |
|
12 | and O(changes) merge between branches. | |
13 | """ |
|
13 | """ | |
14 |
|
14 | |||
15 |
|
15 | |||
16 | import binascii |
|
16 | import binascii | |
17 | import collections |
|
17 | import collections | |
18 | import contextlib |
|
18 | import contextlib | |
19 | import io |
|
19 | import io | |
20 | import os |
|
20 | import os | |
21 | import struct |
|
21 | import struct | |
22 | import weakref |
|
22 | import weakref | |
23 | import zlib |
|
23 | import zlib | |
24 |
|
24 | |||
25 | # import stuff from node for others to import from revlog |
|
25 | # import stuff from node for others to import from revlog | |
26 | from .node import ( |
|
26 | from .node import ( | |
27 | bin, |
|
27 | bin, | |
28 | hex, |
|
28 | hex, | |
29 | nullrev, |
|
29 | nullrev, | |
30 | sha1nodeconstants, |
|
30 | sha1nodeconstants, | |
31 | short, |
|
31 | short, | |
32 | wdirrev, |
|
32 | wdirrev, | |
33 | ) |
|
33 | ) | |
34 | from .i18n import _ |
|
34 | from .i18n import _ | |
35 | from .revlogutils.constants import ( |
|
35 | from .revlogutils.constants import ( | |
36 | ALL_KINDS, |
|
36 | ALL_KINDS, | |
37 | CHANGELOGV2, |
|
37 | CHANGELOGV2, | |
38 | COMP_MODE_DEFAULT, |
|
38 | COMP_MODE_DEFAULT, | |
39 | COMP_MODE_INLINE, |
|
39 | COMP_MODE_INLINE, | |
40 | COMP_MODE_PLAIN, |
|
40 | COMP_MODE_PLAIN, | |
41 | DELTA_BASE_REUSE_NO, |
|
41 | DELTA_BASE_REUSE_NO, | |
42 | DELTA_BASE_REUSE_TRY, |
|
42 | DELTA_BASE_REUSE_TRY, | |
43 | ENTRY_RANK, |
|
43 | ENTRY_RANK, | |
44 | FEATURES_BY_VERSION, |
|
44 | FEATURES_BY_VERSION, | |
45 | FLAG_GENERALDELTA, |
|
45 | FLAG_GENERALDELTA, | |
46 | FLAG_INLINE_DATA, |
|
46 | FLAG_INLINE_DATA, | |
47 | INDEX_HEADER, |
|
47 | INDEX_HEADER, | |
48 | KIND_CHANGELOG, |
|
48 | KIND_CHANGELOG, | |
49 | KIND_FILELOG, |
|
49 | KIND_FILELOG, | |
50 | RANK_UNKNOWN, |
|
50 | RANK_UNKNOWN, | |
51 | REVLOGV0, |
|
51 | REVLOGV0, | |
52 | REVLOGV1, |
|
52 | REVLOGV1, | |
53 | REVLOGV1_FLAGS, |
|
53 | REVLOGV1_FLAGS, | |
54 | REVLOGV2, |
|
54 | REVLOGV2, | |
55 | REVLOGV2_FLAGS, |
|
55 | REVLOGV2_FLAGS, | |
56 | REVLOG_DEFAULT_FLAGS, |
|
56 | REVLOG_DEFAULT_FLAGS, | |
57 | REVLOG_DEFAULT_FORMAT, |
|
57 | REVLOG_DEFAULT_FORMAT, | |
58 | REVLOG_DEFAULT_VERSION, |
|
58 | REVLOG_DEFAULT_VERSION, | |
59 | SUPPORTED_FLAGS, |
|
59 | SUPPORTED_FLAGS, | |
60 | ) |
|
60 | ) | |
61 | from .revlogutils.flagutil import ( |
|
61 | from .revlogutils.flagutil import ( | |
62 | REVIDX_DEFAULT_FLAGS, |
|
62 | REVIDX_DEFAULT_FLAGS, | |
63 | REVIDX_ELLIPSIS, |
|
63 | REVIDX_ELLIPSIS, | |
64 | REVIDX_EXTSTORED, |
|
64 | REVIDX_EXTSTORED, | |
65 | REVIDX_FLAGS_ORDER, |
|
65 | REVIDX_FLAGS_ORDER, | |
66 | REVIDX_HASCOPIESINFO, |
|
66 | REVIDX_HASCOPIESINFO, | |
67 | REVIDX_ISCENSORED, |
|
67 | REVIDX_ISCENSORED, | |
68 | REVIDX_RAWTEXT_CHANGING_FLAGS, |
|
68 | REVIDX_RAWTEXT_CHANGING_FLAGS, | |
69 | ) |
|
69 | ) | |
70 | from .thirdparty import attr |
|
70 | from .thirdparty import attr | |
71 | from . import ( |
|
71 | from . import ( | |
72 | ancestor, |
|
72 | ancestor, | |
73 | dagop, |
|
73 | dagop, | |
74 | error, |
|
74 | error, | |
75 | mdiff, |
|
75 | mdiff, | |
76 | policy, |
|
76 | policy, | |
77 | pycompat, |
|
77 | pycompat, | |
78 | revlogutils, |
|
78 | revlogutils, | |
79 | templatefilters, |
|
79 | templatefilters, | |
80 | util, |
|
80 | util, | |
81 | ) |
|
81 | ) | |
82 | from .interfaces import ( |
|
82 | from .interfaces import ( | |
83 | repository, |
|
83 | repository, | |
84 | util as interfaceutil, |
|
84 | util as interfaceutil, | |
85 | ) |
|
85 | ) | |
86 | from .revlogutils import ( |
|
86 | from .revlogutils import ( | |
87 | deltas as deltautil, |
|
87 | deltas as deltautil, | |
88 | docket as docketutil, |
|
88 | docket as docketutil, | |
89 | flagutil, |
|
89 | flagutil, | |
90 | nodemap as nodemaputil, |
|
90 | nodemap as nodemaputil, | |
91 | randomaccessfile, |
|
91 | randomaccessfile, | |
92 | revlogv0, |
|
92 | revlogv0, | |
93 | rewrite, |
|
93 | rewrite, | |
94 | sidedata as sidedatautil, |
|
94 | sidedata as sidedatautil, | |
95 | ) |
|
95 | ) | |
96 | from .utils import ( |
|
96 | from .utils import ( | |
97 | storageutil, |
|
97 | storageutil, | |
98 | stringutil, |
|
98 | stringutil, | |
99 | ) |
|
99 | ) | |
100 |
|
100 | |||
101 | # blanked usage of all the name to prevent pyflakes constraints |
|
101 | # blanked usage of all the name to prevent pyflakes constraints | |
102 | # We need these name available in the module for extensions. |
|
102 | # We need these name available in the module for extensions. | |
103 |
|
103 | |||
104 | REVLOGV0 |
|
104 | REVLOGV0 | |
105 | REVLOGV1 |
|
105 | REVLOGV1 | |
106 | REVLOGV2 |
|
106 | REVLOGV2 | |
107 | CHANGELOGV2 |
|
107 | CHANGELOGV2 | |
108 | FLAG_INLINE_DATA |
|
108 | FLAG_INLINE_DATA | |
109 | FLAG_GENERALDELTA |
|
109 | FLAG_GENERALDELTA | |
110 | REVLOG_DEFAULT_FLAGS |
|
110 | REVLOG_DEFAULT_FLAGS | |
111 | REVLOG_DEFAULT_FORMAT |
|
111 | REVLOG_DEFAULT_FORMAT | |
112 | REVLOG_DEFAULT_VERSION |
|
112 | REVLOG_DEFAULT_VERSION | |
113 | REVLOGV1_FLAGS |
|
113 | REVLOGV1_FLAGS | |
114 | REVLOGV2_FLAGS |
|
114 | REVLOGV2_FLAGS | |
115 | REVIDX_ISCENSORED |
|
115 | REVIDX_ISCENSORED | |
116 | REVIDX_ELLIPSIS |
|
116 | REVIDX_ELLIPSIS | |
117 | REVIDX_HASCOPIESINFO |
|
117 | REVIDX_HASCOPIESINFO | |
118 | REVIDX_EXTSTORED |
|
118 | REVIDX_EXTSTORED | |
119 | REVIDX_DEFAULT_FLAGS |
|
119 | REVIDX_DEFAULT_FLAGS | |
120 | REVIDX_FLAGS_ORDER |
|
120 | REVIDX_FLAGS_ORDER | |
121 | REVIDX_RAWTEXT_CHANGING_FLAGS |
|
121 | REVIDX_RAWTEXT_CHANGING_FLAGS | |
122 |
|
122 | |||
123 | parsers = policy.importmod('parsers') |
|
123 | parsers = policy.importmod('parsers') | |
124 | rustancestor = policy.importrust('ancestor') |
|
124 | rustancestor = policy.importrust('ancestor') | |
125 | rustdagop = policy.importrust('dagop') |
|
125 | rustdagop = policy.importrust('dagop') | |
126 | rustrevlog = policy.importrust('revlog') |
|
126 | rustrevlog = policy.importrust('revlog') | |
127 |
|
127 | |||
128 | # Aliased for performance. |
|
128 | # Aliased for performance. | |
129 | _zlibdecompress = zlib.decompress |
|
129 | _zlibdecompress = zlib.decompress | |
130 |
|
130 | |||
131 | # max size of inline data embedded into a revlog |
|
131 | # max size of inline data embedded into a revlog | |
132 | _maxinline = 131072 |
|
132 | _maxinline = 131072 | |
133 |
|
133 | |||
134 | # Flag processors for REVIDX_ELLIPSIS. |
|
134 | # Flag processors for REVIDX_ELLIPSIS. | |
135 | def ellipsisreadprocessor(rl, text): |
|
135 | def ellipsisreadprocessor(rl, text): | |
136 | return text, False |
|
136 | return text, False | |
137 |
|
137 | |||
138 |
|
138 | |||
139 | def ellipsiswriteprocessor(rl, text): |
|
139 | def ellipsiswriteprocessor(rl, text): | |
140 | return text, False |
|
140 | return text, False | |
141 |
|
141 | |||
142 |
|
142 | |||
143 | def ellipsisrawprocessor(rl, text): |
|
143 | def ellipsisrawprocessor(rl, text): | |
144 | return False |
|
144 | return False | |
145 |
|
145 | |||
146 |
|
146 | |||
147 | ellipsisprocessor = ( |
|
147 | ellipsisprocessor = ( | |
148 | ellipsisreadprocessor, |
|
148 | ellipsisreadprocessor, | |
149 | ellipsiswriteprocessor, |
|
149 | ellipsiswriteprocessor, | |
150 | ellipsisrawprocessor, |
|
150 | ellipsisrawprocessor, | |
151 | ) |
|
151 | ) | |
152 |
|
152 | |||
153 |
|
153 | |||
154 | def _verify_revision(rl, skipflags, state, node): |
|
154 | def _verify_revision(rl, skipflags, state, node): | |
155 | """Verify the integrity of the given revlog ``node`` while providing a hook |
|
155 | """Verify the integrity of the given revlog ``node`` while providing a hook | |
156 | point for extensions to influence the operation.""" |
|
156 | point for extensions to influence the operation.""" | |
157 | if skipflags: |
|
157 | if skipflags: | |
158 | state[b'skipread'].add(node) |
|
158 | state[b'skipread'].add(node) | |
159 | else: |
|
159 | else: | |
160 | # Side-effect: read content and verify hash. |
|
160 | # Side-effect: read content and verify hash. | |
161 | rl.revision(node) |
|
161 | rl.revision(node) | |
162 |
|
162 | |||
163 |
|
163 | |||
164 | # True if a fast implementation for persistent-nodemap is available |
|
164 | # True if a fast implementation for persistent-nodemap is available | |
165 | # |
|
165 | # | |
166 | # We also consider we have a "fast" implementation in "pure" python because |
|
166 | # We also consider we have a "fast" implementation in "pure" python because | |
167 | # people using pure don't really have performance consideration (and a |
|
167 | # people using pure don't really have performance consideration (and a | |
168 | # wheelbarrow of other slowness source) |
|
168 | # wheelbarrow of other slowness source) | |
169 | HAS_FAST_PERSISTENT_NODEMAP = rustrevlog is not None or hasattr( |
|
169 | HAS_FAST_PERSISTENT_NODEMAP = rustrevlog is not None or hasattr( | |
170 | parsers, 'BaseIndexObject' |
|
170 | parsers, 'BaseIndexObject' | |
171 | ) |
|
171 | ) | |
172 |
|
172 | |||
173 |
|
173 | |||
174 | @interfaceutil.implementer(repository.irevisiondelta) |
|
174 | @interfaceutil.implementer(repository.irevisiondelta) | |
175 | @attr.s(slots=True) |
|
175 | @attr.s(slots=True) | |
176 | class revlogrevisiondelta: |
|
176 | class revlogrevisiondelta: | |
177 | node = attr.ib() |
|
177 | node = attr.ib() | |
178 | p1node = attr.ib() |
|
178 | p1node = attr.ib() | |
179 | p2node = attr.ib() |
|
179 | p2node = attr.ib() | |
180 | basenode = attr.ib() |
|
180 | basenode = attr.ib() | |
181 | flags = attr.ib() |
|
181 | flags = attr.ib() | |
182 | baserevisionsize = attr.ib() |
|
182 | baserevisionsize = attr.ib() | |
183 | revision = attr.ib() |
|
183 | revision = attr.ib() | |
184 | delta = attr.ib() |
|
184 | delta = attr.ib() | |
185 | sidedata = attr.ib() |
|
185 | sidedata = attr.ib() | |
186 | protocol_flags = attr.ib() |
|
186 | protocol_flags = attr.ib() | |
187 | linknode = attr.ib(default=None) |
|
187 | linknode = attr.ib(default=None) | |
188 |
|
188 | |||
189 |
|
189 | |||
190 | @interfaceutil.implementer(repository.iverifyproblem) |
|
190 | @interfaceutil.implementer(repository.iverifyproblem) | |
191 | @attr.s(frozen=True) |
|
191 | @attr.s(frozen=True) | |
192 | class revlogproblem: |
|
192 | class revlogproblem: | |
193 | warning = attr.ib(default=None) |
|
193 | warning = attr.ib(default=None) | |
194 | error = attr.ib(default=None) |
|
194 | error = attr.ib(default=None) | |
195 | node = attr.ib(default=None) |
|
195 | node = attr.ib(default=None) | |
196 |
|
196 | |||
197 |
|
197 | |||
198 | def parse_index_v1(data, inline): |
|
198 | def parse_index_v1(data, inline): | |
199 | # call the C implementation to parse the index data |
|
199 | # call the C implementation to parse the index data | |
200 | index, cache = parsers.parse_index2(data, inline) |
|
200 | index, cache = parsers.parse_index2(data, inline) | |
201 | return index, cache |
|
201 | return index, cache | |
202 |
|
202 | |||
203 |
|
203 | |||
204 | def parse_index_v2(data, inline): |
|
204 | def parse_index_v2(data, inline): | |
205 | # call the C implementation to parse the index data |
|
205 | # call the C implementation to parse the index data | |
206 | index, cache = parsers.parse_index2(data, inline, format=REVLOGV2) |
|
206 | index, cache = parsers.parse_index2(data, inline, format=REVLOGV2) | |
207 | return index, cache |
|
207 | return index, cache | |
208 |
|
208 | |||
209 |
|
209 | |||
210 | def parse_index_cl_v2(data, inline): |
|
210 | def parse_index_cl_v2(data, inline): | |
211 | # call the C implementation to parse the index data |
|
211 | # call the C implementation to parse the index data | |
212 | index, cache = parsers.parse_index2(data, inline, format=CHANGELOGV2) |
|
212 | index, cache = parsers.parse_index2(data, inline, format=CHANGELOGV2) | |
213 | return index, cache |
|
213 | return index, cache | |
214 |
|
214 | |||
215 |
|
215 | |||
216 | if hasattr(parsers, 'parse_index_devel_nodemap'): |
|
216 | if hasattr(parsers, 'parse_index_devel_nodemap'): | |
217 |
|
217 | |||
218 | def parse_index_v1_nodemap(data, inline): |
|
218 | def parse_index_v1_nodemap(data, inline): | |
219 | index, cache = parsers.parse_index_devel_nodemap(data, inline) |
|
219 | index, cache = parsers.parse_index_devel_nodemap(data, inline) | |
220 | return index, cache |
|
220 | return index, cache | |
221 |
|
221 | |||
222 |
|
222 | |||
223 | else: |
|
223 | else: | |
224 | parse_index_v1_nodemap = None |
|
224 | parse_index_v1_nodemap = None | |
225 |
|
225 | |||
226 |
|
226 | |||
227 | def parse_index_v1_mixed(data, inline): |
|
227 | def parse_index_v1_mixed(data, inline): | |
228 | index, cache = parse_index_v1(data, inline) |
|
228 | index, cache = parse_index_v1(data, inline) | |
229 | return rustrevlog.MixedIndex(index), cache |
|
229 | return rustrevlog.MixedIndex(index), cache | |
230 |
|
230 | |||
231 |
|
231 | |||
232 | # corresponds to uncompressed length of indexformatng (2 gigs, 4-byte |
|
232 | # corresponds to uncompressed length of indexformatng (2 gigs, 4-byte | |
233 | # signed integer) |
|
233 | # signed integer) | |
234 | _maxentrysize = 0x7FFFFFFF |
|
234 | _maxentrysize = 0x7FFFFFFF | |
235 |
|
235 | |||
236 | FILE_TOO_SHORT_MSG = _( |
|
236 | FILE_TOO_SHORT_MSG = _( | |
237 | b'cannot read from revlog %s;' |
|
237 | b'cannot read from revlog %s;' | |
238 | b' expected %d bytes from offset %d, data size is %d' |
|
238 | b' expected %d bytes from offset %d, data size is %d' | |
239 | ) |
|
239 | ) | |
240 |
|
240 | |||
241 | hexdigits = b'0123456789abcdefABCDEF' |
|
241 | hexdigits = b'0123456789abcdefABCDEF' | |
242 |
|
242 | |||
243 |
|
243 | |||
244 | class _Config: |
|
244 | class _Config: | |
245 | def copy(self): |
|
245 | def copy(self): | |
246 | return self.__class__(**self.__dict__) |
|
246 | return self.__class__(**self.__dict__) | |
247 |
|
247 | |||
248 |
|
248 | |||
249 | @attr.s() |
|
249 | @attr.s() | |
250 | class FeatureConfig(_Config): |
|
250 | class FeatureConfig(_Config): | |
251 | """Hold configuration values about the available revlog features""" |
|
251 | """Hold configuration values about the available revlog features""" | |
252 |
|
252 | |||
253 | # the default compression engine |
|
253 | # the default compression engine | |
254 | compression_engine = attr.ib(default=b'zlib') |
|
254 | compression_engine = attr.ib(default=b'zlib') | |
255 | # compression engines options |
|
255 | # compression engines options | |
256 | compression_engine_options = attr.ib(default=attr.Factory(dict)) |
|
256 | compression_engine_options = attr.ib(default=attr.Factory(dict)) | |
257 |
|
257 | |||
258 | # can we use censor on this revlog |
|
258 | # can we use censor on this revlog | |
259 | censorable = attr.ib(default=False) |
|
259 | censorable = attr.ib(default=False) | |
260 | # does this revlog use the "side data" feature |
|
260 | # does this revlog use the "side data" feature | |
261 | has_side_data = attr.ib(default=False) |
|
261 | has_side_data = attr.ib(default=False) | |
262 | # might remove rank configuration once the computation has no impact |
|
262 | # might remove rank configuration once the computation has no impact | |
263 | compute_rank = attr.ib(default=False) |
|
263 | compute_rank = attr.ib(default=False) | |
264 | # parent order is supposed to be semantically irrelevant, so we |
|
264 | # parent order is supposed to be semantically irrelevant, so we | |
265 | # normally resort parents to ensure that the first parent is non-null, |
|
265 | # normally resort parents to ensure that the first parent is non-null, | |
266 | # if there is a non-null parent at all. |
|
266 | # if there is a non-null parent at all. | |
267 | # filelog abuses the parent order as flag to mark some instances of |
|
267 | # filelog abuses the parent order as flag to mark some instances of | |
268 | # meta-encoded files, so allow it to disable this behavior. |
|
268 | # meta-encoded files, so allow it to disable this behavior. | |
269 | canonical_parent_order = attr.ib(default=False) |
|
269 | canonical_parent_order = attr.ib(default=False) | |
270 | # can ellipsis commit be used |
|
270 | # can ellipsis commit be used | |
271 | enable_ellipsis = attr.ib(default=False) |
|
271 | enable_ellipsis = attr.ib(default=False) | |
272 |
|
272 | |||
273 | def copy(self): |
|
273 | def copy(self): | |
274 | new = super().copy() |
|
274 | new = super().copy() | |
275 | new.compression_engine_options = self.compression_engine_options.copy() |
|
275 | new.compression_engine_options = self.compression_engine_options.copy() | |
276 | return new |
|
276 | return new | |
277 |
|
277 | |||
278 |
|
278 | |||
279 | @attr.s() |
|
279 | @attr.s() | |
280 | class DataConfig(_Config): |
|
280 | class DataConfig(_Config): | |
281 | """Hold configuration value about how the revlog data are read""" |
|
281 | """Hold configuration value about how the revlog data are read""" | |
282 |
|
282 | |||
283 | # should we try to open the "pending" version of the revlog |
|
283 | # should we try to open the "pending" version of the revlog | |
284 | try_pending = attr.ib(default=False) |
|
284 | try_pending = attr.ib(default=False) | |
285 | # should we try to open the "splitted" version of the revlog |
|
285 | # should we try to open the "splitted" version of the revlog | |
286 | try_split = attr.ib(default=False) |
|
286 | try_split = attr.ib(default=False) | |
287 | # When True, indexfile should be opened with checkambig=True at writing, |
|
287 | # When True, indexfile should be opened with checkambig=True at writing, | |
288 | # to avoid file stat ambiguity. |
|
288 | # to avoid file stat ambiguity. | |
289 | check_ambig = attr.ib(default=False) |
|
289 | check_ambig = attr.ib(default=False) | |
290 |
|
290 | |||
291 | # If true, use mmap instead of reading to deal with large index |
|
291 | # If true, use mmap instead of reading to deal with large index | |
292 | mmap_large_index = attr.ib(default=False) |
|
292 | mmap_large_index = attr.ib(default=False) | |
293 | # how much data is large |
|
293 | # how much data is large | |
294 | mmap_index_threshold = attr.ib(default=None) |
|
294 | mmap_index_threshold = attr.ib(default=None) | |
295 | # How much data to read and cache into the raw revlog data cache. |
|
295 | # How much data to read and cache into the raw revlog data cache. | |
296 | chunk_cache_size = attr.ib(default=65536) |
|
296 | chunk_cache_size = attr.ib(default=65536) | |
297 |
|
297 | |||
298 | # Allow sparse reading of the revlog data |
|
298 | # Allow sparse reading of the revlog data | |
299 | with_sparse_read = attr.ib(default=False) |
|
299 | with_sparse_read = attr.ib(default=False) | |
300 | # minimal density of a sparse read chunk |
|
300 | # minimal density of a sparse read chunk | |
301 | sr_density_threshold = attr.ib(default=0.50) |
|
301 | sr_density_threshold = attr.ib(default=0.50) | |
302 | # minimal size of data we skip when performing sparse read |
|
302 | # minimal size of data we skip when performing sparse read | |
303 | sr_min_gap_size = attr.ib(default=262144) |
|
303 | sr_min_gap_size = attr.ib(default=262144) | |
304 |
|
304 | |||
305 | # are delta encoded against arbitrary bases. |
|
305 | # are delta encoded against arbitrary bases. | |
306 | generaldelta = attr.ib(default=False) |
|
306 | generaldelta = attr.ib(default=False) | |
307 |
|
307 | |||
308 |
|
308 | |||
309 | @attr.s() |
|
309 | @attr.s() | |
310 | class DeltaConfig(_Config): |
|
310 | class DeltaConfig(_Config): | |
311 | """Hold configuration value about how new delta are computed |
|
311 | """Hold configuration value about how new delta are computed | |
312 |
|
312 | |||
313 | Some attributes are duplicated from DataConfig to help havign each object |
|
313 | Some attributes are duplicated from DataConfig to help havign each object | |
314 | self contained. |
|
314 | self contained. | |
315 | """ |
|
315 | """ | |
316 |
|
316 | |||
317 | # can delta be encoded against arbitrary bases. |
|
317 | # can delta be encoded against arbitrary bases. | |
318 | general_delta = attr.ib(default=False) |
|
318 | general_delta = attr.ib(default=False) | |
319 | # Allow sparse writing of the revlog data |
|
319 | # Allow sparse writing of the revlog data | |
320 | sparse_revlog = attr.ib(default=False) |
|
320 | sparse_revlog = attr.ib(default=False) | |
321 | # maximum length of a delta chain |
|
321 | # maximum length of a delta chain | |
322 | max_chain_len = attr.ib(default=None) |
|
322 | max_chain_len = attr.ib(default=None) | |
323 | # Maximum distance between delta chain base start and end |
|
323 | # Maximum distance between delta chain base start and end | |
324 | max_deltachain_span = attr.ib(default=-1) |
|
324 | max_deltachain_span = attr.ib(default=-1) | |
325 | # If `upper_bound_comp` is not None, this is the expected maximal gain from |
|
325 | # If `upper_bound_comp` is not None, this is the expected maximal gain from | |
326 | # compression for the data content. |
|
326 | # compression for the data content. | |
327 | upper_bound_comp = attr.ib(default=None) |
|
327 | upper_bound_comp = attr.ib(default=None) | |
328 | # Should we try a delta against both parent |
|
328 | # Should we try a delta against both parent | |
329 | delta_both_parents = attr.ib(default=True) |
|
329 | delta_both_parents = attr.ib(default=True) | |
330 | # Test delta base candidate group by chunk of this maximal size. |
|
330 | # Test delta base candidate group by chunk of this maximal size. | |
331 | candidate_group_chunk_size = attr.ib(default=0) |
|
331 | candidate_group_chunk_size = attr.ib(default=0) | |
332 | # Should we display debug information about delta computation |
|
332 | # Should we display debug information about delta computation | |
333 | debug_delta = attr.ib(default=False) |
|
333 | debug_delta = attr.ib(default=False) | |
334 | # trust incoming delta by default |
|
334 | # trust incoming delta by default | |
335 | lazy_delta = attr.ib(default=True) |
|
335 | lazy_delta = attr.ib(default=True) | |
336 | # trust the base of incoming delta by default |
|
336 | # trust the base of incoming delta by default | |
337 | lazy_delta_base = attr.ib(default=False) |
|
337 | lazy_delta_base = attr.ib(default=False) | |
338 |
|
338 | |||
339 |
|
339 | |||
340 | class _InnerRevlog: |
|
340 | class _InnerRevlog: | |
341 | """An inner layer of the revlog object |
|
341 | """An inner layer of the revlog object | |
342 |
|
342 | |||
343 | That layer exist to be able to delegate some operation to Rust, its |
|
343 | That layer exist to be able to delegate some operation to Rust, its | |
344 | boundaries are arbitrary and based on what we can delegate to Rust. |
|
344 | boundaries are arbitrary and based on what we can delegate to Rust. | |
345 | """ |
|
345 | """ | |
346 |
|
346 | |||
347 | def __init__( |
|
347 | def __init__( | |
348 | self, |
|
348 | self, | |
349 | opener, |
|
349 | opener, | |
350 | index, |
|
350 | index, | |
351 | index_file, |
|
351 | index_file, | |
352 | data_file, |
|
352 | data_file, | |
353 | sidedata_file, |
|
353 | sidedata_file, | |
354 | inline, |
|
354 | inline, | |
355 | data_config, |
|
355 | data_config, | |
356 | delta_config, |
|
356 | delta_config, | |
357 | feature_config, |
|
357 | feature_config, | |
358 | chunk_cache, |
|
358 | chunk_cache, | |
359 | default_compression_header, |
|
359 | default_compression_header, | |
360 | ): |
|
360 | ): | |
361 | self.opener = opener |
|
361 | self.opener = opener | |
362 | self.index = index |
|
362 | self.index = index | |
363 |
|
363 | |||
364 | self.__index_file = index_file |
|
364 | self.__index_file = index_file | |
365 | self.data_file = data_file |
|
365 | self.data_file = data_file | |
366 | self.sidedata_file = sidedata_file |
|
366 | self.sidedata_file = sidedata_file | |
367 | self.inline = inline |
|
367 | self.inline = inline | |
368 | self.data_config = data_config |
|
368 | self.data_config = data_config | |
369 | self.delta_config = delta_config |
|
369 | self.delta_config = delta_config | |
370 | self.feature_config = feature_config |
|
370 | self.feature_config = feature_config | |
371 |
|
371 | |||
372 | self._default_compression_header = default_compression_header |
|
372 | self._default_compression_header = default_compression_header | |
373 |
|
373 | |||
374 | # index |
|
374 | # index | |
375 |
|
375 | |||
376 | # 3-tuple of file handles being used for active writing. |
|
376 | # 3-tuple of file handles being used for active writing. | |
377 | self._writinghandles = None |
|
377 | self._writinghandles = None | |
378 |
|
378 | |||
379 | self._segmentfile = randomaccessfile.randomaccessfile( |
|
379 | self._segmentfile = randomaccessfile.randomaccessfile( | |
380 | self.opener, |
|
380 | self.opener, | |
381 | (self.index_file if self.inline else self.data_file), |
|
381 | (self.index_file if self.inline else self.data_file), | |
382 | self.data_config.chunk_cache_size, |
|
382 | self.data_config.chunk_cache_size, | |
383 | chunk_cache, |
|
383 | chunk_cache, | |
384 | ) |
|
384 | ) | |
385 | self._segmentfile_sidedata = randomaccessfile.randomaccessfile( |
|
385 | self._segmentfile_sidedata = randomaccessfile.randomaccessfile( | |
386 | self.opener, |
|
386 | self.opener, | |
387 | self.sidedata_file, |
|
387 | self.sidedata_file, | |
388 | self.data_config.chunk_cache_size, |
|
388 | self.data_config.chunk_cache_size, | |
389 | ) |
|
389 | ) | |
390 |
|
390 | |||
391 | # revlog header -> revlog compressor |
|
391 | # revlog header -> revlog compressor | |
392 | self._decompressors = {} |
|
392 | self._decompressors = {} | |
|
393 | # 3-tuple of (node, rev, text) for a raw revision. | |||
|
394 | self._revisioncache = None | |||
393 |
|
395 | |||
394 | @property |
|
396 | @property | |
395 | def index_file(self): |
|
397 | def index_file(self): | |
396 | return self.__index_file |
|
398 | return self.__index_file | |
397 |
|
399 | |||
398 | @index_file.setter |
|
400 | @index_file.setter | |
399 | def index_file(self, new_index_file): |
|
401 | def index_file(self, new_index_file): | |
400 | self.__index_file = new_index_file |
|
402 | self.__index_file = new_index_file | |
401 | if self.inline: |
|
403 | if self.inline: | |
402 | self._segmentfile.filename = new_index_file |
|
404 | self._segmentfile.filename = new_index_file | |
403 |
|
405 | |||
404 | def __len__(self): |
|
406 | def __len__(self): | |
405 | return len(self.index) |
|
407 | return len(self.index) | |
406 |
|
408 | |||
407 | # Derived from index values. |
|
409 | # Derived from index values. | |
408 |
|
410 | |||
409 | def start(self, rev): |
|
411 | def start(self, rev): | |
410 | """the offset of the data chunk for this revision""" |
|
412 | """the offset of the data chunk for this revision""" | |
411 | return int(self.index[rev][0] >> 16) |
|
413 | return int(self.index[rev][0] >> 16) | |
412 |
|
414 | |||
413 | def length(self, rev): |
|
415 | def length(self, rev): | |
414 | """the length of the data chunk for this revision""" |
|
416 | """the length of the data chunk for this revision""" | |
415 | return self.index[rev][1] |
|
417 | return self.index[rev][1] | |
416 |
|
418 | |||
417 | def end(self, rev): |
|
419 | def end(self, rev): | |
418 | """the end of the data chunk for this revision""" |
|
420 | """the end of the data chunk for this revision""" | |
419 | return self.start(rev) + self.length(rev) |
|
421 | return self.start(rev) + self.length(rev) | |
420 |
|
422 | |||
421 | def deltaparent(self, rev): |
|
423 | def deltaparent(self, rev): | |
422 | """return deltaparent of the given revision""" |
|
424 | """return deltaparent of the given revision""" | |
423 | base = self.index[rev][3] |
|
425 | base = self.index[rev][3] | |
424 | if base == rev: |
|
426 | if base == rev: | |
425 | return nullrev |
|
427 | return nullrev | |
426 | elif self.delta_config.general_delta: |
|
428 | elif self.delta_config.general_delta: | |
427 | return base |
|
429 | return base | |
428 | else: |
|
430 | else: | |
429 | return rev - 1 |
|
431 | return rev - 1 | |
430 |
|
432 | |||
431 | def issnapshot(self, rev): |
|
433 | def issnapshot(self, rev): | |
432 | """tells whether rev is a snapshot""" |
|
434 | """tells whether rev is a snapshot""" | |
433 | if not self.delta_config.sparse_revlog: |
|
435 | if not self.delta_config.sparse_revlog: | |
434 | return self.deltaparent(rev) == nullrev |
|
436 | return self.deltaparent(rev) == nullrev | |
435 | elif hasattr(self.index, 'issnapshot'): |
|
437 | elif hasattr(self.index, 'issnapshot'): | |
436 | # directly assign the method to cache the testing and access |
|
438 | # directly assign the method to cache the testing and access | |
437 | self.issnapshot = self.index.issnapshot |
|
439 | self.issnapshot = self.index.issnapshot | |
438 | return self.issnapshot(rev) |
|
440 | return self.issnapshot(rev) | |
439 | if rev == nullrev: |
|
441 | if rev == nullrev: | |
440 | return True |
|
442 | return True | |
441 | entry = self.index[rev] |
|
443 | entry = self.index[rev] | |
442 | base = entry[3] |
|
444 | base = entry[3] | |
443 | if base == rev: |
|
445 | if base == rev: | |
444 | return True |
|
446 | return True | |
445 | if base == nullrev: |
|
447 | if base == nullrev: | |
446 | return True |
|
448 | return True | |
447 | p1 = entry[5] |
|
449 | p1 = entry[5] | |
448 | while self.length(p1) == 0: |
|
450 | while self.length(p1) == 0: | |
449 | b = self.deltaparent(p1) |
|
451 | b = self.deltaparent(p1) | |
450 | if b == p1: |
|
452 | if b == p1: | |
451 | break |
|
453 | break | |
452 | p1 = b |
|
454 | p1 = b | |
453 | p2 = entry[6] |
|
455 | p2 = entry[6] | |
454 | while self.length(p2) == 0: |
|
456 | while self.length(p2) == 0: | |
455 | b = self.deltaparent(p2) |
|
457 | b = self.deltaparent(p2) | |
456 | if b == p2: |
|
458 | if b == p2: | |
457 | break |
|
459 | break | |
458 | p2 = b |
|
460 | p2 = b | |
459 | if base == p1 or base == p2: |
|
461 | if base == p1 or base == p2: | |
460 | return False |
|
462 | return False | |
461 | return self.issnapshot(base) |
|
463 | return self.issnapshot(base) | |
462 |
|
464 | |||
463 | def _deltachain(self, rev, stoprev=None): |
|
465 | def _deltachain(self, rev, stoprev=None): | |
464 | """Obtain the delta chain for a revision. |
|
466 | """Obtain the delta chain for a revision. | |
465 |
|
467 | |||
466 | ``stoprev`` specifies a revision to stop at. If not specified, we |
|
468 | ``stoprev`` specifies a revision to stop at. If not specified, we | |
467 | stop at the base of the chain. |
|
469 | stop at the base of the chain. | |
468 |
|
470 | |||
469 | Returns a 2-tuple of (chain, stopped) where ``chain`` is a list of |
|
471 | Returns a 2-tuple of (chain, stopped) where ``chain`` is a list of | |
470 | revs in ascending order and ``stopped`` is a bool indicating whether |
|
472 | revs in ascending order and ``stopped`` is a bool indicating whether | |
471 | ``stoprev`` was hit. |
|
473 | ``stoprev`` was hit. | |
472 | """ |
|
474 | """ | |
473 | generaldelta = self.delta_config.general_delta |
|
475 | generaldelta = self.delta_config.general_delta | |
474 | # Try C implementation. |
|
476 | # Try C implementation. | |
475 | try: |
|
477 | try: | |
476 | return self.index.deltachain(rev, stoprev, generaldelta) |
|
478 | return self.index.deltachain(rev, stoprev, generaldelta) | |
477 | except AttributeError: |
|
479 | except AttributeError: | |
478 | pass |
|
480 | pass | |
479 |
|
481 | |||
480 | chain = [] |
|
482 | chain = [] | |
481 |
|
483 | |||
482 | # Alias to prevent attribute lookup in tight loop. |
|
484 | # Alias to prevent attribute lookup in tight loop. | |
483 | index = self.index |
|
485 | index = self.index | |
484 |
|
486 | |||
485 | iterrev = rev |
|
487 | iterrev = rev | |
486 | e = index[iterrev] |
|
488 | e = index[iterrev] | |
487 | while iterrev != e[3] and iterrev != stoprev: |
|
489 | while iterrev != e[3] and iterrev != stoprev: | |
488 | chain.append(iterrev) |
|
490 | chain.append(iterrev) | |
489 | if generaldelta: |
|
491 | if generaldelta: | |
490 | iterrev = e[3] |
|
492 | iterrev = e[3] | |
491 | else: |
|
493 | else: | |
492 | iterrev -= 1 |
|
494 | iterrev -= 1 | |
493 | e = index[iterrev] |
|
495 | e = index[iterrev] | |
494 |
|
496 | |||
495 | if iterrev == stoprev: |
|
497 | if iterrev == stoprev: | |
496 | stopped = True |
|
498 | stopped = True | |
497 | else: |
|
499 | else: | |
498 | chain.append(iterrev) |
|
500 | chain.append(iterrev) | |
499 | stopped = False |
|
501 | stopped = False | |
500 |
|
502 | |||
501 | chain.reverse() |
|
503 | chain.reverse() | |
502 | return chain, stopped |
|
504 | return chain, stopped | |
503 |
|
505 | |||
504 | @util.propertycache |
|
506 | @util.propertycache | |
505 | def _compressor(self): |
|
507 | def _compressor(self): | |
506 | engine = util.compengines[self.feature_config.compression_engine] |
|
508 | engine = util.compengines[self.feature_config.compression_engine] | |
507 | return engine.revlogcompressor( |
|
509 | return engine.revlogcompressor( | |
508 | self.feature_config.compression_engine_options |
|
510 | self.feature_config.compression_engine_options | |
509 | ) |
|
511 | ) | |
510 |
|
512 | |||
511 | @util.propertycache |
|
513 | @util.propertycache | |
512 | def _decompressor(self): |
|
514 | def _decompressor(self): | |
513 | """the default decompressor""" |
|
515 | """the default decompressor""" | |
514 | if self._default_compression_header is None: |
|
516 | if self._default_compression_header is None: | |
515 | return None |
|
517 | return None | |
516 | t = self._default_compression_header |
|
518 | t = self._default_compression_header | |
517 | c = self._get_decompressor(t) |
|
519 | c = self._get_decompressor(t) | |
518 | return c.decompress |
|
520 | return c.decompress | |
519 |
|
521 | |||
520 | def _get_decompressor(self, t): |
|
522 | def _get_decompressor(self, t): | |
521 | try: |
|
523 | try: | |
522 | compressor = self._decompressors[t] |
|
524 | compressor = self._decompressors[t] | |
523 | except KeyError: |
|
525 | except KeyError: | |
524 | try: |
|
526 | try: | |
525 | engine = util.compengines.forrevlogheader(t) |
|
527 | engine = util.compengines.forrevlogheader(t) | |
526 | compressor = engine.revlogcompressor( |
|
528 | compressor = engine.revlogcompressor( | |
527 | self.feature_config.compression_engine_options |
|
529 | self.feature_config.compression_engine_options | |
528 | ) |
|
530 | ) | |
529 | self._decompressors[t] = compressor |
|
531 | self._decompressors[t] = compressor | |
530 | except KeyError: |
|
532 | except KeyError: | |
531 | raise error.RevlogError( |
|
533 | raise error.RevlogError( | |
532 | _(b'unknown compression type %s') % binascii.hexlify(t) |
|
534 | _(b'unknown compression type %s') % binascii.hexlify(t) | |
533 | ) |
|
535 | ) | |
534 | return compressor |
|
536 | return compressor | |
535 |
|
537 | |||
536 | def compress(self, data): |
|
538 | def compress(self, data): | |
537 | """Generate a possibly-compressed representation of data.""" |
|
539 | """Generate a possibly-compressed representation of data.""" | |
538 | if not data: |
|
540 | if not data: | |
539 | return b'', data |
|
541 | return b'', data | |
540 |
|
542 | |||
541 | compressed = self._compressor.compress(data) |
|
543 | compressed = self._compressor.compress(data) | |
542 |
|
544 | |||
543 | if compressed: |
|
545 | if compressed: | |
544 | # The revlog compressor added the header in the returned data. |
|
546 | # The revlog compressor added the header in the returned data. | |
545 | return b'', compressed |
|
547 | return b'', compressed | |
546 |
|
548 | |||
547 | if data[0:1] == b'\0': |
|
549 | if data[0:1] == b'\0': | |
548 | return b'', data |
|
550 | return b'', data | |
549 | return b'u', data |
|
551 | return b'u', data | |
550 |
|
552 | |||
551 | def decompress(self, data): |
|
553 | def decompress(self, data): | |
552 | """Decompress a revlog chunk. |
|
554 | """Decompress a revlog chunk. | |
553 |
|
555 | |||
554 | The chunk is expected to begin with a header identifying the |
|
556 | The chunk is expected to begin with a header identifying the | |
555 | format type so it can be routed to an appropriate decompressor. |
|
557 | format type so it can be routed to an appropriate decompressor. | |
556 | """ |
|
558 | """ | |
557 | if not data: |
|
559 | if not data: | |
558 | return data |
|
560 | return data | |
559 |
|
561 | |||
560 | # Revlogs are read much more frequently than they are written and many |
|
562 | # Revlogs are read much more frequently than they are written and many | |
561 | # chunks only take microseconds to decompress, so performance is |
|
563 | # chunks only take microseconds to decompress, so performance is | |
