|
@@
-1,3466
+1,3468
b''
|
|
1
|
# revlog.py - storage back-end for mercurial
|
|
1
|
# revlog.py - storage back-end for mercurial
|
|
2
|
#
|
|
2
|
#
|
|
3
|
# Copyright 2005-2007 Olivia Mackall <olivia@selenic.com>
|
|
3
|
# Copyright 2005-2007 Olivia Mackall <olivia@selenic.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
|
"""Storage back-end for Mercurial.
|
|
8
|
"""Storage back-end for Mercurial.
|
|
9
|
|
|
9
|
|
|
10
|
This provides efficient delta storage with O(1) retrieve and append
|
|
10
|
This provides efficient delta storage with O(1) retrieve and append
|
|
11
|
and O(changes) merge between branches.
|
|
11
|
and O(changes) merge between branches.
|
|
12
|
"""
|
|
12
|
"""
|
|
13
|
|
|
13
|
|
|
14
|
from __future__ import absolute_import
|
|
14
|
from __future__ import absolute_import
|
|
15
|
|
|
15
|
|
|
16
|
import binascii
|
|
16
|
import binascii
|
|
17
|
import collections
|
|
17
|
import collections
|
|
18
|
import contextlib
|
|
18
|
import contextlib
|
|
19
|
import errno
|
|
19
|
import errno
|
|
20
|
import io
|
|
20
|
import io
|
|
21
|
import os
|
|
21
|
import os
|
|
22
|
import struct
|
|
22
|
import struct
|
|
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 .pycompat import getattr
|
|
35
|
from .pycompat import getattr
|
|
36
|
from .revlogutils.constants import (
|
|
36
|
from .revlogutils.constants import (
|
|
37
|
ALL_KINDS,
|
|
37
|
ALL_KINDS,
|
|
38
|
CHANGELOGV2,
|
|
38
|
CHANGELOGV2,
|
|
39
|
COMP_MODE_DEFAULT,
|
|
39
|
COMP_MODE_DEFAULT,
|
|
40
|
COMP_MODE_INLINE,
|
|
40
|
COMP_MODE_INLINE,
|
|
41
|
COMP_MODE_PLAIN,
|
|
41
|
COMP_MODE_PLAIN,
|
|
42
|
FEATURES_BY_VERSION,
|
|
42
|
FEATURES_BY_VERSION,
|
|
43
|
FLAG_GENERALDELTA,
|
|
43
|
FLAG_GENERALDELTA,
|
|
44
|
FLAG_INLINE_DATA,
|
|
44
|
FLAG_INLINE_DATA,
|
|
45
|
INDEX_HEADER,
|
|
45
|
INDEX_HEADER,
|
|
46
|
KIND_CHANGELOG,
|
|
46
|
KIND_CHANGELOG,
|
|
47
|
REVLOGV0,
|
|
47
|
REVLOGV0,
|
|
48
|
REVLOGV1,
|
|
48
|
REVLOGV1,
|
|
49
|
REVLOGV1_FLAGS,
|
|
49
|
REVLOGV1_FLAGS,
|
|
50
|
REVLOGV2,
|
|
50
|
REVLOGV2,
|
|
51
|
REVLOGV2_FLAGS,
|
|
51
|
REVLOGV2_FLAGS,
|
|
52
|
REVLOG_DEFAULT_FLAGS,
|
|
52
|
REVLOG_DEFAULT_FLAGS,
|
|
53
|
REVLOG_DEFAULT_FORMAT,
|
|
53
|
REVLOG_DEFAULT_FORMAT,
|
|
54
|
REVLOG_DEFAULT_VERSION,
|
|
54
|
REVLOG_DEFAULT_VERSION,
|
|
55
|
SUPPORTED_FLAGS,
|
|
55
|
SUPPORTED_FLAGS,
|
|
56
|
)
|
|
56
|
)
|
|
57
|
from .revlogutils.flagutil import (
|
|
57
|
from .revlogutils.flagutil import (
|
|
58
|
REVIDX_DEFAULT_FLAGS,
|
|
58
|
REVIDX_DEFAULT_FLAGS,
|
|
59
|
REVIDX_ELLIPSIS,
|
|
59
|
REVIDX_ELLIPSIS,
|
|
60
|
REVIDX_EXTSTORED,
|
|
60
|
REVIDX_EXTSTORED,
|
|
61
|
REVIDX_FLAGS_ORDER,
|
|
61
|
REVIDX_FLAGS_ORDER,
|
|
62
|
REVIDX_HASCOPIESINFO,
|
|
62
|
REVIDX_HASCOPIESINFO,
|
|
63
|
REVIDX_ISCENSORED,
|
|
63
|
REVIDX_ISCENSORED,
|
|
64
|
REVIDX_RAWTEXT_CHANGING_FLAGS,
|
|
64
|
REVIDX_RAWTEXT_CHANGING_FLAGS,
|
|
65
|
)
|
|
65
|
)
|
|
66
|
from .thirdparty import attr
|
|
66
|
from .thirdparty import attr
|
|
67
|
from . import (
|
|
67
|
from . import (
|
|
68
|
ancestor,
|
|
68
|
ancestor,
|
|
69
|
dagop,
|
|
69
|
dagop,
|
|
70
|
error,
|
|
70
|
error,
|
|
71
|
mdiff,
|
|
71
|
mdiff,
|
|
72
|
policy,
|
|
72
|
policy,
|
|
73
|
pycompat,
|
|
73
|
pycompat,
|
|
74
|
templatefilters,
|
|
74
|
templatefilters,
|
|
75
|
util,
|
|
75
|
util,
|
|
76
|
)
|
|
76
|
)
|
|
77
|
from .interfaces import (
|
|
77
|
from .interfaces import (
|
|
78
|
repository,
|
|
78
|
repository,
|
|
79
|
util as interfaceutil,
|
|
79
|
util as interfaceutil,
|
|
80
|
)
|
|
80
|
)
|
|
81
|
from .revlogutils import (
|
|
81
|
from .revlogutils import (
|
|
82
|
deltas as deltautil,
|
|
82
|
deltas as deltautil,
|
|
83
|
docket as docketutil,
|
|
83
|
docket as docketutil,
|
|
84
|
flagutil,
|
|
84
|
flagutil,
|
|
85
|
nodemap as nodemaputil,
|
|
85
|
nodemap as nodemaputil,
|
|
86
|
revlogv0,
|
|
86
|
revlogv0,
|
|
87
|
sidedata as sidedatautil,
|
|
87
|
sidedata as sidedatautil,
|
|
88
|
)
|
|
88
|
)
|
|
89
|
from .utils import (
|
|
89
|
from .utils import (
|
|
90
|
storageutil,
|
|
90
|
storageutil,
|
|
91
|
stringutil,
|
|
91
|
stringutil,
|
|
92
|
)
|
|
92
|
)
|
|
93
|
|
|
93
|
|
|
94
|
# blanked usage of all the name to prevent pyflakes constraints
|
|
94
|
# blanked usage of all the name to prevent pyflakes constraints
|
|
95
|
# We need these name available in the module for extensions.
|
|
95
|
# We need these name available in the module for extensions.
|
|
96
|
|
|
96
|
|
|
97
|
REVLOGV0
|
|
97
|
REVLOGV0
|
|
98
|
REVLOGV1
|
|
98
|
REVLOGV1
|
|
99
|
REVLOGV2
|
|
99
|
REVLOGV2
|
|
100
|
FLAG_INLINE_DATA
|
|
100
|
FLAG_INLINE_DATA
|
|
101
|
FLAG_GENERALDELTA
|
|
101
|
FLAG_GENERALDELTA
|
|
102
|
REVLOG_DEFAULT_FLAGS
|
|
102
|
REVLOG_DEFAULT_FLAGS
|
|
103
|
REVLOG_DEFAULT_FORMAT
|
|
103
|
REVLOG_DEFAULT_FORMAT
|
|
104
|
REVLOG_DEFAULT_VERSION
|
|
104
|
REVLOG_DEFAULT_VERSION
|
|
105
|
REVLOGV1_FLAGS
|
|
105
|
REVLOGV1_FLAGS
|
|
106
|
REVLOGV2_FLAGS
|
|
106
|
REVLOGV2_FLAGS
|
|
107
|
REVIDX_ISCENSORED
|
|
107
|
REVIDX_ISCENSORED
|
|
108
|
REVIDX_ELLIPSIS
|
|
108
|
REVIDX_ELLIPSIS
|
|
109
|
REVIDX_HASCOPIESINFO
|
|
109
|
REVIDX_HASCOPIESINFO
|
|
110
|
REVIDX_EXTSTORED
|
|
110
|
REVIDX_EXTSTORED
|
|
111
|
REVIDX_DEFAULT_FLAGS
|
|
111
|
REVIDX_DEFAULT_FLAGS
|
|
112
|
REVIDX_FLAGS_ORDER
|
|
112
|
REVIDX_FLAGS_ORDER
|
|
113
|
REVIDX_RAWTEXT_CHANGING_FLAGS
|
|
113
|
REVIDX_RAWTEXT_CHANGING_FLAGS
|
|
114
|
|
|
114
|
|
|
115
|
parsers = policy.importmod('parsers')
|
|
115
|
parsers = policy.importmod('parsers')
|
|
116
|
rustancestor = policy.importrust('ancestor')
|
|
116
|
rustancestor = policy.importrust('ancestor')
|
|
117
|
rustdagop = policy.importrust('dagop')
|
|
117
|
rustdagop = policy.importrust('dagop')
|
|
118
|
rustrevlog = policy.importrust('revlog')
|
|
118
|
rustrevlog = policy.importrust('revlog')
|
|
119
|
|
|
119
|
|
|
120
|
# Aliased for performance.
|
|
120
|
# Aliased for performance.
|
|
121
|
_zlibdecompress = zlib.decompress
|
|
121
|
_zlibdecompress = zlib.decompress
|
|
122
|
|
|
122
|
|
|
123
|
# max size of revlog with inline data
|
|
123
|
# max size of revlog with inline data
|
|
124
|
_maxinline = 131072
|
|
124
|
_maxinline = 131072
|
|
125
|
_chunksize = 1048576
|
|
125
|
_chunksize = 1048576
|
|
126
|
|
|
126
|
|
|
127
|
# Flag processors for REVIDX_ELLIPSIS.
|
|
127
|
# Flag processors for REVIDX_ELLIPSIS.
|
|
128
|
def ellipsisreadprocessor(rl, text):
|
|
128
|
def ellipsisreadprocessor(rl, text):
|
|
129
|
return text, False
|
|
129
|
return text, False
|
|
130
|
|
|
130
|
|
|
131
|
|
|
131
|
|
|
132
|
def ellipsiswriteprocessor(rl, text):
|
|
132
|
def ellipsiswriteprocessor(rl, text):
|
|
133
|
return text, False
|
|
133
|
return text, False
|
|
134
|
|
|
134
|
|
|
135
|
|
|
135
|
|
|
136
|
def ellipsisrawprocessor(rl, text):
|
|
136
|
def ellipsisrawprocessor(rl, text):
|
|
137
|
return False
|
|
137
|
return False
|
|
138
|
|
|
138
|
|
|
139
|
|
|
139
|
|
|
140
|
ellipsisprocessor = (
|
|
140
|
ellipsisprocessor = (
|
|
141
|
ellipsisreadprocessor,
|
|
141
|
ellipsisreadprocessor,
|
|
142
|
ellipsiswriteprocessor,
|
|
142
|
ellipsiswriteprocessor,
|
|
143
|
ellipsisrawprocessor,
|
|
143
|
ellipsisrawprocessor,
|
|
144
|
)
|
|
144
|
)
|
|
145
|
|
|
145
|
|
|
146
|
|
|
146
|
|
|
147
|
def offset_type(offset, type):
|
|
147
|
def offset_type(offset, type):
|
|
148
|
if (type & ~flagutil.REVIDX_KNOWN_FLAGS) != 0:
|
|
148
|
if (type & ~flagutil.REVIDX_KNOWN_FLAGS) != 0:
|
|
149
|
raise ValueError(b'unknown revlog index flags')
|
|
149
|
raise ValueError(b'unknown revlog index flags')
|
|
150
|
return int(int(offset) << 16 | type)
|
|
150
|
return int(int(offset) << 16 | type)
|
|
151
|
|
|
151
|
|
|
152
|
|
|
152
|
|
|
153
|
def _verify_revision(rl, skipflags, state, node):
|
|
153
|
def _verify_revision(rl, skipflags, state, node):
|
|
154
|
"""Verify the integrity of the given revlog ``node`` while providing a hook
|
|
154
|
"""Verify the integrity of the given revlog ``node`` while providing a hook
|
|
155
|
point for extensions to influence the operation."""
|
|
155
|
point for extensions to influence the operation."""
|
|
156
|
if skipflags:
|
|
156
|
if skipflags:
|
|
157
|
state[b'skipread'].add(node)
|
|
157
|
state[b'skipread'].add(node)
|
|
158
|
else:
|
|
158
|
else:
|
|
159
|
# Side-effect: read content and verify hash.
|
|
159
|
# Side-effect: read content and verify hash.
|
|
160
|
rl.revision(node)
|
|
160
|
rl.revision(node)
|
|
161
|
|
|
161
|
|
|
162
|
|
|
162
|
|
|
163
|
# True if a fast implementation for persistent-nodemap is available
|
|
163
|
# True if a fast implementation for persistent-nodemap is available
|
|
164
|
#
|
|
164
|
#
|
|
165
|
# We also consider we have a "fast" implementation in "pure" python because
|
|
165
|
# We also consider we have a "fast" implementation in "pure" python because
|
|
166
|
# people using pure don't really have performance consideration (and a
|
|
166
|
# people using pure don't really have performance consideration (and a
|
|
167
|
# wheelbarrow of other slowness source)
|
|
167
|
# wheelbarrow of other slowness source)
|
|
168
|
HAS_FAST_PERSISTENT_NODEMAP = rustrevlog is not None or util.safehasattr(
|
|
168
|
HAS_FAST_PERSISTENT_NODEMAP = rustrevlog is not None or util.safehasattr(
|
|
169
|
parsers, 'BaseIndexObject'
|
|
169
|
parsers, 'BaseIndexObject'
|
|
170
|
)
|
|
170
|
)
|
|
171
|
|
|
171
|
|
|
172
|
|
|
172
|
|
|
173
|
@attr.s(slots=True, frozen=True)
|
|
173
|
@attr.s(slots=True, frozen=True)
|
|
174
|
class _revisioninfo(object):
|
|
174
|
class _revisioninfo(object):
|
|
175
|
"""Information about a revision that allows building its fulltext
|
|
175
|
"""Information about a revision that allows building its fulltext
|
|
176
|
node: expected hash of the revision
|
|
176
|
node: expected hash of the revision
|
|
177
|
p1, p2: parent revs of the revision
|
|
177
|
p1, p2: parent revs of the revision
|
|
178
|
btext: built text cache consisting of a one-element list
|
|
178
|
btext: built text cache consisting of a one-element list
|
|
179
|
cachedelta: (baserev, uncompressed_delta) or None
|
|
179
|
cachedelta: (baserev, uncompressed_delta) or None
|
|
180
|
flags: flags associated to the revision storage
|
|
180
|
flags: flags associated to the revision storage
|
|
181
|
|
|
181
|
|
|
182
|
One of btext[0] or cachedelta must be set.
|
|
182
|
One of btext[0] or cachedelta must be set.
|
|
183
|
"""
|
|
183
|
"""
|
|
184
|
|
|
184
|
|
|
185
|
node = attr.ib()
|
|
185
|
node = attr.ib()
|
|
186
|
p1 = attr.ib()
|
|
186
|
p1 = attr.ib()
|
|
187
|
p2 = attr.ib()
|
|
187
|
p2 = attr.ib()
|
|
188
|
btext = attr.ib()
|
|
188
|
btext = attr.ib()
|
|
189
|
textlen = attr.ib()
|
|
189
|
textlen = attr.ib()
|
|
190
|
cachedelta = attr.ib()
|
|
190
|
cachedelta = attr.ib()
|
|
191
|
flags = attr.ib()
|
|
191
|
flags = attr.ib()
|
|
192
|
|
|
192
|
|
|
193
|
|
|
193
|
|
|
194
|
@interfaceutil.implementer(repository.irevisiondelta)
|
|
194
|
@interfaceutil.implementer(repository.irevisiondelta)
|
|
195
|
@attr.s(slots=True)
|
|
195
|
@attr.s(slots=True)
|
|
196
|
class revlogrevisiondelta(object):
|
|
196
|
class revlogrevisiondelta(object):
|
|
197
|
node = attr.ib()
|
|
197
|
node = attr.ib()
|
|
198
|
p1node = attr.ib()
|
|
198
|
p1node = attr.ib()
|
|
199
|
p2node = attr.ib()
|
|
199
|
p2node = attr.ib()
|
|
200
|
basenode = attr.ib()
|
|
200
|
basenode = attr.ib()
|
|
201
|
flags = attr.ib()
|
|
201
|
flags = attr.ib()
|
|
202
|
baserevisionsize = attr.ib()
|
|
202
|
baserevisionsize = attr.ib()
|
|
203
|
revision = attr.ib()
|
|
203
|
revision = attr.ib()
|
|
204
|
delta = attr.ib()
|
|
204
|
delta = attr.ib()
|
|
205
|
sidedata = attr.ib()
|
|
205
|
sidedata = attr.ib()
|
|
206
|
protocol_flags = attr.ib()
|
|
206
|
protocol_flags = attr.ib()
|
|
207
|
linknode = attr.ib(default=None)
|
|
207
|
linknode = attr.ib(default=None)
|
|
208
|
|
|
208
|
|
|
209
|
|
|
209
|
|
|
210
|
@interfaceutil.implementer(repository.iverifyproblem)
|
|
210
|
@interfaceutil.implementer(repository.iverifyproblem)
|
|
211
|
@attr.s(frozen=True)
|
|
211
|
@attr.s(frozen=True)
|
|
212
|
class revlogproblem(object):
|
|
212
|
class revlogproblem(object):
|
|
213
|
warning = attr.ib(default=None)
|
|
213
|
warning = attr.ib(default=None)
|
|
214
|
error = attr.ib(default=None)
|
|
214
|
error = attr.ib(default=None)
|
|
215
|
node = attr.ib(default=None)
|
|
215
|
node = attr.ib(default=None)
|
|
216
|
|
|
216
|
|
|
217
|
|
|
217
|
|
|
218
|
def parse_index_v1(data, inline):
|
|
218
|
def parse_index_v1(data, inline):
|
|
219
|
# call the C implementation to parse the index data
|
|
219
|
# call the C implementation to parse the index data
|
|
220
|
index, cache = parsers.parse_index2(data, inline)
|
|
220
|
index, cache = parsers.parse_index2(data, inline)
|
|
221
|
return index, cache
|
|
221
|
return index, cache
|
|
222
|
|
|
222
|
|
|
223
|
|
|
223
|
|
|
224
|
def parse_index_v2(data, inline):
|
|
224
|
def parse_index_v2(data, inline):
|
|
225
|
# call the C implementation to parse the index data
|
|
225
|
# call the C implementation to parse the index data
|
|
226
|
index, cache = parsers.parse_index2(data, inline, revlogv2=True)
|
|
226
|
index, cache = parsers.parse_index2(data, inline, revlogv2=True)
|
|
227
|
return index, cache
|
|
227
|
return index, cache
|
|
228
|
|
|
228
|
|
|
229
|
|
|
229
|
|
|
230
|
def parse_index_cl_v2(data, inline):
|
|
230
|
def parse_index_cl_v2(data, inline):
|
|
231
|
# call the C implementation to parse the index data
|
|
231
|
# call the C implementation to parse the index data
|
|
232
|
assert not inline
|
|
232
|
assert not inline
|
|
233
|
from .pure.parsers import parse_index_cl_v2
|
|
233
|
from .pure.parsers import parse_index_cl_v2
|
|
234
|
|
|
234
|
|
|
235
|
index, cache = parse_index_cl_v2(data)
|
|
235
|
index, cache = parse_index_cl_v2(data)
|
|
236
|
return index, cache
|
|
236
|
return index, cache
|
|
237
|
|
|
237
|
|
|
238
|
|
|
238
|
|
|
239
|
if util.safehasattr(parsers, 'parse_index_devel_nodemap'):
|
|
239
|
if util.safehasattr(parsers, 'parse_index_devel_nodemap'):
|
|
240
|
|
|
240
|
|
|
241
|
def parse_index_v1_nodemap(data, inline):
|
|
241
|
def parse_index_v1_nodemap(data, inline):
|
|
242
|
index, cache = parsers.parse_index_devel_nodemap(data, inline)
|
|
242
|
index, cache = parsers.parse_index_devel_nodemap(data, inline)
|
|
243
|
return index, cache
|
|
243
|
return index, cache
|
|
244
|
|
|
244
|
|
|
245
|
|
|
245
|
|
|
246
|
else:
|
|
246
|
else:
|
|
247
|
parse_index_v1_nodemap = None
|
|
247
|
parse_index_v1_nodemap = None
|
|
248
|
|
|
248
|
|
|
249
|
|
|
249
|
|
|
250
|
def parse_index_v1_mixed(data, inline):
|
|
250
|
def parse_index_v1_mixed(data, inline):
|
|
251
|
index, cache = parse_index_v1(data, inline)
|
|
251
|
index, cache = parse_index_v1(data, inline)
|
|
252
|
return rustrevlog.MixedIndex(index), cache
|
|
252
|
return rustrevlog.MixedIndex(index), cache
|
|
253
|
|
|
253
|
|
|
254
|
|
|
254
|
|
|
255
|
# corresponds to uncompressed length of indexformatng (2 gigs, 4-byte
|
|
255
|
# corresponds to uncompressed length of indexformatng (2 gigs, 4-byte
|
|
256
|
# signed integer)
|
|
256
|
# signed integer)
|
|
257
|
_maxentrysize = 0x7FFFFFFF
|
|
257
|
_maxentrysize = 0x7FFFFFFF
|
|
258
|
|
|
258
|
|
|
259
|
|
|
259
|
|
|
260
|
class revlog(object):
|
|
260
|
class revlog(object):
|
|
261
|
"""
|
|
261
|
"""
|
|
262
|
the underlying revision storage object
|
|
262
|
the underlying revision storage object
|
|
263
|
|
|
263
|
|
|
264
|
A revlog consists of two parts, an index and the revision data.
|
|
264
|
A revlog consists of two parts, an index and the revision data.
|
|
265
|
|
|
265
|
|
|
266
|
The index is a file with a fixed record size containing
|
|
266
|
The index is a file with a fixed record size containing
|
|
267
|
information on each revision, including its nodeid (hash), the
|
|
267
|
information on each revision, including its nodeid (hash), the
|
|
268
|
nodeids of its parents, the position and offset of its data within
|
|
268
|
nodeids of its parents, the position and offset of its data within
|
|
269
|
the data file, and the revision it's based on. Finally, each entry
|
|
269
|
the data file, and the revision it's based on. Finally, each entry
|
|
270
|
contains a linkrev entry that can serve as a pointer to external
|
|
270
|
contains a linkrev entry that can serve as a pointer to external
|
|
271
|
data.
|
|
271
|
data.
|
|
272
|
|
|
272
|
|
|
273
|
The revision data itself is a linear collection of data chunks.
|
|
273
|
The revision data itself is a linear collection of data chunks.
|
|
274
|
Each chunk represents a revision and is usually represented as a
|
|
274
|
Each chunk represents a revision and is usually represented as a
|
|
275
|
delta against the previous chunk. To bound lookup time, runs of
|
|
275
|
delta against the previous chunk. To bound lookup time, runs of
|
|
276
|
deltas are limited to about 2 times the length of the original
|
|
276
|
deltas are limited to about 2 times the length of the original
|
|
277
|
version data. This makes retrieval of a version proportional to
|
|
277
|
version data. This makes retrieval of a version proportional to
|
|
278
|
its size, or O(1) relative to the number of revisions.
|
|
278
|
its size, or O(1) relative to the number of revisions.
|
|
279
|
|
|
279
|
|
|
280
|
Both pieces of the revlog are written to in an append-only
|
|
280
|
Both pieces of the revlog are written to in an append-only
|
|
281
|
fashion, which means we never need to rewrite a file to insert or
|
|
281
|
fashion, which means we never need to rewrite a file to insert or
|
|
282
|
remove data, and can use some simple techniques to avoid the need
|
|
282
|
remove data, and can use some simple techniques to avoid the need
|
|
283
|
for locking while reading.
|
|
283
|
for locking while reading.
|
|
284
|
|
|
284
|
|
|
285
|
If checkambig, indexfile is opened with checkambig=True at
|
|
285
|
If checkambig, indexfile is opened with checkambig=True at
|
|
286
|
writing, to avoid file stat ambiguity.
|
|
286
|
writing, to avoid file stat ambiguity.
|
|
287
|
|
|
287
|
|
|
288
|
If mmaplargeindex is True, and an mmapindexthreshold is set, the
|
|
288
|
If mmaplargeindex is True, and an mmapindexthreshold is set, the
|
|
289
|
index will be mmapped rather than read if it is larger than the
|
|
289
|
index will be mmapped rather than read if it is larger than the
|
|
290
|
configured threshold.
|
|
290
|
configured threshold.
|
|
291
|
|
|
291
|
|
|
292
|
If censorable is True, the revlog can have censored revisions.
|
|
292
|
If censorable is True, the revlog can have censored revisions.
|
|
293
|
|
|
293
|
|
|
294
|
If `upperboundcomp` is not None, this is the expected maximal gain from
|
|
294
|
If `upperboundcomp` is not None, this is the expected maximal gain from
|
|
295
|
compression for the data content.
|
|
295
|
compression for the data content.
|
|
296
|
|
|
296
|
|
|
297
|
`concurrencychecker` is an optional function that receives 3 arguments: a
|
|
297
|
`concurrencychecker` is an optional function that receives 3 arguments: a
|
|
298
|
file handle, a filename, and an expected position. It should check whether
|
|
298
|
file handle, a filename, and an expected position. It should check whether
|
|
299
|
the current position in the file handle is valid, and log/warn/fail (by
|
|
299
|
the current position in the file handle is valid, and log/warn/fail (by
|
|
300
|
raising).
|
|
300
|
raising).
|
|
301
|
|
|
301
|
|
|
302
|
|
|
302
|
|
|
303
|
Internal details
|
|
303
|
Internal details
|
|
304
|
----------------
|
|
304
|
----------------
|
|
305
|
|
|
305
|
|
|
306
|
A large part of the revlog logic deals with revisions' "index entries", tuple
|
|
306
|
A large part of the revlog logic deals with revisions' "index entries", tuple
|
|
307
|
objects that contains the same "items" whatever the revlog version.
|
|
307
|
objects that contains the same "items" whatever the revlog version.
|
|
308
|
Different versions will have different ways of storing these items (sometimes
|
|
308
|
Different versions will have different ways of storing these items (sometimes
|
|
309
|
not having them at all), but the tuple will always be the same. New fields
|
|
309
|
not having them at all), but the tuple will always be the same. New fields
|
|
310
|
are usually added at the end to avoid breaking existing code that relies
|
|
310
|
are usually added at the end to avoid breaking existing code that relies
|
|
311
|
on the existing order. The field are defined as follows:
|
|
311
|
on the existing order. The field are defined as follows:
|
|
312
|
|
|
312
|
|
|
313
|
[0] offset:
|
|
313
|
[0] offset:
|
|
314
|
The byte index of the start of revision data chunk.
|
|
314
|
The byte index of the start of revision data chunk.
|
|
315
|
That value is shifted up by 16 bits. use "offset = field >> 16" to
|
|
315
|
That value is shifted up by 16 bits. use "offset = field >> 16" to
|
|
316
|
retrieve it.
|
|
316
|
retrieve it.
|
|
317
|
|
|
317
|
|
|
318
|
flags:
|
|
318
|
flags:
|
|
319
|
A flag field that carries special information or changes the behavior
|
|
319
|
A flag field that carries special information or changes the behavior
|
|
320
|
of the revision. (see `REVIDX_*` constants for details)
|
|
320
|
of the revision. (see `REVIDX_*` constants for details)
|
|
321
|
The flag field only occupies the first 16 bits of this field,
|
|
321
|
The flag field only occupies the first 16 bits of this field,
|
|
322
|
use "flags = field & 0xFFFF" to retrieve the value.
|
|
322
|
use "flags = field & 0xFFFF" to retrieve the value.
|
|
323
|
|
|
323
|
|
|
324
|
[1] compressed length:
|
|
324
|
[1] compressed length:
|
|
325
|
The size, in bytes, of the chunk on disk
|
|
325
|
The size, in bytes, of the chunk on disk
|
|
326
|
|
|
326
|
|
|
327
|
[2] uncompressed length:
|
|
327
|
[2] uncompressed length:
|
|
328
|
The size, in bytes, of the full revision once reconstructed.
|
|
328
|
The size, in bytes, of the full revision once reconstructed.
|
|
329
|
|
|
329
|
|
|
330
|
[3] base rev:
|
|
330
|
[3] base rev:
|
|
331
|
Either the base of the revision delta chain (without general
|
|
331
|
Either the base of the revision delta chain (without general
|
|
332
|
delta), or the base of the delta (stored in the data chunk)
|
|
332
|
delta), or the base of the delta (stored in the data chunk)
|
|
333
|
with general delta.
|
|
333
|
with general delta.
|
|
334
|
|
|
334
|
|
|
335
|
[4] link rev:
|
|
335
|
[4] link rev:
|
|
336
|
Changelog revision number of the changeset introducing this
|
|
336
|
Changelog revision number of the changeset introducing this
|
|
337
|
revision.
|
|
337
|
revision.
|
|
338
|
|
|
338
|
|
|
339
|
[5] parent 1 rev:
|
|
339
|
[5] parent 1 rev:
|
|
340
|
Revision number of the first parent
|
|
340
|
Revision number of the first parent
|
|
341
|
|
|
341
|
|
|
342
|
[6] parent 2 rev:
|
|
342
|
[6] parent 2 rev:
|
|
343
|
Revision number of the second parent
|
|
343
|
Revision number of the second parent
|
|
344
|
|
|
344
|
|
|
345
|
[7] node id:
|
|
345
|
[7] node id:
|
|
346
|
The node id of the current revision
|
|
346
|
The node id of the current revision
|
|
347
|
|
|
347
|
|
|
348
|
[8] sidedata offset:
|
|
348
|
[8] sidedata offset:
|
|
349
|
The byte index of the start of the revision's side-data chunk.
|
|
349
|
The byte index of the start of the revision's side-data chunk.
|
|
350
|
|
|
350
|
|
|
351
|
[9] sidedata chunk length:
|
|
351
|
[9] sidedata chunk length:
|
|
352
|
The size, in bytes, of the revision's side-data chunk.
|
|
352
|
The size, in bytes, of the revision's side-data chunk.
|
|
353
|
|
|
353
|
|
|
354
|
[10] data compression mode:
|
|
354
|
[10] data compression mode:
|
|
355
|
two bits that detail the way the data chunk is compressed on disk.
|
|
355
|
two bits that detail the way the data chunk is compressed on disk.
|
|
356
|
(see "COMP_MODE_*" constants for details). For revlog version 0 and
|
|
356
|
(see "COMP_MODE_*" constants for details). For revlog version 0 and
|
|
357
|
1 this will always be COMP_MODE_INLINE.
|
|
357
|
1 this will always be COMP_MODE_INLINE.
|
|
358
|
|
|
358
|
|
|
359
|
[11] side-data compression mode:
|
|
359
|
[11] side-data compression mode:
|
|
360
|
two bits that detail the way the sidedata chunk is compressed on disk.
|
|
360
|
two bits that detail the way the sidedata chunk is compressed on disk.
|
|
361
|
(see "COMP_MODE_*" constants for details)
|
|
361
|
(see "COMP_MODE_*" constants for details)
|
|
362
|
"""
|
|
362
|
"""
|
|
363
|
|
|
363
|
|
|
364
|
_flagserrorclass = error.RevlogError
|
|
364
|
_flagserrorclass = error.RevlogError
|
|
365
|
|
|
365
|
|
|
366
|
def __init__(
|
|
366
|
def __init__(
|
|
367
|
self,
|
|
367
|
self,
|
|
368
|
opener,
|
|
368
|
opener,
|
|
369
|
target,
|
|
369
|
target,
|
|
370
|
radix,
|
|
370
|
radix,
|
|
371
|
postfix=None, # only exist for `tmpcensored` now
|
|
371
|
postfix=None, # only exist for `tmpcensored` now
|
|
372
|
checkambig=False,
|
|
372
|
checkambig=False,
|
|
373
|
mmaplargeindex=False,
|
|
373
|
mmaplargeindex=False,
|
|
374
|
censorable=False,
|
|
374
|
censorable=False,
|
|
375
|
upperboundcomp=None,
|
|
375
|
upperboundcomp=None,
|
|
376
|
persistentnodemap=False,
|
|
376
|
persistentnodemap=False,
|
|
377
|
concurrencychecker=None,
|
|
377
|
concurrencychecker=None,
|
|
378
|
trypending=False,
|
|
378
|
trypending=False,
|
|
379
|
):
|
|
379
|
):
|
|
380
|
"""
|
|
380
|
"""
|
|
381
|
create a revlog object
|
|
381
|
create a revlog object
|
|
382
|
|
|
382
|
|
|
383
|
opener is a function that abstracts the file opening operation
|
|
383
|
opener is a function that abstracts the file opening operation
|
|
384
|
and can be used to implement COW semantics or the like.
|
|
384
|
and can be used to implement COW semantics or the like.
|
|
385
|
|
|
385
|
|
|
386
|
`target`: a (KIND, ID) tuple that identify the content stored in
|
|
386
|
`target`: a (KIND, ID) tuple that identify the content stored in
|
|
387
|
this revlog. It help the rest of the code to understand what the revlog
|
|
387
|
this revlog. It help the rest of the code to understand what the revlog
|
|
388
|
is about without having to resort to heuristic and index filename
|
|
388
|
is about without having to resort to heuristic and index filename
|
|
389
|
analysis. Note: that this must be reliably be set by normal code, but
|
|
389
|
analysis. Note: that this must be reliably be set by normal code, but
|
|
390
|
that test, debug, or performance measurement code might not set this to
|
|
390
|
that test, debug, or performance measurement code might not set this to
|
|
391
|
accurate value.
|
|
391
|
accurate value.
|
|
392
|
"""
|
|
392
|
"""
|
|
393
|
self.upperboundcomp = upperboundcomp
|
|
393
|
self.upperboundcomp = upperboundcomp
|
|
394
|
|
|
394
|
|
|
395
|
self.radix = radix
|
|
395
|
self.radix = radix
|
|
396
|
|
|
396
|
|
|
397
|
self._docket_file = None
|
|
397
|
self._docket_file = None
|
|
398
|
self._indexfile = None
|
|
398
|
self._indexfile = None
|
|
399
|
self._datafile = None
|
|
399
|
self._datafile = None
|
|
400
|
self._nodemap_file = None
|
|
400
|
self._nodemap_file = None
|
|
401
|
self.postfix = postfix
|
|
401
|
self.postfix = postfix
|
|
402
|
self._trypending = trypending
|
|
402
|
self._trypending = trypending
|
|
403
|
self.opener = opener
|
|
403
|
self.opener = opener
|
|
404
|
if persistentnodemap:
|
|
404
|
if persistentnodemap:
|
|
405
|
self._nodemap_file = nodemaputil.get_nodemap_file(self)
|
|
405
|
self._nodemap_file = nodemaputil.get_nodemap_file(self)
|
|
406
|
|
|
406
|
|
|
407
|
assert target[0] in ALL_KINDS
|
|
407
|
assert target[0] in ALL_KINDS
|
|
408
|
assert len(target) == 2
|
|
408
|
assert len(target) == 2
|
|
409
|
self.target = target
|
|
409
|
self.target = target
|
|
410
|
# When True, indexfile is opened with checkambig=True at writing, to
|
|
410
|
# When True, indexfile is opened with checkambig=True at writing, to
|
|
411
|
# avoid file stat ambiguity.
|
|
411
|
# avoid file stat ambiguity.
|
|
412
|
self._checkambig = checkambig
|
|
412
|
self._checkambig = checkambig
|
|
413
|
self._mmaplargeindex = mmaplargeindex
|
|
413
|
self._mmaplargeindex = mmaplargeindex
|
|
414
|
self._censorable = censorable
|
|
414
|
self._censorable = censorable
|
|
415
|
# 3-tuple of (node, rev, text) for a raw revision.
|
|
415
|
# 3-tuple of (node, rev, text) for a raw revision.
|
|
416
|
self._revisioncache = None
|
|
416
|
self._revisioncache = None
|
|
417
|
# Maps rev to chain base rev.
|
|
417
|
# Maps rev to chain base rev.
|
|
418
|
self._chainbasecache = util.lrucachedict(100)
|
|
418
|
self._chainbasecache = util.lrucachedict(100)
|
|
419
|
# 2-tuple of (offset, data) of raw data from the revlog at an offset.
|
|
419
|
# 2-tuple of (offset, data) of raw data from the revlog at an offset.
|
|
420
|
self._chunkcache = (0, b'')
|
|
420
|
self._chunkcache = (0, b'')
|
|
421
|
# How much data to read and cache into the raw revlog data cache.
|
|
421
|
# How much data to read and cache into the raw revlog data cache.
|
|
422
|
self._chunkcachesize = 65536
|
|
422
|
self._chunkcachesize = 65536
|
|
423
|
self._maxchainlen = None
|
|
423
|
self._maxchainlen = None
|
|
424
|
self._deltabothparents = True
|
|
424
|
self._deltabothparents = True
|
|
425
|
self.index = None
|
|
425
|
self.index = None
|
|
426
|
self._docket = None
|
|
426
|
self._docket = None
|
|
427
|
self._nodemap_docket = None
|
|
427
|
self._nodemap_docket = None
|
|
428
|
# Mapping of partial identifiers to full nodes.
|
|
428
|
# Mapping of partial identifiers to full nodes.
|
|
429
|
self._pcache = {}
|
|
429
|
self._pcache = {}
|
|
430
|
# Mapping of revision integer to full node.
|
|
430
|
# Mapping of revision integer to full node.
|
|
431
|
self._compengine = b'zlib'
|
|
431
|
self._compengine = b'zlib'
|
|
432
|
self._compengineopts = {}
|
|
432
|
self._compengineopts = {}
|
|
433
|
self._maxdeltachainspan = -1
|
|
433
|
self._maxdeltachainspan = -1
|
|
434
|
self._withsparseread = False
|
|
434
|
self._withsparseread = False
|
|
435
|
self._sparserevlog = False
|
|
435
|
self._sparserevlog = False
|
|
436
|
self.hassidedata = False
|
|
436
|
self.hassidedata = False
|
|
437
|
self._srdensitythreshold = 0.50
|
|
437
|
self._srdensitythreshold = 0.50
|
|
438
|
self._srmingapsize = 262144
|
|
438
|
self._srmingapsize = 262144
|
|
439
|
|
|
439
|
|
|
440
|
# Make copy of flag processors so each revlog instance can support
|
|
440
|
# Make copy of flag processors so each revlog instance can support
|
|
441
|
# custom flags.
|
|
441
|
# custom flags.
|
|
442
|
self._flagprocessors = dict(flagutil.flagprocessors)
|
|
442
|
self._flagprocessors = dict(flagutil.flagprocessors)
|
|
443
|
|
|
443
|
|
|
444
|
# 2-tuple of file handles being used for active writing.
|
|
444
|
# 2-tuple of file handles being used for active writing.
|
|
445
|
self._writinghandles = None
|
|
445
|
self._writinghandles = None
|
|
446
|
# prevent nesting of addgroup
|
|
446
|
# prevent nesting of addgroup
|
|
447
|
self._adding_group = None
|
|
447
|
self._adding_group = None
|
|
448
|
|
|
448
|
|
|
449
|
self._loadindex()
|
|
449
|
self._loadindex()
|
|
450
|
|
|
450
|
|
|
451
|
self._concurrencychecker = concurrencychecker
|
|
451
|
self._concurrencychecker = concurrencychecker
|
|
452
|
|
|
452
|
|
|
453
|
def _init_opts(self):
|
|
453
|
def _init_opts(self):
|
|
454
|
"""process options (from above/config) to setup associated default revlog mode
|
|
454
|
"""process options (from above/config) to setup associated default revlog mode
|
|
455
|
|
|
455
|
|
|
456
|
These values might be affected when actually reading on disk information.
|
|
456
|
These values might be affected when actually reading on disk information.
|
|
457
|
|
|
457
|
|
|
458
|
The relevant values are returned for use in _loadindex().
|
|
458
|
The relevant values are returned for use in _loadindex().
