##// END OF EJS Templates
revlog: catch more specific exception in shortest()...
Yuya Nishihara -
r39856:7a9e2d85 default
parent child Browse files
Show More
@@ -1,2583 +1,2583 b''
1 # revlog.py - storage back-end for mercurial
1 # revlog.py - storage back-end for mercurial
2 #
2 #
3 # Copyright 2005-2007 Matt Mackall <mpm@selenic.com>
3 # Copyright 2005-2007 Matt Mackall <mpm@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 collections
16 import collections
17 import contextlib
17 import contextlib
18 import errno
18 import errno
19 import hashlib
19 import hashlib
20 import os
20 import os
21 import re
21 import re
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 nullhex,
29 nullhex,
30 nullid,
30 nullid,
31 nullrev,
31 nullrev,
32 wdirfilenodeids,
32 wdirfilenodeids,
33 wdirhex,
33 wdirhex,
34 wdirid,
34 wdirid,
35 wdirrev,
35 wdirrev,
36 )
36 )
37 from .i18n import _
37 from .i18n import _
38 from .revlogutils.constants import (
38 from .revlogutils.constants import (
39 FLAG_GENERALDELTA,
39 FLAG_GENERALDELTA,
40 FLAG_INLINE_DATA,
40 FLAG_INLINE_DATA,
41 REVIDX_DEFAULT_FLAGS,
41 REVIDX_DEFAULT_FLAGS,
42 REVIDX_ELLIPSIS,
42 REVIDX_ELLIPSIS,
43 REVIDX_EXTSTORED,
43 REVIDX_EXTSTORED,
44 REVIDX_FLAGS_ORDER,
44 REVIDX_FLAGS_ORDER,
45 REVIDX_ISCENSORED,
45 REVIDX_ISCENSORED,
46 REVIDX_KNOWN_FLAGS,
46 REVIDX_KNOWN_FLAGS,
47 REVIDX_RAWTEXT_CHANGING_FLAGS,
47 REVIDX_RAWTEXT_CHANGING_FLAGS,
48 REVLOGV0,
48 REVLOGV0,
49 REVLOGV1,
49 REVLOGV1,
50 REVLOGV1_FLAGS,
50 REVLOGV1_FLAGS,
51 REVLOGV2,
51 REVLOGV2,
52 REVLOGV2_FLAGS,
52 REVLOGV2_FLAGS,
53 REVLOG_DEFAULT_FLAGS,
53 REVLOG_DEFAULT_FLAGS,
54 REVLOG_DEFAULT_FORMAT,
54 REVLOG_DEFAULT_FORMAT,
55 REVLOG_DEFAULT_VERSION,
55 REVLOG_DEFAULT_VERSION,
56 )
56 )
57 from .thirdparty import (
57 from .thirdparty import (
58 attr,
58 attr,
59 )
59 )
60 from . import (
60 from . import (
61 ancestor,
61 ancestor,
62 error,
62 error,
63 mdiff,
63 mdiff,
64 policy,
64 policy,
65 pycompat,
65 pycompat,
66 repository,
66 repository,
67 templatefilters,
67 templatefilters,
68 util,
68 util,
69 )
69 )
70 from .revlogutils import (
70 from .revlogutils import (
71 deltas as deltautil,
71 deltas as deltautil,
72 )
72 )
73 from .utils import (
73 from .utils import (
74 interfaceutil,
74 interfaceutil,
75 stringutil,
75 stringutil,
76 )
76 )
77
77
78 # blanked usage of all the name to prevent pyflakes constraints
78 # blanked usage of all the name to prevent pyflakes constraints
79 # We need these name available in the module for extensions.
79 # We need these name available in the module for extensions.
80 REVLOGV0
80 REVLOGV0
81 REVLOGV1
81 REVLOGV1
82 REVLOGV2
82 REVLOGV2
83 FLAG_INLINE_DATA
83 FLAG_INLINE_DATA
84 FLAG_GENERALDELTA
84 FLAG_GENERALDELTA
85 REVLOG_DEFAULT_FLAGS
85 REVLOG_DEFAULT_FLAGS
86 REVLOG_DEFAULT_FORMAT
86 REVLOG_DEFAULT_FORMAT
87 REVLOG_DEFAULT_VERSION
87 REVLOG_DEFAULT_VERSION
88 REVLOGV1_FLAGS
88 REVLOGV1_FLAGS
89 REVLOGV2_FLAGS
89 REVLOGV2_FLAGS
90 REVIDX_ISCENSORED
90 REVIDX_ISCENSORED
91 REVIDX_ELLIPSIS
91 REVIDX_ELLIPSIS
92 REVIDX_EXTSTORED
92 REVIDX_EXTSTORED
93 REVIDX_DEFAULT_FLAGS
93 REVIDX_DEFAULT_FLAGS
94 REVIDX_FLAGS_ORDER
94 REVIDX_FLAGS_ORDER
95 REVIDX_KNOWN_FLAGS
95 REVIDX_KNOWN_FLAGS
96 REVIDX_RAWTEXT_CHANGING_FLAGS
96 REVIDX_RAWTEXT_CHANGING_FLAGS
97
97
98 parsers = policy.importmod(r'parsers')
98 parsers = policy.importmod(r'parsers')
99
99
100 # Aliased for performance.
100 # Aliased for performance.
101 _zlibdecompress = zlib.decompress
101 _zlibdecompress = zlib.decompress
102
102
103 # max size of revlog with inline data
103 # max size of revlog with inline data
104 _maxinline = 131072
104 _maxinline = 131072
105 _chunksize = 1048576
105 _chunksize = 1048576
106
106
107 # Store flag processors (cf. 'addflagprocessor()' to register)
107 # Store flag processors (cf. 'addflagprocessor()' to register)
108 _flagprocessors = {
108 _flagprocessors = {
109 REVIDX_ISCENSORED: None,
109 REVIDX_ISCENSORED: None,
110 }
110 }
111
111
112 # Flag processors for REVIDX_ELLIPSIS.
112 # Flag processors for REVIDX_ELLIPSIS.
113 def ellipsisreadprocessor(rl, text):
113 def ellipsisreadprocessor(rl, text):
114 return text, False
114 return text, False
115
115
116 def ellipsiswriteprocessor(rl, text):
116 def ellipsiswriteprocessor(rl, text):
117 return text, False
117 return text, False
118
118
119 def ellipsisrawprocessor(rl, text):
119 def ellipsisrawprocessor(rl, text):
120 return False
120 return False
121
121
122 ellipsisprocessor = (
122 ellipsisprocessor = (
123 ellipsisreadprocessor,
123 ellipsisreadprocessor,
124 ellipsiswriteprocessor,
124 ellipsiswriteprocessor,
125 ellipsisrawprocessor,
125 ellipsisrawprocessor,
126 )
126 )
127
127
128 _mdre = re.compile('\1\n')
128 _mdre = re.compile('\1\n')
129 def parsemeta(text):
129 def parsemeta(text):
130 """return (metadatadict, metadatasize)"""
130 """return (metadatadict, metadatasize)"""
131 # text can be buffer, so we can't use .startswith or .index
131 # text can be buffer, so we can't use .startswith or .index
132 if text[:2] != '\1\n':
132 if text[:2] != '\1\n':
133 return None, None
133 return None, None
134 s = _mdre.search(text, 2).start()
134 s = _mdre.search(text, 2).start()
135 mtext = text[2:s]
135 mtext = text[2:s]
136 meta = {}
136 meta = {}
137 for l in mtext.splitlines():
137 for l in mtext.splitlines():
138 k, v = l.split(": ", 1)
138 k, v = l.split(": ", 1)
139 meta[k] = v
139 meta[k] = v
140 return meta, (s + 2)
140 return meta, (s + 2)
141
141
142 def packmeta(meta, text):
142 def packmeta(meta, text):
143 keys = sorted(meta)
143 keys = sorted(meta)
144 metatext = "".join("%s: %s\n" % (k, meta[k]) for k in keys)
144 metatext = "".join("%s: %s\n" % (k, meta[k]) for k in keys)
145 return "\1\n%s\1\n%s" % (metatext, text)
145 return "\1\n%s\1\n%s" % (metatext, text)
146
146
147 def _censoredtext(text):
147 def _censoredtext(text):
148 m, offs = parsemeta(text)
148 m, offs = parsemeta(text)
149 return m and "censored" in m
149 return m and "censored" in m
150
150
151 def addflagprocessor(flag, processor):
151 def addflagprocessor(flag, processor):
152 """Register a flag processor on a revision data flag.
152 """Register a flag processor on a revision data flag.
153
153
154 Invariant:
154 Invariant:
155 - Flags need to be defined in REVIDX_KNOWN_FLAGS and REVIDX_FLAGS_ORDER,
155 - Flags need to be defined in REVIDX_KNOWN_FLAGS and REVIDX_FLAGS_ORDER,
156 and REVIDX_RAWTEXT_CHANGING_FLAGS if they can alter rawtext.
156 and REVIDX_RAWTEXT_CHANGING_FLAGS if they can alter rawtext.
157 - Only one flag processor can be registered on a specific flag.
157 - Only one flag processor can be registered on a specific flag.
158 - flagprocessors must be 3-tuples of functions (read, write, raw) with the
158 - flagprocessors must be 3-tuples of functions (read, write, raw) with the
159 following signatures:
159 following signatures:
160 - (read) f(self, rawtext) -> text, bool
160 - (read) f(self, rawtext) -> text, bool
161 - (write) f(self, text) -> rawtext, bool
161 - (write) f(self, text) -> rawtext, bool
162 - (raw) f(self, rawtext) -> bool
162 - (raw) f(self, rawtext) -> bool
163 "text" is presented to the user. "rawtext" is stored in revlog data, not
163 "text" is presented to the user. "rawtext" is stored in revlog data, not
164 directly visible to the user.
164 directly visible to the user.
165 The boolean returned by these transforms is used to determine whether
165 The boolean returned by these transforms is used to determine whether
166 the returned text can be used for hash integrity checking. For example,
166 the returned text can be used for hash integrity checking. For example,
167 if "write" returns False, then "text" is used to generate hash. If
167 if "write" returns False, then "text" is used to generate hash. If
168 "write" returns True, that basically means "rawtext" returned by "write"
168 "write" returns True, that basically means "rawtext" returned by "write"
169 should be used to generate hash. Usually, "write" and "read" return
169 should be used to generate hash. Usually, "write" and "read" return
170 different booleans. And "raw" returns a same boolean as "write".
170 different booleans. And "raw" returns a same boolean as "write".
171
171
172 Note: The 'raw' transform is used for changegroup generation and in some
172 Note: The 'raw' transform is used for changegroup generation and in some
173 debug commands. In this case the transform only indicates whether the
173 debug commands. In this case the transform only indicates whether the
174 contents can be used for hash integrity checks.
174 contents can be used for hash integrity checks.