562 | # important here. |
|
564 | # important here. | |
563 | # |
|
565 | # | |
564 | # We can make a few assumptions about revlogs: |
|
566 | # We can make a few assumptions about revlogs: | |
565 | # |
|
567 | # | |
566 | # 1) the majority of chunks will be compressed (as opposed to inline |
|
568 | # 1) the majority of chunks will be compressed (as opposed to inline | |
567 | # raw data). |
|
569 | # raw data). | |
568 | # 2) decompressing *any* data will likely by at least 10x slower than |
|
570 | # 2) decompressing *any* data will likely by at least 10x slower than | |
569 | # returning raw inline data. |
|
571 | # returning raw inline data. | |
570 | # 3) we want to prioritize common and officially supported compression |
|
572 | # 3) we want to prioritize common and officially supported compression | |
571 | # engines |
|
573 | # engines | |
572 | # |
|
574 | # | |
573 | # It follows that we want to optimize for "decompress compressed data |
|
575 | # It follows that we want to optimize for "decompress compressed data | |
574 | # when encoded with common and officially supported compression engines" |
|
576 | # when encoded with common and officially supported compression engines" | |
575 | # case over "raw data" and "data encoded by less common or non-official |
|
577 | # case over "raw data" and "data encoded by less common or non-official | |
576 | # compression engines." That is why we have the inline lookup first |
|
578 | # compression engines." That is why we have the inline lookup first | |
577 | # followed by the compengines lookup. |
|
579 | # followed by the compengines lookup. | |
578 | # |
|
580 | # | |
579 | # According to `hg perfrevlogchunks`, this is ~0.5% faster for zlib |
|
581 | # According to `hg perfrevlogchunks`, this is ~0.5% faster for zlib | |
580 | # compressed chunks. And this matters for changelog and manifest reads. |
|
582 | # compressed chunks. And this matters for changelog and manifest reads. | |
581 | t = data[0:1] |
|
583 | t = data[0:1] | |
582 |
|
584 | |||
583 | if t == b'x': |
|
585 | if t == b'x': | |
584 | try: |
|
586 | try: | |
585 | return _zlibdecompress(data) |
|
587 | return _zlibdecompress(data) | |
586 | except zlib.error as e: |
|
588 | except zlib.error as e: | |
587 | raise error.RevlogError( |
|
589 | raise error.RevlogError( | |
588 | _(b'revlog decompress error: %s') |
|
590 | _(b'revlog decompress error: %s') | |
589 | % stringutil.forcebytestr(e) |
|
591 | % stringutil.forcebytestr(e) | |
590 | ) |
|
592 | ) | |
591 | # '\0' is more common than 'u' so it goes first. |
|
593 | # '\0' is more common than 'u' so it goes first. | |
592 | elif t == b'\0': |
|
594 | elif t == b'\0': | |
593 | return data |
|
595 | return data | |
594 | elif t == b'u': |
|
596 | elif t == b'u': | |
595 | return util.buffer(data, 1) |
|
597 | return util.buffer(data, 1) | |
596 |
|
598 | |||
597 | compressor = self._get_decompressor(t) |
|
599 | compressor = self._get_decompressor(t) | |
598 |
|
600 | |||
599 | return compressor.decompress(data) |
|
601 | return compressor.decompress(data) | |
600 |
|
602 | |||
601 | @contextlib.contextmanager |
|
603 | @contextlib.contextmanager | |
602 | def reading(self): |
|
604 | def reading(self): | |
603 | """Context manager that keeps data and sidedata files open for reading""" |
|
605 | """Context manager that keeps data and sidedata files open for reading""" | |
604 | if len(self.index) == 0: |
|
606 | if len(self.index) == 0: | |
605 | yield # nothing to be read |
|
607 | yield # nothing to be read | |
606 | else: |
|
608 | else: | |
607 | with self._segmentfile.reading(): |
|
609 | with self._segmentfile.reading(): | |
608 | with self._segmentfile_sidedata.reading(): |
|
610 | with self._segmentfile_sidedata.reading(): | |
609 | yield |
|
611 | yield | |
610 |
|
612 | |||
611 | @property |
|
613 | @property | |
612 | def is_writing(self): |
|
614 | def is_writing(self): | |
613 | """True is a writing context is open""" |
|
615 | """True is a writing context is open""" | |
614 | return self._writinghandles is not None |
|
616 | return self._writinghandles is not None | |
615 |
|
617 | |||
616 | @contextlib.contextmanager |
|
618 | @contextlib.contextmanager | |
617 | def writing(self, transaction, data_end=None, sidedata_end=None): |
|
619 | def writing(self, transaction, data_end=None, sidedata_end=None): | |
618 | """Open the revlog files for writing |
|
620 | """Open the revlog files for writing | |
619 |
|
621 | |||
620 | Add content to a revlog should be done within such context. |
|
622 | Add content to a revlog should be done within such context. | |
621 | """ |
|
623 | """ | |
622 | if self.is_writing: |
|
624 | if self.is_writing: | |
623 | yield |
|
625 | yield | |
624 | else: |
|
626 | else: | |
625 | ifh = dfh = sdfh = None |
|
627 | ifh = dfh = sdfh = None | |
626 | try: |
|
628 | try: | |
627 | r = len(self.index) |
|
629 | r = len(self.index) | |
628 | # opening the data file. |
|
630 | # opening the data file. | |
629 | dsize = 0 |
|
631 | dsize = 0 | |
630 | if r: |
|
632 | if r: | |
631 | dsize = self.end(r - 1) |
|
633 | dsize = self.end(r - 1) | |
632 | dfh = None |
|
634 | dfh = None | |
633 | if not self.inline: |
|
635 | if not self.inline: | |
634 | try: |
|
636 | try: | |
635 | dfh = self.opener(self.data_file, mode=b"r+") |
|
637 | dfh = self.opener(self.data_file, mode=b"r+") | |
636 | if data_end is None: |
|
638 | if data_end is None: | |
637 | dfh.seek(0, os.SEEK_END) |
|
639 | dfh.seek(0, os.SEEK_END) | |
638 | else: |
|
640 | else: | |
639 | dfh.seek(data_end, os.SEEK_SET) |
|
641 | dfh.seek(data_end, os.SEEK_SET) | |
640 | except FileNotFoundError: |
|
642 | except FileNotFoundError: | |
641 | dfh = self.opener(self.data_file, mode=b"w+") |
|
643 | dfh = self.opener(self.data_file, mode=b"w+") | |
642 | transaction.add(self.data_file, dsize) |
|
644 | transaction.add(self.data_file, dsize) | |
643 | if self.sidedata_file is not None: |
|
645 | if self.sidedata_file is not None: | |
644 | assert sidedata_end is not None |
|
646 | assert sidedata_end is not None | |
645 | # revlog-v2 does not inline, help Pytype |
|
647 | # revlog-v2 does not inline, help Pytype | |
646 | assert dfh is not None |
|
648 | assert dfh is not None | |
647 | try: |
|
649 | try: | |
648 | sdfh = self.opener(self.sidedata_file, mode=b"r+") |
|
650 | sdfh = self.opener(self.sidedata_file, mode=b"r+") | |
649 | dfh.seek(sidedata_end, os.SEEK_SET) |
|
651 | dfh.seek(sidedata_end, os.SEEK_SET) | |
650 | except FileNotFoundError: |
|
652 | except FileNotFoundError: | |
651 | sdfh = self.opener(self.sidedata_file, mode=b"w+") |
|
653 | sdfh = self.opener(self.sidedata_file, mode=b"w+") | |
652 | transaction.add(self.sidedata_file, sidedata_end) |
|
654 | transaction.add(self.sidedata_file, sidedata_end) | |
653 |
|
655 | |||
654 | # opening the index file. |
|
656 | # opening the index file. | |
655 | isize = r * self.index.entry_size |
|
657 | isize = r * self.index.entry_size | |
656 | ifh = self.__index_write_fp() |
|
658 | ifh = self.__index_write_fp() | |
657 | if self.inline: |
|
659 | if self.inline: | |
658 | transaction.add(self.index_file, dsize + isize) |
|
660 | transaction.add(self.index_file, dsize + isize) | |
659 | else: |
|
661 | else: | |
660 | transaction.add(self.index_file, isize) |
|
662 | transaction.add(self.index_file, isize) | |
661 | # exposing all file handle for writing. |
|
663 | # exposing all file handle for writing. | |
662 | self._writinghandles = (ifh, dfh, sdfh) |
|
664 | self._writinghandles = (ifh, dfh, sdfh) | |
663 | self._segmentfile.writing_handle = ifh if self.inline else dfh |
|
665 | self._segmentfile.writing_handle = ifh if self.inline else dfh | |
664 | self._segmentfile_sidedata.writing_handle = sdfh |
|
666 | self._segmentfile_sidedata.writing_handle = sdfh | |
665 | yield |
|
667 | yield | |
666 | finally: |
|
668 | finally: | |
667 | self._writinghandles = None |
|
669 | self._writinghandles = None | |
668 | self._segmentfile.writing_handle = None |
|
670 | self._segmentfile.writing_handle = None | |
669 | self._segmentfile_sidedata.writing_handle = None |
|
671 | self._segmentfile_sidedata.writing_handle = None | |
670 | if dfh is not None: |
|
672 | if dfh is not None: | |
671 | dfh.close() |
|
673 | dfh.close() | |
672 | if sdfh is not None: |
|
674 | if sdfh is not None: | |
673 | sdfh.close() |
|
675 | sdfh.close() | |
674 | # closing the index file last to avoid exposing referent to |
|
676 | # closing the index file last to avoid exposing referent to | |
675 | # potential unflushed data content. |
|
677 | # potential unflushed data content. | |
676 | if ifh is not None: |
|
678 | if ifh is not None: | |
677 | ifh.close() |
|
679 | ifh.close() | |
678 |
|
680 | |||
679 | def __index_write_fp(self, index_end=None): |
|
681 | def __index_write_fp(self, index_end=None): | |
680 | """internal method to open the index file for writing |
|
682 | """internal method to open the index file for writing | |
681 |
|
683 | |||
682 | You should not use this directly and use `_writing` instead |
|
684 | You should not use this directly and use `_writing` instead | |
683 | """ |
|
685 | """ | |
684 | try: |
|
686 | try: | |
685 | f = self.opener( |
|
687 | f = self.opener( | |
686 | self.index_file, |
|
688 | self.index_file, | |
687 | mode=b"r+", |
|
689 | mode=b"r+", | |
688 | checkambig=self.data_config.check_ambig, |
|
690 | checkambig=self.data_config.check_ambig, | |
689 | ) |
|
691 | ) | |
690 | if index_end is None: |
|
692 | if index_end is None: | |
691 | f.seek(0, os.SEEK_END) |
|
693 | f.seek(0, os.SEEK_END) | |
692 | else: |
|
694 | else: | |
693 | f.seek(index_end, os.SEEK_SET) |
|
695 | f.seek(index_end, os.SEEK_SET) | |
694 | return f |
|
696 | return f | |
695 | except FileNotFoundError: |
|
697 | except FileNotFoundError: | |
696 | return self.opener( |
|
698 | return self.opener( | |
697 | self.index_file, |
|
699 | self.index_file, | |
698 | mode=b"w+", |
|
700 | mode=b"w+", | |
699 | checkambig=self.data_config.check_ambig, |
|
701 | checkambig=self.data_config.check_ambig, | |
700 | ) |
|
702 | ) | |
701 |
|
703 | |||
702 | def __index_new_fp(self): |
|
704 | def __index_new_fp(self): | |
703 | """internal method to create a new index file for writing |
|
705 | """internal method to create a new index file for writing | |
704 |
|
706 | |||
705 | You should not use this unless you are upgrading from inline revlog |
|
707 | You should not use this unless you are upgrading from inline revlog | |
706 | """ |
|
708 | """ | |
707 | return self.opener( |
|
709 | return self.opener( | |
708 | self.index_file, |
|
710 | self.index_file, | |
709 | mode=b"w", |
|
711 | mode=b"w", | |
710 | checkambig=self.data_config.check_ambig, |
|
712 | checkambig=self.data_config.check_ambig, | |
711 | atomictemp=True, |
|
713 | atomictemp=True, | |
712 | ) |
|
714 | ) | |
713 |
|
715 | |||
714 | def split_inline(self, tr, header, new_index_file_path=None): |
|
716 | def split_inline(self, tr, header, new_index_file_path=None): | |
715 | """split the data of an inline revlog into an index and a data file""" |
|
717 | """split the data of an inline revlog into an index and a data file""" | |
716 | existing_handles = False |
|
718 | existing_handles = False | |
717 | if self._writinghandles is not None: |
|
719 | if self._writinghandles is not None: | |
718 | existing_handles = True |
|
720 | existing_handles = True | |
719 | fp = self._writinghandles[0] |
|
721 | fp = self._writinghandles[0] | |
720 | fp.flush() |
|
722 | fp.flush() | |
721 | fp.close() |
|
723 | fp.close() | |
722 | # We can't use the cached file handle after close(). So prevent |
|
724 | # We can't use the cached file handle after close(). So prevent | |
723 | # its usage. |
|
725 | # its usage. | |
724 | self._writinghandles = None |
|
726 | self._writinghandles = None | |
725 | self._segmentfile.writing_handle = None |
|
727 | self._segmentfile.writing_handle = None | |
726 | # No need to deal with sidedata writing handle as it is only |
|
728 | # No need to deal with sidedata writing handle as it is only | |
727 | # relevant with revlog-v2 which is never inline, not reaching |
|
729 | # relevant with revlog-v2 which is never inline, not reaching | |
728 | # this code |
|
730 | # this code | |
729 |
|
731 | |||
730 | new_dfh = self.opener(self.data_file, mode=b"w+") |
|
732 | new_dfh = self.opener(self.data_file, mode=b"w+") | |
731 | new_dfh.truncate(0) # drop any potentially existing data |
|
733 | new_dfh.truncate(0) # drop any potentially existing data | |
732 | try: |
|
734 | try: | |
733 | with self.reading(): |
|
735 | with self.reading(): | |
734 | for r in range(len(self.index)): |
|
736 | for r in range(len(self.index)): | |
735 | new_dfh.write(self.get_segment_for_revs(r, r)[1]) |
|
737 | new_dfh.write(self.get_segment_for_revs(r, r)[1]) | |
736 | new_dfh.flush() |
|
738 | new_dfh.flush() | |
737 |
|
739 | |||
738 | if new_index_file_path is not None: |
|
740 | if new_index_file_path is not None: | |
739 | self.index_file = new_index_file_path |
|
741 | self.index_file = new_index_file_path | |
740 | with self.__index_new_fp() as fp: |
|
742 | with self.__index_new_fp() as fp: | |
741 | self.inline = False |
|
743 | self.inline = False | |
742 | for i in range(len(self.index)): |
|
744 | for i in range(len(self.index)): | |
743 | e = self.index.entry_binary(i) |
|
745 | e = self.index.entry_binary(i) | |
744 | if i == 0: |
|
746 | if i == 0: | |
745 | packed_header = self.index.pack_header(header) |
|
747 | packed_header = self.index.pack_header(header) | |
746 | e = packed_header + e |
|
748 | e = packed_header + e | |
747 | fp.write(e) |
|
749 | fp.write(e) | |
748 |
|
750 | |||
749 | # If we don't use side-write, the temp file replace the real |
|
751 | # If we don't use side-write, the temp file replace the real | |
750 | # index when we exit the context manager |
|
752 | # index when we exit the context manager | |
751 |
|
753 | |||
752 | self._segmentfile = randomaccessfile.randomaccessfile( |
|
754 | self._segmentfile = randomaccessfile.randomaccessfile( | |
753 | self.opener, |
|
755 | self.opener, | |
754 | self.data_file, |
|
756 | self.data_file, | |
755 | self.data_config.chunk_cache_size, |
|
757 | self.data_config.chunk_cache_size, | |
756 | ) |
|
758 | ) | |
757 |
|
759 | |||
758 | if existing_handles: |
|
760 | if existing_handles: | |
759 | # switched from inline to conventional reopen the index |
|
761 | # switched from inline to conventional reopen the index | |
760 | ifh = self.__index_write_fp() |
|
762 | ifh = self.__index_write_fp() | |
761 | self._writinghandles = (ifh, new_dfh, None) |
|
763 | self._writinghandles = (ifh, new_dfh, None) | |
762 | self._segmentfile.writing_handle = new_dfh |
|
764 | self._segmentfile.writing_handle = new_dfh | |
763 | new_dfh = None |
|
765 | new_dfh = None | |
764 | # No need to deal with sidedata writing handle as it is only |
|
766 | # No need to deal with sidedata writing handle as it is only | |
765 | # relevant with revlog-v2 which is never inline, not reaching |
|
767 | # relevant with revlog-v2 which is never inline, not reaching | |
766 | # this code |
|
768 | # this code | |
767 | finally: |
|
769 | finally: | |
768 | if new_dfh is not None: |
|
770 | if new_dfh is not None: | |
769 | new_dfh.close() |
|
771 | new_dfh.close() | |
770 | return self.index_file |
|
772 | return self.index_file | |
771 |
|
773 | |||
772 | def get_segment_for_revs(self, startrev, endrev): |
|
774 | def get_segment_for_revs(self, startrev, endrev): | |
773 | """Obtain a segment of raw data corresponding to a range of revisions. |
|
775 | """Obtain a segment of raw data corresponding to a range of revisions. | |
774 |
|
776 | |||
775 | Accepts the start and end revisions and an optional already-open |
|
777 | Accepts the start and end revisions and an optional already-open | |
776 | file handle to be used for reading. If the file handle is read, its |
|
778 | file handle to be used for reading. If the file handle is read, its | |
777 | seek position will not be preserved. |
|
779 | seek position will not be preserved. | |
778 |
|
780 | |||
779 | Requests for data may be satisfied by a cache. |
|
781 | Requests for data may be satisfied by a cache. | |
780 |
|
782 | |||
781 | Returns a 2-tuple of (offset, data) for the requested range of |
|
783 | Returns a 2-tuple of (offset, data) for the requested range of | |
782 | revisions. Offset is the integer offset from the beginning of the |
|
784 | revisions. Offset is the integer offset from the beginning of the | |
783 | revlog and data is a str or buffer of the raw byte data. |
|
785 | revlog and data is a str or buffer of the raw byte data. | |
784 |
|
786 | |||
785 | Callers will need to call ``self.start(rev)`` and ``self.length(rev)`` |
|
787 | Callers will need to call ``self.start(rev)`` and ``self.length(rev)`` | |
786 | to determine where each revision's data begins and ends. |
|
788 | to determine where each revision's data begins and ends. | |
787 |
|
789 | |||
788 | API: we should consider making this a private part of the InnerRevlog |
|
790 | API: we should consider making this a private part of the InnerRevlog | |
789 | at some point. |
|
791 | at some point. | |
790 | """ |
|
792 | """ | |
791 | # Inlined self.start(startrev) & self.end(endrev) for perf reasons |
|
793 | # Inlined self.start(startrev) & self.end(endrev) for perf reasons | |
792 | # (functions are expensive). |
|
794 | # (functions are expensive). | |
793 | index = self.index |
|
795 | index = self.index | |
794 | istart = index[startrev] |
|
796 | istart = index[startrev] | |
795 | start = int(istart[0] >> 16) |
|
797 | start = int(istart[0] >> 16) | |
796 | if startrev == endrev: |
|
798 | if startrev == endrev: | |
797 | end = start + istart[1] |
|
799 | end = start + istart[1] | |
798 | else: |
|
800 | else: | |
799 | iend = index[endrev] |
|
801 | iend = index[endrev] | |
800 | end = int(iend[0] >> 16) + iend[1] |
|
802 | end = int(iend[0] >> 16) + iend[1] | |
801 |
|
803 | |||
802 | if self.inline: |
|
804 | if self.inline: | |
803 | start += (startrev + 1) * self.index.entry_size |
|
805 | start += (startrev + 1) * self.index.entry_size | |
804 | end += (endrev + 1) * self.index.entry_size |
|
806 | end += (endrev + 1) * self.index.entry_size | |
805 | length = end - start |
|
807 | length = end - start | |
806 |
|
808 | |||
807 | return start, self._segmentfile.read_chunk(start, length) |
|
809 | return start, self._segmentfile.read_chunk(start, length) | |
808 |
|
810 | |||
809 | def _chunk(self, rev): |
|
811 | def _chunk(self, rev): | |
810 | """Obtain a single decompressed chunk for a revision. |
|
812 | """Obtain a single decompressed chunk for a revision. | |
811 |
|
813 | |||
812 | Accepts an integer revision and an optional already-open file handle |
|
814 | Accepts an integer revision and an optional already-open file handle | |
813 | to be used for reading. If used, the seek position of the file will not |
|
815 | to be used for reading. If used, the seek position of the file will not | |
814 | be preserved. |
|
816 | be preserved. | |
815 |
|
817 | |||
816 | Returns a str holding uncompressed data for the requested revision. |
|
818 | Returns a str holding uncompressed data for the requested revision. | |
817 | """ |
|
819 | """ | |
818 | compression_mode = self.index[rev][10] |
|
820 | compression_mode = self.index[rev][10] | |
819 | data = self.get_segment_for_revs(rev, rev)[1] |
|
821 | data = self.get_segment_for_revs(rev, rev)[1] | |
820 | if compression_mode == COMP_MODE_PLAIN: |
|
822 | if compression_mode == COMP_MODE_PLAIN: | |
821 | return data |
|
823 | return data | |
822 | elif compression_mode == COMP_MODE_DEFAULT: |
|
824 | elif compression_mode == COMP_MODE_DEFAULT: | |
823 | return self._decompressor(data) |
|
825 | return self._decompressor(data) | |
824 | elif compression_mode == COMP_MODE_INLINE: |
|
826 | elif compression_mode == COMP_MODE_INLINE: | |
825 | return self.decompress(data) |
|
827 | return self.decompress(data) | |
826 | else: |
|
828 | else: | |
827 | msg = b'unknown compression mode %d' |
|
829 | msg = b'unknown compression mode %d' | |
828 | msg %= compression_mode |
|
830 | msg %= compression_mode | |
829 | raise error.RevlogError(msg) |
|
831 | raise error.RevlogError(msg) | |
830 |
|
832 | |||
831 | def _chunks(self, revs, targetsize=None): |
|
833 | def _chunks(self, revs, targetsize=None): | |
832 | """Obtain decompressed chunks for the specified revisions. |
|
834 | """Obtain decompressed chunks for the specified revisions. | |
833 |
|
835 | |||
834 | Accepts an iterable of numeric revisions that are assumed to be in |
|
836 | Accepts an iterable of numeric revisions that are assumed to be in | |
835 | ascending order. Also accepts an optional already-open file handle |
|
837 | ascending order. Also accepts an optional already-open file handle | |
836 | to be used for reading. If used, the seek position of the file will |
|
838 | to be used for reading. If used, the seek position of the file will | |
837 | not be preserved. |
|
839 | not be preserved. | |
838 |
|
840 | |||
839 | This function is similar to calling ``self._chunk()`` multiple times, |
|
841 | This function is similar to calling ``self._chunk()`` multiple times, | |
840 | but is faster. |
|
842 | but is faster. | |
841 |
|
843 | |||
842 | Returns a list with decompressed data for each requested revision. |
|
844 | Returns a list with decompressed data for each requested revision. | |
843 | """ |
|
845 | """ | |
844 | if not revs: |
|
846 | if not revs: | |
845 | return [] |
|
847 | return [] | |
846 | start = self.start |
|
848 | start = self.start | |
847 | length = self.length |
|
849 | length = self.length | |
848 | inline = self.inline |
|
850 | inline = self.inline | |
849 | iosize = self.index.entry_size |
|
851 | iosize = self.index.entry_size | |
850 | buffer = util.buffer |
|
852 | buffer = util.buffer | |
851 |
|
853 | |||
852 | l = [] |
|
854 | l = [] | |
853 | ladd = l.append |
|
855 | ladd = l.append | |
854 |
|
856 | |||
855 | if not self.data_config.with_sparse_read: |
|
857 | if not self.data_config.with_sparse_read: | |
856 | slicedchunks = (revs,) |
|
858 | slicedchunks = (revs,) | |
857 | else: |
|
859 | else: | |
858 | slicedchunks = deltautil.slicechunk( |
|
860 | slicedchunks = deltautil.slicechunk( | |
859 | self, |
|
861 | self, | |
860 | revs, |
|
862 | revs, | |
861 | targetsize=targetsize, |
|
863 | targetsize=targetsize, | |
862 | ) |
|
864 | ) | |
863 |
|
865 | |||
864 | for revschunk in slicedchunks: |
|
866 | for revschunk in slicedchunks: | |
865 | firstrev = revschunk[0] |
|
867 | firstrev = revschunk[0] | |
866 | # Skip trailing revisions with empty diff |
|
868 | # Skip trailing revisions with empty diff | |
867 | for lastrev in revschunk[::-1]: |
|
869 | for lastrev in revschunk[::-1]: | |
868 | if length(lastrev) != 0: |
|
870 | if length(lastrev) != 0: | |
869 | break |
|
871 | break | |
870 |
|
872 | |||
871 | try: |
|
873 | try: | |
872 | offset, data = self.get_segment_for_revs(firstrev, lastrev) |
|
874 | offset, data = self.get_segment_for_revs(firstrev, lastrev) | |
873 | except OverflowError: |
|
875 | except OverflowError: | |
874 | # issue4215 - we can't cache a run of chunks greater than |
|
876 | # issue4215 - we can't cache a run of chunks greater than | |
875 | # 2G on Windows |
|
877 | # 2G on Windows | |
876 | return [self._chunk(rev) for rev in revschunk] |
|
878 | return [self._chunk(rev) for rev in revschunk] | |
877 |
|
879 | |||
878 | decomp = self.decompress |
|
880 | decomp = self.decompress | |
879 | # self._decompressor might be None, but will not be used in that case |
|
881 | # self._decompressor might be None, but will not be used in that case | |
880 | def_decomp = self._decompressor |
|
882 | def_decomp = self._decompressor | |
881 | for rev in revschunk: |
|
883 | for rev in revschunk: | |
882 | chunkstart = start(rev) |
|
884 | chunkstart = start(rev) | |
883 | if inline: |
|
885 | if inline: | |
884 | chunkstart += (rev + 1) * iosize |
|
886 | chunkstart += (rev + 1) * iosize | |
885 | chunklength = length(rev) |
|
887 | chunklength = length(rev) | |
886 | comp_mode = self.index[rev][10] |
|
888 | comp_mode = self.index[rev][10] | |
887 | c = buffer(data, chunkstart - offset, chunklength) |
|
889 | c = buffer(data, chunkstart - offset, chunklength) | |
888 | if comp_mode == COMP_MODE_PLAIN: |
|
890 | if comp_mode == COMP_MODE_PLAIN: | |
889 | ladd(c) |
|
891 | ladd(c) | |
890 | elif comp_mode == COMP_MODE_INLINE: |
|
892 | elif comp_mode == COMP_MODE_INLINE: | |
891 | ladd(decomp(c)) |
|
893 | ladd(decomp(c)) | |
892 | elif comp_mode == COMP_MODE_DEFAULT: |
|
894 | elif comp_mode == COMP_MODE_DEFAULT: | |
893 | ladd(def_decomp(c)) |
|
895 | ladd(def_decomp(c)) | |
894 | else: |
|
896 | else: | |
895 | msg = b'unknown compression mode %d' |
|
897 | msg = b'unknown compression mode %d' | |
896 | msg %= comp_mode |
|
898 | msg %= comp_mode | |
897 | raise error.RevlogError(msg) |
|
899 | raise error.RevlogError(msg) | |
898 |
|
900 | |||
899 | return l |
|
901 | return l | |
900 |
|
902 | |||
901 |
|
903 | |||
902 | class revlog: |
|
904 | class revlog: | |
903 | """ |
|
905 | """ | |
904 | the underlying revision storage object |
|
906 | the underlying revision storage object | |
905 |
|
907 | |||
906 | A revlog consists of two parts, an index and the revision data. |
|
908 | A revlog consists of two parts, an index and the revision data. | |
907 |
|
909 | |||
908 | The index is a file with a fixed record size containing |
|
910 | The index is a file with a fixed record size containing | |
909 | information on each revision, including its nodeid (hash), the |
|
911 | information on each revision, including its nodeid (hash), the | |
910 | nodeids of its parents, the position and offset of its data within |
|
912 | nodeids of its parents, the position and offset of its data within | |
911 | the data file, and the revision it's based on. Finally, each entry |
|
913 | the data file, and the revision it's based on. Finally, each entry | |
912 | contains a linkrev entry that can serve as a pointer to external |
|
914 | contains a linkrev entry that can serve as a pointer to external | |
913 | data. |
|
915 | data. | |
914 |
|
916 | |||
915 | The revision data itself is a linear collection of data chunks. |
|
917 | The revision data itself is a linear collection of data chunks. | |
916 | Each chunk represents a revision and is usually represented as a |
|
918 | Each chunk represents a revision and is usually represented as a | |
917 | delta against the previous chunk. To bound lookup time, runs of |
|
919 | delta against the previous chunk. To bound lookup time, runs of | |
918 | deltas are limited to about 2 times the length of the original |
|
920 | deltas are limited to about 2 times the length of the original | |
919 | version data. This makes retrieval of a version proportional to |
|
921 | version data. This makes retrieval of a version proportional to | |
920 | its size, or O(1) relative to the number of revisions. |
|
922 | its size, or O(1) relative to the number of revisions. | |
921 |
|
923 | |||
922 | Both pieces of the revlog are written to in an append-only |
|
924 | Both pieces of the revlog are written to in an append-only | |
923 | fashion, which means we never need to rewrite a file to insert or |
|
925 | fashion, which means we never need to rewrite a file to insert or | |
924 | remove data, and can use some simple techniques to avoid the need |
|
926 | remove data, and can use some simple techniques to avoid the need | |
925 | for locking while reading. |
|
927 | for locking while reading. | |
926 |
|
928 | |||
927 | If checkambig, indexfile is opened with checkambig=True at |
|
929 | If checkambig, indexfile is opened with checkambig=True at | |
928 | writing, to avoid file stat ambiguity. |
|
930 | writing, to avoid file stat ambiguity. | |
929 |
|
931 | |||
930 | If mmaplargeindex is True, and an mmapindexthreshold is set, the |
|
932 | If mmaplargeindex is True, and an mmapindexthreshold is set, the | |
931 | index will be mmapped rather than read if it is larger than the |
|
933 | index will be mmapped rather than read if it is larger than the | |
932 | configured threshold. |
|
934 | configured threshold. | |
933 |
|
935 | |||
934 | If censorable is True, the revlog can have censored revisions. |
|
936 | If censorable is True, the revlog can have censored revisions. | |
935 |
|
937 | |||
936 | If `upperboundcomp` is not None, this is the expected maximal gain from |
|
938 | If `upperboundcomp` is not None, this is the expected maximal gain from | |
937 | compression for the data content. |
|
939 | compression for the data content. | |
938 |
|
940 | |||
939 | `concurrencychecker` is an optional function that receives 3 arguments: a |
|
941 | `concurrencychecker` is an optional function that receives 3 arguments: a | |
940 | file handle, a filename, and an expected position. It should check whether |
|
942 | file handle, a filename, and an expected position. It should check whether | |
941 | the current position in the file handle is valid, and log/warn/fail (by |
|
943 | the current position in the file handle is valid, and log/warn/fail (by | |
942 | raising). |
|
944 | raising). | |
943 |
|
945 | |||
944 | See mercurial/revlogutils/contants.py for details about the content of an |
|
946 | See mercurial/revlogutils/contants.py for details about the content of an | |
945 | index entry. |
|
947 | index entry. | |
946 | """ |
|
948 | """ | |
947 |
|
949 | |||
948 | _flagserrorclass = error.RevlogError |
|
950 | _flagserrorclass = error.RevlogError | |
949 |
|
951 | |||
950 | @staticmethod |
|
952 | @staticmethod | |
951 | def is_inline_index(header_bytes): |
|
953 | def is_inline_index(header_bytes): | |
952 | """Determine if a revlog is inline from the initial bytes of the index""" |
|
954 | """Determine if a revlog is inline from the initial bytes of the index""" | |
953 | header = INDEX_HEADER.unpack(header_bytes)[0] |
|
955 | header = INDEX_HEADER.unpack(header_bytes)[0] | |
954 |
|
956 | |||
955 | _format_flags = header & ~0xFFFF |
|
957 | _format_flags = header & ~0xFFFF | |
956 | _format_version = header & 0xFFFF |
|
958 | _format_version = header & 0xFFFF | |
957 |
|
959 | |||
958 | features = FEATURES_BY_VERSION[_format_version] |
|
960 | features = FEATURES_BY_VERSION[_format_version] | |
959 | return features[b'inline'](_format_flags) |
|
961 | return features[b'inline'](_format_flags) | |
960 |
|
962 | |||
961 | def __init__( |
|
963 | def __init__( | |
962 | self, |
|
964 | self, | |
963 | opener, |
|
965 | opener, | |
964 | target, |
|
966 | target, | |
965 | radix, |
|
967 | radix, | |
966 | postfix=None, # only exist for `tmpcensored` now |
|
968 | postfix=None, # only exist for `tmpcensored` now | |
967 | checkambig=False, |
|
969 | checkambig=False, | |
968 | mmaplargeindex=False, |
|
970 | mmaplargeindex=False, | |
969 | censorable=False, |
|
971 | censorable=False, | |
970 | upperboundcomp=None, |
|
972 | upperboundcomp=None, | |
971 | persistentnodemap=False, |
|
973 | persistentnodemap=False, | |
972 | concurrencychecker=None, |
|
974 | concurrencychecker=None, | |
973 | trypending=False, |
|
975 | trypending=False, | |
974 | try_split=False, |
|
976 | try_split=False, | |
975 | canonical_parent_order=True, |
|
977 | canonical_parent_order=True, | |
976 | ): |
|
978 | ): | |
977 | """ |
|
979 | """ | |
978 | create a revlog object |
|
980 | create a revlog object | |
979 |
|
981 | |||
980 | opener is a function that abstracts the file opening operation |
|
982 | opener is a function that abstracts the file opening operation | |
981 | and can be used to implement COW semantics or the like. |
|
983 | and can be used to implement COW semantics or the like. | |
982 |
|
984 | |||
983 | `target`: a (KIND, ID) tuple that identify the content stored in |
|
985 | `target`: a (KIND, ID) tuple that identify the content stored in | |
984 | this revlog. It help the rest of the code to understand what the revlog |
|
986 | this revlog. It help the rest of the code to understand what the revlog | |
985 | is about without having to resort to heuristic and index filename |
|
987 | is about without having to resort to heuristic and index filename | |
986 | analysis. Note: that this must be reliably be set by normal code, but |
|
988 | analysis. Note: that this must be reliably be set by normal code, but | |
987 | that test, debug, or performance measurement code might not set this to |
|
989 | that test, debug, or performance measurement code might not set this to | |
988 | accurate value. |
|
990 | accurate value. | |
989 | """ |
|
991 | """ | |
990 |
|
992 | |||
991 | self.radix = radix |
|
993 | self.radix = radix | |
992 |
|
994 | |||
993 | self._docket_file = None |
|
995 | self._docket_file = None | |
994 | self._indexfile = None |
|
996 | self._indexfile = None | |
995 | self._datafile = None |
|
997 | self._datafile = None | |
996 | self._sidedatafile = None |
|
998 | self._sidedatafile = None | |
997 | self._nodemap_file = None |
|
999 | self._nodemap_file = None | |
998 | self.postfix = postfix |
|
1000 | self.postfix = postfix | |
999 | self._trypending = trypending |
|
1001 | self._trypending = trypending | |
1000 | self._try_split = try_split |
|
1002 | self._try_split = try_split | |
1001 | self.opener = opener |
|
1003 | self.opener = opener | |
1002 | if persistentnodemap: |
|
1004 | if persistentnodemap: | |
1003 | self._nodemap_file = nodemaputil.get_nodemap_file(self) |
|
1005 | self._nodemap_file = nodemaputil.get_nodemap_file(self) | |
1004 |
|
1006 | |||
1005 | assert target[0] in ALL_KINDS |
|
1007 | assert target[0] in ALL_KINDS | |
1006 | assert len(target) == 2 |
|
1008 | assert len(target) == 2 | |
1007 | self.target = target |
|
1009 | self.target = target | |
1008 | if b'feature-config' in self.opener.options: |
|
1010 | if b'feature-config' in self.opener.options: | |
1009 | self.feature_config = self.opener.options[b'feature-config'].copy() |
|
1011 | self.feature_config = self.opener.options[b'feature-config'].copy() | |
1010 | else: |
|
1012 | else: | |
1011 | self.feature_config = FeatureConfig() |
|
1013 | self.feature_config = FeatureConfig() | |
1012 | self.feature_config.censorable = censorable |
|
1014 | self.feature_config.censorable = censorable | |
1013 | self.feature_config.canonical_parent_order = canonical_parent_order |
|
1015 | self.feature_config.canonical_parent_order = canonical_parent_order | |
1014 | if b'data-config' in self.opener.options: |
|
1016 | if b'data-config' in self.opener.options: | |
1015 | self.data_config = self.opener.options[b'data-config'].copy() |
|
1017 | self.data_config = self.opener.options[b'data-config'].copy() | |
1016 | else: |
|
1018 | else: | |
1017 | self.data_config = DataConfig() |
|
1019 | self.data_config = DataConfig() | |
1018 | self.data_config.check_ambig = checkambig |
|
1020 | self.data_config.check_ambig = checkambig | |
1019 | self.data_config.mmap_large_index = mmaplargeindex |
|
1021 | self.data_config.mmap_large_index = mmaplargeindex | |
1020 | if b'delta-config' in self.opener.options: |
|
1022 | if b'delta-config' in self.opener.options: | |
1021 | self.delta_config = self.opener.options[b'delta-config'].copy() |
|
1023 | self.delta_config = self.opener.options[b'delta-config'].copy() | |
1022 | else: |
|
1024 | else: | |
1023 | self.delta_config = DeltaConfig() |
|
1025 | self.delta_config = DeltaConfig() | |
1024 | self.delta_config.upper_bound_comp = upperboundcomp |
|
1026 | self.delta_config.upper_bound_comp = upperboundcomp | |
1025 |
|
1027 | |||
1026 | # 3-tuple of (node, rev, text) for a raw revision. |
|
|||
1027 | self._revisioncache = None |
|
|||
1028 | # Maps rev to chain base rev. |
|
1028 | # Maps rev to chain base rev. | |
1029 | self._chainbasecache = util.lrucachedict(100) |
|
1029 | self._chainbasecache = util.lrucachedict(100) | |
1030 |
|
1030 | |||
1031 | self.index = None |
|
1031 | self.index = None | |
1032 | self._docket = None |
|
1032 | self._docket = None | |
1033 | self._nodemap_docket = None |
|
1033 | self._nodemap_docket = None | |
1034 | # Mapping of partial identifiers to full nodes. |
|
1034 | # Mapping of partial identifiers to full nodes. | |
1035 | self._pcache = {} |
|
1035 | self._pcache = {} | |
1036 |
|
1036 | |||
1037 | # other optionnals features |
|
1037 | # other optionnals features | |
1038 |
|
1038 | |||