|
|
459
|
|
|
459
|
|
|
460
|
* newversionflags:
|
|
460
|
* newversionflags:
|
|
461
|
version header to use if we need to create a new revlog
|
|
461
|
version header to use if we need to create a new revlog
|
|
462
|
|
|
462
|
|
|
463
|
* mmapindexthreshold:
|
|
463
|
* mmapindexthreshold:
|
|
464
|
minimal index size for start to use mmap
|
|
464
|
minimal index size for start to use mmap
|
|
465
|
|
|
465
|
|
|
466
|
* force_nodemap:
|
|
466
|
* force_nodemap:
|
|
467
|
force the usage of a "development" version of the nodemap code
|
|
467
|
force the usage of a "development" version of the nodemap code
|
|
468
|
"""
|
|
468
|
"""
|
|
469
|
mmapindexthreshold = None
|
|
469
|
mmapindexthreshold = None
|
|
470
|
opts = self.opener.options
|
|
470
|
opts = self.opener.options
|
|
471
|
|
|
471
|
|
|
472
|
if b'changelogv2' in opts and self.revlog_kind == KIND_CHANGELOG:
|
|
472
|
if b'changelogv2' in opts and self.revlog_kind == KIND_CHANGELOG:
|
|
473
|
new_header = CHANGELOGV2
|
|
473
|
new_header = CHANGELOGV2
|
|
474
|
elif b'revlogv2' in opts:
|
|
474
|
elif b'revlogv2' in opts:
|
|
475
|
new_header = REVLOGV2
|
|
475
|
new_header = REVLOGV2
|
|
476
|
elif b'revlogv1' in opts:
|
|
476
|
elif b'revlogv1' in opts:
|
|
477
|
new_header = REVLOGV1 | FLAG_INLINE_DATA
|
|
477
|
new_header = REVLOGV1 | FLAG_INLINE_DATA
|
|
478
|
if b'generaldelta' in opts:
|
|
478
|
if b'generaldelta' in opts:
|
|
479
|
new_header |= FLAG_GENERALDELTA
|
|
479
|
new_header |= FLAG_GENERALDELTA
|
|
480
|
elif b'revlogv0' in self.opener.options:
|
|
480
|
elif b'revlogv0' in self.opener.options:
|
|
481
|
new_header = REVLOGV0
|
|
481
|
new_header = REVLOGV0
|
|
482
|
else:
|
|
482
|
else:
|
|
483
|
new_header = REVLOG_DEFAULT_VERSION
|
|
483
|
new_header = REVLOG_DEFAULT_VERSION
|
|
484
|
|
|
484
|
|
|
485
|
if b'chunkcachesize' in opts:
|
|
485
|
if b'chunkcachesize' in opts:
|
|
486
|
self._chunkcachesize = opts[b'chunkcachesize']
|
|
486
|
self._chunkcachesize = opts[b'chunkcachesize']
|
|
487
|
if b'maxchainlen' in opts:
|
|
487
|
if b'maxchainlen' in opts:
|
|
488
|
self._maxchainlen = opts[b'maxchainlen']
|
|
488
|
self._maxchainlen = opts[b'maxchainlen']
|
|
489
|
if b'deltabothparents' in opts:
|
|
489
|
if b'deltabothparents' in opts:
|
|
490
|
self._deltabothparents = opts[b'deltabothparents']
|
|
490
|
self._deltabothparents = opts[b'deltabothparents']
|
|
491
|
self._lazydelta = bool(opts.get(b'lazydelta', True))
|
|
491
|
self._lazydelta = bool(opts.get(b'lazydelta', True))
|
|
492
|
self._lazydeltabase = False
|
|
492
|
self._lazydeltabase = False
|
|
493
|
if self._lazydelta:
|
|
493
|
if self._lazydelta:
|
|
494
|
self._lazydeltabase = bool(opts.get(b'lazydeltabase', False))
|
|
494
|
self._lazydeltabase = bool(opts.get(b'lazydeltabase', False))
|
|
495
|
if b'compengine' in opts:
|
|
495
|
if b'compengine' in opts:
|
|
496
|
self._compengine = opts[b'compengine']
|
|
496
|
self._compengine = opts[b'compengine']
|
|
497
|
if b'zlib.level' in opts:
|
|
497
|
if b'zlib.level' in opts:
|
|
498
|
self._compengineopts[b'zlib.level'] = opts[b'zlib.level']
|
|
498
|
self._compengineopts[b'zlib.level'] = opts[b'zlib.level']
|
|
499
|
if b'zstd.level' in opts:
|
|
499
|
if b'zstd.level' in opts:
|
|
500
|
self._compengineopts[b'zstd.level'] = opts[b'zstd.level']
|
|
500
|
self._compengineopts[b'zstd.level'] = opts[b'zstd.level']
|
|
501
|
if b'maxdeltachainspan' in opts:
|
|
501
|
if b'maxdeltachainspan' in opts:
|
|
502
|
self._maxdeltachainspan = opts[b'maxdeltachainspan']
|
|
502
|
self._maxdeltachainspan = opts[b'maxdeltachainspan']
|
|
503
|
if self._mmaplargeindex and b'mmapindexthreshold' in opts:
|
|
503
|
if self._mmaplargeindex and b'mmapindexthreshold' in opts:
|
|
504
|
mmapindexthreshold = opts[b'mmapindexthreshold']
|
|
504
|
mmapindexthreshold = opts[b'mmapindexthreshold']
|
|
505
|
self._sparserevlog = bool(opts.get(b'sparse-revlog', False))
|
|
505
|
self._sparserevlog = bool(opts.get(b'sparse-revlog', False))
|
|
506
|
withsparseread = bool(opts.get(b'with-sparse-read', False))
|
|
506
|
withsparseread = bool(opts.get(b'with-sparse-read', False))
|
|
507
|
# sparse-revlog forces sparse-read
|
|
507
|
# sparse-revlog forces sparse-read
|
|
508
|
self._withsparseread = self._sparserevlog or withsparseread
|
|
508
|
self._withsparseread = self._sparserevlog or withsparseread
|
|
509
|
if b'sparse-read-density-threshold' in opts:
|
|
509
|
if b'sparse-read-density-threshold' in opts:
|
|
510
|
self._srdensitythreshold = opts[b'sparse-read-density-threshold']
|
|
510
|
self._srdensitythreshold = opts[b'sparse-read-density-threshold']
|
|
511
|
if b'sparse-read-min-gap-size' in opts:
|
|
511
|
if b'sparse-read-min-gap-size' in opts:
|
|
512
|
self._srmingapsize = opts[b'sparse-read-min-gap-size']
|
|
512
|
self._srmingapsize = opts[b'sparse-read-min-gap-size']
|
|
513
|
if opts.get(b'enableellipsis'):
|
|
513
|
if opts.get(b'enableellipsis'):
|
|
514
|
self._flagprocessors[REVIDX_ELLIPSIS] = ellipsisprocessor
|
|
514
|
self._flagprocessors[REVIDX_ELLIPSIS] = ellipsisprocessor
|
|
515
|
|
|
515
|
|
|
516
|
# revlog v0 doesn't have flag processors
|
|
516
|
# revlog v0 doesn't have flag processors
|
|
517
|
for flag, processor in pycompat.iteritems(
|
|
517
|
for flag, processor in pycompat.iteritems(
|
|
518
|
opts.get(b'flagprocessors', {})
|
|
518
|
opts.get(b'flagprocessors', {})
|
|
519
|
):
|
|
519
|
):
|
|
520
|
flagutil.insertflagprocessor(flag, processor, self._flagprocessors)
|
|
520
|
flagutil.insertflagprocessor(flag, processor, self._flagprocessors)
|
|
521
|
|
|
521
|
|
|
522
|
if self._chunkcachesize <= 0:
|
|
522
|
if self._chunkcachesize <= 0:
|
|
523
|
raise error.RevlogError(
|
|
523
|
raise error.RevlogError(
|
|
524
|
_(b'revlog chunk cache size %r is not greater than 0')
|
|
524
|
_(b'revlog chunk cache size %r is not greater than 0')
|
|
525
|
% self._chunkcachesize
|
|
525
|
% self._chunkcachesize
|
|
526
|
)
|
|
526
|
)
|
|
527
|
elif self._chunkcachesize & (self._chunkcachesize - 1):
|
|
527
|
elif self._chunkcachesize & (self._chunkcachesize - 1):
|
|
528
|
raise error.RevlogError(
|
|
528
|
raise error.RevlogError(
|
|
529
|
_(b'revlog chunk cache size %r is not a power of 2')
|
|
529
|
_(b'revlog chunk cache size %r is not a power of 2')
|
|
530
|
% self._chunkcachesize
|
|
530
|
% self._chunkcachesize
|
|
531
|
)
|
|
531
|
)
|
|
532
|
force_nodemap = opts.get(b'devel-force-nodemap', False)
|
|
532
|
force_nodemap = opts.get(b'devel-force-nodemap', False)
|
|
533
|
return new_header, mmapindexthreshold, force_nodemap
|
|
533
|
return new_header, mmapindexthreshold, force_nodemap
|
|
534
|
|
|
534
|
|
|
535
|
def _get_data(self, filepath, mmap_threshold, size=None):
|
|
535
|
def _get_data(self, filepath, mmap_threshold, size=None):
|
|
536
|
"""return a file content with or without mmap
|
|
536
|
"""return a file content with or without mmap
|
|
537
|
|
|
537
|
|
|
538
|
If the file is missing return the empty string"""
|
|
538
|
If the file is missing return the empty string"""
|
|
539
|
try:
|
|
539
|
try:
|
|
540
|
with self.opener(filepath) as fp:
|
|
540
|
with self.opener(filepath) as fp:
|
|
541
|
if mmap_threshold is not None:
|
|
541
|
if mmap_threshold is not None:
|
|
542
|
file_size = self.opener.fstat(fp).st_size
|
|
542
|
file_size = self.opener.fstat(fp).st_size
|
|
543
|
if file_size >= mmap_threshold:
|
|
543
|
if file_size >= mmap_threshold:
|
|
544
|
if size is not None:
|
|
544
|
if size is not None:
|
|
545
|
# avoid potentiel mmap crash
|
|
545
|
# avoid potentiel mmap crash
|
|
546
|
size = min(file_size, size)
|
|
546
|
size = min(file_size, size)
|
|
547
|
# TODO: should .close() to release resources without
|
|
547
|
# TODO: should .close() to release resources without
|
|
548
|
# relying on Python GC
|
|
548
|
# relying on Python GC
|
|
549
|
if size is None:
|
|
549
|
if size is None:
|
|
550
|
return util.buffer(util.mmapread(fp))
|
|
550
|
return util.buffer(util.mmapread(fp))
|
|
551
|
else:
|
|
551
|
else:
|
|
552
|
return util.buffer(util.mmapread(fp, size))
|
|
552
|
return util.buffer(util.mmapread(fp, size))
|
|
553
|
if size is None:
|
|
553
|
if size is None:
|
|
554
|
return fp.read()
|
|
554
|
return fp.read()
|
|
555
|
else:
|
|
555
|
else:
|
|
556
|
return fp.read(size)
|
|
556
|
return fp.read(size)
|
|
557
|
except IOError as inst:
|
|
557
|
except IOError as inst:
|
|
558
|
if inst.errno != errno.ENOENT:
|
|
558
|
if inst.errno != errno.ENOENT:
|
|
559
|
raise
|
|
559
|
raise
|
|
560
|
return b''
|
|
560
|
return b''
|
|
561
|
|
|
561
|
|
|
562
|
def _loadindex(self):
|
|
562
|
def _loadindex(self):
|
|
563
|
|
|
563
|
|
|
564
|
new_header, mmapindexthreshold, force_nodemap = self._init_opts()
|
|
564
|
new_header, mmapindexthreshold, force_nodemap = self._init_opts()
|
|
565
|
|
|
565
|
|
|
566
|
if self.postfix is not None:
|
|
566
|
if self.postfix is not None:
|
|
567
|
entry_point = b'%s.i.%s' % (self.radix, self.postfix)
|
|
567
|
entry_point = b'%s.i.%s' % (self.radix, self.postfix)
|
|
568
|
elif self._trypending and self.opener.exists(b'%s.i.a' % self.radix):
|
|
568
|
elif self._trypending and self.opener.exists(b'%s.i.a' % self.radix):
|
|
569
|
entry_point = b'%s.i.a' % self.radix
|
|
569
|
entry_point = b'%s.i.a' % self.radix
|
|
570
|
else:
|
|
570
|
else:
|
|
571
|
entry_point = b'%s.i' % self.radix
|
|
571
|
entry_point = b'%s.i' % self.radix
|
|
572
|
|
|
572
|
|
|
573
|
entry_data = b''
|
|
573
|
entry_data = b''
|
|
574
|
self._initempty = True
|
|
574
|
self._initempty = True
|
|
575
|
entry_data = self._get_data(entry_point, mmapindexthreshold)
|
|
575
|
entry_data = self._get_data(entry_point, mmapindexthreshold)
|
|
576
|
if len(entry_data) > 0:
|
|
576
|
if len(entry_data) > 0:
|
|
577
|
header = INDEX_HEADER.unpack(entry_data[:4])[0]
|
|
577
|
header = INDEX_HEADER.unpack(entry_data[:4])[0]
|
|
578
|
self._initempty = False
|
|
578
|
self._initempty = False
|
|
579
|
else:
|
|
579
|
else:
|
|
580
|
header = new_header
|
|
580
|
header = new_header
|
|
581
|
|
|
581
|
|
|
582
|
self._format_flags = header & ~0xFFFF
|
|
582
|
self._format_flags = header & ~0xFFFF
|
|
583
|
self._format_version = header & 0xFFFF
|
|
583
|
self._format_version = header & 0xFFFF
|
|
584
|
|
|
584
|
|
|
585
|
supported_flags = SUPPORTED_FLAGS.get(self._format_version)
|
|
585
|
supported_flags = SUPPORTED_FLAGS.get(self._format_version)
|
|
586
|
if supported_flags is None:
|
|
586
|
if supported_flags is None:
|
|
587
|
msg = _(b'unknown version (%d) in revlog %s')
|
|
587
|
msg = _(b'unknown version (%d) in revlog %s')
|
|
588
|
msg %= (self._format_version, self.display_id)
|
|
588
|
msg %= (self._format_version, self.display_id)
|
|
589
|
raise error.RevlogError(msg)
|
|
589
|
raise error.RevlogError(msg)
|
|
590
|
elif self._format_flags & ~supported_flags:
|
|
590
|
elif self._format_flags & ~supported_flags:
|
|
591
|
msg = _(b'unknown flags (%#04x) in version %d revlog %s')
|
|
591
|
msg = _(b'unknown flags (%#04x) in version %d revlog %s')
|
|
592
|
display_flag = self._format_flags >> 16
|
|
592
|
display_flag = self._format_flags >> 16
|
|
593
|
msg %= (display_flag, self._format_version, self.display_id)
|
|
593
|
msg %= (display_flag, self._format_version, self.display_id)
|
|
594
|
raise error.RevlogError(msg)
|
|
594
|
raise error.RevlogError(msg)
|
|
595
|
|
|
595
|
|
|
596
|
features = FEATURES_BY_VERSION[self._format_version]
|
|
596
|
features = FEATURES_BY_VERSION[self._format_version]
|
|
597
|
self._inline = features[b'inline'](self._format_flags)
|
|
597
|
self._inline = features[b'inline'](self._format_flags)
|
|
598
|
self._generaldelta = features[b'generaldelta'](self._format_flags)
|
|
598
|
self._generaldelta = features[b'generaldelta'](self._format_flags)
|
|
599
|
self.hassidedata = features[b'sidedata']
|
|
599
|
self.hassidedata = features[b'sidedata']
|
|
600
|
|
|
600
|
|
|
601
|
if not features[b'docket']:
|
|
601
|
if not features[b'docket']:
|
|
602
|
self._indexfile = entry_point
|
|
602
|
self._indexfile = entry_point
|
|
603
|
index_data = entry_data
|
|
603
|
index_data = entry_data
|
|
604
|
else:
|
|
604
|
else:
|
|
605
|
self._docket_file = entry_point
|
|
605
|
self._docket_file = entry_point
|
|
606
|
if self._initempty:
|
|
606
|
if self._initempty:
|
|
607
|
self._docket = docketutil.default_docket(self, header)
|
|
607
|
self._docket = docketutil.default_docket(self, header)
|
|
608
|
else:
|
|
608
|
else:
|
|
609
|
self._docket = docketutil.parse_docket(
|
|
609
|
self._docket = docketutil.parse_docket(
|
|
610
|
self, entry_data, use_pending=self._trypending
|
|
610
|
self, entry_data, use_pending=self._trypending
|
|
611
|
)
|
|
611
|
)
|
|
612
|
self._indexfile = self._docket.index_filepath()
|
|
612
|
self._indexfile = self._docket.index_filepath()
|
|
613
|
index_data = b''
|
|
613
|
index_data = b''
|
|
614
|
index_size = self._docket.index_end
|
|
614
|
index_size = self._docket.index_end
|
|
615
|
if index_size > 0:
|
|
615
|
if index_size > 0:
|
|
616
|
index_data = self._get_data(
|
|
616
|
index_data = self._get_data(
|
|
617
|
self._indexfile, mmapindexthreshold, size=index_size
|
|
617
|
self._indexfile, mmapindexthreshold, size=index_size
|
|
618
|
)
|
|
618
|
)
|
|
619
|
if len(index_data) < index_size:
|
|
619
|
if len(index_data) < index_size:
|
|
620
|
msg = _(b'too few index data for %s: got %d, expected %d')
|
|
620
|
msg = _(b'too few index data for %s: got %d, expected %d')
|
|
621
|
msg %= (self.display_id, len(index_data), index_size)
|
|
621
|
msg %= (self.display_id, len(index_data), index_size)
|
|
622
|
raise error.RevlogError(msg)
|
|
622
|
raise error.RevlogError(msg)
|
|
623
|
|
|
623
|
|
|
624
|
self._inline = False
|
|
624
|
self._inline = False
|
|
625
|
# generaldelta implied by version 2 revlogs.
|
|
625
|
# generaldelta implied by version 2 revlogs.
|
|
626
|
self._generaldelta = True
|
|
626
|
self._generaldelta = True
|
|
627
|
# the logic for persistent nodemap will be dealt with within the
|
|
627
|
# the logic for persistent nodemap will be dealt with within the
|
|
628
|
# main docket, so disable it for now.
|
|
628
|
# main docket, so disable it for now.
|
|
629
|
self._nodemap_file = None
|
|
629
|
self._nodemap_file = None
|
|
630
|
|
|
630
|
|
|
631
|
if self.postfix is None:
|
|
631
|
if self._docket is not None:
|
|
|
|
|
632
|
self._datafile = self._docket.data_filepath()
|
|
|
|
|
633
|
elif self.postfix is None:
|
|
632
|
self._datafile = b'%s.d' % self.radix
|
|
634
|
self._datafile = b'%s.d' % self.radix
|
|
633
|
else:
|
|
635
|
else:
|
|
634
|
self._datafile = b'%s.d.%s' % (self.radix, self.postfix)
|
|
636
|
self._datafile = b'%s.d.%s' % (self.radix, self.postfix)
|
|
635
|
|
|
637
|
|
|
636
|
self.nodeconstants = sha1nodeconstants
|
|
638
|
self.nodeconstants = sha1nodeconstants
|
|
637
|
self.nullid = self.nodeconstants.nullid
|
|
639
|
self.nullid = self.nodeconstants.nullid
|
|
638
|
|
|
640
|
|
|
639
|
# sparse-revlog can't be on without general-delta (issue6056)
|
|
641
|
# sparse-revlog can't be on without general-delta (issue6056)
|
|
640
|
if not self._generaldelta:
|
|
642
|
if not self._generaldelta:
|
|
641
|
self._sparserevlog = False
|
|
643
|
self._sparserevlog = False
|
|
642
|
|
|
644
|
|
|
643
|
self._storedeltachains = True
|
|
645
|
self._storedeltachains = True
|
|
644
|
|
|
646
|
|
|
645
|
devel_nodemap = (
|
|
647
|
devel_nodemap = (
|
|
646
|
self._nodemap_file
|
|
648
|
self._nodemap_file
|
|
647
|
and force_nodemap
|
|
649
|
and force_nodemap
|
|
648
|
and parse_index_v1_nodemap is not None
|
|
650
|
and parse_index_v1_nodemap is not None
|
|
649
|
)
|
|
651
|
)
|
|
650
|
|
|
652
|
|
|
651
|
use_rust_index = False
|
|
653
|
use_rust_index = False
|
|
652
|
if rustrevlog is not None:
|
|
654
|
if rustrevlog is not None:
|
|
653
|
if self._nodemap_file is not None:
|
|
655
|
if self._nodemap_file is not None:
|
|
654
|
use_rust_index = True
|
|
656
|
use_rust_index = True
|
|
655
|
else:
|
|
657
|
else:
|
|
656
|
use_rust_index = self.opener.options.get(b'rust.index')
|
|
658
|
use_rust_index = self.opener.options.get(b'rust.index')
|
|
657
|
|
|
659
|
|
|
658
|
self._parse_index = parse_index_v1
|
|
660
|
self._parse_index = parse_index_v1
|
|
659
|
if self._format_version == REVLOGV0:
|
|
661
|
if self._format_version == REVLOGV0:
|
|
660
|
self._parse_index = revlogv0.parse_index_v0
|
|
662
|
self._parse_index = revlogv0.parse_index_v0
|
|
661
|
elif self._format_version == REVLOGV2:
|
|
663
|
elif self._format_version == REVLOGV2:
|
|
662
|
self._parse_index = parse_index_v2
|
|
664
|
self._parse_index = parse_index_v2
|
|
663
|
elif self._format_version == CHANGELOGV2:
|
|
665
|
elif self._format_version == CHANGELOGV2:
|
|
664
|
self._parse_index = parse_index_cl_v2
|
|
666
|
self._parse_index = parse_index_cl_v2
|
|
665
|
elif devel_nodemap:
|
|
667
|
elif devel_nodemap:
|
|
666
|
self._parse_index = parse_index_v1_nodemap
|
|
668
|
self._parse_index = parse_index_v1_nodemap
|
|
667
|
elif use_rust_index:
|
|
669
|
elif use_rust_index:
|
|
668
|
self._parse_index = parse_index_v1_mixed
|
|
670
|
self._parse_index = parse_index_v1_mixed
|
|
669
|
try:
|
|
671
|
try:
|
|
670
|
d = self._parse_index(index_data, self._inline)
|
|
672
|
d = self._parse_index(index_data, self._inline)
|
|
671
|
index, _chunkcache = d
|
|
673
|
index, _chunkcache = d
|
|
672
|
use_nodemap = (
|
|
674
|
use_nodemap = (
|
|
673
|
not self._inline
|
|
675
|
not self._inline
|
|
674
|
and self._nodemap_file is not None
|
|
676
|
and self._nodemap_file is not None
|
|
675
|
and util.safehasattr(index, 'update_nodemap_data')
|
|
677
|
and util.safehasattr(index, 'update_nodemap_data')
|
|
676
|
)
|
|
678
|
)
|
|
677
|
if use_nodemap:
|
|
679
|
if use_nodemap:
|
|
678
|
nodemap_data = nodemaputil.persisted_data(self)
|
|
680
|
nodemap_data = nodemaputil.persisted_data(self)
|
|
679
|
if nodemap_data is not None:
|
|
681
|
if nodemap_data is not None:
|
|
680
|
docket = nodemap_data[0]
|
|
682
|
docket = nodemap_data[0]
|
|
681
|
if (
|
|
683
|
if (
|
|
682
|
len(d[0]) > docket.tip_rev
|
|
684
|
len(d[0]) > docket.tip_rev
|
|
683
|
and d[0][docket.tip_rev][7] == docket.tip_node
|
|
685
|
and d[0][docket.tip_rev][7] == docket.tip_node
|
|
684
|
):
|
|
686
|
):
|
|
685
|
# no changelog tampering
|
|
687
|
# no changelog tampering
|
|
686
|
self._nodemap_docket = docket
|
|
688
|
self._nodemap_docket = docket
|
|
687
|
index.update_nodemap_data(*nodemap_data)
|
|
689
|
index.update_nodemap_data(*nodemap_data)
|
|
688
|
except (ValueError, IndexError):
|
|
690
|
except (ValueError, IndexError):
|
|
689
|
raise error.RevlogError(
|
|
691
|
raise error.RevlogError(
|
|
690
|
_(b"index %s is corrupted") % self.display_id
|
|
692
|
_(b"index %s is corrupted") % self.display_id
|
|
691
|
)
|
|
693
|
)
|
|
692
|
self.index, self._chunkcache = d
|
|
694
|
self.index, self._chunkcache = d
|
|
693
|
if not self._chunkcache:
|
|
695
|
if not self._chunkcache:
|
|
694
|
self._chunkclear()
|
|
696
|
self._chunkclear()
|
|
695
|
# revnum -> (chain-length, sum-delta-length)
|
|
697
|
# revnum -> (chain-length, sum-delta-length)
|
|
696
|
self._chaininfocache = util.lrucachedict(500)
|
|
698
|
self._chaininfocache = util.lrucachedict(500)
|
|
697
|
# revlog header -> revlog compressor
|
|
699
|
# revlog header -> revlog compressor
|
|
698
|
self._decompressors = {}
|
|
700
|
self._decompressors = {}
|
|
699
|
|
|
701
|
|
|
700
|
@util.propertycache
|
|
702
|
@util.propertycache
|
|
701
|
def revlog_kind(self):
|
|
703
|
def revlog_kind(self):
|
|
702
|
return self.target[0]
|
|
704
|
return self.target[0]
|
|
703
|
|
|
705
|
|
|
704
|
@util.propertycache
|
|
706
|
@util.propertycache
|
|
705
|
def display_id(self):
|
|
707
|
def display_id(self):
|
|
706
|
"""The public facing "ID" of the revlog that we use in message"""
|
|
708
|
"""The public facing "ID" of the revlog that we use in message"""
|
|
707
|
# Maybe we should build a user facing representation of
|
|
709
|
# Maybe we should build a user facing representation of
|
|
708
|
# revlog.target instead of using `self.radix`
|
|
710
|
# revlog.target instead of using `self.radix`
|
|
709
|
return self.radix
|
|
711
|
return self.radix
|
|
710
|
|
|
712
|
|
|
711
|
def _get_decompressor(self, t):
|
|
713
|
def _get_decompressor(self, t):
|
|
712
|
try:
|
|
714
|
try:
|
|
713
|
compressor = self._decompressors[t]
|
|
715
|
compressor = self._decompressors[t]
|
|
714
|
except KeyError:
|
|
716
|
except KeyError:
|
|
715
|
try:
|
|
717
|
try:
|
|
716
|
engine = util.compengines.forrevlogheader(t)
|
|
718
|
engine = util.compengines.forrevlogheader(t)
|
|
717
|
compressor = engine.revlogcompressor(self._compengineopts)
|
|
719
|
compressor = engine.revlogcompressor(self._compengineopts)
|
|
718
|
self._decompressors[t] = compressor
|
|
720
|
self._decompressors[t] = compressor
|
|
719
|
except KeyError:
|
|
721
|
except KeyError:
|
|
720
|
raise error.RevlogError(
|
|
722
|
raise error.RevlogError(
|
|
721
|
_(b'unknown compression type %s') % binascii.hexlify(t)
|
|
723
|
_(b'unknown compression type %s') % binascii.hexlify(t)
|
|
722
|
)
|
|
724
|
)
|
|
723
|
return compressor
|
|
725
|
return compressor
|
|
724
|
|
|
726
|
|
|
725
|
@util.propertycache
|
|
727
|
@util.propertycache
|
|
726
|
def _compressor(self):
|
|
728
|
def _compressor(self):
|
|
727
|
engine = util.compengines[self._compengine]
|
|
729
|
engine = util.compengines[self._compengine]
|
|
728
|
return engine.revlogcompressor(self._compengineopts)
|
|
730
|
return engine.revlogcompressor(self._compengineopts)
|
|
729
|
|
|
731
|
|
|
730
|
@util.propertycache
|
|
732
|
@util.propertycache
|
|
731
|
def _decompressor(self):
|
|
733
|
def _decompressor(self):
|
|
732
|
"""the default decompressor"""
|
|
734
|
"""the default decompressor"""
|
|
733
|
if self._docket is None:
|
|
735
|
if self._docket is None:
|
|
734
|
return None
|
|
736
|
return None
|
|
735
|
t = self._docket.default_compression_header
|
|
737
|
t = self._docket.default_compression_header
|
|
736
|
c = self._get_decompressor(t)
|
|
738
|
c = self._get_decompressor(t)
|
|
737
|
return c.decompress
|
|
739
|
return c.decompress
|
|
738
|
|
|
740
|
|
|
739
|
def _indexfp(self):
|
|
741
|
def _indexfp(self):
|
|
740
|
"""file object for the revlog's index file"""
|
|
742
|
"""file object for the revlog's index file"""
|
|
741
|
return self.opener(self._indexfile, mode=b"r")
|
|
743
|
return self.opener(self._indexfile, mode=b"r")
|
|
742
|
|
|
744
|
|
|
743
|
def __index_write_fp(self):
|
|
745
|
def __index_write_fp(self):
|
|
744
|
# You should not use this directly and use `_writing` instead
|
|
746
|
# You should not use this directly and use `_writing` instead
|
|
745
|
try:
|
|
747
|
try:
|
|
746
|
f = self.opener(
|
|
748
|
f = self.opener(
|
|
747
|
self._indexfile, mode=b"r+", checkambig=self._checkambig
|
|
749
|
self._indexfile, mode=b"r+", checkambig=self._checkambig
|
|
748
|
)
|
|
750
|
)
|
|
749
|
if self._docket is None:
|
|
751
|
if self._docket is None:
|
|
750
|
f.seek(0, os.SEEK_END)
|
|
752
|
f.seek(0, os.SEEK_END)
|
|
751
|
else:
|
|
753
|
else:
|
|
752
|
f.seek(self._docket.index_end, os.SEEK_SET)
|
|
754
|
f.seek(self._docket.index_end, os.SEEK_SET)
|
|
753
|
return f
|
|
755
|
return f
|
|
754
|
except IOError as inst:
|
|
756
|
except IOError as inst:
|
|
755
|
if inst.errno != errno.ENOENT:
|
|
757
|
if inst.errno != errno.ENOENT:
|
|
756
|
raise
|
|
758
|
raise
|
|
757
|
return self.opener(
|
|
759
|
return self.opener(
|
|
758
|
self._indexfile, mode=b"w+", checkambig=self._checkambig
|
|
760
|
self._indexfile, mode=b"w+", checkambig=self._checkambig
|
|
759
|
)
|
|
761
|
)
|
|
760
|
|
|
762
|
|
|
761
|
def __index_new_fp(self):
|
|
763
|
def __index_new_fp(self):
|
|
762
|
# You should not use this unless you are upgrading from inline revlog
|
|
764
|
# You should not use this unless you are upgrading from inline revlog
|
|
763
|
return self.opener(
|
|
765
|
return self.opener(
|
|
764
|
self._indexfile,
|
|
766
|
self._indexfile,
|
|
765
|
mode=b"w",
|
|
767
|
mode=b"w",
|
|
766
|
checkambig=self._checkambig,
|
|
768
|
checkambig=self._checkambig,
|
|
767
|
atomictemp=True,
|
|
769
|
atomictemp=True,
|
|
768
|
)
|
|
770
|
)
|
|
769
|
|
|
771
|
|
|
770
|
def _datafp(self, mode=b'r'):
|
|
772
|
def _datafp(self, mode=b'r'):
|
|
771
|
"""file object for the revlog's data file"""
|
|
773
|
"""file object for the revlog's data file"""
|
|
772
|
return self.opener(self._datafile, mode=mode)
|
|
774
|
return self.opener(self._datafile, mode=mode)
|
|
773
|
|
|
775
|
|
|
774
|
@contextlib.contextmanager
|
|
776
|
@contextlib.contextmanager
|
|
775
|
def _datareadfp(self, existingfp=None):
|
|
777
|
def _datareadfp(self, existingfp=None):
|
|
776
|
"""file object suitable to read data"""
|
|
778
|
"""file object suitable to read data"""
|
|
777
|
# Use explicit file handle, if given.
|
|
779
|
# Use explicit file handle, if given.
|
|
778
|
if existingfp is not None:
|
|
780
|
if existingfp is not None:
|
|
779
|
yield existingfp
|
|
781
|
yield existingfp
|
|
780
|
|
|
782
|
|
|
781
|
# Use a file handle being actively used for writes, if available.
|
|
783
|
# Use a file handle being actively used for writes, if available.
|
|
782
|
# There is some danger to doing this because reads will seek the
|
|
784
|
# There is some danger to doing this because reads will seek the
|
|
783
|
# file. However, _writeentry() performs a SEEK_END before all writes,
|
|
785
|
# file. However, _writeentry() performs a SEEK_END before all writes,
|
|
784
|
# so we should be safe.
|
|
786
|
# so we should be safe.
|
|
785
|
elif self._writinghandles:
|
|
787
|
elif self._writinghandles:
|
|
786
|
if self._inline:
|
|
788
|
if self._inline:
|
|
787
|
yield self._writinghandles[0]
|
|
789
|
yield self._writinghandles[0]
|
|
788
|
else:
|
|
790
|
else:
|
|
789
|
yield self._writinghandles[1]
|
|
791
|
yield self._writinghandles[1]
|
|
790
|
|
|
792
|
|
|
791
|
# Otherwise open a new file handle.
|
|
793
|
# Otherwise open a new file handle.
|
|
792
|
else:
|
|
794
|
else:
|
|
793
|
if self._inline:
|
|
795
|
if self._inline:
|
|
794
|
func = self._indexfp
|
|
796
|
func = self._indexfp
|
|
795
|
else:
|
|
797
|
else:
|
|
796
|
func = self._datafp
|
|
798
|
func = self._datafp
|
|
797
|
with func() as fp:
|
|
799
|
with func() as fp:
|
|
798
|
yield fp
|
|
800
|
yield fp
|
|
799
|
|
|
801
|
|
|
800
|
def tiprev(self):
|
|
802
|
def tiprev(self):
|
|
801
|
return len(self.index) - 1
|
|
803
|
return len(self.index) - 1
|
|
802
|
|
|
804
|
|
|
803
|
def tip(self):
|
|
805
|
def tip(self):
|
|
804
|
return self.node(self.tiprev())
|
|
806
|
return self.node(self.tiprev())
|
|
805
|
|
|
807
|
|
|
806
|
def __contains__(self, rev):
|
|
808
|
def __contains__(self, rev):
|
|
807
|
return 0 <= rev < len(self)
|
|
809
|
return 0 <= rev < len(self)
|
|
808
|
|
|
810
|
|
|
809
|
def __len__(self):
|
|
811
|
def __len__(self):
|
|
810
|
return len(self.index)
|
|
812
|
return len(self.index)
|
|
811
|
|
|
813
|
|
|
812
|
def __iter__(self):
|
|
814
|
def __iter__(self):
|
|
813
|
return iter(pycompat.xrange(len(self)))
|
|
815
|
return iter(pycompat.xrange(len(self)))
|
|
814
|
|
|
816
|
|
|
815
|
def revs(self, start=0, stop=None):
|
|
817
|
def revs(self, start=0, stop=None):
|
|
816
|
"""iterate over all rev in this revlog (from start to stop)"""
|
|
818
|
"""iterate over all rev in this revlog (from start to stop)"""
|
|
817
|
return storageutil.iterrevs(len(self), start=start, stop=stop)
|
|
819
|
return storageutil.iterrevs(len(self), start=start, stop=stop)
|
|
818
|
|
|
820
|
|
|
819
|
@property
|
|
821
|
@property
|
|
820
|
def nodemap(self):
|
|
822
|
def nodemap(self):
|
|
821
|
msg = (
|
|
823
|
msg = (
|
|
822
|
b"revlog.nodemap is deprecated, "
|
|
824
|
b"revlog.nodemap is deprecated, "
|
|
823
|
b"use revlog.index.[has_node|rev|get_rev]"
|
|
825
|
b"use revlog.index.[has_node|rev|get_rev]"
|
|
824
|
)
|
|
826
|
)
|
|
825
|
util.nouideprecwarn(msg, b'5.3', stacklevel=2)
|
|
827
|
util.nouideprecwarn(msg, b'5.3', stacklevel=2)
|
|
826
|
return self.index.nodemap
|
|
828
|
return self.index.nodemap
|
|
827
|
|
|
829
|
|
|
828
|
@property
|
|
830
|
@property
|
|
829
|
def _nodecache(self):
|
|
831
|
def _nodecache(self):
|
|
830
|
msg = b"revlog._nodecache is deprecated, use revlog.index.nodemap"
|
|
832
|
msg = b"revlog._nodecache is deprecated, use revlog.index.nodemap"
|
|
831
|
util.nouideprecwarn(msg, b'5.3', stacklevel=2)
|
|
833
|
util.nouideprecwarn(msg, b'5.3', stacklevel=2)
|
|
832
|
return self.index.nodemap
|
|
834
|
return self.index.nodemap
|
|
833
|
|
|
835
|
|
|
834
|
def hasnode(self, node):
|
|
836
|
def hasnode(self, node):
|
|
835
|
try:
|
|
837
|
try:
|
|
836
|
self.rev(node)
|
|
838
|
self.rev(node)
|
|
837
|
return True
|
|
839
|
return True
|
|
838
|
except KeyError:
|
|
840
|
except KeyError:
|
|
839
|
return False
|
|
841
|
return False
|
|
840
|
|
|
842
|
|
|
841
|
def candelta(self, baserev, rev):
|
|
843
|
def candelta(self, baserev, rev):
|
|
842
|
"""whether two revisions (baserev, rev) can be delta-ed or not"""
|
|
844
|
"""whether two revisions (baserev, rev) can be delta-ed or not"""
|
|
843
|
# Disable delta if either rev requires a content-changing flag
|
|
845
|
# Disable delta if either rev requires a content-changing flag
|
|
844
|
# processor (ex. LFS). This is because such flag processor can alter
|
|
846
|
# processor (ex. LFS). This is because such flag processor can alter
|
|
845
|
# the rawtext content that the delta will be based on, and two clients
|
|
847
|
# the rawtext content that the delta will be based on, and two clients
|
|
846
|
# could have a same revlog node with different flags (i.e. different
|
|
848
|
# could have a same revlog node with different flags (i.e. different
|
|
847
|
# rawtext contents) and the delta could be incompatible.
|
|
849
|
# rawtext contents) and the delta could be incompatible.
|
|
848
|
if (self.flags(baserev) & REVIDX_RAWTEXT_CHANGING_FLAGS) or (
|
|
850
|
if (self.flags(baserev) & REVIDX_RAWTEXT_CHANGING_FLAGS) or (
|
|
849
|
self.flags(rev) & REVIDX_RAWTEXT_CHANGING_FLAGS
|
|
851
|
self.flags(rev) & REVIDX_RAWTEXT_CHANGING_FLAGS
|
|
850
|
):
|
|
852
|
):
|
|
851
|
return False
|
|
853
|
return False
|
|
852
|
return True
|
|
854
|
return True
|
|
853
|
|
|
855
|
|
|
854
|
def update_caches(self, transaction):
|
|
856
|
def update_caches(self, transaction):
|
|
855
|
if self._nodemap_file is not None:
|
|
857
|
if self._nodemap_file is not None:
|
|
856
|
if transaction is None:
|
|
858
|
if transaction is None:
|
|
857
|
nodemaputil.update_persistent_nodemap(self)
|
|
859
|
nodemaputil.update_persistent_nodemap(self)
|
|
858
|
else:
|
|
860
|
else:
|
|
859
|
nodemaputil.setup_persistent_nodemap(transaction, self)
|
|
861
|
nodemaputil.setup_persistent_nodemap(transaction, self)
|
|
860
|
|
|
862
|
|
|
861
|
def clearcaches(self):
|
|
863
|
def clearcaches(self):
|
|
862
|
self._revisioncache = None
|
|
864
|
self._revisioncache = None
|
|
863
|
self._chainbasecache.clear()
|
|
865
|
self._chainbasecache.clear()
|
|
864
|
self._chunkcache = (0, b'')
|
|
866
|
self._chunkcache = (0, b'')
|
|
865
|
self._pcache = {}
|
|
867
|
self._pcache = {}
|
|
866
|
self._nodemap_docket = None
|
|
868
|
self._nodemap_docket = None
|
|
867
|
self.index.clearcaches()
|
|
869
|
self.index.clearcaches()
|
|
868
|
# The python code is the one responsible for validating the docket, we
|
|
870
|
# The python code is the one responsible for validating the docket, we
|
|
869
|
# end up having to refresh it here.
|
|
871
|
# end up having to refresh it here.
|
|
870
|
use_nodemap = (
|
|
872
|
use_nodemap = (
|
|
871
|
not self._inline
|
|
873
|
not self._inline
|
|
872
|
and self._nodemap_file is not None
|
|
874
|
and self._nodemap_file is not None
|
|
873
|
and util.safehasattr(self.index, 'update_nodemap_data')
|
|
875
|
and util.safehasattr(self.index, 'update_nodemap_data')
|
|
874
|
)
|
|
876
|
)
|
|
875
|
if use_nodemap:
|
|
877
|
if use_nodemap:
|
|
876
|
nodemap_data = nodemaputil.persisted_data(self)
|
|
878
|
nodemap_data = nodemaputil.persisted_data(self)
|
|
877
|
if nodemap_data is not None:
|
|
879
|
if nodemap_data is not None:
|
|
878
|
self._nodemap_docket = nodemap_data[0]
|
|
880
|
self._nodemap_docket = nodemap_data[0]
|
|
879
|
self.index.update_nodemap_data(*nodemap_data)
|
|
881
|
self.index.update_nodemap_data(*nodemap_data)
|
|
880
|
|
|
882
|
|
|
881
|
def rev(self, node):
|
|
883
|
def rev(self, node):
|
|
882
|
try:
|
|
884
|
try:
|
|
883
|
return self.index.rev(node)
|
|
885
|
return self.index.rev(node)
|
|
884
|
except TypeError:
|
|
886
|
except TypeError:
|
|
885
|
raise
|
|
887
|
raise
|
|
886
|
except error.RevlogError:
|
|
888
|
except error.RevlogError:
|
|
887
|
# parsers.c radix tree lookup failed
|
|
889
|
# parsers.c radix tree lookup failed
|
|
888
|
if (
|
|
890
|
if (
|
|
889
|
node == self.nodeconstants.wdirid
|
|
891
|
node == self.nodeconstants.wdirid
|
|
890
|
or node in self.nodeconstants.wdirfilenodeids
|
|
892
|
or node in self.nodeconstants.wdirfilenodeids
|
|
891
|
):
|
|
893
|
):
|
|
892
|
raise error.WdirUnsupported
|
|
894
|
raise error.WdirUnsupported
|
|
893
|
raise error.LookupError(node, self.display_id, _(b'no node'))
|
|
895
|
raise error.LookupError(node, self.display_id, _(b'no node'))