175 """
175 """
176 if not flag & REVIDX_KNOWN_FLAGS:
176 if not flag & REVIDX_KNOWN_FLAGS:
177 msg = _("cannot register processor on unknown flag '%#x'.") % (flag)
177 msg = _("cannot register processor on unknown flag '%#x'.") % (flag)
178 raise error.ProgrammingError(msg)
178 raise error.ProgrammingError(msg)
179 if flag not in REVIDX_FLAGS_ORDER:
179 if flag not in REVIDX_FLAGS_ORDER:
180 msg = _("flag '%#x' undefined in REVIDX_FLAGS_ORDER.") % (flag)
180 msg = _("flag '%#x' undefined in REVIDX_FLAGS_ORDER.") % (flag)
181 raise error.ProgrammingError(msg)
181 raise error.ProgrammingError(msg)
182 if flag in _flagprocessors:
182 if flag in _flagprocessors:
183 msg = _("cannot register multiple processors on flag '%#x'.") % (flag)
183 msg = _("cannot register multiple processors on flag '%#x'.") % (flag)
184 raise error.Abort(msg)
184 raise error.Abort(msg)
185 _flagprocessors[flag] = processor
185 _flagprocessors[flag] = processor
186
186
187 def getoffset(q):
187 def getoffset(q):
188 return int(q >> 16)
188 return int(q >> 16)
189
189
190 def gettype(q):
190 def gettype(q):
191 return int(q & 0xFFFF)
191 return int(q & 0xFFFF)
192
192
193 def offset_type(offset, type):
193 def offset_type(offset, type):
194 if (type & ~REVIDX_KNOWN_FLAGS) != 0:
194 if (type & ~REVIDX_KNOWN_FLAGS) != 0:
195 raise ValueError('unknown revlog index flags')
195 raise ValueError('unknown revlog index flags')
196 return int(int(offset) << 16 | type)
196 return int(int(offset) << 16 | type)
197
197
198 _nullhash = hashlib.sha1(nullid)
198 _nullhash = hashlib.sha1(nullid)
199
199
200 def hash(text, p1, p2):
200 def hash(text, p1, p2):
201 """generate a hash from the given text and its parent hashes
201 """generate a hash from the given text and its parent hashes
202
202
203 This hash combines both the current file contents and its history
203 This hash combines both the current file contents and its history
204 in a manner that makes it easy to distinguish nodes with the same
204 in a manner that makes it easy to distinguish nodes with the same
205 content in the revision graph.
205 content in the revision graph.
206 """
206 """
207 # As of now, if one of the parent node is null, p2 is null
207 # As of now, if one of the parent node is null, p2 is null
208 if p2 == nullid:
208 if p2 == nullid:
209 # deep copy of a hash is faster than creating one
209 # deep copy of a hash is faster than creating one
210 s = _nullhash.copy()
210 s = _nullhash.copy()
211 s.update(p1)
211 s.update(p1)
212 else:
212 else:
213 # none of the parent nodes are nullid
213 # none of the parent nodes are nullid
214 if p1 < p2:
214 if p1 < p2:
215 a = p1
215 a = p1
216 b = p2
216 b = p2
217 else:
217 else:
218 a = p2
218 a = p2
219 b = p1
219 b = p1
220 s = hashlib.sha1(a)
220 s = hashlib.sha1(a)
221 s.update(b)
221 s.update(b)
222 s.update(text)
222 s.update(text)
223 return s.digest()
223 return s.digest()
224
224
225 @attr.s(slots=True, frozen=True)
225 @attr.s(slots=True, frozen=True)
226 class _revisioninfo(object):
226 class _revisioninfo(object):
227 """Information about a revision that allows building its fulltext
227 """Information about a revision that allows building its fulltext
228 node: expected hash of the revision
228 node: expected hash of the revision
229 p1, p2: parent revs of the revision
229 p1, p2: parent revs of the revision
230 btext: built text cache consisting of a one-element list
230 btext: built text cache consisting of a one-element list
231 cachedelta: (baserev, uncompressed_delta) or None
231 cachedelta: (baserev, uncompressed_delta) or None
232 flags: flags associated to the revision storage
232 flags: flags associated to the revision storage
233
233
234 One of btext[0] or cachedelta must be set.
234 One of btext[0] or cachedelta must be set.
235 """
235 """
236 node = attr.ib()
236 node = attr.ib()
237 p1 = attr.ib()
237 p1 = attr.ib()
238 p2 = attr.ib()
238 p2 = attr.ib()
239 btext = attr.ib()
239 btext = attr.ib()
240 textlen = attr.ib()
240 textlen = attr.ib()
241 cachedelta = attr.ib()
241 cachedelta = attr.ib()
242 flags = attr.ib()
242 flags = attr.ib()
243
243
244 @interfaceutil.implementer(repository.irevisiondelta)
244 @interfaceutil.implementer(repository.irevisiondelta)
245 @attr.s(slots=True, frozen=True)
245 @attr.s(slots=True, frozen=True)
246 class revlogrevisiondelta(object):
246 class revlogrevisiondelta(object):
247 node = attr.ib()
247 node = attr.ib()
248 p1node = attr.ib()
248 p1node = attr.ib()
249 p2node = attr.ib()
249 p2node = attr.ib()
250 basenode = attr.ib()
250 basenode = attr.ib()
251 linknode = attr.ib()
251 linknode = attr.ib()
252 flags = attr.ib()
252 flags = attr.ib()
253 baserevisionsize = attr.ib()
253 baserevisionsize = attr.ib()
254 revision = attr.ib()
254 revision = attr.ib()
255 delta = attr.ib()
255 delta = attr.ib()
256
256
257 # index v0:
257 # index v0:
258 # 4 bytes: offset
258 # 4 bytes: offset
259 # 4 bytes: compressed length
259 # 4 bytes: compressed length
260 # 4 bytes: base rev
260 # 4 bytes: base rev
261 # 4 bytes: link rev
261 # 4 bytes: link rev
262 # 20 bytes: parent 1 nodeid
262 # 20 bytes: parent 1 nodeid
263 # 20 bytes: parent 2 nodeid
263 # 20 bytes: parent 2 nodeid
264 # 20 bytes: nodeid
264 # 20 bytes: nodeid
265 indexformatv0 = struct.Struct(">4l20s20s20s")
265 indexformatv0 = struct.Struct(">4l20s20s20s")
266 indexformatv0_pack = indexformatv0.pack
266 indexformatv0_pack = indexformatv0.pack
267 indexformatv0_unpack = indexformatv0.unpack
267 indexformatv0_unpack = indexformatv0.unpack
268
268
269 class revlogoldindex(list):
269 class revlogoldindex(list):
270 def __getitem__(self, i):
270 def __getitem__(self, i):
271 if i == -1:
271 if i == -1:
272 return (0, 0, 0, -1, -1, -1, -1, nullid)
272 return (0, 0, 0, -1, -1, -1, -1, nullid)
273 return list.__getitem__(self, i)
273 return list.__getitem__(self, i)
274
274
275 class revlogoldio(object):
275 class revlogoldio(object):
276 def __init__(self):
276 def __init__(self):
277 self.size = indexformatv0.size
277 self.size = indexformatv0.size
278
278
279 def parseindex(self, data, inline):
279 def parseindex(self, data, inline):
280 s = self.size
280 s = self.size
281 index = []
281 index = []
282 nodemap = {nullid: nullrev}
282 nodemap = {nullid: nullrev}
283 n = off = 0
283 n = off = 0
284 l = len(data)
284 l = len(data)
285 while off + s <= l:
285 while off + s <= l:
286 cur = data[off:off + s]
286 cur = data[off:off + s]
287 off += s
287 off += s
288 e = indexformatv0_unpack(cur)
288 e = indexformatv0_unpack(cur)
289 # transform to revlogv1 format
289 # transform to revlogv1 format
290 e2 = (offset_type(e[0], 0), e[1], -1, e[2], e[3],
290 e2 = (offset_type(e[0], 0), e[1], -1, e[2], e[3],
291 nodemap.get(e[4], nullrev), nodemap.get(e[5], nullrev), e[6])
291 nodemap.get(e[4], nullrev), nodemap.get(e[5], nullrev), e[6])
292 index.append(e2)
292 index.append(e2)
293 nodemap[e[6]] = n
293 nodemap[e[6]] = n
294 n += 1
294 n += 1
295
295
296 return revlogoldindex(index), nodemap, None
296 return revlogoldindex(index), nodemap, None
297
297
298 def packentry(self, entry, node, version, rev):
298 def packentry(self, entry, node, version, rev):
299 if gettype(entry[0]):
299 if gettype(entry[0]):
300 raise error.RevlogError(_('index entry flags need revlog '
300 raise error.RevlogError(_('index entry flags need revlog '
301 'version 1'))
301 'version 1'))
302 e2 = (getoffset(entry[0]), entry[1], entry[3], entry[4],
302 e2 = (getoffset(entry[0]), entry[1], entry[3], entry[4],
303 node(entry[5]), node(entry[6]), entry[7])
303 node(entry[5]), node(entry[6]), entry[7])
304 return indexformatv0_pack(*e2)
304 return indexformatv0_pack(*e2)
305
305
306 # index ng:
306 # index ng:
307 # 6 bytes: offset
307 # 6 bytes: offset
308 # 2 bytes: flags
308 # 2 bytes: flags
309 # 4 bytes: compressed length
309 # 4 bytes: compressed length
310 # 4 bytes: uncompressed length
310 # 4 bytes: uncompressed length
311 # 4 bytes: base rev
311 # 4 bytes: base rev
312 # 4 bytes: link rev
312 # 4 bytes: link rev
313 # 4 bytes: parent 1 rev
313 # 4 bytes: parent 1 rev
314 # 4 bytes: parent 2 rev
314 # 4 bytes: parent 2 rev
315 # 32 bytes: nodeid
315 # 32 bytes: nodeid
316 indexformatng = struct.Struct(">Qiiiiii20s12x")
316 indexformatng = struct.Struct(">Qiiiiii20s12x")
317 indexformatng_pack = indexformatng.pack
317 indexformatng_pack = indexformatng.pack
318 versionformat = struct.Struct(">I")
318 versionformat = struct.Struct(">I")
319 versionformat_pack = versionformat.pack
319 versionformat_pack = versionformat.pack
320 versionformat_unpack = versionformat.unpack
320 versionformat_unpack = versionformat.unpack
321
321
322 # corresponds to uncompressed length of indexformatng (2 gigs, 4-byte
322 # corresponds to uncompressed length of indexformatng (2 gigs, 4-byte
323 # signed integer)
323 # signed integer)
324 _maxentrysize = 0x7fffffff
324 _maxentrysize = 0x7fffffff
325
325
326 class revlogio(object):
326 class revlogio(object):
327 def __init__(self):
327 def __init__(self):
328 self.size = indexformatng.size
328 self.size = indexformatng.size
329
329
330 def parseindex(self, data, inline):
330 def parseindex(self, data, inline):
331 # call the C implementation to parse the index data
331 # call the C implementation to parse the index data
332 index, cache = parsers.parse_index2(data, inline)
332 index, cache = parsers.parse_index2(data, inline)
333 return index, getattr(index, 'nodemap', None), cache
333 return index, getattr(index, 'nodemap', None), cache
334
334
335 def packentry(self, entry, node, version, rev):
335 def packentry(self, entry, node, version, rev):
336 p = indexformatng_pack(*entry)
336 p = indexformatng_pack(*entry)
337 if rev == 0:
337 if rev == 0:
338 p = versionformat_pack(version) + p[4:]
338 p = versionformat_pack(version) + p[4:]
339 return p
339 return p
340
340
341 class revlog(object):
341 class revlog(object):
342 """
342 """
343 the underlying revision storage object
343 the underlying revision storage object
344
344
345 A revlog consists of two parts, an index and the revision data.
345 A revlog consists of two parts, an index and the revision data.
346
346
347 The index is a file with a fixed record size containing
347 The index is a file with a fixed record size containing
348 information on each revision, including its nodeid (hash), the
348 information on each revision, including its nodeid (hash), the
349 nodeids of its parents, the position and offset of its data within
349 nodeids of its parents, the position and offset of its data within
350 the data file, and the revision it's based on. Finally, each entry
350 the data file, and the revision it's based on. Finally, each entry
351 contains a linkrev entry that can serve as a pointer to external
351 contains a linkrev entry that can serve as a pointer to external
352 data.