1039 | # Make copy of flag processors so each revlog instance can support |
|
1039 | # Make copy of flag processors so each revlog instance can support | |
1040 | # custom flags. |
|
1040 | # custom flags. | |
1041 | self._flagprocessors = dict(flagutil.flagprocessors) |
|
1041 | self._flagprocessors = dict(flagutil.flagprocessors) | |
1042 | # prevent nesting of addgroup |
|
1042 | # prevent nesting of addgroup | |
1043 | self._adding_group = None |
|
1043 | self._adding_group = None | |
1044 |
|
1044 | |||
1045 | chunk_cache = self._loadindex() |
|
1045 | chunk_cache = self._loadindex() | |
1046 | self._load_inner(chunk_cache) |
|
1046 | self._load_inner(chunk_cache) | |
1047 | self._concurrencychecker = concurrencychecker |
|
1047 | self._concurrencychecker = concurrencychecker | |
1048 |
|
1048 | |||
1049 | @property |
|
1049 | @property | |
1050 | def _generaldelta(self): |
|
1050 | def _generaldelta(self): | |
1051 | """temporary compatibility proxy""" |
|
1051 | """temporary compatibility proxy""" | |
1052 | util.nouideprecwarn( |
|
1052 | util.nouideprecwarn( | |
1053 | b"use revlog.delta_config.general_delta", b"6.6", stacklevel=2 |
|
1053 | b"use revlog.delta_config.general_delta", b"6.6", stacklevel=2 | |
1054 | ) |
|
1054 | ) | |
1055 | return self.delta_config.general_delta |
|
1055 | return self.delta_config.general_delta | |
1056 |
|
1056 | |||
1057 | @property |
|
1057 | @property | |
1058 | def _checkambig(self): |
|
1058 | def _checkambig(self): | |
1059 | """temporary compatibility proxy""" |
|
1059 | """temporary compatibility proxy""" | |
1060 | util.nouideprecwarn( |
|
1060 | util.nouideprecwarn( | |
1061 | b"use revlog.data_config.checkambig", b"6.6", stacklevel=2 |
|
1061 | b"use revlog.data_config.checkambig", b"6.6", stacklevel=2 | |
1062 | ) |
|
1062 | ) | |
1063 | return self.data_config.check_ambig |
|
1063 | return self.data_config.check_ambig | |
1064 |
|
1064 | |||
1065 | @property |
|
1065 | @property | |
1066 | def _mmaplargeindex(self): |
|
1066 | def _mmaplargeindex(self): | |
1067 | """temporary compatibility proxy""" |
|
1067 | """temporary compatibility proxy""" | |
1068 | util.nouideprecwarn( |
|
1068 | util.nouideprecwarn( | |
1069 | b"use revlog.data_config.mmap_large_index", b"6.6", stacklevel=2 |
|
1069 | b"use revlog.data_config.mmap_large_index", b"6.6", stacklevel=2 | |
1070 | ) |
|
1070 | ) | |
1071 | return self.data_config.mmap_large_index |
|
1071 | return self.data_config.mmap_large_index | |
1072 |
|
1072 | |||
1073 | @property |
|
1073 | @property | |
1074 | def _censorable(self): |
|
1074 | def _censorable(self): | |
1075 | """temporary compatibility proxy""" |
|
1075 | """temporary compatibility proxy""" | |
1076 | util.nouideprecwarn( |
|
1076 | util.nouideprecwarn( | |
1077 | b"use revlog.feature_config.censorable", b"6.6", stacklevel=2 |
|
1077 | b"use revlog.feature_config.censorable", b"6.6", stacklevel=2 | |
1078 | ) |
|
1078 | ) | |
1079 | return self.feature_config.censorable |
|
1079 | return self.feature_config.censorable | |
1080 |
|
1080 | |||
1081 | @property |
|
1081 | @property | |
1082 | def _chunkcachesize(self): |
|
1082 | def _chunkcachesize(self): | |
1083 | """temporary compatibility proxy""" |
|
1083 | """temporary compatibility proxy""" | |
1084 | util.nouideprecwarn( |
|
1084 | util.nouideprecwarn( | |
1085 | b"use revlog.data_config.chunk_cache_size", b"6.6", stacklevel=2 |
|
1085 | b"use revlog.data_config.chunk_cache_size", b"6.6", stacklevel=2 | |
1086 | ) |
|
1086 | ) | |
1087 | return self.data_config.chunk_cache_size |
|
1087 | return self.data_config.chunk_cache_size | |
1088 |
|
1088 | |||
1089 | @property |
|
1089 | @property | |
1090 | def _maxchainlen(self): |
|
1090 | def _maxchainlen(self): | |
1091 | """temporary compatibility proxy""" |
|
1091 | """temporary compatibility proxy""" | |
1092 | util.nouideprecwarn( |
|
1092 | util.nouideprecwarn( | |
1093 | b"use revlog.delta_config.max_chain_len", b"6.6", stacklevel=2 |
|
1093 | b"use revlog.delta_config.max_chain_len", b"6.6", stacklevel=2 | |
1094 | ) |
|
1094 | ) | |
1095 | return self.delta_config.max_chain_len |
|
1095 | return self.delta_config.max_chain_len | |
1096 |
|
1096 | |||
1097 | @property |
|
1097 | @property | |
1098 | def _deltabothparents(self): |
|
1098 | def _deltabothparents(self): | |
1099 | """temporary compatibility proxy""" |
|
1099 | """temporary compatibility proxy""" | |
1100 | util.nouideprecwarn( |
|
1100 | util.nouideprecwarn( | |
1101 | b"use revlog.delta_config.delta_both_parents", b"6.6", stacklevel=2 |
|
1101 | b"use revlog.delta_config.delta_both_parents", b"6.6", stacklevel=2 | |
1102 | ) |
|
1102 | ) | |
1103 | return self.delta_config.delta_both_parents |
|
1103 | return self.delta_config.delta_both_parents | |
1104 |
|
1104 | |||
1105 | @property |
|
1105 | @property | |
1106 | def _candidate_group_chunk_size(self): |
|
1106 | def _candidate_group_chunk_size(self): | |
1107 | """temporary compatibility proxy""" |
|
1107 | """temporary compatibility proxy""" | |
1108 | util.nouideprecwarn( |
|
1108 | util.nouideprecwarn( | |
1109 | b"use revlog.delta_config.candidate_group_chunk_size", |
|
1109 | b"use revlog.delta_config.candidate_group_chunk_size", | |
1110 | b"6.6", |
|
1110 | b"6.6", | |
1111 | stacklevel=2, |
|
1111 | stacklevel=2, | |
1112 | ) |
|
1112 | ) | |
1113 | return self.delta_config.candidate_group_chunk_size |
|
1113 | return self.delta_config.candidate_group_chunk_size | |
1114 |
|
1114 | |||
1115 | @property |
|
1115 | @property | |
1116 | def _debug_delta(self): |
|
1116 | def _debug_delta(self): | |
1117 | """temporary compatibility proxy""" |
|
1117 | """temporary compatibility proxy""" | |
1118 | util.nouideprecwarn( |
|
1118 | util.nouideprecwarn( | |
1119 | b"use revlog.delta_config.debug_delta", b"6.6", stacklevel=2 |
|
1119 | b"use revlog.delta_config.debug_delta", b"6.6", stacklevel=2 | |
1120 | ) |
|
1120 | ) | |
1121 | return self.delta_config.debug_delta |
|
1121 | return self.delta_config.debug_delta | |
1122 |
|
1122 | |||
1123 | @property |
|
1123 | @property | |
1124 | def _compengine(self): |
|
1124 | def _compengine(self): | |
1125 | """temporary compatibility proxy""" |
|
1125 | """temporary compatibility proxy""" | |
1126 | util.nouideprecwarn( |
|
1126 | util.nouideprecwarn( | |
1127 | b"use revlog.feature_config.compression_engine", |
|
1127 | b"use revlog.feature_config.compression_engine", | |
1128 | b"6.6", |
|
1128 | b"6.6", | |
1129 | stacklevel=2, |
|
1129 | stacklevel=2, | |
1130 | ) |
|
1130 | ) | |
1131 | return self.feature_config.compression_engine |
|
1131 | return self.feature_config.compression_engine | |
1132 |
|
1132 | |||
1133 | @property |
|
1133 | @property | |
1134 | def upperboundcomp(self): |
|
1134 | def upperboundcomp(self): | |
1135 | """temporary compatibility proxy""" |
|
1135 | """temporary compatibility proxy""" | |
1136 | util.nouideprecwarn( |
|
1136 | util.nouideprecwarn( | |
1137 | b"use revlog.delta_config.upper_bound_comp", |
|
1137 | b"use revlog.delta_config.upper_bound_comp", | |
1138 | b"6.6", |
|
1138 | b"6.6", | |
1139 | stacklevel=2, |
|
1139 | stacklevel=2, | |
1140 | ) |
|
1140 | ) | |
1141 | return self.delta_config.upper_bound_comp |
|
1141 | return self.delta_config.upper_bound_comp | |
1142 |
|
1142 | |||
1143 | @property |
|
1143 | @property | |
1144 | def _compengineopts(self): |
|
1144 | def _compengineopts(self): | |
1145 | """temporary compatibility proxy""" |
|
1145 | """temporary compatibility proxy""" | |
1146 | util.nouideprecwarn( |
|
1146 | util.nouideprecwarn( | |
1147 | b"use revlog.feature_config.compression_engine_options", |
|
1147 | b"use revlog.feature_config.compression_engine_options", | |
1148 | b"6.6", |
|
1148 | b"6.6", | |
1149 | stacklevel=2, |
|
1149 | stacklevel=2, | |
1150 | ) |
|
1150 | ) | |
1151 | return self.feature_config.compression_engine_options |
|
1151 | return self.feature_config.compression_engine_options | |
1152 |
|
1152 | |||
1153 | @property |
|
1153 | @property | |
1154 | def _maxdeltachainspan(self): |
|
1154 | def _maxdeltachainspan(self): | |
1155 | """temporary compatibility proxy""" |
|
1155 | """temporary compatibility proxy""" | |
1156 | util.nouideprecwarn( |
|
1156 | util.nouideprecwarn( | |
1157 | b"use revlog.delta_config.max_deltachain_span", b"6.6", stacklevel=2 |
|
1157 | b"use revlog.delta_config.max_deltachain_span", b"6.6", stacklevel=2 | |
1158 | ) |
|
1158 | ) | |
1159 | return self.delta_config.max_deltachain_span |
|
1159 | return self.delta_config.max_deltachain_span | |
1160 |
|
1160 | |||
1161 | @property |
|
1161 | @property | |
1162 | def _withsparseread(self): |
|
1162 | def _withsparseread(self): | |
1163 | """temporary compatibility proxy""" |
|
1163 | """temporary compatibility proxy""" | |
1164 | util.nouideprecwarn( |
|
1164 | util.nouideprecwarn( | |
1165 | b"use revlog.data_config.with_sparse_read", b"6.6", stacklevel=2 |
|
1165 | b"use revlog.data_config.with_sparse_read", b"6.6", stacklevel=2 | |
1166 | ) |
|
1166 | ) | |
1167 | return self.data_config.with_sparse_read |
|
1167 | return self.data_config.with_sparse_read | |
1168 |
|
1168 | |||
1169 | @property |
|
1169 | @property | |
1170 | def _sparserevlog(self): |
|
1170 | def _sparserevlog(self): | |
1171 | """temporary compatibility proxy""" |
|
1171 | """temporary compatibility proxy""" | |
1172 | util.nouideprecwarn( |
|
1172 | util.nouideprecwarn( | |
1173 | b"use revlog.delta_config.sparse_revlog", b"6.6", stacklevel=2 |
|
1173 | b"use revlog.delta_config.sparse_revlog", b"6.6", stacklevel=2 | |
1174 | ) |
|
1174 | ) | |
1175 | return self.delta_config.sparse_revlog |
|
1175 | return self.delta_config.sparse_revlog | |
1176 |
|
1176 | |||
1177 | @property |
|
1177 | @property | |
1178 | def hassidedata(self): |
|
1178 | def hassidedata(self): | |
1179 | """temporary compatibility proxy""" |
|
1179 | """temporary compatibility proxy""" | |
1180 | util.nouideprecwarn( |
|
1180 | util.nouideprecwarn( | |
1181 | b"use revlog.feature_config.has_side_data", b"6.6", stacklevel=2 |
|
1181 | b"use revlog.feature_config.has_side_data", b"6.6", stacklevel=2 | |
1182 | ) |
|
1182 | ) | |
1183 | return self.feature_config.has_side_data |
|
1183 | return self.feature_config.has_side_data | |
1184 |
|
1184 | |||
1185 | @property |
|
1185 | @property | |
1186 | def _srdensitythreshold(self): |
|
1186 | def _srdensitythreshold(self): | |
1187 | """temporary compatibility proxy""" |
|
1187 | """temporary compatibility proxy""" | |
1188 | util.nouideprecwarn( |
|
1188 | util.nouideprecwarn( | |
1189 | b"use revlog.data_config.sr_density_threshold", |
|
1189 | b"use revlog.data_config.sr_density_threshold", | |
1190 | b"6.6", |
|
1190 | b"6.6", | |
1191 | stacklevel=2, |
|
1191 | stacklevel=2, | |
1192 | ) |
|
1192 | ) | |
1193 | return self.data_config.sr_density_threshold |
|
1193 | return self.data_config.sr_density_threshold | |
1194 |
|
1194 | |||
1195 | @property |
|
1195 | @property | |
1196 | def _srmingapsize(self): |
|
1196 | def _srmingapsize(self): | |
1197 | """temporary compatibility proxy""" |
|
1197 | """temporary compatibility proxy""" | |
1198 | util.nouideprecwarn( |
|
1198 | util.nouideprecwarn( | |
1199 | b"use revlog.data_config.sr_min_gap_size", b"6.6", stacklevel=2 |
|
1199 | b"use revlog.data_config.sr_min_gap_size", b"6.6", stacklevel=2 | |
1200 | ) |
|
1200 | ) | |
1201 | return self.data_config.sr_min_gap_size |
|
1201 | return self.data_config.sr_min_gap_size | |
1202 |
|
1202 | |||
1203 | @property |
|
1203 | @property | |
1204 | def _compute_rank(self): |
|
1204 | def _compute_rank(self): | |
1205 | """temporary compatibility proxy""" |
|
1205 | """temporary compatibility proxy""" | |
1206 | util.nouideprecwarn( |
|
1206 | util.nouideprecwarn( | |
1207 | b"use revlog.feature_config.compute_rank", b"6.6", stacklevel=2 |
|
1207 | b"use revlog.feature_config.compute_rank", b"6.6", stacklevel=2 | |
1208 | ) |
|
1208 | ) | |
1209 | return self.feature_config.compute_rank |
|
1209 | return self.feature_config.compute_rank | |
1210 |
|
1210 | |||
1211 | @property |
|
1211 | @property | |
1212 | def canonical_parent_order(self): |
|
1212 | def canonical_parent_order(self): | |
1213 | """temporary compatibility proxy""" |
|
1213 | """temporary compatibility proxy""" | |
1214 | util.nouideprecwarn( |
|
1214 | util.nouideprecwarn( | |
1215 | b"use revlog.feature_config.canonical_parent_order", |
|
1215 | b"use revlog.feature_config.canonical_parent_order", | |
1216 | b"6.6", |
|
1216 | b"6.6", | |
1217 | stacklevel=2, |
|
1217 | stacklevel=2, | |
1218 | ) |
|
1218 | ) | |
1219 | return self.feature_config.canonical_parent_order |
|
1219 | return self.feature_config.canonical_parent_order | |
1220 |
|
1220 | |||
1221 | @property |
|
1221 | @property | |
1222 | def _lazydelta(self): |
|
1222 | def _lazydelta(self): | |
1223 | """temporary compatibility proxy""" |
|
1223 | """temporary compatibility proxy""" | |
1224 | util.nouideprecwarn( |
|
1224 | util.nouideprecwarn( | |
1225 | b"use revlog.delta_config.lazy_delta", b"6.6", stacklevel=2 |
|
1225 | b"use revlog.delta_config.lazy_delta", b"6.6", stacklevel=2 | |
1226 | ) |
|
1226 | ) | |
1227 | return self.delta_config.lazy_delta |
|
1227 | return self.delta_config.lazy_delta | |
1228 |
|
1228 | |||
1229 | @property |
|
1229 | @property | |
1230 | def _lazydeltabase(self): |
|
1230 | def _lazydeltabase(self): | |
1231 | """temporary compatibility proxy""" |
|
1231 | """temporary compatibility proxy""" | |
1232 | util.nouideprecwarn( |
|
1232 | util.nouideprecwarn( | |
1233 | b"use revlog.delta_config.lazy_delta_base", b"6.6", stacklevel=2 |
|
1233 | b"use revlog.delta_config.lazy_delta_base", b"6.6", stacklevel=2 | |
1234 | ) |
|
1234 | ) | |
1235 | return self.delta_config.lazy_delta_base |
|
1235 | return self.delta_config.lazy_delta_base | |
1236 |
|
1236 | |||
1237 | def _init_opts(self): |
|
1237 | def _init_opts(self): | |
1238 | """process options (from above/config) to setup associated default revlog mode |
|
1238 | """process options (from above/config) to setup associated default revlog mode | |
1239 |
|
1239 | |||
1240 | These values might be affected when actually reading on disk information. |
|
1240 | These values might be affected when actually reading on disk information. | |
1241 |
|
1241 | |||
1242 | The relevant values are returned for use in _loadindex(). |
|
1242 | The relevant values are returned for use in _loadindex(). | |
1243 |
|
1243 | |||
1244 | * newversionflags: |
|
1244 | * newversionflags: | |
1245 | version header to use if we need to create a new revlog |
|
1245 | version header to use if we need to create a new revlog | |
1246 |
|
1246 | |||
1247 | * mmapindexthreshold: |
|
1247 | * mmapindexthreshold: | |
1248 | minimal index size for start to use mmap |
|
1248 | minimal index size for start to use mmap | |
1249 |
|
1249 | |||
1250 | * force_nodemap: |
|
1250 | * force_nodemap: | |
1251 | force the usage of a "development" version of the nodemap code |
|
1251 | force the usage of a "development" version of the nodemap code | |
1252 | """ |
|
1252 | """ | |
1253 | opts = self.opener.options |
|
1253 | opts = self.opener.options | |
1254 |
|
1254 | |||
1255 | if b'changelogv2' in opts and self.revlog_kind == KIND_CHANGELOG: |
|
1255 | if b'changelogv2' in opts and self.revlog_kind == KIND_CHANGELOG: | |
1256 | new_header = CHANGELOGV2 |
|
1256 | new_header = CHANGELOGV2 | |
1257 | compute_rank = opts.get(b'changelogv2.compute-rank', True) |
|
1257 | compute_rank = opts.get(b'changelogv2.compute-rank', True) | |
1258 | self.feature_config.compute_rank = compute_rank |
|
1258 | self.feature_config.compute_rank = compute_rank | |
1259 | elif b'revlogv2' in opts: |
|
1259 | elif b'revlogv2' in opts: | |
1260 | new_header = REVLOGV2 |
|
1260 | new_header = REVLOGV2 | |
1261 | elif b'revlogv1' in opts: |
|
1261 | elif b'revlogv1' in opts: | |
1262 | new_header = REVLOGV1 | FLAG_INLINE_DATA |
|
1262 | new_header = REVLOGV1 | FLAG_INLINE_DATA | |
1263 | if b'generaldelta' in opts: |
|
1263 | if b'generaldelta' in opts: | |
1264 | new_header |= FLAG_GENERALDELTA |
|
1264 | new_header |= FLAG_GENERALDELTA | |
1265 | elif b'revlogv0' in self.opener.options: |
|
1265 | elif b'revlogv0' in self.opener.options: | |
1266 | new_header = REVLOGV0 |
|
1266 | new_header = REVLOGV0 | |
1267 | else: |
|
1267 | else: | |
1268 | new_header = REVLOG_DEFAULT_VERSION |
|
1268 | new_header = REVLOG_DEFAULT_VERSION | |
1269 |
|
1269 | |||
1270 | mmapindexthreshold = None |
|
1270 | mmapindexthreshold = None | |
1271 | if self.data_config.mmap_large_index: |
|
1271 | if self.data_config.mmap_large_index: | |
1272 | mmapindexthreshold = self.data_config.mmap_index_threshold |
|
1272 | mmapindexthreshold = self.data_config.mmap_index_threshold | |
1273 | if self.feature_config.enable_ellipsis: |
|
1273 | if self.feature_config.enable_ellipsis: | |
1274 | self._flagprocessors[REVIDX_ELLIPSIS] = ellipsisprocessor |
|
1274 | self._flagprocessors[REVIDX_ELLIPSIS] = ellipsisprocessor | |
1275 |
|
1275 | |||
1276 | # revlog v0 doesn't have flag processors |
|
1276 | # revlog v0 doesn't have flag processors | |
1277 | for flag, processor in opts.get(b'flagprocessors', {}).items(): |
|
1277 | for flag, processor in opts.get(b'flagprocessors', {}).items(): | |
1278 | flagutil.insertflagprocessor(flag, processor, self._flagprocessors) |
|
1278 | flagutil.insertflagprocessor(flag, processor, self._flagprocessors) | |
1279 |
|
1279 | |||
1280 | chunk_cache_size = self.data_config.chunk_cache_size |
|
1280 | chunk_cache_size = self.data_config.chunk_cache_size | |
1281 | if chunk_cache_size <= 0: |
|
1281 | if chunk_cache_size <= 0: | |
1282 | raise error.RevlogError( |
|
1282 | raise error.RevlogError( | |
1283 | _(b'revlog chunk cache size %r is not greater than 0') |
|
1283 | _(b'revlog chunk cache size %r is not greater than 0') | |
1284 | % chunk_cache_size |
|
1284 | % chunk_cache_size | |
1285 | ) |
|
1285 | ) | |
1286 | elif chunk_cache_size & (chunk_cache_size - 1): |
|
1286 | elif chunk_cache_size & (chunk_cache_size - 1): | |
1287 | raise error.RevlogError( |
|
1287 | raise error.RevlogError( | |
1288 | _(b'revlog chunk cache size %r is not a power of 2') |
|
1288 | _(b'revlog chunk cache size %r is not a power of 2') | |
1289 | % chunk_cache_size |
|
1289 | % chunk_cache_size | |
1290 | ) |
|
1290 | ) | |
1291 | force_nodemap = opts.get(b'devel-force-nodemap', False) |
|
1291 | force_nodemap = opts.get(b'devel-force-nodemap', False) | |
1292 | return new_header, mmapindexthreshold, force_nodemap |
|
1292 | return new_header, mmapindexthreshold, force_nodemap | |
1293 |
|
1293 | |||
1294 | def _get_data(self, filepath, mmap_threshold, size=None): |
|
1294 | def _get_data(self, filepath, mmap_threshold, size=None): | |
1295 | """return a file content with or without mmap |
|
1295 | """return a file content with or without mmap | |
1296 |
|
1296 | |||
1297 | If the file is missing return the empty string""" |
|
1297 | If the file is missing return the empty string""" | |
1298 | try: |
|
1298 | try: | |
1299 | with self.opener(filepath) as fp: |
|
1299 | with self.opener(filepath) as fp: | |
1300 | if mmap_threshold is not None: |
|
1300 | if mmap_threshold is not None: | |
1301 | file_size = self.opener.fstat(fp).st_size |
|
1301 | file_size = self.opener.fstat(fp).st_size | |
1302 | if file_size >= mmap_threshold: |
|
1302 | if file_size >= mmap_threshold: | |
1303 | if size is not None: |
|
1303 | if size is not None: | |
1304 | # avoid potentiel mmap crash |
|
1304 | # avoid potentiel mmap crash | |
1305 | size = min(file_size, size) |
|
1305 | size = min(file_size, size) | |
1306 | # TODO: should .close() to release resources without |
|
1306 | # TODO: should .close() to release resources without | |
1307 | # relying on Python GC |
|
1307 | # relying on Python GC | |
1308 | if size is None: |
|
1308 | if size is None: | |
1309 | return util.buffer(util.mmapread(fp)) |
|
1309 | return util.buffer(util.mmapread(fp)) | |
1310 | else: |
|
1310 | else: | |
1311 | return util.buffer(util.mmapread(fp, size)) |
|
1311 | return util.buffer(util.mmapread(fp, size)) | |
1312 | if size is None: |
|
1312 | if size is None: | |
1313 | return fp.read() |
|
1313 | return fp.read() | |
1314 | else: |
|
1314 | else: | |
1315 | return fp.read(size) |
|
1315 | return fp.read(size) | |
1316 | except FileNotFoundError: |
|
1316 | except FileNotFoundError: | |
1317 | return b'' |
|
1317 | return b'' | |
1318 |
|
1318 | |||
1319 | def get_streams(self, max_linkrev, force_inline=False): |
|
1319 | def get_streams(self, max_linkrev, force_inline=False): | |
1320 | """return a list of streams that represent this revlog |
|
1320 | """return a list of streams that represent this revlog | |
1321 |
|
1321 | |||
1322 | This is used by stream-clone to do bytes to bytes copies of a repository. |
|
1322 | This is used by stream-clone to do bytes to bytes copies of a repository. | |
1323 |
|
1323 | |||
1324 | This streams data for all revisions that refer to a changelog revision up |
|
1324 | This streams data for all revisions that refer to a changelog revision up | |
1325 | to `max_linkrev`. |
|
1325 | to `max_linkrev`. | |
1326 |
|
1326 | |||
1327 | If `force_inline` is set, it enforces that the stream will represent an inline revlog. |
|
1327 | If `force_inline` is set, it enforces that the stream will represent an inline revlog. | |
1328 |
|
1328 | |||
1329 | It returns is a list of three-tuple: |
|
1329 | It returns is a list of three-tuple: | |
1330 |
|
1330 | |||
1331 | [ |
|
1331 | [ | |
1332 | (filename, bytes_stream, stream_size), |
|
1332 | (filename, bytes_stream, stream_size), | |
1333 | β¦ |
|
1333 | β¦ | |
1334 | ] |
|
1334 | ] | |
1335 | """ |
|
1335 | """ | |
1336 | n = len(self) |
|
1336 | n = len(self) | |
1337 | index = self.index |
|
1337 | index = self.index | |
1338 | while n > 0: |
|
1338 | while n > 0: | |
1339 | linkrev = index[n - 1][4] |
|
1339 | linkrev = index[n - 1][4] | |
1340 | if linkrev < max_linkrev: |
|
1340 | if linkrev < max_linkrev: | |
1341 | break |
|
1341 | break | |
1342 | # note: this loop will rarely go through multiple iterations, since |
|
1342 | # note: this loop will rarely go through multiple iterations, since | |
1343 | # it only traverses commits created during the current streaming |
|
1343 | # it only traverses commits created during the current streaming | |
1344 | # pull operation. |
|
1344 | # pull operation. | |
1345 | # |
|
1345 | # | |
1346 | # If this become a problem, using a binary search should cap the |
|
1346 | # If this become a problem, using a binary search should cap the | |
1347 | # runtime of this. |
|
1347 | # runtime of this. | |
1348 | n = n - 1 |
|
1348 | n = n - 1 | |
1349 | if n == 0: |
|
1349 | if n == 0: | |
1350 | # no data to send |
|
1350 | # no data to send | |
1351 | return [] |
|
1351 | return [] | |
1352 | index_size = n * index.entry_size |
|
1352 | index_size = n * index.entry_size | |
1353 | data_size = self.end(n - 1) |
|
1353 | data_size = self.end(n - 1) | |
1354 |
|
1354 | |||
1355 | # XXX we might have been split (or stripped) since the object |
|
1355 | # XXX we might have been split (or stripped) since the object | |
1356 | # initialization, We need to close this race too, but having a way to |
|
1356 | # initialization, We need to close this race too, but having a way to | |
1357 | # pre-open the file we feed to the revlog and never closing them before |
|
1357 | # pre-open the file we feed to the revlog and never closing them before | |
1358 | # we are done streaming. |
|
1358 | # we are done streaming. | |
1359 |
|
1359 | |||
1360 | if self._inline: |
|
1360 | if self._inline: | |
1361 |
|
1361 | |||
1362 | def get_stream(): |
|
1362 | def get_stream(): | |
1363 | with self.opener(self._indexfile, mode=b"r") as fp: |
|
1363 | with self.opener(self._indexfile, mode=b"r") as fp: | |
1364 | yield None |
|
1364 | yield None | |
1365 | size = index_size + data_size |
|
1365 | size = index_size + data_size | |
1366 | if size <= 65536: |
|
1366 | if size <= 65536: | |
1367 | yield fp.read(size) |
|
1367 | yield fp.read(size) | |
1368 | else: |
|
1368 | else: | |
1369 | yield from util.filechunkiter(fp, limit=size) |
|
1369 | yield from util.filechunkiter(fp, limit=size) | |
1370 |
|
1370 | |||
1371 | inline_stream = get_stream() |
|
1371 | inline_stream = get_stream() | |
1372 | next(inline_stream) |
|
1372 | next(inline_stream) | |
1373 | return [ |
|
1373 | return [ | |
1374 | (self._indexfile, inline_stream, index_size + data_size), |
|
1374 | (self._indexfile, inline_stream, index_size + data_size), | |
1375 | ] |
|
1375 | ] | |
1376 | elif force_inline: |
|
1376 | elif force_inline: | |
1377 |
|
1377 | |||
1378 | def get_stream(): |
|
1378 | def get_stream(): | |
1379 | with self.reading(): |
|
1379 | with self.reading(): | |
1380 | yield None |
|
1380 | yield None | |
1381 |
|
1381 | |||
1382 | for rev in range(n): |
|
1382 | for rev in range(n): | |
1383 | idx = self.index.entry_binary(rev) |
|
1383 | idx = self.index.entry_binary(rev) | |
1384 | if rev == 0 and self._docket is None: |
|
1384 | if rev == 0 and self._docket is None: | |
1385 | # re-inject the inline flag |
|
1385 | # re-inject the inline flag | |
1386 | header = self._format_flags |
|
1386 | header = self._format_flags | |
1387 | header |= self._format_version |
|
1387 | header |= self._format_version | |
1388 | header |= FLAG_INLINE_DATA |
|
1388 | header |= FLAG_INLINE_DATA | |
1389 | header = self.index.pack_header(header) |
|
1389 | header = self.index.pack_header(header) | |
1390 | idx = header + idx |
|
1390 | idx = header + idx | |
1391 | yield idx |
|
1391 | yield idx | |
1392 | yield self._inner.get_segment_for_revs(rev, rev)[1] |
|
1392 | yield self._inner.get_segment_for_revs(rev, rev)[1] | |
1393 |
|
1393 | |||
1394 | inline_stream = get_stream() |
|
1394 | inline_stream = get_stream() | |
1395 | next(inline_stream) |
|
1395 | next(inline_stream) | |
1396 | return [ |
|
1396 | return [ | |
1397 | (self._indexfile, inline_stream, index_size + data_size), |
|
1397 | (self._indexfile, inline_stream, index_size + data_size), | |
1398 | ] |
|
1398 | ] | |
1399 | else: |
|
1399 | else: | |
1400 |
|
1400 | |||
1401 | def get_index_stream(): |
|
1401 | def get_index_stream(): | |
1402 | with self.opener(self._indexfile, mode=b"r") as fp: |
|
1402 | with self.opener(self._indexfile, mode=b"r") as fp: | |
1403 | yield None |
|
1403 | yield None | |
1404 | if index_size <= 65536: |
|
1404 | if index_size <= 65536: | |
1405 | yield fp.read(index_size) |
|
1405 | yield fp.read(index_size) | |
1406 | else: |
|
1406 | else: | |
1407 | yield from util.filechunkiter(fp, limit=index_size) |
|
1407 | yield from util.filechunkiter(fp, limit=index_size) | |
1408 |
|
1408 | |||
1409 | def get_data_stream(): |
|
1409 | def get_data_stream(): | |
1410 | with self._datafp() as fp: |
|
1410 | with self._datafp() as fp: | |
1411 | yield None |
|
1411 | yield None | |
1412 | if data_size <= 65536: |
|
1412 | if data_size <= 65536: | |
1413 | yield fp.read(data_size) |
|
1413 | yield fp.read(data_size) | |
1414 | else: |
|
1414 | else: | |
1415 | yield from util.filechunkiter(fp, limit=data_size) |
|
1415 | yield from util.filechunkiter(fp, limit=data_size) | |
1416 |
|
1416 | |||
1417 | index_stream = get_index_stream() |
|
1417 | index_stream = get_index_stream() | |
1418 | next(index_stream) |
|
1418 | next(index_stream) | |
1419 | data_stream = get_data_stream() |
|
1419 | data_stream = get_data_stream() | |
1420 | next(data_stream) |
|
1420 | next(data_stream) | |
1421 | return [ |
|
1421 | return [ | |
1422 | (self._datafile, data_stream, data_size), |
|
1422 | (self._datafile, data_stream, data_size), | |
1423 | (self._indexfile, index_stream, index_size), |
|
1423 | (self._indexfile, index_stream, index_size), | |
1424 | ] |
|
1424 | ] | |
1425 |
|
1425 | |||
1426 | def _loadindex(self, docket=None): |
|
1426 | def _loadindex(self, docket=None): | |
1427 |
|
1427 | |||
1428 | new_header, mmapindexthreshold, force_nodemap = self._init_opts() |
|
1428 | new_header, mmapindexthreshold, force_nodemap = self._init_opts() | |
1429 |
|
1429 | |||
1430 | if self.postfix is not None: |
|
1430 | if self.postfix is not None: | |
1431 | entry_point = b'%s.i.%s' % (self.radix, self.postfix) |
|
1431 | entry_point = b'%s.i.%s' % (self.radix, self.postfix) | |
1432 | elif self._trypending and self.opener.exists(b'%s.i.a' % self.radix): |
|
1432 | elif self._trypending and self.opener.exists(b'%s.i.a' % self.radix): | |
1433 | entry_point = b'%s.i.a' % self.radix |
|
1433 | entry_point = b'%s.i.a' % self.radix | |
1434 | elif self._try_split and self.opener.exists(self._split_index_file): |
|
1434 | elif self._try_split and self.opener.exists(self._split_index_file): | |
1435 | entry_point = self._split_index_file |
|
1435 | entry_point = self._split_index_file | |
1436 | else: |
|
1436 | else: | |
1437 | entry_point = b'%s.i' % self.radix |
|
1437 | entry_point = b'%s.i' % self.radix | |
1438 |
|
1438 | |||
1439 | if docket is not None: |
|
1439 | if docket is not None: | |
1440 | self._docket = docket |
|
1440 | self._docket = docket | |
1441 | self._docket_file = entry_point |
|
1441 | self._docket_file = entry_point | |
1442 | else: |
|
1442 | else: | |
1443 | self._initempty = True |
|
1443 | self._initempty = True | |
1444 | entry_data = self._get_data(entry_point, mmapindexthreshold) |
|
1444 | entry_data = self._get_data(entry_point, mmapindexthreshold) | |
1445 | if len(entry_data) > 0: |
|
1445 | if len(entry_data) > 0: | |
1446 | header = INDEX_HEADER.unpack(entry_data[:4])[0] |
|
1446 | header = INDEX_HEADER.unpack(entry_data[:4])[0] | |
1447 | self._initempty = False |
|
1447 | self._initempty = False | |
1448 | else: |
|
1448 | else: | |
1449 | header = new_header |
|
1449 | header = new_header | |
1450 |
|
1450 | |||
1451 | self._format_flags = header & ~0xFFFF |
|
1451 | self._format_flags = header & ~0xFFFF | |
1452 | self._format_version = header & 0xFFFF |
|
1452 | self._format_version = header & 0xFFFF | |
1453 |
|
1453 | |||
1454 | supported_flags = SUPPORTED_FLAGS.get(self._format_version) |
|
1454 | supported_flags = SUPPORTED_FLAGS.get(self._format_version) | |
1455 | if supported_flags is None: |
|
1455 | if supported_flags is None: | |
1456 | msg = _(b'unknown version (%d) in revlog %s') |
|
1456 | msg = _(b'unknown version (%d) in revlog %s') | |
1457 | msg %= (self._format_version, self.display_id) |
|
1457 | msg %= (self._format_version, self.display_id) | |
1458 | raise error.RevlogError(msg) |
|
1458 | raise error.RevlogError(msg) | |
1459 | elif self._format_flags & ~supported_flags: |
|
1459 | elif self._format_flags & ~supported_flags: | |
1460 | msg = _(b'unknown flags (%#04x) in version %d revlog %s') |
|
1460 | msg = _(b'unknown flags (%#04x) in version %d revlog %s') | |
1461 | display_flag = self._format_flags >> 16 |
|
1461 | display_flag = self._format_flags >> 16 | |
1462 | msg %= (display_flag, self._format_version, self.display_id) |
|
1462 | msg %= (display_flag, self._format_version, self.display_id) | |
1463 | raise error.RevlogError(msg) |
|
1463 | raise error.RevlogError(msg) | |
1464 |
|
1464 | |||
1465 | features = FEATURES_BY_VERSION[self._format_version] |
|
1465 | features = FEATURES_BY_VERSION[self._format_version] | |
1466 | self._inline = features[b'inline'](self._format_flags) |
|
1466 | self._inline = features[b'inline'](self._format_flags) | |
1467 | self.delta_config.general_delta = features[b'generaldelta']( |
|
1467 | self.delta_config.general_delta = features[b'generaldelta']( | |
1468 | self._format_flags |
|
1468 | self._format_flags | |
1469 | ) |
|
1469 | ) | |
1470 | self.feature_config.has_side_data = features[b'sidedata'] |
|
1470 | self.feature_config.has_side_data = features[b'sidedata'] | |
1471 |
|
1471 | |||
1472 | if not features[b'docket']: |
|
1472 | if not features[b'docket']: | |
1473 | self._indexfile = entry_point |
|
1473 | self._indexfile = entry_point | |
1474 | index_data = entry_data |
|
1474 | index_data = entry_data | |
1475 | else: |
|
1475 | else: | |
1476 | self._docket_file = entry_point |
|
1476 | self._docket_file = entry_point | |
1477 | if self._initempty: |
|
1477 | if self._initempty: | |
1478 | self._docket = docketutil.default_docket(self, header) |
|
1478 | self._docket = docketutil.default_docket(self, header) | |
1479 | else: |
|
1479 | else: | |
1480 | self._docket = docketutil.parse_docket( |
|
1480 | self._docket = docketutil.parse_docket( | |
1481 | self, entry_data, use_pending=self._trypending |
|
1481 | self, entry_data, use_pending=self._trypending | |
1482 | ) |
|
1482 | ) | |
1483 |
|
1483 | |||
1484 | if self._docket is not None: |
|
1484 | if self._docket is not None: | |
1485 | self._indexfile = self._docket.index_filepath() |
|
1485 | self._indexfile = self._docket.index_filepath() | |
1486 | index_data = b'' |
|
1486 | index_data = b'' | |
1487 | index_size = self._docket.index_end |
|
1487 | index_size = self._docket.index_end | |
1488 | if index_size > 0: |
|
1488 | if index_size > 0: | |
1489 | index_data = self._get_data( |
|
1489 | index_data = self._get_data( | |
1490 | self._indexfile, mmapindexthreshold, size=index_size |
|
1490 | self._indexfile, mmapindexthreshold, size=index_size | |
1491 | ) |
|
1491 | ) | |
1492 | if len(index_data) < index_size: |
|
1492 | if len(index_data) < index_size: | |
1493 | msg = _(b'too few index data for %s: got %d, expected %d') |
|
1493 | msg = _(b'too few index data for %s: got %d, expected %d') | |
1494 | msg %= (self.display_id, len(index_data), index_size) |
|
1494 | msg %= (self.display_id, len(index_data), index_size) | |
1495 | raise error.RevlogError(msg) |
|
1495 | raise error.RevlogError(msg) | |
1496 |
|
1496 | |||
1497 | self._inline = False |
|
1497 | self._inline = False | |
1498 | # generaldelta implied by version 2 revlogs. |
|
1498 | # generaldelta implied by version 2 revlogs. | |
1499 | self.delta_config.general_delta = True |
|
1499 | self.delta_config.general_delta = True | |
1500 | # the logic for persistent nodemap will be dealt with within the |
|
1500 | # the logic for persistent nodemap will be dealt with within the | |
1501 | # main docket, so disable it for now. |
|
1501 | # main docket, so disable it for now. | |
1502 | self._nodemap_file = None |
|
1502 | self._nodemap_file = None | |
1503 |
|
1503 | |||
1504 | if self._docket is not None: |
|
1504 | if self._docket is not None: | |
1505 | self._datafile = self._docket.data_filepath() |
|
1505 | self._datafile = self._docket.data_filepath() | |
1506 | self._sidedatafile = self._docket.sidedata_filepath() |
|
1506 | self._sidedatafile = self._docket.sidedata_filepath() | |
1507 | elif self.postfix is None: |
|
1507 | elif self.postfix is None: | |
1508 | self._datafile = b'%s.d' % self.radix |
|
1508 | self._datafile = b'%s.d' % self.radix | |
1509 | else: |
|
1509 | else: | |
1510 | self._datafile = b'%s.d.%s' % (self.radix, self.postfix) |
|
1510 | self._datafile = b'%s.d.%s' % (self.radix, self.postfix) | |
1511 |
|
1511 | |||
1512 | self.nodeconstants = sha1nodeconstants |
|
1512 | self.nodeconstants = sha1nodeconstants | |
1513 | self.nullid = self.nodeconstants.nullid |
|
1513 | self.nullid = self.nodeconstants.nullid | |
1514 |
|
1514 | |||
1515 | # sparse-revlog can't be on without general-delta (issue6056) |
|
1515 | # sparse-revlog can't be on without general-delta (issue6056) | |
1516 | if not self.delta_config.general_delta: |
|
1516 | if not self.delta_config.general_delta: | |
1517 | self.delta_config.sparse_revlog = False |
|
1517 | self.delta_config.sparse_revlog = False | |
1518 |
|
1518 | |||
1519 | self._storedeltachains = True |
|
1519 | self._storedeltachains = True | |
1520 |
|
1520 | |||
1521 | devel_nodemap = ( |
|
1521 | devel_nodemap = ( | |
1522 | self._nodemap_file |
|
1522 | self._nodemap_file | |
1523 | and force_nodemap |
|
1523 | and force_nodemap | |
1524 | and parse_index_v1_nodemap is not None |
|
1524 | and parse_index_v1_nodemap is not None | |
1525 | ) |
|
1525 | ) | |
1526 |
|
1526 | |||
1527 | use_rust_index = False |
|
1527 | use_rust_index = False | |
1528 | if rustrevlog is not None: |
|
1528 | if rustrevlog is not None: | |
1529 | if self._nodemap_file is not None: |
|
1529 | if self._nodemap_file is not None: | |
1530 | use_rust_index = True |
|
1530 | use_rust_index = True | |
1531 | else: |
|
1531 | else: | |
1532 | use_rust_index = self.opener.options.get(b'rust.index') |
|
1532 | use_rust_index = self.opener.options.get(b'rust.index') | |