|
|
894
|
|
|
896
|
|
|
895
|
# Accessors for index entries.
|
|
897
|
# Accessors for index entries.
|
|
896
|
|
|
898
|
|
|
897
|
# First tuple entry is 8 bytes. First 6 bytes are offset. Last 2 bytes
|
|
899
|
# First tuple entry is 8 bytes. First 6 bytes are offset. Last 2 bytes
|
|
898
|
# are flags.
|
|
900
|
# are flags.
|
|
899
|
def start(self, rev):
|
|
901
|
def start(self, rev):
|
|
900
|
return int(self.index[rev][0] >> 16)
|
|
902
|
return int(self.index[rev][0] >> 16)
|
|
901
|
|
|
903
|
|
|
902
|
def flags(self, rev):
|
|
904
|
def flags(self, rev):
|
|
903
|
return self.index[rev][0] & 0xFFFF
|
|
905
|
return self.index[rev][0] & 0xFFFF
|
|
904
|
|
|
906
|
|
|
905
|
def length(self, rev):
|
|
907
|
def length(self, rev):
|
|
906
|
return self.index[rev][1]
|
|
908
|
return self.index[rev][1]
|
|
907
|
|
|
909
|
|
|
908
|
def sidedata_length(self, rev):
|
|
910
|
def sidedata_length(self, rev):
|
|
909
|
if not self.hassidedata:
|
|
911
|
if not self.hassidedata:
|
|
910
|
return 0
|
|
912
|
return 0
|
|
911
|
return self.index[rev][9]
|
|
913
|
return self.index[rev][9]
|
|
912
|
|
|
914
|
|
|
913
|
def rawsize(self, rev):
|
|
915
|
def rawsize(self, rev):
|
|
914
|
"""return the length of the uncompressed text for a given revision"""
|
|
916
|
"""return the length of the uncompressed text for a given revision"""
|
|
915
|
l = self.index[rev][2]
|
|
917
|
l = self.index[rev][2]
|
|
916
|
if l >= 0:
|
|
918
|
if l >= 0:
|
|
917
|
return l
|
|
919
|
return l
|
|
918
|
|
|
920
|
|
|
919
|
t = self.rawdata(rev)
|
|
921
|
t = self.rawdata(rev)
|
|
920
|
return len(t)
|
|
922
|
return len(t)
|
|
921
|
|
|
923
|
|
|
922
|
def size(self, rev):
|
|
924
|
def size(self, rev):
|
|
923
|
"""length of non-raw text (processed by a "read" flag processor)"""
|
|
925
|
"""length of non-raw text (processed by a "read" flag processor)"""
|
|
924
|
# fast path: if no "read" flag processor could change the content,
|
|
926
|
# fast path: if no "read" flag processor could change the content,
|
|
925
|
# size is rawsize. note: ELLIPSIS is known to not change the content.
|
|
927
|
# size is rawsize. note: ELLIPSIS is known to not change the content.
|
|
926
|
flags = self.flags(rev)
|
|
928
|
flags = self.flags(rev)
|
|
927
|
if flags & (flagutil.REVIDX_KNOWN_FLAGS ^ REVIDX_ELLIPSIS) == 0:
|
|
929
|
if flags & (flagutil.REVIDX_KNOWN_FLAGS ^ REVIDX_ELLIPSIS) == 0:
|
|
928
|
return self.rawsize(rev)
|
|
930
|
return self.rawsize(rev)
|
|
929
|
|
|
931
|
|
|
930
|
return len(self.revision(rev, raw=False))
|
|
932
|
return len(self.revision(rev, raw=False))
|
|
931
|
|
|
933
|
|
|
932
|
def chainbase(self, rev):
|
|
934
|
def chainbase(self, rev):
|
|
933
|
base = self._chainbasecache.get(rev)
|
|
935
|
base = self._chainbasecache.get(rev)
|
|
934
|
if base is not None:
|
|
936
|
if base is not None:
|
|
935
|
return base
|
|
937
|
return base
|
|
936
|
|
|
938
|
|
|
937
|
index = self.index
|
|
939
|
index = self.index
|
|
938
|
iterrev = rev
|
|
940
|
iterrev = rev
|
|
939
|
base = index[iterrev][3]
|
|
941
|
base = index[iterrev][3]
|
|
940
|
while base != iterrev:
|
|
942
|
while base != iterrev:
|
|
941
|
iterrev = base
|
|
943
|
iterrev = base
|
|
942
|
base = index[iterrev][3]
|
|
944
|
base = index[iterrev][3]
|
|
943
|
|
|
945
|
|
|
944
|
self._chainbasecache[rev] = base
|
|
946
|
self._chainbasecache[rev] = base
|
|
945
|
return base
|
|
947
|
return base
|
|
946
|
|
|
948
|
|
|
947
|
def linkrev(self, rev):
|
|
949
|
def linkrev(self, rev):
|
|
948
|
return self.index[rev][4]
|
|
950
|
return self.index[rev][4]
|
|
949
|
|
|
951
|
|
|
950
|
def parentrevs(self, rev):
|
|
952
|
def parentrevs(self, rev):
|
|
951
|
try:
|
|
953
|
try:
|
|
952
|
entry = self.index[rev]
|
|
954
|
entry = self.index[rev]
|
|
953
|
except IndexError:
|
|
955
|
except IndexError:
|
|
954
|
if rev == wdirrev:
|
|
956
|
if rev == wdirrev:
|
|
955
|
raise error.WdirUnsupported
|
|
957
|
raise error.WdirUnsupported
|
|
956
|
raise
|
|
958
|
raise
|
|
957
|
if entry[5] == nullrev:
|
|
959
|
if entry[5] == nullrev:
|
|
958
|
return entry[6], entry[5]
|
|
960
|
return entry[6], entry[5]
|
|
959
|
else:
|
|
961
|
else:
|
|
960
|
return entry[5], entry[6]
|
|
962
|
return entry[5], entry[6]
|
|
961
|
|
|
963
|
|
|
962
|
# fast parentrevs(rev) where rev isn't filtered
|
|
964
|
# fast parentrevs(rev) where rev isn't filtered
|
|
963
|
_uncheckedparentrevs = parentrevs
|
|
965
|
_uncheckedparentrevs = parentrevs
|
|
964
|
|
|
966
|
|
|
965
|
def node(self, rev):
|
|
967
|
def node(self, rev):
|
|
966
|
try:
|
|
968
|
try:
|
|
967
|
return self.index[rev][7]
|
|
969
|
return self.index[rev][7]
|
|
968
|
except IndexError:
|
|
970
|
except IndexError:
|
|
969
|
if rev == wdirrev:
|
|
971
|
if rev == wdirrev:
|
|
970
|
raise error.WdirUnsupported
|
|
972
|
raise error.WdirUnsupported
|
|
971
|
raise
|
|
973
|
raise
|
|
972
|
|
|
974
|
|
|
973
|
# Derived from index values.
|
|
975
|
# Derived from index values.
|
|
974
|
|
|
976
|
|
|
975
|
def end(self, rev):
|
|
977
|
def end(self, rev):
|
|
976
|
return self.start(rev) + self.length(rev)
|
|
978
|
return self.start(rev) + self.length(rev)
|
|
977
|
|
|
979
|
|
|
978
|
def parents(self, node):
|
|
980
|
def parents(self, node):
|
|
979
|
i = self.index
|
|
981
|
i = self.index
|
|
980
|
d = i[self.rev(node)]
|
|
982
|
d = i[self.rev(node)]
|
|
981
|
# inline node() to avoid function call overhead
|
|
983
|
# inline node() to avoid function call overhead
|
|
982
|
if d[5] == self.nullid:
|
|
984
|
if d[5] == self.nullid:
|
|
983
|
return i[d[6]][7], i[d[5]][7]
|
|
985
|
return i[d[6]][7], i[d[5]][7]
|
|
984
|
else:
|
|
986
|
else:
|
|
985
|
return i[d[5]][7], i[d[6]][7]
|
|
987
|
return i[d[5]][7], i[d[6]][7]
|
|
986
|
|
|
988
|
|
|
987
|
def chainlen(self, rev):
|
|
989
|
def chainlen(self, rev):
|
|
988
|
return self._chaininfo(rev)[0]
|
|
990
|
return self._chaininfo(rev)[0]
|
|
989
|
|
|
991
|
|
|
990
|
def _chaininfo(self, rev):
|
|
992
|
def _chaininfo(self, rev):
|
|
991
|
chaininfocache = self._chaininfocache
|
|
993
|
chaininfocache = self._chaininfocache
|
|
992
|
if rev in chaininfocache:
|
|
994
|
if rev in chaininfocache:
|
|
993
|
return chaininfocache[rev]
|
|
995
|
return chaininfocache[rev]
|
|
994
|
index = self.index
|
|
996
|
index = self.index
|
|
995
|
generaldelta = self._generaldelta
|
|
997
|
generaldelta = self._generaldelta
|
|
996
|
iterrev = rev
|
|
998
|
iterrev = rev
|
|
997
|
e = index[iterrev]
|
|
999
|
e = index[iterrev]
|
|
998
|
clen = 0
|
|
1000
|
clen = 0
|
|
999
|
compresseddeltalen = 0
|
|
1001
|
compresseddeltalen = 0
|
|
1000
|
while iterrev != e[3]:
|
|
1002
|
while iterrev != e[3]:
|
|
1001
|
clen += 1
|
|
1003
|
clen += 1
|
|
1002
|
compresseddeltalen += e[1]
|
|
1004
|
compresseddeltalen += e[1]
|
|
1003
|
if generaldelta:
|
|
1005
|
if generaldelta:
|
|
1004
|
iterrev = e[3]
|
|
1006
|
iterrev = e[3]
|
|
1005
|
else:
|
|
1007
|
else:
|
|
1006
|
iterrev -= 1
|
|
1008
|
iterrev -= 1
|
|
1007
|
if iterrev in chaininfocache:
|
|
1009
|
if iterrev in chaininfocache:
|
|
1008
|
t = chaininfocache[iterrev]
|
|
1010
|
t = chaininfocache[iterrev]
|
|
1009
|
clen += t[0]
|
|
1011
|
clen += t[0]
|
|
1010
|
compresseddeltalen += t[1]
|
|
1012
|
compresseddeltalen += t[1]
|
|
1011
|
break
|
|
1013
|
break
|
|
1012
|
e = index[iterrev]
|
|
1014
|
e = index[iterrev]
|
|
1013
|
else:
|
|
1015
|
else:
|
|
1014
|
# Add text length of base since decompressing that also takes
|
|
1016
|
# Add text length of base since decompressing that also takes
|
|
1015
|
# work. For cache hits the length is already included.
|
|
1017
|
# work. For cache hits the length is already included.
|
|
1016
|
compresseddeltalen += e[1]
|
|
1018
|
compresseddeltalen += e[1]
|
|
1017
|
r = (clen, compresseddeltalen)
|
|
1019
|
r = (clen, compresseddeltalen)
|
|
1018
|
chaininfocache[rev] = r
|
|
1020
|
chaininfocache[rev] = r
|
|
1019
|
return r
|
|
1021
|
return r
|
|
1020
|
|
|
1022
|
|
|
1021
|
def _deltachain(self, rev, stoprev=None):
|
|
1023
|
def _deltachain(self, rev, stoprev=None):
|
|
1022
|
"""Obtain the delta chain for a revision.
|
|
1024
|
"""Obtain the delta chain for a revision.
|
|
1023
|
|
|
1025
|
|
|
1024
|
``stoprev`` specifies a revision to stop at. If not specified, we
|
|
1026
|
``stoprev`` specifies a revision to stop at. If not specified, we
|
|
1025
|
stop at the base of the chain.
|
|
1027
|
stop at the base of the chain.
|
|
1026
|
|
|
1028
|
|
|
1027
|
Returns a 2-tuple of (chain, stopped) where ``chain`` is a list of
|
|
1029
|
Returns a 2-tuple of (chain, stopped) where ``chain`` is a list of
|
|
1028
|
revs in ascending order and ``stopped`` is a bool indicating whether
|
|
1030
|
revs in ascending order and ``stopped`` is a bool indicating whether
|
|
1029
|
``stoprev`` was hit.
|
|
1031
|
``stoprev`` was hit.
|
|
1030
|
"""
|
|
1032
|
"""
|
|
1031
|
# Try C implementation.
|
|
1033
|
# Try C implementation.
|
|
1032
|
try:
|
|
1034
|
try:
|
|
1033
|
return self.index.deltachain(rev, stoprev, self._generaldelta)
|
|
1035
|
return self.index.deltachain(rev, stoprev, self._generaldelta)
|
|
1034
|
except AttributeError:
|
|
1036
|
except AttributeError:
|
|
1035
|
pass
|
|
1037
|
pass
|
|
1036
|
|
|
1038
|
|
|
1037
|
chain = []
|
|
1039
|
chain = []
|
|
1038
|
|
|
1040
|
|
|
1039
|
# Alias to prevent attribute lookup in tight loop.
|
|
1041
|
# Alias to prevent attribute lookup in tight loop.
|
|
1040
|
index = self.index
|
|
1042
|
index = self.index
|
|
1041
|
generaldelta = self._generaldelta
|
|
1043
|
generaldelta = self._generaldelta
|
|
1042
|
|
|
1044
|
|
|
1043
|
iterrev = rev
|
|
1045
|
iterrev = rev
|
|
1044
|
e = index[iterrev]
|
|
1046
|
e = index[iterrev]
|
|
1045
|
while iterrev != e[3] and iterrev != stoprev:
|
|
1047
|
while iterrev != e[3] and iterrev != stoprev:
|
|
1046
|
chain.append(iterrev)
|
|
1048
|
chain.append(iterrev)
|
|
1047
|
if generaldelta:
|
|
1049
|
if generaldelta:
|
|
1048
|
iterrev = e[3]
|
|
1050
|
iterrev = e[3]
|
|
1049
|
else:
|
|
1051
|
else:
|
|
1050
|
iterrev -= 1
|
|
1052
|
iterrev -= 1
|
|
1051
|
e = index[iterrev]
|
|
1053
|
e = index[iterrev]
|
|
1052
|
|
|
1054
|
|
|
1053
|
if iterrev == stoprev:
|
|
1055
|
if iterrev == stoprev:
|
|
1054
|
stopped = True
|
|
1056
|
stopped = True
|
|
1055
|
else:
|
|
1057
|
else:
|
|
1056
|
chain.append(iterrev)
|
|
1058
|
chain.append(iterrev)
|
|
1057
|
stopped = False
|
|
1059
|
stopped = False
|
|
1058
|
|
|
1060
|
|
|
1059
|
chain.reverse()
|
|
1061
|
chain.reverse()
|
|
1060
|
return chain, stopped
|
|
1062
|
return chain, stopped
|
|
1061
|
|
|
1063
|
|
|
1062
|
def ancestors(self, revs, stoprev=0, inclusive=False):
|
|
1064
|
def ancestors(self, revs, stoprev=0, inclusive=False):
|
|
1063
|
"""Generate the ancestors of 'revs' in reverse revision order.
|
|
1065
|
"""Generate the ancestors of 'revs' in reverse revision order.
|
|
1064
|
Does not generate revs lower than stoprev.
|
|
1066
|
Does not generate revs lower than stoprev.
|
|
1065
|
|
|
1067
|
|
|
1066
|
See the documentation for ancestor.lazyancestors for more details."""
|
|
1068
|
See the documentation for ancestor.lazyancestors for more details."""
|
|
1067
|
|
|
1069
|
|
|
1068
|
# first, make sure start revisions aren't filtered
|
|
1070
|
# first, make sure start revisions aren't filtered
|
|
1069
|
revs = list(revs)
|
|
1071
|
revs = list(revs)
|
|
1070
|
checkrev = self.node
|
|
1072
|
checkrev = self.node
|
|
1071
|
for r in revs:
|
|
1073
|
for r in revs:
|
|
1072
|
checkrev(r)
|
|
1074
|
checkrev(r)
|
|
1073
|
# and we're sure ancestors aren't filtered as well
|
|
1075
|
# and we're sure ancestors aren't filtered as well
|
|
1074
|
|
|
1076
|
|
|
1075
|
if rustancestor is not None and self.index.rust_ext_compat:
|
|
1077
|
if rustancestor is not None and self.index.rust_ext_compat:
|
|
1076
|
lazyancestors = rustancestor.LazyAncestors
|
|
1078
|
lazyancestors = rustancestor.LazyAncestors
|
|
1077
|
arg = self.index
|
|
1079
|
arg = self.index
|
|
1078
|
else:
|
|
1080
|
else:
|
|
1079
|
lazyancestors = ancestor.lazyancestors
|
|
1081
|
lazyancestors = ancestor.lazyancestors
|
|
1080
|
arg = self._uncheckedparentrevs
|
|
1082
|
arg = self._uncheckedparentrevs
|
|
1081
|
return lazyancestors(arg, revs, stoprev=stoprev, inclusive=inclusive)
|
|
1083
|
return lazyancestors(arg, revs, stoprev=stoprev, inclusive=inclusive)
|
|
1082
|
|
|
1084
|
|
|
1083
|
def descendants(self, revs):
|
|
1085
|
def descendants(self, revs):
|
|
1084
|
return dagop.descendantrevs(revs, self.revs, self.parentrevs)
|
|
1086
|
return dagop.descendantrevs(revs, self.revs, self.parentrevs)
|
|
1085
|
|
|
1087
|
|
|
1086
|
def findcommonmissing(self, common=None, heads=None):
|
|
1088
|
def findcommonmissing(self, common=None, heads=None):
|
|
1087
|
"""Return a tuple of the ancestors of common and the ancestors of heads
|
|
1089
|
"""Return a tuple of the ancestors of common and the ancestors of heads
|
|
1088
|
that are not ancestors of common. In revset terminology, we return the
|
|
1090
|
that are not ancestors of common. In revset terminology, we return the
|
|
1089
|
tuple:
|
|
1091
|
tuple:
|
|
1090
|
|
|
1092
|
|
|
1091
|
::common, (::heads) - (::common)
|
|
1093
|
::common, (::heads) - (::common)
|
|
1092
|
|
|
1094
|
|
|
1093
|
The list is sorted by revision number, meaning it is
|
|
1095
|
The list is sorted by revision number, meaning it is
|
|
1094
|
topologically sorted.
|
|
1096
|
topologically sorted.
|
|
1095
|
|
|
1097
|
|
|
1096
|
'heads' and 'common' are both lists of node IDs. If heads is
|
|
1098
|
'heads' and 'common' are both lists of node IDs. If heads is
|
|
1097
|
not supplied, uses all of the revlog's heads. If common is not
|
|
1099
|
not supplied, uses all of the revlog's heads. If common is not
|
|
1098
|
supplied, uses nullid."""
|
|
1100
|
supplied, uses nullid."""
|
|
1099
|
if common is None:
|
|
1101
|
if common is None:
|
|
1100
|
common = [self.nullid]
|
|
1102
|
common = [self.nullid]
|
|
1101
|
if heads is None:
|
|
1103
|
if heads is None:
|
|
1102
|
heads = self.heads()
|
|
1104
|
heads = self.heads()
|
|
1103
|
|
|
1105
|
|
|
1104
|
common = [self.rev(n) for n in common]
|
|
1106
|
common = [self.rev(n) for n in common]
|
|
1105
|
heads = [self.rev(n) for n in heads]
|
|
1107
|
heads = [self.rev(n) for n in heads]
|
|
1106
|
|
|
1108
|
|
|
1107
|
# we want the ancestors, but inclusive
|
|
1109
|
# we want the ancestors, but inclusive
|
|
1108
|
class lazyset(object):
|
|
1110
|
class lazyset(object):
|
|
1109
|
def __init__(self, lazyvalues):
|
|
1111
|
def __init__(self, lazyvalues):
|
|
1110
|
self.addedvalues = set()
|
|
1112
|
self.addedvalues = set()
|
|
1111
|
self.lazyvalues = lazyvalues
|
|
1113
|
self.lazyvalues = lazyvalues
|
|
1112
|
|
|
1114
|
|
|
1113
|
def __contains__(self, value):
|
|
1115
|
def __contains__(self, value):
|
|
1114
|
return value in self.addedvalues or value in self.lazyvalues
|
|
1116
|
return value in self.addedvalues or value in self.lazyvalues
|
|
1115
|
|
|
1117
|
|
|
1116
|
def __iter__(self):
|
|
1118
|
def __iter__(self):
|
|
1117
|
added = self.addedvalues
|
|
1119
|
added = self.addedvalues
|
|
1118
|
for r in added:
|
|
1120
|
for r in added:
|
|
1119
|
yield r
|
|
1121
|
yield r
|
|
1120
|
for r in self.lazyvalues:
|
|
1122
|
for r in self.lazyvalues:
|
|
1121
|
if not r in added:
|
|
1123
|
if not r in added:
|
|
1122
|
yield r
|
|
1124
|
yield r
|
|
1123
|
|
|
1125
|
|
|
1124
|
def add(self, value):
|
|
1126
|
def add(self, value):
|
|
1125
|
self.addedvalues.add(value)
|
|
1127
|
self.addedvalues.add(value)
|
|
1126
|
|
|
1128
|
|
|
1127
|
def update(self, values):
|
|
1129
|
def update(self, values):
|
|
1128
|
self.addedvalues.update(values)
|
|
1130
|
self.addedvalues.update(values)
|
|
1129
|
|
|
1131
|
|
|
1130
|
has = lazyset(self.ancestors(common))
|
|
1132
|
has = lazyset(self.ancestors(common))
|
|
1131
|
has.add(nullrev)
|
|
1133
|
has.add(nullrev)
|
|
1132
|
has.update(common)
|
|
1134
|
has.update(common)
|
|
1133
|
|
|
1135
|
|
|
1134
|
# take all ancestors from heads that aren't in has
|
|
1136
|
# take all ancestors from heads that aren't in has
|
|
1135
|
missing = set()
|
|
1137
|
missing = set()
|
|
1136
|
visit = collections.deque(r for r in heads if r not in has)
|
|
1138
|
visit = collections.deque(r for r in heads if r not in has)
|
|
1137
|
while visit:
|
|
1139
|
while visit:
|
|
1138
|
r = visit.popleft()
|
|
1140
|
r = visit.popleft()
|
|
1139
|
if r in missing:
|
|
1141
|
if r in missing:
|
|
1140
|
continue
|
|
1142
|
continue
|
|
1141
|
else:
|
|
1143
|
else:
|
|
1142
|
missing.add(r)
|
|
1144
|
missing.add(r)
|
|
1143
|
for p in self.parentrevs(r):
|
|
1145
|
for p in self.parentrevs(r):
|
|
1144
|
if p not in has:
|
|
1146
|
if p not in has:
|
|
1145
|
visit.append(p)
|
|
1147
|
visit.append(p)
|
|
1146
|
missing = list(missing)
|
|
1148
|
missing = list(missing)
|
|
1147
|
missing.sort()
|
|
1149
|
missing.sort()
|
|
1148
|
return has, [self.node(miss) for miss in missing]
|
|
1150
|
return has, [self.node(miss) for miss in missing]
|
|
1149
|
|
|
1151
|
|
|
1150
|
def incrementalmissingrevs(self, common=None):
|
|
1152
|
def incrementalmissingrevs(self, common=None):
|
|
1151
|
"""Return an object that can be used to incrementally compute the
|
|
1153
|
"""Return an object that can be used to incrementally compute the
|
|
1152
|
revision numbers of the ancestors of arbitrary sets that are not
|
|
1154
|
revision numbers of the ancestors of arbitrary sets that are not
|
|
1153
|
ancestors of common. This is an ancestor.incrementalmissingancestors
|
|
1155
|
ancestors of common. This is an ancestor.incrementalmissingancestors
|
|
1154
|
object.
|
|
1156
|
object.
|
|
1155
|
|
|
1157
|
|
|
1156
|
'common' is a list of revision numbers. If common is not supplied, uses
|
|
1158
|
'common' is a list of revision numbers. If common is not supplied, uses
|
|
1157
|
nullrev.
|
|
1159
|
nullrev.
|
|
1158
|
"""
|
|
1160
|
"""
|
|
1159
|
if common is None:
|
|
1161
|
if common is None:
|
|
1160
|
common = [nullrev]
|
|
1162
|
common = [nullrev]
|
|
1161
|
|
|
1163
|
|
|
1162
|
if rustancestor is not None and self.index.rust_ext_compat:
|
|
1164
|
if rustancestor is not None and self.index.rust_ext_compat:
|
|
1163
|
return rustancestor.MissingAncestors(self.index, common)
|
|
1165
|
return rustancestor.MissingAncestors(self.index, common)
|
|
1164
|
return ancestor.incrementalmissingancestors(self.parentrevs, common)
|
|
1166
|
return ancestor.incrementalmissingancestors(self.parentrevs, common)
|
|
1165
|
|
|
1167
|
|
|
1166
|
def findmissingrevs(self, common=None, heads=None):
|
|
1168
|
def findmissingrevs(self, common=None, heads=None):
|
|
1167
|
"""Return the revision numbers of the ancestors of heads that
|
|
1169
|
"""Return the revision numbers of the ancestors of heads that
|
|
1168
|
are not ancestors of common.
|
|
1170
|
are not ancestors of common.
|
|
1169
|
|
|
1171
|
|
|
1170
|
More specifically, return a list of revision numbers corresponding to
|
|
1172
|
More specifically, return a list of revision numbers corresponding to
|
|
1171
|
nodes N such that every N satisfies the following constraints:
|
|
1173
|
nodes N such that every N satisfies the following constraints:
|
|
1172
|
|
|
1174
|
|
|
1173
|
1. N is an ancestor of some node in 'heads'
|
|
1175
|
1. N is an ancestor of some node in 'heads'
|
|
1174
|
2. N is not an ancestor of any node in 'common'
|
|
1176
|
2. N is not an ancestor of any node in 'common'
|
|
1175
|
|
|
1177
|
|
|
1176
|
The list is sorted by revision number, meaning it is
|
|
1178
|
The list is sorted by revision number, meaning it is
|
|
1177
|
topologically sorted.
|
|
1179
|
topologically sorted.
|
|
1178
|
|
|
1180
|
|
|
1179
|
'heads' and 'common' are both lists of revision numbers. If heads is
|
|
1181
|
'heads' and 'common' are both lists of revision numbers. If heads is
|
|
1180
|
not supplied, uses all of the revlog's heads. If common is not
|
|
1182
|
not supplied, uses all of the revlog's heads. If common is not
|
|
1181
|
supplied, uses nullid."""
|
|
1183
|
supplied, uses nullid."""
|
|
1182
|
if common is None:
|
|
1184
|
if common is None:
|
|
1183
|
common = [nullrev]
|
|
1185
|
common = [nullrev]
|
|
1184
|
if heads is None:
|
|
1186
|
if heads is None:
|
|
1185
|
heads = self.headrevs()
|
|
1187
|
heads = self.headrevs()
|
|
1186
|
|
|
1188
|
|
|
1187
|
inc = self.incrementalmissingrevs(common=common)
|
|
1189
|
inc = self.incrementalmissingrevs(common=common)
|
|
1188
|
return inc.missingancestors(heads)
|
|
1190
|
return inc.missingancestors(heads)
|
|
1189
|
|
|
1191
|
|
|
1190
|
def findmissing(self, common=None, heads=None):
|
|
1192
|
def findmissing(self, common=None, heads=None):
|
|
1191
|
"""Return the ancestors of heads that are not ancestors of common.
|
|
1193
|
"""Return the ancestors of heads that are not ancestors of common.
|
|
1192
|
|
|
1194
|
|
|
1193
|
More specifically, return a list of nodes N such that every N
|
|
1195
|
More specifically, return a list of nodes N such that every N
|
|
1194
|
satisfies the following constraints:
|
|
1196
|
satisfies the following constraints:
|
|
1195
|
|
|
1197
|
|
|
1196
|
1. N is an ancestor of some node in 'heads'
|
|
1198
|
1. N is an ancestor of some node in 'heads'
|
|
1197
|
2. N is not an ancestor of any node in 'common'
|
|
1199
|
2. N is not an ancestor of any node in 'common'
|
|
1198
|
|
|
1200
|
|
|
1199
|
The list is sorted by revision number, meaning it is
|
|
1201
|
The list is sorted by revision number, meaning it is
|
|
1200
|
topologically sorted.
|
|
1202
|
topologically sorted.
|
|
1201
|
|
|
1203
|
|
|
1202
|
'heads' and 'common' are both lists of node IDs. If heads is
|
|
1204
|
'heads' and 'common' are both lists of node IDs. If heads is
|
|
1203
|
not supplied, uses all of the revlog's heads. If common is not
|
|
1205
|
not supplied, uses all of the revlog's heads. If common is not
|
|
1204
|
supplied, uses nullid."""
|
|
1206
|
supplied, uses nullid."""
|
|
1205
|
if common is None:
|
|
1207
|
if common is None:
|
|
1206
|
common = [self.nullid]
|
|
1208
|
common = [self.nullid]
|
|
1207
|
if heads is None:
|
|
1209
|
if heads is None:
|
|
1208
|
heads = self.heads()
|
|
1210
|
heads = self.heads()
|
|
1209
|
|
|
1211
|
|
|
1210
|
common = [self.rev(n) for n in common]
|
|
1212
|
common = [self.rev(n) for n in common]
|
|
1211
|
heads = [self.rev(n) for n in heads]
|
|
1213
|
heads = [self.rev(n) for n in heads]
|
|
1212
|
|
|
1214
|
|
|
1213
|
inc = self.incrementalmissingrevs(common=common)
|
|
1215
|
inc = self.incrementalmissingrevs(common=common)
|
|
1214
|
return [self.node(r) for r in inc.missingancestors(heads)]
|
|
1216
|
return [self.node(r) for r in inc.missingancestors(heads)]
|
|
1215
|
|
|
1217
|
|
|
1216
|
def nodesbetween(self, roots=None, heads=None):
|
|
1218
|
def nodesbetween(self, roots=None, heads=None):
|
|
1217
|
"""Return a topological path from 'roots' to 'heads'.
|
|
1219
|
"""Return a topological path from 'roots' to 'heads'.
|
|
1218
|
|
|
1220
|
|
|
1219
|
Return a tuple (nodes, outroots, outheads) where 'nodes' is a
|
|
1221
|
Return a tuple (nodes, outroots, outheads) where 'nodes' is a
|
|
1220
|
topologically sorted list of all nodes N that satisfy both of
|
|
1222
|
topologically sorted list of all nodes N that satisfy both of
|
|
1221
|
these constraints:
|
|
1223
|
these constraints:
|
|
1222
|
|
|
1224
|
|
|
1223
|
1. N is a descendant of some node in 'roots'
|
|
1225
|
1. N is a descendant of some node in 'roots'
|
|
1224
|
2. N is an ancestor of some node in 'heads'
|
|
1226
|
2. N is an ancestor of some node in 'heads'
|
|
1225
|
|
|
1227
|
|
|
1226
|
Every node is considered to be both a descendant and an ancestor
|
|
1228
|
Every node is considered to be both a descendant and an ancestor
|
|
1227
|
of itself, so every reachable node in 'roots' and 'heads' will be
|
|
1229
|
of itself, so every reachable node in 'roots' and 'heads' will be
|
|
1228
|
included in 'nodes'.
|
|
1230
|
included in 'nodes'.
|
|
1229
|
|
|
1231
|
|
|
1230
|
'outroots' is the list of reachable nodes in 'roots', i.e., the
|
|
1232
|
'outroots' is the list of reachable nodes in 'roots', i.e., the
|
|
1231
|
subset of 'roots' that is returned in 'nodes'. Likewise,
|
|
1233
|
subset of 'roots' that is returned in 'nodes'. Likewise,
|
|
1232
|
'outheads' is the subset of 'heads' that is also in 'nodes'.
|
|
1234
|
'outheads' is the subset of 'heads' that is also in 'nodes'.
|
|
1233
|
|
|
1235
|
|
|
1234
|
'roots' and 'heads' are both lists of node IDs. If 'roots' is
|
|
1236
|
'roots' and 'heads' are both lists of node IDs. If 'roots' is
|
|
1235
|
unspecified, uses nullid as the only root. If 'heads' is
|
|
1237
|
unspecified, uses nullid as the only root. If 'heads' is
|
|
1236
|
unspecified, uses list of all of the revlog's heads."""
|
|
1238
|
unspecified, uses list of all of the revlog's heads."""
|
|
1237
|
nonodes = ([], [], [])
|
|
1239
|
nonodes = ([], [], [])
|
|
1238
|
if roots is not None:
|
|
1240
|
if roots is not None:
|
|
1239
|
roots = list(roots)
|
|
1241
|
roots = list(roots)
|
|
1240
|
if not roots:
|
|
1242
|
if not roots:
|
|
1241
|
return nonodes
|
|
1243
|
return nonodes
|
|
1242
|
lowestrev = min([self.rev(n) for n in roots])
|
|
1244
|
lowestrev = min([self.rev(n) for n in roots])
|
|
1243
|
else:
|
|
1245
|
else:
|
|
1244
|
roots = [self.nullid] # Everybody's a descendant of nullid
|
|
1246
|
roots = [self.nullid] # Everybody's a descendant of nullid
|
|
1245
|
lowestrev = nullrev
|
|
1247
|
lowestrev = nullrev
|
|
1246
|
if (lowestrev == nullrev) and (heads is None):
|
|
1248
|
if (lowestrev == nullrev) and (heads is None):
|
|
1247
|
# We want _all_ the nodes!
|
|
1249
|
# We want _all_ the nodes!
|
|
1248
|
return (
|
|
1250
|
return (
|
|
1249
|
[self.node(r) for r in self],
|
|
1251
|
[self.node(r) for r in self],
|
|
1250
|
[self.nullid],
|
|
1252
|
[self.nullid],
|
|
1251
|
list(self.heads()),
|
|
1253
|
list(self.heads()),
|
|
1252
|
)
|
|
1254
|
)
|
|
1253
|
if heads is None:
|
|
1255
|
if heads is None:
|
|
1254
|
# All nodes are ancestors, so the latest ancestor is the last
|
|
1256
|
# All nodes are ancestors, so the latest ancestor is the last
|
|
1255
|
# node.
|
|
1257
|
# node.
|
|
1256
|
highestrev = len(self) - 1
|
|
1258
|
highestrev = len(self) - 1
|
|
1257
|
# Set ancestors to None to signal that every node is an ancestor.
|
|
1259
|
# Set ancestors to None to signal that every node is an ancestor.
|
|
1258
|
ancestors = None
|
|
1260
|
ancestors = None
|
|
1259
|
# Set heads to an empty dictionary for later discovery of heads
|
|
1261
|
# Set heads to an empty dictionary for later discovery of heads
|
|
1260
|
heads = {}
|
|
1262
|
heads = {}
|
|
1261
|
else:
|
|
1263
|
else:
|
|
1262
|
heads = list(heads)
|
|
1264
|
heads = list(heads)
|
|
1263
|
if not heads:
|
|
1265
|
if not heads:
|
|
1264
|
return nonodes
|
|
1266
|
return nonodes
|
|
1265
|
ancestors = set()
|
|
1267
|
ancestors = set()
|
|
1266
|
# Turn heads into a dictionary so we can remove 'fake' heads.
|
|
1268
|
# Turn heads into a dictionary so we can remove 'fake' heads.
|
|
1267
|
# Also, later we will be using it to filter out the heads we can't
|
|
1269
|
# Also, later we will be using it to filter out the heads we can't
|
|
1268
|
# find from roots.
|
|
1270
|
# find from roots.
|
|
1269
|
heads = dict.fromkeys(heads, False)
|
|
1271
|
heads = dict.fromkeys(heads, False)
|
|
1270
|
# Start at the top and keep marking parents until we're done.
|
|
1272
|
# Start at the top and keep marking parents until we're done.
|
|
1271
|
nodestotag = set(heads)
|
|
1273
|
nodestotag = set(heads)
|
|
1272
|
# Remember where the top was so we can use it as a limit later.
|
|
1274
|
# Remember where the top was so we can use it as a limit later.
|
|
1273
|
highestrev = max([self.rev(n) for n in nodestotag])
|
|
1275
|
highestrev = max([self.rev(n) for n in nodestotag])
|
|
1274
|
while nodestotag:
|
|
1276
|
while nodestotag:
|
|
1275
|
# grab a node to tag
|
|
1277
|
# grab a node to tag
|
|
1276
|
n = nodestotag.pop()
|
|
1278
|
n = nodestotag.pop()
|
|
1277
|
# Never tag nullid
|
|
1279
|
# Never tag nullid
|
|
1278
|
if n == self.nullid:
|
|
1280
|
if n == self.nullid:
|
|
1279
|
continue
|
|
1281
|
continue
|
|
1280
|
# A node's revision number represents its place in a
|
|
1282
|
# A node's revision number represents its place in a
|
|
1281
|
# topologically sorted list of nodes.
|
|
1283
|
# topologically sorted list of nodes.
|
|
1282
|
r = self.rev(n)
|
|
1284
|
r = self.rev(n)
|
|
1283
|
if r >= lowestrev:
|
|
1285
|
if r >= lowestrev:
|
|
1284
|
if n not in ancestors:
|
|
1286
|
if n not in ancestors:
|
|
1285
|
# If we are possibly a descendant of one of the roots
|
|
1287
|
# If we are possibly a descendant of one of the roots
|
|
1286
|
# and we haven't already been marked as an ancestor
|
|
1288
|
# and we haven't already been marked as an ancestor
|
|
1287
|
ancestors.add(n) # Mark as ancestor
|
|
1289
|
ancestors.add(n) # Mark as ancestor
|
|
1288
|
# Add non-nullid parents to list of nodes to tag.
|
|
1290
|
# Add non-nullid parents to list of nodes to tag.
|
|
1289
|
nodestotag.update(
|
|
1291
|
nodestotag.update(
|
|
1290
|
[p for p in self.parents(n) if p != self.nullid]
|
|
1292
|
[p for p in self.parents(n) if p != self.nullid]
|
|
1291
|
)
|
|
1293
|
)
|
|
1292
|
elif n in heads: # We've seen it before, is it a fake head?
|
|
1294
|
elif n in heads: # We've seen it before, is it a fake head?