352 data.
353
353
354 The revision data itself is a linear collection of data chunks.
354 The revision data itself is a linear collection of data chunks.
355 Each chunk represents a revision and is usually represented as a
355 Each chunk represents a revision and is usually represented as a
356 delta against the previous chunk. To bound lookup time, runs of
356 delta against the previous chunk. To bound lookup time, runs of
357 deltas are limited to about 2 times the length of the original
357 deltas are limited to about 2 times the length of the original
358 version data. This makes retrieval of a version proportional to
358 version data. This makes retrieval of a version proportional to
359 its size, or O(1) relative to the number of revisions.
359 its size, or O(1) relative to the number of revisions.
360
360
361 Both pieces of the revlog are written to in an append-only
361 Both pieces of the revlog are written to in an append-only
362 fashion, which means we never need to rewrite a file to insert or
362 fashion, which means we never need to rewrite a file to insert or
363 remove data, and can use some simple techniques to avoid the need
363 remove data, and can use some simple techniques to avoid the need
364 for locking while reading.
364 for locking while reading.
365
365
366 If checkambig, indexfile is opened with checkambig=True at
366 If checkambig, indexfile is opened with checkambig=True at
367 writing, to avoid file stat ambiguity.
367 writing, to avoid file stat ambiguity.
368
368
369 If mmaplargeindex is True, and an mmapindexthreshold is set, the
369 If mmaplargeindex is True, and an mmapindexthreshold is set, the
370 index will be mmapped rather than read if it is larger than the
370 index will be mmapped rather than read if it is larger than the
371 configured threshold.
371 configured threshold.
372
372
373 If censorable is True, the revlog can have censored revisions.
373 If censorable is True, the revlog can have censored revisions.
374 """
374 """
375 def __init__(self, opener, indexfile, datafile=None, checkambig=False,
375 def __init__(self, opener, indexfile, datafile=None, checkambig=False,
376 mmaplargeindex=False, censorable=False):
376 mmaplargeindex=False, censorable=False):
377 """
377 """
378 create a revlog object
378 create a revlog object
379
379
380 opener is a function that abstracts the file opening operation
380 opener is a function that abstracts the file opening operation
381 and can be used to implement COW semantics or the like.
381 and can be used to implement COW semantics or the like.
382 """
382 """
383 self.indexfile = indexfile
383 self.indexfile = indexfile
384 self.datafile = datafile or (indexfile[:-2] + ".d")
384 self.datafile = datafile or (indexfile[:-2] + ".d")
385 self.opener = opener
385 self.opener = opener
386 # When True, indexfile is opened with checkambig=True at writing, to
386 # When True, indexfile is opened with checkambig=True at writing, to
387 # avoid file stat ambiguity.
387 # avoid file stat ambiguity.
388 self._checkambig = checkambig
388 self._checkambig = checkambig
389 self._censorable = censorable
389 self._censorable = censorable
390 # 3-tuple of (node, rev, text) for a raw revision.
390 # 3-tuple of (node, rev, text) for a raw revision.
391 self._cache = None
391 self._cache = None
392 # Maps rev to chain base rev.
392 # Maps rev to chain base rev.
393 self._chainbasecache = util.lrucachedict(100)
393 self._chainbasecache = util.lrucachedict(100)
394 # 2-tuple of (offset, data) of raw data from the revlog at an offset.
394 # 2-tuple of (offset, data) of raw data from the revlog at an offset.
395 self._chunkcache = (0, '')
395 self._chunkcache = (0, '')
396 # How much data to read and cache into the raw revlog data cache.
396 # How much data to read and cache into the raw revlog data cache.
397 self._chunkcachesize = 65536
397 self._chunkcachesize = 65536
398 self._maxchainlen = None
398 self._maxchainlen = None
399 self._deltabothparents = True
399 self._deltabothparents = True
400 self.index = []
400 self.index = []
401 # Mapping of partial identifiers to full nodes.
401 # Mapping of partial identifiers to full nodes.
402 self._pcache = {}
402 self._pcache = {}
403 # Mapping of revision integer to full node.
403 # Mapping of revision integer to full node.
404 self._nodecache = {nullid: nullrev}
404 self._nodecache = {nullid: nullrev}
405 self._nodepos = None
405 self._nodepos = None
406 self._compengine = 'zlib'
406 self._compengine = 'zlib'
407 self._maxdeltachainspan = -1
407 self._maxdeltachainspan = -1
408 self._withsparseread = False
408 self._withsparseread = False
409 self._sparserevlog = False
409 self._sparserevlog = False
410 self._srdensitythreshold = 0.50
410 self._srdensitythreshold = 0.50
411 self._srmingapsize = 262144
411 self._srmingapsize = 262144
412
412
413 # Make copy of flag processors so each revlog instance can support
413 # Make copy of flag processors so each revlog instance can support
414 # custom flags.
414 # custom flags.
415 self._flagprocessors = dict(_flagprocessors)
415 self._flagprocessors = dict(_flagprocessors)
416
416
417 mmapindexthreshold = None
417 mmapindexthreshold = None
418 v = REVLOG_DEFAULT_VERSION
418 v = REVLOG_DEFAULT_VERSION
419 opts = getattr(opener, 'options', None)
419 opts = getattr(opener, 'options', None)
420 if opts is not None:
420 if opts is not None:
421 if 'revlogv2' in opts:
421 if 'revlogv2' in opts:
422 # version 2 revlogs always use generaldelta.
422 # version 2 revlogs always use generaldelta.
423 v = REVLOGV2 | FLAG_GENERALDELTA | FLAG_INLINE_DATA
423 v = REVLOGV2 | FLAG_GENERALDELTA | FLAG_INLINE_DATA
424 elif 'revlogv1' in opts:
424 elif 'revlogv1' in opts:
425 if 'generaldelta' in opts:
425 if 'generaldelta' in opts:
426 v |= FLAG_GENERALDELTA
426 v |= FLAG_GENERALDELTA
427 else:
427 else:
428 v = 0
428 v = 0
429 if 'chunkcachesize' in opts:
429 if 'chunkcachesize' in opts:
430 self._chunkcachesize = opts['chunkcachesize']
430 self._chunkcachesize = opts['chunkcachesize']
431 if 'maxchainlen' in opts:
431 if 'maxchainlen' in opts:
432 self._maxchainlen = opts['maxchainlen']
432 self._maxchainlen = opts['maxchainlen']
433 if 'deltabothparents' in opts:
433 if 'deltabothparents' in opts:
434 self._deltabothparents = opts['deltabothparents']
434 self._deltabothparents = opts['deltabothparents']
435 self._lazydeltabase = bool(opts.get('lazydeltabase', False))
435 self._lazydeltabase = bool(opts.get('lazydeltabase', False))
436 if 'compengine' in opts:
436 if 'compengine' in opts:
437 self._compengine = opts['compengine']
437 self._compengine = opts['compengine']
438 if 'maxdeltachainspan' in opts:
438 if 'maxdeltachainspan' in opts:
439 self._maxdeltachainspan = opts['maxdeltachainspan']
439 self._maxdeltachainspan = opts['maxdeltachainspan']
440 if mmaplargeindex and 'mmapindexthreshold' in opts:
440 if mmaplargeindex and 'mmapindexthreshold' in opts:
441 mmapindexthreshold = opts['mmapindexthreshold']
441 mmapindexthreshold = opts['mmapindexthreshold']
442 self._sparserevlog = bool(opts.get('sparse-revlog', False))
442 self._sparserevlog = bool(opts.get('sparse-revlog', False))
443 withsparseread = bool(opts.get('with-sparse-read', False))
443 withsparseread = bool(opts.get('with-sparse-read', False))
444 # sparse-revlog forces sparse-read
444 # sparse-revlog forces sparse-read
445 self._withsparseread = self._sparserevlog or withsparseread
445 self._withsparseread = self._sparserevlog or withsparseread
446 if 'sparse-read-density-threshold' in opts:
446 if 'sparse-read-density-threshold' in opts:
447 self._srdensitythreshold = opts['sparse-read-density-threshold']
447 self._srdensitythreshold = opts['sparse-read-density-threshold']
448 if 'sparse-read-min-gap-size' in opts:
448 if 'sparse-read-min-gap-size' in opts:
449 self._srmingapsize = opts['sparse-read-min-gap-size']
449 self._srmingapsize = opts['sparse-read-min-gap-size']
450 if opts.get('enableellipsis'):
450 if opts.get('enableellipsis'):
451 self._flagprocessors[REVIDX_ELLIPSIS] = ellipsisprocessor
451 self._flagprocessors[REVIDX_ELLIPSIS] = ellipsisprocessor
452
452
453 if self._chunkcachesize <= 0:
453 if self._chunkcachesize <= 0:
454 raise error.RevlogError(_('revlog chunk cache size %r is not '
454 raise error.RevlogError(_('revlog chunk cache size %r is not '
455 'greater than 0') % self._chunkcachesize)
455 'greater than 0') % self._chunkcachesize)
456 elif self._chunkcachesize & (self._chunkcachesize - 1):
456 elif self._chunkcachesize & (self._chunkcachesize - 1):
457 raise error.RevlogError(_('revlog chunk cache size %r is not a '
457 raise error.RevlogError(_('revlog chunk cache size %r is not a '
458 'power of 2') % self._chunkcachesize)
458 'power of 2') % self._chunkcachesize)
459
459
460 indexdata = ''
460 indexdata = ''
461 self._initempty = True
461 self._initempty = True
462 try:
462 try:
463 with self._indexfp() as f:
463 with self._indexfp() as f:
464 if (mmapindexthreshold is not None and
464 if (mmapindexthreshold is not None and
465 self.opener.fstat(f).st_size >= mmapindexthreshold):
465 self.opener.fstat(f).st_size >= mmapindexthreshold):
466 indexdata = util.buffer(util.mmapread(f))
466 indexdata = util.buffer(util.mmapread(f))
467 else:
467 else:
468 indexdata = f.read()
468 indexdata = f.read()
469 if len(indexdata) > 0:
469 if len(indexdata) > 0:
470 v = versionformat_unpack(indexdata[:4])[0]
470 v = versionformat_unpack(indexdata[:4])[0]
471 self._initempty = False
471 self._initempty = False
472 except IOError as inst:
472 except IOError as inst:
473 if inst.errno != errno.ENOENT:
473 if inst.errno != errno.ENOENT:
474 raise
474 raise
475
475
476 self.version = v
476 self.version = v
477 self._inline = v & FLAG_INLINE_DATA
477 self._inline = v & FLAG_INLINE_DATA
478 self._generaldelta = v & FLAG_GENERALDELTA
478 self._generaldelta = v & FLAG_GENERALDELTA
479 flags = v & ~0xFFFF
479 flags = v & ~0xFFFF
480 fmt = v & 0xFFFF
480 fmt = v & 0xFFFF
481 if fmt == REVLOGV0:
481 if fmt == REVLOGV0:
482 if flags:
482 if flags:
483 raise error.RevlogError(_('unknown flags (%#04x) in version %d '
483 raise error.RevlogError(_('unknown flags (%#04x) in version %d '
484 'revlog %s') %
484 'revlog %s') %
485 (flags >> 16, fmt, self.indexfile))
485 (flags >> 16, fmt, self.indexfile))
486 elif fmt == REVLOGV1:
486 elif fmt == REVLOGV1:
487 if flags & ~REVLOGV1_FLAGS:
487 if flags & ~REVLOGV1_FLAGS:
488 raise error.RevlogError(_('unknown flags (%#04x) in version %d '