1533 |
|
1533 | |||
1534 | self._parse_index = parse_index_v1 |
|
1534 | self._parse_index = parse_index_v1 | |
1535 | if self._format_version == REVLOGV0: |
|
1535 | if self._format_version == REVLOGV0: | |
1536 | self._parse_index = revlogv0.parse_index_v0 |
|
1536 | self._parse_index = revlogv0.parse_index_v0 | |
1537 | elif self._format_version == REVLOGV2: |
|
1537 | elif self._format_version == REVLOGV2: | |
1538 | self._parse_index = parse_index_v2 |
|
1538 | self._parse_index = parse_index_v2 | |
1539 | elif self._format_version == CHANGELOGV2: |
|
1539 | elif self._format_version == CHANGELOGV2: | |
1540 | self._parse_index = parse_index_cl_v2 |
|
1540 | self._parse_index = parse_index_cl_v2 | |
1541 | elif devel_nodemap: |
|
1541 | elif devel_nodemap: | |
1542 | self._parse_index = parse_index_v1_nodemap |
|
1542 | self._parse_index = parse_index_v1_nodemap | |
1543 | elif use_rust_index: |
|
1543 | elif use_rust_index: | |
1544 | self._parse_index = parse_index_v1_mixed |
|
1544 | self._parse_index = parse_index_v1_mixed | |
1545 | try: |
|
1545 | try: | |
1546 | d = self._parse_index(index_data, self._inline) |
|
1546 | d = self._parse_index(index_data, self._inline) | |
1547 | index, chunkcache = d |
|
1547 | index, chunkcache = d | |
1548 | use_nodemap = ( |
|
1548 | use_nodemap = ( | |
1549 | not self._inline |
|
1549 | not self._inline | |
1550 | and self._nodemap_file is not None |
|
1550 | and self._nodemap_file is not None | |
1551 | and hasattr(index, 'update_nodemap_data') |
|
1551 | and hasattr(index, 'update_nodemap_data') | |
1552 | ) |
|
1552 | ) | |
1553 | if use_nodemap: |
|
1553 | if use_nodemap: | |
1554 | nodemap_data = nodemaputil.persisted_data(self) |
|
1554 | nodemap_data = nodemaputil.persisted_data(self) | |
1555 | if nodemap_data is not None: |
|
1555 | if nodemap_data is not None: | |
1556 | docket = nodemap_data[0] |
|
1556 | docket = nodemap_data[0] | |
1557 | if ( |
|
1557 | if ( | |
1558 | len(d[0]) > docket.tip_rev |
|
1558 | len(d[0]) > docket.tip_rev | |
1559 | and d[0][docket.tip_rev][7] == docket.tip_node |
|
1559 | and d[0][docket.tip_rev][7] == docket.tip_node | |
1560 | ): |
|
1560 | ): | |
1561 | # no changelog tampering |
|
1561 | # no changelog tampering | |
1562 | self._nodemap_docket = docket |
|
1562 | self._nodemap_docket = docket | |
1563 | index.update_nodemap_data(*nodemap_data) |
|
1563 | index.update_nodemap_data(*nodemap_data) | |
1564 | except (ValueError, IndexError): |
|
1564 | except (ValueError, IndexError): | |
1565 | raise error.RevlogError( |
|
1565 | raise error.RevlogError( | |
1566 | _(b"index %s is corrupted") % self.display_id |
|
1566 | _(b"index %s is corrupted") % self.display_id | |
1567 | ) |
|
1567 | ) | |
1568 | self.index = index |
|
1568 | self.index = index | |
1569 | # revnum -> (chain-length, sum-delta-length) |
|
1569 | # revnum -> (chain-length, sum-delta-length) | |
1570 | self._chaininfocache = util.lrucachedict(500) |
|
1570 | self._chaininfocache = util.lrucachedict(500) | |
1571 |
|
1571 | |||
1572 | return chunkcache |
|
1572 | return chunkcache | |
1573 |
|
1573 | |||
1574 | def _load_inner(self, chunk_cache): |
|
1574 | def _load_inner(self, chunk_cache): | |
1575 | if self._docket is None: |
|
1575 | if self._docket is None: | |
1576 | default_compression_header = None |
|
1576 | default_compression_header = None | |
1577 | else: |
|
1577 | else: | |
1578 | default_compression_header = self._docket.default_compression_header |
|
1578 | default_compression_header = self._docket.default_compression_header | |
1579 |
|
1579 | |||
1580 | self._inner = _InnerRevlog( |
|
1580 | self._inner = _InnerRevlog( | |
1581 | opener=self.opener, |
|
1581 | opener=self.opener, | |
1582 | index=self.index, |
|
1582 | index=self.index, | |
1583 | index_file=self._indexfile, |
|
1583 | index_file=self._indexfile, | |
1584 | data_file=self._datafile, |
|
1584 | data_file=self._datafile, | |
1585 | sidedata_file=self._sidedatafile, |
|
1585 | sidedata_file=self._sidedatafile, | |
1586 | inline=self._inline, |
|
1586 | inline=self._inline, | |
1587 | data_config=self.data_config, |
|
1587 | data_config=self.data_config, | |
1588 | delta_config=self.delta_config, |
|
1588 | delta_config=self.delta_config, | |
1589 | feature_config=self.feature_config, |
|
1589 | feature_config=self.feature_config, | |
1590 | chunk_cache=chunk_cache, |
|
1590 | chunk_cache=chunk_cache, | |
1591 | default_compression_header=default_compression_header, |
|
1591 | default_compression_header=default_compression_header, | |
1592 | ) |
|
1592 | ) | |
1593 |
|
1593 | |||
1594 | def get_revlog(self): |
|
1594 | def get_revlog(self): | |
1595 | """simple function to mirror API of other not-really-revlog API""" |
|
1595 | """simple function to mirror API of other not-really-revlog API""" | |
1596 | return self |
|
1596 | return self | |
1597 |
|
1597 | |||
1598 | @util.propertycache |
|
1598 | @util.propertycache | |
1599 | def revlog_kind(self): |
|
1599 | def revlog_kind(self): | |
1600 | return self.target[0] |
|
1600 | return self.target[0] | |
1601 |
|
1601 | |||
1602 | @util.propertycache |
|
1602 | @util.propertycache | |
1603 | def display_id(self): |
|
1603 | def display_id(self): | |
1604 | """The public facing "ID" of the revlog that we use in message""" |
|
1604 | """The public facing "ID" of the revlog that we use in message""" | |
1605 | if self.revlog_kind == KIND_FILELOG: |
|
1605 | if self.revlog_kind == KIND_FILELOG: | |
1606 | # Reference the file without the "data/" prefix, so it is familiar |
|
1606 | # Reference the file without the "data/" prefix, so it is familiar | |
1607 | # to the user. |
|
1607 | # to the user. | |
1608 | return self.target[1] |
|
1608 | return self.target[1] | |
1609 | else: |
|
1609 | else: | |
1610 | return self.radix |
|
1610 | return self.radix | |
1611 |
|
1611 | |||
1612 | def _datafp(self, mode=b'r'): |
|
1612 | def _datafp(self, mode=b'r'): | |
1613 | """file object for the revlog's data file""" |
|
1613 | """file object for the revlog's data file""" | |
1614 | return self.opener(self._datafile, mode=mode) |
|
1614 | return self.opener(self._datafile, mode=mode) | |
1615 |
|
1615 | |||
1616 | def tiprev(self): |
|
1616 | def tiprev(self): | |
1617 | return len(self.index) - 1 |
|
1617 | return len(self.index) - 1 | |
1618 |
|
1618 | |||
1619 | def tip(self): |
|
1619 | def tip(self): | |
1620 | return self.node(self.tiprev()) |
|
1620 | return self.node(self.tiprev()) | |
1621 |
|
1621 | |||
1622 | def __contains__(self, rev): |
|
1622 | def __contains__(self, rev): | |
1623 | return 0 <= rev < len(self) |
|
1623 | return 0 <= rev < len(self) | |
1624 |
|
1624 | |||
1625 | def __len__(self): |
|
1625 | def __len__(self): | |
1626 | return len(self.index) |
|
1626 | return len(self.index) | |
1627 |
|
1627 | |||
1628 | def __iter__(self): |
|
1628 | def __iter__(self): | |
1629 | return iter(range(len(self))) |
|
1629 | return iter(range(len(self))) | |
1630 |
|
1630 | |||
1631 | def revs(self, start=0, stop=None): |
|
1631 | def revs(self, start=0, stop=None): | |
1632 | """iterate over all rev in this revlog (from start to stop)""" |
|
1632 | """iterate over all rev in this revlog (from start to stop)""" | |
1633 | return storageutil.iterrevs(len(self), start=start, stop=stop) |
|
1633 | return storageutil.iterrevs(len(self), start=start, stop=stop) | |
1634 |
|
1634 | |||
1635 | def hasnode(self, node): |
|
1635 | def hasnode(self, node): | |
1636 | try: |
|
1636 | try: | |
1637 | self.rev(node) |
|
1637 | self.rev(node) | |
1638 | return True |
|
1638 | return True | |
1639 | except KeyError: |
|
1639 | except KeyError: | |
1640 | return False |
|
1640 | return False | |
1641 |
|
1641 | |||
1642 | def _candelta(self, baserev, rev): |
|
1642 | def _candelta(self, baserev, rev): | |
1643 | """whether two revisions (baserev, rev) can be delta-ed or not""" |
|
1643 | """whether two revisions (baserev, rev) can be delta-ed or not""" | |
1644 | # Disable delta if either rev requires a content-changing flag |
|
1644 | # Disable delta if either rev requires a content-changing flag | |
1645 | # processor (ex. LFS). This is because such flag processor can alter |
|
1645 | # processor (ex. LFS). This is because such flag processor can alter | |
1646 | # the rawtext content that the delta will be based on, and two clients |
|
1646 | # the rawtext content that the delta will be based on, and two clients | |
1647 | # could have a same revlog node with different flags (i.e. different |
|
1647 | # could have a same revlog node with different flags (i.e. different | |
1648 | # rawtext contents) and the delta could be incompatible. |
|
1648 | # rawtext contents) and the delta could be incompatible. | |
1649 | if (self.flags(baserev) & REVIDX_RAWTEXT_CHANGING_FLAGS) or ( |
|
1649 | if (self.flags(baserev) & REVIDX_RAWTEXT_CHANGING_FLAGS) or ( | |
1650 | self.flags(rev) & REVIDX_RAWTEXT_CHANGING_FLAGS |
|
1650 | self.flags(rev) & REVIDX_RAWTEXT_CHANGING_FLAGS | |
1651 | ): |
|
1651 | ): | |
1652 | return False |
|
1652 | return False | |
1653 | return True |
|
1653 | return True | |
1654 |
|
1654 | |||
1655 | def update_caches(self, transaction): |
|
1655 | def update_caches(self, transaction): | |
1656 | """update on disk cache |
|
1656 | """update on disk cache | |
1657 |
|
1657 | |||
1658 | If a transaction is passed, the update may be delayed to transaction |
|
1658 | If a transaction is passed, the update may be delayed to transaction | |
1659 | commit.""" |
|
1659 | commit.""" | |
1660 | if self._nodemap_file is not None: |
|
1660 | if self._nodemap_file is not None: | |
1661 | if transaction is None: |
|
1661 | if transaction is None: | |
1662 | nodemaputil.update_persistent_nodemap(self) |
|
1662 | nodemaputil.update_persistent_nodemap(self) | |
1663 | else: |
|
1663 | else: | |
1664 | nodemaputil.setup_persistent_nodemap(transaction, self) |
|
1664 | nodemaputil.setup_persistent_nodemap(transaction, self) | |
1665 |
|
1665 | |||
1666 | def clearcaches(self): |
|
1666 | def clearcaches(self): | |
1667 | """Clear in-memory caches""" |
|
1667 | """Clear in-memory caches""" | |
1668 | self._revisioncache = None |
|
1668 | self._inner._revisioncache = None | |
1669 | self._chainbasecache.clear() |
|
1669 | self._chainbasecache.clear() | |
1670 | self._inner._segmentfile.clear_cache() |
|
1670 | self._inner._segmentfile.clear_cache() | |
1671 | self._inner._segmentfile_sidedata.clear_cache() |
|
1671 | self._inner._segmentfile_sidedata.clear_cache() | |
1672 | self._pcache = {} |
|
1672 | self._pcache = {} | |
1673 | self._nodemap_docket = None |
|
1673 | self._nodemap_docket = None | |
1674 | self.index.clearcaches() |
|
1674 | self.index.clearcaches() | |
1675 | # The python code is the one responsible for validating the docket, we |
|
1675 | # The python code is the one responsible for validating the docket, we | |
1676 | # end up having to refresh it here. |
|
1676 | # end up having to refresh it here. | |
1677 | use_nodemap = ( |
|
1677 | use_nodemap = ( | |
1678 | not self._inline |
|
1678 | not self._inline | |
1679 | and self._nodemap_file is not None |
|
1679 | and self._nodemap_file is not None | |
1680 | and hasattr(self.index, 'update_nodemap_data') |
|
1680 | and hasattr(self.index, 'update_nodemap_data') | |
1681 | ) |
|
1681 | ) | |
1682 | if use_nodemap: |
|
1682 | if use_nodemap: | |
1683 | nodemap_data = nodemaputil.persisted_data(self) |
|
1683 | nodemap_data = nodemaputil.persisted_data(self) | |
1684 | if nodemap_data is not None: |
|
1684 | if nodemap_data is not None: | |
1685 | self._nodemap_docket = nodemap_data[0] |
|
1685 | self._nodemap_docket = nodemap_data[0] | |
1686 | self.index.update_nodemap_data(*nodemap_data) |
|
1686 | self.index.update_nodemap_data(*nodemap_data) | |
1687 |
|
1687 | |||
1688 | def rev(self, node): |
|
1688 | def rev(self, node): | |
1689 | """return the revision number associated with a <nodeid>""" |
|
1689 | """return the revision number associated with a <nodeid>""" | |
1690 | try: |
|
1690 | try: | |
1691 | return self.index.rev(node) |
|
1691 | return self.index.rev(node) | |
1692 | except TypeError: |
|
1692 | except TypeError: | |
1693 | raise |
|
1693 | raise | |
1694 | except error.RevlogError: |
|
1694 | except error.RevlogError: | |
1695 | # parsers.c radix tree lookup failed |
|
1695 | # parsers.c radix tree lookup failed | |
1696 | if ( |
|
1696 | if ( | |
1697 | node == self.nodeconstants.wdirid |
|
1697 | node == self.nodeconstants.wdirid | |
1698 | or node in self.nodeconstants.wdirfilenodeids |
|
1698 | or node in self.nodeconstants.wdirfilenodeids | |
1699 | ): |
|
1699 | ): | |
1700 | raise error.WdirUnsupported |
|
1700 | raise error.WdirUnsupported | |
1701 | raise error.LookupError(node, self.display_id, _(b'no node')) |
|
1701 | raise error.LookupError(node, self.display_id, _(b'no node')) | |
1702 |
|
1702 | |||
1703 | # Accessors for index entries. |
|
1703 | # Accessors for index entries. | |
1704 |
|
1704 | |||
1705 | # First tuple entry is 8 bytes. First 6 bytes are offset. Last 2 bytes |
|
1705 | # First tuple entry is 8 bytes. First 6 bytes are offset. Last 2 bytes | |
1706 | # are flags. |
|
1706 | # are flags. | |
1707 | def start(self, rev): |
|
1707 | def start(self, rev): | |
1708 | return int(self.index[rev][0] >> 16) |
|
1708 | return int(self.index[rev][0] >> 16) | |
1709 |
|
1709 | |||
1710 | def sidedata_cut_off(self, rev): |
|
1710 | def sidedata_cut_off(self, rev): | |
1711 | sd_cut_off = self.index[rev][8] |
|
1711 | sd_cut_off = self.index[rev][8] | |
1712 | if sd_cut_off != 0: |
|
1712 | if sd_cut_off != 0: | |
1713 | return sd_cut_off |
|
1713 | return sd_cut_off | |
1714 | # This is some annoying dance, because entries without sidedata |
|
1714 | # This is some annoying dance, because entries without sidedata | |
1715 | # currently use 0 as their ofsset. (instead of previous-offset + |
|
1715 | # currently use 0 as their ofsset. (instead of previous-offset + | |
1716 | # previous-size) |
|
1716 | # previous-size) | |
1717 | # |
|
1717 | # | |
1718 | # We should reconsider this sidedata β 0 sidata_offset policy. |
|
1718 | # We should reconsider this sidedata β 0 sidata_offset policy. | |
1719 | # In the meantime, we need this. |
|
1719 | # In the meantime, we need this. | |
1720 | while 0 <= rev: |
|
1720 | while 0 <= rev: | |
1721 | e = self.index[rev] |
|
1721 | e = self.index[rev] | |
1722 | if e[9] != 0: |
|
1722 | if e[9] != 0: | |
1723 | return e[8] + e[9] |
|
1723 | return e[8] + e[9] | |
1724 | rev -= 1 |
|
1724 | rev -= 1 | |
1725 | return 0 |
|
1725 | return 0 | |
1726 |
|
1726 | |||
1727 | def flags(self, rev): |
|
1727 | def flags(self, rev): | |
1728 | return self.index[rev][0] & 0xFFFF |
|
1728 | return self.index[rev][0] & 0xFFFF | |
1729 |
|
1729 | |||
1730 | def length(self, rev): |
|
1730 | def length(self, rev): | |
1731 | return self.index[rev][1] |
|
1731 | return self.index[rev][1] | |
1732 |
|
1732 | |||
1733 | def sidedata_length(self, rev): |
|
1733 | def sidedata_length(self, rev): | |
1734 | if not self.feature_config.has_side_data: |
|
1734 | if not self.feature_config.has_side_data: | |
1735 | return 0 |
|
1735 | return 0 | |
1736 | return self.index[rev][9] |
|
1736 | return self.index[rev][9] | |
1737 |
|
1737 | |||
1738 | def rawsize(self, rev): |
|
1738 | def rawsize(self, rev): | |
1739 | """return the length of the uncompressed text for a given revision""" |
|
1739 | """return the length of the uncompressed text for a given revision""" | |
1740 | l = self.index[rev][2] |
|
1740 | l = self.index[rev][2] | |
1741 | if l >= 0: |
|
1741 | if l >= 0: | |
1742 | return l |
|
1742 | return l | |
1743 |
|
1743 | |||
1744 | t = self.rawdata(rev) |
|
1744 | t = self.rawdata(rev) | |
1745 | return len(t) |
|
1745 | return len(t) | |
1746 |
|
1746 | |||
1747 | def size(self, rev): |
|
1747 | def size(self, rev): | |
1748 | """length of non-raw text (processed by a "read" flag processor)""" |
|
1748 | """length of non-raw text (processed by a "read" flag processor)""" | |
1749 | # fast path: if no "read" flag processor could change the content, |
|
1749 | # fast path: if no "read" flag processor could change the content, | |
1750 | # size is rawsize. note: ELLIPSIS is known to not change the content. |
|
1750 | # size is rawsize. note: ELLIPSIS is known to not change the content. | |
1751 | flags = self.flags(rev) |
|
1751 | flags = self.flags(rev) | |
1752 | if flags & (flagutil.REVIDX_KNOWN_FLAGS ^ REVIDX_ELLIPSIS) == 0: |
|
1752 | if flags & (flagutil.REVIDX_KNOWN_FLAGS ^ REVIDX_ELLIPSIS) == 0: | |
1753 | return self.rawsize(rev) |
|
1753 | return self.rawsize(rev) | |
1754 |
|
1754 | |||
1755 | return len(self.revision(rev)) |
|
1755 | return len(self.revision(rev)) | |
1756 |
|
1756 | |||
1757 | def fast_rank(self, rev): |
|
1757 | def fast_rank(self, rev): | |
1758 | """Return the rank of a revision if already known, or None otherwise. |
|
1758 | """Return the rank of a revision if already known, or None otherwise. | |
1759 |
|
1759 | |||
1760 | The rank of a revision is the size of the sub-graph it defines as a |
|
1760 | The rank of a revision is the size of the sub-graph it defines as a | |
1761 | head. Equivalently, the rank of a revision `r` is the size of the set |
|
1761 | head. Equivalently, the rank of a revision `r` is the size of the set | |
1762 | `ancestors(r)`, `r` included. |
|
1762 | `ancestors(r)`, `r` included. | |
1763 |
|
1763 | |||
1764 | This method returns the rank retrieved from the revlog in constant |
|
1764 | This method returns the rank retrieved from the revlog in constant | |
1765 | time. It makes no attempt at computing unknown values for versions of |
|
1765 | time. It makes no attempt at computing unknown values for versions of | |
1766 | the revlog which do not persist the rank. |
|
1766 | the revlog which do not persist the rank. | |
1767 | """ |
|
1767 | """ | |
1768 | rank = self.index[rev][ENTRY_RANK] |
|
1768 | rank = self.index[rev][ENTRY_RANK] | |
1769 | if self._format_version != CHANGELOGV2 or rank == RANK_UNKNOWN: |
|
1769 | if self._format_version != CHANGELOGV2 or rank == RANK_UNKNOWN: | |
1770 | return None |
|
1770 | return None | |
1771 | if rev == nullrev: |
|
1771 | if rev == nullrev: | |
1772 | return 0 # convention |
|
1772 | return 0 # convention | |
1773 | return rank |
|
1773 | return rank | |
1774 |
|
1774 | |||
1775 | def chainbase(self, rev): |
|
1775 | def chainbase(self, rev): | |
1776 | base = self._chainbasecache.get(rev) |
|
1776 | base = self._chainbasecache.get(rev) | |
1777 | if base is not None: |
|
1777 | if base is not None: | |
1778 | return base |
|
1778 | return base | |
1779 |
|
1779 | |||
1780 | index = self.index |
|
1780 | index = self.index | |
1781 | iterrev = rev |
|
1781 | iterrev = rev | |
1782 | base = index[iterrev][3] |
|
1782 | base = index[iterrev][3] | |
1783 | while base != iterrev: |
|
1783 | while base != iterrev: | |
1784 | iterrev = base |
|
1784 | iterrev = base | |
1785 | base = index[iterrev][3] |
|
1785 | base = index[iterrev][3] | |
1786 |
|
1786 | |||
1787 | self._chainbasecache[rev] = base |
|
1787 | self._chainbasecache[rev] = base | |
1788 | return base |
|
1788 | return base | |
1789 |
|
1789 | |||
1790 | def linkrev(self, rev): |
|
1790 | def linkrev(self, rev): | |
1791 | return self.index[rev][4] |
|
1791 | return self.index[rev][4] | |
1792 |
|
1792 | |||
1793 | def parentrevs(self, rev): |
|
1793 | def parentrevs(self, rev): | |
1794 | try: |
|
1794 | try: | |
1795 | entry = self.index[rev] |
|
1795 | entry = self.index[rev] | |
1796 | except IndexError: |
|
1796 | except IndexError: | |
1797 | if rev == wdirrev: |
|
1797 | if rev == wdirrev: | |
1798 | raise error.WdirUnsupported |
|
1798 | raise error.WdirUnsupported | |
1799 | raise |
|
1799 | raise | |
1800 |
|
1800 | |||
1801 | if self.feature_config.canonical_parent_order and entry[5] == nullrev: |
|
1801 | if self.feature_config.canonical_parent_order and entry[5] == nullrev: | |
1802 | return entry[6], entry[5] |
|
1802 | return entry[6], entry[5] | |
1803 | else: |
|
1803 | else: | |
1804 | return entry[5], entry[6] |
|
1804 | return entry[5], entry[6] | |
1805 |
|
1805 | |||
1806 | # fast parentrevs(rev) where rev isn't filtered |
|
1806 | # fast parentrevs(rev) where rev isn't filtered | |
1807 | _uncheckedparentrevs = parentrevs |
|
1807 | _uncheckedparentrevs = parentrevs | |
1808 |
|
1808 | |||
1809 | def node(self, rev): |
|
1809 | def node(self, rev): | |
1810 | try: |
|
1810 | try: | |
1811 | return self.index[rev][7] |
|
1811 | return self.index[rev][7] | |
1812 | except IndexError: |
|
1812 | except IndexError: | |
1813 | if rev == wdirrev: |
|
1813 | if rev == wdirrev: | |
1814 | raise error.WdirUnsupported |
|
1814 | raise error.WdirUnsupported | |
1815 | raise |
|
1815 | raise | |
1816 |
|
1816 | |||
1817 | # Derived from index values. |
|
1817 | # Derived from index values. | |
1818 |
|
1818 | |||
1819 | def end(self, rev): |
|
1819 | def end(self, rev): | |
1820 | return self.start(rev) + self.length(rev) |
|
1820 | return self.start(rev) + self.length(rev) | |
1821 |
|
1821 | |||
1822 | def parents(self, node): |
|
1822 | def parents(self, node): | |
1823 | i = self.index |
|
1823 | i = self.index | |
1824 | d = i[self.rev(node)] |
|
1824 | d = i[self.rev(node)] | |
1825 | # inline node() to avoid function call overhead |
|
1825 | # inline node() to avoid function call overhead | |
1826 | if self.feature_config.canonical_parent_order and d[5] == self.nullid: |
|
1826 | if self.feature_config.canonical_parent_order and d[5] == self.nullid: | |
1827 | return i[d[6]][7], i[d[5]][7] |
|
1827 | return i[d[6]][7], i[d[5]][7] | |
1828 | else: |
|
1828 | else: | |
1829 | return i[d[5]][7], i[d[6]][7] |
|
1829 | return i[d[5]][7], i[d[6]][7] | |
1830 |
|
1830 | |||
1831 | def chainlen(self, rev): |
|
1831 | def chainlen(self, rev): | |
1832 | return self._chaininfo(rev)[0] |
|
1832 | return self._chaininfo(rev)[0] | |
1833 |
|
1833 | |||
1834 | def _chaininfo(self, rev): |
|
1834 | def _chaininfo(self, rev): | |
1835 | chaininfocache = self._chaininfocache |
|
1835 | chaininfocache = self._chaininfocache | |
1836 | if rev in chaininfocache: |
|
1836 | if rev in chaininfocache: | |
1837 | return chaininfocache[rev] |
|
1837 | return chaininfocache[rev] | |
1838 | index = self.index |
|
1838 | index = self.index | |
1839 | generaldelta = self.delta_config.general_delta |
|
1839 | generaldelta = self.delta_config.general_delta | |
1840 | iterrev = rev |
|
1840 | iterrev = rev | |
1841 | e = index[iterrev] |
|
1841 | e = index[iterrev] | |
1842 | clen = 0 |
|
1842 | clen = 0 | |
1843 | compresseddeltalen = 0 |
|
1843 | compresseddeltalen = 0 | |
1844 | while iterrev != e[3]: |
|
1844 | while iterrev != e[3]: | |
1845 | clen += 1 |
|
1845 | clen += 1 | |
1846 | compresseddeltalen += e[1] |
|
1846 | compresseddeltalen += e[1] | |
1847 | if generaldelta: |
|
1847 | if generaldelta: | |
1848 | iterrev = e[3] |
|
1848 | iterrev = e[3] | |
1849 | else: |
|
1849 | else: | |
1850 | iterrev -= 1 |
|
1850 | iterrev -= 1 | |
1851 | if iterrev in chaininfocache: |
|
1851 | if iterrev in chaininfocache: | |
1852 | t = chaininfocache[iterrev] |
|
1852 | t = chaininfocache[iterrev] | |
1853 | clen += t[0] |
|
1853 | clen += t[0] | |
1854 | compresseddeltalen += t[1] |
|
1854 | compresseddeltalen += t[1] | |
1855 | break |
|
1855 | break | |
1856 | e = index[iterrev] |
|
1856 | e = index[iterrev] | |
1857 | else: |
|
1857 | else: | |
1858 | # Add text length of base since decompressing that also takes |
|
1858 | # Add text length of base since decompressing that also takes | |
1859 | # work. For cache hits the length is already included. |
|
1859 | # work. For cache hits the length is already included. | |
1860 | compresseddeltalen += e[1] |
|
1860 | compresseddeltalen += e[1] | |
1861 | r = (clen, compresseddeltalen) |
|
1861 | r = (clen, compresseddeltalen) | |
1862 | chaininfocache[rev] = r |
|
1862 | chaininfocache[rev] = r | |
1863 | return r |
|
1863 | return r | |
1864 |
|
1864 | |||
1865 | def _deltachain(self, rev, stoprev=None): |
|
1865 | def _deltachain(self, rev, stoprev=None): | |
1866 | return self._inner._deltachain(rev, stoprev=stoprev) |
|
1866 | return self._inner._deltachain(rev, stoprev=stoprev) | |
1867 |
|
1867 | |||
1868 | def ancestors(self, revs, stoprev=0, inclusive=False): |
|
1868 | def ancestors(self, revs, stoprev=0, inclusive=False): | |
1869 | """Generate the ancestors of 'revs' in reverse revision order. |
|
1869 | """Generate the ancestors of 'revs' in reverse revision order. | |
1870 | Does not generate revs lower than stoprev. |
|
1870 | Does not generate revs lower than stoprev. | |
1871 |
|
1871 | |||
1872 | See the documentation for ancestor.lazyancestors for more details.""" |
|
1872 | See the documentation for ancestor.lazyancestors for more details.""" | |
1873 |
|
1873 | |||
1874 | # first, make sure start revisions aren't filtered |
|
1874 | # first, make sure start revisions aren't filtered | |
1875 | revs = list(revs) |
|
1875 | revs = list(revs) | |
1876 | checkrev = self.node |
|
1876 | checkrev = self.node | |
1877 | for r in revs: |
|
1877 | for r in revs: | |
1878 | checkrev(r) |
|
1878 | checkrev(r) | |
1879 | # and we're sure ancestors aren't filtered as well |
|
1879 | # and we're sure ancestors aren't filtered as well | |
1880 |
|
1880 | |||
1881 | if rustancestor is not None and self.index.rust_ext_compat: |
|
1881 | if rustancestor is not None and self.index.rust_ext_compat: | |
1882 | lazyancestors = rustancestor.LazyAncestors |
|
1882 | lazyancestors = rustancestor.LazyAncestors | |
1883 | arg = self.index |
|
1883 | arg = self.index | |
1884 | else: |
|
1884 | else: | |
1885 | lazyancestors = ancestor.lazyancestors |
|
1885 | lazyancestors = ancestor.lazyancestors | |
1886 | arg = self._uncheckedparentrevs |
|
1886 | arg = self._uncheckedparentrevs | |
1887 | return lazyancestors(arg, revs, stoprev=stoprev, inclusive=inclusive) |
|
1887 | return lazyancestors(arg, revs, stoprev=stoprev, inclusive=inclusive) | |
1888 |
|
1888 | |||
1889 | def descendants(self, revs): |
|
1889 | def descendants(self, revs): | |
1890 | return dagop.descendantrevs(revs, self.revs, self.parentrevs) |
|
1890 | return dagop.descendantrevs(revs, self.revs, self.parentrevs) | |
1891 |
|
1891 | |||
1892 | def findcommonmissing(self, common=None, heads=None): |
|
1892 | def findcommonmissing(self, common=None, heads=None): | |
1893 | """Return a tuple of the ancestors of common and the ancestors of heads |
|
1893 | """Return a tuple of the ancestors of common and the ancestors of heads | |
1894 | that are not ancestors of common. In revset terminology, we return the |
|
1894 | that are not ancestors of common. In revset terminology, we return the | |
1895 | tuple: |
|
1895 | tuple: | |
1896 |
|
1896 | |||
1897 | ::common, (::heads) - (::common) |
|
1897 | ::common, (::heads) - (::common) | |
1898 |
|
1898 | |||
1899 | The list is sorted by revision number, meaning it is |
|
1899 | The list is sorted by revision number, meaning it is | |
1900 | topologically sorted. |
|
1900 | topologically sorted. | |
1901 |
|
1901 | |||
1902 | 'heads' and 'common' are both lists of node IDs. If heads is |
|
1902 | 'heads' and 'common' are both lists of node IDs. If heads is | |
1903 | not supplied, uses all of the revlog's heads. If common is not |
|
1903 | not supplied, uses all of the revlog's heads. If common is not | |
1904 | supplied, uses nullid.""" |
|
1904 | supplied, uses nullid.""" | |
1905 | if common is None: |
|
1905 | if common is None: | |
1906 | common = [self.nullid] |
|
1906 | common = [self.nullid] | |
1907 | if heads is None: |
|
1907 | if heads is None: | |
1908 | heads = self.heads() |
|
1908 | heads = self.heads() | |
1909 |
|
1909 | |||
1910 | common = [self.rev(n) for n in common] |
|
1910 | common = [self.rev(n) for n in common] | |
1911 | heads = [self.rev(n) for n in heads] |
|
1911 | heads = [self.rev(n) for n in heads] | |
1912 |
|
1912 | |||
1913 | # we want the ancestors, but inclusive |
|
1913 | # we want the ancestors, but inclusive | |
1914 | class lazyset: |
|
1914 | class lazyset: | |
1915 | def __init__(self, lazyvalues): |
|
1915 | def __init__(self, lazyvalues): | |
1916 | self.addedvalues = set() |
|
1916 | self.addedvalues = set() | |
1917 | self.lazyvalues = lazyvalues |
|
1917 | self.lazyvalues = lazyvalues | |
1918 |
|
1918 | |||
1919 | def __contains__(self, value): |
|
1919 | def __contains__(self, value): | |
1920 | return value in self.addedvalues or value in self.lazyvalues |
|
1920 | return value in self.addedvalues or value in self.lazyvalues | |
1921 |
|
1921 | |||
1922 | def __iter__(self): |
|
1922 | def __iter__(self): | |
1923 | added = self.addedvalues |
|
1923 | added = self.addedvalues | |
1924 | for r in added: |
|
1924 | for r in added: | |
1925 | yield r |
|
1925 | yield r | |
1926 | for r in self.lazyvalues: |
|
1926 | for r in self.lazyvalues: | |
1927 | if not r in added: |
|
1927 | if not r in added: | |
1928 | yield r |
|
1928 | yield r | |
1929 |
|
1929 | |||
1930 | def add(self, value): |
|
1930 | def add(self, value): | |
1931 | self.addedvalues.add(value) |
|
1931 | self.addedvalues.add(value) | |
1932 |
|
1932 | |||
1933 | def update(self, values): |
|
1933 | def update(self, values): | |
1934 | self.addedvalues.update(values) |
|
1934 | self.addedvalues.update(values) | |
1935 |
|
1935 | |||
1936 | has = lazyset(self.ancestors(common)) |
|
1936 | has = lazyset(self.ancestors(common)) | |
1937 | has.add(nullrev) |
|
1937 | has.add(nullrev) | |
1938 | has.update(common) |
|
1938 | has.update(common) | |
1939 |
|
1939 | |||
1940 | # take all ancestors from heads that aren't in has |
|
1940 | # take all ancestors from heads that aren't in has | |
1941 | missing = set() |
|
1941 | missing = set() | |
1942 | visit = collections.deque(r for r in heads if r not in has) |
|
1942 | visit = collections.deque(r for r in heads if r not in has) | |
1943 | while visit: |
|
1943 | while visit: | |
1944 | r = visit.popleft() |
|
1944 | r = visit.popleft() | |
1945 | if r in missing: |
|
1945 | if r in missing: | |
1946 | continue |
|
1946 | continue | |
1947 | else: |
|
1947 | else: | |
1948 | missing.add(r) |
|
1948 | missing.add(r) | |
1949 | for p in self.parentrevs(r): |
|
1949 | for p in self.parentrevs(r): | |
1950 | if p not in has: |
|
1950 | if p not in has: | |
1951 | visit.append(p) |
|
1951 | visit.append(p) | |
1952 | missing = list(missing) |
|
1952 | missing = list(missing) | |
1953 | missing.sort() |
|
1953 | missing.sort() | |
1954 | return has, [self.node(miss) for miss in missing] |
|
1954 | return has, [self.node(miss) for miss in missing] | |
1955 |
|
1955 | |||
1956 | def incrementalmissingrevs(self, common=None): |
|
1956 | def incrementalmissingrevs(self, common=None): | |
1957 | """Return an object that can be used to incrementally compute the |
|
1957 | """Return an object that can be used to incrementally compute the | |
1958 | revision numbers of the ancestors of arbitrary sets that are not |
|
1958 | revision numbers of the ancestors of arbitrary sets that are not | |
1959 | ancestors of common. This is an ancestor.incrementalmissingancestors |
|
1959 | ancestors of common. This is an ancestor.incrementalmissingancestors | |
1960 | object. |
|
1960 | object. | |
1961 |
|
1961 | |||
1962 | 'common' is a list of revision numbers. If common is not supplied, uses |
|
1962 | 'common' is a list of revision numbers. If common is not supplied, uses | |
1963 | nullrev. |
|
1963 | nullrev. | |
1964 | """ |
|
1964 | """ | |
1965 | if common is None: |
|
1965 | if common is None: | |
1966 | common = [nullrev] |
|
1966 | common = [nullrev] | |
1967 |
|
1967 | |||
1968 | if rustancestor is not None and self.index.rust_ext_compat: |
|
1968 | if rustancestor is not None and self.index.rust_ext_compat: | |
1969 | return rustancestor.MissingAncestors(self.index, common) |
|
1969 | return rustancestor.MissingAncestors(self.index, common) | |
1970 | return ancestor.incrementalmissingancestors(self.parentrevs, common) |
|
1970 | return ancestor.incrementalmissingancestors(self.parentrevs, common) | |
1971 |
|
1971 | |||
1972 | def findmissingrevs(self, common=None, heads=None): |
|
1972 | def findmissingrevs(self, common=None, heads=None): | |
1973 | """Return the revision numbers of the ancestors of heads that |
|
1973 | """Return the revision numbers of the ancestors of heads that | |
1974 | are not ancestors of common. |
|
1974 | are not ancestors of common. | |
1975 |
|
1975 | |||
1976 | More specifically, return a list of revision numbers corresponding to |
|
1976 | More specifically, return a list of revision numbers corresponding to | |
1977 | nodes N such that every N satisfies the following constraints: |
|
1977 | nodes N such that every N satisfies the following constraints: | |
1978 |
|
1978 | |||
1979 | 1. N is an ancestor of some node in 'heads' |
|
1979 | 1. N is an ancestor of some node in 'heads' | |
1980 | 2. N is not an ancestor of any node in 'common' |
|
1980 | 2. N is not an ancestor of any node in 'common' | |
1981 |
|
1981 | |||
1982 | The list is sorted by revision number, meaning it is |
|
1982 | The list is sorted by revision number, meaning it is | |
1983 | topologically sorted. |
|
1983 | topologically sorted. | |
1984 |
|
1984 | |||
1985 | 'heads' and 'common' are both lists of revision numbers. If heads is |
|
1985 | 'heads' and 'common' are both lists of revision numbers. If heads is | |
1986 | not supplied, uses all of the revlog's heads. If common is not |
|
1986 | not supplied, uses all of the revlog's heads. If common is not | |
1987 | supplied, uses nullid.""" |
|
1987 | supplied, uses nullid.""" | |
1988 | if common is None: |
|
1988 | if common is None: | |
1989 | common = [nullrev] |
|
1989 | common = [nullrev] | |
1990 | if heads is None: |
|
1990 | if heads is None: | |
1991 | heads = self.headrevs() |
|
1991 | heads = self.headrevs() | |
1992 |
|
1992 | |||
1993 | inc = self.incrementalmissingrevs(common=common) |
|
1993 | inc = self.incrementalmissingrevs(common=common) | |
1994 | return inc.missingancestors(heads) |
|
1994 | return inc.missingancestors(heads) | |
1995 |
|
1995 | |||
1996 | def findmissing(self, common=None, heads=None): |
|
1996 | def findmissing(self, common=None, heads=None): | |
1997 | """Return the ancestors of heads that are not ancestors of common. |
|
1997 | """Return the ancestors of heads that are not ancestors of common. | |
1998 |
|
1998 | |||
1999 | More specifically, return a list of nodes N such that every N |
|
1999 | More specifically, return a list of nodes N such that every N | |
2000 | satisfies the following constraints: |
|
2000 | satisfies the following constraints: | |
2001 |
|
2001 | |||
2002 | 1. N is an ancestor of some node in 'heads' |
|
2002 | 1. N is an ancestor of some node in 'heads' | |
2003 | 2. N is not an ancestor of any node in 'common' |
|
2003 | 2. N is not an ancestor of any node in 'common' | |
2004 |
|
2004 | |||
2005 | The list is sorted by revision number, meaning it is |
|
2005 | The list is sorted by revision number, meaning it is | |
2006 | topologically sorted. |
|
2006 | topologically sorted. | |
2007 |
|
2007 | |||
2008 | 'heads' and 'common' are both lists of node IDs. If heads is |
|
2008 | 'heads' and 'common' are both lists of node IDs. If heads is | |
2009 | not supplied, uses all of the revlog's heads. If common is not |
|
2009 | not supplied, uses all of the revlog's heads. If common is not | |
2010 | supplied, uses nullid.""" |
|
2010 | supplied, uses nullid.""" | |
2011 | if common is None: |
|
2011 | if common is None: | |
2012 | common = [self.nullid] |
|
2012 | common = [self.nullid] | |
2013 | if heads is None: |
|
2013 | if heads is None: | |
2014 | heads = self.heads() |
|
2014 | heads = self.heads() | |
2015 |
|
2015 | |||
2016 | common = [self.rev(n) for n in common] |
|
2016 | common = [self.rev(n) for n in common] | |
2017 | heads = [self.rev(n) for n in heads] |
|
2017 | heads = [self.rev(n) for n in heads] | |
2018 |
|
2018 | |||
2019 | inc = self.incrementalmissingrevs(common=common) |
|
2019 | inc = self.incrementalmissingrevs(common=common) | |
2020 | return [self.node(r) for r in inc.missingancestors(heads)] |
|
2020 | return [self.node(r) for r in inc.missingancestors(heads)] | |
2021 |
|
2021 | |||
2022 | def nodesbetween(self, roots=None, heads=None): |
|
2022 | def nodesbetween(self, roots=None, heads=None): | |
2023 | """Return a topological path from 'roots' to 'heads'. |
|