|
|
1293
|
# So it is, real heads should not be the ancestors of
|
|
1295
|
# So it is, real heads should not be the ancestors of
|
|
1294
|
# any other heads.
|
|
1296
|
# any other heads.
|
|
1295
|
heads.pop(n)
|
|
1297
|
heads.pop(n)
|
|
1296
|
if not ancestors:
|
|
1298
|
if not ancestors:
|
|
1297
|
return nonodes
|
|
1299
|
return nonodes
|
|
1298
|
# Now that we have our set of ancestors, we want to remove any
|
|
1300
|
# Now that we have our set of ancestors, we want to remove any
|
|
1299
|
# roots that are not ancestors.
|
|
1301
|
# roots that are not ancestors.
|
|
1300
|
|
|
1302
|
|
|
1301
|
# If one of the roots was nullid, everything is included anyway.
|
|
1303
|
# If one of the roots was nullid, everything is included anyway.
|
|
1302
|
if lowestrev > nullrev:
|
|
1304
|
if lowestrev > nullrev:
|
|
1303
|
# But, since we weren't, let's recompute the lowest rev to not
|
|
1305
|
# But, since we weren't, let's recompute the lowest rev to not
|
|
1304
|
# include roots that aren't ancestors.
|
|
1306
|
# include roots that aren't ancestors.
|
|
1305
|
|
|
1307
|
|
|
1306
|
# Filter out roots that aren't ancestors of heads
|
|
1308
|
# Filter out roots that aren't ancestors of heads
|
|
1307
|
roots = [root for root in roots if root in ancestors]
|
|
1309
|
roots = [root for root in roots if root in ancestors]
|
|
1308
|
# Recompute the lowest revision
|
|
1310
|
# Recompute the lowest revision
|
|
1309
|
if roots:
|
|
1311
|
if roots:
|
|
1310
|
lowestrev = min([self.rev(root) for root in roots])
|
|
1312
|
lowestrev = min([self.rev(root) for root in roots])
|
|
1311
|
else:
|
|
1313
|
else:
|
|
1312
|
# No more roots? Return empty list
|
|
1314
|
# No more roots? Return empty list
|
|
1313
|
return nonodes
|
|
1315
|
return nonodes
|
|
1314
|
else:
|
|
1316
|
else:
|
|
1315
|
# We are descending from nullid, and don't need to care about
|
|
1317
|
# We are descending from nullid, and don't need to care about
|
|
1316
|
# any other roots.
|
|
1318
|
# any other roots.
|
|
1317
|
lowestrev = nullrev
|
|
1319
|
lowestrev = nullrev
|
|
1318
|
roots = [self.nullid]
|
|
1320
|
roots = [self.nullid]
|
|
1319
|
# Transform our roots list into a set.
|
|
1321
|
# Transform our roots list into a set.
|
|
1320
|
descendants = set(roots)
|
|
1322
|
descendants = set(roots)
|
|
1321
|
# Also, keep the original roots so we can filter out roots that aren't
|
|
1323
|
# Also, keep the original roots so we can filter out roots that aren't
|
|
1322
|
# 'real' roots (i.e. are descended from other roots).
|
|
1324
|
# 'real' roots (i.e. are descended from other roots).
|
|
1323
|
roots = descendants.copy()
|
|
1325
|
roots = descendants.copy()
|
|
1324
|
# Our topologically sorted list of output nodes.
|
|
1326
|
# Our topologically sorted list of output nodes.
|
|
1325
|
orderedout = []
|
|
1327
|
orderedout = []
|
|
1326
|
# Don't start at nullid since we don't want nullid in our output list,
|
|
1328
|
# Don't start at nullid since we don't want nullid in our output list,
|
|
1327
|
# and if nullid shows up in descendants, empty parents will look like
|
|
1329
|
# and if nullid shows up in descendants, empty parents will look like
|
|
1328
|
# they're descendants.
|
|
1330
|
# they're descendants.
|
|
1329
|
for r in self.revs(start=max(lowestrev, 0), stop=highestrev + 1):
|
|
1331
|
for r in self.revs(start=max(lowestrev, 0), stop=highestrev + 1):
|
|
1330
|
n = self.node(r)
|
|
1332
|
n = self.node(r)
|
|
1331
|
isdescendant = False
|
|
1333
|
isdescendant = False
|
|
1332
|
if lowestrev == nullrev: # Everybody is a descendant of nullid
|
|
1334
|
if lowestrev == nullrev: # Everybody is a descendant of nullid
|
|
1333
|
isdescendant = True
|
|
1335
|
isdescendant = True
|
|
1334
|
elif n in descendants:
|
|
1336
|
elif n in descendants:
|
|
1335
|
# n is already a descendant
|
|
1337
|
# n is already a descendant
|
|
1336
|
isdescendant = True
|
|
1338
|
isdescendant = True
|
|
1337
|
# This check only needs to be done here because all the roots
|
|
1339
|
# This check only needs to be done here because all the roots
|
|
1338
|
# will start being marked is descendants before the loop.
|
|
1340
|
# will start being marked is descendants before the loop.
|
|
1339
|
if n in roots:
|
|
1341
|
if n in roots:
|
|
1340
|
# If n was a root, check if it's a 'real' root.
|
|
1342
|
# If n was a root, check if it's a 'real' root.
|
|
1341
|
p = tuple(self.parents(n))
|
|
1343
|
p = tuple(self.parents(n))
|
|
1342
|
# If any of its parents are descendants, it's not a root.
|
|
1344
|
# If any of its parents are descendants, it's not a root.
|
|
1343
|
if (p[0] in descendants) or (p[1] in descendants):
|
|
1345
|
if (p[0] in descendants) or (p[1] in descendants):
|
|
1344
|
roots.remove(n)
|
|
1346
|
roots.remove(n)
|
|
1345
|
else:
|
|
1347
|
else:
|
|
1346
|
p = tuple(self.parents(n))
|
|
1348
|
p = tuple(self.parents(n))
|
|
1347
|
# A node is a descendant if either of its parents are
|
|
1349
|
# A node is a descendant if either of its parents are
|
|
1348
|
# descendants. (We seeded the dependents list with the roots
|
|
1350
|
# descendants. (We seeded the dependents list with the roots
|
|
1349
|
# up there, remember?)
|
|
1351
|
# up there, remember?)
|
|
1350
|
if (p[0] in descendants) or (p[1] in descendants):
|
|
1352
|
if (p[0] in descendants) or (p[1] in descendants):
|
|
1351
|
descendants.add(n)
|
|
1353
|
descendants.add(n)
|
|
1352
|
isdescendant = True
|
|
1354
|
isdescendant = True
|
|
1353
|
if isdescendant and ((ancestors is None) or (n in ancestors)):
|
|
1355
|
if isdescendant and ((ancestors is None) or (n in ancestors)):
|
|
1354
|
# Only include nodes that are both descendants and ancestors.
|
|
1356
|
# Only include nodes that are both descendants and ancestors.
|
|
1355
|
orderedout.append(n)
|
|
1357
|
orderedout.append(n)
|
|
1356
|
if (ancestors is not None) and (n in heads):
|
|
1358
|
if (ancestors is not None) and (n in heads):
|
|
1357
|
# We're trying to figure out which heads are reachable
|
|
1359
|
# We're trying to figure out which heads are reachable
|
|
1358
|
# from roots.
|
|
1360
|
# from roots.
|
|
1359
|
# Mark this head as having been reached
|
|
1361
|
# Mark this head as having been reached
|
|
1360
|
heads[n] = True
|
|
1362
|
heads[n] = True
|
|
1361
|
elif ancestors is None:
|
|
1363
|
elif ancestors is None:
|
|
1362
|
# Otherwise, we're trying to discover the heads.
|
|
1364
|
# Otherwise, we're trying to discover the heads.
|
|
1363
|
# Assume this is a head because if it isn't, the next step
|
|
1365
|
# Assume this is a head because if it isn't, the next step
|
|
1364
|
# will eventually remove it.
|
|
1366
|
# will eventually remove it.
|
|
1365
|
heads[n] = True
|
|
1367
|
heads[n] = True
|
|
1366
|
# But, obviously its parents aren't.
|
|
1368
|
# But, obviously its parents aren't.
|
|
1367
|
for p in self.parents(n):
|
|
1369
|
for p in self.parents(n):
|
|
1368
|
heads.pop(p, None)
|
|
1370
|
heads.pop(p, None)
|
|
1369
|
heads = [head for head, flag in pycompat.iteritems(heads) if flag]
|
|
1371
|
heads = [head for head, flag in pycompat.iteritems(heads) if flag]
|
|
1370
|
roots = list(roots)
|
|
1372
|
roots = list(roots)
|
|
1371
|
assert orderedout
|
|
1373
|
assert orderedout
|
|
1372
|
assert roots
|
|
1374
|
assert roots
|
|
1373
|
assert heads
|
|
1375
|
assert heads
|
|
1374
|
return (orderedout, roots, heads)
|
|
1376
|
return (orderedout, roots, heads)
|
|
1375
|
|
|
1377
|
|
|
1376
|
def headrevs(self, revs=None):
|
|
1378
|
def headrevs(self, revs=None):
|
|
1377
|
if revs is None:
|
|
1379
|
if revs is None:
|
|
1378
|
try:
|
|
1380
|
try:
|
|
1379
|
return self.index.headrevs()
|
|
1381
|
return self.index.headrevs()
|
|
1380
|
except AttributeError:
|
|
1382
|
except AttributeError:
|
|
1381
|
return self._headrevs()
|
|
1383
|
return self._headrevs()
|
|
1382
|
if rustdagop is not None and self.index.rust_ext_compat:
|
|
1384
|
if rustdagop is not None and self.index.rust_ext_compat:
|
|
1383
|
return rustdagop.headrevs(self.index, revs)
|
|
1385
|
return rustdagop.headrevs(self.index, revs)
|
|
1384
|
return dagop.headrevs(revs, self._uncheckedparentrevs)
|
|
1386
|
return dagop.headrevs(revs, self._uncheckedparentrevs)
|
|
1385
|
|
|
1387
|
|
|
1386
|
def computephases(self, roots):
|
|
1388
|
def computephases(self, roots):
|
|
1387
|
return self.index.computephasesmapsets(roots)
|
|
1389
|
return self.index.computephasesmapsets(roots)
|
|
1388
|
|
|
1390
|
|
|
1389
|
def _headrevs(self):
|
|
1391
|
def _headrevs(self):
|
|
1390
|
count = len(self)
|
|
1392
|
count = len(self)
|
|
1391
|
if not count:
|
|
1393
|
if not count:
|
|
1392
|
return [nullrev]
|
|
1394
|
return [nullrev]
|
|
1393
|
# we won't iter over filtered rev so nobody is a head at start
|
|
1395
|
# we won't iter over filtered rev so nobody is a head at start
|
|
1394
|
ishead = [0] * (count + 1)
|
|
1396
|
ishead = [0] * (count + 1)
|
|
1395
|
index = self.index
|
|
1397
|
index = self.index
|
|
1396
|
for r in self:
|
|
1398
|
for r in self:
|
|
1397
|
ishead[r] = 1 # I may be an head
|
|
1399
|
ishead[r] = 1 # I may be an head
|
|
1398
|
e = index[r]
|
|
1400
|
e = index[r]
|
|
1399
|
ishead[e[5]] = ishead[e[6]] = 0 # my parent are not
|
|
1401
|
ishead[e[5]] = ishead[e[6]] = 0 # my parent are not
|
|
1400
|
return [r for r, val in enumerate(ishead) if val]
|
|
1402
|
return [r for r, val in enumerate(ishead) if val]
|
|
1401
|
|
|
1403
|
|
|
1402
|
def heads(self, start=None, stop=None):
|
|
1404
|
def heads(self, start=None, stop=None):
|
|
1403
|
"""return the list of all nodes that have no children
|
|
1405
|
"""return the list of all nodes that have no children
|
|
1404
|
|
|
1406
|
|
|
1405
|
if start is specified, only heads that are descendants of
|
|
1407
|
if start is specified, only heads that are descendants of
|
|
1406
|
start will be returned
|
|
1408
|
start will be returned
|
|
1407
|
if stop is specified, it will consider all the revs from stop
|
|
1409
|
if stop is specified, it will consider all the revs from stop
|
|
1408
|
as if they had no children
|
|
1410
|
as if they had no children
|
|
1409
|
"""
|
|
1411
|
"""
|
|
1410
|
if start is None and stop is None:
|
|
1412
|
if start is None and stop is None:
|
|
1411
|
if not len(self):
|
|
1413
|
if not len(self):
|
|
1412
|
return [self.nullid]
|
|
1414
|
return [self.nullid]
|
|
1413
|
return [self.node(r) for r in self.headrevs()]
|
|
1415
|
return [self.node(r) for r in self.headrevs()]
|
|
1414
|
|
|
1416
|
|
|
1415
|
if start is None:
|
|
1417
|
if start is None:
|
|
1416
|
start = nullrev
|
|
1418
|
start = nullrev
|
|
1417
|
else:
|
|
1419
|
else:
|
|
1418
|
start = self.rev(start)
|
|
1420
|
start = self.rev(start)
|
|
1419
|
|
|
1421
|
|
|
1420
|
stoprevs = {self.rev(n) for n in stop or []}
|
|
1422
|
stoprevs = {self.rev(n) for n in stop or []}
|
|
1421
|
|
|
1423
|
|
|
1422
|
revs = dagop.headrevssubset(
|
|
1424
|
revs = dagop.headrevssubset(
|
|
1423
|
self.revs, self.parentrevs, startrev=start, stoprevs=stoprevs
|
|
1425
|
self.revs, self.parentrevs, startrev=start, stoprevs=stoprevs
|
|
1424
|
)
|
|
1426
|
)
|
|
1425
|
|
|
1427
|
|
|
1426
|
return [self.node(rev) for rev in revs]
|
|
1428
|
return [self.node(rev) for rev in revs]
|
|
1427
|
|
|
1429
|
|
|
1428
|
def children(self, node):
|
|
1430
|
def children(self, node):
|
|
1429
|
"""find the children of a given node"""
|
|
1431
|
"""find the children of a given node"""
|
|
1430
|
c = []
|
|
1432
|
c = []
|
|
1431
|
p = self.rev(node)
|
|
1433
|
p = self.rev(node)
|
|
1432
|
for r in self.revs(start=p + 1):
|
|
1434
|
for r in self.revs(start=p + 1):
|
|
1433
|
prevs = [pr for pr in self.parentrevs(r) if pr != nullrev]
|
|
1435
|
prevs = [pr for pr in self.parentrevs(r) if pr != nullrev]
|
|
1434
|
if prevs:
|
|
1436
|
if prevs:
|
|
1435
|
for pr in prevs:
|
|
1437
|
for pr in prevs:
|
|
1436
|
if pr == p:
|
|
1438
|
if pr == p:
|
|
1437
|
c.append(self.node(r))
|
|
1439
|
c.append(self.node(r))
|
|
1438
|
elif p == nullrev:
|
|
1440
|
elif p == nullrev:
|
|
1439
|
c.append(self.node(r))
|
|
1441
|
c.append(self.node(r))
|
|
1440
|
return c
|
|
1442
|
return c
|
|
1441
|
|
|
1443
|
|
|
1442
|
def commonancestorsheads(self, a, b):
|
|
1444
|
def commonancestorsheads(self, a, b):
|
|
1443
|
"""calculate all the heads of the common ancestors of nodes a and b"""
|
|
1445
|
"""calculate all the heads of the common ancestors of nodes a and b"""
|
|
1444
|
a, b = self.rev(a), self.rev(b)
|
|
1446
|
a, b = self.rev(a), self.rev(b)
|
|
1445
|
ancs = self._commonancestorsheads(a, b)
|
|
1447
|
ancs = self._commonancestorsheads(a, b)
|
|
1446
|
return pycompat.maplist(self.node, ancs)
|
|
1448
|
return pycompat.maplist(self.node, ancs)
|
|
1447
|
|
|
1449
|
|
|
1448
|
def _commonancestorsheads(self, *revs):
|
|
1450
|
def _commonancestorsheads(self, *revs):
|
|
1449
|
"""calculate all the heads of the common ancestors of revs"""
|
|
1451
|
"""calculate all the heads of the common ancestors of revs"""
|
|
1450
|
try:
|
|
1452
|
try:
|
|
1451
|
ancs = self.index.commonancestorsheads(*revs)
|
|
1453
|
ancs = self.index.commonancestorsheads(*revs)
|
|
1452
|
except (AttributeError, OverflowError): # C implementation failed
|
|
1454
|
except (AttributeError, OverflowError): # C implementation failed
|
|
1453
|
ancs = ancestor.commonancestorsheads(self.parentrevs, *revs)
|
|
1455
|
ancs = ancestor.commonancestorsheads(self.parentrevs, *revs)
|
|
1454
|
return ancs
|
|
1456
|
return ancs
|
|
1455
|
|
|
1457
|
|
|
1456
|
def isancestor(self, a, b):
|
|
1458
|
def isancestor(self, a, b):
|
|
1457
|
"""return True if node a is an ancestor of node b
|
|
1459
|
"""return True if node a is an ancestor of node b
|
|
1458
|
|
|
1460
|
|
|
1459
|
A revision is considered an ancestor of itself."""
|
|
1461
|
A revision is considered an ancestor of itself."""
|
|
1460
|
a, b = self.rev(a), self.rev(b)
|
|
1462
|
a, b = self.rev(a), self.rev(b)
|
|
1461
|
return self.isancestorrev(a, b)
|
|
1463
|
return self.isancestorrev(a, b)
|
|
1462
|
|
|
1464
|
|
|
1463
|
def isancestorrev(self, a, b):
|
|
1465
|
def isancestorrev(self, a, b):
|
|
1464
|
"""return True if revision a is an ancestor of revision b
|
|
1466
|
"""return True if revision a is an ancestor of revision b
|
|
1465
|
|
|
1467
|
|
|
1466
|
A revision is considered an ancestor of itself.
|
|
1468
|
A revision is considered an ancestor of itself.
|
|
1467
|
|
|
1469
|
|
|
1468
|
The implementation of this is trivial but the use of
|
|
1470
|
The implementation of this is trivial but the use of
|
|
1469
|
reachableroots is not."""
|
|
1471
|
reachableroots is not."""
|
|
1470
|
if a == nullrev:
|
|
1472
|
if a == nullrev:
|
|
1471
|
return True
|
|
1473
|
return True
|
|
1472
|
elif a == b:
|
|
1474
|
elif a == b:
|
|
1473
|
return True
|
|
1475
|
return True
|
|
1474
|
elif a > b:
|
|
1476
|
elif a > b:
|
|
1475
|
return False
|
|
1477
|
return False
|
|
1476
|
return bool(self.reachableroots(a, [b], [a], includepath=False))
|
|
1478
|
return bool(self.reachableroots(a, [b], [a], includepath=False))
|
|
1477
|
|
|
1479
|
|
|
1478
|
def reachableroots(self, minroot, heads, roots, includepath=False):
|
|
1480
|
def reachableroots(self, minroot, heads, roots, includepath=False):
|
|
1479
|
"""return (heads(::(<roots> and <roots>::<heads>)))
|
|
1481
|
"""return (heads(::(<roots> and <roots>::<heads>)))
|
|
1480
|
|
|
1482
|
|
|
1481
|
If includepath is True, return (<roots>::<heads>)."""
|
|
1483
|
If includepath is True, return (<roots>::<heads>)."""
|
|
1482
|
try:
|
|
1484
|
try:
|
|
1483
|
return self.index.reachableroots2(
|
|
1485
|
return self.index.reachableroots2(
|
|
1484
|
minroot, heads, roots, includepath
|
|
1486
|
minroot, heads, roots, includepath
|
|
1485
|
)
|
|
1487
|
)
|
|
1486
|
except AttributeError:
|
|
1488
|
except AttributeError:
|
|
1487
|
return dagop._reachablerootspure(
|
|
1489
|
return dagop._reachablerootspure(
|
|
1488
|
self.parentrevs, minroot, roots, heads, includepath
|
|
1490
|
self.parentrevs, minroot, roots, heads, includepath
|
|
1489
|
)
|
|
1491
|
)
|
|
1490
|
|
|
1492
|
|
|
1491
|
def ancestor(self, a, b):
|
|
1493
|
def ancestor(self, a, b):
|
|
1492
|
"""calculate the "best" common ancestor of nodes a and b"""
|
|
1494
|
"""calculate the "best" common ancestor of nodes a and b"""
|
|
1493
|
|
|
1495
|
|
|
1494
|
a, b = self.rev(a), self.rev(b)
|
|
1496
|
a, b = self.rev(a), self.rev(b)
|
|
1495
|
try:
|
|
1497
|
try:
|
|
1496
|
ancs = self.index.ancestors(a, b)
|
|
1498
|
ancs = self.index.ancestors(a, b)
|
|
1497
|
except (AttributeError, OverflowError):
|
|
1499
|
except (AttributeError, OverflowError):
|
|
1498
|
ancs = ancestor.ancestors(self.parentrevs, a, b)
|
|
1500
|
ancs = ancestor.ancestors(self.parentrevs, a, b)
|
|
1499
|
if ancs:
|
|
1501
|
if ancs:
|
|
1500
|
# choose a consistent winner when there's a tie
|
|
1502
|
# choose a consistent winner when there's a tie
|
|
1501
|
return min(map(self.node, ancs))
|
|
1503
|
return min(map(self.node, ancs))
|
|
1502
|
return self.nullid
|
|
1504
|
return self.nullid
|
|
1503
|
|
|
1505
|
|
|
1504
|
def _match(self, id):
|
|
1506
|
def _match(self, id):
|
|
1505
|
if isinstance(id, int):
|
|
1507
|
if isinstance(id, int):
|
|
1506
|
# rev
|
|
1508
|
# rev
|
|
1507
|
return self.node(id)
|
|
1509
|
return self.node(id)
|
|
1508
|
if len(id) == self.nodeconstants.nodelen:
|
|
1510
|
if len(id) == self.nodeconstants.nodelen:
|
|
1509
|
# possibly a binary node
|
|
1511
|
# possibly a binary node
|
|
1510
|
# odds of a binary node being all hex in ASCII are 1 in 10**25
|
|
1512
|
# odds of a binary node being all hex in ASCII are 1 in 10**25
|
|
1511
|
try:
|
|
1513
|
try:
|
|
1512
|
node = id
|
|
1514
|
node = id
|
|
1513
|
self.rev(node) # quick search the index
|
|
1515
|
self.rev(node) # quick search the index
|
|
1514
|
return node
|
|
1516
|
return node
|
|
1515
|
except error.LookupError:
|
|
1517
|
except error.LookupError:
|
|
1516
|
pass # may be partial hex id
|
|
1518
|
pass # may be partial hex id
|
|
1517
|
try:
|
|
1519
|
try:
|
|
1518
|
# str(rev)
|
|
1520
|
# str(rev)
|
|
1519
|
rev = int(id)
|
|
1521
|
rev = int(id)
|
|
1520
|
if b"%d" % rev != id:
|
|
1522
|
if b"%d" % rev != id:
|
|
1521
|
raise ValueError
|
|
1523
|
raise ValueError
|
|
1522
|
if rev < 0:
|
|
1524
|
if rev < 0:
|
|
1523
|
rev = len(self) + rev
|
|
1525
|
rev = len(self) + rev
|
|
1524
|
if rev < 0 or rev >= len(self):
|
|
1526
|
if rev < 0 or rev >= len(self):
|
|
1525
|
raise ValueError
|
|
1527
|
raise ValueError
|
|
1526
|
return self.node(rev)
|
|
1528
|
return self.node(rev)
|
|
1527
|
except (ValueError, OverflowError):
|
|
1529
|
except (ValueError, OverflowError):
|
|
1528
|
pass
|
|
1530
|
pass
|
|
1529
|
if len(id) == 2 * self.nodeconstants.nodelen:
|
|
1531
|
if len(id) == 2 * self.nodeconstants.nodelen:
|
|
1530
|
try:
|
|
1532
|
try:
|
|
1531
|
# a full hex nodeid?
|
|
1533
|
# a full hex nodeid?
|
|
1532
|
node = bin(id)
|
|
1534
|
node = bin(id)
|
|
1533
|
self.rev(node)
|
|
1535
|
self.rev(node)
|
|
1534
|
return node
|
|
1536
|
return node
|
|
1535
|
except (TypeError, error.LookupError):
|
|
1537
|
except (TypeError, error.LookupError):
|
|
1536
|
pass
|
|
1538
|
pass
|
|
1537
|
|
|
1539
|
|
|
1538
|
def _partialmatch(self, id):
|
|
1540
|
def _partialmatch(self, id):
|
|
1539
|
# we don't care wdirfilenodeids as they should be always full hash
|
|
1541
|
# we don't care wdirfilenodeids as they should be always full hash
|
|
1540
|
maybewdir = self.nodeconstants.wdirhex.startswith(id)
|
|
1542
|
maybewdir = self.nodeconstants.wdirhex.startswith(id)
|
|
1541
|
ambiguous = False
|
|
1543
|
ambiguous = False
|
|
1542
|
try:
|
|
1544
|
try:
|
|
1543
|
partial = self.index.partialmatch(id)
|
|
1545
|
partial = self.index.partialmatch(id)
|
|
1544
|
if partial and self.hasnode(partial):
|
|
1546
|
if partial and self.hasnode(partial):
|
|
1545
|
if maybewdir:
|
|
1547
|
if maybewdir:
|
|
1546
|
# single 'ff...' match in radix tree, ambiguous with wdir
|
|
1548
|
# single 'ff...' match in radix tree, ambiguous with wdir
|
|
1547
|
ambiguous = True
|
|
1549
|
ambiguous = True
|
|
1548
|
else:
|
|
1550
|
else:
|
|
1549
|
return partial
|
|
1551
|
return partial
|
|
1550
|
elif maybewdir:
|
|
1552
|
elif maybewdir:
|
|
1551
|
# no 'ff...' match in radix tree, wdir identified
|
|
1553
|
# no 'ff...' match in radix tree, wdir identified
|
|
1552
|
raise error.WdirUnsupported
|
|
1554
|
raise error.WdirUnsupported
|
|
1553
|
else:
|
|
1555
|
else:
|
|
1554
|
return None
|
|
1556
|
return None
|
|
1555
|
except error.RevlogError:
|
|
1557
|
except error.RevlogError:
|
|
1556
|
# parsers.c radix tree lookup gave multiple matches
|
|
1558
|
# parsers.c radix tree lookup gave multiple matches
|
|
1557
|
# fast path: for unfiltered changelog, radix tree is accurate
|
|
1559
|
# fast path: for unfiltered changelog, radix tree is accurate
|
|
1558
|
if not getattr(self, 'filteredrevs', None):
|
|
1560
|
if not getattr(self, 'filteredrevs', None):
|
|
1559
|
ambiguous = True
|
|
1561
|
ambiguous = True
|
|
1560
|
# fall through to slow path that filters hidden revisions
|
|
1562
|
# fall through to slow path that filters hidden revisions
|
|
1561
|
except (AttributeError, ValueError):
|
|
1563
|
except (AttributeError, ValueError):
|
|
1562
|
# we are pure python, or key was too short to search radix tree
|
|
1564
|
# we are pure python, or key was too short to search radix tree
|
|
1563
|
pass
|
|
1565
|
pass
|
|
1564
|
if ambiguous:
|
|
1566
|
if ambiguous:
|
|
1565
|
raise error.AmbiguousPrefixLookupError(
|
|
1567
|
raise error.AmbiguousPrefixLookupError(
|
|
1566
|
id, self.display_id, _(b'ambiguous identifier')
|
|
1568
|
id, self.display_id, _(b'ambiguous identifier')
|
|
1567
|
)
|
|
1569
|
)
|
|
1568
|
|
|
1570
|
|
|
1569
|
if id in self._pcache:
|
|
1571
|
if id in self._pcache:
|
|
1570
|
return self._pcache[id]
|
|
1572
|
return self._pcache[id]
|
|
1571
|
|
|
1573
|
|
|
1572
|
if len(id) <= 40:
|
|
1574
|
if len(id) <= 40:
|
|
1573
|
try:
|
|
1575
|
try:
|
|
1574
|
# hex(node)[:...]
|
|
1576
|
# hex(node)[:...]
|
|
1575
|
l = len(id) // 2 # grab an even number of digits
|
|
1577
|
l = len(id) // 2 # grab an even number of digits
|
|
1576
|
prefix = bin(id[: l * 2])
|
|
1578
|
prefix = bin(id[: l * 2])
|
|
1577
|
nl = [e[7] for e in self.index if e[7].startswith(prefix)]
|
|
1579
|
nl = [e[7] for e in self.index if e[7].startswith(prefix)]
|
|
1578
|
nl = [
|
|
1580
|
nl = [
|
|
1579
|
n for n in nl if hex(n).startswith(id) and self.hasnode(n)
|
|
1581
|
n for n in nl if hex(n).startswith(id) and self.hasnode(n)
|
|
1580
|
]
|
|
1582
|
]
|
|
1581
|
if self.nodeconstants.nullhex.startswith(id):
|
|
1583
|
if self.nodeconstants.nullhex.startswith(id):
|
|
1582
|
nl.append(self.nullid)
|
|
1584
|
nl.append(self.nullid)
|
|
1583
|
if len(nl) > 0:
|
|
1585
|
if len(nl) > 0:
|
|
1584
|
if len(nl) == 1 and not maybewdir:
|
|
1586
|
if len(nl) == 1 and not maybewdir:
|
|
1585
|
self._pcache[id] = nl[0]
|
|
1587
|
self._pcache[id] = nl[0]
|
|
1586
|
return nl[0]
|
|
1588
|
return nl[0]
|
|
1587
|
raise error.AmbiguousPrefixLookupError(
|
|
1589
|
raise error.AmbiguousPrefixLookupError(
|
|
1588
|
id, self.display_id, _(b'ambiguous identifier')
|
|
1590
|
id, self.display_id, _(b'ambiguous identifier')
|
|
1589
|
)
|
|
1591
|
)
|
|
1590
|
if maybewdir:
|
|
1592
|
if maybewdir:
|
|
1591
|
raise error.WdirUnsupported
|
|
1593
|
raise error.WdirUnsupported
|
|
1592
|
return None
|
|
1594
|
return None
|
|
1593
|
except TypeError:
|
|
1595
|
except TypeError:
|
|
1594
|
pass
|
|
1596
|
pass
|
|
1595
|
|
|
1597
|
|
|
1596
|
def lookup(self, id):
|
|
1598
|
def lookup(self, id):
|
|
1597
|
"""locate a node based on:
|
|
1599
|
"""locate a node based on:
|
|
1598
|
- revision number or str(revision number)
|
|
1600
|
- revision number or str(revision number)
|
|
1599
|
- nodeid or subset of hex nodeid
|
|
1601
|
- nodeid or subset of hex nodeid
|
|
1600
|
"""
|
|
1602
|
"""
|
|
1601
|
n = self._match(id)
|
|
1603
|
n = self._match(id)
|
|
1602
|
if n is not None:
|
|
1604
|
if n is not None:
|
|
1603
|
return n
|
|
1605
|
return n
|
|
1604
|
n = self._partialmatch(id)
|
|
1606
|
n = self._partialmatch(id)
|
|
1605
|
if n:
|
|
1607
|
if n:
|
|
1606
|
return n
|
|
1608
|
return n
|
|
1607
|
|
|
1609
|
|
|
1608
|
raise error.LookupError(id, self.display_id, _(b'no match found'))
|
|
1610
|
raise error.LookupError(id, self.display_id, _(b'no match found'))
|
|
1609
|
|
|
1611
|
|
|
1610
|
def shortest(self, node, minlength=1):
|
|
1612
|
def shortest(self, node, minlength=1):
|
|
1611
|
"""Find the shortest unambiguous prefix that matches node."""
|
|
1613
|
"""Find the shortest unambiguous prefix that matches node."""
|
|
1612
|
|
|
1614
|
|
|
1613
|
def isvalid(prefix):
|
|
1615
|
def isvalid(prefix):
|
|
1614
|
try:
|
|
1616
|
try:
|
|
1615
|
matchednode = self._partialmatch(prefix)
|
|
1617
|
matchednode = self._partialmatch(prefix)
|
|
1616
|
except error.AmbiguousPrefixLookupError:
|
|
1618
|
except error.AmbiguousPrefixLookupError:
|
|
1617
|
return False
|
|
1619
|
return False
|
|
1618
|
except error.WdirUnsupported:
|
|
1620
|
except error.WdirUnsupported:
|
|
1619
|
# single 'ff...' match
|
|
1621
|
# single 'ff...' match
|
|
1620
|
return True
|
|
1622
|
return True
|
|
1621
|
if matchednode is None:
|
|
1623
|
if matchednode is None:
|
|
1622
|
raise error.LookupError(node, self.display_id, _(b'no node'))
|
|
1624
|
raise error.LookupError(node, self.display_id, _(b'no node'))
|
|
1623
|
return True
|
|
1625
|
return True
|
|
1624
|
|
|
1626
|
|
|
1625
|
def maybewdir(prefix):
|
|
1627
|
def maybewdir(prefix):
|
|
1626
|
return all(c == b'f' for c in pycompat.iterbytestr(prefix))
|
|
1628
|
return all(c == b'f' for c in pycompat.iterbytestr(prefix))
|
|
1627
|
|
|
1629
|
|
|
1628
|
hexnode = hex(node)
|
|
1630
|
hexnode = hex(node)
|
|
1629
|
|
|
1631
|
|
|
1630
|
def disambiguate(hexnode, minlength):
|
|
1632
|
def disambiguate(hexnode, minlength):
|
|
1631
|
"""Disambiguate against wdirid."""
|
|
1633
|
"""Disambiguate against wdirid."""
|
|
1632
|
for length in range(minlength, len(hexnode) + 1):
|
|
1634
|
for length in range(minlength, len(hexnode) + 1):
|
|
1633
|
prefix = hexnode[:length]
|
|
1635
|
prefix = hexnode[:length]
|
|
1634
|
if not maybewdir(prefix):
|
|
1636
|
if not maybewdir(prefix):
|
|
1635
|
return prefix
|
|
1637
|
return prefix
|
|
1636
|
|
|
1638
|
|
|
1637
|
if not getattr(self, 'filteredrevs', None):
|
|
1639
|
if not getattr(self, 'filteredrevs', None):
|
|
1638
|
try:
|
|
1640
|
try:
|
|
1639
|
length = max(self.index.shortest(node), minlength)
|
|
1641
|
length = max(self.index.shortest(node), minlength)
|
|
1640
|
return disambiguate(hexnode, length)
|
|
1642
|
return disambiguate(hexnode, length)
|
|
1641
|
except error.RevlogError:
|
|
1643
|
except error.RevlogError:
|
|
1642
|
if node != self.nodeconstants.wdirid:
|
|
1644
|
if node != self.nodeconstants.wdirid:
|
|
1643
|
raise error.LookupError(
|
|
1645
|
raise error.LookupError(
|
|
1644
|
node, self.display_id, _(b'no node')
|
|
1646
|
node, self.display_id, _(b'no node')
|
|
1645
|
)
|
|
1647
|
)
|
|
1646
|
except AttributeError:
|
|
1648
|
except AttributeError:
|
|
1647
|
# Fall through to pure code
|
|
1649
|
# Fall through to pure code
|
|
1648
|
pass
|
|
1650
|
pass
|
|
1649
|
|
|
1651
|
|
|
1650
|
if node == self.nodeconstants.wdirid:
|
|
1652
|
if node == self.nodeconstants.wdirid:
|
|
1651
|
for length in range(minlength, len(hexnode) + 1):
|
|
1653
|
for length in range(minlength, len(hexnode) + 1):
|
|
1652
|
prefix = hexnode[:length]
|
|
1654
|
prefix = hexnode[:length]
|
|
1653
|
if isvalid(prefix):
|
|
1655
|
if isvalid(prefix):
|
|
1654
|
return prefix
|
|
1656
|
return prefix
|
|
1655
|
|
|
1657
|
|
|
1656
|
for length in range(minlength, len(hexnode) + 1):
|
|
1658
|
for length in range(minlength, len(hexnode) + 1):
|
|
1657
|
prefix = hexnode[:length]
|
|
1659
|
prefix = hexnode[:length]
|
|
1658
|
if isvalid(prefix):
|
|
1660
|
if isvalid(prefix):
|
|
1659
|
return disambiguate(hexnode, length)
|
|
1661
|
return disambiguate(hexnode, length)
|
|
1660
|
|
|
1662
|
|
|
1661
|
def cmp(self, node, text):
|
|
1663
|
def cmp(self, node, text):
|
|
1662
|
"""compare text with a given file revision
|
|
1664
|
"""compare text with a given file revision
|
|
1663
|
|
|
1665
|
|
|
1664
|
returns True if text is different than what is stored.
|
|
1666
|
returns True if text is different than what is stored.
|
|
1665
|
"""
|
|
1667
|
"""
|
|
1666
|
p1, p2 = self.parents(node)
|
|
1668
|
p1, p2 = self.parents(node)
|
|
1667
|
return storageutil.hashrevisionsha1(text, p1, p2) != node
|
|
1669
|
return storageutil.hashrevisionsha1(text, p1, p2) != node
|
|
1668
|
|
|
1670
|
|
|
1669
|
def _cachesegment(self, offset, data):
|
|
1671
|
def _cachesegment(self, offset, data):
|
|
1670
|
"""Add a segment to the revlog cache.
|
|
1672
|
"""Add a segment to the revlog cache.
|
|
1671
|
|
|
1673
|
|
|
1672
|
Accepts an absolute offset and the data that is at that location.
|
|
1674
|
Accepts an absolute offset and the data that is at that location.
|
|
1673
|
"""
|
|
1675
|
"""
|
|
1674
|
o, d = self._chunkcache
|
|
1676
|
o, d = self._chunkcache
|
|
1675
|
# try to add to existing cache
|
|
1677
|
# try to add to existing cache
|
|
1676
|
if o + len(d) == offset and len(d) + len(data) < _chunksize:
|
|
1678
|
if o + len(d) == offset and len(d) + len(data) < _chunksize:
|
|
1677
|
self._chunkcache = o, d + data
|
|
1679
|
self._chunkcache = o, d + data
|
|
1678
|
else:
|
|
1680
|
else:
|
|
1679
|
self._chunkcache = offset, data
|
|
1681
|
self._chunkcache = offset, data
|
|
1680
|
|
|
1682
|
|
|
1681
|
def _readsegment(self, offset, length, df=None):
|
|
1683
|
def _readsegment(self, offset, length, df=None):
|
|
1682
|
"""Load a segment of raw data from the revlog.
|
|
1684
|
"""Load a segment of raw data from the revlog.
|
|
1683
|
|
|
1685
|
|
|
1684
|
Accepts an absolute offset, length to read, and an optional existing
|
|
1686
|
Accepts an absolute offset, length to read, and an optional existing
|
|
1685
|
file handle to read from.
|
|
1687
|
file handle to read from.
|
|
1686
|
|
|
1688
|
|
|
1687
|
If an existing file handle is passed, it will be seeked and the
|
|
1689
|
If an existing file handle is passed, it will be seeked and the
|
|
1688
|
original seek position will NOT be restored.
|
|
1690
|
original seek position will NOT be restored.
|
|
1689
|
|
|
1691
|
|
|
1690
|
Returns a str or buffer of raw byte data.
|
|
1692
|
Returns a str or buffer of raw byte data.
|
|
1691
|
|
|
1693
|
|
|
1692
|
Raises if the requested number of bytes could not be read.
|
|
1694
|
Raises if the requested number of bytes could not be read.
|
|
1693
|
"""
|
|
1695
|
"""
|
|
1694
|
# Cache data both forward and backward around the requested
|
|
1696
|
# Cache data both forward and backward around the requested
|
|
1695
|
# data, in a fixed size window. This helps speed up operations
|
|
1697
|
# data, in a fixed size window. This helps speed up operations
|
|
1696
|
# involving reading the revlog backwards.
|
|
1698
|
# involving reading the revlog backwards.
|
|
1697
|
cachesize = self._chunkcachesize
|
|
1699
|
cachesize = self._chunkcachesize
|
|
1698
|
realoffset = offset & ~(cachesize - 1)
|
|
1700
|
realoffset = offset & ~(cachesize - 1)
|
|
1699
|
reallength = (
|
|
1701
|
reallength = (
|
|
1700
|
(offset + length + cachesize) & ~(cachesize - 1)
|
|
1702
|
(offset + length + cachesize) & ~(cachesize - 1)
|
|
1701
|
) - realoffset
|
|
1703
|
) - realoffset
|
|
1702
|
with self._datareadfp(df) as df:
|
|
1704
|
with self._datareadfp(df) as df:
|
|
1703
|
df.seek(realoffset)
|
|
1705
|
df.seek(realoffset)
|
|
1704
|
d = df.read(reallength)
|
|
1706
|
d = df.read(reallength)
|
|
1705
|
|
|
1707
|
|
|
1706
|
self._cachesegment(realoffset, d)
|
|
1708
|
self._cachesegment(realoffset, d)
|
|
1707
|
if offset != realoffset or reallength != length:
|
|
1709
|
if offset != realoffset or reallength != length:
|
|
1708
|
startoffset = offset - realoffset
|
|
1710
|
startoffset = offset - realoffset
|
|
1709
|
if len(d) - startoffset < length:
|
|
1711
|
if len(d) - startoffset < length:
|
|
1710
|
raise error.RevlogError(
|
|
1712
|
raise error.RevlogError(
|
|
1711
|
_(
|
|
1713
|
_(
|
|
1712
|
b'partial read of revlog %s; expected %d bytes from '
|
|
1714
|
b'partial read of revlog %s; expected %d bytes from '
|
|
1713
|
b'offset %d, got %d'
|
|
1715
|
b'offset %d, got %d'
|
|
1714
|
)
|
|
1716
|
)
|
|
1715
|
% (
|
|
1717
|
% (
|
|
1716
|
self._indexfile if self._inline else self._datafile,
|
|
1718
|
self._indexfile if self._inline else self._datafile,
|
|
1717
|
length,
|
|
1719
|
length,
|
|
1718
|
offset,
|
|
1720
|
offset,
|
|
1719
|
len(d) - startoffset,
|
|
1721
|
len(d) - startoffset,
|
|
1720
|
)
|
|
1722
|
)
|
|
1721
|
)
|
|
1723
|
)
|
|
1722
|
|
|
1724
|
|
|
1723
|
return util.buffer(d, startoffset, length)
|
|
1725
|
return util.buffer(d, startoffset, length)
|
|
1724
|
|
|
1726
|
|
|
1725
|
if len(d) < length:
|
|
1727
|
if len(d) < length:
|
|
1726
|
raise error.RevlogError(
|
|
1728
|
raise error.RevlogError(
|
|
1727
|
_(
|
|
1729
|
_(
|
|
1728
|
b'partial read of revlog %s; expected %d bytes from offset '
|
|
1730
|
b'partial read of revlog %s; expected %d bytes from offset '
|
|
1729
|
b'%d, got %d'
|
|
1731
|
b'%d, got %d'
|
|
1730
|
)
|
|
1732
|
)
|
|
1731
|
% (
|
|
1733
|
% (
|
|
1732
|
self._indexfile if self._inline else self._datafile,
|
|
1734
|
self._indexfile if self._inline else self._datafile,
|
|
1733
|
length,
|
|
1735
|
length,
|
|
1734
|
offset,
|
|
1736
|
offset,
|
|
1735
|
len(d),
|
|
1737
|
len(d),
|
|
1736
|
)
|
|
1738
|
)
|
|
1737
|
)
|
|
1739
|
)
|
|
1738
|
|
|
1740
|
|
|
1739
|
return d
|
|
1741
|
return d
|
|
1740
|
|
|
1742
|
|
|
1741
|
def _getsegment(self, offset, length, df=None):
|
|
1743
|
def _getsegment(self, offset, length, df=None):
|
|
1742
|
"""Obtain a segment of raw data from the revlog.
|
|
1744
|
"""Obtain a segment of raw data from the revlog.
|
|
1743
|
|
|
1745
|
|
|
1744
|
Accepts an absolute offset, length of bytes to obtain, and an
|
|
1746
|
Accepts an absolute offset, length of bytes to obtain, and an
|
|
1745
|
optional file handle to the already-opened revlog. If the file
|
|
1747
|
optional file handle to the already-opened revlog. If the file
|
|
1746
|
handle is used, it's original seek position will not be preserved.
|
|
1748
|
handle is used, it's original seek position will not be preserved.
|
|
1747
|
|
|
1749
|
|
|
1748
|
Requests for data may be returned from a cache.
|
|
1750
|
Requests for data may be returned from a cache.
|
|
1749
|
|
|
1751
|
|
|
1750
|
Returns a str or a buffer instance of raw byte data.
|
|
1752
|
Returns a str or a buffer instance of raw byte data.