488 raise error.RevlogError(_('unknown flags (%#04x) in version %d '
489 'revlog %s') %
489 'revlog %s') %
490 (flags >> 16, fmt, self.indexfile))
490 (flags >> 16, fmt, self.indexfile))
491 elif fmt == REVLOGV2:
491 elif fmt == REVLOGV2:
492 if flags & ~REVLOGV2_FLAGS:
492 if flags & ~REVLOGV2_FLAGS:
493 raise error.RevlogError(_('unknown flags (%#04x) in version %d '
493 raise error.RevlogError(_('unknown flags (%#04x) in version %d '
494 'revlog %s') %
494 'revlog %s') %
495 (flags >> 16, fmt, self.indexfile))
495 (flags >> 16, fmt, self.indexfile))
496 else:
496 else:
497 raise error.RevlogError(_('unknown version (%d) in revlog %s') %
497 raise error.RevlogError(_('unknown version (%d) in revlog %s') %
498 (fmt, self.indexfile))
498 (fmt, self.indexfile))
499
499
500 self._storedeltachains = True
500 self._storedeltachains = True
501
501
502 self._io = revlogio()
502 self._io = revlogio()
503 if self.version == REVLOGV0:
503 if self.version == REVLOGV0:
504 self._io = revlogoldio()
504 self._io = revlogoldio()
505 try:
505 try:
506 d = self._io.parseindex(indexdata, self._inline)
506 d = self._io.parseindex(indexdata, self._inline)
507 except (ValueError, IndexError):
507 except (ValueError, IndexError):
508 raise error.RevlogError(_("index %s is corrupted") %
508 raise error.RevlogError(_("index %s is corrupted") %
509 self.indexfile)
509 self.indexfile)
510 self.index, nodemap, self._chunkcache = d
510 self.index, nodemap, self._chunkcache = d
511 if nodemap is not None:
511 if nodemap is not None:
512 self.nodemap = self._nodecache = nodemap
512 self.nodemap = self._nodecache = nodemap
513 if not self._chunkcache:
513 if not self._chunkcache:
514 self._chunkclear()
514 self._chunkclear()
515 # revnum -> (chain-length, sum-delta-length)
515 # revnum -> (chain-length, sum-delta-length)
516 self._chaininfocache = {}
516 self._chaininfocache = {}
517 # revlog header -> revlog compressor
517 # revlog header -> revlog compressor
518 self._decompressors = {}
518 self._decompressors = {}
519
519
520 @util.propertycache
520 @util.propertycache
521 def _compressor(self):
521 def _compressor(self):
522 return util.compengines[self._compengine].revlogcompressor()
522 return util.compengines[self._compengine].revlogcompressor()
523
523
524 def _indexfp(self, mode='r'):
524 def _indexfp(self, mode='r'):
525 """file object for the revlog's index file"""
525 """file object for the revlog's index file"""
526 args = {r'mode': mode}
526 args = {r'mode': mode}
527 if mode != 'r':
527 if mode != 'r':
528 args[r'checkambig'] = self._checkambig
528 args[r'checkambig'] = self._checkambig
529 if mode == 'w':
529 if mode == 'w':
530 args[r'atomictemp'] = True
530 args[r'atomictemp'] = True
531 return self.opener(self.indexfile, **args)
531 return self.opener(self.indexfile, **args)
532
532
533 def _datafp(self, mode='r'):
533 def _datafp(self, mode='r'):
534 """file object for the revlog's data file"""
534 """file object for the revlog's data file"""
535 return self.opener(self.datafile, mode=mode)
535 return self.opener(self.datafile, mode=mode)
536
536
537 @contextlib.contextmanager
537 @contextlib.contextmanager
538 def _datareadfp(self, existingfp=None):
538 def _datareadfp(self, existingfp=None):
539 """file object suitable to read data"""
539 """file object suitable to read data"""
540 if existingfp is not None:
540 if existingfp is not None:
541 yield existingfp
541 yield existingfp
542 else:
542 else:
543 if self._inline:
543 if self._inline:
544 func = self._indexfp
544 func = self._indexfp
545 else:
545 else:
546 func = self._datafp
546 func = self._datafp
547 with func() as fp:
547 with func() as fp:
548 yield fp
548 yield fp
549
549
550 def tip(self):
550 def tip(self):
551 return self.node(len(self.index) - 1)
551 return self.node(len(self.index) - 1)
552 def __contains__(self, rev):
552 def __contains__(self, rev):
553 return 0 <= rev < len(self)
553 return 0 <= rev < len(self)
554 def __len__(self):
554 def __len__(self):
555 return len(self.index)
555 return len(self.index)
556 def __iter__(self):
556 def __iter__(self):
557 return iter(pycompat.xrange(len(self)))
557 return iter(pycompat.xrange(len(self)))
558 def revs(self, start=0, stop=None):
558 def revs(self, start=0, stop=None):
559 """iterate over all rev in this revlog (from start to stop)"""
559 """iterate over all rev in this revlog (from start to stop)"""
560 step = 1
560 step = 1
561 length = len(self)
561 length = len(self)
562 if stop is not None:
562 if stop is not None:
563 if start > stop:
563 if start > stop:
564 step = -1
564 step = -1
565 stop += step
565 stop += step
566 if stop > length:
566 if stop > length:
567 stop = length
567 stop = length
568 else:
568 else:
569 stop = length
569 stop = length
570 return pycompat.xrange(start, stop, step)
570 return pycompat.xrange(start, stop, step)
571
571
572 @util.propertycache
572 @util.propertycache
573 def nodemap(self):
573 def nodemap(self):
574 if self.index:
574 if self.index:
575 # populate mapping down to the initial node
575 # populate mapping down to the initial node
576 node0 = self.index[0][7] # get around changelog filtering
576 node0 = self.index[0][7] # get around changelog filtering
577 self.rev(node0)
577 self.rev(node0)
578 return self._nodecache
578 return self._nodecache
579
579
580 def hasnode(self, node):
580 def hasnode(self, node):
581 try:
581 try:
582 self.rev(node)
582 self.rev(node)
583 return True
583 return True
584 except KeyError:
584 except KeyError:
585 return False
585 return False
586
586
587 def candelta(self, baserev, rev):
587 def candelta(self, baserev, rev):
588 """whether two revisions (baserev, rev) can be delta-ed or not"""
588 """whether two revisions (baserev, rev) can be delta-ed or not"""
589 # Disable delta if either rev requires a content-changing flag
589 # Disable delta if either rev requires a content-changing flag
590 # processor (ex. LFS). This is because such flag processor can alter
590 # processor (ex. LFS). This is because such flag processor can alter
591 # the rawtext content that the delta will be based on, and two clients
591 # the rawtext content that the delta will be based on, and two clients
592 # could have a same revlog node with different flags (i.e. different
592 # could have a same revlog node with different flags (i.e. different
593 # rawtext contents) and the delta could be incompatible.
593 # rawtext contents) and the delta could be incompatible.
594 if ((self.flags(baserev) & REVIDX_RAWTEXT_CHANGING_FLAGS)
594 if ((self.flags(baserev) & REVIDX_RAWTEXT_CHANGING_FLAGS)
595 or (self.flags(rev) & REVIDX_RAWTEXT_CHANGING_FLAGS)):
595 or (self.flags(rev) & REVIDX_RAWTEXT_CHANGING_FLAGS)):
596 return False
596 return False
597 return True
597 return True
598
598
599 def clearcaches(self):
599 def clearcaches(self):
600 self._cache = None
600 self._cache = None
601 self._chainbasecache.clear()
601 self._chainbasecache.clear()
602 self._chunkcache = (0, '')
602 self._chunkcache = (0, '')
603 self._pcache = {}
603 self._pcache = {}
604
604
605 try:
605 try:
606 self._nodecache.clearcaches()
606 self._nodecache.clearcaches()
607 except AttributeError:
607 except AttributeError:
608 self._nodecache = {nullid: nullrev}
608 self._nodecache = {nullid: nullrev}
609 self._nodepos = None
609 self._nodepos = None
610
610
611 def rev(self, node):
611 def rev(self, node):
612 try:
612 try:
613 return self._nodecache[node]
613 return self._nodecache[node]
614 except TypeError:
614 except TypeError:
615 raise
615 raise
616 except error.RevlogError:
616 except error.RevlogError:
617 # parsers.c radix tree lookup failed
617 # parsers.c radix tree lookup failed
618 if node == wdirid or node in wdirfilenodeids:
618 if node == wdirid or node in wdirfilenodeids:
619 raise error.WdirUnsupported
619 raise error.WdirUnsupported
620 raise error.LookupError(node, self.indexfile, _('no node'))
620 raise error.LookupError(node, self.indexfile, _('no node'))
621 except KeyError:
621 except KeyError:
622 # pure python cache lookup failed
622 # pure python cache lookup failed
623 n = self._nodecache
623 n = self._nodecache
624 i = self.index
624 i = self.index
625 p = self._nodepos
625 p = self._nodepos
626 if p is None:
626 if p is None:
627 p = len(i) - 1
627 p = len(i) - 1
628 else:
628 else:
629 assert p < len(i)
629 assert p < len(i)
630 for r in pycompat.xrange(p, -1, -1):
630 for r in pycompat.xrange(p, -1, -1):
631 v = i[r][7]
631 v = i[r][7]
632 n[v] = r
632 n[v] = r
633 if v == node:
633 if v == node:
634 self._nodepos = r - 1
634 self._nodepos = r - 1
635 return r
635 return r
636 if node == wdirid or node in wdirfilenodeids:
636 if node == wdirid or node in wdirfilenodeids:
637 raise error.WdirUnsupported
637 raise error.WdirUnsupported
638 raise error.LookupError(node, self.indexfile, _('no node'))
638 raise error.LookupError(node, self.indexfile, _('no node'))
639
639
640 # Accessors for index entries.
640 # Accessors for index entries.
641
641
642 # First tuple entry is 8 bytes. First 6 bytes are offset. Last 2 bytes
642 # First tuple entry is 8 bytes. First 6 bytes are offset. Last 2 bytes
643 # are flags.
643 # are flags.
644 def start(self, rev):
644 def start(self, rev):
645 return int(self.index[rev][0] >> 16)
645 return int(self.index[rev][0] >> 16)
646
646
647 def flags(self, rev):
647 def flags(self, rev):
648 return self.index[rev][0] & 0xFFFF
648 return self.index[rev][0] & 0xFFFF
649
649
650 def length(self, rev):
650 def length(self, rev):
651 return self.index[rev][1]
651 return self.index[rev][1]
652
652
653 def rawsize(self, rev):
653 def rawsize(self, rev):
654 """return the length of the uncompressed text for a given revision"""
654 """return the length of the uncompressed text for a given revision"""
655 l = self.index[rev][2]
655 l = self.index[rev][2]
656 if l >= 0:
656 if l >= 0:
657 return l
657 return l
658
658
659 t = self.revision(rev, raw=True)
659 t = self.revision(rev, raw=True)
660 return len(t)
660 return len(t)
661
661
662 def size(self, rev):
662 def size(self, rev):
663 """length of non-raw text (processed by a "read" flag processor)"""
663 """length of non-raw text (processed by a "read" flag processor)"""
664 # fast path: if no "read" flag processor could change the content,
664 # fast path: if no "read" flag processor could change the content,
665 # size is rawsize. note: ELLIPSIS is known to not change the content.
665 # size is rawsize. note: ELLIPSIS is known to not change the content.