2023 | """Return a topological path from 'roots' to 'heads'. | |
2024 |
|
2024 | |||
2025 | Return a tuple (nodes, outroots, outheads) where 'nodes' is a |
|
2025 | Return a tuple (nodes, outroots, outheads) where 'nodes' is a | |
2026 | topologically sorted list of all nodes N that satisfy both of |
|
2026 | topologically sorted list of all nodes N that satisfy both of | |
2027 | these constraints: |
|
2027 | these constraints: | |
2028 |
|
2028 | |||
2029 | 1. N is a descendant of some node in 'roots' |
|
2029 | 1. N is a descendant of some node in 'roots' | |
2030 | 2. N is an ancestor of some node in 'heads' |
|
2030 | 2. N is an ancestor of some node in 'heads' | |
2031 |
|
2031 | |||
2032 | Every node is considered to be both a descendant and an ancestor |
|
2032 | Every node is considered to be both a descendant and an ancestor | |
2033 | of itself, so every reachable node in 'roots' and 'heads' will be |
|
2033 | of itself, so every reachable node in 'roots' and 'heads' will be | |
2034 | included in 'nodes'. |
|
2034 | included in 'nodes'. | |
2035 |
|
2035 | |||
2036 | 'outroots' is the list of reachable nodes in 'roots', i.e., the |
|
2036 | 'outroots' is the list of reachable nodes in 'roots', i.e., the | |
2037 | subset of 'roots' that is returned in 'nodes'. Likewise, |
|
2037 | subset of 'roots' that is returned in 'nodes'. Likewise, | |
2038 | 'outheads' is the subset of 'heads' that is also in 'nodes'. |
|
2038 | 'outheads' is the subset of 'heads' that is also in 'nodes'. | |
2039 |
|
2039 | |||
2040 | 'roots' and 'heads' are both lists of node IDs. If 'roots' is |
|
2040 | 'roots' and 'heads' are both lists of node IDs. If 'roots' is | |
2041 | unspecified, uses nullid as the only root. If 'heads' is |
|
2041 | unspecified, uses nullid as the only root. If 'heads' is | |
2042 | unspecified, uses list of all of the revlog's heads.""" |
|
2042 | unspecified, uses list of all of the revlog's heads.""" | |
2043 | nonodes = ([], [], []) |
|
2043 | nonodes = ([], [], []) | |
2044 | if roots is not None: |
|
2044 | if roots is not None: | |
2045 | roots = list(roots) |
|
2045 | roots = list(roots) | |
2046 | if not roots: |
|
2046 | if not roots: | |
2047 | return nonodes |
|
2047 | return nonodes | |
2048 | lowestrev = min([self.rev(n) for n in roots]) |
|
2048 | lowestrev = min([self.rev(n) for n in roots]) | |
2049 | else: |
|
2049 | else: | |
2050 | roots = [self.nullid] # Everybody's a descendant of nullid |
|
2050 | roots = [self.nullid] # Everybody's a descendant of nullid | |
2051 | lowestrev = nullrev |
|
2051 | lowestrev = nullrev | |
2052 | if (lowestrev == nullrev) and (heads is None): |
|
2052 | if (lowestrev == nullrev) and (heads is None): | |
2053 | # We want _all_ the nodes! |
|
2053 | # We want _all_ the nodes! | |
2054 | return ( |
|
2054 | return ( | |
2055 | [self.node(r) for r in self], |
|
2055 | [self.node(r) for r in self], | |
2056 | [self.nullid], |
|
2056 | [self.nullid], | |
2057 | list(self.heads()), |
|
2057 | list(self.heads()), | |
2058 | ) |
|
2058 | ) | |
2059 | if heads is None: |
|
2059 | if heads is None: | |
2060 | # All nodes are ancestors, so the latest ancestor is the last |
|
2060 | # All nodes are ancestors, so the latest ancestor is the last | |
2061 | # node. |
|
2061 | # node. | |
2062 | highestrev = len(self) - 1 |
|
2062 | highestrev = len(self) - 1 | |
2063 | # Set ancestors to None to signal that every node is an ancestor. |
|
2063 | # Set ancestors to None to signal that every node is an ancestor. | |
2064 | ancestors = None |
|
2064 | ancestors = None | |
2065 | # Set heads to an empty dictionary for later discovery of heads |
|
2065 | # Set heads to an empty dictionary for later discovery of heads | |
2066 | heads = {} |
|
2066 | heads = {} | |
2067 | else: |
|
2067 | else: | |
2068 | heads = list(heads) |
|
2068 | heads = list(heads) | |
2069 | if not heads: |
|
2069 | if not heads: | |
2070 | return nonodes |
|
2070 | return nonodes | |
2071 | ancestors = set() |
|
2071 | ancestors = set() | |
2072 | # Turn heads into a dictionary so we can remove 'fake' heads. |
|
2072 | # Turn heads into a dictionary so we can remove 'fake' heads. | |
2073 | # Also, later we will be using it to filter out the heads we can't |
|
2073 | # Also, later we will be using it to filter out the heads we can't | |
2074 | # find from roots. |
|
2074 | # find from roots. | |
2075 | heads = dict.fromkeys(heads, False) |
|
2075 | heads = dict.fromkeys(heads, False) | |
2076 | # Start at the top and keep marking parents until we're done. |
|
2076 | # Start at the top and keep marking parents until we're done. | |
2077 | nodestotag = set(heads) |
|
2077 | nodestotag = set(heads) | |
2078 | # Remember where the top was so we can use it as a limit later. |
|
2078 | # Remember where the top was so we can use it as a limit later. | |
2079 | highestrev = max([self.rev(n) for n in nodestotag]) |
|
2079 | highestrev = max([self.rev(n) for n in nodestotag]) | |
2080 | while nodestotag: |
|
2080 | while nodestotag: | |
2081 | # grab a node to tag |
|
2081 | # grab a node to tag | |
2082 | n = nodestotag.pop() |
|
2082 | n = nodestotag.pop() | |
2083 | # Never tag nullid |
|
2083 | # Never tag nullid | |
2084 | if n == self.nullid: |
|
2084 | if n == self.nullid: | |
2085 | continue |
|
2085 | continue | |
2086 | # A node's revision number represents its place in a |
|
2086 | # A node's revision number represents its place in a | |
2087 | # topologically sorted list of nodes. |
|
2087 | # topologically sorted list of nodes. | |
2088 | r = self.rev(n) |
|
2088 | r = self.rev(n) | |
2089 | if r >= lowestrev: |
|
2089 | if r >= lowestrev: | |
2090 | if n not in ancestors: |
|
2090 | if n not in ancestors: | |
2091 | # If we are possibly a descendant of one of the roots |
|
2091 | # If we are possibly a descendant of one of the roots | |
2092 | # and we haven't already been marked as an ancestor |
|
2092 | # and we haven't already been marked as an ancestor | |
2093 | ancestors.add(n) # Mark as ancestor |
|
2093 | ancestors.add(n) # Mark as ancestor | |
2094 | # Add non-nullid parents to list of nodes to tag. |
|
2094 | # Add non-nullid parents to list of nodes to tag. | |
2095 | nodestotag.update( |
|
2095 | nodestotag.update( | |
2096 | [p for p in self.parents(n) if p != self.nullid] |
|
2096 | [p for p in self.parents(n) if p != self.nullid] | |
2097 | ) |
|
2097 | ) | |
2098 | elif n in heads: # We've seen it before, is it a fake head? |
|
2098 | elif n in heads: # We've seen it before, is it a fake head? | |
2099 | # So it is, real heads should not be the ancestors of |
|
2099 | # So it is, real heads should not be the ancestors of | |
2100 | # any other heads. |
|
2100 | # any other heads. | |
2101 | heads.pop(n) |
|
2101 | heads.pop(n) | |
2102 | if not ancestors: |
|
2102 | if not ancestors: | |
2103 | return nonodes |
|
2103 | return nonodes | |
2104 | # Now that we have our set of ancestors, we want to remove any |
|
2104 | # Now that we have our set of ancestors, we want to remove any | |
2105 | # roots that are not ancestors. |
|
2105 | # roots that are not ancestors. | |
2106 |
|
2106 | |||
2107 | # If one of the roots was nullid, everything is included anyway. |
|
2107 | # If one of the roots was nullid, everything is included anyway. | |
2108 | if lowestrev > nullrev: |
|
2108 | if lowestrev > nullrev: | |
2109 | # But, since we weren't, let's recompute the lowest rev to not |
|
2109 | # But, since we weren't, let's recompute the lowest rev to not | |
2110 | # include roots that aren't ancestors. |
|
2110 | # include roots that aren't ancestors. | |
2111 |
|
2111 | |||
2112 | # Filter out roots that aren't ancestors of heads |
|
2112 | # Filter out roots that aren't ancestors of heads | |
2113 | roots = [root for root in roots if root in ancestors] |
|
2113 | roots = [root for root in roots if root in ancestors] | |
2114 | # Recompute the lowest revision |
|
2114 | # Recompute the lowest revision | |
2115 | if roots: |
|
2115 | if roots: | |
2116 | lowestrev = min([self.rev(root) for root in roots]) |
|
2116 | lowestrev = min([self.rev(root) for root in roots]) | |
2117 | else: |
|
2117 | else: | |
2118 | # No more roots? Return empty list |
|
2118 | # No more roots? Return empty list | |
2119 | return nonodes |
|
2119 | return nonodes | |
2120 | else: |
|
2120 | else: | |
2121 | # We are descending from nullid, and don't need to care about |
|
2121 | # We are descending from nullid, and don't need to care about | |
2122 | # any other roots. |
|
2122 | # any other roots. | |
2123 | lowestrev = nullrev |
|
2123 | lowestrev = nullrev | |
2124 | roots = [self.nullid] |
|
2124 | roots = [self.nullid] | |
2125 | # Transform our roots list into a set. |
|
2125 | # Transform our roots list into a set. | |
2126 | descendants = set(roots) |
|
2126 | descendants = set(roots) | |
2127 | # Also, keep the original roots so we can filter out roots that aren't |
|
2127 | # Also, keep the original roots so we can filter out roots that aren't | |
2128 | # 'real' roots (i.e. are descended from other roots). |
|
2128 | # 'real' roots (i.e. are descended from other roots). | |
2129 | roots = descendants.copy() |
|
2129 | roots = descendants.copy() | |
2130 | # Our topologically sorted list of output nodes. |
|
2130 | # Our topologically sorted list of output nodes. | |
2131 | orderedout = [] |
|
2131 | orderedout = [] | |
2132 | # Don't start at nullid since we don't want nullid in our output list, |
|
2132 | # Don't start at nullid since we don't want nullid in our output list, | |
2133 | # and if nullid shows up in descendants, empty parents will look like |
|
2133 | # and if nullid shows up in descendants, empty parents will look like | |
2134 | # they're descendants. |
|
2134 | # they're descendants. | |
2135 | for r in self.revs(start=max(lowestrev, 0), stop=highestrev + 1): |
|
2135 | for r in self.revs(start=max(lowestrev, 0), stop=highestrev + 1): | |
2136 | n = self.node(r) |
|
2136 | n = self.node(r) | |
2137 | isdescendant = False |
|
2137 | isdescendant = False | |
2138 | if lowestrev == nullrev: # Everybody is a descendant of nullid |
|
2138 | if lowestrev == nullrev: # Everybody is a descendant of nullid | |
2139 | isdescendant = True |
|
2139 | isdescendant = True | |
2140 | elif n in descendants: |
|
2140 | elif n in descendants: | |
2141 | # n is already a descendant |
|
2141 | # n is already a descendant | |
2142 | isdescendant = True |
|
2142 | isdescendant = True | |
2143 | # This check only needs to be done here because all the roots |
|
2143 | # This check only needs to be done here because all the roots | |
2144 | # will start being marked is descendants before the loop. |
|
2144 | # will start being marked is descendants before the loop. | |
2145 | if n in roots: |
|
2145 | if n in roots: | |
2146 | # If n was a root, check if it's a 'real' root. |
|
2146 | # If n was a root, check if it's a 'real' root. | |
2147 | p = tuple(self.parents(n)) |
|
2147 | p = tuple(self.parents(n)) | |
2148 | # If any of its parents are descendants, it's not a root. |
|
2148 | # If any of its parents are descendants, it's not a root. | |
2149 | if (p[0] in descendants) or (p[1] in descendants): |
|
2149 | if (p[0] in descendants) or (p[1] in descendants): | |
2150 | roots.remove(n) |
|
2150 | roots.remove(n) | |
2151 | else: |
|
2151 | else: | |
2152 | p = tuple(self.parents(n)) |
|
2152 | p = tuple(self.parents(n)) | |
2153 | # A node is a descendant if either of its parents are |
|
2153 | # A node is a descendant if either of its parents are | |
2154 | # descendants. (We seeded the dependents list with the roots |
|
2154 | # descendants. (We seeded the dependents list with the roots | |
2155 | # up there, remember?) |
|
2155 | # up there, remember?) | |
2156 | if (p[0] in descendants) or (p[1] in descendants): |
|
2156 | if (p[0] in descendants) or (p[1] in descendants): | |
2157 | descendants.add(n) |
|
2157 | descendants.add(n) | |
2158 | isdescendant = True |
|
2158 | isdescendant = True | |
2159 | if isdescendant and ((ancestors is None) or (n in ancestors)): |
|
2159 | if isdescendant and ((ancestors is None) or (n in ancestors)): | |
2160 | # Only include nodes that are both descendants and ancestors. |
|
2160 | # Only include nodes that are both descendants and ancestors. | |
2161 | orderedout.append(n) |
|
2161 | orderedout.append(n) | |
2162 | if (ancestors is not None) and (n in heads): |
|
2162 | if (ancestors is not None) and (n in heads): | |
2163 | # We're trying to figure out which heads are reachable |
|
2163 | # We're trying to figure out which heads are reachable | |
2164 | # from roots. |
|
2164 | # from roots. | |
2165 | # Mark this head as having been reached |
|
2165 | # Mark this head as having been reached | |
2166 | heads[n] = True |
|
2166 | heads[n] = True | |
2167 | elif ancestors is None: |
|
2167 | elif ancestors is None: | |
2168 | # Otherwise, we're trying to discover the heads. |
|
2168 | # Otherwise, we're trying to discover the heads. | |
2169 | # Assume this is a head because if it isn't, the next step |
|
2169 | # Assume this is a head because if it isn't, the next step | |
2170 | # will eventually remove it. |
|
2170 | # will eventually remove it. | |
2171 | heads[n] = True |
|
2171 | heads[n] = True | |
2172 | # But, obviously its parents aren't. |
|
2172 | # But, obviously its parents aren't. | |
2173 | for p in self.parents(n): |
|
2173 | for p in self.parents(n): | |
2174 | heads.pop(p, None) |
|
2174 | heads.pop(p, None) | |
2175 | heads = [head for head, flag in heads.items() if flag] |
|
2175 | heads = [head for head, flag in heads.items() if flag] | |
2176 | roots = list(roots) |
|
2176 | roots = list(roots) | |
2177 | assert orderedout |
|
2177 | assert orderedout | |
2178 | assert roots |
|
2178 | assert roots | |
2179 | assert heads |
|
2179 | assert heads | |
2180 | return (orderedout, roots, heads) |
|
2180 | return (orderedout, roots, heads) | |
2181 |
|
2181 | |||
2182 | def headrevs(self, revs=None): |
|
2182 | def headrevs(self, revs=None): | |
2183 | if revs is None: |
|
2183 | if revs is None: | |
2184 | try: |
|
2184 | try: | |
2185 | return self.index.headrevs() |
|
2185 | return self.index.headrevs() | |
2186 | except AttributeError: |
|
2186 | except AttributeError: | |
2187 | return self._headrevs() |
|
2187 | return self._headrevs() | |
2188 | if rustdagop is not None and self.index.rust_ext_compat: |
|
2188 | if rustdagop is not None and self.index.rust_ext_compat: | |
2189 | return rustdagop.headrevs(self.index, revs) |
|
2189 | return rustdagop.headrevs(self.index, revs) | |
2190 | return dagop.headrevs(revs, self._uncheckedparentrevs) |
|
2190 | return dagop.headrevs(revs, self._uncheckedparentrevs) | |
2191 |
|
2191 | |||
2192 | def computephases(self, roots): |
|
2192 | def computephases(self, roots): | |
2193 | return self.index.computephasesmapsets(roots) |
|
2193 | return self.index.computephasesmapsets(roots) | |
2194 |
|
2194 | |||
2195 | def _headrevs(self): |
|
2195 | def _headrevs(self): | |
2196 | count = len(self) |
|
2196 | count = len(self) | |
2197 | if not count: |
|
2197 | if not count: | |
2198 | return [nullrev] |
|
2198 | return [nullrev] | |
2199 | # we won't iter over filtered rev so nobody is a head at start |
|
2199 | # we won't iter over filtered rev so nobody is a head at start | |
2200 | ishead = [0] * (count + 1) |
|
2200 | ishead = [0] * (count + 1) | |
2201 | index = self.index |
|
2201 | index = self.index | |
2202 | for r in self: |
|
2202 | for r in self: | |
2203 | ishead[r] = 1 # I may be an head |
|
2203 | ishead[r] = 1 # I may be an head | |
2204 | e = index[r] |
|
2204 | e = index[r] | |
2205 | ishead[e[5]] = ishead[e[6]] = 0 # my parent are not |
|
2205 | ishead[e[5]] = ishead[e[6]] = 0 # my parent are not | |
2206 | return [r for r, val in enumerate(ishead) if val] |
|
2206 | return [r for r, val in enumerate(ishead) if val] | |
2207 |
|
2207 | |||
2208 | def heads(self, start=None, stop=None): |
|
2208 | def heads(self, start=None, stop=None): | |
2209 | """return the list of all nodes that have no children |
|
2209 | """return the list of all nodes that have no children | |
2210 |
|
2210 | |||
2211 | if start is specified, only heads that are descendants of |
|
2211 | if start is specified, only heads that are descendants of | |
2212 | start will be returned |
|
2212 | start will be returned | |
2213 | if stop is specified, it will consider all the revs from stop |
|
2213 | if stop is specified, it will consider all the revs from stop | |
2214 | as if they had no children |
|
2214 | as if they had no children | |
2215 | """ |
|
2215 | """ | |
2216 | if start is None and stop is None: |
|
2216 | if start is None and stop is None: | |
2217 | if not len(self): |
|
2217 | if not len(self): | |
2218 | return [self.nullid] |
|
2218 | return [self.nullid] | |
2219 | return [self.node(r) for r in self.headrevs()] |
|
2219 | return [self.node(r) for r in self.headrevs()] | |
2220 |
|
2220 | |||
2221 | if start is None: |
|
2221 | if start is None: | |
2222 | start = nullrev |
|
2222 | start = nullrev | |
2223 | else: |
|
2223 | else: | |
2224 | start = self.rev(start) |
|
2224 | start = self.rev(start) | |
2225 |
|
2225 | |||
2226 | stoprevs = {self.rev(n) for n in stop or []} |
|
2226 | stoprevs = {self.rev(n) for n in stop or []} | |
2227 |
|
2227 | |||
2228 | revs = dagop.headrevssubset( |
|
2228 | revs = dagop.headrevssubset( | |
2229 | self.revs, self.parentrevs, startrev=start, stoprevs=stoprevs |
|
2229 | self.revs, self.parentrevs, startrev=start, stoprevs=stoprevs | |
2230 | ) |
|
2230 | ) | |
2231 |
|
2231 | |||
2232 | return [self.node(rev) for rev in revs] |
|
2232 | return [self.node(rev) for rev in revs] | |
2233 |
|
2233 | |||
2234 | def children(self, node): |
|
2234 | def children(self, node): | |
2235 | """find the children of a given node""" |
|
2235 | """find the children of a given node""" | |
2236 | c = [] |
|
2236 | c = [] | |
2237 | p = self.rev(node) |
|
2237 | p = self.rev(node) | |
2238 | for r in self.revs(start=p + 1): |
|
2238 | for r in self.revs(start=p + 1): | |
2239 | prevs = [pr for pr in self.parentrevs(r) if pr != nullrev] |
|
2239 | prevs = [pr for pr in self.parentrevs(r) if pr != nullrev] | |
2240 | if prevs: |
|
2240 | if prevs: | |
2241 | for pr in prevs: |
|
2241 | for pr in prevs: | |
2242 | if pr == p: |
|
2242 | if pr == p: | |
2243 | c.append(self.node(r)) |
|
2243 | c.append(self.node(r)) | |
2244 | elif p == nullrev: |
|
2244 | elif p == nullrev: | |
2245 | c.append(self.node(r)) |
|
2245 | c.append(self.node(r)) | |
2246 | return c |
|
2246 | return c | |
2247 |
|
2247 | |||
2248 | def commonancestorsheads(self, a, b): |
|
2248 | def commonancestorsheads(self, a, b): | |
2249 | """calculate all the heads of the common ancestors of nodes a and b""" |
|
2249 | """calculate all the heads of the common ancestors of nodes a and b""" | |
2250 | a, b = self.rev(a), self.rev(b) |
|
2250 | a, b = self.rev(a), self.rev(b) | |
2251 | ancs = self._commonancestorsheads(a, b) |
|
2251 | ancs = self._commonancestorsheads(a, b) | |
2252 | return pycompat.maplist(self.node, ancs) |
|
2252 | return pycompat.maplist(self.node, ancs) | |
2253 |
|
2253 | |||
2254 | def _commonancestorsheads(self, *revs): |
|
2254 | def _commonancestorsheads(self, *revs): | |
2255 | """calculate all the heads of the common ancestors of revs""" |
|
2255 | """calculate all the heads of the common ancestors of revs""" | |
2256 | try: |
|
2256 | try: | |
2257 | ancs = self.index.commonancestorsheads(*revs) |
|
2257 | ancs = self.index.commonancestorsheads(*revs) | |
2258 | except (AttributeError, OverflowError): # C implementation failed |
|
2258 | except (AttributeError, OverflowError): # C implementation failed | |
2259 | ancs = ancestor.commonancestorsheads(self.parentrevs, *revs) |
|
2259 | ancs = ancestor.commonancestorsheads(self.parentrevs, *revs) | |
2260 | return ancs |
|
2260 | return ancs | |
2261 |
|
2261 | |||
2262 | def isancestor(self, a, b): |
|
2262 | def isancestor(self, a, b): | |
2263 | """return True if node a is an ancestor of node b |
|
2263 | """return True if node a is an ancestor of node b | |
2264 |
|
2264 | |||
2265 | A revision is considered an ancestor of itself.""" |
|
2265 | A revision is considered an ancestor of itself.""" | |
2266 | a, b = self.rev(a), self.rev(b) |
|
2266 | a, b = self.rev(a), self.rev(b) | |
2267 | return self.isancestorrev(a, b) |
|
2267 | return self.isancestorrev(a, b) | |
2268 |
|
2268 | |||
2269 | def isancestorrev(self, a, b): |
|
2269 | def isancestorrev(self, a, b): | |
2270 | """return True if revision a is an ancestor of revision b |
|
2270 | """return True if revision a is an ancestor of revision b | |
2271 |
|
2271 | |||
2272 | A revision is considered an ancestor of itself. |
|
2272 | A revision is considered an ancestor of itself. | |
2273 |
|
2273 | |||
2274 | The implementation of this is trivial but the use of |
|
2274 | The implementation of this is trivial but the use of | |
2275 | reachableroots is not.""" |
|
2275 | reachableroots is not.""" | |
2276 | if a == nullrev: |
|
2276 | if a == nullrev: | |
2277 | return True |
|
2277 | return True | |
2278 | elif a == b: |
|
2278 | elif a == b: | |
2279 | return True |
|
2279 | return True | |
2280 | elif a > b: |
|
2280 | elif a > b: | |
2281 | return False |
|
2281 | return False | |
2282 | return bool(self.reachableroots(a, [b], [a], includepath=False)) |
|
2282 | return bool(self.reachableroots(a, [b], [a], includepath=False)) | |
2283 |
|
2283 | |||
2284 | def reachableroots(self, minroot, heads, roots, includepath=False): |
|
2284 | def reachableroots(self, minroot, heads, roots, includepath=False): | |
2285 | """return (heads(::(<roots> and <roots>::<heads>))) |
|
2285 | """return (heads(::(<roots> and <roots>::<heads>))) | |
2286 |
|
2286 | |||
2287 | If includepath is True, return (<roots>::<heads>).""" |
|
2287 | If includepath is True, return (<roots>::<heads>).""" | |
2288 | try: |
|
2288 | try: | |
2289 | return self.index.reachableroots2( |
|
2289 | return self.index.reachableroots2( | |
2290 | minroot, heads, roots, includepath |
|
2290 | minroot, heads, roots, includepath | |
2291 | ) |
|
2291 | ) | |
2292 | except AttributeError: |
|
2292 | except AttributeError: | |
2293 | return dagop._reachablerootspure( |
|
2293 | return dagop._reachablerootspure( | |
2294 | self.parentrevs, minroot, roots, heads, includepath |
|
2294 | self.parentrevs, minroot, roots, heads, includepath | |
2295 | ) |
|
2295 | ) | |
2296 |
|
2296 | |||
2297 | def ancestor(self, a, b): |
|
2297 | def ancestor(self, a, b): | |
2298 | """calculate the "best" common ancestor of nodes a and b""" |
|
2298 | """calculate the "best" common ancestor of nodes a and b""" | |
2299 |
|
2299 | |||
2300 | a, b = self.rev(a), self.rev(b) |
|
2300 | a, b = self.rev(a), self.rev(b) | |
2301 | try: |
|
2301 | try: | |
2302 | ancs = self.index.ancestors(a, b) |
|
2302 | ancs = self.index.ancestors(a, b) | |
2303 | except (AttributeError, OverflowError): |
|
2303 | except (AttributeError, OverflowError): | |
2304 | ancs = ancestor.ancestors(self.parentrevs, a, b) |
|
2304 | ancs = ancestor.ancestors(self.parentrevs, a, b) | |
2305 | if ancs: |
|
2305 | if ancs: | |
2306 | # choose a consistent winner when there's a tie |
|
2306 | # choose a consistent winner when there's a tie | |
2307 | return min(map(self.node, ancs)) |
|
2307 | return min(map(self.node, ancs)) | |
2308 | return self.nullid |
|
2308 | return self.nullid | |
2309 |
|
2309 | |||
2310 | def _match(self, id): |
|
2310 | def _match(self, id): | |
2311 | if isinstance(id, int): |
|
2311 | if isinstance(id, int): | |
2312 | # rev |
|
2312 | # rev | |
2313 | return self.node(id) |
|
2313 | return self.node(id) | |
2314 | if len(id) == self.nodeconstants.nodelen: |
|
2314 | if len(id) == self.nodeconstants.nodelen: | |
2315 | # possibly a binary node |
|
2315 | # possibly a binary node | |
2316 | # odds of a binary node being all hex in ASCII are 1 in 10**25 |
|
2316 | # odds of a binary node being all hex in ASCII are 1 in 10**25 | |
2317 | try: |
|
2317 | try: | |
2318 | node = id |
|
2318 | node = id | |
2319 | self.rev(node) # quick search the index |
|
2319 | self.rev(node) # quick search the index | |
2320 | return node |
|
2320 | return node | |
2321 | except error.LookupError: |
|
2321 | except error.LookupError: | |
2322 | pass # may be partial hex id |
|
2322 | pass # may be partial hex id | |
2323 | try: |
|
2323 | try: | |
2324 | # str(rev) |
|
2324 | # str(rev) | |
2325 | rev = int(id) |
|
2325 | rev = int(id) | |
2326 | if b"%d" % rev != id: |
|
2326 | if b"%d" % rev != id: | |
2327 | raise ValueError |
|
2327 | raise ValueError | |
2328 | if rev < 0: |
|
2328 | if rev < 0: | |
2329 | rev = len(self) + rev |
|
2329 | rev = len(self) + rev | |
2330 | if rev < 0 or rev >= len(self): |
|
2330 | if rev < 0 or rev >= len(self): | |
2331 | raise ValueError |
|
2331 | raise ValueError | |
2332 | return self.node(rev) |
|
2332 | return self.node(rev) | |
2333 | except (ValueError, OverflowError): |
|
2333 | except (ValueError, OverflowError): | |
2334 | pass |
|
2334 | pass | |
2335 | if len(id) == 2 * self.nodeconstants.nodelen: |
|
2335 | if len(id) == 2 * self.nodeconstants.nodelen: | |
2336 | try: |
|
2336 | try: | |
2337 | # a full hex nodeid? |
|
2337 | # a full hex nodeid? | |
2338 | node = bin(id) |
|
2338 | node = bin(id) | |
2339 | self.rev(node) |
|
2339 | self.rev(node) | |
2340 | return node |
|
2340 | return node | |
2341 | except (binascii.Error, error.LookupError): |
|
2341 | except (binascii.Error, error.LookupError): | |
2342 | pass |
|
2342 | pass | |
2343 |
|
2343 | |||
2344 | def _partialmatch(self, id): |
|
2344 | def _partialmatch(self, id): | |
2345 | # we don't care wdirfilenodeids as they should be always full hash |
|
2345 | # we don't care wdirfilenodeids as they should be always full hash | |
2346 | maybewdir = self.nodeconstants.wdirhex.startswith(id) |
|
2346 | maybewdir = self.nodeconstants.wdirhex.startswith(id) | |
2347 | ambiguous = False |
|
2347 | ambiguous = False | |
2348 | try: |
|
2348 | try: | |
2349 | partial = self.index.partialmatch(id) |
|
2349 | partial = self.index.partialmatch(id) | |
2350 | if partial and self.hasnode(partial): |
|
2350 | if partial and self.hasnode(partial): | |
2351 | if maybewdir: |
|
2351 | if maybewdir: | |
2352 | # single 'ff...' match in radix tree, ambiguous with wdir |
|
2352 | # single 'ff...' match in radix tree, ambiguous with wdir | |
2353 | ambiguous = True |
|
2353 | ambiguous = True | |
2354 | else: |
|
2354 | else: | |
2355 | return partial |
|
2355 | return partial | |
2356 | elif maybewdir: |
|
2356 | elif maybewdir: | |
2357 | # no 'ff...' match in radix tree, wdir identified |
|
2357 | # no 'ff...' match in radix tree, wdir identified | |
2358 | raise error.WdirUnsupported |
|
2358 | raise error.WdirUnsupported | |
2359 | else: |
|
2359 | else: | |
2360 | return None |
|
2360 | return None | |
2361 | except error.RevlogError: |
|
2361 | except error.RevlogError: | |
2362 | # parsers.c radix tree lookup gave multiple matches |
|
2362 | # parsers.c radix tree lookup gave multiple matches | |
2363 | # fast path: for unfiltered changelog, radix tree is accurate |
|
2363 | # fast path: for unfiltered changelog, radix tree is accurate | |
2364 | if not getattr(self, 'filteredrevs', None): |
|
2364 | if not getattr(self, 'filteredrevs', None): | |
2365 | ambiguous = True |
|
2365 | ambiguous = True | |
2366 | # fall through to slow path that filters hidden revisions |
|
2366 | # fall through to slow path that filters hidden revisions | |
2367 | except (AttributeError, ValueError): |
|
2367 | except (AttributeError, ValueError): | |
2368 | # we are pure python, or key is not hex |
|
2368 | # we are pure python, or key is not hex | |
2369 | pass |
|
2369 | pass | |
2370 | if ambiguous: |
|
2370 | if ambiguous: | |
2371 | raise error.AmbiguousPrefixLookupError( |
|
2371 | raise error.AmbiguousPrefixLookupError( | |
2372 | id, self.display_id, _(b'ambiguous identifier') |
|
2372 | id, self.display_id, _(b'ambiguous identifier') | |
2373 | ) |
|
2373 | ) | |
2374 |
|
2374 | |||
2375 | if id in self._pcache: |
|
2375 | if id in self._pcache: | |
2376 | return self._pcache[id] |
|
2376 | return self._pcache[id] | |
2377 |
|
2377 | |||
2378 | if len(id) <= 40: |
|
2378 | if len(id) <= 40: | |
2379 | # hex(node)[:...] |
|
2379 | # hex(node)[:...] | |
2380 | l = len(id) // 2 * 2 # grab an even number of digits |
|
2380 | l = len(id) // 2 * 2 # grab an even number of digits | |
2381 | try: |
|
2381 | try: | |
2382 | # we're dropping the last digit, so let's check that it's hex, |
|
2382 | # we're dropping the last digit, so let's check that it's hex, | |
2383 | # to avoid the expensive computation below if it's not |
|
2383 | # to avoid the expensive computation below if it's not | |
2384 | if len(id) % 2 > 0: |
|
2384 | if len(id) % 2 > 0: | |
2385 | if not (id[-1] in hexdigits): |
|
2385 | if not (id[-1] in hexdigits): | |
2386 | return None |
|
2386 | return None | |
2387 | prefix = bin(id[:l]) |
|
2387 | prefix = bin(id[:l]) | |
2388 | except binascii.Error: |
|
2388 | except binascii.Error: | |
2389 | pass |
|
2389 | pass | |
2390 | else: |
|
2390 | else: | |
2391 | nl = [e[7] for e in self.index if e[7].startswith(prefix)] |
|
2391 | nl = [e[7] for e in self.index if e[7].startswith(prefix)] | |
2392 | nl = [ |
|
2392 | nl = [ | |
2393 | n for n in nl if hex(n).startswith(id) and self.hasnode(n) |
|
2393 | n for n in nl if hex(n).startswith(id) and self.hasnode(n) | |
2394 | ] |
|
2394 | ] | |
2395 | if self.nodeconstants.nullhex.startswith(id): |
|
2395 | if self.nodeconstants.nullhex.startswith(id): | |
2396 | nl.append(self.nullid) |
|
2396 | nl.append(self.nullid) | |
2397 | if len(nl) > 0: |
|
2397 | if len(nl) > 0: | |
2398 | if len(nl) == 1 and not maybewdir: |
|
2398 | if len(nl) == 1 and not maybewdir: | |
2399 | self._pcache[id] = nl[0] |
|
2399 | self._pcache[id] = nl[0] | |
2400 | return nl[0] |
|
2400 | return nl[0] | |
2401 | raise error.AmbiguousPrefixLookupError( |
|
2401 | raise error.AmbiguousPrefixLookupError( | |
2402 | id, self.display_id, _(b'ambiguous identifier') |
|
2402 | id, self.display_id, _(b'ambiguous identifier') | |
2403 | ) |
|
2403 | ) | |
2404 | if maybewdir: |
|
2404 | if maybewdir: | |
2405 | raise error.WdirUnsupported |
|
2405 | raise error.WdirUnsupported | |
2406 | return None |
|
2406 | return None | |
2407 |
|
2407 | |||
2408 | def lookup(self, id): |
|
2408 | def lookup(self, id): | |
2409 | """locate a node based on: |
|
2409 | """locate a node based on: | |
2410 | - revision number or str(revision number) |
|
2410 | - revision number or str(revision number) | |
2411 | - nodeid or subset of hex nodeid |
|
2411 | - nodeid or subset of hex nodeid | |
2412 | """ |
|
2412 | """ | |
2413 | n = self._match(id) |
|
2413 | n = self._match(id) | |
2414 | if n is not None: |
|
2414 | if n is not None: | |
2415 | return n |
|
2415 | return n | |
2416 | n = self._partialmatch(id) |
|
2416 | n = self._partialmatch(id) | |
2417 | if n: |
|
2417 | if n: | |
2418 | return n |
|
2418 | return n | |
2419 |
|
2419 | |||
2420 | raise error.LookupError(id, self.display_id, _(b'no match found')) |
|
2420 | raise error.LookupError(id, self.display_id, _(b'no match found')) | |
2421 |
|
2421 | |||
2422 | def shortest(self, node, minlength=1): |
|
2422 | def shortest(self, node, minlength=1): | |
2423 | """Find the shortest unambiguous prefix that matches node.""" |
|
2423 | """Find the shortest unambiguous prefix that matches node.""" | |
2424 |
|
2424 | |||
2425 | def isvalid(prefix): |
|
2425 | def isvalid(prefix): | |
2426 | try: |
|
2426 | try: | |
2427 | matchednode = self._partialmatch(prefix) |
|
2427 | matchednode = self._partialmatch(prefix) | |
2428 | except error.AmbiguousPrefixLookupError: |
|
2428 | except error.AmbiguousPrefixLookupError: | |
2429 | return False |
|
2429 | return False | |
2430 | except error.WdirUnsupported: |
|
2430 | except error.WdirUnsupported: | |
2431 | # single 'ff...' match |
|
2431 | # single 'ff...' match | |
2432 | return True |
|
2432 | return True | |
2433 | if matchednode is None: |
|
2433 | if matchednode is None: | |
2434 | raise error.LookupError(node, self.display_id, _(b'no node')) |
|
2434 | raise error.LookupError(node, self.display_id, _(b'no node')) | |
2435 | return True |
|
2435 | return True | |
2436 |
|
2436 | |||
2437 | def maybewdir(prefix): |
|
2437 | def maybewdir(prefix): | |
2438 | return all(c == b'f' for c in pycompat.iterbytestr(prefix)) |
|
2438 | return all(c == b'f' for c in pycompat.iterbytestr(prefix)) | |
2439 |
|
2439 | |||
2440 | hexnode = hex(node) |
|
2440 | hexnode = hex(node) | |
2441 |
|
2441 | |||
2442 | def disambiguate(hexnode, minlength): |
|
2442 | def disambiguate(hexnode, minlength): | |
2443 | """Disambiguate against wdirid.""" |
|
2443 | """Disambiguate against wdirid.""" | |
2444 | for length in range(minlength, len(hexnode) + 1): |
|
2444 | for length in range(minlength, len(hexnode) + 1): | |
2445 | prefix = hexnode[:length] |
|
2445 | prefix = hexnode[:length] | |
2446 | if not maybewdir(prefix): |
|
2446 | if not maybewdir(prefix): | |
2447 | return prefix |
|
2447 | return prefix | |
2448 |
|
2448 | |||
2449 | if not getattr(self, 'filteredrevs', None): |
|
2449 | if not getattr(self, 'filteredrevs', None): | |
2450 | try: |
|
2450 | try: | |
2451 | length = max(self.index.shortest(node), minlength) |
|
2451 | length = max(self.index.shortest(node), minlength) | |
2452 | return disambiguate(hexnode, length) |
|
2452 | return disambiguate(hexnode, length) | |
2453 | except error.RevlogError: |
|
2453 | except error.RevlogError: | |
2454 | if node != self.nodeconstants.wdirid: |
|
2454 | if node != self.nodeconstants.wdirid: | |
2455 | raise error.LookupError( |
|
2455 | raise error.LookupError( | |
2456 | node, self.display_id, _(b'no node') |
|
2456 | node, self.display_id, _(b'no node') | |
2457 | ) |
|
2457 | ) | |
2458 | except AttributeError: |
|
2458 | except AttributeError: | |
2459 | # Fall through to pure code |
|
2459 | # Fall through to pure code | |
2460 | pass |
|
2460 | pass | |
2461 |
|
2461 | |||
2462 | if node == self.nodeconstants.wdirid: |
|
2462 | if node == self.nodeconstants.wdirid: | |
2463 | for length in range(minlength, len(hexnode) + 1): |
|
2463 | for length in range(minlength, len(hexnode) + 1): | |
2464 | prefix = hexnode[:length] |
|
2464 | prefix = hexnode[:length] | |
2465 | if isvalid(prefix): |
|
2465 | if isvalid(prefix): | |
2466 | return prefix |
|
2466 | return prefix | |
2467 |
|
2467 | |||
2468 | for length in range(minlength, len(hexnode) + 1): |
|
2468 | for length in range(minlength, len(hexnode) + 1): | |
2469 | prefix = hexnode[:length] |
|
2469 | prefix = hexnode[:length] | |
2470 | if isvalid(prefix): |
|
2470 | if isvalid(prefix): | |
2471 | return disambiguate(hexnode, length) |
|
2471 | return disambiguate(hexnode, length) | |
2472 |
|
2472 | |||
2473 | def cmp(self, node, text): |
|
2473 | def cmp(self, node, text): | |
2474 | """compare text with a given file revision |
|
2474 | """compare text with a given file revision | |
2475 |
|
2475 | |||
2476 | returns True if text is different than what is stored. |
|
2476 | returns True if text is different than what is stored. | |
2477 | """ |
|
2477 | """ | |
2478 | p1, p2 = self.parents(node) |
|
2478 | p1, p2 = self.parents(node) | |
2479 | return storageutil.hashrevisionsha1(text, p1, p2) != node |
|
2479 | return storageutil.hashrevisionsha1(text, p1, p2) != node | |
2480 |
|
2480 | |||
2481 | def deltaparent(self, rev): |
|
2481 | def deltaparent(self, rev): | |
2482 | """return deltaparent of the given revision""" |
|
2482 | """return deltaparent of the given revision""" | |
2483 | base = self.index[rev][3] |
|
2483 | base = self.index[rev][3] | |
2484 | if base == rev: |
|
2484 | if base == rev: | |
2485 | return nullrev |
|
2485 | return nullrev | |
2486 | elif self.delta_config.general_delta: |
|
2486 | elif self.delta_config.general_delta: | |
2487 | return base |
|
2487 | return base | |
2488 | else: |
|
2488 | else: | |
2489 | return rev - 1 |
|
2489 | return rev - 1 | |
2490 |
|
2490 | |||
2491 | def issnapshot(self, rev): |
|
2491 | def issnapshot(self, rev): | |
2492 | """tells whether rev is a snapshot""" |
|
2492 | """tells whether rev is a snapshot""" | |
2493 | ret = self._inner.issnapshot(rev) |
|
2493 | ret = self._inner.issnapshot(rev) | |
2494 | self.issnapshot = self._inner.issnapshot |
|