|
|
1751
|
"""
|
|
1753
|
"""
|
|
1752
|
o, d = self._chunkcache
|
|
1754
|
o, d = self._chunkcache
|
|
1753
|
l = len(d)
|
|
1755
|
l = len(d)
|
|
1754
|
|
|
1756
|
|
|
1755
|
# is it in the cache?
|
|
1757
|
# is it in the cache?
|
|
1756
|
cachestart = offset - o
|
|
1758
|
cachestart = offset - o
|
|
1757
|
cacheend = cachestart + length
|
|
1759
|
cacheend = cachestart + length
|
|
1758
|
if cachestart >= 0 and cacheend <= l:
|
|
1760
|
if cachestart >= 0 and cacheend <= l:
|
|
1759
|
if cachestart == 0 and cacheend == l:
|
|
1761
|
if cachestart == 0 and cacheend == l:
|
|
1760
|
return d # avoid a copy
|
|
1762
|
return d # avoid a copy
|
|
1761
|
return util.buffer(d, cachestart, cacheend - cachestart)
|
|
1763
|
return util.buffer(d, cachestart, cacheend - cachestart)
|
|
1762
|
|
|
1764
|
|
|
1763
|
return self._readsegment(offset, length, df=df)
|
|
1765
|
return self._readsegment(offset, length, df=df)
|
|
1764
|
|
|
1766
|
|
|
1765
|
def _getsegmentforrevs(self, startrev, endrev, df=None):
|
|
1767
|
def _getsegmentforrevs(self, startrev, endrev, df=None):
|
|
1766
|
"""Obtain a segment of raw data corresponding to a range of revisions.
|
|
1768
|
"""Obtain a segment of raw data corresponding to a range of revisions.
|
|
1767
|
|
|
1769
|
|
|
1768
|
Accepts the start and end revisions and an optional already-open
|
|
1770
|
Accepts the start and end revisions and an optional already-open
|
|
1769
|
file handle to be used for reading. If the file handle is read, its
|
|
1771
|
file handle to be used for reading. If the file handle is read, its
|
|
1770
|
seek position will not be preserved.
|
|
1772
|
seek position will not be preserved.
|
|
1771
|
|
|
1773
|
|
|
1772
|
Requests for data may be satisfied by a cache.
|
|
1774
|
Requests for data may be satisfied by a cache.
|
|
1773
|
|
|
1775
|
|
|
1774
|
Returns a 2-tuple of (offset, data) for the requested range of
|
|
1776
|
Returns a 2-tuple of (offset, data) for the requested range of
|
|
1775
|
revisions. Offset is the integer offset from the beginning of the
|
|
1777
|
revisions. Offset is the integer offset from the beginning of the
|
|
1776
|
revlog and data is a str or buffer of the raw byte data.
|
|
1778
|
revlog and data is a str or buffer of the raw byte data.
|
|
1777
|
|
|
1779
|
|
|
1778
|
Callers will need to call ``self.start(rev)`` and ``self.length(rev)``
|
|
1780
|
Callers will need to call ``self.start(rev)`` and ``self.length(rev)``
|
|
1779
|
to determine where each revision's data begins and ends.
|
|
1781
|
to determine where each revision's data begins and ends.
|
|
1780
|
"""
|
|
1782
|
"""
|
|
1781
|
# Inlined self.start(startrev) & self.end(endrev) for perf reasons
|
|
1783
|
# Inlined self.start(startrev) & self.end(endrev) for perf reasons
|
|
1782
|
# (functions are expensive).
|
|
1784
|
# (functions are expensive).
|
|
1783
|
index = self.index
|
|
1785
|
index = self.index
|
|
1784
|
istart = index[startrev]
|
|
1786
|
istart = index[startrev]
|
|
1785
|
start = int(istart[0] >> 16)
|
|
1787
|
start = int(istart[0] >> 16)
|
|
1786
|
if startrev == endrev:
|
|
1788
|
if startrev == endrev:
|
|
1787
|
end = start + istart[1]
|
|
1789
|
end = start + istart[1]
|
|
1788
|
else:
|
|
1790
|
else:
|
|
1789
|
iend = index[endrev]
|
|
1791
|
iend = index[endrev]
|
|
1790
|
end = int(iend[0] >> 16) + iend[1]
|
|
1792
|
end = int(iend[0] >> 16) + iend[1]
|
|
1791
|
|
|
1793
|
|
|
1792
|
if self._inline:
|
|
1794
|
if self._inline:
|
|
1793
|
start += (startrev + 1) * self.index.entry_size
|
|
1795
|
start += (startrev + 1) * self.index.entry_size
|
|
1794
|
end += (endrev + 1) * self.index.entry_size
|
|
1796
|
end += (endrev + 1) * self.index.entry_size
|
|
1795
|
length = end - start
|
|
1797
|
length = end - start
|
|
1796
|
|
|
1798
|
|
|
1797
|
return start, self._getsegment(start, length, df=df)
|
|
1799
|
return start, self._getsegment(start, length, df=df)
|
|
1798
|
|
|
1800
|
|
|
1799
|
def _chunk(self, rev, df=None):
|
|
1801
|
def _chunk(self, rev, df=None):
|
|
1800
|
"""Obtain a single decompressed chunk for a revision.
|
|
1802
|
"""Obtain a single decompressed chunk for a revision.
|
|
1801
|
|
|
1803
|
|
|
1802
|
Accepts an integer revision and an optional already-open file handle
|
|
1804
|
Accepts an integer revision and an optional already-open file handle
|
|
1803
|
to be used for reading. If used, the seek position of the file will not
|
|
1805
|
to be used for reading. If used, the seek position of the file will not
|
|
1804
|
be preserved.
|
|
1806
|
be preserved.
|
|
1805
|
|
|
1807
|
|
|
1806
|
Returns a str holding uncompressed data for the requested revision.
|
|
1808
|
Returns a str holding uncompressed data for the requested revision.
|
|
1807
|
"""
|
|
1809
|
"""
|
|
1808
|
compression_mode = self.index[rev][10]
|
|
1810
|
compression_mode = self.index[rev][10]
|
|
1809
|
data = self._getsegmentforrevs(rev, rev, df=df)[1]
|
|
1811
|
data = self._getsegmentforrevs(rev, rev, df=df)[1]
|
|
1810
|
if compression_mode == COMP_MODE_PLAIN:
|
|
1812
|
if compression_mode == COMP_MODE_PLAIN:
|
|
1811
|
return data
|
|
1813
|
return data
|
|
1812
|
elif compression_mode == COMP_MODE_DEFAULT:
|
|
1814
|
elif compression_mode == COMP_MODE_DEFAULT:
|
|
1813
|
return self._decompressor(data)
|
|
1815
|
return self._decompressor(data)
|
|
1814
|
elif compression_mode == COMP_MODE_INLINE:
|
|
1816
|
elif compression_mode == COMP_MODE_INLINE:
|
|
1815
|
return self.decompress(data)
|
|
1817
|
return self.decompress(data)
|
|
1816
|
else:
|
|
1818
|
else:
|
|
1817
|
msg = 'unknown compression mode %d'
|
|
1819
|
msg = 'unknown compression mode %d'
|
|
1818
|
msg %= compression_mode
|
|
1820
|
msg %= compression_mode
|
|
1819
|
raise error.RevlogError(msg)
|
|
1821
|
raise error.RevlogError(msg)
|
|
1820
|
|
|
1822
|
|
|
1821
|
def _chunks(self, revs, df=None, targetsize=None):
|
|
1823
|
def _chunks(self, revs, df=None, targetsize=None):
|
|
1822
|
"""Obtain decompressed chunks for the specified revisions.
|
|
1824
|
"""Obtain decompressed chunks for the specified revisions.
|
|
1823
|
|
|
1825
|
|
|
1824
|
Accepts an iterable of numeric revisions that are assumed to be in
|
|
1826
|
Accepts an iterable of numeric revisions that are assumed to be in
|
|
1825
|
ascending order. Also accepts an optional already-open file handle
|
|
1827
|
ascending order. Also accepts an optional already-open file handle
|
|
1826
|
to be used for reading. If used, the seek position of the file will
|
|
1828
|
to be used for reading. If used, the seek position of the file will
|
|
1827
|
not be preserved.
|
|
1829
|
not be preserved.
|
|
1828
|
|
|
1830
|
|
|
1829
|
This function is similar to calling ``self._chunk()`` multiple times,
|
|
1831
|
This function is similar to calling ``self._chunk()`` multiple times,
|
|
1830
|
but is faster.
|
|
1832
|
but is faster.
|
|
1831
|
|
|
1833
|
|
|
1832
|
Returns a list with decompressed data for each requested revision.
|
|
1834
|
Returns a list with decompressed data for each requested revision.
|
|
1833
|
"""
|
|
1835
|
"""
|
|
1834
|
if not revs:
|
|
1836
|
if not revs:
|
|
1835
|
return []
|
|
1837
|
return []
|
|
1836
|
start = self.start
|
|
1838
|
start = self.start
|
|
1837
|
length = self.length
|
|
1839
|
length = self.length
|
|
1838
|
inline = self._inline
|
|
1840
|
inline = self._inline
|
|
1839
|
iosize = self.index.entry_size
|
|
1841
|
iosize = self.index.entry_size
|
|
1840
|
buffer = util.buffer
|
|
1842
|
buffer = util.buffer
|
|
1841
|
|
|
1843
|
|
|
1842
|
l = []
|
|
1844
|
l = []
|
|
1843
|
ladd = l.append
|
|
1845
|
ladd = l.append
|
|
1844
|
|
|
1846
|
|
|
1845
|
if not self._withsparseread:
|
|
1847
|
if not self._withsparseread:
|
|
1846
|
slicedchunks = (revs,)
|
|
1848
|
slicedchunks = (revs,)
|
|
1847
|
else:
|
|
1849
|
else:
|
|
1848
|
slicedchunks = deltautil.slicechunk(
|
|
1850
|
slicedchunks = deltautil.slicechunk(
|
|
1849
|
self, revs, targetsize=targetsize
|
|
1851
|
self, revs, targetsize=targetsize
|
|
1850
|
)
|
|
1852
|
)
|
|
1851
|
|
|
1853
|
|
|
1852
|
for revschunk in slicedchunks:
|
|
1854
|
for revschunk in slicedchunks:
|
|
1853
|
firstrev = revschunk[0]
|
|
1855
|
firstrev = revschunk[0]
|
|
1854
|
# Skip trailing revisions with empty diff
|
|
1856
|
# Skip trailing revisions with empty diff
|
|
1855
|
for lastrev in revschunk[::-1]:
|
|
1857
|
for lastrev in revschunk[::-1]:
|
|
1856
|
if length(lastrev) != 0:
|
|
1858
|
if length(lastrev) != 0:
|
|
1857
|
break
|
|
1859
|
break
|
|
1858
|
|
|
1860
|
|
|
1859
|
try:
|
|
1861
|
try:
|
|
1860
|
offset, data = self._getsegmentforrevs(firstrev, lastrev, df=df)
|
|
1862
|
offset, data = self._getsegmentforrevs(firstrev, lastrev, df=df)
|
|
1861
|
except OverflowError:
|
|
1863
|
except OverflowError:
|
|
1862
|
# issue4215 - we can't cache a run of chunks greater than
|
|
1864
|
# issue4215 - we can't cache a run of chunks greater than
|
|
1863
|
# 2G on Windows
|
|
1865
|
# 2G on Windows
|
|
1864
|
return [self._chunk(rev, df=df) for rev in revschunk]
|
|
1866
|
return [self._chunk(rev, df=df) for rev in revschunk]
|
|
1865
|
|
|
1867
|
|
|
1866
|
decomp = self.decompress
|
|
1868
|
decomp = self.decompress
|
|
1867
|
# self._decompressor might be None, but will not be used in that case
|
|
1869
|
# self._decompressor might be None, but will not be used in that case
|
|
1868
|
def_decomp = self._decompressor
|
|
1870
|
def_decomp = self._decompressor
|
|
1869
|
for rev in revschunk:
|
|
1871
|
for rev in revschunk:
|
|
1870
|
chunkstart = start(rev)
|
|
1872
|
chunkstart = start(rev)
|
|
1871
|
if inline:
|
|
1873
|
if inline:
|
|
1872
|
chunkstart += (rev + 1) * iosize
|
|
1874
|
chunkstart += (rev + 1) * iosize
|
|
1873
|
chunklength = length(rev)
|
|
1875
|
chunklength = length(rev)
|
|
1874
|
comp_mode = self.index[rev][10]
|
|
1876
|
comp_mode = self.index[rev][10]
|
|
1875
|
c = buffer(data, chunkstart - offset, chunklength)
|
|
1877
|
c = buffer(data, chunkstart - offset, chunklength)
|
|
1876
|
if comp_mode == COMP_MODE_PLAIN:
|
|
1878
|
if comp_mode == COMP_MODE_PLAIN:
|
|
1877
|
ladd(c)
|
|
1879
|
ladd(c)
|
|
1878
|
elif comp_mode == COMP_MODE_INLINE:
|
|
1880
|
elif comp_mode == COMP_MODE_INLINE:
|
|
1879
|
ladd(decomp(c))
|
|
1881
|
ladd(decomp(c))
|
|
1880
|
elif comp_mode == COMP_MODE_DEFAULT:
|
|
1882
|
elif comp_mode == COMP_MODE_DEFAULT:
|
|
1881
|
ladd(def_decomp(c))
|
|
1883
|
ladd(def_decomp(c))
|
|
1882
|
else:
|
|
1884
|
else:
|
|
1883
|
msg = 'unknown compression mode %d'
|
|
1885
|
msg = 'unknown compression mode %d'
|
|
1884
|
msg %= comp_mode
|
|
1886
|
msg %= comp_mode
|
|
1885
|
raise error.RevlogError(msg)
|
|
1887
|
raise error.RevlogError(msg)
|
|
1886
|
|
|
1888
|
|
|
1887
|
return l
|
|
1889
|
return l
|
|
1888
|
|
|
1890
|
|
|
1889
|
def _chunkclear(self):
|
|
1891
|
def _chunkclear(self):
|
|
1890
|
"""Clear the raw chunk cache."""
|
|
1892
|
"""Clear the raw chunk cache."""
|
|
1891
|
self._chunkcache = (0, b'')
|
|
1893
|
self._chunkcache = (0, b'')
|
|
1892
|
|
|
1894
|
|
|
1893
|
def deltaparent(self, rev):
|
|
1895
|
def deltaparent(self, rev):
|
|
1894
|
"""return deltaparent of the given revision"""
|
|
1896
|
"""return deltaparent of the given revision"""
|
|
1895
|
base = self.index[rev][3]
|
|
1897
|
base = self.index[rev][3]
|
|
1896
|
if base == rev:
|
|
1898
|
if base == rev:
|
|
1897
|
return nullrev
|
|
1899
|
return nullrev
|
|
1898
|
elif self._generaldelta:
|
|
1900
|
elif self._generaldelta:
|
|
1899
|
return base
|
|
1901
|
return base
|
|
1900
|
else:
|
|
1902
|
else:
|
|
1901
|
return rev - 1
|
|
1903
|
return rev - 1
|
|
1902
|
|
|
1904
|
|
|
1903
|
def issnapshot(self, rev):
|
|
1905
|
def issnapshot(self, rev):
|
|
1904
|
"""tells whether rev is a snapshot"""
|
|
1906
|
"""tells whether rev is a snapshot"""
|
|
1905
|
if not self._sparserevlog:
|
|
1907
|
if not self._sparserevlog:
|
|
1906
|
return self.deltaparent(rev) == nullrev
|
|
1908
|
return self.deltaparent(rev) == nullrev
|
|
1907
|
elif util.safehasattr(self.index, b'issnapshot'):
|
|
1909
|
elif util.safehasattr(self.index, b'issnapshot'):
|
|
1908
|
# directly assign the method to cache the testing and access
|
|
1910
|
# directly assign the method to cache the testing and access
|
|
1909
|
self.issnapshot = self.index.issnapshot
|
|
1911
|
self.issnapshot = self.index.issnapshot
|
|
1910
|
return self.issnapshot(rev)
|
|
1912
|
return self.issnapshot(rev)
|
|
1911
|
if rev == nullrev:
|
|
1913
|
if rev == nullrev:
|
|
1912
|
return True
|
|
1914
|
return True
|
|
1913
|
entry = self.index[rev]
|
|
1915
|
entry = self.index[rev]
|
|
1914
|
base = entry[3]
|
|
1916
|
base = entry[3]
|
|
1915
|
if base == rev:
|
|
1917
|
if base == rev:
|
|
1916
|
return True
|
|
1918
|
return True
|
|
1917
|
if base == nullrev:
|
|
1919
|
if base == nullrev:
|
|
1918
|
return True
|
|
1920
|
return True
|
|
1919
|
p1 = entry[5]
|
|
1921
|
p1 = entry[5]
|
|
1920
|
p2 = entry[6]
|
|
1922
|
p2 = entry[6]
|
|
1921
|
if base == p1 or base == p2:
|
|
1923
|
if base == p1 or base == p2:
|
|
1922
|
return False
|
|
1924
|
return False
|
|
1923
|
return self.issnapshot(base)
|
|
1925
|
return self.issnapshot(base)
|
|
1924
|
|
|
1926
|
|
|
1925
|
def snapshotdepth(self, rev):
|
|
1927
|
def snapshotdepth(self, rev):
|
|
1926
|
"""number of snapshot in the chain before this one"""
|
|
1928
|
"""number of snapshot in the chain before this one"""
|
|
1927
|
if not self.issnapshot(rev):
|
|
1929
|
if not self.issnapshot(rev):
|
|
1928
|
raise error.ProgrammingError(b'revision %d not a snapshot')
|
|
1930
|
raise error.ProgrammingError(b'revision %d not a snapshot')
|
|
1929
|
return len(self._deltachain(rev)[0]) - 1
|
|
1931
|
return len(self._deltachain(rev)[0]) - 1
|
|
1930
|
|
|
1932
|
|
|
1931
|
def revdiff(self, rev1, rev2):
|
|
1933
|
def revdiff(self, rev1, rev2):
|
|
1932
|
"""return or calculate a delta between two revisions
|
|
1934
|
"""return or calculate a delta between two revisions
|
|
1933
|
|
|
1935
|
|
|
1934
|
The delta calculated is in binary form and is intended to be written to
|
|
1936
|
The delta calculated is in binary form and is intended to be written to
|
|
1935
|
revlog data directly. So this function needs raw revision data.
|
|
1937
|
revlog data directly. So this function needs raw revision data.
|
|
1936
|
"""
|
|
1938
|
"""
|
|
1937
|
if rev1 != nullrev and self.deltaparent(rev2) == rev1:
|
|
1939
|
if rev1 != nullrev and self.deltaparent(rev2) == rev1:
|
|
1938
|
return bytes(self._chunk(rev2))
|
|
1940
|
return bytes(self._chunk(rev2))
|
|
1939
|
|
|
1941
|
|
|
1940
|
return mdiff.textdiff(self.rawdata(rev1), self.rawdata(rev2))
|
|
1942
|
return mdiff.textdiff(self.rawdata(rev1), self.rawdata(rev2))
|
|
1941
|
|
|
1943
|
|
|
1942
|
def _processflags(self, text, flags, operation, raw=False):
|
|
1944
|
def _processflags(self, text, flags, operation, raw=False):
|
|
1943
|
"""deprecated entry point to access flag processors"""
|
|
1945
|
"""deprecated entry point to access flag processors"""
|
|
1944
|
msg = b'_processflag(...) use the specialized variant'
|
|
1946
|
msg = b'_processflag(...) use the specialized variant'
|
|
1945
|
util.nouideprecwarn(msg, b'5.2', stacklevel=2)
|
|
1947
|
util.nouideprecwarn(msg, b'5.2', stacklevel=2)
|
|
1946
|
if raw:
|
|
1948
|
if raw:
|
|
1947
|
return text, flagutil.processflagsraw(self, text, flags)
|
|
1949
|
return text, flagutil.processflagsraw(self, text, flags)
|
|
1948
|
elif operation == b'read':
|
|
1950
|
elif operation == b'read':
|
|
1949
|
return flagutil.processflagsread(self, text, flags)
|
|
1951
|
return flagutil.processflagsread(self, text, flags)
|
|
1950
|
else: # write operation
|
|
1952
|
else: # write operation
|
|
1951
|
return flagutil.processflagswrite(self, text, flags)
|
|
1953
|
return flagutil.processflagswrite(self, text, flags)
|
|
1952
|
|
|
1954
|
|
|
1953
|
def revision(self, nodeorrev, _df=None, raw=False):
|
|
1955
|
def revision(self, nodeorrev, _df=None, raw=False):
|
|
1954
|
"""return an uncompressed revision of a given node or revision
|
|
1956
|
"""return an uncompressed revision of a given node or revision
|
|
1955
|
number.
|
|
1957
|
number.
|
|
1956
|
|
|
1958
|
|
|
1957
|
_df - an existing file handle to read from. (internal-only)
|
|
1959
|
_df - an existing file handle to read from. (internal-only)
|
|
1958
|
raw - an optional argument specifying if the revision data is to be
|
|
1960
|
raw - an optional argument specifying if the revision data is to be
|
|
1959
|
treated as raw data when applying flag transforms. 'raw' should be set
|
|
1961
|
treated as raw data when applying flag transforms. 'raw' should be set
|
|
1960
|
to True when generating changegroups or in debug commands.
|
|
1962
|
to True when generating changegroups or in debug commands.
|
|
1961
|
"""
|
|
1963
|
"""
|
|
1962
|
if raw:
|
|
1964
|
if raw:
|
|
1963
|
msg = (
|
|
1965
|
msg = (
|
|
1964
|
b'revlog.revision(..., raw=True) is deprecated, '
|
|
1966
|
b'revlog.revision(..., raw=True) is deprecated, '
|
|
1965
|
b'use revlog.rawdata(...)'
|
|
1967
|
b'use revlog.rawdata(...)'
|
|
1966
|
)
|
|
1968
|
)
|
|
1967
|
util.nouideprecwarn(msg, b'5.2', stacklevel=2)
|
|
1969
|
util.nouideprecwarn(msg, b'5.2', stacklevel=2)
|
|
1968
|
return self._revisiondata(nodeorrev, _df, raw=raw)[0]
|
|
1970
|
return self._revisiondata(nodeorrev, _df, raw=raw)[0]
|
|
1969
|
|
|
1971
|
|
|
1970
|
def sidedata(self, nodeorrev, _df=None):
|
|
1972
|
def sidedata(self, nodeorrev, _df=None):
|
|
1971
|
"""a map of extra data related to the changeset but not part of the hash
|
|
1973
|
"""a map of extra data related to the changeset but not part of the hash
|
|
1972
|
|
|
1974
|
|
|
1973
|
This function currently return a dictionary. However, more advanced
|
|
1975
|
This function currently return a dictionary. However, more advanced
|
|
1974
|
mapping object will likely be used in the future for a more
|
|
1976
|
mapping object will likely be used in the future for a more
|
|
1975
|
efficient/lazy code.
|
|
1977
|
efficient/lazy code.
|
|
1976
|
"""
|
|
1978
|
"""
|
|
1977
|
return self._revisiondata(nodeorrev, _df)[1]
|
|
1979
|
return self._revisiondata(nodeorrev, _df)[1]
|
|
1978
|
|
|
1980
|
|
|
1979
|
def _revisiondata(self, nodeorrev, _df=None, raw=False):
|
|
1981
|
def _revisiondata(self, nodeorrev, _df=None, raw=False):
|
|
1980
|
# deal with <nodeorrev> argument type
|
|
1982
|
# deal with <nodeorrev> argument type
|
|
1981
|
if isinstance(nodeorrev, int):
|
|
1983
|
if isinstance(nodeorrev, int):
|
|
1982
|
rev = nodeorrev
|
|
1984
|
rev = nodeorrev
|
|
1983
|
node = self.node(rev)
|
|
1985
|
node = self.node(rev)
|
|
1984
|
else:
|
|
1986
|
else:
|
|
1985
|
node = nodeorrev
|
|
1987
|
node = nodeorrev
|
|
1986
|
rev = None
|
|
1988
|
rev = None
|
|
1987
|
|
|
1989
|
|
|
1988
|
# fast path the special `nullid` rev
|
|
1990
|
# fast path the special `nullid` rev
|
|
1989
|
if node == self.nullid:
|
|
1991
|
if node == self.nullid:
|
|
1990
|
return b"", {}
|
|
1992
|
return b"", {}
|
|
1991
|
|
|
1993
|
|
|
1992
|
# ``rawtext`` is the text as stored inside the revlog. Might be the
|
|
1994
|
# ``rawtext`` is the text as stored inside the revlog. Might be the
|
|
1993
|
# revision or might need to be processed to retrieve the revision.
|
|
1995
|
# revision or might need to be processed to retrieve the revision.
|
|
1994
|
rev, rawtext, validated = self._rawtext(node, rev, _df=_df)
|
|
1996
|
rev, rawtext, validated = self._rawtext(node, rev, _df=_df)
|
|
1995
|
|
|
1997
|
|
|
1996
|
if self.hassidedata:
|
|
1998
|
if self.hassidedata:
|
|
1997
|
if rev is None:
|
|
1999
|
if rev is None:
|
|
1998
|
rev = self.rev(node)
|
|
2000
|
rev = self.rev(node)
|
|
1999
|
sidedata = self._sidedata(rev)
|
|
2001
|
sidedata = self._sidedata(rev)
|
|
2000
|
else:
|
|
2002
|
else:
|
|
2001
|
sidedata = {}
|
|
2003
|
sidedata = {}
|
|
2002
|
|
|
2004
|
|
|
2003
|
if raw and validated:
|
|
2005
|
if raw and validated:
|
|
2004
|
# if we don't want to process the raw text and that raw
|
|
2006
|
# if we don't want to process the raw text and that raw
|
|
2005
|
# text is cached, we can exit early.
|
|
2007
|
# text is cached, we can exit early.
|
|
2006
|
return rawtext, sidedata
|
|
2008
|
return rawtext, sidedata
|
|
2007
|
if rev is None:
|
|
2009
|
if rev is None:
|
|
2008
|
rev = self.rev(node)
|
|
2010
|
rev = self.rev(node)
|
|
2009
|
# the revlog's flag for this revision
|
|
2011
|
# the revlog's flag for this revision
|
|
2010
|
# (usually alter its state or content)
|
|
2012
|
# (usually alter its state or content)
|
|
2011
|
flags = self.flags(rev)
|
|
2013
|
flags = self.flags(rev)
|
|
2012
|
|
|
2014
|
|
|
2013
|
if validated and flags == REVIDX_DEFAULT_FLAGS:
|
|
2015
|
if validated and flags == REVIDX_DEFAULT_FLAGS:
|
|
2014
|
# no extra flags set, no flag processor runs, text = rawtext
|
|
2016
|
# no extra flags set, no flag processor runs, text = rawtext
|
|
2015
|
return rawtext, sidedata
|
|
2017
|
return rawtext, sidedata
|
|
2016
|
|
|
2018
|
|
|
2017
|
if raw:
|
|
2019
|
if raw:
|
|
2018
|
validatehash = flagutil.processflagsraw(self, rawtext, flags)
|
|
2020
|
validatehash = flagutil.processflagsraw(self, rawtext, flags)
|
|
2019
|
text = rawtext
|
|
2021
|
text = rawtext
|
|
2020
|
else:
|
|
2022
|
else:
|
|
2021
|
r = flagutil.processflagsread(self, rawtext, flags)
|
|
2023
|
r = flagutil.processflagsread(self, rawtext, flags)
|
|
2022
|
text, validatehash = r
|
|
2024
|
text, validatehash = r
|
|
2023
|
if validatehash:
|
|
2025
|
if validatehash:
|
|
2024
|
self.checkhash(text, node, rev=rev)
|
|
2026
|
self.checkhash(text, node, rev=rev)
|
|
2025
|
if not validated:
|
|
2027
|
if not validated:
|
|
2026
|
self._revisioncache = (node, rev, rawtext)
|
|
2028
|
self._revisioncache = (node, rev, rawtext)
|
|
2027
|
|
|
2029
|
|
|
2028
|
return text, sidedata
|
|
2030
|
return text, sidedata
|
|
2029
|
|
|
2031
|
|
|
2030
|
def _rawtext(self, node, rev, _df=None):
|
|
2032
|
def _rawtext(self, node, rev, _df=None):
|
|
2031
|
"""return the possibly unvalidated rawtext for a revision
|
|
2033
|
"""return the possibly unvalidated rawtext for a revision
|
|
2032
|
|
|
2034
|
|
|
2033
|
returns (rev, rawtext, validated)
|
|
2035
|
returns (rev, rawtext, validated)
|
|
2034
|
"""
|
|
2036
|
"""
|
|
2035
|
|
|
2037
|
|
|
2036
|
# revision in the cache (could be useful to apply delta)
|
|
2038
|
# revision in the cache (could be useful to apply delta)
|
|
2037
|
cachedrev = None
|
|
2039
|
cachedrev = None
|
|
2038
|
# An intermediate text to apply deltas to
|
|
2040
|
# An intermediate text to apply deltas to
|
|
2039
|
basetext = None
|
|
2041
|
basetext = None
|
|
2040
|
|
|
2042
|
|
|
2041
|
# Check if we have the entry in cache
|
|
2043
|
# Check if we have the entry in cache
|
|
2042
|
# The cache entry looks like (node, rev, rawtext)
|
|
2044
|
# The cache entry looks like (node, rev, rawtext)
|
|
2043
|
if self._revisioncache:
|
|
2045
|
if self._revisioncache:
|
|
2044
|
if self._revisioncache[0] == node:
|
|
2046
|
if self._revisioncache[0] == node:
|
|
2045
|
return (rev, self._revisioncache[2], True)
|
|
2047
|
return (rev, self._revisioncache[2], True)
|
|
2046
|
cachedrev = self._revisioncache[1]
|
|
2048
|
cachedrev = self._revisioncache[1]
|
|
2047
|
|
|
2049
|
|
|
2048
|
if rev is None:
|
|
2050
|
if rev is None:
|
|
2049
|
rev = self.rev(node)
|
|
2051
|
rev = self.rev(node)
|
|
2050
|
|
|
2052
|
|
|
2051
|
chain, stopped = self._deltachain(rev, stoprev=cachedrev)
|
|
2053
|
chain, stopped = self._deltachain(rev, stoprev=cachedrev)
|
|
2052
|
if stopped:
|
|
2054
|
if stopped:
|
|
2053
|
basetext = self._revisioncache[2]
|
|
2055
|
basetext = self._revisioncache[2]
|
|
2054
|
|
|
2056
|
|
|
2055
|
# drop cache to save memory, the caller is expected to
|
|
2057
|
# drop cache to save memory, the caller is expected to
|
|
2056
|
# update self._revisioncache after validating the text
|
|
2058
|
# update self._revisioncache after validating the text
|
|
2057
|
self._revisioncache = None
|
|
2059
|
self._revisioncache = None
|
|
2058
|
|
|
2060
|
|
|
2059
|
targetsize = None
|
|
2061
|
targetsize = None
|
|
2060
|
rawsize = self.index[rev][2]
|
|
2062
|
rawsize = self.index[rev][2]
|
|
2061
|
if 0 <= rawsize:
|
|
2063
|
if 0 <= rawsize:
|
|
2062
|
targetsize = 4 * rawsize
|
|
2064
|
targetsize = 4 * rawsize
|
|
2063
|
|
|
2065
|
|
|
2064
|
bins = self._chunks(chain, df=_df, targetsize=targetsize)
|
|
2066
|
bins = self._chunks(chain, df=_df, targetsize=targetsize)
|
|
2065
|
if basetext is None:
|
|
2067
|
if basetext is None:
|
|
2066
|
basetext = bytes(bins[0])
|
|
2068
|
basetext = bytes(bins[0])
|
|
2067
|
bins = bins[1:]
|
|
2069
|
bins = bins[1:]
|
|
2068
|
|
|
2070
|
|
|
2069
|
rawtext = mdiff.patches(basetext, bins)
|
|
2071
|
rawtext = mdiff.patches(basetext, bins)
|
|
2070
|
del basetext # let us have a chance to free memory early
|
|
2072
|
del basetext # let us have a chance to free memory early
|
|
2071
|
return (rev, rawtext, False)
|
|
2073
|
return (rev, rawtext, False)
|
|
2072
|
|
|
2074
|
|
|
2073
|
def _sidedata(self, rev):
|
|
2075
|
def _sidedata(self, rev):
|
|
2074
|
"""Return the sidedata for a given revision number."""
|
|
2076
|
"""Return the sidedata for a given revision number."""
|
|
2075
|
index_entry = self.index[rev]
|
|
2077
|
index_entry = self.index[rev]
|
|
2076
|
sidedata_offset = index_entry[8]
|
|
2078
|
sidedata_offset = index_entry[8]
|
|
2077
|
sidedata_size = index_entry[9]
|
|
2079
|
sidedata_size = index_entry[9]
|
|
2078
|
|
|
2080
|
|
|
2079
|
if self._inline:
|
|
2081
|
if self._inline:
|
|
2080
|
sidedata_offset += self.index.entry_size * (1 + rev)
|
|
2082
|
sidedata_offset += self.index.entry_size * (1 + rev)
|
|
2081
|
if sidedata_size == 0:
|
|
2083
|
if sidedata_size == 0:
|
|
2082
|
return {}
|
|
2084
|
return {}
|
|
2083
|
|
|
2085
|
|
|
2084
|
comp_segment = self._getsegment(sidedata_offset, sidedata_size)
|
|
2086
|
comp_segment = self._getsegment(sidedata_offset, sidedata_size)
|
|
2085
|
comp = self.index[rev][11]
|
|
2087
|
comp = self.index[rev][11]
|
|
2086
|
if comp == COMP_MODE_PLAIN:
|
|
2088
|
if comp == COMP_MODE_PLAIN:
|
|
2087
|
segment = comp_segment
|
|
2089
|
segment = comp_segment
|
|
2088
|
elif comp == COMP_MODE_DEFAULT:
|
|
2090
|
elif comp == COMP_MODE_DEFAULT:
|
|
2089
|
segment = self._decompressor(comp_segment)
|
|
2091
|
segment = self._decompressor(comp_segment)
|
|
2090
|
elif comp == COMP_MODE_INLINE:
|
|
2092
|
elif comp == COMP_MODE_INLINE:
|
|
2091
|
segment = self.decompress(comp_segment)
|
|
2093
|
segment = self.decompress(comp_segment)
|
|
2092
|
else:
|
|
2094
|
else:
|
|
2093
|
msg = 'unknown compression mode %d'
|
|
2095
|
msg = 'unknown compression mode %d'
|
|
2094
|
msg %= comp
|
|
2096
|
msg %= comp
|
|
2095
|
raise error.RevlogError(msg)
|
|
2097
|
raise error.RevlogError(msg)
|
|
2096
|
|
|
2098
|
|
|
2097
|
sidedata = sidedatautil.deserialize_sidedata(segment)
|
|
2099
|
sidedata = sidedatautil.deserialize_sidedata(segment)
|
|
2098
|
return sidedata
|
|
2100
|
return sidedata
|
|
2099
|
|
|
2101
|
|
|
2100
|
def rawdata(self, nodeorrev, _df=None):
|
|
2102
|
def rawdata(self, nodeorrev, _df=None):
|
|
2101
|
"""return an uncompressed raw data of a given node or revision number.
|
|
2103
|
"""return an uncompressed raw data of a given node or revision number.
|
|
2102
|
|
|
2104
|
|
|
2103
|
_df - an existing file handle to read from. (internal-only)
|
|
2105
|
_df - an existing file handle to read from. (internal-only)
|
|
2104
|
"""
|
|
2106
|
"""
|
|
2105
|
return self._revisiondata(nodeorrev, _df, raw=True)[0]
|
|
2107
|
return self._revisiondata(nodeorrev, _df, raw=True)[0]
|
|
2106
|
|
|
2108
|
|
|
2107
|
def hash(self, text, p1, p2):
|
|
2109
|
def hash(self, text, p1, p2):
|
|
2108
|
"""Compute a node hash.
|
|
2110
|
"""Compute a node hash.
|
|
2109
|
|
|
2111
|
|
|
2110
|
Available as a function so that subclasses can replace the hash
|
|
2112
|
Available as a function so that subclasses can replace the hash
|
|
2111
|
as needed.
|
|
2113
|
as needed.
|
|
2112
|
"""
|
|
2114
|
"""
|
|
2113
|
return storageutil.hashrevisionsha1(text, p1, p2)
|
|
2115
|
return storageutil.hashrevisionsha1(text, p1, p2)
|
|
2114
|
|
|
2116
|
|
|
2115
|
def checkhash(self, text, node, p1=None, p2=None, rev=None):
|
|
2117
|
def checkhash(self, text, node, p1=None, p2=None, rev=None):
|
|
2116
|
"""Check node hash integrity.
|
|
2118
|
"""Check node hash integrity.
|
|
2117
|
|
|
2119
|
|
|
2118
|
Available as a function so that subclasses can extend hash mismatch
|
|
2120
|
Available as a function so that subclasses can extend hash mismatch
|
|
2119
|
behaviors as needed.
|
|
2121
|
behaviors as needed.
|
|
2120
|
"""
|
|
2122
|
"""
|
|
2121
|
try:
|
|
2123
|
try:
|
|
2122
|
if p1 is None and p2 is None:
|
|
2124
|
if p1 is None and p2 is None:
|
|
2123
|
p1, p2 = self.parents(node)
|
|
2125
|
p1, p2 = self.parents(node)
|
|
2124
|
if node != self.hash(text, p1, p2):
|
|
2126
|
if node != self.hash(text, p1, p2):
|
|
2125
|
# Clear the revision cache on hash failure. The revision cache
|
|
2127
|
# Clear the revision cache on hash failure. The revision cache
|
|
2126
|
# only stores the raw revision and clearing the cache does have
|
|
2128
|
# only stores the raw revision and clearing the cache does have
|
|
2127
|
# the side-effect that we won't have a cache hit when the raw
|
|
2129
|
# the side-effect that we won't have a cache hit when the raw
|
|
2128
|
# revision data is accessed. But this case should be rare and
|
|
2130
|
# revision data is accessed. But this case should be rare and
|
|
2129
|
# it is extra work to teach the cache about the hash
|
|
2131
|
# it is extra work to teach the cache about the hash
|
|
2130
|
# verification state.
|
|
2132
|
# verification state.
|
|
2131
|
if self._revisioncache and self._revisioncache[0] == node:
|
|
2133
|
if self._revisioncache and self._revisioncache[0] == node:
|
|
2132
|
self._revisioncache = None
|
|
2134
|
self._revisioncache = None
|
|
2133
|
|
|
2135
|
|
|
2134
|
revornode = rev
|
|
2136
|
revornode = rev
|
|
2135
|
if revornode is None:
|
|
2137
|
if revornode is None:
|
|
2136
|
revornode = templatefilters.short(hex(node))
|
|
2138
|
revornode = templatefilters.short(hex(node))
|
|
2137
|
raise error.RevlogError(
|
|
2139
|
raise error.RevlogError(
|
|
2138
|
_(b"integrity check failed on %s:%s")
|
|
2140
|
_(b"integrity check failed on %s:%s")
|
|
2139
|
% (self.display_id, pycompat.bytestr(revornode))
|
|
2141
|
% (self.display_id, pycompat.bytestr(revornode))
|
|
2140
|
)
|
|
2142
|
)
|
|
2141
|
except error.RevlogError:
|
|
2143
|
except error.RevlogError:
|
|
2142
|
if self._censorable and storageutil.iscensoredtext(text):
|
|
2144
|
if self._censorable and storageutil.iscensoredtext(text):
|
|
2143
|
raise error.CensoredNodeError(self.display_id, node, text)
|
|
2145
|
raise error.CensoredNodeError(self.display_id, node, text)
|
|
2144
|
raise
|
|
2146
|
raise
|
|
2145
|
|
|
2147
|
|
|
2146
|
def _enforceinlinesize(self, tr):
|
|
2148
|
def _enforceinlinesize(self, tr):
|
|
2147
|
"""Check if the revlog is too big for inline and convert if so.
|
|
2149
|
"""Check if the revlog is too big for inline and convert if so.
|
|
2148
|
|
|
2150
|
|
|
2149
|
This should be called after revisions are added to the revlog. If the
|
|
2151
|
This should be called after revisions are added to the revlog. If the
|
|
2150
|
revlog has grown too large to be an inline revlog, it will convert it
|
|
2152
|
revlog has grown too large to be an inline revlog, it will convert it
|
|
2151
|
to use multiple index and data files.
|
|
2153
|
to use multiple index and data files.