666 flags = self.flags(rev)
666 flags = self.flags(rev)
667 if flags & (REVIDX_KNOWN_FLAGS ^ REVIDX_ELLIPSIS) == 0:
667 if flags & (REVIDX_KNOWN_FLAGS ^ REVIDX_ELLIPSIS) == 0:
668 return self.rawsize(rev)
668 return self.rawsize(rev)
669
669
670 return len(self.revision(rev, raw=False))
670 return len(self.revision(rev, raw=False))
671
671
672 def chainbase(self, rev):
672 def chainbase(self, rev):
673 base = self._chainbasecache.get(rev)
673 base = self._chainbasecache.get(rev)
674 if base is not None:
674 if base is not None:
675 return base
675 return base
676
676
677 index = self.index
677 index = self.index
678 iterrev = rev
678 iterrev = rev
679 base = index[iterrev][3]
679 base = index[iterrev][3]
680 while base != iterrev:
680 while base != iterrev:
681 iterrev = base
681 iterrev = base
682 base = index[iterrev][3]
682 base = index[iterrev][3]
683
683
684 self._chainbasecache[rev] = base
684 self._chainbasecache[rev] = base
685 return base
685 return base
686
686
687 def linkrev(self, rev):
687 def linkrev(self, rev):
688 return self.index[rev][4]
688 return self.index[rev][4]
689
689
690 def parentrevs(self, rev):
690 def parentrevs(self, rev):
691 try:
691 try:
692 entry = self.index[rev]
692 entry = self.index[rev]
693 except IndexError:
693 except IndexError:
694 if rev == wdirrev:
694 if rev == wdirrev:
695 raise error.WdirUnsupported
695 raise error.WdirUnsupported
696 raise
696 raise
697
697
698 return entry[5], entry[6]
698 return entry[5], entry[6]
699
699
700 def node(self, rev):
700 def node(self, rev):
701 try:
701 try:
702 return self.index[rev][7]
702 return self.index[rev][7]
703 except IndexError:
703 except IndexError:
704 if rev == wdirrev:
704 if rev == wdirrev:
705 raise error.WdirUnsupported
705 raise error.WdirUnsupported
706 raise
706 raise
707
707
708 # Derived from index values.
708 # Derived from index values.
709
709
710 def end(self, rev):
710 def end(self, rev):
711 return self.start(rev) + self.length(rev)
711 return self.start(rev) + self.length(rev)
712
712
713 def parents(self, node):
713 def parents(self, node):
714 i = self.index
714 i = self.index
715 d = i[self.rev(node)]
715 d = i[self.rev(node)]
716 return i[d[5]][7], i[d[6]][7] # map revisions to nodes inline
716 return i[d[5]][7], i[d[6]][7] # map revisions to nodes inline
717
717
718 def chainlen(self, rev):
718 def chainlen(self, rev):
719 return self._chaininfo(rev)[0]
719 return self._chaininfo(rev)[0]
720
720
721 def _chaininfo(self, rev):
721 def _chaininfo(self, rev):
722 chaininfocache = self._chaininfocache
722 chaininfocache = self._chaininfocache
723 if rev in chaininfocache:
723 if rev in chaininfocache:
724 return chaininfocache[rev]
724 return chaininfocache[rev]
725 index = self.index
725 index = self.index
726 generaldelta = self._generaldelta
726 generaldelta = self._generaldelta
727 iterrev = rev
727 iterrev = rev
728 e = index[iterrev]
728 e = index[iterrev]
729 clen = 0
729 clen = 0
730 compresseddeltalen = 0
730 compresseddeltalen = 0
731 while iterrev != e[3]:
731 while iterrev != e[3]:
732 clen += 1
732 clen += 1
733 compresseddeltalen += e[1]
733 compresseddeltalen += e[1]
734 if generaldelta:
734 if generaldelta:
735 iterrev = e[3]
735 iterrev = e[3]
736 else:
736 else:
737 iterrev -= 1
737 iterrev -= 1
738 if iterrev in chaininfocache:
738 if iterrev in chaininfocache:
739 t = chaininfocache[iterrev]
739 t = chaininfocache[iterrev]
740 clen += t[0]
740 clen += t[0]
741 compresseddeltalen += t[1]
741 compresseddeltalen += t[1]
742 break
742 break
743 e = index[iterrev]
743 e = index[iterrev]
744 else:
744 else:
745 # Add text length of base since decompressing that also takes
745 # Add text length of base since decompressing that also takes
746 # work. For cache hits the length is already included.
746 # work. For cache hits the length is already included.
747 compresseddeltalen += e[1]
747 compresseddeltalen += e[1]
748 r = (clen, compresseddeltalen)
748 r = (clen, compresseddeltalen)
749 chaininfocache[rev] = r
749 chaininfocache[rev] = r
750 return r
750 return r
751
751
752 def _deltachain(self, rev, stoprev=None):
752 def _deltachain(self, rev, stoprev=None):
753 """Obtain the delta chain for a revision.
753 """Obtain the delta chain for a revision.
754
754
755 ``stoprev`` specifies a revision to stop at. If not specified, we
755 ``stoprev`` specifies a revision to stop at. If not specified, we
756 stop at the base of the chain.
756 stop at the base of the chain.
757
757
758 Returns a 2-tuple of (chain, stopped) where ``chain`` is a list of
758 Returns a 2-tuple of (chain, stopped) where ``chain`` is a list of
759 revs in ascending order and ``stopped`` is a bool indicating whether
759 revs in ascending order and ``stopped`` is a bool indicating whether
760 ``stoprev`` was hit.
760 ``stoprev`` was hit.
761 """
761 """
762 # Try C implementation.
762 # Try C implementation.
763 try:
763 try:
764 return self.index.deltachain(rev, stoprev, self._generaldelta)
764 return self.index.deltachain(rev, stoprev, self._generaldelta)
765 except AttributeError:
765 except AttributeError:
766 pass
766 pass
767
767
768 chain = []
768 chain = []
769
769
770 # Alias to prevent attribute lookup in tight loop.
770 # Alias to prevent attribute lookup in tight loop.
771 index = self.index
771 index = self.index
772 generaldelta = self._generaldelta
772 generaldelta = self._generaldelta
773
773
774 iterrev = rev
774 iterrev = rev
775 e = index[iterrev]
775 e = index[iterrev]
776 while iterrev != e[3] and iterrev != stoprev:
776 while iterrev != e[3] and iterrev != stoprev:
777 chain.append(iterrev)
777 chain.append(iterrev)
778 if generaldelta:
778 if generaldelta:
779 iterrev = e[3]
779 iterrev = e[3]
780 else:
780 else:
781 iterrev -= 1
781 iterrev -= 1
782 e = index[iterrev]
782 e = index[iterrev]
783
783
784 if iterrev == stoprev:
784 if iterrev == stoprev:
785 stopped = True
785 stopped = True
786 else:
786 else:
787 chain.append(iterrev)
787 chain.append(iterrev)
788 stopped = False
788 stopped = False
789
789
790 chain.reverse()
790 chain.reverse()
791 return chain, stopped
791 return chain, stopped
792
792
793 def ancestors(self, revs, stoprev=0, inclusive=False):
793 def ancestors(self, revs, stoprev=0, inclusive=False):
794 """Generate the ancestors of 'revs' in reverse topological order.
794 """Generate the ancestors of 'revs' in reverse topological order.
795 Does not generate revs lower than stoprev.
795 Does not generate revs lower than stoprev.
796
796
797 See the documentation for ancestor.lazyancestors for more details."""
797 See the documentation for ancestor.lazyancestors for more details."""
798
798
799 return ancestor.lazyancestors(self.parentrevs, revs, stoprev=stoprev,
799 return ancestor.lazyancestors(self.parentrevs, revs, stoprev=stoprev,
800 inclusive=inclusive)
800 inclusive=inclusive)
801
801
802 def descendants(self, revs):
802 def descendants(self, revs):
803 """Generate the descendants of 'revs' in revision order.
803 """Generate the descendants of 'revs' in revision order.
804
804
805 Yield a sequence of revision numbers starting with a child of
805 Yield a sequence of revision numbers starting with a child of
806 some rev in revs, i.e., each revision is *not* considered a
806 some rev in revs, i.e., each revision is *not* considered a
807 descendant of itself. Results are ordered by revision number (a
807 descendant of itself. Results are ordered by revision number (a
808 topological sort)."""
808 topological sort)."""
809 first = min(revs)
809 first = min(revs)
810 if first == nullrev:
810 if first == nullrev:
811 for i in self:
811 for i in self:
812 yield i
812 yield i
813 return
813 return
814
814
815 seen = set(revs)
815 seen = set(revs)
816 for i in self.revs(start=first + 1):
816 for i in self.revs(start=first + 1):
817 for x in self.parentrevs(i):
817 for x in self.parentrevs(i):
818 if x != nullrev and x in seen:
818 if x != nullrev and x in seen:
819 seen.add(i)
819 seen.add(i)
820 yield i
820 yield i
821 break
821 break
822
822
823 def findcommonmissing(self, common=None, heads=None):
823 def findcommonmissing(self, common=None, heads=None):
824 """Return a tuple of the ancestors of common and the ancestors of heads
824 """Return a tuple of the ancestors of common and the ancestors of heads
825 that are not ancestors of common. In revset terminology, we return the
825 that are not ancestors of common. In revset terminology, we return the
826 tuple:
826 tuple:
827
827
828 ::common, (::heads) - (::common)
828 ::common, (::heads) - (::common)
829
829
830 The list is sorted by revision number, meaning it is
830 The list is sorted by revision number, meaning it is
831 topologically sorted.
831 topologically sorted.
832
832
833 'heads' and 'common' are both lists of node IDs. If heads is
833 'heads' and 'common' are both lists of node IDs. If heads is
834 not supplied, uses all of the revlog's heads. If common is not
834 not supplied, uses all of the revlog's heads. If common is not
835 supplied, uses nullid."""
835 supplied, uses nullid."""
836 if common is None:
836 if common is None:
837 common = [nullid]
837 common = [nullid]
838 if heads is None:
838 if heads is None:
839 heads = self.heads()
839 heads = self.heads()
840
840
841 common = [self.rev(n) for n in common]
841 common = [self.rev(n) for n in common]
842 heads = [self.rev(n) for n in heads]
842 heads = [self.rev(n) for n in heads]
843
843
844 # we want the ancestors, but inclusive
844 # we want the ancestors, but inclusive
845 class lazyset(object):
845 class lazyset(object):
846 def __init__(self, lazyvalues):
846 def __init__(self, lazyvalues):
847 self.addedvalues = set()
847 self.addedvalues = set()
848 self.lazyvalues = lazyvalues
848 self.lazyvalues = lazyvalues
849
849
850 def __contains__(self, value):
850 def __contains__(self, value):
851 return value in self.addedvalues or value in self.lazyvalues
851 return value in self.addedvalues or value in self.lazyvalues
852
852
853 def __iter__(self):
853 def __iter__(self):
854 added = self.addedvalues
854 added = self.addedvalues
855 for r in added:
855 for r in added:
856 yield r
856 yield r
857 for r in self.lazyvalues:
857 for r in self.lazyvalues:
858 if not r in added:
858 if not r in added:
859 yield r
859 yield r
860
860
861 def add(self, value):
861 def add(self, value):
862 self.addedvalues.add(value)
862 self.addedvalues.add(value)
863
863
864 def update(self, values):
864 def update(self, values):
865 self.addedvalues.update(values)
865 self.addedvalues.update(values)
866
866
867 has = lazyset(self.ancestors(common))
867 has = lazyset(self.ancestors(common))
868 has.add(nullrev)
868 has.add(nullrev)
869 has.update(common)
869 has.update(common)
870
870
871 # take all ancestors from heads that aren't in has
871 # take all ancestors from heads that aren't in has
872 missing = set()
872 missing = set()
873 visit = collections.deque(r for r in heads if r not in has)
873 visit = collections.deque(r for r in heads if r not in has)
874 while visit:
874 while visit:
875 r = visit.popleft()
875 r = visit.popleft()
876 if r in missing:
876 if r in missing:
877 continue
877 continue
878 else:
878 else:
879 missing.add(r)
879 missing.add(r)
880 for p in self.parentrevs(r):
880 for p in self.parentrevs(r):
881 if p not in has:
881 if p not in has:
882 visit.append(p)
882 visit.append(p)
883 missing = list(missing)
883 missing = list(missing)
884 missing.sort()
884 missing.sort()
885 return has, [self.node(miss) for miss in missing]
885 return has, [self.node(miss) for miss in missing]
886
886
887 def incrementalmissingrevs(self, common=None):
887 def incrementalmissingrevs(self, common=None):
888 """Return an object that can be used to incrementally compute the
888 """Return an object that can be used to incrementally compute the
889 revision numbers of the ancestors of arbitrary sets that are not
889 revision numbers of the ancestors of arbitrary sets that are not
890 ancestors of common. This is an ancestor.incrementalmissingancestors
890 ancestors of common. This is an ancestor.incrementalmissingancestors
891 object.