2494 | self.issnapshot = self._inner.issnapshot | |
2495 | return ret |
|
2495 | return ret | |
2496 |
|
2496 | |||
2497 | def snapshotdepth(self, rev): |
|
2497 | def snapshotdepth(self, rev): | |
2498 | """number of snapshot in the chain before this one""" |
|
2498 | """number of snapshot in the chain before this one""" | |
2499 | if not self.issnapshot(rev): |
|
2499 | if not self.issnapshot(rev): | |
2500 | raise error.ProgrammingError(b'revision %d not a snapshot') |
|
2500 | raise error.ProgrammingError(b'revision %d not a snapshot') | |
2501 | return len(self._inner._deltachain(rev)[0]) - 1 |
|
2501 | return len(self._inner._deltachain(rev)[0]) - 1 | |
2502 |
|
2502 | |||
2503 | def revdiff(self, rev1, rev2): |
|
2503 | def revdiff(self, rev1, rev2): | |
2504 | """return or calculate a delta between two revisions |
|
2504 | """return or calculate a delta between two revisions | |
2505 |
|
2505 | |||
2506 | The delta calculated is in binary form and is intended to be written to |
|
2506 | The delta calculated is in binary form and is intended to be written to | |
2507 | revlog data directly. So this function needs raw revision data. |
|
2507 | revlog data directly. So this function needs raw revision data. | |
2508 | """ |
|
2508 | """ | |
2509 | if rev1 != nullrev and self.deltaparent(rev2) == rev1: |
|
2509 | if rev1 != nullrev and self.deltaparent(rev2) == rev1: | |
2510 | return bytes(self._inner._chunk(rev2)) |
|
2510 | return bytes(self._inner._chunk(rev2)) | |
2511 |
|
2511 | |||
2512 | return mdiff.textdiff(self.rawdata(rev1), self.rawdata(rev2)) |
|
2512 | return mdiff.textdiff(self.rawdata(rev1), self.rawdata(rev2)) | |
2513 |
|
2513 | |||
2514 | def revision(self, nodeorrev): |
|
2514 | def revision(self, nodeorrev): | |
2515 | """return an uncompressed revision of a given node or revision |
|
2515 | """return an uncompressed revision of a given node or revision | |
2516 | number. |
|
2516 | number. | |
2517 | """ |
|
2517 | """ | |
2518 | return self._revisiondata(nodeorrev) |
|
2518 | return self._revisiondata(nodeorrev) | |
2519 |
|
2519 | |||
2520 | def sidedata(self, nodeorrev): |
|
2520 | def sidedata(self, nodeorrev): | |
2521 | """a map of extra data related to the changeset but not part of the hash |
|
2521 | """a map of extra data related to the changeset but not part of the hash | |
2522 |
|
2522 | |||
2523 | This function currently return a dictionary. However, more advanced |
|
2523 | This function currently return a dictionary. However, more advanced | |
2524 | mapping object will likely be used in the future for a more |
|
2524 | mapping object will likely be used in the future for a more | |
2525 | efficient/lazy code. |
|
2525 | efficient/lazy code. | |
2526 | """ |
|
2526 | """ | |
2527 | # deal with <nodeorrev> argument type |
|
2527 | # deal with <nodeorrev> argument type | |
2528 | if isinstance(nodeorrev, int): |
|
2528 | if isinstance(nodeorrev, int): | |
2529 | rev = nodeorrev |
|
2529 | rev = nodeorrev | |
2530 | else: |
|
2530 | else: | |
2531 | rev = self.rev(nodeorrev) |
|
2531 | rev = self.rev(nodeorrev) | |
2532 | return self._sidedata(rev) |
|
2532 | return self._sidedata(rev) | |
2533 |
|
2533 | |||
2534 | def _revisiondata(self, nodeorrev, raw=False): |
|
2534 | def _revisiondata(self, nodeorrev, raw=False): | |
2535 | # deal with <nodeorrev> argument type |
|
2535 | # deal with <nodeorrev> argument type | |
2536 | if isinstance(nodeorrev, int): |
|
2536 | if isinstance(nodeorrev, int): | |
2537 | rev = nodeorrev |
|
2537 | rev = nodeorrev | |
2538 | node = self.node(rev) |
|
2538 | node = self.node(rev) | |
2539 | else: |
|
2539 | else: | |
2540 | node = nodeorrev |
|
2540 | node = nodeorrev | |
2541 | rev = None |
|
2541 | rev = None | |
2542 |
|
2542 | |||
2543 | # fast path the special `nullid` rev |
|
2543 | # fast path the special `nullid` rev | |
2544 | if node == self.nullid: |
|
2544 | if node == self.nullid: | |
2545 | return b"" |
|
2545 | return b"" | |
2546 |
|
2546 | |||
2547 | # ``rawtext`` is the text as stored inside the revlog. Might be the |
|
2547 | # ``rawtext`` is the text as stored inside the revlog. Might be the | |
2548 | # revision or might need to be processed to retrieve the revision. |
|
2548 | # revision or might need to be processed to retrieve the revision. | |
2549 | rev, rawtext, validated = self._rawtext(node, rev) |
|
2549 | rev, rawtext, validated = self._rawtext(node, rev) | |
2550 |
|
2550 | |||
2551 | if raw and validated: |
|
2551 | if raw and validated: | |
2552 | # if we don't want to process the raw text and that raw |
|
2552 | # if we don't want to process the raw text and that raw | |
2553 | # text is cached, we can exit early. |
|
2553 | # text is cached, we can exit early. | |
2554 | return rawtext |
|
2554 | return rawtext | |
2555 | if rev is None: |
|
2555 | if rev is None: | |
2556 | rev = self.rev(node) |
|
2556 | rev = self.rev(node) | |
2557 | # the revlog's flag for this revision |
|
2557 | # the revlog's flag for this revision | |
2558 | # (usually alter its state or content) |
|
2558 | # (usually alter its state or content) | |
2559 | flags = self.flags(rev) |
|
2559 | flags = self.flags(rev) | |
2560 |
|
2560 | |||
2561 | if validated and flags == REVIDX_DEFAULT_FLAGS: |
|
2561 | if validated and flags == REVIDX_DEFAULT_FLAGS: | |
2562 | # no extra flags set, no flag processor runs, text = rawtext |
|
2562 | # no extra flags set, no flag processor runs, text = rawtext | |
2563 | return rawtext |
|
2563 | return rawtext | |
2564 |
|
2564 | |||
2565 | if raw: |
|
2565 | if raw: | |
2566 | validatehash = flagutil.processflagsraw(self, rawtext, flags) |
|
2566 | validatehash = flagutil.processflagsraw(self, rawtext, flags) | |
2567 | text = rawtext |
|
2567 | text = rawtext | |
2568 | else: |
|
2568 | else: | |
2569 | r = flagutil.processflagsread(self, rawtext, flags) |
|
2569 | r = flagutil.processflagsread(self, rawtext, flags) | |
2570 | text, validatehash = r |
|
2570 | text, validatehash = r | |
2571 | if validatehash: |
|
2571 | if validatehash: | |
2572 | self.checkhash(text, node, rev=rev) |
|
2572 | self.checkhash(text, node, rev=rev) | |
2573 | if not validated: |
|
2573 | if not validated: | |
2574 | self._revisioncache = (node, rev, rawtext) |
|
2574 | self._inner._revisioncache = (node, rev, rawtext) | |
2575 |
|
2575 | |||
2576 | return text |
|
2576 | return text | |
2577 |
|
2577 | |||
2578 | def _rawtext(self, node, rev): |
|
2578 | def _rawtext(self, node, rev): | |
2579 | """return the possibly unvalidated rawtext for a revision |
|
2579 | """return the possibly unvalidated rawtext for a revision | |
2580 |
|
2580 | |||
2581 | returns (rev, rawtext, validated) |
|
2581 | returns (rev, rawtext, validated) | |
2582 | """ |
|
2582 | """ | |
2583 |
|
2583 | |||
2584 | # revision in the cache (could be useful to apply delta) |
|
2584 | # revision in the cache (could be useful to apply delta) | |
2585 | cachedrev = None |
|
2585 | cachedrev = None | |
2586 | # An intermediate text to apply deltas to |
|
2586 | # An intermediate text to apply deltas to | |
2587 | basetext = None |
|
2587 | basetext = None | |
2588 |
|
2588 | |||
2589 | # Check if we have the entry in cache |
|
2589 | # Check if we have the entry in cache | |
2590 | # The cache entry looks like (node, rev, rawtext) |
|
2590 | # The cache entry looks like (node, rev, rawtext) | |
2591 | if self._revisioncache: |
|
2591 | if self._inner._revisioncache: | |
2592 | if self._revisioncache[0] == node: |
|
2592 | if self._inner._revisioncache[0] == node: | |
2593 | return (rev, self._revisioncache[2], True) |
|
2593 | return (rev, self._inner._revisioncache[2], True) | |
2594 | cachedrev = self._revisioncache[1] |
|
2594 | cachedrev = self._inner._revisioncache[1] | |
2595 |
|
2595 | |||
2596 | if rev is None: |
|
2596 | if rev is None: | |
2597 | rev = self.rev(node) |
|
2597 | rev = self.rev(node) | |
2598 |
|
2598 | |||
2599 | chain, stopped = self._inner._deltachain(rev, stoprev=cachedrev) |
|
2599 | chain, stopped = self._inner._deltachain(rev, stoprev=cachedrev) | |
2600 | if stopped: |
|
2600 | if stopped: | |
2601 | basetext = self._revisioncache[2] |
|
2601 | basetext = self._inner._revisioncache[2] | |
2602 |
|
2602 | |||
2603 | # drop cache to save memory, the caller is expected to |
|
2603 | # drop cache to save memory, the caller is expected to | |
2604 | # update self._revisioncache after validating the text |
|
2604 | # update self._inner._revisioncache after validating the text | |
2605 | self._revisioncache = None |
|
2605 | self._inner._revisioncache = None | |
2606 |
|
2606 | |||
2607 | targetsize = None |
|
2607 | targetsize = None | |
2608 | rawsize = self.index[rev][2] |
|
2608 | rawsize = self.index[rev][2] | |
2609 | if 0 <= rawsize: |
|
2609 | if 0 <= rawsize: | |
2610 | targetsize = 4 * rawsize |
|
2610 | targetsize = 4 * rawsize | |
2611 |
|
2611 | |||
2612 | bins = self._inner._chunks(chain, targetsize=targetsize) |
|
2612 | bins = self._inner._chunks(chain, targetsize=targetsize) | |
2613 | if basetext is None: |
|
2613 | if basetext is None: | |
2614 | basetext = bytes(bins[0]) |
|
2614 | basetext = bytes(bins[0]) | |
2615 | bins = bins[1:] |
|
2615 | bins = bins[1:] | |
2616 |
|
2616 | |||
2617 | rawtext = mdiff.patches(basetext, bins) |
|
2617 | rawtext = mdiff.patches(basetext, bins) | |
2618 | del basetext # let us have a chance to free memory early |
|
2618 | del basetext # let us have a chance to free memory early | |
2619 | return (rev, rawtext, False) |
|
2619 | return (rev, rawtext, False) | |
2620 |
|
2620 | |||
2621 | def _sidedata(self, rev): |
|
2621 | def _sidedata(self, rev): | |
2622 | """Return the sidedata for a given revision number.""" |
|
2622 | """Return the sidedata for a given revision number.""" | |
2623 | index_entry = self.index[rev] |
|
2623 | index_entry = self.index[rev] | |
2624 | sidedata_offset = index_entry[8] |
|
2624 | sidedata_offset = index_entry[8] | |
2625 | sidedata_size = index_entry[9] |
|
2625 | sidedata_size = index_entry[9] | |
2626 |
|
2626 | |||
2627 | if self._inline: |
|
2627 | if self._inline: | |
2628 | sidedata_offset += self.index.entry_size * (1 + rev) |
|
2628 | sidedata_offset += self.index.entry_size * (1 + rev) | |
2629 | if sidedata_size == 0: |
|
2629 | if sidedata_size == 0: | |
2630 | return {} |
|
2630 | return {} | |
2631 |
|
2631 | |||
2632 | if self._docket.sidedata_end < sidedata_offset + sidedata_size: |
|
2632 | if self._docket.sidedata_end < sidedata_offset + sidedata_size: | |
2633 | filename = self._sidedatafile |
|
2633 | filename = self._sidedatafile | |
2634 | end = self._docket.sidedata_end |
|
2634 | end = self._docket.sidedata_end | |
2635 | offset = sidedata_offset |
|
2635 | offset = sidedata_offset | |
2636 | length = sidedata_size |
|
2636 | length = sidedata_size | |
2637 | m = FILE_TOO_SHORT_MSG % (filename, length, offset, end) |
|
2637 | m = FILE_TOO_SHORT_MSG % (filename, length, offset, end) | |
2638 | raise error.RevlogError(m) |
|
2638 | raise error.RevlogError(m) | |
2639 |
|
2639 | |||
2640 | comp_segment = self._inner._segmentfile_sidedata.read_chunk( |
|
2640 | comp_segment = self._inner._segmentfile_sidedata.read_chunk( | |
2641 | sidedata_offset, sidedata_size |
|
2641 | sidedata_offset, sidedata_size | |
2642 | ) |
|
2642 | ) | |
2643 |
|
2643 | |||
2644 | comp = self.index[rev][11] |
|
2644 | comp = self.index[rev][11] | |
2645 | if comp == COMP_MODE_PLAIN: |
|
2645 | if comp == COMP_MODE_PLAIN: | |
2646 | segment = comp_segment |
|
2646 | segment = comp_segment | |
2647 | elif comp == COMP_MODE_DEFAULT: |
|
2647 | elif comp == COMP_MODE_DEFAULT: | |
2648 | segment = self._inner._decompressor(comp_segment) |
|
2648 | segment = self._inner._decompressor(comp_segment) | |
2649 | elif comp == COMP_MODE_INLINE: |
|
2649 | elif comp == COMP_MODE_INLINE: | |
2650 | segment = self._inner.decompress(comp_segment) |
|
2650 | segment = self._inner.decompress(comp_segment) | |
2651 | else: |
|
2651 | else: | |
2652 | msg = b'unknown compression mode %d' |
|
2652 | msg = b'unknown compression mode %d' | |
2653 | msg %= comp |
|
2653 | msg %= comp | |
2654 | raise error.RevlogError(msg) |
|
2654 | raise error.RevlogError(msg) | |
2655 |
|
2655 | |||
2656 | sidedata = sidedatautil.deserialize_sidedata(segment) |
|
2656 | sidedata = sidedatautil.deserialize_sidedata(segment) | |
2657 | return sidedata |
|
2657 | return sidedata | |
2658 |
|
2658 | |||
2659 | def rawdata(self, nodeorrev): |
|
2659 | def rawdata(self, nodeorrev): | |
2660 | """return an uncompressed raw data of a given node or revision number.""" |
|
2660 | """return an uncompressed raw data of a given node or revision number.""" | |
2661 | return self._revisiondata(nodeorrev, raw=True) |
|
2661 | return self._revisiondata(nodeorrev, raw=True) | |
2662 |
|
2662 | |||
2663 | def hash(self, text, p1, p2): |
|
2663 | def hash(self, text, p1, p2): | |
2664 | """Compute a node hash. |
|
2664 | """Compute a node hash. | |
2665 |
|
2665 | |||
2666 | Available as a function so that subclasses can replace the hash |
|
2666 | Available as a function so that subclasses can replace the hash | |
2667 | as needed. |
|
2667 | as needed. | |
2668 | """ |
|
2668 | """ | |
2669 | return storageutil.hashrevisionsha1(text, p1, p2) |
|
2669 | return storageutil.hashrevisionsha1(text, p1, p2) | |
2670 |
|
2670 | |||
2671 | def checkhash(self, text, node, p1=None, p2=None, rev=None): |
|
2671 | def checkhash(self, text, node, p1=None, p2=None, rev=None): | |
2672 | """Check node hash integrity. |
|
2672 | """Check node hash integrity. | |
2673 |
|
2673 | |||
2674 | Available as a function so that subclasses can extend hash mismatch |
|
2674 | Available as a function so that subclasses can extend hash mismatch | |
2675 | behaviors as needed. |
|
2675 | behaviors as needed. | |
2676 | """ |
|
2676 | """ | |
2677 | try: |
|
2677 | try: | |
2678 | if p1 is None and p2 is None: |
|
2678 | if p1 is None and p2 is None: | |
2679 | p1, p2 = self.parents(node) |
|
2679 | p1, p2 = self.parents(node) | |
2680 | if node != self.hash(text, p1, p2): |
|
2680 | if node != self.hash(text, p1, p2): | |
2681 | # Clear the revision cache on hash failure. The revision cache |
|
2681 | # Clear the revision cache on hash failure. The revision cache | |
2682 | # only stores the raw revision and clearing the cache does have |
|
2682 | # only stores the raw revision and clearing the cache does have | |
2683 | # the side-effect that we won't have a cache hit when the raw |
|
2683 | # the side-effect that we won't have a cache hit when the raw | |
2684 | # revision data is accessed. But this case should be rare and |
|
2684 | # revision data is accessed. But this case should be rare and | |
2685 | # it is extra work to teach the cache about the hash |
|
2685 | # it is extra work to teach the cache about the hash | |
2686 | # verification state. |
|
2686 | # verification state. | |
2687 | if self._revisioncache and self._revisioncache[0] == node: |
|
2687 | if ( | |
2688 |
self._revisioncache |
|
2688 | self._inner._revisioncache | |
|
2689 | and self._inner._revisioncache[0] == node | |||
|
2690 | ): | |||
|
2691 | self._inner._revisioncache = None | |||
2689 |
|
2692 | |||
2690 | revornode = rev |
|
2693 | revornode = rev | |
2691 | if revornode is None: |
|
2694 | if revornode is None: | |
2692 | revornode = templatefilters.short(hex(node)) |
|
2695 | revornode = templatefilters.short(hex(node)) | |
2693 | raise error.RevlogError( |
|
2696 | raise error.RevlogError( | |
2694 | _(b"integrity check failed on %s:%s") |
|
2697 | _(b"integrity check failed on %s:%s") | |
2695 | % (self.display_id, pycompat.bytestr(revornode)) |
|
2698 | % (self.display_id, pycompat.bytestr(revornode)) | |
2696 | ) |
|
2699 | ) | |
2697 | except error.RevlogError: |
|
2700 | except error.RevlogError: | |
2698 | if self.feature_config.censorable and storageutil.iscensoredtext( |
|
2701 | if self.feature_config.censorable and storageutil.iscensoredtext( | |
2699 | text |
|
2702 | text | |
2700 | ): |
|
2703 | ): | |
2701 | raise error.CensoredNodeError(self.display_id, node, text) |
|
2704 | raise error.CensoredNodeError(self.display_id, node, text) | |
2702 | raise |
|
2705 | raise | |
2703 |
|
2706 | |||
2704 | @property |
|
2707 | @property | |
2705 | def _split_index_file(self): |
|
2708 | def _split_index_file(self): | |
2706 | """the path where to expect the index of an ongoing splitting operation |
|
2709 | """the path where to expect the index of an ongoing splitting operation | |
2707 |
|
2710 | |||
2708 | The file will only exist if a splitting operation is in progress, but |
|
2711 | The file will only exist if a splitting operation is in progress, but | |
2709 | it is always expected at the same location.""" |
|
2712 | it is always expected at the same location.""" | |
2710 | parts = self.radix.split(b'/') |
|
2713 | parts = self.radix.split(b'/') | |
2711 | if len(parts) > 1: |
|
2714 | if len(parts) > 1: | |
2712 | # adds a '-s' prefix to the ``data/` or `meta/` base |
|
2715 | # adds a '-s' prefix to the ``data/` or `meta/` base | |
2713 | head = parts[0] + b'-s' |
|
2716 | head = parts[0] + b'-s' | |
2714 | mids = parts[1:-1] |
|
2717 | mids = parts[1:-1] | |
2715 | tail = parts[-1] + b'.i' |
|
2718 | tail = parts[-1] + b'.i' | |
2716 | pieces = [head] + mids + [tail] |
|
2719 | pieces = [head] + mids + [tail] | |
2717 | return b'/'.join(pieces) |
|
2720 | return b'/'.join(pieces) | |
2718 | else: |
|
2721 | else: | |
2719 | # the revlog is stored at the root of the store (changelog or |
|
2722 | # the revlog is stored at the root of the store (changelog or | |
2720 | # manifest), no risk of collision. |
|
2723 | # manifest), no risk of collision. | |
2721 | return self.radix + b'.i.s' |
|
2724 | return self.radix + b'.i.s' | |
2722 |
|
2725 | |||
2723 | def _enforceinlinesize(self, tr, side_write=True): |
|
2726 | def _enforceinlinesize(self, tr, side_write=True): | |
2724 | """Check if the revlog is too big for inline and convert if so. |
|
2727 | """Check if the revlog is too big for inline and convert if so. | |
2725 |
|
2728 | |||
2726 | This should be called after revisions are added to the revlog. If the |
|
2729 | This should be called after revisions are added to the revlog. If the | |
2727 | revlog has grown too large to be an inline revlog, it will convert it |
|
2730 | revlog has grown too large to be an inline revlog, it will convert it | |
2728 | to use multiple index and data files. |
|
2731 | to use multiple index and data files. | |
2729 | """ |
|
2732 | """ | |
2730 | tiprev = len(self) - 1 |
|
2733 | tiprev = len(self) - 1 | |
2731 | total_size = self.start(tiprev) + self.length(tiprev) |
|
2734 | total_size = self.start(tiprev) + self.length(tiprev) | |
2732 | if not self._inline or total_size < _maxinline: |
|
2735 | if not self._inline or total_size < _maxinline: | |
2733 | return |
|
2736 | return | |
2734 |
|
2737 | |||
2735 | if self._docket is not None: |
|
2738 | if self._docket is not None: | |
2736 | msg = b"inline revlog should not have a docket" |
|
2739 | msg = b"inline revlog should not have a docket" | |
2737 | raise error.ProgrammingError(msg) |
|
2740 | raise error.ProgrammingError(msg) | |
2738 |
|
2741 | |||
2739 | troffset = tr.findoffset(self._indexfile) |
|
2742 | troffset = tr.findoffset(self._indexfile) | |
2740 | if troffset is None: |
|
2743 | if troffset is None: | |
2741 | raise error.RevlogError( |
|
2744 | raise error.RevlogError( | |
2742 | _(b"%s not found in the transaction") % self._indexfile |
|
2745 | _(b"%s not found in the transaction") % self._indexfile | |
2743 | ) |
|
2746 | ) | |
2744 | if troffset: |
|
2747 | if troffset: | |
2745 | tr.addbackup(self._indexfile, for_offset=True) |
|
2748 | tr.addbackup(self._indexfile, for_offset=True) | |
2746 | tr.add(self._datafile, 0) |
|
2749 | tr.add(self._datafile, 0) | |
2747 |
|
2750 | |||
2748 | new_index_file_path = None |
|
2751 | new_index_file_path = None | |
2749 | if side_write: |
|
2752 | if side_write: | |
2750 | old_index_file_path = self._indexfile |
|
2753 | old_index_file_path = self._indexfile | |
2751 | new_index_file_path = self._split_index_file |
|
2754 | new_index_file_path = self._split_index_file | |
2752 | opener = self.opener |
|
2755 | opener = self.opener | |
2753 | weak_self = weakref.ref(self) |
|
2756 | weak_self = weakref.ref(self) | |
2754 |
|
2757 | |||
2755 | # the "split" index replace the real index when the transaction is |
|
2758 | # the "split" index replace the real index when the transaction is | |
2756 | # finalized |
|
2759 | # finalized | |
2757 | def finalize_callback(tr): |
|
2760 | def finalize_callback(tr): | |
2758 | opener.rename( |
|
2761 | opener.rename( | |
2759 | new_index_file_path, |
|
2762 | new_index_file_path, | |
2760 | old_index_file_path, |
|
2763 | old_index_file_path, | |
2761 | checkambig=True, |
|
2764 | checkambig=True, | |
2762 | ) |
|
2765 | ) | |
2763 | maybe_self = weak_self() |
|
2766 | maybe_self = weak_self() | |
2764 | if maybe_self is not None: |
|
2767 | if maybe_self is not None: | |
2765 | maybe_self._indexfile = old_index_file_path |
|
2768 | maybe_self._indexfile = old_index_file_path | |
2766 | maybe_self._inner.index_file = maybe_self._indexfile |
|
2769 | maybe_self._inner.index_file = maybe_self._indexfile | |
2767 |
|
2770 | |||
2768 | def abort_callback(tr): |
|
2771 | def abort_callback(tr): | |
2769 | maybe_self = weak_self() |
|
2772 | maybe_self = weak_self() | |
2770 | if maybe_self is not None: |
|
2773 | if maybe_self is not None: | |
2771 | maybe_self._indexfile = old_index_file_path |
|
2774 | maybe_self._indexfile = old_index_file_path | |
2772 | maybe_self._inner.inline = True |
|
2775 | maybe_self._inner.inline = True | |
2773 | maybe_self._inner.index_file = old_index_file_path |
|
2776 | maybe_self._inner.index_file = old_index_file_path | |
2774 |
|
2777 | |||
2775 | tr.registertmp(new_index_file_path) |
|
2778 | tr.registertmp(new_index_file_path) | |
2776 | if self.target[1] is not None: |
|
2779 | if self.target[1] is not None: | |
2777 | callback_id = b'000-revlog-split-%d-%s' % self.target |
|
2780 | callback_id = b'000-revlog-split-%d-%s' % self.target | |
2778 | else: |
|
2781 | else: | |
2779 | callback_id = b'000-revlog-split-%d' % self.target[0] |
|
2782 | callback_id = b'000-revlog-split-%d' % self.target[0] | |
2780 | tr.addfinalize(callback_id, finalize_callback) |
|
2783 | tr.addfinalize(callback_id, finalize_callback) | |
2781 | tr.addabort(callback_id, abort_callback) |
|
2784 | tr.addabort(callback_id, abort_callback) | |
2782 |
|
2785 | |||
2783 | self._format_flags &= ~FLAG_INLINE_DATA |
|
2786 | self._format_flags &= ~FLAG_INLINE_DATA | |
2784 | self._inner.split_inline( |
|
2787 | self._inner.split_inline( | |
2785 | tr, |
|
2788 | tr, | |
2786 | self._format_flags | self._format_version, |
|
2789 | self._format_flags | self._format_version, | |
2787 | new_index_file_path=new_index_file_path, |
|
2790 | new_index_file_path=new_index_file_path, | |
2788 | ) |
|
2791 | ) | |
2789 |
|
2792 | |||
2790 | self._inline = False |
|
2793 | self._inline = False | |
2791 | if new_index_file_path is not None: |
|
2794 | if new_index_file_path is not None: | |
2792 | self._indexfile = new_index_file_path |
|
2795 | self._indexfile = new_index_file_path | |
2793 |
|
2796 | |||
2794 | nodemaputil.setup_persistent_nodemap(tr, self) |
|
2797 | nodemaputil.setup_persistent_nodemap(tr, self) | |
2795 |
|
2798 | |||
2796 | def _nodeduplicatecallback(self, transaction, node): |
|
2799 | def _nodeduplicatecallback(self, transaction, node): | |
2797 | """called when trying to add a node already stored.""" |
|
2800 | """called when trying to add a node already stored.""" | |
2798 |
|
2801 | |||
2799 | @contextlib.contextmanager |
|
2802 | @contextlib.contextmanager | |
2800 | def reading(self): |
|
2803 | def reading(self): | |
2801 | with self._inner.reading(): |
|
2804 | with self._inner.reading(): | |
2802 | yield |
|
2805 | yield | |
2803 |
|
2806 | |||
2804 | @contextlib.contextmanager |
|
2807 | @contextlib.contextmanager | |
2805 | def _writing(self, transaction): |
|
2808 | def _writing(self, transaction): | |
2806 | if self._trypending: |
|
2809 | if self._trypending: | |
2807 | msg = b'try to write in a `trypending` revlog: %s' |
|
2810 | msg = b'try to write in a `trypending` revlog: %s' | |
2808 | msg %= self.display_id |
|
2811 | msg %= self.display_id | |
2809 | raise error.ProgrammingError(msg) |
|
2812 | raise error.ProgrammingError(msg) | |
2810 | if self._inner.is_writing: |
|
2813 | if self._inner.is_writing: | |
2811 | yield |
|
2814 | yield | |
2812 | else: |
|
2815 | else: | |
2813 | data_end = None |
|
2816 | data_end = None | |
2814 | sidedata_end = None |
|
2817 | sidedata_end = None | |
2815 | if self._docket is not None: |
|
2818 | if self._docket is not None: | |
2816 | data_end = self._docket.data_end |
|
2819 | data_end = self._docket.data_end | |
2817 | sidedata_end = self._docket.sidedata_end |
|
2820 | sidedata_end = self._docket.sidedata_end | |
2818 | with self._inner.writing( |
|
2821 | with self._inner.writing( | |
2819 | transaction, |
|
2822 | transaction, | |
2820 | data_end=data_end, |
|
2823 | data_end=data_end, | |
2821 | sidedata_end=sidedata_end, |
|
2824 | sidedata_end=sidedata_end, | |
2822 | ): |
|
2825 | ): | |
2823 | yield |
|
2826 | yield | |
2824 | if self._docket is not None: |
|
2827 | if self._docket is not None: | |
2825 | self._write_docket(transaction) |
|
2828 | self._write_docket(transaction) | |
2826 |
|
2829 | |||
2827 | def _write_docket(self, transaction): |
|
2830 | def _write_docket(self, transaction): | |
2828 | """write the current docket on disk |
|
2831 | """write the current docket on disk | |
2829 |
|
2832 | |||
2830 | Exist as a method to help changelog to implement transaction logic |
|
2833 | Exist as a method to help changelog to implement transaction logic | |
2831 |
|
2834 | |||
2832 | We could also imagine using the same transaction logic for all revlog |
|
2835 | We could also imagine using the same transaction logic for all revlog | |
2833 | since docket are cheap.""" |
|
2836 | since docket are cheap.""" | |
2834 | self._docket.write(transaction) |
|
2837 | self._docket.write(transaction) | |
2835 |
|
2838 | |||
2836 | def addrevision( |
|
2839 | def addrevision( | |
2837 | self, |
|
2840 | self, | |
2838 | text, |
|
2841 | text, | |
2839 | transaction, |
|
2842 | transaction, | |
2840 | link, |
|
2843 | link, | |
2841 | p1, |
|
2844 | p1, | |
2842 | p2, |
|
2845 | p2, | |
2843 | cachedelta=None, |
|
2846 | cachedelta=None, | |
2844 | node=None, |
|
2847 | node=None, | |
2845 | flags=REVIDX_DEFAULT_FLAGS, |
|
2848 | flags=REVIDX_DEFAULT_FLAGS, | |
2846 | deltacomputer=None, |
|
2849 | deltacomputer=None, | |
2847 | sidedata=None, |
|
2850 | sidedata=None, | |
2848 | ): |
|
2851 | ): | |
2849 | """add a revision to the log |
|
2852 | """add a revision to the log | |
2850 |
|
2853 | |||
2851 | text - the revision data to add |
|
2854 | text - the revision data to add | |
2852 | transaction - the transaction object used for rollback |
|
2855 | transaction - the transaction object used for rollback | |
2853 | link - the linkrev data to add |
|
2856 | link - the linkrev data to add | |
2854 | p1, p2 - the parent nodeids of the revision |
|
2857 | p1, p2 - the parent nodeids of the revision | |
2855 | cachedelta - an optional precomputed delta |
|
2858 | cachedelta - an optional precomputed delta | |
2856 | node - nodeid of revision; typically node is not specified, and it is |
|
2859 | node - nodeid of revision; typically node is not specified, and it is | |
2857 | computed by default as hash(text, p1, p2), however subclasses might |
|
2860 | computed by default as hash(text, p1, p2), however subclasses might | |
2858 | use different hashing method (and override checkhash() in such case) |
|
2861 | use different hashing method (and override checkhash() in such case) | |
2859 | flags - the known flags to set on the revision |
|
2862 | flags - the known flags to set on the revision | |
2860 | deltacomputer - an optional deltacomputer instance shared between |
|
2863 | deltacomputer - an optional deltacomputer instance shared between | |
2861 | multiple calls |
|
2864 | multiple calls | |
2862 | """ |
|
2865 | """ | |
2863 | if link == nullrev: |
|
2866 | if link == nullrev: | |
2864 | raise error.RevlogError( |
|
2867 | raise error.RevlogError( | |
2865 | _(b"attempted to add linkrev -1 to %s") % self.display_id |
|
2868 | _(b"attempted to add linkrev -1 to %s") % self.display_id | |
2866 | ) |
|
2869 | ) | |
2867 |
|
2870 | |||
2868 | if sidedata is None: |
|
2871 | if sidedata is None: | |
2869 | sidedata = {} |
|
2872 | sidedata = {} | |
2870 | elif sidedata and not self.feature_config.has_side_data: |
|
2873 | elif sidedata and not self.feature_config.has_side_data: | |
2871 | raise error.ProgrammingError( |
|
2874 | raise error.ProgrammingError( | |
2872 | _(b"trying to add sidedata to a revlog who don't support them") |
|
2875 | _(b"trying to add sidedata to a revlog who don't support them") | |
2873 | ) |
|
2876 | ) | |
2874 |
|
2877 | |||
2875 | if flags: |
|
2878 | if flags: | |
2876 | node = node or self.hash(text, p1, p2) |
|
2879 | node = node or self.hash(text, p1, p2) | |
2877 |
|
2880 | |||
2878 | rawtext, validatehash = flagutil.processflagswrite(self, text, flags) |
|
2881 | rawtext, validatehash = flagutil.processflagswrite(self, text, flags) | |
2879 |
|
2882 | |||
2880 | # If the flag processor modifies the revision data, ignore any provided |
|
2883 | # If the flag processor modifies the revision data, ignore any provided | |
2881 | # cachedelta. |
|
2884 | # cachedelta. | |
2882 | if rawtext != text: |
|
2885 | if rawtext != text: | |
2883 | cachedelta = None |
|
2886 | cachedelta = None | |
2884 |
|
2887 | |||
2885 | if len(rawtext) > _maxentrysize: |
|
2888 | if len(rawtext) > _maxentrysize: | |
2886 | raise error.RevlogError( |
|
2889 | raise error.RevlogError( | |
2887 | _( |
|
2890 | _( | |
2888 | b"%s: size of %d bytes exceeds maximum revlog storage of 2GiB" |
|
2891 | b"%s: size of %d bytes exceeds maximum revlog storage of 2GiB" | |
2889 | ) |
|
2892 | ) | |
2890 | % (self.display_id, len(rawtext)) |
|
2893 | % (self.display_id, len(rawtext)) | |
2891 | ) |
|
2894 | ) | |
2892 |
|
2895 | |||
2893 | node = node or self.hash(rawtext, p1, p2) |
|
2896 | node = node or self.hash(rawtext, p1, p2) | |
2894 | rev = self.index.get_rev(node) |
|
2897 | rev = self.index.get_rev(node) | |
2895 | if rev is not None: |
|
2898 | if rev is not None: | |
2896 | return rev |
|
2899 | return rev | |
2897 |
|
2900 | |||
2898 | if validatehash: |
|
2901 | if validatehash: | |
2899 | self.checkhash(rawtext, node, p1=p1, p2=p2) |
|
2902 | self.checkhash(rawtext, node, p1=p1, p2=p2) | |
2900 |
|
2903 | |||
2901 | return self.addrawrevision( |
|
2904 | return self.addrawrevision( | |
2902 | rawtext, |
|
2905 | rawtext, | |
2903 | transaction, |
|
2906 | transaction, | |
2904 | link, |
|
2907 | link, | |
2905 | p1, |
|
2908 | p1, | |
2906 | p2, |
|
2909 | p2, | |
2907 | node, |
|
2910 | node, | |
2908 | flags, |
|
2911 | flags, | |
2909 | cachedelta=cachedelta, |
|
2912 | cachedelta=cachedelta, | |
2910 | deltacomputer=deltacomputer, |
|
2913 | deltacomputer=deltacomputer, | |
2911 | sidedata=sidedata, |
|
2914 | sidedata=sidedata, | |
2912 | ) |
|
2915 | ) | |
2913 |
|
2916 | |||
2914 | def addrawrevision( |
|
2917 | def addrawrevision( | |
2915 | self, |
|
2918 | self, | |
2916 | rawtext, |
|
2919 | rawtext, | |
2917 | transaction, |
|
2920 | transaction, | |
2918 | link, |
|
2921 | link, | |
2919 | p1, |
|
2922 | p1, | |
2920 | p2, |
|
2923 | p2, | |
2921 | node, |
|
2924 | node, | |
2922 | flags, |
|
2925 | flags, | |
2923 | cachedelta=None, |
|
2926 | cachedelta=None, | |
2924 | deltacomputer=None, |
|
2927 | deltacomputer=None, | |
2925 | sidedata=None, |
|
2928 | sidedata=None, | |
2926 | ): |
|
2929 | ): | |
2927 | """add a raw revision with known flags, node and parents |
|
2930 | """add a raw revision with known flags, node and parents | |
2928 | useful when reusing a revision not stored in this revlog (ex: received |
|
2931 | useful when reusing a revision not stored in this revlog (ex: received | |
2929 | over wire, or read from an external bundle). |
|
2932 | over wire, or read from an external bundle). | |
2930 | """ |
|
2933 | """ | |
2931 | with self._writing(transaction): |
|
2934 | with self._writing(transaction): | |
2932 | return self._addrevision( |
|
2935 | return self._addrevision( | |
2933 | node, |
|
2936 | node, | |
2934 | rawtext, |
|
2937 | rawtext, | |
2935 | transaction, |
|
2938 | transaction, | |
2936 | link, |
|
2939 | link, | |
2937 | p1, |
|
2940 | p1, | |
2938 | p2, |
|
2941 | p2, | |
2939 | flags, |
|
2942 | flags, | |
2940 | cachedelta, |
|
2943 | cachedelta, | |
2941 | deltacomputer=deltacomputer, |
|
2944 | deltacomputer=deltacomputer, | |
2942 | sidedata=sidedata, |
|
2945 | sidedata=sidedata, | |
2943 | ) |
|
2946 | ) | |
2944 |
|
2947 | |||
2945 | def compress(self, data): |
|
2948 | def compress(self, data): | |
2946 | return self._inner.compress(data) |
|
2949 | return self._inner.compress(data) | |
2947 |
|
2950 | |||
2948 | def decompress(self, data): |
|
2951 | def decompress(self, data): | |
2949 | return self._inner.decompress(data) |
|
2952 | return self._inner.decompress(data) | |
2950 |
|
2953 | |||
2951 | def _addrevision( |
|
2954 | def _addrevision( | |
2952 | self, |
|
2955 | self, | |
2953 | node, |
|
2956 | node, | |
2954 | rawtext, |
|
2957 | rawtext, | |
2955 | transaction, |
|
2958 | transaction, | |
2956 | link, |
|
2959 | link, | |
2957 | p1, |
|
2960 | p1, | |
2958 | p2, |
|
2961 | p2, | |
2959 | flags, |
|
2962 | flags, | |
2960 | cachedelta, |
|
2963 | cachedelta, | |
2961 | alwayscache=False, |
|
2964 | alwayscache=False, | |
2962 | deltacomputer=None, |
|
2965 | deltacomputer=None, | |
2963 | sidedata=None, |
|
2966 | sidedata=None, | |
2964 | ): |
|
2967 | ): | |
2965 | """internal function to add revisions to the log |
|
2968 | """internal function to add revisions to the log | |
2966 |
|
2969 | |||
2967 | see addrevision for argument descriptions. |
|
2970 | see addrevision for argument descriptions. | |
2968 |
|
2971 | |||
2969 | note: "addrevision" takes non-raw text, "_addrevision" takes raw text. |
|
2972 | note: "addrevision" takes non-raw text, "_addrevision" takes raw text. | |
2970 |
|
2973 | |||
2971 | if "deltacomputer" is not provided or None, a defaultdeltacomputer will |
|
2974 | if "deltacomputer" is not provided or None, a defaultdeltacomputer will | |
2972 | be used. |
|
2975 | be used. | |
2973 |
|
2976 | |||
2974 | invariants: |
|
2977 | invariants: | |
2975 | - rawtext is optional (can be None); if not set, cachedelta must be set. |
|
2978 | - rawtext is optional (can be None); if not set, cachedelta must be set. | |
2976 | if both are set, they must correspond to each other. |
|
2979 | if both are set, they must correspond to each other. | |
2977 | """ |
|
2980 | """ | |
2978 | if node == self.nullid: |
|
2981 | if node == self.nullid: | |
2979 | raise error.RevlogError( |
|
2982 | raise error.RevlogError( | |
2980 | _(b"%s: attempt to add null revision") % self.display_id |
|
2983 | _(b"%s: attempt to add null revision") % self.display_id | |
2981 | ) |
|
2984 | ) | |
2982 | if ( |
|
2985 | if ( | |
2983 | node == self.nodeconstants.wdirid |
|
2986 | node == self.nodeconstants.wdirid | |
2984 | or node in self.nodeconstants.wdirfilenodeids |
|
2987 | or node in self.nodeconstants.wdirfilenodeids | |
2985 | ): |
|
2988 | ): | |
2986 | raise error.RevlogError( |
|
2989 | raise error.RevlogError( | |
2987 | _(b"%s: attempt to add wdir revision") % self.display_id |
|
2990 | _(b"%s: attempt to add wdir revision") % self.display_id | |
2988 | ) |
|
2991 | ) | |
2989 | if self._inner._writinghandles is None: |
|
2992 | if self._inner._writinghandles is None: | |
2990 | msg = b'adding revision outside `revlog._writing` context' |
|
2993 | msg = b'adding revision outside `revlog._writing` context' | |
2991 | raise error.ProgrammingError(msg) |
|
2994 | raise error.ProgrammingError(msg) | |
2992 |
|
2995 | |||
2993 | btext = [rawtext] |
|
2996 | btext = [rawtext] | |
2994 |
|
2997 | |||
2995 | curr = len(self) |
|
2998 | curr = len(self) | |
2996 | prev = curr - 1 |
|
2999 | prev = curr - 1 | |
2997 |
|
3000 | |||
2998 | offset = self._get_data_offset(prev) |
|
3001 | offset = self._get_data_offset(prev) | |
2999 |
|
3002 | |||