|
|
2152
|
"""
|
|
2154
|
"""
|
|
2153
|
tiprev = len(self) - 1
|
|
2155
|
tiprev = len(self) - 1
|
|
2154
|
total_size = self.start(tiprev) + self.length(tiprev)
|
|
2156
|
total_size = self.start(tiprev) + self.length(tiprev)
|
|
2155
|
if not self._inline or total_size < _maxinline:
|
|
2157
|
if not self._inline or total_size < _maxinline:
|
|
2156
|
return
|
|
2158
|
return
|
|
2157
|
|
|
2159
|
|
|
2158
|
troffset = tr.findoffset(self._indexfile)
|
|
2160
|
troffset = tr.findoffset(self._indexfile)
|
|
2159
|
if troffset is None:
|
|
2161
|
if troffset is None:
|
|
2160
|
raise error.RevlogError(
|
|
2162
|
raise error.RevlogError(
|
|
2161
|
_(b"%s not found in the transaction") % self._indexfile
|
|
2163
|
_(b"%s not found in the transaction") % self._indexfile
|
|
2162
|
)
|
|
2164
|
)
|
|
2163
|
trindex = 0
|
|
2165
|
trindex = 0
|
|
2164
|
tr.add(self._datafile, 0)
|
|
2166
|
tr.add(self._datafile, 0)
|
|
2165
|
|
|
2167
|
|
|
2166
|
existing_handles = False
|
|
2168
|
existing_handles = False
|
|
2167
|
if self._writinghandles is not None:
|
|
2169
|
if self._writinghandles is not None:
|
|
2168
|
existing_handles = True
|
|
2170
|
existing_handles = True
|
|
2169
|
fp = self._writinghandles[0]
|
|
2171
|
fp = self._writinghandles[0]
|
|
2170
|
fp.flush()
|
|
2172
|
fp.flush()
|
|
2171
|
fp.close()
|
|
2173
|
fp.close()
|
|
2172
|
# We can't use the cached file handle after close(). So prevent
|
|
2174
|
# We can't use the cached file handle after close(). So prevent
|
|
2173
|
# its usage.
|
|
2175
|
# its usage.
|
|
2174
|
self._writinghandles = None
|
|
2176
|
self._writinghandles = None
|
|
2175
|
|
|
2177
|
|
|
2176
|
new_dfh = self._datafp(b'w+')
|
|
2178
|
new_dfh = self._datafp(b'w+')
|
|
2177
|
new_dfh.truncate(0) # drop any potentially existing data
|
|
2179
|
new_dfh.truncate(0) # drop any potentially existing data
|
|
2178
|
try:
|
|
2180
|
try:
|
|
2179
|
with self._indexfp() as read_ifh:
|
|
2181
|
with self._indexfp() as read_ifh:
|
|
2180
|
for r in self:
|
|
2182
|
for r in self:
|
|
2181
|
new_dfh.write(self._getsegmentforrevs(r, r, df=read_ifh)[1])
|
|
2183
|
new_dfh.write(self._getsegmentforrevs(r, r, df=read_ifh)[1])
|
|
2182
|
if troffset <= self.start(r) + r * self.index.entry_size:
|
|
2184
|
if troffset <= self.start(r) + r * self.index.entry_size:
|
|
2183
|
trindex = r
|
|
2185
|
trindex = r
|
|
2184
|
new_dfh.flush()
|
|
2186
|
new_dfh.flush()
|
|
2185
|
|
|
2187
|
|
|
2186
|
with self.__index_new_fp() as fp:
|
|
2188
|
with self.__index_new_fp() as fp:
|
|
2187
|
self._format_flags &= ~FLAG_INLINE_DATA
|
|
2189
|
self._format_flags &= ~FLAG_INLINE_DATA
|
|
2188
|
self._inline = False
|
|
2190
|
self._inline = False
|
|
2189
|
for i in self:
|
|
2191
|
for i in self:
|
|
2190
|
e = self.index.entry_binary(i)
|
|
2192
|
e = self.index.entry_binary(i)
|
|
2191
|
if i == 0 and self._docket is None:
|
|
2193
|
if i == 0 and self._docket is None:
|
|
2192
|
header = self._format_flags | self._format_version
|
|
2194
|
header = self._format_flags | self._format_version
|
|
2193
|
header = self.index.pack_header(header)
|
|
2195
|
header = self.index.pack_header(header)
|
|
2194
|
e = header + e
|
|
2196
|
e = header + e
|
|
2195
|
fp.write(e)
|
|
2197
|
fp.write(e)
|
|
2196
|
if self._docket is not None:
|
|
2198
|
if self._docket is not None:
|
|
2197
|
self._docket.index_end = fp.tell()
|
|
2199
|
self._docket.index_end = fp.tell()
|
|
2198
|
|
|
2200
|
|
|
2199
|
# There is a small transactional race here. If the rename of
|
|
2201
|
# There is a small transactional race here. If the rename of
|
|
2200
|
# the index fails, we should remove the datafile. It is more
|
|
2202
|
# the index fails, we should remove the datafile. It is more
|
|
2201
|
# important to ensure that the data file is not truncated
|
|
2203
|
# important to ensure that the data file is not truncated
|
|
2202
|
# when the index is replaced as otherwise data is lost.
|
|
2204
|
# when the index is replaced as otherwise data is lost.
|
|
2203
|
tr.replace(self._datafile, self.start(trindex))
|
|
2205
|
tr.replace(self._datafile, self.start(trindex))
|
|
2204
|
|
|
2206
|
|
|
2205
|
# the temp file replace the real index when we exit the context
|
|
2207
|
# the temp file replace the real index when we exit the context
|
|
2206
|
# manager
|
|
2208
|
# manager
|
|
2207
|
|
|
2209
|
|
|
2208
|
tr.replace(self._indexfile, trindex * self.index.entry_size)
|
|
2210
|
tr.replace(self._indexfile, trindex * self.index.entry_size)
|
|
2209
|
nodemaputil.setup_persistent_nodemap(tr, self)
|
|
2211
|
nodemaputil.setup_persistent_nodemap(tr, self)
|
|
2210
|
self._chunkclear()
|
|
2212
|
self._chunkclear()
|
|
2211
|
|
|
2213
|
|
|
2212
|
if existing_handles:
|
|
2214
|
if existing_handles:
|
|
2213
|
# switched from inline to conventional reopen the index
|
|
2215
|
# switched from inline to conventional reopen the index
|
|
2214
|
ifh = self.__index_write_fp()
|
|
2216
|
ifh = self.__index_write_fp()
|
|
2215
|
self._writinghandles = (ifh, new_dfh)
|
|
2217
|
self._writinghandles = (ifh, new_dfh)
|
|
2216
|
new_dfh = None
|
|
2218
|
new_dfh = None
|
|
2217
|
finally:
|
|
2219
|
finally:
|
|
2218
|
if new_dfh is not None:
|
|
2220
|
if new_dfh is not None:
|
|
2219
|
new_dfh.close()
|
|
2221
|
new_dfh.close()
|
|
2220
|
|
|
2222
|
|
|
2221
|
def _nodeduplicatecallback(self, transaction, node):
|
|
2223
|
def _nodeduplicatecallback(self, transaction, node):
|
|
2222
|
"""called when trying to add a node already stored."""
|
|
2224
|
"""called when trying to add a node already stored."""
|
|
2223
|
|
|
2225
|
|
|
2224
|
@contextlib.contextmanager
|
|
2226
|
@contextlib.contextmanager
|
|
2225
|
def _writing(self, transaction):
|
|
2227
|
def _writing(self, transaction):
|
|
2226
|
if self._trypending:
|
|
2228
|
if self._trypending:
|
|
2227
|
msg = b'try to write in a `trypending` revlog: %s'
|
|
2229
|
msg = b'try to write in a `trypending` revlog: %s'
|
|
2228
|
msg %= self.display_id
|
|
2230
|
msg %= self.display_id
|
|
2229
|
raise error.ProgrammingError(msg)
|
|
2231
|
raise error.ProgrammingError(msg)
|
|
2230
|
if self._writinghandles is not None:
|
|
2232
|
if self._writinghandles is not None:
|
|
2231
|
yield
|
|
2233
|
yield
|
|
2232
|
else:
|
|
2234
|
else:
|
|
2233
|
r = len(self)
|
|
2235
|
r = len(self)
|
|
2234
|
dsize = 0
|
|
2236
|
dsize = 0
|
|
2235
|
if r:
|
|
2237
|
if r:
|
|
2236
|
dsize = self.end(r - 1)
|
|
2238
|
dsize = self.end(r - 1)
|
|
2237
|
dfh = None
|
|
2239
|
dfh = None
|
|
2238
|
if not self._inline:
|
|
2240
|
if not self._inline:
|
|
2239
|
try:
|
|
2241
|
try:
|
|
2240
|
dfh = self._datafp(b"r+")
|
|
2242
|
dfh = self._datafp(b"r+")
|
|
2241
|
if self._docket is None:
|
|
2243
|
if self._docket is None:
|
|
2242
|
dfh.seek(0, os.SEEK_END)
|
|
2244
|
dfh.seek(0, os.SEEK_END)
|
|
2243
|
else:
|
|
2245
|
else:
|
|
2244
|
dfh.seek(self._docket.data_end, os.SEEK_SET)
|
|
2246
|
dfh.seek(self._docket.data_end, os.SEEK_SET)
|
|
2245
|
except IOError as inst:
|
|
2247
|
except IOError as inst:
|
|
2246
|
if inst.errno != errno.ENOENT:
|
|
2248
|
if inst.errno != errno.ENOENT:
|
|
2247
|
raise
|
|
2249
|
raise
|
|
2248
|
dfh = self._datafp(b"w+")
|
|
2250
|
dfh = self._datafp(b"w+")
|
|
2249
|
transaction.add(self._datafile, dsize)
|
|
2251
|
transaction.add(self._datafile, dsize)
|
|
2250
|
try:
|
|
2252
|
try:
|
|
2251
|
isize = r * self.index.entry_size
|
|
2253
|
isize = r * self.index.entry_size
|
|
2252
|
ifh = self.__index_write_fp()
|
|
2254
|
ifh = self.__index_write_fp()
|
|
2253
|
if self._inline:
|
|
2255
|
if self._inline:
|
|
2254
|
transaction.add(self._indexfile, dsize + isize)
|
|
2256
|
transaction.add(self._indexfile, dsize + isize)
|
|
2255
|
else:
|
|
2257
|
else:
|
|
2256
|
transaction.add(self._indexfile, isize)
|
|
2258
|
transaction.add(self._indexfile, isize)
|
|
2257
|
try:
|
|
2259
|
try:
|
|
2258
|
self._writinghandles = (ifh, dfh)
|
|
2260
|
self._writinghandles = (ifh, dfh)
|
|
2259
|
try:
|
|
2261
|
try:
|
|
2260
|
yield
|
|
2262
|
yield
|
|
2261
|
if self._docket is not None:
|
|
2263
|
if self._docket is not None:
|
|
2262
|
self._write_docket(transaction)
|
|
2264
|
self._write_docket(transaction)
|
|
2263
|
finally:
|
|
2265
|
finally:
|
|
2264
|
self._writinghandles = None
|
|
2266
|
self._writinghandles = None
|
|
2265
|
finally:
|
|
2267
|
finally:
|
|
2266
|
ifh.close()
|
|
2268
|
ifh.close()
|
|
2267
|
finally:
|
|
2269
|
finally:
|
|
2268
|
if dfh is not None:
|
|
2270
|
if dfh is not None:
|
|
2269
|
dfh.close()
|
|
2271
|
dfh.close()
|
|
2270
|
|
|
2272
|
|
|
2271
|
def _write_docket(self, transaction):
|
|
2273
|
def _write_docket(self, transaction):
|
|
2272
|
"""write the current docket on disk
|
|
2274
|
"""write the current docket on disk
|
|
2273
|
|
|
2275
|
|
|
2274
|
Exist as a method to help changelog to implement transaction logic
|
|
2276
|
Exist as a method to help changelog to implement transaction logic
|
|
2275
|
|
|
2277
|
|
|
2276
|
We could also imagine using the same transaction logic for all revlog
|
|
2278
|
We could also imagine using the same transaction logic for all revlog
|
|
2277
|
since docket are cheap."""
|
|
2279
|
since docket are cheap."""
|
|
2278
|
self._docket.write(transaction)
|
|
2280
|
self._docket.write(transaction)
|
|
2279
|
|
|
2281
|
|
|
2280
|
def addrevision(
|
|
2282
|
def addrevision(
|
|
2281
|
self,
|
|
2283
|
self,
|
|
2282
|
text,
|
|
2284
|
text,
|
|
2283
|
transaction,
|
|
2285
|
transaction,
|
|
2284
|
link,
|
|
2286
|
link,
|
|
2285
|
p1,
|
|
2287
|
p1,
|
|
2286
|
p2,
|
|
2288
|
p2,
|
|
2287
|
cachedelta=None,
|
|
2289
|
cachedelta=None,
|
|
2288
|
node=None,
|
|
2290
|
node=None,
|
|
2289
|
flags=REVIDX_DEFAULT_FLAGS,
|
|
2291
|
flags=REVIDX_DEFAULT_FLAGS,
|
|
2290
|
deltacomputer=None,
|
|
2292
|
deltacomputer=None,
|
|
2291
|
sidedata=None,
|
|
2293
|
sidedata=None,
|
|
2292
|
):
|
|
2294
|
):
|
|
2293
|
"""add a revision to the log
|
|
2295
|
"""add a revision to the log
|
|
2294
|
|
|
2296
|
|
|
2295
|
text - the revision data to add
|
|
2297
|
text - the revision data to add
|
|
2296
|
transaction - the transaction object used for rollback
|
|
2298
|
transaction - the transaction object used for rollback
|
|
2297
|
link - the linkrev data to add
|
|
2299
|
link - the linkrev data to add
|
|
2298
|
p1, p2 - the parent nodeids of the revision
|
|
2300
|
p1, p2 - the parent nodeids of the revision
|
|
2299
|
cachedelta - an optional precomputed delta
|
|
2301
|
cachedelta - an optional precomputed delta
|
|
2300
|
node - nodeid of revision; typically node is not specified, and it is
|
|
2302
|
node - nodeid of revision; typically node is not specified, and it is
|
|
2301
|
computed by default as hash(text, p1, p2), however subclasses might
|
|
2303
|
computed by default as hash(text, p1, p2), however subclasses might
|
|
2302
|
use different hashing method (and override checkhash() in such case)
|
|
2304
|
use different hashing method (and override checkhash() in such case)
|
|
2303
|
flags - the known flags to set on the revision
|
|
2305
|
flags - the known flags to set on the revision
|
|
2304
|
deltacomputer - an optional deltacomputer instance shared between
|
|
2306
|
deltacomputer - an optional deltacomputer instance shared between
|
|
2305
|
multiple calls
|
|
2307
|
multiple calls
|
|
2306
|
"""
|
|
2308
|
"""
|
|
2307
|
if link == nullrev:
|
|
2309
|
if link == nullrev:
|
|
2308
|
raise error.RevlogError(
|
|
2310
|
raise error.RevlogError(
|
|
2309
|
_(b"attempted to add linkrev -1 to %s") % self.display_id
|
|
2311
|
_(b"attempted to add linkrev -1 to %s") % self.display_id
|
|
2310
|
)
|
|
2312
|
)
|
|
2311
|
|
|
2313
|
|
|
2312
|
if sidedata is None:
|
|
2314
|
if sidedata is None:
|
|
2313
|
sidedata = {}
|
|
2315
|
sidedata = {}
|
|
2314
|
elif sidedata and not self.hassidedata:
|
|
2316
|
elif sidedata and not self.hassidedata:
|
|
2315
|
raise error.ProgrammingError(
|
|
2317
|
raise error.ProgrammingError(
|
|
2316
|
_(b"trying to add sidedata to a revlog who don't support them")
|
|
2318
|
_(b"trying to add sidedata to a revlog who don't support them")
|
|
2317
|
)
|
|
2319
|
)
|
|
2318
|
|
|
2320
|
|
|
2319
|
if flags:
|
|
2321
|
if flags:
|
|
2320
|
node = node or self.hash(text, p1, p2)
|
|
2322
|
node = node or self.hash(text, p1, p2)
|
|
2321
|
|
|
2323
|
|
|
2322
|
rawtext, validatehash = flagutil.processflagswrite(self, text, flags)
|
|
2324
|
rawtext, validatehash = flagutil.processflagswrite(self, text, flags)
|
|
2323
|
|
|
2325
|
|
|
2324
|
# If the flag processor modifies the revision data, ignore any provided
|
|
2326
|
# If the flag processor modifies the revision data, ignore any provided
|
|
2325
|
# cachedelta.
|
|
2327
|
# cachedelta.
|
|
2326
|
if rawtext != text:
|
|
2328
|
if rawtext != text:
|
|
2327
|
cachedelta = None
|
|
2329
|
cachedelta = None
|
|
2328
|
|
|
2330
|
|
|
2329
|
if len(rawtext) > _maxentrysize:
|
|
2331
|
if len(rawtext) > _maxentrysize:
|
|
2330
|
raise error.RevlogError(
|
|
2332
|
raise error.RevlogError(
|
|
2331
|
_(
|
|
2333
|
_(
|
|
2332
|
b"%s: size of %d bytes exceeds maximum revlog storage of 2GiB"
|
|
2334
|
b"%s: size of %d bytes exceeds maximum revlog storage of 2GiB"
|
|
2333
|
)
|
|
2335
|
)
|
|
2334
|
% (self.display_id, len(rawtext))
|
|
2336
|
% (self.display_id, len(rawtext))
|
|
2335
|
)
|
|
2337
|
)
|
|
2336
|
|
|
2338
|
|
|
2337
|
node = node or self.hash(rawtext, p1, p2)
|
|
2339
|
node = node or self.hash(rawtext, p1, p2)
|
|
2338
|
rev = self.index.get_rev(node)
|
|
2340
|
rev = self.index.get_rev(node)
|
|
2339
|
if rev is not None:
|
|
2341
|
if rev is not None:
|
|
2340
|
return rev
|
|
2342
|
return rev
|
|
2341
|
|
|
2343
|
|
|
2342
|
if validatehash:
|
|
2344
|
if validatehash:
|
|
2343
|
self.checkhash(rawtext, node, p1=p1, p2=p2)
|
|
2345
|
self.checkhash(rawtext, node, p1=p1, p2=p2)
|
|
2344
|
|
|
2346
|
|
|
2345
|
return self.addrawrevision(
|
|
2347
|
return self.addrawrevision(
|
|
2346
|
rawtext,
|
|
2348
|
rawtext,
|
|
2347
|
transaction,
|
|
2349
|
transaction,
|
|
2348
|
link,
|
|
2350
|
link,
|
|
2349
|
p1,
|
|
2351
|
p1,
|
|
2350
|
p2,
|
|
2352
|
p2,
|
|
2351
|
node,
|
|
2353
|
node,
|
|
2352
|
flags,
|
|
2354
|
flags,
|
|
2353
|
cachedelta=cachedelta,
|
|
2355
|
cachedelta=cachedelta,
|
|
2354
|
deltacomputer=deltacomputer,
|
|
2356
|
deltacomputer=deltacomputer,
|
|
2355
|
sidedata=sidedata,
|
|
2357
|
sidedata=sidedata,
|
|
2356
|
)
|
|
2358
|
)
|
|
2357
|
|
|
2359
|
|
|
2358
|
def addrawrevision(
|
|
2360
|
def addrawrevision(
|
|
2359
|
self,
|
|
2361
|
self,
|
|
2360
|
rawtext,
|
|
2362
|
rawtext,
|
|
2361
|
transaction,
|
|
2363
|
transaction,
|
|
2362
|
link,
|
|
2364
|
link,
|
|
2363
|
p1,
|
|
2365
|
p1,
|
|
2364
|
p2,
|
|
2366
|
p2,
|
|
2365
|
node,
|
|
2367
|
node,
|
|
2366
|
flags,
|
|
2368
|
flags,
|
|
2367
|
cachedelta=None,
|
|
2369
|
cachedelta=None,
|
|
2368
|
deltacomputer=None,
|
|
2370
|
deltacomputer=None,
|
|
2369
|
sidedata=None,
|
|
2371
|
sidedata=None,
|
|
2370
|
):
|
|
2372
|
):
|
|
2371
|
"""add a raw revision with known flags, node and parents
|
|
2373
|
"""add a raw revision with known flags, node and parents
|
|
2372
|
useful when reusing a revision not stored in this revlog (ex: received
|
|
2374
|
useful when reusing a revision not stored in this revlog (ex: received
|
|
2373
|
over wire, or read from an external bundle).
|
|
2375
|
over wire, or read from an external bundle).
|
|
2374
|
"""
|
|
2376
|
"""
|
|
2375
|
with self._writing(transaction):
|
|
2377
|
with self._writing(transaction):
|
|
2376
|
return self._addrevision(
|
|
2378
|
return self._addrevision(
|
|
2377
|
node,
|
|
2379
|
node,
|
|
2378
|
rawtext,
|
|
2380
|
rawtext,
|
|
2379
|
transaction,
|
|
2381
|
transaction,
|
|
2380
|
link,
|
|
2382
|
link,
|
|
2381
|
p1,
|
|
2383
|
p1,
|
|
2382
|
p2,
|
|
2384
|
p2,
|
|
2383
|
flags,
|
|
2385
|
flags,
|
|
2384
|
cachedelta,
|
|
2386
|
cachedelta,
|
|
2385
|
deltacomputer=deltacomputer,
|
|
2387
|
deltacomputer=deltacomputer,
|
|
2386
|
sidedata=sidedata,
|
|
2388
|
sidedata=sidedata,
|
|
2387
|
)
|
|
2389
|
)
|
|
2388
|
|
|
2390
|
|
|
2389
|
def compress(self, data):
|
|
2391
|
def compress(self, data):
|
|
2390
|
"""Generate a possibly-compressed representation of data."""
|
|
2392
|
"""Generate a possibly-compressed representation of data."""
|
|
2391
|
if not data:
|
|
2393
|
if not data:
|
|
2392
|
return b'', data
|
|
2394
|
return b'', data
|
|
2393
|
|
|
2395
|
|
|
2394
|
compressed = self._compressor.compress(data)
|
|
2396
|
compressed = self._compressor.compress(data)
|
|
2395
|
|
|
2397
|
|
|
2396
|
if compressed:
|
|
2398
|
if compressed:
|
|
2397
|
# The revlog compressor added the header in the returned data.
|
|
2399
|
# The revlog compressor added the header in the returned data.
|
|
2398
|
return b'', compressed
|
|
2400
|
return b'', compressed
|
|
2399
|
|
|
2401
|
|
|
2400
|
if data[0:1] == b'\0':
|
|
2402
|
if data[0:1] == b'\0':
|
|
2401
|
return b'', data
|
|
2403
|
return b'', data
|
|
2402
|
return b'u', data
|
|
2404
|
return b'u', data
|
|
2403
|
|
|
2405
|
|
|
2404
|
def decompress(self, data):
|
|
2406
|
def decompress(self, data):
|
|
2405
|
"""Decompress a revlog chunk.
|
|
2407
|
"""Decompress a revlog chunk.
|
|
2406
|
|
|
2408
|
|
|
2407
|
The chunk is expected to begin with a header identifying the
|
|
2409
|
The chunk is expected to begin with a header identifying the
|
|
2408
|
format type so it can be routed to an appropriate decompressor.
|
|
2410
|
format type so it can be routed to an appropriate decompressor.
|
|
2409
|
"""
|
|
2411
|
"""
|
|
2410
|
if not data:
|
|
2412
|
if not data:
|
|
2411
|
return data
|
|
2413
|
return data
|
|
2412
|
|
|
2414
|
|
|
2413
|
# Revlogs are read much more frequently than they are written and many
|
|
2415
|
# Revlogs are read much more frequently than they are written and many
|
|
2414
|
# chunks only take microseconds to decompress, so performance is
|
|
2416
|
# chunks only take microseconds to decompress, so performance is
|
|
2415
|
# important here.
|
|
2417
|
# important here.
|
|
2416
|
#
|
|
2418
|
#
|
|
2417
|
# We can make a few assumptions about revlogs:
|
|
2419
|
# We can make a few assumptions about revlogs:
|
|
2418
|
#
|
|
2420
|
#
|
|
2419
|
# 1) the majority of chunks will be compressed (as opposed to inline
|
|
2421
|
# 1) the majority of chunks will be compressed (as opposed to inline
|
|
2420
|
# raw data).
|
|
2422
|
# raw data).
|
|
2421
|
# 2) decompressing *any* data will likely by at least 10x slower than
|
|
2423
|
# 2) decompressing *any* data will likely by at least 10x slower than
|
|
2422
|
# returning raw inline data.
|
|
2424
|
# returning raw inline data.
|
|
2423
|
# 3) we want to prioritize common and officially supported compression
|
|
2425
|
# 3) we want to prioritize common and officially supported compression
|
|
2424
|
# engines
|
|
2426
|
# engines
|
|
2425
|
#
|
|
2427
|
#
|
|
2426
|
# It follows that we want to optimize for "decompress compressed data
|
|
2428
|
# It follows that we want to optimize for "decompress compressed data
|
|
2427
|
# when encoded with common and officially supported compression engines"
|
|
2429
|
# when encoded with common and officially supported compression engines"
|
|
2428
|
# case over "raw data" and "data encoded by less common or non-official
|
|
2430
|
# case over "raw data" and "data encoded by less common or non-official
|
|
2429
|
# compression engines." That is why we have the inline lookup first
|
|
2431
|
# compression engines." That is why we have the inline lookup first
|
|
2430
|
# followed by the compengines lookup.
|
|
2432
|
# followed by the compengines lookup.
|
|
2431
|
#
|
|
2433
|
#
|
|
2432
|
# According to `hg perfrevlogchunks`, this is ~0.5% faster for zlib
|
|
2434
|
# According to `hg perfrevlogchunks`, this is ~0.5% faster for zlib
|
|
2433
|
# compressed chunks. And this matters for changelog and manifest reads.
|
|
2435
|
# compressed chunks. And this matters for changelog and manifest reads.
|
|
2434
|
t = data[0:1]
|
|
2436
|
t = data[0:1]
|
|
2435
|
|
|
2437
|
|
|
2436
|
if t == b'x':
|
|
2438
|
if t == b'x':
|
|
2437
|
try:
|
|
2439
|
try:
|
|
2438
|
return _zlibdecompress(data)
|
|
2440
|
return _zlibdecompress(data)
|
|
2439
|
except zlib.error as e:
|
|
2441
|
except zlib.error as e:
|
|
2440
|
raise error.RevlogError(
|
|
2442
|
raise error.RevlogError(
|
|
2441
|
_(b'revlog decompress error: %s')
|
|
2443
|
_(b'revlog decompress error: %s')
|
|
2442
|
% stringutil.forcebytestr(e)
|
|
2444
|
% stringutil.forcebytestr(e)
|
|
2443
|
)
|
|
2445
|
)
|
|
2444
|
# '\0' is more common than 'u' so it goes first.
|
|
2446
|
# '\0' is more common than 'u' so it goes first.
|
|
2445
|
elif t == b'\0':
|
|
2447
|
elif t == b'\0':
|
|
2446
|
return data
|
|
2448
|
return data
|
|
2447
|
elif t == b'u':
|
|
2449
|
elif t == b'u':
|
|
2448
|
return util.buffer(data, 1)
|
|
2450
|
return util.buffer(data, 1)
|
|
2449
|
|
|
2451
|
|
|
2450
|
compressor = self._get_decompressor(t)
|
|
2452
|
compressor = self._get_decompressor(t)
|
|
2451
|
|
|
2453
|
|
|
2452
|
return compressor.decompress(data)
|
|
2454
|
return compressor.decompress(data)
|
|
2453
|
|
|
2455
|
|
|
2454
|
def _addrevision(
|
|
2456
|
def _addrevision(
|
|
2455
|
self,
|
|
2457
|
self,
|
|
2456
|
node,
|
|
2458
|
node,
|
|
2457
|
rawtext,
|
|
2459
|
rawtext,
|
|
2458
|
transaction,
|
|
2460
|
transaction,
|
|
2459
|
link,
|
|
2461
|
link,
|
|
2460
|
p1,
|
|
2462
|
p1,
|
|
2461
|
p2,
|
|
2463
|
p2,
|
|
2462
|
flags,
|
|
2464
|
flags,
|
|
2463
|
cachedelta,
|
|
2465
|
cachedelta,
|
|
2464
|
alwayscache=False,
|
|
2466
|
alwayscache=False,
|
|
2465
|
deltacomputer=None,
|
|
2467
|
deltacomputer=None,
|
|
2466
|
sidedata=None,
|
|
2468
|
sidedata=None,
|
|
2467
|
):
|
|
2469
|
):
|
|
2468
|
"""internal function to add revisions to the log
|
|
2470
|
"""internal function to add revisions to the log
|
|
2469
|
|
|
2471
|
|
|
2470
|
see addrevision for argument descriptions.
|
|
2472
|
see addrevision for argument descriptions.
|
|
2471
|
|
|
2473
|
|
|
2472
|
note: "addrevision" takes non-raw text, "_addrevision" takes raw text.
|
|
2474
|
note: "addrevision" takes non-raw text, "_addrevision" takes raw text.
|
|
2473
|
|
|
2475
|
|
|
2474
|
if "deltacomputer" is not provided or None, a defaultdeltacomputer will
|
|
2476
|
if "deltacomputer" is not provided or None, a defaultdeltacomputer will
|
|
2475
|
be used.
|
|
2477
|
be used.
|
|
2476
|
|
|
2478
|
|
|
2477
|
invariants:
|
|
2479
|
invariants:
|
|
2478
|
- rawtext is optional (can be None); if not set, cachedelta must be set.
|
|
2480
|
- rawtext is optional (can be None); if not set, cachedelta must be set.
|
|
2479
|
if both are set, they must correspond to each other.
|
|
2481
|
if both are set, they must correspond to each other.
|
|
2480
|
"""
|
|
2482
|
"""
|
|
2481
|
if node == self.nullid:
|
|
2483
|
if node == self.nullid:
|
|
2482
|
raise error.RevlogError(
|
|
2484
|
raise error.RevlogError(
|
|
2483
|
_(b"%s: attempt to add null revision") % self.display_id
|
|
2485
|
_(b"%s: attempt to add null revision") % self.display_id
|
|
2484
|
)
|
|
2486
|
)
|
|
2485
|
if (
|
|
2487
|
if (
|
|
2486
|
node == self.nodeconstants.wdirid
|
|
2488
|
node == self.nodeconstants.wdirid
|
|
2487
|
or node in self.nodeconstants.wdirfilenodeids
|
|
2489
|
or node in self.nodeconstants.wdirfilenodeids
|
|
2488
|
):
|
|
2490
|
):
|
|
2489
|
raise error.RevlogError(
|
|
2491
|
raise error.RevlogError(
|
|
2490
|
_(b"%s: attempt to add wdir revision") % self.display_id
|
|
2492
|
_(b"%s: attempt to add wdir revision") % self.display_id
|
|
2491
|
)
|
|
2493
|
)
|
|
2492
|
if self._writinghandles is None:
|
|
2494
|
if self._writinghandles is None:
|
|
2493
|
msg = b'adding revision outside `revlog._writing` context'
|
|
2495
|
msg = b'adding revision outside `revlog._writing` context'
|
|
2494
|
raise error.ProgrammingError(msg)
|
|
2496
|
raise error.ProgrammingError(msg)
|
|
2495
|
|
|
2497
|
|
|
2496
|
if self._inline:
|
|
2498
|
if self._inline:
|
|
2497
|
fh = self._writinghandles[0]
|
|
2499
|
fh = self._writinghandles[0]
|
|
2498
|
else:
|
|
2500
|
else:
|
|
2499
|
fh = self._writinghandles[1]
|
|
2501
|
fh = self._writinghandles[1]
|
|
2500
|
|
|
2502
|
|
|
2501
|
btext = [rawtext]
|
|
2503
|
btext = [rawtext]
|
|
2502
|
|
|
2504
|
|
|
2503
|
curr = len(self)
|
|
2505
|
curr = len(self)
|
|
2504
|
prev = curr - 1
|
|
2506
|
prev = curr - 1
|
|
2505
|
|
|
2507
|
|
|
2506
|
offset = self._get_data_offset(prev)
|
|
2508
|
offset = self._get_data_offset(prev)
|
|
2507
|
|
|
2509
|
|
|
2508
|
if self._concurrencychecker:
|
|
2510
|
if self._concurrencychecker:
|
|
2509
|
ifh, dfh = self._writinghandles
|
|
2511
|
ifh, dfh = self._writinghandles
|
|
2510
|
if self._inline:
|
|
2512
|
if self._inline:
|
|
2511
|
# offset is "as if" it were in the .d file, so we need to add on
|
|
2513
|
# offset is "as if" it were in the .d file, so we need to add on
|
|
2512
|
# the size of the entry metadata.
|
|
2514
|
# the size of the entry metadata.
|
|
2513
|
self._concurrencychecker(
|
|
2515
|
self._concurrencychecker(
|
|
2514
|
ifh, self._indexfile, offset + curr * self.index.entry_size
|
|
2516
|
ifh, self._indexfile, offset + curr * self.index.entry_size
|
|
2515
|
)
|
|
2517
|
)
|
|
2516
|
else:
|
|
2518
|
else:
|
|
2517
|
# Entries in the .i are a consistent size.
|
|
2519
|
# Entries in the .i are a consistent size.
|
|
2518
|
self._concurrencychecker(
|
|
2520
|
self._concurrencychecker(
|
|
2519
|
ifh, self._indexfile, curr * self.index.entry_size
|
|
2521
|
ifh, self._indexfile, curr * self.index.entry_size
|
|
2520
|
)
|
|
2522
|
)
|
|
2521
|
self._concurrencychecker(dfh, self._datafile, offset)
|
|
2523
|
self._concurrencychecker(dfh, self._datafile, offset)
|
|
2522
|
|
|
2524
|
|
|
2523
|
p1r, p2r = self.rev(p1), self.rev(p2)
|
|
2525
|
p1r, p2r = self.rev(p1), self.rev(p2)
|
|
2524
|
|
|
2526
|
|
|
2525
|
# full versions are inserted when the needed deltas
|
|
2527
|
# full versions are inserted when the needed deltas
|
|
2526
|
# become comparable to the uncompressed text
|
|
2528
|
# become comparable to the uncompressed text
|
|
2527
|
if rawtext is None:
|
|
2529
|
if rawtext is None:
|
|
2528
|
# need rawtext size, before changed by flag processors, which is
|
|
2530
|
# need rawtext size, before changed by flag processors, which is
|
|
2529
|
# the non-raw size. use revlog explicitly to avoid filelog's extra
|
|
2531
|
# the non-raw size. use revlog explicitly to avoid filelog's extra
|
|
2530
|
# logic that might remove metadata size.
|
|
2532
|
# logic that might remove metadata size.
|
|
2531
|
textlen = mdiff.patchedsize(
|
|
2533
|
textlen = mdiff.patchedsize(
|
|
2532
|
revlog.size(self, cachedelta[0]), cachedelta[1]
|
|
2534
|
revlog.size(self, cachedelta[0]), cachedelta[1]
|
|
2533
|
)
|
|
2535
|
)
|
|
2534
|
else:
|
|
2536
|
else:
|
|
2535
|
textlen = len(rawtext)
|
|
2537
|
textlen = len(rawtext)
|
|
2536
|
|
|
2538
|
|
|
2537
|
if deltacomputer is None:
|
|
2539
|
if deltacomputer is None:
|
|
2538
|
deltacomputer = deltautil.deltacomputer(self)
|
|
2540
|
deltacomputer = deltautil.deltacomputer(self)
|
|
2539
|
|
|
2541
|
|
|
2540
|
revinfo = _revisioninfo(node, p1, p2, btext, textlen, cachedelta, flags)
|
|
2542
|
revinfo = _revisioninfo(node, p1, p2, btext, textlen, cachedelta, flags)
|
|
2541
|
|
|
2543
|
|
|
2542
|
deltainfo = deltacomputer.finddeltainfo(revinfo, fh)
|
|
2544
|
deltainfo = deltacomputer.finddeltainfo(revinfo, fh)
|
|
2543
|
|
|
2545
|
|
|
2544
|
compression_mode = COMP_MODE_INLINE
|
|
2546
|
compression_mode = COMP_MODE_INLINE
|
|
2545
|
if self._docket is not None:
|
|
2547
|
if self._docket is not None:
|
|
2546
|
h, d = deltainfo.data
|
|
2548
|
h, d = deltainfo.data
|
|
2547
|
if not h and not d:
|
|
2549
|
if not h and not d:
|
|
2548
|
# not data to store at all... declare them uncompressed
|
|
2550
|
# not data to store at all... declare them uncompressed
|
|
2549
|
compression_mode = COMP_MODE_PLAIN
|
|
2551
|
compression_mode = COMP_MODE_PLAIN
|
|
2550
|
elif not h:
|
|
2552
|
elif not h:
|
|
2551
|
t = d[0:1]
|
|
2553
|
t = d[0:1]
|
|
2552
|
if t == b'\0':
|
|
2554
|
if t == b'\0':
|
|
2553
|
compression_mode = COMP_MODE_PLAIN
|
|
2555
|
compression_mode = COMP_MODE_PLAIN
|
|
2554
|
elif t == self._docket.default_compression_header:
|
|
2556
|
elif t == self._docket.default_compression_header:
|
|
2555
|
compression_mode = COMP_MODE_DEFAULT
|
|
2557
|
compression_mode = COMP_MODE_DEFAULT
|
|
2556
|
elif h == b'u':
|
|
2558
|
elif h == b'u':
|
|
2557
|
# we have a more efficient way to declare uncompressed
|
|
2559
|
# we have a more efficient way to declare uncompressed
|
|
2558
|
h = b''
|
|
2560
|
h = b''
|
|
2559
|
compression_mode = COMP_MODE_PLAIN
|
|
2561
|
compression_mode = COMP_MODE_PLAIN
|
|
2560
|
deltainfo = deltautil.drop_u_compression(deltainfo)
|
|
2562
|
deltainfo = deltautil.drop_u_compression(deltainfo)
|
|
2561
|
|
|
2563
|
|
|
2562
|
sidedata_compression_mode = COMP_MODE_INLINE
|
|
2564
|
sidedata_compression_mode = COMP_MODE_INLINE
|
|
2563
|
if sidedata and self.hassidedata:
|
|
2565
|
if sidedata and self.hassidedata:
|
|
2564
|
sidedata_compression_mode = COMP_MODE_PLAIN
|
|
2566
|
sidedata_compression_mode = COMP_MODE_PLAIN
|
|
2565
|
serialized_sidedata = sidedatautil.serialize_sidedata(sidedata)
|
|
2567
|
serialized_sidedata = sidedatautil.serialize_sidedata(sidedata)
|
|
2566
|
sidedata_offset = offset + deltainfo.deltalen
|
|
2568
|
sidedata_offset = offset + deltainfo.deltalen
|
|
2567
|
h, comp_sidedata = self.compress(serialized_sidedata)
|
|
2569
|
h, comp_sidedata = self.compress(serialized_sidedata)
|
|
2568
|
if (
|
|
2570
|
if (
|
|
2569
|
h != b'u'
|
|
2571
|
h != b'u'
|
|
2570
|
and comp_sidedata[0:1] != b'\0'
|
|
2572
|
and comp_sidedata[0:1] != b'\0'
|
|
2571
|
and len(comp_sidedata) < len(serialized_sidedata)
|
|
2573
|
and len(comp_sidedata) < len(serialized_sidedata)
|
|
2572
|
):
|
|
2574
|
):
|
|
2573
|
assert not h
|
|
2575
|
assert not h
|
|
2574
|
if (
|
|
2576
|
if (
|
|
2575
|
comp_sidedata[0:1]
|
|
2577
|
comp_sidedata[0:1]
|
|
2576
|
== self._docket.default_compression_header
|
|
2578
|
== self._docket.default_compression_header
|
|
2577
|
):
|
|
2579
|
):
|
|
2578
|
sidedata_compression_mode = COMP_MODE_DEFAULT
|
|
2580
|
sidedata_compression_mode = COMP_MODE_DEFAULT
|
|
2579
|
serialized_sidedata = comp_sidedata
|
|
2581
|
serialized_sidedata = comp_sidedata
|
|
2580
|
else:
|
|
2582
|
else:
|
|
2581
|
sidedata_compression_mode = COMP_MODE_INLINE
|
|
2583
|
sidedata_compression_mode = COMP_MODE_INLINE
|
|
2582
|
serialized_sidedata = comp_sidedata
|
|
2584
|
serialized_sidedata = comp_sidedata
|
|
2583
|
else:
|
|
2585
|
else:
|
|
2584
|
serialized_sidedata = b""
|
|
2586
|
serialized_sidedata = b""
|
|
2585
|
# Don't store the offset if the sidedata is empty, that way
|
|
2587
|
# Don't store the offset if the sidedata is empty, that way
|
|
2586
|
# we can easily detect empty sidedata and they will be no different
|
|
2588
|
# we can easily detect empty sidedata and they will be no different
|
|
2587
|
# than ones we manually add.
|
|
2589
|
# than ones we manually add.
|
|
2588
|
sidedata_offset = 0
|
|
2590
|
sidedata_offset = 0
|
|
2589
|
|
|
2591
|
|
|
2590
|
e = (
|
|
2592
|
e = (
|
|
2591
|
offset_type(offset, flags),
|
|
2593
|
offset_type(offset, flags),
|
|
2592
|
deltainfo.deltalen,
|
|
2594
|
deltainfo.deltalen,
|
|
2593
|
textlen,
|
|
2595
|
textlen,
|
|
2594
|
deltainfo.base,
|
|
2596
|
deltainfo.base,
|
|
2595
|
link,
|
|
2597
|
link,
|
|
2596
|
p1r,
|
|
2598
|
p1r,
|
|
2597
|
p2r,
|
|
2599
|
p2r,
|
|
2598
|
node,
|
|
2600
|
node,
|
|
2599
|
sidedata_offset,
|
|
2601
|
sidedata_offset,
|
|
2600
|
len(serialized_sidedata),
|
|
2602
|
len(serialized_sidedata),
|
|
2601
|
compression_mode,
|
|
2603
|
compression_mode,
|
|
2602
|
sidedata_compression_mode,
|
|
2604
|
sidedata_compression_mode,
|
|
2603
|
)
|
|
2605
|
)
|
|
2604
|
|
|
2606
|
|
|
2605
|
self.index.append(e)
|
|
2607
|
self.index.append(e)
|
|
2606
|
entry = self.index.entry_binary(curr)
|
|
2608
|
entry = self.index.entry_binary(curr)
|
|
2607
|
if curr == 0 and self._docket is None:
|
|
2609
|
if curr == 0 and self._docket is None:
|
|
2608
|
header = self._format_flags | self._format_version
|
|
2610
|
header = self._format_flags | self._format_version
|
|
2609
|
header = self.index.pack_header(header)
|
|
2611
|
header = self.index.pack_header(header)
|
|
2610
|
entry = header + entry
|
|
2612
|
entry = header + entry
|
|
2611
|
self._writeentry(
|
|
2613
|
self._writeentry(
|
|
2612
|
transaction,
|
|
2614
|
transaction,
|
|
2613
|
entry,
|
|
2615
|
entry,
|
|
2614
|
deltainfo.data,
|
|
2616
|
deltainfo.data,
|
|
2615
|
link,
|
|
2617
|
link,
|
|
2616
|
offset,
|
|
2618
|
offset,
|
|
2617
|
serialized_sidedata,
|
|
2619
|
serialized_sidedata,
|
|
2618
|
)
|
|
2620
|
)
|
|
2619
|
|
|
2621
|
|
|
2620
|
rawtext = btext[0]
|
|
2622
|
rawtext = btext[0]
|
|
2621
|
|
|
2623
|
|
|
2622
|
if alwayscache and rawtext is None:
|
|
2624
|
if alwayscache and rawtext is None:
|
|
2623
|
rawtext = deltacomputer.buildtext(revinfo, fh)
|
|
2625
|
rawtext = deltacomputer.buildtext(revinfo, fh)
|
|
2624
|
|
|
2626
|
|
|
2625
|
if type(rawtext) == bytes: # only accept immutable objects
|
|
2627
|
if type(rawtext) == bytes: # only accept immutable objects
|
|
2626
|
self._revisioncache = (node, curr, rawtext)
|
|
2628
|
self._revisioncache = (node, curr, rawtext)
|
|
2627
|
self._chainbasecache[curr] = deltainfo.chainbase
|
|
2629
|
self._chainbasecache[curr] = deltainfo.chainbase
|
|
2628
|
return curr
|
|
2630
|
return curr
|
|
2629
|
|
|
2631
|
|
|
2630
|
def _get_data_offset(self, prev):
|
|
2632
|
def _get_data_offset(self, prev):
|
|
2631
|
"""Returns the current offset in the (in-transaction) data file.