891 object.
892
892
893 'common' is a list of revision numbers. If common is not supplied, uses
893 'common' is a list of revision numbers. If common is not supplied, uses
894 nullrev.
894 nullrev.
895 """
895 """
896 if common is None:
896 if common is None:
897 common = [nullrev]
897 common = [nullrev]
898
898
899 return ancestor.incrementalmissingancestors(self.parentrevs, common)
899 return ancestor.incrementalmissingancestors(self.parentrevs, common)
900
900
901 def findmissingrevs(self, common=None, heads=None):
901 def findmissingrevs(self, common=None, heads=None):
902 """Return the revision numbers of the ancestors of heads that
902 """Return the revision numbers of the ancestors of heads that
903 are not ancestors of common.
903 are not ancestors of common.
904
904
905 More specifically, return a list of revision numbers corresponding to
905 More specifically, return a list of revision numbers corresponding to
906 nodes N such that every N satisfies the following constraints:
906 nodes N such that every N satisfies the following constraints:
907
907
908 1. N is an ancestor of some node in 'heads'
908 1. N is an ancestor of some node in 'heads'
909 2. N is not an ancestor of any node in 'common'
909 2. N is not an ancestor of any node in 'common'
910
910
911 The list is sorted by revision number, meaning it is
911 The list is sorted by revision number, meaning it is
912 topologically sorted.
912 topologically sorted.
913
913
914 'heads' and 'common' are both lists of revision numbers. If heads is
914 'heads' and 'common' are both lists of revision numbers. If heads is
915 not supplied, uses all of the revlog's heads. If common is not
915 not supplied, uses all of the revlog's heads. If common is not
916 supplied, uses nullid."""
916 supplied, uses nullid."""
917 if common is None:
917 if common is None:
918 common = [nullrev]
918 common = [nullrev]
919 if heads is None:
919 if heads is None:
920 heads = self.headrevs()
920 heads = self.headrevs()
921
921
922 inc = self.incrementalmissingrevs(common=common)
922 inc = self.incrementalmissingrevs(common=common)
923 return inc.missingancestors(heads)
923 return inc.missingancestors(heads)
924
924
925 def findmissing(self, common=None, heads=None):
925 def findmissing(self, common=None, heads=None):
926 """Return the ancestors of heads that are not ancestors of common.
926 """Return the ancestors of heads that are not ancestors of common.
927
927
928 More specifically, return a list of nodes N such that every N
928 More specifically, return a list of nodes N such that every N
929 satisfies the following constraints:
929 satisfies the following constraints:
930
930
931 1. N is an ancestor of some node in 'heads'
931 1. N is an ancestor of some node in 'heads'
932 2. N is not an ancestor of any node in 'common'
932 2. N is not an ancestor of any node in 'common'
933
933
934 The list is sorted by revision number, meaning it is
934 The list is sorted by revision number, meaning it is
935 topologically sorted.
935 topologically sorted.
936
936
937 'heads' and 'common' are both lists of node IDs. If heads is
937 'heads' and 'common' are both lists of node IDs. If heads is
938 not supplied, uses all of the revlog's heads. If common is not
938 not supplied, uses all of the revlog's heads. If common is not
939 supplied, uses nullid."""
939 supplied, uses nullid."""
940 if common is None:
940 if common is None:
941 common = [nullid]
941 common = [nullid]
942 if heads is None:
942 if heads is None:
943 heads = self.heads()
943 heads = self.heads()
944
944
945 common = [self.rev(n) for n in common]
945 common = [self.rev(n) for n in common]
946 heads = [self.rev(n) for n in heads]
946 heads = [self.rev(n) for n in heads]
947
947
948 inc = self.incrementalmissingrevs(common=common)
948 inc = self.incrementalmissingrevs(common=common)
949 return [self.node(r) for r in inc.missingancestors(heads)]
949 return [self.node(r) for r in inc.missingancestors(heads)]
950
950
951 def nodesbetween(self, roots=None, heads=None):
951 def nodesbetween(self, roots=None, heads=None):
952 """Return a topological path from 'roots' to 'heads'.
952 """Return a topological path from 'roots' to 'heads'.
953
953
954 Return a tuple (nodes, outroots, outheads) where 'nodes' is a
954 Return a tuple (nodes, outroots, outheads) where 'nodes' is a
955 topologically sorted list of all nodes N that satisfy both of
955 topologically sorted list of all nodes N that satisfy both of
956 these constraints:
956 these constraints:
957
957
958 1. N is a descendant of some node in 'roots'
958 1. N is a descendant of some node in 'roots'
959 2. N is an ancestor of some node in 'heads'
959 2. N is an ancestor of some node in 'heads'
960
960
961 Every node is considered to be both a descendant and an ancestor
961 Every node is considered to be both a descendant and an ancestor
962 of itself, so every reachable node in 'roots' and 'heads' will be
962 of itself, so every reachable node in 'roots' and 'heads' will be
963 included in 'nodes'.
963 included in 'nodes'.
964
964
965 'outroots' is the list of reachable nodes in 'roots', i.e., the
965 'outroots' is the list of reachable nodes in 'roots', i.e., the
966 subset of 'roots' that is returned in 'nodes'. Likewise,
966 subset of 'roots' that is returned in 'nodes'. Likewise,
967 'outheads' is the subset of 'heads' that is also in 'nodes'.
967 'outheads' is the subset of 'heads' that is also in 'nodes'.
968
968
969 'roots' and 'heads' are both lists of node IDs. If 'roots' is
969 'roots' and 'heads' are both lists of node IDs. If 'roots' is
970 unspecified, uses nullid as the only root. If 'heads' is
970 unspecified, uses nullid as the only root. If 'heads' is
971 unspecified, uses list of all of the revlog's heads."""
971 unspecified, uses list of all of the revlog's heads."""
972 nonodes = ([], [], [])
972 nonodes = ([], [], [])
973 if roots is not None:
973 if roots is not None:
974 roots = list(roots)
974 roots = list(roots)
975 if not roots:
975 if not roots:
976 return nonodes
976 return nonodes
977 lowestrev = min([self.rev(n) for n in roots])
977 lowestrev = min([self.rev(n) for n in roots])
978 else:
978 else:
979 roots = [nullid] # Everybody's a descendant of nullid
979 roots = [nullid] # Everybody's a descendant of nullid
980 lowestrev = nullrev
980 lowestrev = nullrev
981 if (lowestrev == nullrev) and (heads is None):
981 if (lowestrev == nullrev) and (heads is None):
982 # We want _all_ the nodes!
982 # We want _all_ the nodes!
983 return ([self.node(r) for r in self], [nullid], list(self.heads()))
983 return ([self.node(r) for r in self], [nullid], list(self.heads()))
984 if heads is None:
984 if heads is None:
985 # All nodes are ancestors, so the latest ancestor is the last
985 # All nodes are ancestors, so the latest ancestor is the last
986 # node.
986 # node.
987 highestrev = len(self) - 1
987 highestrev = len(self) - 1
988 # Set ancestors to None to signal that every node is an ancestor.
988 # Set ancestors to None to signal that every node is an ancestor.
989 ancestors = None
989 ancestors = None
990 # Set heads to an empty dictionary for later discovery of heads
990 # Set heads to an empty dictionary for later discovery of heads
991 heads = {}
991 heads = {}
992 else:
992 else:
993 heads = list(heads)
993 heads = list(heads)
994 if not heads:
994 if not heads:
995 return nonodes
995 return nonodes
996 ancestors = set()
996 ancestors = set()
997 # Turn heads into a dictionary so we can remove 'fake' heads.
997 # Turn heads into a dictionary so we can remove 'fake' heads.
998 # Also, later we will be using it to filter out the heads we can't
998 # Also, later we will be using it to filter out the heads we can't
999 # find from roots.
999 # find from roots.
1000 heads = dict.fromkeys(heads, False)
1000 heads = dict.fromkeys(heads, False)
1001 # Start at the top and keep marking parents until we're done.
1001 # Start at the top and keep marking parents until we're done.
1002 nodestotag = set(heads)
1002 nodestotag = set(heads)
1003 # Remember where the top was so we can use it as a limit later.
1003 # Remember where the top was so we can use it as a limit later.
1004 highestrev = max([self.rev(n) for n in nodestotag])
1004 highestrev = max([self.rev(n) for n in nodestotag])
1005 while nodestotag:
1005 while nodestotag:
1006 # grab a node to tag
1006 # grab a node to tag
1007 n = nodestotag.pop()
1007 n = nodestotag.pop()
1008 # Never tag nullid
1008 # Never tag nullid
1009 if n == nullid:
1009 if n == nullid:
1010 continue
1010 continue
1011 # A node's revision number represents its place in a
1011 # A node's revision number represents its place in a
1012 # topologically sorted list of nodes.
1012 # topologically sorted list of nodes.
1013 r = self.rev(n)
1013 r = self.rev(n)
1014 if r >= lowestrev:
1014 if r >= lowestrev:
1015 if n not in ancestors:
1015 if n not in ancestors:
1016 # If we are possibly a descendant of one of the roots
1016 # If we are possibly a descendant of one of the roots
1017 # and we haven't already been marked as an ancestor
1017 # and we haven't already been marked as an ancestor
1018 ancestors.add(n) # Mark as ancestor
1018 ancestors.add(n) # Mark as ancestor
1019 # Add non-nullid parents to list of nodes to tag.
1019 # Add non-nullid parents to list of nodes to tag.
1020 nodestotag.update([p for p in self.parents(n) if
1020 nodestotag.update([p for p in self.parents(n) if
1021 p != nullid])
1021 p != nullid])
1022 elif n in heads: # We've seen it before, is it a fake head?
1022 elif n in heads: # We've seen it before, is it a fake head?
1023 # So it is, real heads should not be the ancestors of
1023 # So it is, real heads should not be the ancestors of
1024 # any other heads.
1024 # any other heads.
1025 heads.pop(n)
1025 heads.pop(n)
1026 if not ancestors:
1026 if not ancestors:
1027 return nonodes
1027 return nonodes
1028 # Now that we have our set of ancestors, we want to remove any
1028 # Now that we have our set of ancestors, we want to remove any
1029 # roots that are not ancestors.