3000 | if self._concurrencychecker: |
|
3003 | if self._concurrencychecker: | |
3001 | ifh, dfh, sdfh = self._inner._writinghandles |
|
3004 | ifh, dfh, sdfh = self._inner._writinghandles | |
3002 | # XXX no checking for the sidedata file |
|
3005 | # XXX no checking for the sidedata file | |
3003 | if self._inline: |
|
3006 | if self._inline: | |
3004 | # offset is "as if" it were in the .d file, so we need to add on |
|
3007 | # offset is "as if" it were in the .d file, so we need to add on | |
3005 | # the size of the entry metadata. |
|
3008 | # the size of the entry metadata. | |
3006 | self._concurrencychecker( |
|
3009 | self._concurrencychecker( | |
3007 | ifh, self._indexfile, offset + curr * self.index.entry_size |
|
3010 | ifh, self._indexfile, offset + curr * self.index.entry_size | |
3008 | ) |
|
3011 | ) | |
3009 | else: |
|
3012 | else: | |
3010 | # Entries in the .i are a consistent size. |
|
3013 | # Entries in the .i are a consistent size. | |
3011 | self._concurrencychecker( |
|
3014 | self._concurrencychecker( | |
3012 | ifh, self._indexfile, curr * self.index.entry_size |
|
3015 | ifh, self._indexfile, curr * self.index.entry_size | |
3013 | ) |
|
3016 | ) | |
3014 | self._concurrencychecker(dfh, self._datafile, offset) |
|
3017 | self._concurrencychecker(dfh, self._datafile, offset) | |
3015 |
|
3018 | |||
3016 | p1r, p2r = self.rev(p1), self.rev(p2) |
|
3019 | p1r, p2r = self.rev(p1), self.rev(p2) | |
3017 |
|
3020 | |||
3018 | # full versions are inserted when the needed deltas |
|
3021 | # full versions are inserted when the needed deltas | |
3019 | # become comparable to the uncompressed text |
|
3022 | # become comparable to the uncompressed text | |
3020 | if rawtext is None: |
|
3023 | if rawtext is None: | |
3021 | # need rawtext size, before changed by flag processors, which is |
|
3024 | # need rawtext size, before changed by flag processors, which is | |
3022 | # the non-raw size. use revlog explicitly to avoid filelog's extra |
|
3025 | # the non-raw size. use revlog explicitly to avoid filelog's extra | |
3023 | # logic that might remove metadata size. |
|
3026 | # logic that might remove metadata size. | |
3024 | textlen = mdiff.patchedsize( |
|
3027 | textlen = mdiff.patchedsize( | |
3025 | revlog.size(self, cachedelta[0]), cachedelta[1] |
|
3028 | revlog.size(self, cachedelta[0]), cachedelta[1] | |
3026 | ) |
|
3029 | ) | |
3027 | else: |
|
3030 | else: | |
3028 | textlen = len(rawtext) |
|
3031 | textlen = len(rawtext) | |
3029 |
|
3032 | |||
3030 | if deltacomputer is None: |
|
3033 | if deltacomputer is None: | |
3031 | write_debug = None |
|
3034 | write_debug = None | |
3032 | if self.delta_config.debug_delta: |
|
3035 | if self.delta_config.debug_delta: | |
3033 | write_debug = transaction._report |
|
3036 | write_debug = transaction._report | |
3034 | deltacomputer = deltautil.deltacomputer( |
|
3037 | deltacomputer = deltautil.deltacomputer( | |
3035 | self, write_debug=write_debug |
|
3038 | self, write_debug=write_debug | |
3036 | ) |
|
3039 | ) | |
3037 |
|
3040 | |||
3038 | if cachedelta is not None and len(cachedelta) == 2: |
|
3041 | if cachedelta is not None and len(cachedelta) == 2: | |
3039 | # If the cached delta has no information about how it should be |
|
3042 | # If the cached delta has no information about how it should be | |
3040 | # reused, add the default reuse instruction according to the |
|
3043 | # reused, add the default reuse instruction according to the | |
3041 | # revlog's configuration. |
|
3044 | # revlog's configuration. | |
3042 | if ( |
|
3045 | if ( | |
3043 | self.delta_config.general_delta |
|
3046 | self.delta_config.general_delta | |
3044 | and self.delta_config.lazy_delta_base |
|
3047 | and self.delta_config.lazy_delta_base | |
3045 | ): |
|
3048 | ): | |
3046 | delta_base_reuse = DELTA_BASE_REUSE_TRY |
|
3049 | delta_base_reuse = DELTA_BASE_REUSE_TRY | |
3047 | else: |
|
3050 | else: | |
3048 | delta_base_reuse = DELTA_BASE_REUSE_NO |
|
3051 | delta_base_reuse = DELTA_BASE_REUSE_NO | |
3049 | cachedelta = (cachedelta[0], cachedelta[1], delta_base_reuse) |
|
3052 | cachedelta = (cachedelta[0], cachedelta[1], delta_base_reuse) | |
3050 |
|
3053 | |||
3051 | revinfo = revlogutils.revisioninfo( |
|
3054 | revinfo = revlogutils.revisioninfo( | |
3052 | node, |
|
3055 | node, | |
3053 | p1, |
|
3056 | p1, | |
3054 | p2, |
|
3057 | p2, | |
3055 | btext, |
|
3058 | btext, | |
3056 | textlen, |
|
3059 | textlen, | |
3057 | cachedelta, |
|
3060 | cachedelta, | |
3058 | flags, |
|
3061 | flags, | |
3059 | ) |
|
3062 | ) | |
3060 |
|
3063 | |||
3061 | deltainfo = deltacomputer.finddeltainfo(revinfo) |
|
3064 | deltainfo = deltacomputer.finddeltainfo(revinfo) | |
3062 |
|
3065 | |||
3063 | compression_mode = COMP_MODE_INLINE |
|
3066 | compression_mode = COMP_MODE_INLINE | |
3064 | if self._docket is not None: |
|
3067 | if self._docket is not None: | |
3065 | default_comp = self._docket.default_compression_header |
|
3068 | default_comp = self._docket.default_compression_header | |
3066 | r = deltautil.delta_compression(default_comp, deltainfo) |
|
3069 | r = deltautil.delta_compression(default_comp, deltainfo) | |
3067 | compression_mode, deltainfo = r |
|
3070 | compression_mode, deltainfo = r | |
3068 |
|
3071 | |||
3069 | sidedata_compression_mode = COMP_MODE_INLINE |
|
3072 | sidedata_compression_mode = COMP_MODE_INLINE | |
3070 | if sidedata and self.feature_config.has_side_data: |
|
3073 | if sidedata and self.feature_config.has_side_data: | |
3071 | sidedata_compression_mode = COMP_MODE_PLAIN |
|
3074 | sidedata_compression_mode = COMP_MODE_PLAIN | |
3072 | serialized_sidedata = sidedatautil.serialize_sidedata(sidedata) |
|
3075 | serialized_sidedata = sidedatautil.serialize_sidedata(sidedata) | |
3073 | sidedata_offset = self._docket.sidedata_end |
|
3076 | sidedata_offset = self._docket.sidedata_end | |
3074 | h, comp_sidedata = self._inner.compress(serialized_sidedata) |
|
3077 | h, comp_sidedata = self._inner.compress(serialized_sidedata) | |
3075 | if ( |
|
3078 | if ( | |
3076 | h != b'u' |
|
3079 | h != b'u' | |
3077 | and comp_sidedata[0:1] != b'\0' |
|
3080 | and comp_sidedata[0:1] != b'\0' | |
3078 | and len(comp_sidedata) < len(serialized_sidedata) |
|
3081 | and len(comp_sidedata) < len(serialized_sidedata) | |
3079 | ): |
|
3082 | ): | |
3080 | assert not h |
|
3083 | assert not h | |
3081 | if ( |
|
3084 | if ( | |
3082 | comp_sidedata[0:1] |
|
3085 | comp_sidedata[0:1] | |
3083 | == self._docket.default_compression_header |
|
3086 | == self._docket.default_compression_header | |
3084 | ): |
|
3087 | ): | |
3085 | sidedata_compression_mode = COMP_MODE_DEFAULT |
|
3088 | sidedata_compression_mode = COMP_MODE_DEFAULT | |
3086 | serialized_sidedata = comp_sidedata |
|
3089 | serialized_sidedata = comp_sidedata | |
3087 | else: |
|
3090 | else: | |
3088 | sidedata_compression_mode = COMP_MODE_INLINE |
|
3091 | sidedata_compression_mode = COMP_MODE_INLINE | |
3089 | serialized_sidedata = comp_sidedata |
|
3092 | serialized_sidedata = comp_sidedata | |
3090 | else: |
|
3093 | else: | |
3091 | serialized_sidedata = b"" |
|
3094 | serialized_sidedata = b"" | |
3092 | # Don't store the offset if the sidedata is empty, that way |
|
3095 | # Don't store the offset if the sidedata is empty, that way | |
3093 | # we can easily detect empty sidedata and they will be no different |
|
3096 | # we can easily detect empty sidedata and they will be no different | |
3094 | # than ones we manually add. |
|
3097 | # than ones we manually add. | |
3095 | sidedata_offset = 0 |
|
3098 | sidedata_offset = 0 | |
3096 |
|
3099 | |||
3097 | rank = RANK_UNKNOWN |
|
3100 | rank = RANK_UNKNOWN | |
3098 | if self.feature_config.compute_rank: |
|
3101 | if self.feature_config.compute_rank: | |
3099 | if (p1r, p2r) == (nullrev, nullrev): |
|
3102 | if (p1r, p2r) == (nullrev, nullrev): | |
3100 | rank = 1 |
|
3103 | rank = 1 | |
3101 | elif p1r != nullrev and p2r == nullrev: |
|
3104 | elif p1r != nullrev and p2r == nullrev: | |
3102 | rank = 1 + self.fast_rank(p1r) |
|
3105 | rank = 1 + self.fast_rank(p1r) | |
3103 | elif p1r == nullrev and p2r != nullrev: |
|
3106 | elif p1r == nullrev and p2r != nullrev: | |
3104 | rank = 1 + self.fast_rank(p2r) |
|
3107 | rank = 1 + self.fast_rank(p2r) | |
3105 | else: # merge node |
|
3108 | else: # merge node | |
3106 | if rustdagop is not None and self.index.rust_ext_compat: |
|
3109 | if rustdagop is not None and self.index.rust_ext_compat: | |
3107 | rank = rustdagop.rank(self.index, p1r, p2r) |
|
3110 | rank = rustdagop.rank(self.index, p1r, p2r) | |
3108 | else: |
|
3111 | else: | |
3109 | pmin, pmax = sorted((p1r, p2r)) |
|
3112 | pmin, pmax = sorted((p1r, p2r)) | |
3110 | rank = 1 + self.fast_rank(pmax) |
|
3113 | rank = 1 + self.fast_rank(pmax) | |
3111 | rank += sum(1 for _ in self.findmissingrevs([pmax], [pmin])) |
|
3114 | rank += sum(1 for _ in self.findmissingrevs([pmax], [pmin])) | |
3112 |
|
3115 | |||
3113 | e = revlogutils.entry( |
|
3116 | e = revlogutils.entry( | |
3114 | flags=flags, |
|
3117 | flags=flags, | |
3115 | data_offset=offset, |
|
3118 | data_offset=offset, | |
3116 | data_compressed_length=deltainfo.deltalen, |
|
3119 | data_compressed_length=deltainfo.deltalen, | |
3117 | data_uncompressed_length=textlen, |
|
3120 | data_uncompressed_length=textlen, | |
3118 | data_compression_mode=compression_mode, |
|
3121 | data_compression_mode=compression_mode, | |
3119 | data_delta_base=deltainfo.base, |
|
3122 | data_delta_base=deltainfo.base, | |
3120 | link_rev=link, |
|
3123 | link_rev=link, | |
3121 | parent_rev_1=p1r, |
|
3124 | parent_rev_1=p1r, | |
3122 | parent_rev_2=p2r, |
|
3125 | parent_rev_2=p2r, | |
3123 | node_id=node, |
|
3126 | node_id=node, | |
3124 | sidedata_offset=sidedata_offset, |
|
3127 | sidedata_offset=sidedata_offset, | |
3125 | sidedata_compressed_length=len(serialized_sidedata), |
|
3128 | sidedata_compressed_length=len(serialized_sidedata), | |
3126 | sidedata_compression_mode=sidedata_compression_mode, |
|
3129 | sidedata_compression_mode=sidedata_compression_mode, | |
3127 | rank=rank, |
|
3130 | rank=rank, | |
3128 | ) |
|
3131 | ) | |
3129 |
|
3132 | |||
3130 | self.index.append(e) |
|
3133 | self.index.append(e) | |
3131 | entry = self.index.entry_binary(curr) |
|
3134 | entry = self.index.entry_binary(curr) | |
3132 | if curr == 0 and self._docket is None: |
|
3135 | if curr == 0 and self._docket is None: | |
3133 | header = self._format_flags | self._format_version |
|
3136 | header = self._format_flags | self._format_version | |
3134 | header = self.index.pack_header(header) |
|
3137 | header = self.index.pack_header(header) | |
3135 | entry = header + entry |
|
3138 | entry = header + entry | |
3136 | self._writeentry( |
|
3139 | self._writeentry( | |
3137 | transaction, |
|
3140 | transaction, | |
3138 | entry, |
|
3141 | entry, | |
3139 | deltainfo.data, |
|
3142 | deltainfo.data, | |
3140 | link, |
|
3143 | link, | |
3141 | offset, |
|
3144 | offset, | |
3142 | serialized_sidedata, |
|
3145 | serialized_sidedata, | |
3143 | sidedata_offset, |
|
3146 | sidedata_offset, | |
3144 | ) |
|
3147 | ) | |
3145 |
|
3148 | |||
3146 | rawtext = btext[0] |
|
3149 | rawtext = btext[0] | |
3147 |
|
3150 | |||
3148 | if alwayscache and rawtext is None: |
|
3151 | if alwayscache and rawtext is None: | |
3149 | rawtext = deltacomputer.buildtext(revinfo) |
|
3152 | rawtext = deltacomputer.buildtext(revinfo) | |
3150 |
|
3153 | |||
3151 | if type(rawtext) == bytes: # only accept immutable objects |
|
3154 | if type(rawtext) == bytes: # only accept immutable objects | |
3152 | self._revisioncache = (node, curr, rawtext) |
|
3155 | self._inner._revisioncache = (node, curr, rawtext) | |
3153 | self._chainbasecache[curr] = deltainfo.chainbase |
|
3156 | self._chainbasecache[curr] = deltainfo.chainbase | |
3154 | return curr |
|
3157 | return curr | |
3155 |
|
3158 | |||
3156 | def _get_data_offset(self, prev): |
|
3159 | def _get_data_offset(self, prev): | |
3157 | """Returns the current offset in the (in-transaction) data file. |
|
3160 | """Returns the current offset in the (in-transaction) data file. | |
3158 | Versions < 2 of the revlog can get this 0(1), revlog v2 needs a docket |
|
3161 | Versions < 2 of the revlog can get this 0(1), revlog v2 needs a docket | |
3159 | file to store that information: since sidedata can be rewritten to the |
|
3162 | file to store that information: since sidedata can be rewritten to the | |
3160 | end of the data file within a transaction, you can have cases where, for |
|
3163 | end of the data file within a transaction, you can have cases where, for | |
3161 | example, rev `n` does not have sidedata while rev `n - 1` does, leading |
|
3164 | example, rev `n` does not have sidedata while rev `n - 1` does, leading | |
3162 | to `n - 1`'s sidedata being written after `n`'s data. |
|
3165 | to `n - 1`'s sidedata being written after `n`'s data. | |
3163 |
|
3166 | |||
3164 | TODO cache this in a docket file before getting out of experimental.""" |
|
3167 | TODO cache this in a docket file before getting out of experimental.""" | |
3165 | if self._docket is None: |
|
3168 | if self._docket is None: | |
3166 | return self.end(prev) |
|
3169 | return self.end(prev) | |
3167 | else: |
|
3170 | else: | |
3168 | return self._docket.data_end |
|
3171 | return self._docket.data_end | |
3169 |
|
3172 | |||
3170 | def _writeentry( |
|
3173 | def _writeentry( | |
3171 | self, transaction, entry, data, link, offset, sidedata, sidedata_offset |
|
3174 | self, transaction, entry, data, link, offset, sidedata, sidedata_offset | |
3172 | ): |
|
3175 | ): | |
3173 | # Files opened in a+ mode have inconsistent behavior on various |
|
3176 | # Files opened in a+ mode have inconsistent behavior on various | |
3174 | # platforms. Windows requires that a file positioning call be made |
|
3177 | # platforms. Windows requires that a file positioning call be made | |
3175 | # when the file handle transitions between reads and writes. See |
|
3178 | # when the file handle transitions between reads and writes. See | |
3176 | # 3686fa2b8eee and the mixedfilemodewrapper in windows.py. On other |
|
3179 | # 3686fa2b8eee and the mixedfilemodewrapper in windows.py. On other | |
3177 | # platforms, Python or the platform itself can be buggy. Some versions |
|
3180 | # platforms, Python or the platform itself can be buggy. Some versions | |
3178 | # of Solaris have been observed to not append at the end of the file |
|
3181 | # of Solaris have been observed to not append at the end of the file | |
3179 | # if the file was seeked to before the end. See issue4943 for more. |
|
3182 | # if the file was seeked to before the end. See issue4943 for more. | |
3180 | # |
|
3183 | # | |
3181 | # We work around this issue by inserting a seek() before writing. |
|
3184 | # We work around this issue by inserting a seek() before writing. | |
3182 | # Note: This is likely not necessary on Python 3. However, because |
|
3185 | # Note: This is likely not necessary on Python 3. However, because | |
3183 | # the file handle is reused for reads and may be seeked there, we need |
|
3186 | # the file handle is reused for reads and may be seeked there, we need | |
3184 | # to be careful before changing this. |
|
3187 | # to be careful before changing this. | |
3185 | if self._inner._writinghandles is None: |
|
3188 | if self._inner._writinghandles is None: | |
3186 | msg = b'adding revision outside `revlog._writing` context' |
|
3189 | msg = b'adding revision outside `revlog._writing` context' | |
3187 | raise error.ProgrammingError(msg) |
|
3190 | raise error.ProgrammingError(msg) | |
3188 | ifh, dfh, sdfh = self._inner._writinghandles |
|
3191 | ifh, dfh, sdfh = self._inner._writinghandles | |
3189 | if self._docket is None: |
|
3192 | if self._docket is None: | |
3190 | ifh.seek(0, os.SEEK_END) |
|
3193 | ifh.seek(0, os.SEEK_END) | |
3191 | else: |
|
3194 | else: | |
3192 | ifh.seek(self._docket.index_end, os.SEEK_SET) |
|
3195 | ifh.seek(self._docket.index_end, os.SEEK_SET) | |
3193 | if dfh: |
|
3196 | if dfh: | |
3194 | if self._docket is None: |
|
3197 | if self._docket is None: | |
3195 | dfh.seek(0, os.SEEK_END) |
|
3198 | dfh.seek(0, os.SEEK_END) | |
3196 | else: |
|
3199 | else: | |
3197 | dfh.seek(self._docket.data_end, os.SEEK_SET) |
|
3200 | dfh.seek(self._docket.data_end, os.SEEK_SET) | |
3198 | if sdfh: |
|
3201 | if sdfh: | |
3199 | sdfh.seek(self._docket.sidedata_end, os.SEEK_SET) |
|
3202 | sdfh.seek(self._docket.sidedata_end, os.SEEK_SET) | |
3200 |
|
3203 | |||
3201 | curr = len(self) - 1 |
|
3204 | curr = len(self) - 1 | |
3202 | if not self._inline: |
|
3205 | if not self._inline: | |
3203 | transaction.add(self._datafile, offset) |
|
3206 | transaction.add(self._datafile, offset) | |
3204 | if self._sidedatafile: |
|
3207 | if self._sidedatafile: | |
3205 | transaction.add(self._sidedatafile, sidedata_offset) |
|
3208 | transaction.add(self._sidedatafile, sidedata_offset) | |
3206 | transaction.add(self._indexfile, curr * len(entry)) |
|
3209 | transaction.add(self._indexfile, curr * len(entry)) | |
3207 | if data[0]: |
|
3210 | if data[0]: | |
3208 | dfh.write(data[0]) |
|
3211 | dfh.write(data[0]) | |
3209 | dfh.write(data[1]) |
|
3212 | dfh.write(data[1]) | |
3210 | if sidedata: |
|
3213 | if sidedata: | |
3211 | sdfh.write(sidedata) |
|
3214 | sdfh.write(sidedata) | |
3212 | ifh.write(entry) |
|
3215 | ifh.write(entry) | |
3213 | else: |
|
3216 | else: | |
3214 | offset += curr * self.index.entry_size |
|
3217 | offset += curr * self.index.entry_size | |
3215 | transaction.add(self._indexfile, offset) |
|
3218 | transaction.add(self._indexfile, offset) | |
3216 | ifh.write(entry) |
|
3219 | ifh.write(entry) | |
3217 | ifh.write(data[0]) |
|
3220 | ifh.write(data[0]) | |
3218 | ifh.write(data[1]) |
|
3221 | ifh.write(data[1]) | |
3219 | assert not sidedata |
|
3222 | assert not sidedata | |
3220 | self._enforceinlinesize(transaction) |
|
3223 | self._enforceinlinesize(transaction) | |
3221 | if self._docket is not None: |
|
3224 | if self._docket is not None: | |
3222 | # revlog-v2 always has 3 writing handles, help Pytype |
|
3225 | # revlog-v2 always has 3 writing handles, help Pytype | |
3223 | wh1 = self._inner._writinghandles[0] |
|
3226 | wh1 = self._inner._writinghandles[0] | |
3224 | wh2 = self._inner._writinghandles[1] |
|
3227 | wh2 = self._inner._writinghandles[1] | |
3225 | wh3 = self._inner._writinghandles[2] |
|
3228 | wh3 = self._inner._writinghandles[2] | |
3226 | assert wh1 is not None |
|
3229 | assert wh1 is not None | |
3227 | assert wh2 is not None |
|
3230 | assert wh2 is not None | |
3228 | assert wh3 is not None |
|
3231 | assert wh3 is not None | |
3229 | self._docket.index_end = wh1.tell() |
|
3232 | self._docket.index_end = wh1.tell() | |
3230 | self._docket.data_end = wh2.tell() |
|
3233 | self._docket.data_end = wh2.tell() | |
3231 | self._docket.sidedata_end = wh3.tell() |
|
3234 | self._docket.sidedata_end = wh3.tell() | |
3232 |
|
3235 | |||
3233 | nodemaputil.setup_persistent_nodemap(transaction, self) |
|
3236 | nodemaputil.setup_persistent_nodemap(transaction, self) | |
3234 |
|
3237 | |||
3235 | def addgroup( |
|
3238 | def addgroup( | |
3236 | self, |
|
3239 | self, | |
3237 | deltas, |
|
3240 | deltas, | |
3238 | linkmapper, |
|
3241 | linkmapper, | |
3239 | transaction, |
|
3242 | transaction, | |
3240 | alwayscache=False, |
|
3243 | alwayscache=False, | |
3241 | addrevisioncb=None, |
|
3244 | addrevisioncb=None, | |
3242 | duplicaterevisioncb=None, |
|
3245 | duplicaterevisioncb=None, | |
3243 | debug_info=None, |
|
3246 | debug_info=None, | |
3244 | delta_base_reuse_policy=None, |
|
3247 | delta_base_reuse_policy=None, | |
3245 | ): |
|
3248 | ): | |
3246 | """ |
|
3249 | """ | |
3247 | add a delta group |
|
3250 | add a delta group | |
3248 |
|
3251 | |||
3249 | given a set of deltas, add them to the revision log. the |
|
3252 | given a set of deltas, add them to the revision log. the | |
3250 | first delta is against its parent, which should be in our |
|
3253 | first delta is against its parent, which should be in our | |
3251 | log, the rest are against the previous delta. |
|
3254 | log, the rest are against the previous delta. | |
3252 |
|
3255 | |||
3253 | If ``addrevisioncb`` is defined, it will be called with arguments of |
|
3256 | If ``addrevisioncb`` is defined, it will be called with arguments of | |
3254 | this revlog and the node that was added. |
|
3257 | this revlog and the node that was added. | |
3255 | """ |
|
3258 | """ | |
3256 |
|
3259 | |||
3257 | if self._adding_group: |
|
3260 | if self._adding_group: | |
3258 | raise error.ProgrammingError(b'cannot nest addgroup() calls') |
|
3261 | raise error.ProgrammingError(b'cannot nest addgroup() calls') | |
3259 |
|
3262 | |||
3260 | # read the default delta-base reuse policy from revlog config if the |
|
3263 | # read the default delta-base reuse policy from revlog config if the | |
3261 | # group did not specify one. |
|
3264 | # group did not specify one. | |
3262 | if delta_base_reuse_policy is None: |
|
3265 | if delta_base_reuse_policy is None: | |
3263 | if ( |
|
3266 | if ( | |
3264 | self.delta_config.general_delta |
|
3267 | self.delta_config.general_delta | |
3265 | and self.delta_config.lazy_delta_base |
|
3268 | and self.delta_config.lazy_delta_base | |
3266 | ): |
|
3269 | ): | |
3267 | delta_base_reuse_policy = DELTA_BASE_REUSE_TRY |
|
3270 | delta_base_reuse_policy = DELTA_BASE_REUSE_TRY | |
3268 | else: |
|
3271 | else: | |
3269 | delta_base_reuse_policy = DELTA_BASE_REUSE_NO |
|
3272 | delta_base_reuse_policy = DELTA_BASE_REUSE_NO | |
3270 |
|
3273 | |||
3271 | self._adding_group = True |
|
3274 | self._adding_group = True | |
3272 | empty = True |
|
3275 | empty = True | |
3273 | try: |
|
3276 | try: | |
3274 | with self._writing(transaction): |
|
3277 | with self._writing(transaction): | |
3275 | write_debug = None |
|
3278 | write_debug = None | |
3276 | if self.delta_config.debug_delta: |
|
3279 | if self.delta_config.debug_delta: | |
3277 | write_debug = transaction._report |
|
3280 | write_debug = transaction._report | |
3278 | deltacomputer = deltautil.deltacomputer( |
|
3281 | deltacomputer = deltautil.deltacomputer( | |
3279 | self, |
|
3282 | self, | |
3280 | write_debug=write_debug, |
|
3283 | write_debug=write_debug, | |
3281 | debug_info=debug_info, |
|
3284 | debug_info=debug_info, | |
3282 | ) |
|
3285 | ) | |
3283 | # loop through our set of deltas |
|
3286 | # loop through our set of deltas | |
3284 | for data in deltas: |
|
3287 | for data in deltas: | |
3285 | ( |
|
3288 | ( | |
3286 | node, |
|
3289 | node, | |
3287 | p1, |
|
3290 | p1, | |
3288 | p2, |
|
3291 | p2, | |
3289 | linknode, |
|
3292 | linknode, | |
3290 | deltabase, |
|
3293 | deltabase, | |
3291 | delta, |
|
3294 | delta, | |
3292 | flags, |
|
3295 | flags, | |
3293 | sidedata, |
|
3296 | sidedata, | |
3294 | ) = data |
|
3297 | ) = data | |
3295 | link = linkmapper(linknode) |
|
3298 | link = linkmapper(linknode) | |
3296 | flags = flags or REVIDX_DEFAULT_FLAGS |
|
3299 | flags = flags or REVIDX_DEFAULT_FLAGS | |
3297 |
|
3300 | |||
3298 | rev = self.index.get_rev(node) |
|
3301 | rev = self.index.get_rev(node) | |
3299 | if rev is not None: |
|
3302 | if rev is not None: | |
3300 | # this can happen if two branches make the same change |
|
3303 | # this can happen if two branches make the same change | |
3301 | self._nodeduplicatecallback(transaction, rev) |
|
3304 | self._nodeduplicatecallback(transaction, rev) | |
3302 | if duplicaterevisioncb: |
|
3305 | if duplicaterevisioncb: | |
3303 | duplicaterevisioncb(self, rev) |
|
3306 | duplicaterevisioncb(self, rev) | |
3304 | empty = False |
|
3307 | empty = False | |
3305 | continue |
|
3308 | continue | |
3306 |
|
3309 | |||
3307 | for p in (p1, p2): |
|
3310 | for p in (p1, p2): | |
3308 | if not self.index.has_node(p): |
|
3311 | if not self.index.has_node(p): | |
3309 | raise error.LookupError( |
|
3312 | raise error.LookupError( | |
3310 | p, self.radix, _(b'unknown parent') |
|
3313 | p, self.radix, _(b'unknown parent') | |
3311 | ) |
|
3314 | ) | |
3312 |
|
3315 | |||
3313 | if not self.index.has_node(deltabase): |
|
3316 | if not self.index.has_node(deltabase): | |
3314 | raise error.LookupError( |
|
3317 | raise error.LookupError( | |
3315 | deltabase, self.display_id, _(b'unknown delta base') |
|
3318 | deltabase, self.display_id, _(b'unknown delta base') | |
3316 | ) |
|
3319 | ) | |
3317 |
|
3320 | |||
3318 | baserev = self.rev(deltabase) |
|
3321 | baserev = self.rev(deltabase) | |
3319 |
|
3322 | |||
3320 | if baserev != nullrev and self.iscensored(baserev): |
|
3323 | if baserev != nullrev and self.iscensored(baserev): | |
3321 | # if base is censored, delta must be full replacement in a |
|
3324 | # if base is censored, delta must be full replacement in a | |
3322 | # single patch operation |
|
3325 | # single patch operation | |
3323 | hlen = struct.calcsize(b">lll") |
|
3326 | hlen = struct.calcsize(b">lll") | |
3324 | oldlen = self.rawsize(baserev) |
|
3327 | oldlen = self.rawsize(baserev) | |
3325 | newlen = len(delta) - hlen |
|
3328 | newlen = len(delta) - hlen | |
3326 | if delta[:hlen] != mdiff.replacediffheader( |
|
3329 | if delta[:hlen] != mdiff.replacediffheader( | |
3327 | oldlen, newlen |
|
3330 | oldlen, newlen | |
3328 | ): |
|
3331 | ): | |
3329 | raise error.CensoredBaseError( |
|
3332 | raise error.CensoredBaseError( | |
3330 | self.display_id, self.node(baserev) |
|
3333 | self.display_id, self.node(baserev) | |
3331 | ) |
|
3334 | ) | |
3332 |
|
3335 | |||
3333 | if not flags and self._peek_iscensored(baserev, delta): |
|
3336 | if not flags and self._peek_iscensored(baserev, delta): | |
3334 | flags |= REVIDX_ISCENSORED |
|
3337 | flags |= REVIDX_ISCENSORED | |
3335 |
|
3338 | |||
3336 | # We assume consumers of addrevisioncb will want to retrieve |
|
3339 | # We assume consumers of addrevisioncb will want to retrieve | |
3337 | # the added revision, which will require a call to |
|
3340 | # the added revision, which will require a call to | |
3338 | # revision(). revision() will fast path if there is a cache |
|
3341 | # revision(). revision() will fast path if there is a cache | |
3339 | # hit. So, we tell _addrevision() to always cache in this case. |
|
3342 | # hit. So, we tell _addrevision() to always cache in this case. | |
3340 | # We're only using addgroup() in the context of changegroup |
|
3343 | # We're only using addgroup() in the context of changegroup | |
3341 | # generation so the revision data can always be handled as raw |
|
3344 | # generation so the revision data can always be handled as raw | |
3342 | # by the flagprocessor. |
|
3345 | # by the flagprocessor. | |
3343 | rev = self._addrevision( |
|
3346 | rev = self._addrevision( | |
3344 | node, |
|
3347 | node, | |
3345 | None, |
|
3348 | None, | |
3346 | transaction, |
|
3349 | transaction, | |
3347 | link, |
|
3350 | link, | |
3348 | p1, |
|
3351 | p1, | |
3349 | p2, |
|
3352 | p2, | |
3350 | flags, |
|
3353 | flags, | |
3351 | (baserev, delta, delta_base_reuse_policy), |
|
3354 | (baserev, delta, delta_base_reuse_policy), | |
3352 | alwayscache=alwayscache, |
|
3355 | alwayscache=alwayscache, | |
3353 | deltacomputer=deltacomputer, |
|
3356 | deltacomputer=deltacomputer, | |
3354 | sidedata=sidedata, |
|
3357 | sidedata=sidedata, | |
3355 | ) |
|
3358 | ) | |
3356 |
|
3359 | |||
3357 | if addrevisioncb: |
|
3360 | if addrevisioncb: | |
3358 | addrevisioncb(self, rev) |
|
3361 | addrevisioncb(self, rev) | |
3359 | empty = False |
|
3362 | empty = False | |
3360 | finally: |
|
3363 | finally: | |
3361 | self._adding_group = False |
|
3364 | self._adding_group = False | |
3362 | return not empty |
|
3365 | return not empty | |
3363 |
|
3366 | |||
3364 | def iscensored(self, rev): |
|
3367 | def iscensored(self, rev): | |
3365 | """Check if a file revision is censored.""" |
|
3368 | """Check if a file revision is censored.""" | |
3366 | if not self.feature_config.censorable: |
|
3369 | if not self.feature_config.censorable: | |
3367 | return False |
|
3370 | return False | |
3368 |
|
3371 | |||
3369 | return self.flags(rev) & REVIDX_ISCENSORED |
|
3372 | return self.flags(rev) & REVIDX_ISCENSORED | |
3370 |
|
3373 | |||
3371 | def _peek_iscensored(self, baserev, delta): |
|
3374 | def _peek_iscensored(self, baserev, delta): | |
3372 | """Quickly check if a delta produces a censored revision.""" |
|
3375 | """Quickly check if a delta produces a censored revision.""" | |
3373 | if not self.feature_config.censorable: |
|
3376 | if not self.feature_config.censorable: | |
3374 | return False |
|
3377 | return False | |
3375 |
|
3378 | |||
3376 | return storageutil.deltaiscensored(delta, baserev, self.rawsize) |
|
3379 | return storageutil.deltaiscensored(delta, baserev, self.rawsize) | |
3377 |
|
3380 | |||
3378 | def getstrippoint(self, minlink): |
|
3381 | def getstrippoint(self, minlink): | |
3379 | """find the minimum rev that must be stripped to strip the linkrev |
|
3382 | """find the minimum rev that must be stripped to strip the linkrev | |
3380 |
|
3383 | |||
3381 | Returns a tuple containing the minimum rev and a set of all revs that |
|
3384 | Returns a tuple containing the minimum rev and a set of all revs that | |
3382 | have linkrevs that will be broken by this strip. |
|
3385 | have linkrevs that will be broken by this strip. | |
3383 | """ |
|
3386 | """ | |
3384 | return storageutil.resolvestripinfo( |
|
3387 | return storageutil.resolvestripinfo( | |
3385 | minlink, |
|
3388 | minlink, | |
3386 | len(self) - 1, |
|
3389 | len(self) - 1, | |
3387 | self.headrevs(), |
|
3390 | self.headrevs(), | |
3388 | self.linkrev, |
|
3391 | self.linkrev, | |
3389 | self.parentrevs, |
|
3392 | self.parentrevs, | |
3390 | ) |
|
3393 | ) | |
3391 |
|
3394 | |||
3392 | def strip(self, minlink, transaction): |
|
3395 | def strip(self, minlink, transaction): | |
3393 | """truncate the revlog on the first revision with a linkrev >= minlink |
|
3396 | """truncate the revlog on the first revision with a linkrev >= minlink | |
3394 |
|
3397 | |||
3395 | This function is called when we're stripping revision minlink and |
|
3398 | This function is called when we're stripping revision minlink and | |
3396 | its descendants from the repository. |
|
3399 | its descendants from the repository. | |
3397 |
|
3400 | |||
3398 | We have to remove all revisions with linkrev >= minlink, because |
|
3401 | We have to remove all revisions with linkrev >= minlink, because | |
3399 | the equivalent changelog revisions will be renumbered after the |
|
3402 | the equivalent changelog revisions will be renumbered after the | |
3400 | strip. |
|
3403 | strip. | |
3401 |
|
3404 | |||
3402 | So we truncate the revlog on the first of these revisions, and |
|
3405 | So we truncate the revlog on the first of these revisions, and | |
3403 | trust that the caller has saved the revisions that shouldn't be |
|
3406 | trust that the caller has saved the revisions that shouldn't be | |
3404 | removed and that it'll re-add them after this truncation. |
|
3407 | removed and that it'll re-add them after this truncation. | |
3405 | """ |
|
3408 | """ | |
3406 | if len(self) == 0: |
|
3409 | if len(self) == 0: | |
3407 | return |
|
3410 | return | |
3408 |
|
3411 | |||
3409 | rev, _ = self.getstrippoint(minlink) |
|
3412 | rev, _ = self.getstrippoint(minlink) | |
3410 | if rev == len(self): |
|
3413 | if rev == len(self): | |
3411 | return |
|
3414 | return | |
3412 |
|
3415 | |||
3413 | # first truncate the files on disk |
|
3416 | # first truncate the files on disk | |
3414 | data_end = self.start(rev) |
|
3417 | data_end = self.start(rev) | |
3415 | if not self._inline: |
|
3418 | if not self._inline: | |
3416 | transaction.add(self._datafile, data_end) |
|
3419 | transaction.add(self._datafile, data_end) | |
3417 | end = rev * self.index.entry_size |
|
3420 | end = rev * self.index.entry_size | |
3418 | else: |
|
3421 | else: | |
3419 | end = data_end + (rev * self.index.entry_size) |
|
3422 | end = data_end + (rev * self.index.entry_size) | |
3420 |
|
3423 | |||
3421 | if self._sidedatafile: |
|
3424 | if self._sidedatafile: | |
3422 | sidedata_end = self.sidedata_cut_off(rev) |
|
3425 | sidedata_end = self.sidedata_cut_off(rev) | |
3423 | transaction.add(self._sidedatafile, sidedata_end) |
|
3426 | transaction.add(self._sidedatafile, sidedata_end) | |
3424 |
|
3427 | |||
3425 | transaction.add(self._indexfile, end) |
|
3428 | transaction.add(self._indexfile, end) | |
3426 | if self._docket is not None: |
|
3429 | if self._docket is not None: | |
3427 | # XXX we could, leverage the docket while stripping. However it is |
|
3430 | # XXX we could, leverage the docket while stripping. However it is | |
3428 | # not powerfull enough at the time of this comment |
|
3431 | # not powerfull enough at the time of this comment | |
3429 | self._docket.index_end = end |
|
3432 | self._docket.index_end = end | |
3430 | self._docket.data_end = data_end |
|
3433 | self._docket.data_end = data_end | |
3431 | self._docket.sidedata_end = sidedata_end |
|
3434 | self._docket.sidedata_end = sidedata_end | |
3432 | self._docket.write(transaction, stripping=True) |
|
3435 | self._docket.write(transaction, stripping=True) | |
3433 |
|
3436 | |||
3434 | # then reset internal state in memory to forget those revisions |
|
3437 | # then reset internal state in memory to forget those revisions | |
3435 | self._revisioncache = None |
|
3438 | self._inner._revisioncache = None | |
3436 | self._chaininfocache = util.lrucachedict(500) |
|
3439 | self._chaininfocache = util.lrucachedict(500) | |
3437 | self._inner._segmentfile.clear_cache() |
|
3440 | self._inner._segmentfile.clear_cache() | |
3438 | self._inner._segmentfile_sidedata.clear_cache() |
|
3441 | self._inner._segmentfile_sidedata.clear_cache() | |
3439 |
|
3442 | |||
3440 | del self.index[rev:-1] |
|
3443 | del self.index[rev:-1] | |
3441 |
|
3444 | |||
3442 | def checksize(self): |
|
3445 | def checksize(self): | |
3443 | """Check size of index and data files |
|
3446 | """Check size of index and data files | |
3444 |
|
3447 | |||
3445 | return a (dd, di) tuple. |
|
3448 | return a (dd, di) tuple. | |
3446 | - dd: extra bytes for the "data" file |
|
3449 | - dd: extra bytes for the "data" file | |
3447 | - di: extra bytes for the "index" file |
|
3450 | - di: extra bytes for the "index" file | |
3448 |
|
3451 | |||
3449 | A healthy revlog will return (0, 0). |
|
3452 | A healthy revlog will return (0, 0). | |
3450 | """ |
|
3453 | """ | |
3451 | expected = 0 |
|
3454 | expected = 0 | |
3452 | if len(self): |
|
3455 | if len(self): | |
3453 | expected = max(0, self.end(len(self) - 1)) |
|
3456 | expected = max(0, self.end(len(self) - 1)) | |
3454 |
|
3457 | |||
3455 | try: |
|
3458 | try: | |
3456 | with self._datafp() as f: |
|
3459 | with self._datafp() as f: | |
3457 | f.seek(0, io.SEEK_END) |
|
3460 | f.seek(0, io.SEEK_END) | |
3458 | actual = f.tell() |
|
3461 | actual = f.tell() | |
3459 | dd = actual - expected |
|
3462 | dd = actual - expected | |
3460 | except FileNotFoundError: |
|
3463 | except FileNotFoundError: | |
3461 | dd = 0 |
|
3464 | dd = 0 | |
3462 |
|
3465 | |||
3463 | try: |
|
3466 | try: | |
3464 | f = self.opener(self._indexfile) |
|
3467 | f = self.opener(self._indexfile) | |
3465 | f.seek(0, io.SEEK_END) |
|
3468 | f.seek(0, io.SEEK_END) | |
3466 | actual = f.tell() |
|
3469 | actual = f.tell() | |
3467 | f.close() |
|
3470 | f.close() | |
3468 | s = self.index.entry_size |
|
3471 | s = self.index.entry_size | |
3469 | i = max(0, actual // s) |
|
3472 | i = max(0, actual // s) | |
3470 | di = actual - (i * s) |
|
3473 | di = actual - (i * s) | |
3471 | if self._inline: |
|
3474 | if self._inline: | |
3472 | databytes = 0 |
|
3475 | databytes = 0 | |
3473 | for r in self: |
|
3476 | for r in self: | |
3474 | databytes += max(0, self.length(r)) |
|
3477 | databytes += max(0, self.length(r)) | |
3475 | dd = 0 |
|
3478 | dd = 0 | |
3476 | di = actual - len(self) * s - databytes |
|
3479 | di = actual - len(self) * s - databytes | |
3477 | except FileNotFoundError: |
|
3480 | except FileNotFoundError: | |
3478 | di = 0 |
|
3481 | di = 0 | |
3479 |
|
3482 | |||
3480 | return (dd, di) |
|
3483 | return (dd, di) | |
3481 |
|
3484 | |||
3482 | def files(self): |
|
3485 | def files(self): | |
3483 | """return list of files that compose this revlog""" |
|
3486 | """return list of files that compose this revlog""" | |
3484 | res = [self._indexfile] |
|
3487 | res = [self._indexfile] | |