|
|
2633
|
"""Returns the current offset in the (in-transaction) data file.
|
|
2632
|
Versions < 2 of the revlog can get this 0(1), revlog v2 needs a docket
|
|
2634
|
Versions < 2 of the revlog can get this 0(1), revlog v2 needs a docket
|
|
2633
|
file to store that information: since sidedata can be rewritten to the
|
|
2635
|
file to store that information: since sidedata can be rewritten to the
|
|
2634
|
end of the data file within a transaction, you can have cases where, for
|
|
2636
|
end of the data file within a transaction, you can have cases where, for
|
|
2635
|
example, rev `n` does not have sidedata while rev `n - 1` does, leading
|
|
2637
|
example, rev `n` does not have sidedata while rev `n - 1` does, leading
|
|
2636
|
to `n - 1`'s sidedata being written after `n`'s data.
|
|
2638
|
to `n - 1`'s sidedata being written after `n`'s data.
|
|
2637
|
|
|
2639
|
|
|
2638
|
TODO cache this in a docket file before getting out of experimental."""
|
|
2640
|
TODO cache this in a docket file before getting out of experimental."""
|
|
2639
|
if self._docket is None:
|
|
2641
|
if self._docket is None:
|
|
2640
|
return self.end(prev)
|
|
2642
|
return self.end(prev)
|
|
2641
|
else:
|
|
2643
|
else:
|
|
2642
|
return self._docket.data_end
|
|
2644
|
return self._docket.data_end
|
|
2643
|
|
|
2645
|
|
|
2644
|
def _writeentry(self, transaction, entry, data, link, offset, sidedata):
|
|
2646
|
def _writeentry(self, transaction, entry, data, link, offset, sidedata):
|
|
2645
|
# Files opened in a+ mode have inconsistent behavior on various
|
|
2647
|
# Files opened in a+ mode have inconsistent behavior on various
|
|
2646
|
# platforms. Windows requires that a file positioning call be made
|
|
2648
|
# platforms. Windows requires that a file positioning call be made
|
|
2647
|
# when the file handle transitions between reads and writes. See
|
|
2649
|
# when the file handle transitions between reads and writes. See
|
|
2648
|
# 3686fa2b8eee and the mixedfilemodewrapper in windows.py. On other
|
|
2650
|
# 3686fa2b8eee and the mixedfilemodewrapper in windows.py. On other
|
|
2649
|
# platforms, Python or the platform itself can be buggy. Some versions
|
|
2651
|
# platforms, Python or the platform itself can be buggy. Some versions
|
|
2650
|
# of Solaris have been observed to not append at the end of the file
|
|
2652
|
# of Solaris have been observed to not append at the end of the file
|
|
2651
|
# if the file was seeked to before the end. See issue4943 for more.
|
|
2653
|
# if the file was seeked to before the end. See issue4943 for more.
|
|
2652
|
#
|
|
2654
|
#
|
|
2653
|
# We work around this issue by inserting a seek() before writing.
|
|
2655
|
# We work around this issue by inserting a seek() before writing.
|
|
2654
|
# Note: This is likely not necessary on Python 3. However, because
|
|
2656
|
# Note: This is likely not necessary on Python 3. However, because
|
|
2655
|
# the file handle is reused for reads and may be seeked there, we need
|
|
2657
|
# the file handle is reused for reads and may be seeked there, we need
|
|
2656
|
# to be careful before changing this.
|
|
2658
|
# to be careful before changing this.
|
|
2657
|
if self._writinghandles is None:
|
|
2659
|
if self._writinghandles is None:
|
|
2658
|
msg = b'adding revision outside `revlog._writing` context'
|
|
2660
|
msg = b'adding revision outside `revlog._writing` context'
|
|
2659
|
raise error.ProgrammingError(msg)
|
|
2661
|
raise error.ProgrammingError(msg)
|
|
2660
|
ifh, dfh = self._writinghandles
|
|
2662
|
ifh, dfh = self._writinghandles
|
|
2661
|
if self._docket is None:
|
|
2663
|
if self._docket is None:
|
|
2662
|
ifh.seek(0, os.SEEK_END)
|
|
2664
|
ifh.seek(0, os.SEEK_END)
|
|
2663
|
else:
|
|
2665
|
else:
|
|
2664
|
ifh.seek(self._docket.index_end, os.SEEK_SET)
|
|
2666
|
ifh.seek(self._docket.index_end, os.SEEK_SET)
|
|
2665
|
if dfh:
|
|
2667
|
if dfh:
|
|
2666
|
if self._docket is None:
|
|
2668
|
if self._docket is None:
|
|
2667
|
dfh.seek(0, os.SEEK_END)
|
|
2669
|
dfh.seek(0, os.SEEK_END)
|
|
2668
|
else:
|
|
2670
|
else:
|
|
2669
|
dfh.seek(self._docket.data_end, os.SEEK_SET)
|
|
2671
|
dfh.seek(self._docket.data_end, os.SEEK_SET)
|
|
2670
|
|
|
2672
|
|
|
2671
|
curr = len(self) - 1
|
|
2673
|
curr = len(self) - 1
|
|
2672
|
if not self._inline:
|
|
2674
|
if not self._inline:
|
|
2673
|
transaction.add(self._datafile, offset)
|
|
2675
|
transaction.add(self._datafile, offset)
|
|
2674
|
transaction.add(self._indexfile, curr * len(entry))
|
|
2676
|
transaction.add(self._indexfile, curr * len(entry))
|
|
2675
|
if data[0]:
|
|
2677
|
if data[0]:
|
|
2676
|
dfh.write(data[0])
|
|
2678
|
dfh.write(data[0])
|
|
2677
|
dfh.write(data[1])
|
|
2679
|
dfh.write(data[1])
|
|
2678
|
if sidedata:
|
|
2680
|
if sidedata:
|
|
2679
|
dfh.write(sidedata)
|
|
2681
|
dfh.write(sidedata)
|
|
2680
|
ifh.write(entry)
|
|
2682
|
ifh.write(entry)
|
|
2681
|
else:
|
|
2683
|
else:
|
|
2682
|
offset += curr * self.index.entry_size
|
|
2684
|
offset += curr * self.index.entry_size
|
|
2683
|
transaction.add(self._indexfile, offset)
|
|
2685
|
transaction.add(self._indexfile, offset)
|
|
2684
|
ifh.write(entry)
|
|
2686
|
ifh.write(entry)
|
|
2685
|
ifh.write(data[0])
|
|
2687
|
ifh.write(data[0])
|
|
2686
|
ifh.write(data[1])
|
|
2688
|
ifh.write(data[1])
|
|
2687
|
if sidedata:
|
|
2689
|
if sidedata:
|
|
2688
|
ifh.write(sidedata)
|
|
2690
|
ifh.write(sidedata)
|
|
2689
|
self._enforceinlinesize(transaction)
|
|
2691
|
self._enforceinlinesize(transaction)
|
|
2690
|
if self._docket is not None:
|
|
2692
|
if self._docket is not None:
|
|
2691
|
self._docket.index_end = self._writinghandles[0].tell()
|
|
2693
|
self._docket.index_end = self._writinghandles[0].tell()
|
|
2692
|
self._docket.data_end = self._writinghandles[1].tell()
|
|
2694
|
self._docket.data_end = self._writinghandles[1].tell()
|
|
2693
|
|
|
2695
|
|
|
2694
|
nodemaputil.setup_persistent_nodemap(transaction, self)
|
|
2696
|
nodemaputil.setup_persistent_nodemap(transaction, self)
|
|
2695
|
|
|
2697
|
|
|
2696
|
def addgroup(
|
|
2698
|
def addgroup(
|
|
2697
|
self,
|
|
2699
|
self,
|
|
2698
|
deltas,
|
|
2700
|
deltas,
|
|
2699
|
linkmapper,
|
|
2701
|
linkmapper,
|
|
2700
|
transaction,
|
|
2702
|
transaction,
|
|
2701
|
alwayscache=False,
|
|
2703
|
alwayscache=False,
|
|
2702
|
addrevisioncb=None,
|
|
2704
|
addrevisioncb=None,
|
|
2703
|
duplicaterevisioncb=None,
|
|
2705
|
duplicaterevisioncb=None,
|
|
2704
|
):
|
|
2706
|
):
|
|
2705
|
"""
|
|
2707
|
"""
|
|
2706
|
add a delta group
|
|
2708
|
add a delta group
|
|
2707
|
|
|
2709
|
|
|
2708
|
given a set of deltas, add them to the revision log. the
|
|
2710
|
given a set of deltas, add them to the revision log. the
|
|
2709
|
first delta is against its parent, which should be in our
|
|
2711
|
first delta is against its parent, which should be in our
|
|
2710
|
log, the rest are against the previous delta.
|
|
2712
|
log, the rest are against the previous delta.
|
|
2711
|
|
|
2713
|
|
|
2712
|
If ``addrevisioncb`` is defined, it will be called with arguments of
|
|
2714
|
If ``addrevisioncb`` is defined, it will be called with arguments of
|
|
2713
|
this revlog and the node that was added.
|
|
2715
|
this revlog and the node that was added.
|
|
2714
|
"""
|
|
2716
|
"""
|
|
2715
|
|
|
2717
|
|
|
2716
|
if self._adding_group:
|
|
2718
|
if self._adding_group:
|
|
2717
|
raise error.ProgrammingError(b'cannot nest addgroup() calls')
|
|
2719
|
raise error.ProgrammingError(b'cannot nest addgroup() calls')
|
|
2718
|
|
|
2720
|
|
|
2719
|
self._adding_group = True
|
|
2721
|
self._adding_group = True
|
|
2720
|
empty = True
|
|
2722
|
empty = True
|
|
2721
|
try:
|
|
2723
|
try:
|
|
2722
|
with self._writing(transaction):
|
|
2724
|
with self._writing(transaction):
|
|
2723
|
deltacomputer = deltautil.deltacomputer(self)
|
|
2725
|
deltacomputer = deltautil.deltacomputer(self)
|
|
2724
|
# loop through our set of deltas
|
|
2726
|
# loop through our set of deltas
|
|
2725
|
for data in deltas:
|
|
2727
|
for data in deltas:
|
|
2726
|
(
|
|
2728
|
(
|
|
2727
|
node,
|
|
2729
|
node,
|
|
2728
|
p1,
|
|
2730
|
p1,
|
|
2729
|
p2,
|
|
2731
|
p2,
|
|
2730
|
linknode,
|
|
2732
|
linknode,
|
|
2731
|
deltabase,
|
|
2733
|
deltabase,
|
|
2732
|
delta,
|
|
2734
|
delta,
|
|
2733
|
flags,
|
|
2735
|
flags,
|
|
2734
|
sidedata,
|
|
2736
|
sidedata,
|
|
2735
|
) = data
|
|
2737
|
) = data
|
|
2736
|
link = linkmapper(linknode)
|
|
2738
|
link = linkmapper(linknode)
|
|
2737
|
flags = flags or REVIDX_DEFAULT_FLAGS
|
|
2739
|
flags = flags or REVIDX_DEFAULT_FLAGS
|
|
2738
|
|
|
2740
|
|
|
2739
|
rev = self.index.get_rev(node)
|
|
2741
|
rev = self.index.get_rev(node)
|
|
2740
|
if rev is not None:
|
|
2742
|
if rev is not None:
|
|
2741
|
# this can happen if two branches make the same change
|
|
2743
|
# this can happen if two branches make the same change
|
|
2742
|
self._nodeduplicatecallback(transaction, rev)
|
|
2744
|
self._nodeduplicatecallback(transaction, rev)
|
|
2743
|
if duplicaterevisioncb:
|
|
2745
|
if duplicaterevisioncb:
|
|
2744
|
duplicaterevisioncb(self, rev)
|
|
2746
|
duplicaterevisioncb(self, rev)
|
|
2745
|
empty = False
|
|
2747
|
empty = False
|
|
2746
|
continue
|
|
2748
|
continue
|
|
2747
|
|
|
2749
|
|
|
2748
|
for p in (p1, p2):
|
|
2750
|
for p in (p1, p2):
|
|
2749
|
if not self.index.has_node(p):
|
|
2751
|
if not self.index.has_node(p):
|
|
2750
|
raise error.LookupError(
|
|
2752
|
raise error.LookupError(
|
|
2751
|
p, self.radix, _(b'unknown parent')
|
|
2753
|
p, self.radix, _(b'unknown parent')
|
|
2752
|
)
|
|
2754
|
)
|
|
2753
|
|
|
2755
|
|
|
2754
|
if not self.index.has_node(deltabase):
|
|
2756
|
if not self.index.has_node(deltabase):
|
|
2755
|
raise error.LookupError(
|
|
2757
|
raise error.LookupError(
|
|
2756
|
deltabase, self.display_id, _(b'unknown delta base')
|
|
2758
|
deltabase, self.display_id, _(b'unknown delta base')
|
|
2757
|
)
|
|
2759
|
)
|
|
2758
|
|
|
2760
|
|
|
2759
|
baserev = self.rev(deltabase)
|
|
2761
|
baserev = self.rev(deltabase)
|
|
2760
|
|
|
2762
|
|
|
2761
|
if baserev != nullrev and self.iscensored(baserev):
|
|
2763
|
if baserev != nullrev and self.iscensored(baserev):
|
|
2762
|
# if base is censored, delta must be full replacement in a
|
|
2764
|
# if base is censored, delta must be full replacement in a
|
|
2763
|
# single patch operation
|
|
2765
|
# single patch operation
|
|
2764
|
hlen = struct.calcsize(b">lll")
|
|
2766
|
hlen = struct.calcsize(b">lll")
|
|
2765
|
oldlen = self.rawsize(baserev)
|
|
2767
|
oldlen = self.rawsize(baserev)
|
|
2766
|
newlen = len(delta) - hlen
|
|
2768
|
newlen = len(delta) - hlen
|
|
2767
|
if delta[:hlen] != mdiff.replacediffheader(
|
|
2769
|
if delta[:hlen] != mdiff.replacediffheader(
|
|
2768
|
oldlen, newlen
|
|
2770
|
oldlen, newlen
|
|
2769
|
):
|
|
2771
|
):
|
|
2770
|
raise error.CensoredBaseError(
|
|
2772
|
raise error.CensoredBaseError(
|
|
2771
|
self.display_id, self.node(baserev)
|
|
2773
|
self.display_id, self.node(baserev)
|
|
2772
|
)
|
|
2774
|
)
|
|
2773
|
|
|
2775
|
|
|
2774
|
if not flags and self._peek_iscensored(baserev, delta):
|
|
2776
|
if not flags and self._peek_iscensored(baserev, delta):
|
|
2775
|
flags |= REVIDX_ISCENSORED
|
|
2777
|
flags |= REVIDX_ISCENSORED
|
|
2776
|
|
|
2778
|
|
|
2777
|
# We assume consumers of addrevisioncb will want to retrieve
|
|
2779
|
# We assume consumers of addrevisioncb will want to retrieve
|
|
2778
|
# the added revision, which will require a call to
|
|
2780
|
# the added revision, which will require a call to
|
|
2779
|
# revision(). revision() will fast path if there is a cache
|
|
2781
|
# revision(). revision() will fast path if there is a cache
|
|
2780
|
# hit. So, we tell _addrevision() to always cache in this case.
|
|
2782
|
# hit. So, we tell _addrevision() to always cache in this case.
|
|
2781
|
# We're only using addgroup() in the context of changegroup
|
|
2783
|
# We're only using addgroup() in the context of changegroup
|
|
2782
|
# generation so the revision data can always be handled as raw
|
|
2784
|
# generation so the revision data can always be handled as raw
|
|
2783
|
# by the flagprocessor.
|
|
2785
|
# by the flagprocessor.
|
|
2784
|
rev = self._addrevision(
|
|
2786
|
rev = self._addrevision(
|
|
2785
|
node,
|
|
2787
|
node,
|
|
2786
|
None,
|
|
2788
|
None,
|
|
2787
|
transaction,
|
|
2789
|
transaction,
|
|
2788
|
link,
|
|
2790
|
link,
|
|
2789
|
p1,
|
|
2791
|
p1,
|
|
2790
|
p2,
|
|
2792
|
p2,
|
|
2791
|
flags,
|
|
2793
|
flags,
|
|
2792
|
(baserev, delta),
|
|
2794
|
(baserev, delta),
|
|
2793
|
alwayscache=alwayscache,
|
|
2795
|
alwayscache=alwayscache,
|
|
2794
|
deltacomputer=deltacomputer,
|
|
2796
|
deltacomputer=deltacomputer,
|
|
2795
|
sidedata=sidedata,
|
|
2797
|
sidedata=sidedata,
|
|
2796
|
)
|
|
2798
|
)
|
|
2797
|
|
|
2799
|
|
|
2798
|
if addrevisioncb:
|
|
2800
|
if addrevisioncb:
|
|
2799
|
addrevisioncb(self, rev)
|
|
2801
|
addrevisioncb(self, rev)
|
|
2800
|
empty = False
|
|
2802
|
empty = False
|
|
2801
|
finally:
|
|
2803
|
finally:
|
|
2802
|
self._adding_group = False
|
|
2804
|
self._adding_group = False
|
|
2803
|
return not empty
|
|
2805
|
return not empty
|
|
2804
|
|
|
2806
|
|
|
2805
|
def iscensored(self, rev):
|
|
2807
|
def iscensored(self, rev):
|
|
2806
|
"""Check if a file revision is censored."""
|
|
2808
|
"""Check if a file revision is censored."""
|
|
2807
|
if not self._censorable:
|
|
2809
|
if not self._censorable:
|
|
2808
|
return False
|
|
2810
|
return False
|
|
2809
|
|
|
2811
|
|
|
2810
|
return self.flags(rev) & REVIDX_ISCENSORED
|
|
2812
|
return self.flags(rev) & REVIDX_ISCENSORED
|
|
2811
|
|
|
2813
|
|
|
2812
|
def _peek_iscensored(self, baserev, delta):
|
|
2814
|
def _peek_iscensored(self, baserev, delta):
|
|
2813
|
"""Quickly check if a delta produces a censored revision."""
|
|
2815
|
"""Quickly check if a delta produces a censored revision."""
|
|
2814
|
if not self._censorable:
|
|
2816
|
if not self._censorable:
|
|
2815
|
return False
|
|
2817
|
return False
|
|
2816
|
|
|
2818
|
|
|
2817
|
return storageutil.deltaiscensored(delta, baserev, self.rawsize)
|
|
2819
|
return storageutil.deltaiscensored(delta, baserev, self.rawsize)
|
|
2818
|
|
|
2820
|
|
|
2819
|
def getstrippoint(self, minlink):
|
|
2821
|
def getstrippoint(self, minlink):
|
|
2820
|
"""find the minimum rev that must be stripped to strip the linkrev
|
|
2822
|
"""find the minimum rev that must be stripped to strip the linkrev
|
|
2821
|
|
|
2823
|
|
|
2822
|
Returns a tuple containing the minimum rev and a set of all revs that
|
|
2824
|
Returns a tuple containing the minimum rev and a set of all revs that
|
|
2823
|
have linkrevs that will be broken by this strip.
|
|
2825
|
have linkrevs that will be broken by this strip.
|
|
2824
|
"""
|
|
2826
|
"""
|
|
2825
|
return storageutil.resolvestripinfo(
|
|
2827
|
return storageutil.resolvestripinfo(
|
|
2826
|
minlink,
|
|
2828
|
minlink,
|
|
2827
|
len(self) - 1,
|
|
2829
|
len(self) - 1,
|
|
2828
|
self.headrevs(),
|
|
2830
|
self.headrevs(),
|
|
2829
|
self.linkrev,
|
|
2831
|
self.linkrev,
|
|
2830
|
self.parentrevs,
|
|
2832
|
self.parentrevs,
|
|
2831
|
)
|
|
2833
|
)
|
|
2832
|
|
|
2834
|
|
|
2833
|
def strip(self, minlink, transaction):
|
|
2835
|
def strip(self, minlink, transaction):
|
|
2834
|
"""truncate the revlog on the first revision with a linkrev >= minlink
|
|
2836
|
"""truncate the revlog on the first revision with a linkrev >= minlink
|
|
2835
|
|
|
2837
|
|
|
2836
|
This function is called when we're stripping revision minlink and
|
|
2838
|
This function is called when we're stripping revision minlink and
|
|
2837
|
its descendants from the repository.
|
|
2839
|
its descendants from the repository.
|
|
2838
|
|
|
2840
|
|
|
2839
|
We have to remove all revisions with linkrev >= minlink, because
|
|
2841
|
We have to remove all revisions with linkrev >= minlink, because
|
|
2840
|
the equivalent changelog revisions will be renumbered after the
|
|
2842
|
the equivalent changelog revisions will be renumbered after the
|
|
2841
|
strip.
|
|
2843
|
strip.
|
|
2842
|
|
|
2844
|
|
|
2843
|
So we truncate the revlog on the first of these revisions, and
|
|
2845
|
So we truncate the revlog on the first of these revisions, and
|
|
2844
|
trust that the caller has saved the revisions that shouldn't be
|
|
2846
|
trust that the caller has saved the revisions that shouldn't be
|
|
2845
|
removed and that it'll re-add them after this truncation.
|
|
2847
|
removed and that it'll re-add them after this truncation.
|
|
2846
|
"""
|
|
2848
|
"""
|
|
2847
|
if len(self) == 0:
|
|
2849
|
if len(self) == 0:
|
|
2848
|
return
|
|
2850
|
return
|
|
2849
|
|
|
2851
|
|
|
2850
|
rev, _ = self.getstrippoint(minlink)
|
|
2852
|
rev, _ = self.getstrippoint(minlink)
|
|
2851
|
if rev == len(self):
|
|
2853
|
if rev == len(self):
|
|
2852
|
return
|
|
2854
|
return
|
|
2853
|
|
|
2855
|
|
|
2854
|
# first truncate the files on disk
|
|
2856
|
# first truncate the files on disk
|
|
2855
|
data_end = self.start(rev)
|
|
2857
|
data_end = self.start(rev)
|
|
2856
|
if not self._inline:
|
|
2858
|
if not self._inline:
|
|
2857
|
transaction.add(self._datafile, data_end)
|
|
2859
|
transaction.add(self._datafile, data_end)
|
|
2858
|
end = rev * self.index.entry_size
|
|
2860
|
end = rev * self.index.entry_size
|
|
2859
|
else:
|
|
2861
|
else:
|
|
2860
|
end = data_end + (rev * self.index.entry_size)
|
|
2862
|
end = data_end + (rev * self.index.entry_size)
|
|
2861
|
|
|
2863
|
|
|
2862
|
transaction.add(self._indexfile, end)
|
|
2864
|
transaction.add(self._indexfile, end)
|
|
2863
|
if self._docket is not None:
|
|
2865
|
if self._docket is not None:
|
|
2864
|
# XXX we could, leverage the docket while stripping. However it is
|
|
2866
|
# XXX we could, leverage the docket while stripping. However it is
|
|
2865
|
# not powerfull enough at the time of this comment
|
|
2867
|
# not powerfull enough at the time of this comment
|
|
2866
|
self._docket.index_end = end
|
|
2868
|
self._docket.index_end = end
|
|
2867
|
self._docket.data_end = data_end
|
|
2869
|
self._docket.data_end = data_end
|
|
2868
|
self._docket.write(transaction, stripping=True)
|
|
2870
|
self._docket.write(transaction, stripping=True)
|
|
2869
|
|
|
2871
|
|
|
2870
|
# then reset internal state in memory to forget those revisions
|
|
2872
|
# then reset internal state in memory to forget those revisions
|
|
2871
|
self._revisioncache = None
|
|
2873
|
self._revisioncache = None
|
|
2872
|
self._chaininfocache = util.lrucachedict(500)
|
|
2874
|
self._chaininfocache = util.lrucachedict(500)
|
|
2873
|
self._chunkclear()
|
|
2875
|
self._chunkclear()
|
|
2874
|
|
|
2876
|
|
|
2875
|
del self.index[rev:-1]
|
|
2877
|
del self.index[rev:-1]
|
|
2876
|
|
|
2878
|
|
|
2877
|
def checksize(self):
|
|
2879
|
def checksize(self):
|
|
2878
|
"""Check size of index and data files
|
|
2880
|
"""Check size of index and data files
|
|
2879
|
|
|
2881
|
|
|
2880
|
return a (dd, di) tuple.
|
|
2882
|
return a (dd, di) tuple.
|
|
2881
|
- dd: extra bytes for the "data" file
|
|
2883
|
- dd: extra bytes for the "data" file
|
|
2882
|
- di: extra bytes for the "index" file
|
|
2884
|
- di: extra bytes for the "index" file
|
|
2883
|
|
|
2885
|
|
|
2884
|
A healthy revlog will return (0, 0).
|
|
2886
|
A healthy revlog will return (0, 0).
|
|
2885
|
"""
|
|
2887
|
"""
|
|
2886
|
expected = 0
|
|
2888
|
expected = 0
|
|
2887
|
if len(self):
|
|
2889
|
if len(self):
|
|
2888
|
expected = max(0, self.end(len(self) - 1))
|
|
2890
|
expected = max(0, self.end(len(self) - 1))
|
|
2889
|
|
|
2891
|
|
|
2890
|
try:
|
|
2892
|
try:
|
|
2891
|
with self._datafp() as f:
|
|
2893
|
with self._datafp() as f:
|
|
2892
|
f.seek(0, io.SEEK_END)
|
|
2894
|
f.seek(0, io.SEEK_END)
|
|
2893
|
actual = f.tell()
|
|
2895
|
actual = f.tell()
|
|
2894
|
dd = actual - expected
|
|
2896
|
dd = actual - expected
|
|
2895
|
except IOError as inst:
|
|
2897
|
except IOError as inst:
|
|
2896
|
if inst.errno != errno.ENOENT:
|
|
2898
|
if inst.errno != errno.ENOENT:
|
|
2897
|
raise
|
|
2899
|
raise
|
|
2898
|
dd = 0
|
|
2900
|
dd = 0
|
|
2899
|
|
|
2901
|
|
|
2900
|
try:
|
|
2902
|
try:
|
|
2901
|
f = self.opener(self._indexfile)
|
|
2903
|
f = self.opener(self._indexfile)
|
|
2902
|
f.seek(0, io.SEEK_END)
|
|
2904
|
f.seek(0, io.SEEK_END)
|
|
2903
|
actual = f.tell()
|
|
2905
|
actual = f.tell()
|
|
2904
|
f.close()
|
|
2906
|
f.close()
|
|
2905
|
s = self.index.entry_size
|
|
2907
|
s = self.index.entry_size
|
|
2906
|
i = max(0, actual // s)
|
|
2908
|
i = max(0, actual // s)
|
|
2907
|
di = actual - (i * s)
|
|
2909
|
di = actual - (i * s)
|
|
2908
|
if self._inline:
|
|
2910
|
if self._inline:
|
|
2909
|
databytes = 0
|
|
2911
|
databytes = 0
|
|
2910
|
for r in self:
|
|
2912
|
for r in self:
|
|
2911
|
databytes += max(0, self.length(r))
|
|
2913
|
databytes += max(0, self.length(r))
|
|
2912
|
dd = 0
|
|
2914
|
dd = 0
|
|
2913
|
di = actual - len(self) * s - databytes
|
|
2915
|
di = actual - len(self) * s - databytes
|
|
2914
|
except IOError as inst:
|
|
2916
|
except IOError as inst:
|
|
2915
|
if inst.errno != errno.ENOENT:
|
|
2917
|
if inst.errno != errno.ENOENT:
|
|
2916
|
raise
|
|
2918
|
raise
|
|
2917
|
di = 0
|
|
2919
|
di = 0
|
|
2918
|
|
|
2920
|
|
|
2919
|
return (dd, di)
|
|
2921
|
return (dd, di)
|
|
2920
|
|
|
2922
|
|
|
2921
|
def files(self):
|
|
2923
|
def files(self):
|
|
2922
|
res = [self._indexfile]
|
|
2924
|
res = [self._indexfile]
|
|
2923
|
if not self._inline:
|
|
2925
|
if not self._inline:
|
|
2924
|
res.append(self._datafile)
|
|
2926
|
res.append(self._datafile)
|
|
2925
|
return res
|
|
2927
|
return res
|
|
2926
|
|
|
2928
|
|
|
2927
|
def emitrevisions(
|
|
2929
|
def emitrevisions(
|
|
2928
|
self,
|
|
2930
|
self,
|
|
2929
|
nodes,
|
|
2931
|
nodes,
|
|
2930
|
nodesorder=None,
|
|
2932
|
nodesorder=None,
|
|
2931
|
revisiondata=False,
|
|
2933
|
revisiondata=False,
|
|
2932
|
assumehaveparentrevisions=False,
|
|
2934
|
assumehaveparentrevisions=False,
|
|
2933
|
deltamode=repository.CG_DELTAMODE_STD,
|
|
2935
|
deltamode=repository.CG_DELTAMODE_STD,
|
|
2934
|
sidedata_helpers=None,
|
|
2936
|
sidedata_helpers=None,
|
|
2935
|
):
|
|
2937
|
):
|
|
2936
|
if nodesorder not in (b'nodes', b'storage', b'linear', None):
|
|
2938
|
if nodesorder not in (b'nodes', b'storage', b'linear', None):
|
|
2937
|
raise error.ProgrammingError(
|
|
2939
|
raise error.ProgrammingError(
|
|
2938
|
b'unhandled value for nodesorder: %s' % nodesorder
|
|
2940
|
b'unhandled value for nodesorder: %s' % nodesorder
|
|
2939
|
)
|
|
2941
|
)
|
|
2940
|
|
|
2942
|
|
|
2941
|
if nodesorder is None and not self._generaldelta:
|
|
2943
|
if nodesorder is None and not self._generaldelta:
|
|
2942
|
nodesorder = b'storage'
|
|
2944
|
nodesorder = b'storage'
|
|
2943
|
|
|
2945
|
|
|
2944
|
if (
|
|
2946
|
if (
|
|
2945
|
not self._storedeltachains
|
|
2947
|
not self._storedeltachains
|
|
2946
|
and deltamode != repository.CG_DELTAMODE_PREV
|
|
2948
|
and deltamode != repository.CG_DELTAMODE_PREV
|
|
2947
|
):
|
|
2949
|
):
|
|
2948
|
deltamode = repository.CG_DELTAMODE_FULL
|
|
2950
|
deltamode = repository.CG_DELTAMODE_FULL
|
|
2949
|
|
|
2951
|
|
|
2950
|
return storageutil.emitrevisions(
|
|
2952
|
return storageutil.emitrevisions(
|
|
2951
|
self,
|
|
2953
|
self,
|
|
2952
|
nodes,
|
|
2954
|
nodes,
|
|
2953
|
nodesorder,
|
|
2955
|
nodesorder,
|
|
2954
|
revlogrevisiondelta,
|
|
2956
|
revlogrevisiondelta,
|
|
2955
|
deltaparentfn=self.deltaparent,
|
|
2957
|
deltaparentfn=self.deltaparent,
|
|
2956
|
candeltafn=self.candelta,
|
|
2958
|
candeltafn=self.candelta,
|
|
2957
|
rawsizefn=self.rawsize,
|
|
2959
|
rawsizefn=self.rawsize,
|
|
2958
|
revdifffn=self.revdiff,
|
|
2960
|
revdifffn=self.revdiff,
|
|
2959
|
flagsfn=self.flags,
|
|
2961
|
flagsfn=self.flags,
|
|
2960
|
deltamode=deltamode,
|
|
2962
|
deltamode=deltamode,
|
|
2961
|
revisiondata=revisiondata,
|
|
2963
|
revisiondata=revisiondata,
|
|
2962
|
assumehaveparentrevisions=assumehaveparentrevisions,
|
|
2964
|
assumehaveparentrevisions=assumehaveparentrevisions,
|
|
2963
|
sidedata_helpers=sidedata_helpers,
|
|
2965
|
sidedata_helpers=sidedata_helpers,
|
|
2964
|
)
|
|
2966
|
)
|
|
2965
|
|
|
2967
|
|
|
2966
|
DELTAREUSEALWAYS = b'always'
|
|
2968
|
DELTAREUSEALWAYS = b'always'
|
|
2967
|
DELTAREUSESAMEREVS = b'samerevs'
|
|
2969
|
DELTAREUSESAMEREVS = b'samerevs'
|
|
2968
|
DELTAREUSENEVER = b'never'
|
|
2970
|
DELTAREUSENEVER = b'never'
|
|
2969
|
|
|
2971
|
|
|
2970
|
DELTAREUSEFULLADD = b'fulladd'
|
|
2972
|
DELTAREUSEFULLADD = b'fulladd'
|
|
2971
|
|
|
2973
|
|
|
2972
|
DELTAREUSEALL = {b'always', b'samerevs', b'never', b'fulladd'}
|
|
2974
|
DELTAREUSEALL = {b'always', b'samerevs', b'never', b'fulladd'}
|
|
2973
|
|
|
2975
|
|
|
2974
|
def clone(
|
|
2976
|
def clone(
|
|
2975
|
self,
|
|
2977
|
self,
|
|
2976
|
tr,
|
|
2978
|
tr,
|
|
2977
|
destrevlog,
|
|
2979
|
destrevlog,
|
|
2978
|
addrevisioncb=None,
|
|
2980
|
addrevisioncb=None,
|
|
2979
|
deltareuse=DELTAREUSESAMEREVS,
|
|
2981
|
deltareuse=DELTAREUSESAMEREVS,
|
|
2980
|
forcedeltabothparents=None,
|
|
2982
|
forcedeltabothparents=None,
|
|
2981
|
sidedata_helpers=None,
|
|
2983
|
sidedata_helpers=None,
|
|
2982
|
):
|
|
2984
|
):
|
|
2983
|
"""Copy this revlog to another, possibly with format changes.
|
|
2985
|
"""Copy this revlog to another, possibly with format changes.
|
|
2984
|
|
|
2986
|
|
|
2985
|
The destination revlog will contain the same revisions and nodes.
|
|
2987
|
The destination revlog will contain the same revisions and nodes.
|
|
2986
|
However, it may not be bit-for-bit identical due to e.g. delta encoding
|
|
2988
|
However, it may not be bit-for-bit identical due to e.g. delta encoding
|
|
2987
|
differences.
|
|
2989
|
differences.
|
|
2988
|
|
|
2990
|
|
|
2989
|
The ``deltareuse`` argument control how deltas from the existing revlog
|
|
2991
|
The ``deltareuse`` argument control how deltas from the existing revlog
|
|
2990
|
are preserved in the destination revlog. The argument can have the
|
|
2992
|
are preserved in the destination revlog. The argument can have the
|
|
2991
|
following values:
|
|
2993
|
following values:
|
|
2992
|
|
|
2994
|
|
|
2993
|
DELTAREUSEALWAYS
|
|
2995
|
DELTAREUSEALWAYS
|
|
2994
|
Deltas will always be reused (if possible), even if the destination
|
|
2996
|
Deltas will always be reused (if possible), even if the destination
|
|
2995
|
revlog would not select the same revisions for the delta. This is the
|
|
2997
|
revlog would not select the same revisions for the delta. This is the
|
|
2996
|
fastest mode of operation.
|
|
2998
|
fastest mode of operation.
|
|
2997
|
DELTAREUSESAMEREVS
|
|
2999
|
DELTAREUSESAMEREVS
|
|
2998
|
Deltas will be reused if the destination revlog would pick the same
|
|
3000
|
Deltas will be reused if the destination revlog would pick the same
|
|
2999
|
revisions for the delta. This mode strikes a balance between speed
|
|
3001
|
revisions for the delta. This mode strikes a balance between speed
|
|
3000
|
and optimization.
|
|
3002
|
and optimization.
|
|
3001
|
DELTAREUSENEVER
|
|
3003
|
DELTAREUSENEVER
|
|
3002
|
Deltas will never be reused. This is the slowest mode of execution.
|
|
3004
|
Deltas will never be reused. This is the slowest mode of execution.
|
|
3003
|
This mode can be used to recompute deltas (e.g. if the diff/delta
|
|
3005
|
This mode can be used to recompute deltas (e.g. if the diff/delta
|
|
3004
|
algorithm changes).
|
|
3006
|
algorithm changes).
|
|
3005
|
DELTAREUSEFULLADD
|
|
3007
|
DELTAREUSEFULLADD
|
|
3006
|
Revision will be re-added as if their were new content. This is
|
|
3008
|
Revision will be re-added as if their were new content. This is
|
|
3007
|
slower than DELTAREUSEALWAYS but allow more mechanism to kicks in.
|
|
3009
|
slower than DELTAREUSEALWAYS but allow more mechanism to kicks in.
|
|
3008
|
eg: large file detection and handling.
|
|
3010
|
eg: large file detection and handling.
|
|
3009
|
|
|
3011
|
|
|
3010
|
Delta computation can be slow, so the choice of delta reuse policy can
|
|
3012
|
Delta computation can be slow, so the choice of delta reuse policy can
|
|
3011
|
significantly affect run time.
|
|
3013
|
significantly affect run time.
|
|
3012
|
|
|
3014
|
|
|
3013
|
The default policy (``DELTAREUSESAMEREVS``) strikes a balance between
|
|
3015
|
The default policy (``DELTAREUSESAMEREVS``) strikes a balance between
|
|
3014
|
two extremes. Deltas will be reused if they are appropriate. But if the
|
|
3016
|
two extremes. Deltas will be reused if they are appropriate. But if the
|
|
3015
|
delta could choose a better revision, it will do so. This means if you
|
|
3017
|
delta could choose a better revision, it will do so. This means if you
|
|
3016
|
are converting a non-generaldelta revlog to a generaldelta revlog,
|
|
3018
|
are converting a non-generaldelta revlog to a generaldelta revlog,
|
|
3017
|
deltas will be recomputed if the delta's parent isn't a parent of the
|
|
3019
|
deltas will be recomputed if the delta's parent isn't a parent of the
|
|
3018
|
revision.
|
|
3020
|
revision.
|
|
3019
|
|
|
3021
|
|
|
3020
|
In addition to the delta policy, the ``forcedeltabothparents``
|
|
3022
|
In addition to the delta policy, the ``forcedeltabothparents``
|
|
3021
|
argument controls whether to force compute deltas against both parents
|
|
3023
|
argument controls whether to force compute deltas against both parents
|
|
3022
|
for merges. By default, the current default is used.
|
|
3024
|
for merges. By default, the current default is used.
|
|
3023
|
|
|
3025
|
|
|
3024
|
See `revlogutil.sidedata.get_sidedata_helpers` for the doc on
|
|
3026
|
See `revlogutil.sidedata.get_sidedata_helpers` for the doc on
|
|
3025
|
`sidedata_helpers`.
|
|
3027
|
`sidedata_helpers`.