1029 # roots that are not ancestors.
1030
1030
1031 # If one of the roots was nullid, everything is included anyway.
1031 # If one of the roots was nullid, everything is included anyway.
1032 if lowestrev > nullrev:
1032 if lowestrev > nullrev:
1033 # But, since we weren't, let's recompute the lowest rev to not
1033 # But, since we weren't, let's recompute the lowest rev to not
1034 # include roots that aren't ancestors.
1034 # include roots that aren't ancestors.
1035
1035
1036 # Filter out roots that aren't ancestors of heads
1036 # Filter out roots that aren't ancestors of heads
1037 roots = [root for root in roots if root in ancestors]
1037 roots = [root for root in roots if root in ancestors]
1038 # Recompute the lowest revision
1038 # Recompute the lowest revision
1039 if roots:
1039 if roots:
1040 lowestrev = min([self.rev(root) for root in roots])
1040 lowestrev = min([self.rev(root) for root in roots])
1041 else:
1041 else:
1042 # No more roots? Return empty list
1042 # No more roots? Return empty list
1043 return nonodes
1043 return nonodes
1044 else:
1044 else:
1045 # We are descending from nullid, and don't need to care about
1045 # We are descending from nullid, and don't need to care about
1046 # any other roots.
1046 # any other roots.
1047 lowestrev = nullrev
1047 lowestrev = nullrev
1048 roots = [nullid]
1048 roots = [nullid]
1049 # Transform our roots list into a set.
1049 # Transform our roots list into a set.
1050 descendants = set(roots)
1050 descendants = set(roots)
1051 # Also, keep the original roots so we can filter out roots that aren't
1051 # Also, keep the original roots so we can filter out roots that aren't
1052 # 'real' roots (i.e. are descended from other roots).
1052 # 'real' roots (i.e. are descended from other roots).
1053 roots = descendants.copy()
1053 roots = descendants.copy()
1054 # Our topologically sorted list of output nodes.
1054 # Our topologically sorted list of output nodes.
1055 orderedout = []
1055 orderedout = []
1056 # Don't start at nullid since we don't want nullid in our output list,
1056 # Don't start at nullid since we don't want nullid in our output list,
1057 # and if nullid shows up in descendants, empty parents will look like
1057 # and if nullid shows up in descendants, empty parents will look like
1058 # they're descendants.
1058 # they're descendants.
1059 for r in self.revs(start=max(lowestrev, 0), stop=highestrev + 1):
1059 for r in self.revs(start=max(lowestrev, 0), stop=highestrev + 1):
1060 n = self.node(r)
1060 n = self.node(r)
1061 isdescendant = False
1061 isdescendant = False
1062 if lowestrev == nullrev: # Everybody is a descendant of nullid
1062 if lowestrev == nullrev: # Everybody is a descendant of nullid
1063 isdescendant = True
1063 isdescendant = True
1064 elif n in descendants:
1064 elif n in descendants:
1065 # n is already a descendant
1065 # n is already a descendant
1066 isdescendant = True
1066 isdescendant = True
1067 # This check only needs to be done here because all the roots
1067 # This check only needs to be done here because all the roots
1068 # will start being marked is descendants before the loop.
1068 # will start being marked is descendants before the loop.
1069 if n in roots:
1069 if n in roots:
1070 # If n was a root, check if it's a 'real' root.
1070 # If n was a root, check if it's a 'real' root.
1071 p = tuple(self.parents(n))
1071 p = tuple(self.parents(n))
1072 # If any of its parents are descendants, it's not a root.
1072 # If any of its parents are descendants, it's not a root.
1073 if (p[0] in descendants) or (p[1] in descendants):
1073 if (p[0] in descendants) or (p[1] in descendants):
1074 roots.remove(n)
1074 roots.remove(n)
1075 else:
1075 else:
1076 p = tuple(self.parents(n))
1076 p = tuple(self.parents(n))
1077 # A node is a descendant if either of its parents are
1077 # A node is a descendant if either of its parents are
1078 # descendants. (We seeded the dependents list with the roots
1078 # descendants. (We seeded the dependents list with the roots
1079 # up there, remember?)
1079 # up there, remember?)
1080 if (p[0] in descendants) or (p[1] in descendants):
1080 if (p[0] in descendants) or (p[1] in descendants):
1081 descendants.add(n)
1081 descendants.add(n)
1082 isdescendant = True
1082 isdescendant = True
1083 if isdescendant and ((ancestors is None) or (n in ancestors)):
1083 if isdescendant and ((ancestors is None) or (n in ancestors)):
1084 # Only include nodes that are both descendants and ancestors.
1084 # Only include nodes that are both descendants and ancestors.
1085 orderedout.append(n)
1085 orderedout.append(n)
1086 if (ancestors is not None) and (n in heads):
1086 if (ancestors is not None) and (n in heads):
1087 # We're trying to figure out which heads are reachable
1087 # We're trying to figure out which heads are reachable
1088 # from roots.
1088 # from roots.
1089 # Mark this head as having been reached
1089 # Mark this head as having been reached
1090 heads[n] = True
1090 heads[n] = True
1091 elif ancestors is None:
1091 elif ancestors is None:
1092 # Otherwise, we're trying to discover the heads.
1092 # Otherwise, we're trying to discover the heads.
1093 # Assume this is a head because if it isn't, the next step
1093 # Assume this is a head because if it isn't, the next step
1094 # will eventually remove it.
1094 # will eventually remove it.
1095 heads[n] = True
1095 heads[n] = True
1096 # But, obviously its parents aren't.
1096 # But, obviously its parents aren't.
1097 for p in self.parents(n):
1097 for p in self.parents(n):
1098 heads.pop(p, None)
1098 heads.pop(p, None)
1099 heads = [head for head, flag in heads.iteritems() if flag]
1099 heads = [head for head, flag in heads.iteritems() if flag]
1100 roots = list(roots)
1100 roots = list(roots)
1101 assert orderedout
1101 assert orderedout
1102 assert roots
1102 assert roots
1103 assert heads
1103 assert heads
1104 return (orderedout, roots, heads)
1104 return (orderedout, roots, heads)
1105
1105
1106 def headrevs(self):
1106 def headrevs(self):
1107 try:
1107 try:
1108 return self.index.headrevs()
1108 return self.index.headrevs()
1109 except AttributeError:
1109 except AttributeError:
1110 return self._headrevs()
1110 return self._headrevs()
1111
1111
1112 def computephases(self, roots):
1112 def computephases(self, roots):
1113 return self.index.computephasesmapsets(roots)
1113 return self.index.computephasesmapsets(roots)
1114
1114
1115 def _headrevs(self):
1115 def _headrevs(self):
1116 count = len(self)
1116 count = len(self)
1117 if not count:
1117 if not count:
1118 return [nullrev]
1118 return [nullrev]
1119 # we won't iter over filtered rev so nobody is a head at start
1119 # we won't iter over filtered rev so nobody is a head at start
1120 ishead = [0] * (count + 1)
1120 ishead = [0] * (count + 1)
1121 index = self.index
1121 index = self.index
1122 for r in self:
1122 for r in self:
1123 ishead[r] = 1 # I may be an head
1123 ishead[r] = 1 # I may be an head
1124 e = index[r]
1124 e = index[r]
1125 ishead[e[5]] = ishead[e[6]] = 0 # my parent are not
1125 ishead[e[5]] = ishead[e[6]] = 0 # my parent are not
1126 return [r for r, val in enumerate(ishead) if val]
1126 return [r for r, val in enumerate(ishead) if val]
1127
1127
1128 def heads(self, start=None, stop=None):
1128 def heads(self, start=None, stop=None):
1129 """return the list of all nodes that have no children
1129 """return the list of all nodes that have no children
1130
1130
1131 if start is specified, only heads that are descendants of
1131 if start is specified, only heads that are descendants of
1132 start will be returned
1132 start will be returned
1133 if stop is specified, it will consider all the revs from stop
1133 if stop is specified, it will consider all the revs from stop
1134 as if they had no children
1134 as if they had no children
1135 """
1135 """
1136 if start is None and stop is None:
1136 if start is None and stop is None:
1137 if not len(self):
1137 if not len(self):
1138 return [nullid]
1138 return [nullid]
1139 return [self.node(r) for r in self.headrevs()]
1139 return [self.node(r) for r in self.headrevs()]
1140
1140
1141 if start is None:
1141 if start is None:
1142 start = nullid
1142 start = nullid
1143 if stop is None:
1143 if stop is None:
1144 stop = []
1144 stop = []
1145 stoprevs = set([self.rev(n) for n in stop])
1145 stoprevs = set([self.rev(n) for n in stop])
1146 startrev = self.rev(start)
1146 startrev = self.rev(start)
1147 reachable = {startrev}
1147 reachable = {startrev}
1148 heads = {startrev}
1148 heads = {startrev}
1149
1149
1150 parentrevs = self.parentrevs
1150 parentrevs = self.parentrevs
1151 for r in self.revs(start=startrev + 1):
1151 for r in self.revs(start=startrev + 1):
1152 for p in parentrevs(r):
1152 for p in parentrevs(r):
1153 if p in reachable:
1153 if p in reachable:
1154 if r not in stoprevs:
1154 if r not in stoprevs:
1155 reachable.add(r)
1155 reachable.add(r)
1156 heads.add(r)
1156 heads.add(r)
1157 if p in heads and p not in stoprevs:
1157 if p in heads and p not in stoprevs:
1158 heads.remove(p)
1158 heads.remove(p)
1159
1159
1160 return [self.node(r) for r in heads]
1160 return [self.node(r) for r in heads]
1161
1161
1162 def children(self, node):
1162 def children(self, node):
1163 """find the children of a given node"""
1163 """find the children of a given node"""
1164 c = []
1164 c = []
1165 p = self.rev(node)
1165 p = self.rev(node)
1166 for r in self.revs(start=p + 1):
1166 for r in self.revs(start=p + 1):
1167 prevs = [pr for pr in self.parentrevs(r) if pr != nullrev]
1167 prevs = [pr for pr in self.parentrevs(r) if pr != nullrev]
1168 if prevs:
1168 if prevs:
1169 for pr in prevs:
1169 for pr in prevs:
1170 if pr == p:
1170 if pr == p:
1171 c.append(self.node(r))
1171 c.append(self.node(r))
1172 elif p == nullrev:
1172 elif p == nullrev:
1173 c.append(self.node(r))
1173 c.append(self.node(r))
1174 return c
1174 return c
1175
1175
1176 def commonancestorsheads(self, a, b):
1176 def commonancestorsheads(self, a, b):
1177 """calculate all the heads of the common ancestors of nodes a and b"""
1177 """calculate all the heads of the common ancestors of nodes a and b"""
1178 a, b = self.rev(a), self.rev(b)
1178 a, b = self.rev(a), self.rev(b)
1179 ancs = self._commonancestorsheads(a, b)
1179 ancs = self._commonancestorsheads(a, b)
1180 return pycompat.maplist(self.node, ancs)
1180 return pycompat.maplist(self.node, ancs)
1181
1181
1182 def _commonancestorsheads(self, *revs):
1182 def _commonancestorsheads(self, *revs):
1183 """calculate all the heads of the common ancestors of revs"""
1183 """calculate all the heads of the common ancestors of revs"""
1184 try:
1184 try:
1185 ancs = self.index.commonancestorsheads(*revs)
1185 ancs = self.index.commonancestorsheads(*revs)
1186 except (AttributeError, OverflowError): # C implementation failed
1186 except (AttributeError, OverflowError): # C implementation failed
1187 ancs = ancestor.commonancestorsheads(self.parentrevs, *revs)
1187 ancs = ancestor.commonancestorsheads(self.parentrevs, *revs)
1188 return ancs
1188 return ancs
1189
1189
1190 def isancestor(self, a, b):
1190 def isancestor(self, a, b):
1191 """return True if node a is an ancestor of node b
1191 """return True if node a is an ancestor of node b
1192
1192
1193 A revision is considered an ancestor of itself."""
1193 A revision is considered an ancestor of itself."""
1194 a, b = self.rev(a), self.rev(b)
1194 a, b = self.rev(a), self.rev(b)
1195 return self.isancestorrev(a, b)
1195 return self.isancestorrev(a, b)
1196
1196
1197 def isancestorrev(self, a, b):
1197 def isancestorrev(self, a, b):
1198 """return True if revision a is an ancestor of revision b
1198 """return True if revision a is an ancestor of revision b
1199
1199
1200 A revision is considered an ancestor of itself.
1200 A revision is considered an ancestor of itself.
1201
1201
1202 The implementation of this is trivial but the use of
1202 The implementation of this is trivial but the use of
1203 commonancestorsheads is not."""