3485 | if self._docket_file is None: |
|
3488 | if self._docket_file is None: | |
3486 | if not self._inline: |
|
3489 | if not self._inline: | |
3487 | res.append(self._datafile) |
|
3490 | res.append(self._datafile) | |
3488 | else: |
|
3491 | else: | |
3489 | res.append(self._docket_file) |
|
3492 | res.append(self._docket_file) | |
3490 | res.extend(self._docket.old_index_filepaths(include_empty=False)) |
|
3493 | res.extend(self._docket.old_index_filepaths(include_empty=False)) | |
3491 | if self._docket.data_end: |
|
3494 | if self._docket.data_end: | |
3492 | res.append(self._datafile) |
|
3495 | res.append(self._datafile) | |
3493 | res.extend(self._docket.old_data_filepaths(include_empty=False)) |
|
3496 | res.extend(self._docket.old_data_filepaths(include_empty=False)) | |
3494 | if self._docket.sidedata_end: |
|
3497 | if self._docket.sidedata_end: | |
3495 | res.append(self._sidedatafile) |
|
3498 | res.append(self._sidedatafile) | |
3496 | res.extend(self._docket.old_sidedata_filepaths(include_empty=False)) |
|
3499 | res.extend(self._docket.old_sidedata_filepaths(include_empty=False)) | |
3497 | return res |
|
3500 | return res | |
3498 |
|
3501 | |||
3499 | def emitrevisions( |
|
3502 | def emitrevisions( | |
3500 | self, |
|
3503 | self, | |
3501 | nodes, |
|
3504 | nodes, | |
3502 | nodesorder=None, |
|
3505 | nodesorder=None, | |
3503 | revisiondata=False, |
|
3506 | revisiondata=False, | |
3504 | assumehaveparentrevisions=False, |
|
3507 | assumehaveparentrevisions=False, | |
3505 | deltamode=repository.CG_DELTAMODE_STD, |
|
3508 | deltamode=repository.CG_DELTAMODE_STD, | |
3506 | sidedata_helpers=None, |
|
3509 | sidedata_helpers=None, | |
3507 | debug_info=None, |
|
3510 | debug_info=None, | |
3508 | ): |
|
3511 | ): | |
3509 | if nodesorder not in (b'nodes', b'storage', b'linear', None): |
|
3512 | if nodesorder not in (b'nodes', b'storage', b'linear', None): | |
3510 | raise error.ProgrammingError( |
|
3513 | raise error.ProgrammingError( | |
3511 | b'unhandled value for nodesorder: %s' % nodesorder |
|
3514 | b'unhandled value for nodesorder: %s' % nodesorder | |
3512 | ) |
|
3515 | ) | |
3513 |
|
3516 | |||
3514 | if nodesorder is None and not self.delta_config.general_delta: |
|
3517 | if nodesorder is None and not self.delta_config.general_delta: | |
3515 | nodesorder = b'storage' |
|
3518 | nodesorder = b'storage' | |
3516 |
|
3519 | |||
3517 | if ( |
|
3520 | if ( | |
3518 | not self._storedeltachains |
|
3521 | not self._storedeltachains | |
3519 | and deltamode != repository.CG_DELTAMODE_PREV |
|
3522 | and deltamode != repository.CG_DELTAMODE_PREV | |
3520 | ): |
|
3523 | ): | |
3521 | deltamode = repository.CG_DELTAMODE_FULL |
|
3524 | deltamode = repository.CG_DELTAMODE_FULL | |
3522 |
|
3525 | |||
3523 | return storageutil.emitrevisions( |
|
3526 | return storageutil.emitrevisions( | |
3524 | self, |
|
3527 | self, | |
3525 | nodes, |
|
3528 | nodes, | |
3526 | nodesorder, |
|
3529 | nodesorder, | |
3527 | revlogrevisiondelta, |
|
3530 | revlogrevisiondelta, | |
3528 | deltaparentfn=self.deltaparent, |
|
3531 | deltaparentfn=self.deltaparent, | |
3529 | candeltafn=self._candelta, |
|
3532 | candeltafn=self._candelta, | |
3530 | rawsizefn=self.rawsize, |
|
3533 | rawsizefn=self.rawsize, | |
3531 | revdifffn=self.revdiff, |
|
3534 | revdifffn=self.revdiff, | |
3532 | flagsfn=self.flags, |
|
3535 | flagsfn=self.flags, | |
3533 | deltamode=deltamode, |
|
3536 | deltamode=deltamode, | |
3534 | revisiondata=revisiondata, |
|
3537 | revisiondata=revisiondata, | |
3535 | assumehaveparentrevisions=assumehaveparentrevisions, |
|
3538 | assumehaveparentrevisions=assumehaveparentrevisions, | |
3536 | sidedata_helpers=sidedata_helpers, |
|
3539 | sidedata_helpers=sidedata_helpers, | |
3537 | debug_info=debug_info, |
|
3540 | debug_info=debug_info, | |
3538 | ) |
|
3541 | ) | |
3539 |
|
3542 | |||
3540 | DELTAREUSEALWAYS = b'always' |
|
3543 | DELTAREUSEALWAYS = b'always' | |
3541 | DELTAREUSESAMEREVS = b'samerevs' |
|
3544 | DELTAREUSESAMEREVS = b'samerevs' | |
3542 | DELTAREUSENEVER = b'never' |
|
3545 | DELTAREUSENEVER = b'never' | |
3543 |
|
3546 | |||
3544 | DELTAREUSEFULLADD = b'fulladd' |
|
3547 | DELTAREUSEFULLADD = b'fulladd' | |
3545 |
|
3548 | |||
3546 | DELTAREUSEALL = {b'always', b'samerevs', b'never', b'fulladd'} |
|
3549 | DELTAREUSEALL = {b'always', b'samerevs', b'never', b'fulladd'} | |
3547 |
|
3550 | |||
3548 | def clone( |
|
3551 | def clone( | |
3549 | self, |
|
3552 | self, | |
3550 | tr, |
|
3553 | tr, | |
3551 | destrevlog, |
|
3554 | destrevlog, | |
3552 | addrevisioncb=None, |
|
3555 | addrevisioncb=None, | |
3553 | deltareuse=DELTAREUSESAMEREVS, |
|
3556 | deltareuse=DELTAREUSESAMEREVS, | |
3554 | forcedeltabothparents=None, |
|
3557 | forcedeltabothparents=None, | |
3555 | sidedata_helpers=None, |
|
3558 | sidedata_helpers=None, | |
3556 | ): |
|
3559 | ): | |
3557 | """Copy this revlog to another, possibly with format changes. |
|
3560 | """Copy this revlog to another, possibly with format changes. | |
3558 |
|
3561 | |||
3559 | The destination revlog will contain the same revisions and nodes. |
|
3562 | The destination revlog will contain the same revisions and nodes. | |
3560 | However, it may not be bit-for-bit identical due to e.g. delta encoding |
|
3563 | However, it may not be bit-for-bit identical due to e.g. delta encoding | |
3561 | differences. |
|
3564 | differences. | |
3562 |
|
3565 | |||
3563 | The ``deltareuse`` argument control how deltas from the existing revlog |
|
3566 | The ``deltareuse`` argument control how deltas from the existing revlog | |
3564 | are preserved in the destination revlog. The argument can have the |
|
3567 | are preserved in the destination revlog. The argument can have the | |
3565 | following values: |
|
3568 | following values: | |
3566 |
|
3569 | |||
3567 | DELTAREUSEALWAYS |
|
3570 | DELTAREUSEALWAYS | |
3568 | Deltas will always be reused (if possible), even if the destination |
|
3571 | Deltas will always be reused (if possible), even if the destination | |
3569 | revlog would not select the same revisions for the delta. This is the |
|
3572 | revlog would not select the same revisions for the delta. This is the | |
3570 | fastest mode of operation. |
|
3573 | fastest mode of operation. | |
3571 | DELTAREUSESAMEREVS |
|
3574 | DELTAREUSESAMEREVS | |
3572 | Deltas will be reused if the destination revlog would pick the same |
|
3575 | Deltas will be reused if the destination revlog would pick the same | |
3573 | revisions for the delta. This mode strikes a balance between speed |
|
3576 | revisions for the delta. This mode strikes a balance between speed | |
3574 | and optimization. |
|
3577 | and optimization. | |
3575 | DELTAREUSENEVER |
|
3578 | DELTAREUSENEVER | |
3576 | Deltas will never be reused. This is the slowest mode of execution. |
|
3579 | Deltas will never be reused. This is the slowest mode of execution. | |
3577 | This mode can be used to recompute deltas (e.g. if the diff/delta |
|
3580 | This mode can be used to recompute deltas (e.g. if the diff/delta | |
3578 | algorithm changes). |
|
3581 | algorithm changes). | |
3579 | DELTAREUSEFULLADD |
|
3582 | DELTAREUSEFULLADD | |
3580 | Revision will be re-added as if their were new content. This is |
|
3583 | Revision will be re-added as if their were new content. This is | |
3581 | slower than DELTAREUSEALWAYS but allow more mechanism to kicks in. |
|
3584 | slower than DELTAREUSEALWAYS but allow more mechanism to kicks in. | |
3582 | eg: large file detection and handling. |
|
3585 | eg: large file detection and handling. | |
3583 |
|
3586 | |||
3584 | Delta computation can be slow, so the choice of delta reuse policy can |
|
3587 | Delta computation can be slow, so the choice of delta reuse policy can | |
3585 | significantly affect run time. |
|
3588 | significantly affect run time. | |
3586 |
|
3589 | |||
3587 | The default policy (``DELTAREUSESAMEREVS``) strikes a balance between |
|
3590 | The default policy (``DELTAREUSESAMEREVS``) strikes a balance between | |
3588 | two extremes. Deltas will be reused if they are appropriate. But if the |
|
3591 | two extremes. Deltas will be reused if they are appropriate. But if the | |
3589 | delta could choose a better revision, it will do so. This means if you |
|
3592 | delta could choose a better revision, it will do so. This means if you | |
3590 | are converting a non-generaldelta revlog to a generaldelta revlog, |
|
3593 | are converting a non-generaldelta revlog to a generaldelta revlog, | |
3591 | deltas will be recomputed if the delta's parent isn't a parent of the |
|
3594 | deltas will be recomputed if the delta's parent isn't a parent of the | |
3592 | revision. |
|
3595 | revision. | |
3593 |
|
3596 | |||
3594 | In addition to the delta policy, the ``forcedeltabothparents`` |
|
3597 | In addition to the delta policy, the ``forcedeltabothparents`` | |
3595 | argument controls whether to force compute deltas against both parents |
|
3598 | argument controls whether to force compute deltas against both parents | |
3596 | for merges. By default, the current default is used. |
|
3599 | for merges. By default, the current default is used. | |
3597 |
|
3600 | |||
3598 | See `revlogutil.sidedata.get_sidedata_helpers` for the doc on |
|
3601 | See `revlogutil.sidedata.get_sidedata_helpers` for the doc on | |
3599 | `sidedata_helpers`. |
|
3602 | `sidedata_helpers`. | |
3600 | """ |
|
3603 | """ | |
3601 | if deltareuse not in self.DELTAREUSEALL: |
|
3604 | if deltareuse not in self.DELTAREUSEALL: | |
3602 | raise ValueError( |
|
3605 | raise ValueError( | |
3603 | _(b'value for deltareuse invalid: %s') % deltareuse |
|
3606 | _(b'value for deltareuse invalid: %s') % deltareuse | |
3604 | ) |
|
3607 | ) | |
3605 |
|
3608 | |||
3606 | if len(destrevlog): |
|
3609 | if len(destrevlog): | |
3607 | raise ValueError(_(b'destination revlog is not empty')) |
|
3610 | raise ValueError(_(b'destination revlog is not empty')) | |
3608 |
|
3611 | |||
3609 | if getattr(self, 'filteredrevs', None): |
|
3612 | if getattr(self, 'filteredrevs', None): | |
3610 | raise ValueError(_(b'source revlog has filtered revisions')) |
|
3613 | raise ValueError(_(b'source revlog has filtered revisions')) | |
3611 | if getattr(destrevlog, 'filteredrevs', None): |
|
3614 | if getattr(destrevlog, 'filteredrevs', None): | |
3612 | raise ValueError(_(b'destination revlog has filtered revisions')) |
|
3615 | raise ValueError(_(b'destination revlog has filtered revisions')) | |
3613 |
|
3616 | |||
3614 | # lazydelta and lazydeltabase controls whether to reuse a cached delta, |
|
3617 | # lazydelta and lazydeltabase controls whether to reuse a cached delta, | |
3615 | # if possible. |
|
3618 | # if possible. | |
3616 | old_delta_config = destrevlog.delta_config |
|
3619 | old_delta_config = destrevlog.delta_config | |
3617 | destrevlog.delta_config = destrevlog.delta_config.copy() |
|
3620 | destrevlog.delta_config = destrevlog.delta_config.copy() | |
3618 |
|
3621 | |||
3619 | try: |
|
3622 | try: | |
3620 | if deltareuse == self.DELTAREUSEALWAYS: |
|
3623 | if deltareuse == self.DELTAREUSEALWAYS: | |
3621 | destrevlog.delta_config.lazy_delta_base = True |
|
3624 | destrevlog.delta_config.lazy_delta_base = True | |
3622 | destrevlog.delta_config.lazy_delta = True |
|
3625 | destrevlog.delta_config.lazy_delta = True | |
3623 | elif deltareuse == self.DELTAREUSESAMEREVS: |
|
3626 | elif deltareuse == self.DELTAREUSESAMEREVS: | |
3624 | destrevlog.delta_config.lazy_delta_base = False |
|
3627 | destrevlog.delta_config.lazy_delta_base = False | |
3625 | destrevlog.delta_config.lazy_delta = True |
|
3628 | destrevlog.delta_config.lazy_delta = True | |
3626 | elif deltareuse == self.DELTAREUSENEVER: |
|
3629 | elif deltareuse == self.DELTAREUSENEVER: | |
3627 | destrevlog.delta_config.lazy_delta_base = False |
|
3630 | destrevlog.delta_config.lazy_delta_base = False | |
3628 | destrevlog.delta_config.lazy_delta = False |
|
3631 | destrevlog.delta_config.lazy_delta = False | |
3629 |
|
3632 | |||
3630 | delta_both_parents = ( |
|
3633 | delta_both_parents = ( | |
3631 | forcedeltabothparents or old_delta_config.delta_both_parents |
|
3634 | forcedeltabothparents or old_delta_config.delta_both_parents | |
3632 | ) |
|
3635 | ) | |
3633 | destrevlog.delta_config.delta_both_parents = delta_both_parents |
|
3636 | destrevlog.delta_config.delta_both_parents = delta_both_parents | |
3634 |
|
3637 | |||
3635 | with self.reading(), destrevlog._writing(tr): |
|
3638 | with self.reading(), destrevlog._writing(tr): | |
3636 | self._clone( |
|
3639 | self._clone( | |
3637 | tr, |
|
3640 | tr, | |
3638 | destrevlog, |
|
3641 | destrevlog, | |
3639 | addrevisioncb, |
|
3642 | addrevisioncb, | |
3640 | deltareuse, |
|
3643 | deltareuse, | |
3641 | forcedeltabothparents, |
|
3644 | forcedeltabothparents, | |
3642 | sidedata_helpers, |
|
3645 | sidedata_helpers, | |
3643 | ) |
|
3646 | ) | |
3644 |
|
3647 | |||
3645 | finally: |
|
3648 | finally: | |
3646 | destrevlog.delta_config = old_delta_config |
|
3649 | destrevlog.delta_config = old_delta_config | |
3647 |
|
3650 | |||
3648 | def _clone( |
|
3651 | def _clone( | |
3649 | self, |
|
3652 | self, | |
3650 | tr, |
|
3653 | tr, | |
3651 | destrevlog, |
|
3654 | destrevlog, | |
3652 | addrevisioncb, |
|
3655 | addrevisioncb, | |
3653 | deltareuse, |
|
3656 | deltareuse, | |
3654 | forcedeltabothparents, |
|
3657 | forcedeltabothparents, | |
3655 | sidedata_helpers, |
|
3658 | sidedata_helpers, | |
3656 | ): |
|
3659 | ): | |
3657 | """perform the core duty of `revlog.clone` after parameter processing""" |
|
3660 | """perform the core duty of `revlog.clone` after parameter processing""" | |
3658 | write_debug = None |
|
3661 | write_debug = None | |
3659 | if self.delta_config.debug_delta: |
|
3662 | if self.delta_config.debug_delta: | |
3660 | write_debug = tr._report |
|
3663 | write_debug = tr._report | |
3661 | deltacomputer = deltautil.deltacomputer( |
|
3664 | deltacomputer = deltautil.deltacomputer( | |
3662 | destrevlog, |
|
3665 | destrevlog, | |
3663 | write_debug=write_debug, |
|
3666 | write_debug=write_debug, | |
3664 | ) |
|
3667 | ) | |
3665 | index = self.index |
|
3668 | index = self.index | |
3666 | for rev in self: |
|
3669 | for rev in self: | |
3667 | entry = index[rev] |
|
3670 | entry = index[rev] | |
3668 |
|
3671 | |||
3669 | # Some classes override linkrev to take filtered revs into |
|
3672 | # Some classes override linkrev to take filtered revs into | |
3670 | # account. Use raw entry from index. |
|
3673 | # account. Use raw entry from index. | |
3671 | flags = entry[0] & 0xFFFF |
|
3674 | flags = entry[0] & 0xFFFF | |
3672 | linkrev = entry[4] |
|
3675 | linkrev = entry[4] | |
3673 | p1 = index[entry[5]][7] |
|
3676 | p1 = index[entry[5]][7] | |
3674 | p2 = index[entry[6]][7] |
|
3677 | p2 = index[entry[6]][7] | |
3675 | node = entry[7] |
|
3678 | node = entry[7] | |
3676 |
|
3679 | |||
3677 | # (Possibly) reuse the delta from the revlog if allowed and |
|
3680 | # (Possibly) reuse the delta from the revlog if allowed and | |
3678 | # the revlog chunk is a delta. |
|
3681 | # the revlog chunk is a delta. | |
3679 | cachedelta = None |
|
3682 | cachedelta = None | |
3680 | rawtext = None |
|
3683 | rawtext = None | |
3681 | if deltareuse == self.DELTAREUSEFULLADD: |
|
3684 | if deltareuse == self.DELTAREUSEFULLADD: | |
3682 | text = self._revisiondata(rev) |
|
3685 | text = self._revisiondata(rev) | |
3683 | sidedata = self.sidedata(rev) |
|
3686 | sidedata = self.sidedata(rev) | |
3684 |
|
3687 | |||
3685 | if sidedata_helpers is not None: |
|
3688 | if sidedata_helpers is not None: | |
3686 | (sidedata, new_flags) = sidedatautil.run_sidedata_helpers( |
|
3689 | (sidedata, new_flags) = sidedatautil.run_sidedata_helpers( | |
3687 | self, sidedata_helpers, sidedata, rev |
|
3690 | self, sidedata_helpers, sidedata, rev | |
3688 | ) |
|
3691 | ) | |
3689 | flags = flags | new_flags[0] & ~new_flags[1] |
|
3692 | flags = flags | new_flags[0] & ~new_flags[1] | |
3690 |
|
3693 | |||
3691 | destrevlog.addrevision( |
|
3694 | destrevlog.addrevision( | |
3692 | text, |
|
3695 | text, | |
3693 | tr, |
|
3696 | tr, | |
3694 | linkrev, |
|
3697 | linkrev, | |
3695 | p1, |
|
3698 | p1, | |
3696 | p2, |
|
3699 | p2, | |
3697 | cachedelta=cachedelta, |
|
3700 | cachedelta=cachedelta, | |
3698 | node=node, |
|
3701 | node=node, | |
3699 | flags=flags, |
|
3702 | flags=flags, | |
3700 | deltacomputer=deltacomputer, |
|
3703 | deltacomputer=deltacomputer, | |
3701 | sidedata=sidedata, |
|
3704 | sidedata=sidedata, | |
3702 | ) |
|
3705 | ) | |
3703 | else: |
|
3706 | else: | |
3704 | if destrevlog.delta_config.lazy_delta: |
|
3707 | if destrevlog.delta_config.lazy_delta: | |
3705 | dp = self.deltaparent(rev) |
|
3708 | dp = self.deltaparent(rev) | |
3706 | if dp != nullrev: |
|
3709 | if dp != nullrev: | |
3707 | cachedelta = (dp, bytes(self._inner._chunk(rev))) |
|
3710 | cachedelta = (dp, bytes(self._inner._chunk(rev))) | |
3708 |
|
3711 | |||
3709 | sidedata = None |
|
3712 | sidedata = None | |
3710 | if not cachedelta: |
|
3713 | if not cachedelta: | |
3711 | try: |
|
3714 | try: | |
3712 | rawtext = self._revisiondata(rev) |
|
3715 | rawtext = self._revisiondata(rev) | |
3713 | except error.CensoredNodeError as censored: |
|
3716 | except error.CensoredNodeError as censored: | |
3714 | assert flags & REVIDX_ISCENSORED |
|
3717 | assert flags & REVIDX_ISCENSORED | |
3715 | rawtext = censored.tombstone |
|
3718 | rawtext = censored.tombstone | |
3716 | sidedata = self.sidedata(rev) |
|
3719 | sidedata = self.sidedata(rev) | |
3717 | if sidedata is None: |
|
3720 | if sidedata is None: | |
3718 | sidedata = self.sidedata(rev) |
|
3721 | sidedata = self.sidedata(rev) | |
3719 |
|
3722 | |||
3720 | if sidedata_helpers is not None: |
|
3723 | if sidedata_helpers is not None: | |
3721 | (sidedata, new_flags) = sidedatautil.run_sidedata_helpers( |
|
3724 | (sidedata, new_flags) = sidedatautil.run_sidedata_helpers( | |
3722 | self, sidedata_helpers, sidedata, rev |
|
3725 | self, sidedata_helpers, sidedata, rev | |
3723 | ) |
|
3726 | ) | |
3724 | flags = flags | new_flags[0] & ~new_flags[1] |
|
3727 | flags = flags | new_flags[0] & ~new_flags[1] | |
3725 |
|
3728 | |||
3726 | destrevlog._addrevision( |
|
3729 | destrevlog._addrevision( | |
3727 | node, |
|
3730 | node, | |
3728 | rawtext, |
|
3731 | rawtext, | |
3729 | tr, |
|
3732 | tr, | |
3730 | linkrev, |
|
3733 | linkrev, | |
3731 | p1, |
|
3734 | p1, | |
3732 | p2, |
|
3735 | p2, | |
3733 | flags, |
|
3736 | flags, | |
3734 | cachedelta, |
|
3737 | cachedelta, | |
3735 | deltacomputer=deltacomputer, |
|
3738 | deltacomputer=deltacomputer, | |
3736 | sidedata=sidedata, |
|
3739 | sidedata=sidedata, | |
3737 | ) |
|
3740 | ) | |
3738 |
|
3741 | |||
3739 | if addrevisioncb: |
|
3742 | if addrevisioncb: | |
3740 | addrevisioncb(self, rev, node) |
|
3743 | addrevisioncb(self, rev, node) | |
3741 |
|
3744 | |||
3742 | def censorrevision(self, tr, censornode, tombstone=b''): |
|
3745 | def censorrevision(self, tr, censornode, tombstone=b''): | |
3743 | if self._format_version == REVLOGV0: |
|
3746 | if self._format_version == REVLOGV0: | |
3744 | raise error.RevlogError( |
|
3747 | raise error.RevlogError( | |
3745 | _(b'cannot censor with version %d revlogs') |
|
3748 | _(b'cannot censor with version %d revlogs') | |
3746 | % self._format_version |
|
3749 | % self._format_version | |
3747 | ) |
|
3750 | ) | |
3748 | elif self._format_version == REVLOGV1: |
|
3751 | elif self._format_version == REVLOGV1: | |
3749 | rewrite.v1_censor(self, tr, censornode, tombstone) |
|
3752 | rewrite.v1_censor(self, tr, censornode, tombstone) | |
3750 | else: |
|
3753 | else: | |
3751 | rewrite.v2_censor(self, tr, censornode, tombstone) |
|
3754 | rewrite.v2_censor(self, tr, censornode, tombstone) | |
3752 |
|
3755 | |||
3753 | def verifyintegrity(self, state): |
|
3756 | def verifyintegrity(self, state): | |
3754 | """Verifies the integrity of the revlog. |
|
3757 | """Verifies the integrity of the revlog. | |
3755 |
|
3758 | |||
3756 | Yields ``revlogproblem`` instances describing problems that are |
|
3759 | Yields ``revlogproblem`` instances describing problems that are | |
3757 | found. |
|
3760 | found. | |
3758 | """ |
|
3761 | """ | |
3759 | dd, di = self.checksize() |
|
3762 | dd, di = self.checksize() | |
3760 | if dd: |
|
3763 | if dd: | |
3761 | yield revlogproblem(error=_(b'data length off by %d bytes') % dd) |
|
3764 | yield revlogproblem(error=_(b'data length off by %d bytes') % dd) | |
3762 | if di: |
|
3765 | if di: | |
3763 | yield revlogproblem(error=_(b'index contains %d extra bytes') % di) |
|
3766 | yield revlogproblem(error=_(b'index contains %d extra bytes') % di) | |
3764 |
|
3767 | |||
3765 | version = self._format_version |
|
3768 | version = self._format_version | |
3766 |
|
3769 | |||
3767 | # The verifier tells us what version revlog we should be. |
|
3770 | # The verifier tells us what version revlog we should be. | |
3768 | if version != state[b'expectedversion']: |
|
3771 | if version != state[b'expectedversion']: | |
3769 | yield revlogproblem( |
|
3772 | yield revlogproblem( | |
3770 | warning=_(b"warning: '%s' uses revlog format %d; expected %d") |
|
3773 | warning=_(b"warning: '%s' uses revlog format %d; expected %d") | |
3771 | % (self.display_id, version, state[b'expectedversion']) |
|
3774 | % (self.display_id, version, state[b'expectedversion']) | |
3772 | ) |
|
3775 | ) | |
3773 |
|
3776 | |||
3774 | state[b'skipread'] = set() |
|
3777 | state[b'skipread'] = set() | |
3775 | state[b'safe_renamed'] = set() |
|
3778 | state[b'safe_renamed'] = set() | |
3776 |
|
3779 | |||
3777 | for rev in self: |
|
3780 | for rev in self: | |
3778 | node = self.node(rev) |
|
3781 | node = self.node(rev) | |
3779 |
|
3782 | |||
3780 | # Verify contents. 4 cases to care about: |
|
3783 | # Verify contents. 4 cases to care about: | |
3781 | # |
|
3784 | # | |
3782 | # common: the most common case |
|
3785 | # common: the most common case | |
3783 | # rename: with a rename |
|
3786 | # rename: with a rename | |
3784 | # meta: file content starts with b'\1\n', the metadata |
|
3787 | # meta: file content starts with b'\1\n', the metadata | |
3785 | # header defined in filelog.py, but without a rename |
|
3788 | # header defined in filelog.py, but without a rename | |
3786 | # ext: content stored externally |
|
3789 | # ext: content stored externally | |
3787 | # |
|
3790 | # | |
3788 | # More formally, their differences are shown below: |
|
3791 | # More formally, their differences are shown below: | |
3789 | # |
|
3792 | # | |
3790 | # | common | rename | meta | ext |
|
3793 | # | common | rename | meta | ext | |
3791 | # ------------------------------------------------------- |
|
3794 | # ------------------------------------------------------- | |
3792 | # flags() | 0 | 0 | 0 | not 0 |
|
3795 | # flags() | 0 | 0 | 0 | not 0 | |
3793 | # renamed() | False | True | False | ? |
|
3796 | # renamed() | False | True | False | ? | |
3794 | # rawtext[0:2]=='\1\n'| False | True | True | ? |
|
3797 | # rawtext[0:2]=='\1\n'| False | True | True | ? | |
3795 | # |
|
3798 | # | |
3796 | # "rawtext" means the raw text stored in revlog data, which |
|
3799 | # "rawtext" means the raw text stored in revlog data, which | |
3797 | # could be retrieved by "rawdata(rev)". "text" |
|
3800 | # could be retrieved by "rawdata(rev)". "text" | |
3798 | # mentioned below is "revision(rev)". |
|
3801 | # mentioned below is "revision(rev)". | |
3799 | # |
|
3802 | # | |
3800 | # There are 3 different lengths stored physically: |
|
3803 | # There are 3 different lengths stored physically: | |
3801 | # 1. L1: rawsize, stored in revlog index |
|
3804 | # 1. L1: rawsize, stored in revlog index | |
3802 | # 2. L2: len(rawtext), stored in revlog data |
|
3805 | # 2. L2: len(rawtext), stored in revlog data | |
3803 | # 3. L3: len(text), stored in revlog data if flags==0, or |
|
3806 | # 3. L3: len(text), stored in revlog data if flags==0, or | |
3804 | # possibly somewhere else if flags!=0 |
|
3807 | # possibly somewhere else if flags!=0 | |
3805 | # |
|
3808 | # | |
3806 | # L1 should be equal to L2. L3 could be different from them. |
|
3809 | # L1 should be equal to L2. L3 could be different from them. | |
3807 | # "text" may or may not affect commit hash depending on flag |
|
3810 | # "text" may or may not affect commit hash depending on flag | |
3808 | # processors (see flagutil.addflagprocessor). |
|
3811 | # processors (see flagutil.addflagprocessor). | |
3809 | # |
|
3812 | # | |
3810 | # | common | rename | meta | ext |
|
3813 | # | common | rename | meta | ext | |
3811 | # ------------------------------------------------- |
|
3814 | # ------------------------------------------------- | |
3812 | # rawsize() | L1 | L1 | L1 | L1 |
|
3815 | # rawsize() | L1 | L1 | L1 | L1 | |
3813 | # size() | L1 | L2-LM | L1(*) | L1 (?) |
|
3816 | # size() | L1 | L2-LM | L1(*) | L1 (?) | |
3814 | # len(rawtext) | L2 | L2 | L2 | L2 |
|
3817 | # len(rawtext) | L2 | L2 | L2 | L2 | |
3815 | # len(text) | L2 | L2 | L2 | L3 |
|
3818 | # len(text) | L2 | L2 | L2 | L3 | |
3816 | # len(read()) | L2 | L2-LM | L2-LM | L3 (?) |
|
3819 | # len(read()) | L2 | L2-LM | L2-LM | L3 (?) | |
3817 | # |
|
3820 | # | |
3818 | # LM: length of metadata, depending on rawtext |
|
3821 | # LM: length of metadata, depending on rawtext | |
3819 | # (*): not ideal, see comment in filelog.size |
|
3822 | # (*): not ideal, see comment in filelog.size | |
3820 | # (?): could be "- len(meta)" if the resolved content has |
|
3823 | # (?): could be "- len(meta)" if the resolved content has | |
3821 | # rename metadata |
|
3824 | # rename metadata | |
3822 | # |
|
3825 | # | |
3823 | # Checks needed to be done: |
|
3826 | # Checks needed to be done: | |
3824 | # 1. length check: L1 == L2, in all cases. |
|
3827 | # 1. length check: L1 == L2, in all cases. | |
3825 | # 2. hash check: depending on flag processor, we may need to |
|
3828 | # 2. hash check: depending on flag processor, we may need to | |
3826 | # use either "text" (external), or "rawtext" (in revlog). |
|
3829 | # use either "text" (external), or "rawtext" (in revlog). | |
3827 |
|
3830 | |||
3828 | try: |
|
3831 | try: | |
3829 | skipflags = state.get(b'skipflags', 0) |
|
3832 | skipflags = state.get(b'skipflags', 0) | |
3830 | if skipflags: |
|
3833 | if skipflags: | |
3831 | skipflags &= self.flags(rev) |
|
3834 | skipflags &= self.flags(rev) | |
3832 |
|
3835 | |||
3833 | _verify_revision(self, skipflags, state, node) |
|
3836 | _verify_revision(self, skipflags, state, node) | |
3834 |
|
3837 | |||
3835 | l1 = self.rawsize(rev) |
|
3838 | l1 = self.rawsize(rev) | |
3836 | l2 = len(self.rawdata(node)) |
|
3839 | l2 = len(self.rawdata(node)) | |
3837 |
|
3840 | |||
3838 | if l1 != l2: |
|
3841 | if l1 != l2: | |
3839 | yield revlogproblem( |
|
3842 | yield revlogproblem( | |
3840 | error=_(b'unpacked size is %d, %d expected') % (l2, l1), |
|
3843 | error=_(b'unpacked size is %d, %d expected') % (l2, l1), | |
3841 | node=node, |
|
3844 | node=node, | |
3842 | ) |
|
3845 | ) | |
3843 |
|
3846 | |||
3844 | except error.CensoredNodeError: |
|
3847 | except error.CensoredNodeError: | |
3845 | if state[b'erroroncensored']: |
|
3848 | if state[b'erroroncensored']: | |
3846 | yield revlogproblem( |
|
3849 | yield revlogproblem( | |
3847 | error=_(b'censored file data'), node=node |
|
3850 | error=_(b'censored file data'), node=node | |
3848 | ) |
|
3851 | ) | |
3849 | state[b'skipread'].add(node) |
|
3852 | state[b'skipread'].add(node) | |
3850 | except Exception as e: |
|
3853 | except Exception as e: | |
3851 | yield revlogproblem( |
|
3854 | yield revlogproblem( | |
3852 | error=_(b'unpacking %s: %s') |
|
3855 | error=_(b'unpacking %s: %s') | |
3853 | % (short(node), stringutil.forcebytestr(e)), |
|
3856 | % (short(node), stringutil.forcebytestr(e)), | |
3854 | node=node, |
|
3857 | node=node, | |
3855 | ) |
|
3858 | ) | |
3856 | state[b'skipread'].add(node) |
|
3859 | state[b'skipread'].add(node) | |
3857 |
|
3860 | |||
3858 | def storageinfo( |
|
3861 | def storageinfo( | |
3859 | self, |
|
3862 | self, | |
3860 | exclusivefiles=False, |
|
3863 | exclusivefiles=False, | |
3861 | sharedfiles=False, |
|
3864 | sharedfiles=False, | |
3862 | revisionscount=False, |
|
3865 | revisionscount=False, | |
3863 | trackedsize=False, |
|
3866 | trackedsize=False, | |
3864 | storedsize=False, |
|
3867 | storedsize=False, | |
3865 | ): |
|
3868 | ): | |
3866 | d = {} |
|
3869 | d = {} | |
3867 |
|
3870 | |||
3868 | if exclusivefiles: |
|
3871 | if exclusivefiles: | |
3869 | d[b'exclusivefiles'] = [(self.opener, self._indexfile)] |
|
3872 | d[b'exclusivefiles'] = [(self.opener, self._indexfile)] | |
3870 | if not self._inline: |
|
3873 | if not self._inline: | |
3871 | d[b'exclusivefiles'].append((self.opener, self._datafile)) |
|
3874 | d[b'exclusivefiles'].append((self.opener, self._datafile)) | |
3872 |
|
3875 | |||
3873 | if sharedfiles: |
|
3876 | if sharedfiles: | |
3874 | d[b'sharedfiles'] = [] |
|
3877 | d[b'sharedfiles'] = [] | |
3875 |
|
3878 | |||
3876 | if revisionscount: |
|
3879 | if revisionscount: | |
3877 | d[b'revisionscount'] = len(self) |
|
3880 | d[b'revisionscount'] = len(self) | |
3878 |
|
3881 | |||
3879 | if trackedsize: |
|
3882 | if trackedsize: | |
3880 | d[b'trackedsize'] = sum(map(self.rawsize, iter(self))) |
|
3883 | d[b'trackedsize'] = sum(map(self.rawsize, iter(self))) | |
3881 |
|
3884 | |||
3882 | if storedsize: |
|
3885 | if storedsize: | |
3883 | d[b'storedsize'] = sum( |
|
3886 | d[b'storedsize'] = sum( | |
3884 | self.opener.stat(path).st_size for path in self.files() |
|
3887 | self.opener.stat(path).st_size for path in self.files() | |
3885 | ) |
|
3888 | ) | |
3886 |
|
3889 | |||
3887 | return d |
|
3890 | return d | |
3888 |
|
3891 | |||
3889 | def rewrite_sidedata(self, transaction, helpers, startrev, endrev): |
|
3892 | def rewrite_sidedata(self, transaction, helpers, startrev, endrev): | |
3890 | if not self.feature_config.has_side_data: |
|
3893 | if not self.feature_config.has_side_data: | |
3891 | return |
|
3894 | return | |
3892 | # revlog formats with sidedata support does not support inline |
|
3895 | # revlog formats with sidedata support does not support inline | |
3893 | assert not self._inline |
|
3896 | assert not self._inline | |
3894 | if not helpers[1] and not helpers[2]: |
|
3897 | if not helpers[1] and not helpers[2]: | |
3895 | # Nothing to generate or remove |
|
3898 | # Nothing to generate or remove | |
3896 | return |
|
3899 | return | |
3897 |
|
3900 | |||
3898 | new_entries = [] |
|
3901 | new_entries = [] | |
3899 | # append the new sidedata |
|
3902 | # append the new sidedata | |
3900 | with self._writing(transaction): |
|
3903 | with self._writing(transaction): | |
3901 | ifh, dfh, sdfh = self._inner._writinghandles |
|
3904 | ifh, dfh, sdfh = self._inner._writinghandles | |
3902 | dfh.seek(self._docket.sidedata_end, os.SEEK_SET) |
|
3905 | dfh.seek(self._docket.sidedata_end, os.SEEK_SET) | |
3903 |
|
3906 | |||
3904 | current_offset = sdfh.tell() |
|
3907 | current_offset = sdfh.tell() | |
3905 | for rev in range(startrev, endrev + 1): |
|
3908 | for rev in range(startrev, endrev + 1): | |
3906 | entry = self.index[rev] |
|
3909 | entry = self.index[rev] | |
3907 | new_sidedata, flags = sidedatautil.run_sidedata_helpers( |
|
3910 | new_sidedata, flags = sidedatautil.run_sidedata_helpers( | |
3908 | store=self, |
|
3911 | store=self, | |
3909 | sidedata_helpers=helpers, |
|
3912 | sidedata_helpers=helpers, | |
3910 | sidedata={}, |
|
3913 | sidedata={}, | |
3911 | rev=rev, |
|
3914 | rev=rev, | |
3912 | ) |
|
3915 | ) | |
3913 |
|
3916 | |||
3914 | serialized_sidedata = sidedatautil.serialize_sidedata( |
|
3917 | serialized_sidedata = sidedatautil.serialize_sidedata( | |
3915 | new_sidedata |
|
3918 | new_sidedata | |
3916 | ) |
|
3919 | ) | |
3917 |
|
3920 | |||
3918 | sidedata_compression_mode = COMP_MODE_INLINE |
|
3921 | sidedata_compression_mode = COMP_MODE_INLINE | |
3919 | if serialized_sidedata and self.feature_config.has_side_data: |
|
3922 | if serialized_sidedata and self.feature_config.has_side_data: | |
3920 | sidedata_compression_mode = COMP_MODE_PLAIN |
|
3923 | sidedata_compression_mode = COMP_MODE_PLAIN | |
3921 | h, comp_sidedata = self._inner.compress(serialized_sidedata) |
|
3924 | h, comp_sidedata = self._inner.compress(serialized_sidedata) | |
3922 | if ( |
|
3925 | if ( | |
3923 | h != b'u' |
|
3926 | h != b'u' | |
3924 | and comp_sidedata[0] != b'\0' |
|
3927 | and comp_sidedata[0] != b'\0' | |
3925 | and len(comp_sidedata) < len(serialized_sidedata) |
|
3928 | and len(comp_sidedata) < len(serialized_sidedata) | |
3926 | ): |
|
3929 | ): | |
3927 | assert not h |
|
3930 | assert not h | |
3928 | if ( |
|
3931 | if ( | |
3929 | comp_sidedata[0] |
|
3932 | comp_sidedata[0] | |
3930 | == self._docket.default_compression_header |
|
3933 | == self._docket.default_compression_header | |
3931 | ): |
|
3934 | ): | |
3932 | sidedata_compression_mode = COMP_MODE_DEFAULT |
|
3935 | sidedata_compression_mode = COMP_MODE_DEFAULT | |
3933 | serialized_sidedata = comp_sidedata |
|
3936 | serialized_sidedata = comp_sidedata | |
3934 | else: |
|
3937 | else: | |
3935 | sidedata_compression_mode = COMP_MODE_INLINE |
|
3938 | sidedata_compression_mode = COMP_MODE_INLINE | |
3936 | serialized_sidedata = comp_sidedata |
|
3939 | serialized_sidedata = comp_sidedata | |
3937 | if entry[8] != 0 or entry[9] != 0: |
|
3940 | if entry[8] != 0 or entry[9] != 0: | |
3938 | # rewriting entries that already have sidedata is not |
|
3941 | # rewriting entries that already have sidedata is not | |
3939 | # supported yet, because it introduces garbage data in the |
|
3942 | # supported yet, because it introduces garbage data in the | |
3940 | # revlog. |
|
3943 | # revlog. | |
3941 | msg = b"rewriting existing sidedata is not supported yet" |
|
3944 | msg = b"rewriting existing sidedata is not supported yet" | |
3942 | raise error.Abort(msg) |
|
3945 | raise error.Abort(msg) | |
3943 |
|
3946 | |||
3944 | # Apply (potential) flags to add and to remove after running |
|
3947 | # Apply (potential) flags to add and to remove after running | |
3945 | # the sidedata helpers |
|
3948 | # the sidedata helpers | |
3946 | new_offset_flags = entry[0] | flags[0] & ~flags[1] |
|
3949 | new_offset_flags = entry[0] | flags[0] & ~flags[1] | |
3947 | entry_update = ( |
|
3950 | entry_update = ( | |
3948 | current_offset, |
|
3951 | current_offset, | |
3949 | len(serialized_sidedata), |
|
3952 | len(serialized_sidedata), | |
3950 | new_offset_flags, |
|
3953 | new_offset_flags, | |
3951 | sidedata_compression_mode, |
|
3954 | sidedata_compression_mode, | |
3952 | ) |
|
3955 | ) | |
3953 |
|
3956 | |||
3954 | # the sidedata computation might have move the file cursors around |
|
3957 | # the sidedata computation might have move the file cursors around | |
3955 | sdfh.seek(current_offset, os.SEEK_SET) |
|
3958 | sdfh.seek(current_offset, os.SEEK_SET) | |
3956 | sdfh.write(serialized_sidedata) |
|
3959 | sdfh.write(serialized_sidedata) | |
3957 | new_entries.append(entry_update) |
|
3960 | new_entries.append(entry_update) | |
3958 | current_offset += len(serialized_sidedata) |
|
3961 | current_offset += len(serialized_sidedata) | |
3959 | self._docket.sidedata_end = sdfh.tell() |
|
3962 | self._docket.sidedata_end = sdfh.tell() | |
3960 |
|
3963 | |||
3961 | # rewrite the new index entries |
|
3964 | # rewrite the new index entries | |
3962 | ifh.seek(startrev * self.index.entry_size) |
|
3965 | ifh.seek(startrev * self.index.entry_size) | |
3963 | for i, e in enumerate(new_entries): |
|
3966 | for i, e in enumerate(new_entries): | |
3964 | rev = startrev + i |
|
3967 | rev = startrev + i | |
3965 | self.index.replace_sidedata_info(rev, *e) |
|
3968 | self.index.replace_sidedata_info(rev, *e) | |
3966 | packed = self.index.entry_binary(rev) |
|
3969 | packed = self.index.entry_binary(rev) | |
3967 | if rev == 0 and self._docket is None: |
|
3970 | if rev == 0 and self._docket is None: | |
3968 | header = self._format_flags | self._format_version |
|
3971 | header = self._format_flags | self._format_version | |
3969 | header = self.index.pack_header(header) |
|
3972 | header = self.index.pack_header(header) | |
3970 | packed = header + packed |
|
3973 | packed = header + packed | |
3971 | ifh.write(packed) |
|
3974 | ifh.write(packed) |
General Comments 0
You need to be logged in to leave comments.
Login now