|
|
3026
|
"""
|
|
3028
|
"""
|
|
3027
|
if deltareuse not in self.DELTAREUSEALL:
|
|
3029
|
if deltareuse not in self.DELTAREUSEALL:
|
|
3028
|
raise ValueError(
|
|
3030
|
raise ValueError(
|
|
3029
|
_(b'value for deltareuse invalid: %s') % deltareuse
|
|
3031
|
_(b'value for deltareuse invalid: %s') % deltareuse
|
|
3030
|
)
|
|
3032
|
)
|
|
3031
|
|
|
3033
|
|
|
3032
|
if len(destrevlog):
|
|
3034
|
if len(destrevlog):
|
|
3033
|
raise ValueError(_(b'destination revlog is not empty'))
|
|
3035
|
raise ValueError(_(b'destination revlog is not empty'))
|
|
3034
|
|
|
3036
|
|
|
3035
|
if getattr(self, 'filteredrevs', None):
|
|
3037
|
if getattr(self, 'filteredrevs', None):
|
|
3036
|
raise ValueError(_(b'source revlog has filtered revisions'))
|
|
3038
|
raise ValueError(_(b'source revlog has filtered revisions'))
|
|
3037
|
if getattr(destrevlog, 'filteredrevs', None):
|
|
3039
|
if getattr(destrevlog, 'filteredrevs', None):
|
|
3038
|
raise ValueError(_(b'destination revlog has filtered revisions'))
|
|
3040
|
raise ValueError(_(b'destination revlog has filtered revisions'))
|
|
3039
|
|
|
3041
|
|
|
3040
|
# lazydelta and lazydeltabase controls whether to reuse a cached delta,
|
|
3042
|
# lazydelta and lazydeltabase controls whether to reuse a cached delta,
|
|
3041
|
# if possible.
|
|
3043
|
# if possible.
|
|
3042
|
oldlazydelta = destrevlog._lazydelta
|
|
3044
|
oldlazydelta = destrevlog._lazydelta
|
|
3043
|
oldlazydeltabase = destrevlog._lazydeltabase
|
|
3045
|
oldlazydeltabase = destrevlog._lazydeltabase
|
|
3044
|
oldamd = destrevlog._deltabothparents
|
|
3046
|
oldamd = destrevlog._deltabothparents
|
|
3045
|
|
|
3047
|
|
|
3046
|
try:
|
|
3048
|
try:
|
|
3047
|
if deltareuse == self.DELTAREUSEALWAYS:
|
|
3049
|
if deltareuse == self.DELTAREUSEALWAYS:
|
|
3048
|
destrevlog._lazydeltabase = True
|
|
3050
|
destrevlog._lazydeltabase = True
|
|
3049
|
destrevlog._lazydelta = True
|
|
3051
|
destrevlog._lazydelta = True
|
|
3050
|
elif deltareuse == self.DELTAREUSESAMEREVS:
|
|
3052
|
elif deltareuse == self.DELTAREUSESAMEREVS:
|
|
3051
|
destrevlog._lazydeltabase = False
|
|
3053
|
destrevlog._lazydeltabase = False
|
|
3052
|
destrevlog._lazydelta = True
|
|
3054
|
destrevlog._lazydelta = True
|
|
3053
|
elif deltareuse == self.DELTAREUSENEVER:
|
|
3055
|
elif deltareuse == self.DELTAREUSENEVER:
|
|
3054
|
destrevlog._lazydeltabase = False
|
|
3056
|
destrevlog._lazydeltabase = False
|
|
3055
|
destrevlog._lazydelta = False
|
|
3057
|
destrevlog._lazydelta = False
|
|
3056
|
|
|
3058
|
|
|
3057
|
destrevlog._deltabothparents = forcedeltabothparents or oldamd
|
|
3059
|
destrevlog._deltabothparents = forcedeltabothparents or oldamd
|
|
3058
|
|
|
3060
|
|
|
3059
|
self._clone(
|
|
3061
|
self._clone(
|
|
3060
|
tr,
|
|
3062
|
tr,
|
|
3061
|
destrevlog,
|
|
3063
|
destrevlog,
|
|
3062
|
addrevisioncb,
|
|
3064
|
addrevisioncb,
|
|
3063
|
deltareuse,
|
|
3065
|
deltareuse,
|
|
3064
|
forcedeltabothparents,
|
|
3066
|
forcedeltabothparents,
|
|
3065
|
sidedata_helpers,
|
|
3067
|
sidedata_helpers,
|
|
3066
|
)
|
|
3068
|
)
|
|
3067
|
|
|
3069
|
|
|
3068
|
finally:
|
|
3070
|
finally:
|
|
3069
|
destrevlog._lazydelta = oldlazydelta
|
|
3071
|
destrevlog._lazydelta = oldlazydelta
|
|
3070
|
destrevlog._lazydeltabase = oldlazydeltabase
|
|
3072
|
destrevlog._lazydeltabase = oldlazydeltabase
|
|
3071
|
destrevlog._deltabothparents = oldamd
|
|
3073
|
destrevlog._deltabothparents = oldamd
|
|
3072
|
|
|
3074
|
|
|
3073
|
def _clone(
|
|
3075
|
def _clone(
|
|
3074
|
self,
|
|
3076
|
self,
|
|
3075
|
tr,
|
|
3077
|
tr,
|
|
3076
|
destrevlog,
|
|
3078
|
destrevlog,
|
|
3077
|
addrevisioncb,
|
|
3079
|
addrevisioncb,
|
|
3078
|
deltareuse,
|
|
3080
|
deltareuse,
|
|
3079
|
forcedeltabothparents,
|
|
3081
|
forcedeltabothparents,
|
|
3080
|
sidedata_helpers,
|
|
3082
|
sidedata_helpers,
|
|
3081
|
):
|
|
3083
|
):
|
|
3082
|
"""perform the core duty of `revlog.clone` after parameter processing"""
|
|
3084
|
"""perform the core duty of `revlog.clone` after parameter processing"""
|
|
3083
|
deltacomputer = deltautil.deltacomputer(destrevlog)
|
|
3085
|
deltacomputer = deltautil.deltacomputer(destrevlog)
|
|
3084
|
index = self.index
|
|
3086
|
index = self.index
|
|
3085
|
for rev in self:
|
|
3087
|
for rev in self:
|
|
3086
|
entry = index[rev]
|
|
3088
|
entry = index[rev]
|
|
3087
|
|
|
3089
|
|
|
3088
|
# Some classes override linkrev to take filtered revs into
|
|
3090
|
# Some classes override linkrev to take filtered revs into
|
|
3089
|
# account. Use raw entry from index.
|
|
3091
|
# account. Use raw entry from index.
|
|
3090
|
flags = entry[0] & 0xFFFF
|
|
3092
|
flags = entry[0] & 0xFFFF
|
|
3091
|
linkrev = entry[4]
|
|
3093
|
linkrev = entry[4]
|
|
3092
|
p1 = index[entry[5]][7]
|
|
3094
|
p1 = index[entry[5]][7]
|
|
3093
|
p2 = index[entry[6]][7]
|
|
3095
|
p2 = index[entry[6]][7]
|
|
3094
|
node = entry[7]
|
|
3096
|
node = entry[7]
|
|
3095
|
|
|
3097
|
|
|
3096
|
# (Possibly) reuse the delta from the revlog if allowed and
|
|
3098
|
# (Possibly) reuse the delta from the revlog if allowed and
|
|
3097
|
# the revlog chunk is a delta.
|
|
3099
|
# the revlog chunk is a delta.
|
|
3098
|
cachedelta = None
|
|
3100
|
cachedelta = None
|
|
3099
|
rawtext = None
|
|
3101
|
rawtext = None
|
|
3100
|
if deltareuse == self.DELTAREUSEFULLADD:
|
|
3102
|
if deltareuse == self.DELTAREUSEFULLADD:
|
|
3101
|
text, sidedata = self._revisiondata(rev)
|
|
3103
|
text, sidedata = self._revisiondata(rev)
|
|
3102
|
|
|
3104
|
|
|
3103
|
if sidedata_helpers is not None:
|
|
3105
|
if sidedata_helpers is not None:
|
|
3104
|
(sidedata, new_flags) = sidedatautil.run_sidedata_helpers(
|
|
3106
|
(sidedata, new_flags) = sidedatautil.run_sidedata_helpers(
|
|
3105
|
self, sidedata_helpers, sidedata, rev
|
|
3107
|
self, sidedata_helpers, sidedata, rev
|
|
3106
|
)
|
|
3108
|
)
|
|
3107
|
flags = flags | new_flags[0] & ~new_flags[1]
|
|
3109
|
flags = flags | new_flags[0] & ~new_flags[1]
|
|
3108
|
|
|
3110
|
|
|
3109
|
destrevlog.addrevision(
|
|
3111
|
destrevlog.addrevision(
|
|
3110
|
text,
|
|
3112
|
text,
|
|
3111
|
tr,
|
|
3113
|
tr,
|
|
3112
|
linkrev,
|
|
3114
|
linkrev,
|
|
3113
|
p1,
|
|
3115
|
p1,
|
|
3114
|
p2,
|
|
3116
|
p2,
|
|
3115
|
cachedelta=cachedelta,
|
|
3117
|
cachedelta=cachedelta,
|
|
3116
|
node=node,
|
|
3118
|
node=node,
|
|
3117
|
flags=flags,
|
|
3119
|
flags=flags,
|
|
3118
|
deltacomputer=deltacomputer,
|
|
3120
|
deltacomputer=deltacomputer,
|
|
3119
|
sidedata=sidedata,
|
|
3121
|
sidedata=sidedata,
|
|
3120
|
)
|
|
3122
|
)
|
|
3121
|
else:
|
|
3123
|
else:
|
|
3122
|
if destrevlog._lazydelta:
|
|
3124
|
if destrevlog._lazydelta:
|
|
3123
|
dp = self.deltaparent(rev)
|
|
3125
|
dp = self.deltaparent(rev)
|
|
3124
|
if dp != nullrev:
|
|
3126
|
if dp != nullrev:
|
|
3125
|
cachedelta = (dp, bytes(self._chunk(rev)))
|
|
3127
|
cachedelta = (dp, bytes(self._chunk(rev)))
|
|
3126
|
|
|
3128
|
|
|
3127
|
sidedata = None
|
|
3129
|
sidedata = None
|
|
3128
|
if not cachedelta:
|
|
3130
|
if not cachedelta:
|
|
3129
|
rawtext, sidedata = self._revisiondata(rev)
|
|
3131
|
rawtext, sidedata = self._revisiondata(rev)
|
|
3130
|
if sidedata is None:
|
|
3132
|
if sidedata is None:
|
|
3131
|
sidedata = self.sidedata(rev)
|
|
3133
|
sidedata = self.sidedata(rev)
|
|
3132
|
|
|
3134
|
|
|
3133
|
if sidedata_helpers is not None:
|
|
3135
|
if sidedata_helpers is not None:
|
|
3134
|
(sidedata, new_flags) = sidedatautil.run_sidedata_helpers(
|
|
3136
|
(sidedata, new_flags) = sidedatautil.run_sidedata_helpers(
|
|
3135
|
self, sidedata_helpers, sidedata, rev
|
|
3137
|
self, sidedata_helpers, sidedata, rev
|
|
3136
|
)
|
|
3138
|
)
|
|
3137
|
flags = flags | new_flags[0] & ~new_flags[1]
|
|
3139
|
flags = flags | new_flags[0] & ~new_flags[1]
|
|
3138
|
|
|
3140
|
|
|
3139
|
with destrevlog._writing(tr):
|
|
3141
|
with destrevlog._writing(tr):
|
|
3140
|
destrevlog._addrevision(
|
|
3142
|
destrevlog._addrevision(
|
|
3141
|
node,
|
|
3143
|
node,
|
|
3142
|
rawtext,
|
|
3144
|
rawtext,
|
|
3143
|
tr,
|
|
3145
|
tr,
|
|
3144
|
linkrev,
|
|
3146
|
linkrev,
|
|
3145
|
p1,
|
|
3147
|
p1,
|
|
3146
|
p2,
|
|
3148
|
p2,
|
|
3147
|
flags,
|
|
3149
|
flags,
|
|
3148
|
cachedelta,
|
|
3150
|
cachedelta,
|
|
3149
|
deltacomputer=deltacomputer,
|
|
3151
|
deltacomputer=deltacomputer,
|
|
3150
|
sidedata=sidedata,
|
|
3152
|
sidedata=sidedata,
|
|
3151
|
)
|
|
3153
|
)
|
|
3152
|
|
|
3154
|
|
|
3153
|
if addrevisioncb:
|
|
3155
|
if addrevisioncb:
|
|
3154
|
addrevisioncb(self, rev, node)
|
|
3156
|
addrevisioncb(self, rev, node)
|
|
3155
|
|
|
3157
|
|
|
3156
|
def censorrevision(self, tr, censornode, tombstone=b''):
|
|
3158
|
def censorrevision(self, tr, censornode, tombstone=b''):
|
|
3157
|
if self._format_version == REVLOGV0:
|
|
3159
|
if self._format_version == REVLOGV0:
|
|
3158
|
raise error.RevlogError(
|
|
3160
|
raise error.RevlogError(
|
|
3159
|
_(b'cannot censor with version %d revlogs')
|
|
3161
|
_(b'cannot censor with version %d revlogs')
|
|
3160
|
% self._format_version
|
|
3162
|
% self._format_version
|
|
3161
|
)
|
|
3163
|
)
|
|
3162
|
|
|
3164
|
|
|
3163
|
censorrev = self.rev(censornode)
|
|
3165
|
censorrev = self.rev(censornode)
|
|
3164
|
tombstone = storageutil.packmeta({b'censored': tombstone}, b'')
|
|
3166
|
tombstone = storageutil.packmeta({b'censored': tombstone}, b'')
|
|
3165
|
|
|
3167
|
|
|
3166
|
if len(tombstone) > self.rawsize(censorrev):
|
|
3168
|
if len(tombstone) > self.rawsize(censorrev):
|
|
3167
|
raise error.Abort(
|
|
3169
|
raise error.Abort(
|
|
3168
|
_(b'censor tombstone must be no longer than censored data')
|
|
3170
|
_(b'censor tombstone must be no longer than censored data')
|
|
3169
|
)
|
|
3171
|
)
|
|
3170
|
|
|
3172
|
|
|
3171
|
# Rewriting the revlog in place is hard. Our strategy for censoring is
|
|
3173
|
# Rewriting the revlog in place is hard. Our strategy for censoring is
|
|
3172
|
# to create a new revlog, copy all revisions to it, then replace the
|
|
3174
|
# to create a new revlog, copy all revisions to it, then replace the
|
|
3173
|
# revlogs on transaction close.
|
|
3175
|
# revlogs on transaction close.
|
|
3174
|
#
|
|
3176
|
#
|
|
3175
|
# This is a bit dangerous. We could easily have a mismatch of state.
|
|
3177
|
# This is a bit dangerous. We could easily have a mismatch of state.
|
|
3176
|
newrl = revlog(
|
|
3178
|
newrl = revlog(
|
|
3177
|
self.opener,
|
|
3179
|
self.opener,
|
|
3178
|
target=self.target,
|
|
3180
|
target=self.target,
|
|
3179
|
radix=self.radix,
|
|
3181
|
radix=self.radix,
|
|
3180
|
postfix=b'tmpcensored',
|
|
3182
|
postfix=b'tmpcensored',
|
|
3181
|
censorable=True,
|
|
3183
|
censorable=True,
|
|
3182
|
)
|
|
3184
|
)
|
|
3183
|
newrl._format_version = self._format_version
|
|
3185
|
newrl._format_version = self._format_version
|
|
3184
|
newrl._format_flags = self._format_flags
|
|
3186
|
newrl._format_flags = self._format_flags
|
|
3185
|
newrl._generaldelta = self._generaldelta
|
|
3187
|
newrl._generaldelta = self._generaldelta
|
|
3186
|
newrl._parse_index = self._parse_index
|
|
3188
|
newrl._parse_index = self._parse_index
|
|
3187
|
|
|
3189
|
|
|
3188
|
for rev in self.revs():
|
|
3190
|
for rev in self.revs():
|
|
3189
|
node = self.node(rev)
|
|
3191
|
node = self.node(rev)
|
|
3190
|
p1, p2 = self.parents(node)
|
|
3192
|
p1, p2 = self.parents(node)
|
|
3191
|
|
|
3193
|
|
|
3192
|
if rev == censorrev:
|
|
3194
|
if rev == censorrev:
|
|
3193
|
newrl.addrawrevision(
|
|
3195
|
newrl.addrawrevision(
|
|
3194
|
tombstone,
|
|
3196
|
tombstone,
|
|
3195
|
tr,
|
|
3197
|
tr,
|
|
3196
|
self.linkrev(censorrev),
|
|
3198
|
self.linkrev(censorrev),
|
|
3197
|
p1,
|
|
3199
|
p1,
|
|
3198
|
p2,
|
|
3200
|
p2,
|
|
3199
|
censornode,
|
|
3201
|
censornode,
|
|
3200
|
REVIDX_ISCENSORED,
|
|
3202
|
REVIDX_ISCENSORED,
|
|
3201
|
)
|
|
3203
|
)
|
|
3202
|
|
|
3204
|
|
|
3203
|
if newrl.deltaparent(rev) != nullrev:
|
|
3205
|
if newrl.deltaparent(rev) != nullrev:
|
|
3204
|
raise error.Abort(
|
|
3206
|
raise error.Abort(
|
|
3205
|
_(
|
|
3207
|
_(
|
|
3206
|
b'censored revision stored as delta; '
|
|
3208
|
b'censored revision stored as delta; '
|
|
3207
|
b'cannot censor'
|
|
3209
|
b'cannot censor'
|
|
3208
|
),
|
|
3210
|
),
|
|
3209
|
hint=_(
|
|
3211
|
hint=_(
|
|
3210
|
b'censoring of revlogs is not '
|
|
3212
|
b'censoring of revlogs is not '
|
|
3211
|
b'fully implemented; please report '
|
|
3213
|
b'fully implemented; please report '
|
|
3212
|
b'this bug'
|
|
3214
|
b'this bug'
|
|
3213
|
),
|
|
3215
|
),
|
|
3214
|
)
|
|
3216
|
)
|
|
3215
|
continue
|
|
3217
|
continue
|
|
3216
|
|
|
3218
|
|
|
3217
|
if self.iscensored(rev):
|
|
3219
|
if self.iscensored(rev):
|
|
3218
|
if self.deltaparent(rev) != nullrev:
|
|
3220
|
if self.deltaparent(rev) != nullrev:
|
|
3219
|
raise error.Abort(
|
|
3221
|
raise error.Abort(
|
|
3220
|
_(
|
|
3222
|
_(
|
|
3221
|
b'cannot censor due to censored '
|
|
3223
|
b'cannot censor due to censored '
|
|
3222
|
b'revision having delta stored'
|
|
3224
|
b'revision having delta stored'
|
|
3223
|
)
|
|
3225
|
)
|
|
3224
|
)
|
|
3226
|
)
|
|
3225
|
rawtext = self._chunk(rev)
|
|
3227
|
rawtext = self._chunk(rev)
|
|
3226
|
else:
|
|
3228
|
else:
|
|
3227
|
rawtext = self.rawdata(rev)
|
|
3229
|
rawtext = self.rawdata(rev)
|
|
3228
|
|
|
3230
|
|
|
3229
|
newrl.addrawrevision(
|
|
3231
|
newrl.addrawrevision(
|
|
3230
|
rawtext, tr, self.linkrev(rev), p1, p2, node, self.flags(rev)
|
|
3232
|
rawtext, tr, self.linkrev(rev), p1, p2, node, self.flags(rev)
|
|
3231
|
)
|
|
3233
|
)
|
|
3232
|
|
|
3234
|
|
|
3233
|
tr.addbackup(self._indexfile, location=b'store')
|
|
3235
|
tr.addbackup(self._indexfile, location=b'store')
|
|
3234
|
if not self._inline:
|
|
3236
|
if not self._inline:
|
|
3235
|
tr.addbackup(self._datafile, location=b'store')
|
|
3237
|
tr.addbackup(self._datafile, location=b'store')
|
|
3236
|
|
|
3238
|
|
|
3237
|
self.opener.rename(newrl._indexfile, self._indexfile)
|
|
3239
|
self.opener.rename(newrl._indexfile, self._indexfile)
|
|
3238
|
if not self._inline:
|
|
3240
|
if not self._inline:
|
|
3239
|
self.opener.rename(newrl._datafile, self._datafile)
|
|
3241
|
self.opener.rename(newrl._datafile, self._datafile)
|
|
3240
|
|
|
3242
|
|
|
3241
|
self.clearcaches()
|
|
3243
|
self.clearcaches()
|
|
3242
|
self._loadindex()
|
|
3244
|
self._loadindex()
|
|
3243
|
|
|
3245
|
|
|
3244
|
def verifyintegrity(self, state):
|
|
3246
|
def verifyintegrity(self, state):
|
|
3245
|
"""Verifies the integrity of the revlog.
|
|
3247
|
"""Verifies the integrity of the revlog.
|
|
3246
|
|
|
3248
|
|
|
3247
|
Yields ``revlogproblem`` instances describing problems that are
|
|
3249
|
Yields ``revlogproblem`` instances describing problems that are
|
|
3248
|
found.
|
|
3250
|
found.
|
|
3249
|
"""
|
|
3251
|
"""
|
|
3250
|
dd, di = self.checksize()
|
|
3252
|
dd, di = self.checksize()
|
|
3251
|
if dd:
|
|
3253
|
if dd:
|
|
3252
|
yield revlogproblem(error=_(b'data length off by %d bytes') % dd)
|
|
3254
|
yield revlogproblem(error=_(b'data length off by %d bytes') % dd)
|
|
3253
|
if di:
|
|
3255
|
if di:
|
|
3254
|
yield revlogproblem(error=_(b'index contains %d extra bytes') % di)
|
|
3256
|
yield revlogproblem(error=_(b'index contains %d extra bytes') % di)
|
|
3255
|
|
|
3257
|
|
|
3256
|
version = self._format_version
|
|
3258
|
version = self._format_version
|
|
3257
|
|
|
3259
|
|
|
3258
|
# The verifier tells us what version revlog we should be.
|
|
3260
|
# The verifier tells us what version revlog we should be.
|
|
3259
|
if version != state[b'expectedversion']:
|
|
3261
|
if version != state[b'expectedversion']:
|
|
3260
|
yield revlogproblem(
|
|
3262
|
yield revlogproblem(
|
|
3261
|
warning=_(b"warning: '%s' uses revlog format %d; expected %d")
|
|
3263
|
warning=_(b"warning: '%s' uses revlog format %d; expected %d")
|
|
3262
|
% (self.display_id, version, state[b'expectedversion'])
|
|
3264
|
% (self.display_id, version, state[b'expectedversion'])
|
|
3263
|
)
|
|
3265
|
)
|
|
3264
|
|
|
3266
|
|
|
3265
|
state[b'skipread'] = set()
|
|
3267
|
state[b'skipread'] = set()
|
|
3266
|
state[b'safe_renamed'] = set()
|
|
3268
|
state[b'safe_renamed'] = set()
|
|
3267
|
|
|
3269
|
|
|
3268
|
for rev in self:
|
|
3270
|
for rev in self:
|
|
3269
|
node = self.node(rev)
|
|
3271
|
node = self.node(rev)
|
|
3270
|
|
|
3272
|
|
|
3271
|
# Verify contents. 4 cases to care about:
|
|
3273
|
# Verify contents. 4 cases to care about:
|
|
3272
|
#
|
|
3274
|
#
|
|
3273
|
# common: the most common case
|
|
3275
|
# common: the most common case
|
|
3274
|
# rename: with a rename
|
|
3276
|
# rename: with a rename
|
|
3275
|
# meta: file content starts with b'\1\n', the metadata
|
|
3277
|
# meta: file content starts with b'\1\n', the metadata
|
|
3276
|
# header defined in filelog.py, but without a rename
|
|
3278
|
# header defined in filelog.py, but without a rename
|
|
3277
|
# ext: content stored externally
|
|
3279
|
# ext: content stored externally
|
|
3278
|
#
|
|
3280
|
#
|
|
3279
|
# More formally, their differences are shown below:
|
|
3281
|
# More formally, their differences are shown below:
|
|
3280
|
#
|
|
3282
|
#
|
|
3281
|
# | common | rename | meta | ext
|
|
3283
|
# | common | rename | meta | ext
|
|
3282
|
# -------------------------------------------------------
|
|
3284
|
# -------------------------------------------------------
|
|
3283
|
# flags() | 0 | 0 | 0 | not 0
|
|
3285
|
# flags() | 0 | 0 | 0 | not 0
|
|
3284
|
# renamed() | False | True | False | ?
|
|
3286
|
# renamed() | False | True | False | ?
|
|
3285
|
# rawtext[0:2]=='\1\n'| False | True | True | ?
|
|
3287
|
# rawtext[0:2]=='\1\n'| False | True | True | ?
|
|
3286
|
#
|
|
3288
|
#
|
|
3287
|
# "rawtext" means the raw text stored in revlog data, which
|
|
3289
|
# "rawtext" means the raw text stored in revlog data, which
|
|
3288
|
# could be retrieved by "rawdata(rev)". "text"
|
|
3290
|
# could be retrieved by "rawdata(rev)". "text"
|
|
3289
|
# mentioned below is "revision(rev)".
|
|
3291
|
# mentioned below is "revision(rev)".
|
|
3290
|
#
|
|
3292
|
#
|
|
3291
|
# There are 3 different lengths stored physically:
|
|
3293
|
# There are 3 different lengths stored physically:
|
|
3292
|
# 1. L1: rawsize, stored in revlog index
|
|
3294
|
# 1. L1: rawsize, stored in revlog index
|
|
3293
|
# 2. L2: len(rawtext), stored in revlog data
|
|
3295
|
# 2. L2: len(rawtext), stored in revlog data
|
|
3294
|
# 3. L3: len(text), stored in revlog data if flags==0, or
|
|
3296
|
# 3. L3: len(text), stored in revlog data if flags==0, or
|
|
3295
|
# possibly somewhere else if flags!=0
|
|
3297
|
# possibly somewhere else if flags!=0
|
|
3296
|
#
|
|
3298
|
#
|
|
3297
|
# L1 should be equal to L2. L3 could be different from them.
|
|
3299
|
# L1 should be equal to L2. L3 could be different from them.
|
|
3298
|
# "text" may or may not affect commit hash depending on flag
|
|
3300
|
# "text" may or may not affect commit hash depending on flag
|
|
3299
|
# processors (see flagutil.addflagprocessor).
|
|
3301
|
# processors (see flagutil.addflagprocessor).
|
|
3300
|
#
|
|
3302
|
#
|
|
3301
|
# | common | rename | meta | ext
|
|
3303
|
# | common | rename | meta | ext
|
|
3302
|
# -------------------------------------------------
|
|
3304
|
# -------------------------------------------------
|
|
3303
|
# rawsize() | L1 | L1 | L1 | L1
|
|
3305
|
# rawsize() | L1 | L1 | L1 | L1
|
|
3304
|
# size() | L1 | L2-LM | L1(*) | L1 (?)
|
|
3306
|
# size() | L1 | L2-LM | L1(*) | L1 (?)
|
|
3305
|
# len(rawtext) | L2 | L2 | L2 | L2
|
|
3307
|
# len(rawtext) | L2 | L2 | L2 | L2
|
|
3306
|
# len(text) | L2 | L2 | L2 | L3
|
|
3308
|
# len(text) | L2 | L2 | L2 | L3
|
|
3307
|
# len(read()) | L2 | L2-LM | L2-LM | L3 (?)
|
|
3309
|
# len(read()) | L2 | L2-LM | L2-LM | L3 (?)
|
|
3308
|
#
|
|
3310
|
#
|
|
3309
|
# LM: length of metadata, depending on rawtext
|
|
3311
|
# LM: length of metadata, depending on rawtext
|
|
3310
|
# (*): not ideal, see comment in filelog.size
|
|
3312
|
# (*): not ideal, see comment in filelog.size
|
|
3311
|
# (?): could be "- len(meta)" if the resolved content has
|
|
3313
|
# (?): could be "- len(meta)" if the resolved content has
|
|
3312
|
# rename metadata
|
|
3314
|
# rename metadata
|
|
3313
|
#
|
|
3315
|
#
|
|
3314
|
# Checks needed to be done:
|
|
3316
|
# Checks needed to be done:
|
|
3315
|
# 1. length check: L1 == L2, in all cases.
|
|
3317
|
# 1. length check: L1 == L2, in all cases.
|
|
3316
|
# 2. hash check: depending on flag processor, we may need to
|
|
3318
|
# 2. hash check: depending on flag processor, we may need to
|
|
3317
|
# use either "text" (external), or "rawtext" (in revlog).
|
|
3319
|
# use either "text" (external), or "rawtext" (in revlog).
|
|
3318
|
|
|
3320
|
|
|
3319
|
try:
|
|
3321
|
try:
|
|
3320
|
skipflags = state.get(b'skipflags', 0)
|
|
3322
|
skipflags = state.get(b'skipflags', 0)
|
|
3321
|
if skipflags:
|
|
3323
|
if skipflags:
|
|
3322
|
skipflags &= self.flags(rev)
|
|
3324
|
skipflags &= self.flags(rev)
|
|
3323
|
|
|
3325
|
|
|
3324
|
_verify_revision(self, skipflags, state, node)
|
|
3326
|
_verify_revision(self, skipflags, state, node)
|
|
3325
|
|
|
3327
|
|
|
3326
|
l1 = self.rawsize(rev)
|
|
3328
|
l1 = self.rawsize(rev)
|
|
3327
|
l2 = len(self.rawdata(node))
|
|
3329
|
l2 = len(self.rawdata(node))
|
|
3328
|
|
|
3330
|
|
|
3329
|
if l1 != l2:
|
|
3331
|
if l1 != l2:
|
|
3330
|
yield revlogproblem(
|
|
3332
|
yield revlogproblem(
|
|
3331
|
error=_(b'unpacked size is %d, %d expected') % (l2, l1),
|
|
3333
|
error=_(b'unpacked size is %d, %d expected') % (l2, l1),
|
|
3332
|
node=node,
|
|
3334
|
node=node,
|
|
3333
|
)
|
|
3335
|
)
|
|
3334
|
|
|
3336
|
|
|
3335
|
except error.CensoredNodeError:
|
|
3337
|
except error.CensoredNodeError:
|
|
3336
|
if state[b'erroroncensored']:
|
|
3338
|
if state[b'erroroncensored']:
|
|
3337
|
yield revlogproblem(
|
|
3339
|
yield revlogproblem(
|
|
3338
|
error=_(b'censored file data'), node=node
|
|
3340
|
error=_(b'censored file data'), node=node
|
|
3339
|
)
|
|
3341
|
)
|
|
3340
|
state[b'skipread'].add(node)
|
|
3342
|
state[b'skipread'].add(node)
|
|
3341
|
except Exception as e:
|
|
3343
|
except Exception as e:
|
|
3342
|
yield revlogproblem(
|
|
3344
|
yield revlogproblem(
|
|
3343
|
error=_(b'unpacking %s: %s')
|
|
3345
|
error=_(b'unpacking %s: %s')
|
|
3344
|
% (short(node), stringutil.forcebytestr(e)),
|
|
3346
|
% (short(node), stringutil.forcebytestr(e)),
|
|
3345
|
node=node,
|
|
3347
|
node=node,
|
|
3346
|
)
|
|
3348
|
)
|
|
3347
|
state[b'skipread'].add(node)
|
|
3349
|
state[b'skipread'].add(node)
|
|
3348
|
|
|
3350
|
|
|
3349
|
def storageinfo(
|
|
3351
|
def storageinfo(
|
|
3350
|
self,
|
|
3352
|
self,
|
|
3351
|
exclusivefiles=False,
|
|
3353
|
exclusivefiles=False,
|
|
3352
|
sharedfiles=False,
|
|
3354
|
sharedfiles=False,
|
|
3353
|
revisionscount=False,
|
|
3355
|
revisionscount=False,
|
|
3354
|
trackedsize=False,
|
|
3356
|
trackedsize=False,
|
|
3355
|
storedsize=False,
|
|
3357
|
storedsize=False,
|
|
3356
|
):
|
|
3358
|
):
|
|
3357
|
d = {}
|
|
3359
|
d = {}
|
|
3358
|
|
|
3360
|
|
|
3359
|
if exclusivefiles:
|
|
3361
|
if exclusivefiles:
|
|
3360
|
d[b'exclusivefiles'] = [(self.opener, self._indexfile)]
|
|
3362
|
d[b'exclusivefiles'] = [(self.opener, self._indexfile)]
|
|
3361
|
if not self._inline:
|
|
3363
|
if not self._inline:
|
|
3362
|
d[b'exclusivefiles'].append((self.opener, self._datafile))
|
|
3364
|
d[b'exclusivefiles'].append((self.opener, self._datafile))
|
|
3363
|
|
|
3365
|
|
|
3364
|
if sharedfiles:
|
|
3366
|
if sharedfiles:
|
|
3365
|
d[b'sharedfiles'] = []
|
|
3367
|
d[b'sharedfiles'] = []
|
|
3366
|
|
|
3368
|
|
|
3367
|
if revisionscount:
|
|
3369
|
if revisionscount:
|
|
3368
|
d[b'revisionscount'] = len(self)
|
|
3370
|
d[b'revisionscount'] = len(self)
|
|
3369
|
|
|
3371
|
|
|
3370
|
if trackedsize:
|
|
3372
|
if trackedsize:
|
|
3371
|
d[b'trackedsize'] = sum(map(self.rawsize, iter(self)))
|
|
3373
|
d[b'trackedsize'] = sum(map(self.rawsize, iter(self)))
|
|
3372
|
|
|
3374
|
|
|
3373
|
if storedsize:
|
|
3375
|
if storedsize:
|
|
3374
|
d[b'storedsize'] = sum(
|
|
3376
|
d[b'storedsize'] = sum(
|
|
3375
|
self.opener.stat(path).st_size for path in self.files()
|
|
3377
|
self.opener.stat(path).st_size for path in self.files()
|
|
3376
|
)
|
|
3378
|
)
|
|
3377
|
|
|
3379
|
|
|
3378
|
return d
|
|
3380
|
return d
|
|
3379
|
|
|
3381
|
|
|
3380
|
def rewrite_sidedata(self, transaction, helpers, startrev, endrev):
|
|
3382
|
def rewrite_sidedata(self, transaction, helpers, startrev, endrev):
|
|
3381
|
if not self.hassidedata:
|
|
3383
|
if not self.hassidedata:
|
|
3382
|
return
|
|
3384
|
return
|
|
3383
|
# revlog formats with sidedata support does not support inline
|
|
3385
|
# revlog formats with sidedata support does not support inline
|
|
3384
|
assert not self._inline
|
|
3386
|
assert not self._inline
|
|
3385
|
if not helpers[1] and not helpers[2]:
|
|
3387
|
if not helpers[1] and not helpers[2]:
|
|
3386
|
# Nothing to generate or remove
|
|
3388
|
# Nothing to generate or remove
|
|
3387
|
return
|
|
3389
|
return
|
|
3388
|
|
|
3390
|
|
|
3389
|
new_entries = []
|
|
3391
|
new_entries = []
|
|
3390
|
# append the new sidedata
|
|
3392
|
# append the new sidedata
|
|
3391
|
with self._writing(transaction):
|
|
3393
|
with self._writing(transaction):
|
|
3392
|
ifh, dfh = self._writinghandles
|
|
3394
|
ifh, dfh = self._writinghandles
|
|
3393
|
if self._docket is not None:
|
|
3395
|
if self._docket is not None:
|
|
3394
|
dfh.seek(self._docket.data_end, os.SEEK_SET)
|
|
3396
|
dfh.seek(self._docket.data_end, os.SEEK_SET)
|
|
3395
|
else:
|
|
3397
|
else:
|
|
3396
|
dfh.seek(0, os.SEEK_END)
|
|
3398
|
dfh.seek(0, os.SEEK_END)
|
|
3397
|
|
|
3399
|
|
|
3398
|
current_offset = dfh.tell()
|
|
3400
|
current_offset = dfh.tell()
|
|
3399
|
for rev in range(startrev, endrev + 1):
|
|
3401
|
for rev in range(startrev, endrev + 1):
|
|
3400
|
entry = self.index[rev]
|
|
3402
|
entry = self.index[rev]
|
|
3401
|
new_sidedata, flags = sidedatautil.run_sidedata_helpers(
|
|
3403
|
new_sidedata, flags = sidedatautil.run_sidedata_helpers(
|
|
3402
|
store=self,
|
|
3404
|
store=self,
|
|
3403
|
sidedata_helpers=helpers,
|
|
3405
|
sidedata_helpers=helpers,
|
|
3404
|
sidedata={},
|
|
3406
|
sidedata={},
|
|
3405
|
rev=rev,
|
|
3407
|
rev=rev,
|
|
3406
|
)
|
|
3408
|
)
|
|
3407
|
|
|
3409
|
|
|
3408
|
serialized_sidedata = sidedatautil.serialize_sidedata(
|
|
3410
|
serialized_sidedata = sidedatautil.serialize_sidedata(
|
|
3409
|
new_sidedata
|
|
3411
|
new_sidedata
|
|
3410
|
)
|
|
3412
|
)
|
|
3411
|
|
|
3413
|
|
|
3412
|
sidedata_compression_mode = COMP_MODE_INLINE
|
|
3414
|
sidedata_compression_mode = COMP_MODE_INLINE
|
|
3413
|
if serialized_sidedata and self.hassidedata:
|
|
3415
|
if serialized_sidedata and self.hassidedata:
|
|
3414
|
sidedata_compression_mode = COMP_MODE_PLAIN
|
|
3416
|
sidedata_compression_mode = COMP_MODE_PLAIN
|
|
3415
|
h, comp_sidedata = self.compress(serialized_sidedata)
|
|
3417
|
h, comp_sidedata = self.compress(serialized_sidedata)
|
|
3416
|
if (
|
|
3418
|
if (
|
|
3417
|
h != b'u'
|
|
3419
|
h != b'u'
|
|
3418
|
and comp_sidedata[0] != b'\0'
|
|
3420
|
and comp_sidedata[0] != b'\0'
|
|
3419
|
and len(comp_sidedata) < len(serialized_sidedata)
|
|
3421
|
and len(comp_sidedata) < len(serialized_sidedata)
|
|
3420
|
):
|
|
3422
|
):
|
|
3421
|
assert not h
|
|
3423
|
assert not h
|
|
3422
|
if (
|
|
3424
|
if (
|
|
3423
|
comp_sidedata[0]
|
|
3425
|
comp_sidedata[0]
|
|
3424
|
== self._docket.default_compression_header
|
|
3426
|
== self._docket.default_compression_header
|
|
3425
|
):
|
|
3427
|
):
|
|
3426
|
sidedata_compression_mode = COMP_MODE_DEFAULT
|
|
3428
|
sidedata_compression_mode = COMP_MODE_DEFAULT
|
|
3427
|
serialized_sidedata = comp_sidedata
|
|
3429
|
serialized_sidedata = comp_sidedata
|
|
3428
|
else:
|
|
3430
|
else:
|
|
3429
|
sidedata_compression_mode = COMP_MODE_INLINE
|
|
3431
|
sidedata_compression_mode = COMP_MODE_INLINE
|
|
3430
|
serialized_sidedata = comp_sidedata
|
|
3432
|
serialized_sidedata = comp_sidedata
|
|
3431
|
if entry[8] != 0 or entry[9] != 0:
|
|
3433
|
if entry[8] != 0 or entry[9] != 0:
|
|
3432
|
# rewriting entries that already have sidedata is not
|
|
3434
|
# rewriting entries that already have sidedata is not
|
|
3433
|
# supported yet, because it introduces garbage data in the
|
|
3435
|
# supported yet, because it introduces garbage data in the
|
|
3434
|
# revlog.
|
|
3436
|
# revlog.
|
|
3435
|
msg = b"rewriting existing sidedata is not supported yet"
|
|
3437
|
msg = b"rewriting existing sidedata is not supported yet"
|
|
3436
|
raise error.Abort(msg)
|
|
3438
|
raise error.Abort(msg)
|
|
3437
|
|
|
3439
|
|
|
3438
|
# Apply (potential) flags to add and to remove after running
|
|
3440
|
# Apply (potential) flags to add and to remove after running
|
|
3439
|
# the sidedata helpers
|
|
3441
|
# the sidedata helpers
|
|
3440
|
new_offset_flags = entry[0] | flags[0] & ~flags[1]
|
|
3442
|
new_offset_flags = entry[0] | flags[0] & ~flags[1]
|
|
3441
|
entry_update = (
|
|
3443
|
entry_update = (
|
|
3442
|
current_offset,
|
|
3444
|
current_offset,
|
|
3443
|
len(serialized_sidedata),
|
|
3445
|
len(serialized_sidedata),
|
|
3444
|
new_offset_flags,
|
|
3446
|
new_offset_flags,
|
|
3445
|
sidedata_compression_mode,
|
|
3447
|
sidedata_compression_mode,
|
|
3446
|
)
|
|
3448
|
)
|
|
3447
|
|
|
3449
|
|
|
3448
|
# the sidedata computation might have move the file cursors around
|
|
3450
|
# the sidedata computation might have move the file cursors around
|
|
3449
|
dfh.seek(current_offset, os.SEEK_SET)
|
|
3451
|
dfh.seek(current_offset, os.SEEK_SET)
|
|
3450
|
dfh.write(serialized_sidedata)
|
|
3452
|
dfh.write(serialized_sidedata)
|
|
3451
|
new_entries.append(entry_update)
|
|
3453
|
new_entries.append(entry_update)
|
|
3452
|
current_offset += len(serialized_sidedata)
|
|
3454
|
current_offset += len(serialized_sidedata)
|
|
3453
|
if self._docket is not None:
|
|
3455
|
if self._docket is not None:
|
|
3454
|
self._docket.data_end = dfh.tell()
|
|
3456
|
self._docket.data_end = dfh.tell()
|
|
3455
|
|
|
3457
|
|
|
3456
|
# rewrite the new index entries
|
|
3458
|
# rewrite the new index entries
|
|
3457
|
ifh.seek(startrev * self.index.entry_size)
|
|
3459
|
ifh.seek(startrev * self.index.entry_size)
|
|
3458
|
for i, e in enumerate(new_entries):
|
|
3460
|
for i, e in enumerate(new_entries):
|
|
3459
|
rev = startrev + i
|
|
3461
|
rev = startrev + i
|
|
3460
|
self.index.replace_sidedata_info(rev, *e)
|
|
3462
|
self.index.replace_sidedata_info(rev, *e)
|
|
3461
|
packed = self.index.entry_binary(rev)
|
|
3463
|
packed = self.index.entry_binary(rev)
|
|
3462
|
if rev == 0 and self._docket is None:
|
|
3464
|
if rev == 0 and self._docket is None:
|
|
3463
|
header = self._format_flags | self._format_version
|
|
3465
|
header = self._format_flags | self._format_version
|
|
3464
|
header = self.index.pack_header(header)
|
|
3466
|
header = self.index.pack_header(header)
|
|
3465
|
packed = header + packed
|
|
3467
|
packed = header + packed
|
|
3466
|
ifh.write(packed)
|
|
3468
|
ifh.write(packed)
|