1203 commonancestorsheads is not."""
1204 if a == nullrev:
1204 if a == nullrev:
1205 return True
1205 return True
1206 elif a == b:
1206 elif a == b:
1207 return True
1207 return True
1208 elif a > b:
1208 elif a > b:
1209 return False
1209 return False
1210 return a in self._commonancestorsheads(a, b)
1210 return a in self._commonancestorsheads(a, b)
1211
1211
1212 def ancestor(self, a, b):
1212 def ancestor(self, a, b):
1213 """calculate the "best" common ancestor of nodes a and b"""
1213 """calculate the "best" common ancestor of nodes a and b"""
1214
1214
1215 a, b = self.rev(a), self.rev(b)
1215 a, b = self.rev(a), self.rev(b)
1216 try:
1216 try:
1217 ancs = self.index.ancestors(a, b)
1217 ancs = self.index.ancestors(a, b)
1218 except (AttributeError, OverflowError):
1218 except (AttributeError, OverflowError):
1219 ancs = ancestor.ancestors(self.parentrevs, a, b)
1219 ancs = ancestor.ancestors(self.parentrevs, a, b)
1220 if ancs:
1220 if ancs:
1221 # choose a consistent winner when there's a tie
1221 # choose a consistent winner when there's a tie
1222 return min(map(self.node, ancs))
1222 return min(map(self.node, ancs))
1223 return nullid
1223 return nullid
1224
1224
1225 def _match(self, id):
1225 def _match(self, id):
1226 if isinstance(id, int):
1226 if isinstance(id, int):
1227 # rev
1227 # rev
1228 return self.node(id)
1228 return self.node(id)
1229 if len(id) == 20:
1229 if len(id) == 20:
1230 # possibly a binary node
1230 # possibly a binary node
1231 # odds of a binary node being all hex in ASCII are 1 in 10**25
1231 # odds of a binary node being all hex in ASCII are 1 in 10**25
1232 try:
1232 try:
1233 node = id
1233 node = id
1234 self.rev(node) # quick search the index
1234 self.rev(node) # quick search the index
1235 return node
1235 return node
1236 except error.LookupError:
1236 except error.LookupError:
1237 pass # may be partial hex id
1237 pass # may be partial hex id
1238 try:
1238 try:
1239 # str(rev)
1239 # str(rev)
1240 rev = int(id)
1240 rev = int(id)
1241 if "%d" % rev != id:
1241 if "%d" % rev != id:
1242 raise ValueError
1242 raise ValueError
1243 if rev < 0:
1243 if rev < 0:
1244 rev = len(self) + rev
1244 rev = len(self) + rev
1245 if rev < 0 or rev >= len(self):
1245 if rev < 0 or rev >= len(self):
1246 raise ValueError
1246 raise ValueError
1247 return self.node(rev)
1247 return self.node(rev)
1248 except (ValueError, OverflowError):
1248 except (ValueError, OverflowError):
1249 pass
1249 pass
1250 if len(id) == 40:
1250 if len(id) == 40:
1251 try:
1251 try:
1252 # a full hex nodeid?
1252 # a full hex nodeid?
1253 node = bin(id)
1253 node = bin(id)
1254 self.rev(node)
1254 self.rev(node)
1255 return node
1255 return node
1256 except (TypeError, error.LookupError):
1256 except (TypeError, error.LookupError):
1257 pass
1257 pass
1258
1258
1259 def _partialmatch(self, id):
1259 def _partialmatch(self, id):
1260 # we don't care wdirfilenodeids as they should be always full hash
1260 # we don't care wdirfilenodeids as they should be always full hash
1261 maybewdir = wdirhex.startswith(id)
1261 maybewdir = wdirhex.startswith(id)
1262 try:
1262 try:
1263 partial = self.index.partialmatch(id)
1263 partial = self.index.partialmatch(id)
1264 if partial and self.hasnode(partial):
1264 if partial and self.hasnode(partial):
1265 if maybewdir:
1265 if maybewdir:
1266 # single 'ff...' match in radix tree, ambiguous with wdir
1266 # single 'ff...' match in radix tree, ambiguous with wdir
1267 raise error.RevlogError
1267 raise error.RevlogError
1268 return partial
1268 return partial
1269 if maybewdir:
1269 if maybewdir:
1270 # no 'ff...' match in radix tree, wdir identified
1270 # no 'ff...' match in radix tree, wdir identified
1271 raise error.WdirUnsupported
1271 raise error.WdirUnsupported
1272 return None
1272 return None
1273 except error.RevlogError:
1273 except error.RevlogError:
1274 # parsers.c radix tree lookup gave multiple matches
1274 # parsers.c radix tree lookup gave multiple matches
1275 # fast path: for unfiltered changelog, radix tree is accurate
1275 # fast path: for unfiltered changelog, radix tree is accurate
1276 if not getattr(self, 'filteredrevs', None):
1276 if not getattr(self, 'filteredrevs', None):
1277 raise error.AmbiguousPrefixLookupError(
1277 raise error.AmbiguousPrefixLookupError(
1278 id, self.indexfile, _('ambiguous identifier'))
1278 id, self.indexfile, _('ambiguous identifier'))
1279 # fall through to slow path that filters hidden revisions
1279 # fall through to slow path that filters hidden revisions
1280 except (AttributeError, ValueError):
1280 except (AttributeError, ValueError):
1281 # we are pure python, or key was too short to search radix tree
1281 # we are pure python, or key was too short to search radix tree
1282 pass
1282 pass
1283
1283
1284 if id in self._pcache:
1284 if id in self._pcache:
1285 return self._pcache[id]
1285 return self._pcache[id]
1286
1286
1287 if len(id) <= 40:
1287 if len(id) <= 40:
1288 try:
1288 try:
1289 # hex(node)[:...]
1289 # hex(node)[:...]
1290 l = len(id) // 2 # grab an even number of digits
1290 l = len(id) // 2 # grab an even number of digits
1291 prefix = bin(id[:l * 2])
1291 prefix = bin(id[:l * 2])
1292 nl = [e[7] for e in self.index if e[7].startswith(prefix)]
1292 nl = [e[7] for e in self.index if e[7].startswith(prefix)]
1293 nl = [n for n in nl if hex(n).startswith(id) and
1293 nl = [n for n in nl if hex(n).startswith(id) and
1294 self.hasnode(n)]
1294 self.hasnode(n)]
1295 if nullhex.startswith(id):
1295 if nullhex.startswith(id):
1296 nl.append(nullid)
1296 nl.append(nullid)
1297 if len(nl) > 0:
1297 if len(nl) > 0:
1298 if len(nl) == 1 and not maybewdir:
1298 if len(nl) == 1 and not maybewdir:
1299 self._pcache[id] = nl[0]
1299 self._pcache[id] = nl[0]
1300 return nl[0]
1300 return nl[0]
1301 raise error.AmbiguousPrefixLookupError(
1301 raise error.AmbiguousPrefixLookupError(
1302 id, self.indexfile, _('ambiguous identifier'))
1302 id, self.indexfile, _('ambiguous identifier'))
1303 if maybewdir:
1303 if maybewdir:
1304 raise error.WdirUnsupported
1304 raise error.WdirUnsupported
1305 return None
1305 return None
1306 except TypeError:
1306 except TypeError:
1307 pass
1307 pass
1308
1308
1309 def lookup(self, id):
1309 def lookup(self, id):
1310 """locate a node based on:
1310 """locate a node based on:
1311 - revision number or str(revision number)
1311 - revision number or str(revision number)
1312 - nodeid or subset of hex nodeid
1312 - nodeid or subset of hex nodeid
1313 """
1313 """
1314 n = self._match(id)
1314 n = self._match(id)
1315 if n is not None:
1315 if n is not None:
1316 return n
1316 return n
1317 n = self._partialmatch(id)
1317 n = self._partialmatch(id)
1318 if n:
1318 if n:
1319 return n
1319 return n
1320
1320
1321 raise error.LookupError(id, self.indexfile, _('no match found'))
1321 raise error.LookupError(id, self.indexfile, _('no match found'))
1322
1322
1323 def shortest(self, node, minlength=1):
1323 def shortest(self, node, minlength=1):
1324 """Find the shortest unambiguous prefix that matches node."""
1324 """Find the shortest unambiguous prefix that matches node."""
1325 def isvalid(prefix):
1325 def isvalid(prefix):
1326 try:
1326 try:
1327 node = self._partialmatch(prefix)
1327 node = self._partialmatch(prefix)
1328 except error.RevlogError:
1328 except error.AmbiguousPrefixLookupError:
1329 return False
1329 return False
1330 except error.WdirUnsupported:
1330 except error.WdirUnsupported:
1331 # single 'ff...' match
1331 # single 'ff...' match
1332 return True
1332 return True
1333 if node is None:
1333 if node is None:
1334 raise error.LookupError(node, self.indexfile, _('no node'))
1334 raise error.LookupError(node, self.indexfile, _('no node'))
1335 return True
1335 return True
1336
1336
1337 def maybewdir(prefix):
1337 def maybewdir(prefix):
1338 return all(c == 'f' for c in prefix)
1338 return all(c == 'f' for c in prefix)
1339
1339
1340 hexnode = hex(node)
1340 hexnode = hex(node)
1341
1341
1342 def disambiguate(hexnode, minlength):
1342 def disambiguate(hexnode, minlength):
1343 """Disambiguate against wdirid."""
1343 """Disambiguate against wdirid."""
1344 for length in range(minlength, 41):
1344 for length in range(minlength, 41):
1345 prefix = hexnode[:length]
1345 prefix = hexnode[:length]
1346 if not maybewdir(prefix):
1346 if not maybewdir(prefix):
1347 return prefix
1347 return prefix
1348
1348
1349 if not getattr(self, 'filteredrevs', None):
1349 if not getattr(self, 'filteredrevs', None):
1350 try:
1350 try:
1351 length = max(self.index.shortest(node), minlength)
1351 length = max(self.index.shortest(node), minlength)
1352 return disambiguate(hexnode, length)
1352 return disambiguate(hexnode, length)
1353 except error.RevlogError:
1353 except error.RevlogError:
1354 if node != wdirid:
1354 if node != wdirid:
1355 raise error.LookupError(node, self.indexfile, _('no node'))
1355 raise error.LookupError(node, self.indexfile, _('no node'))
1356 except AttributeError:
1356 except AttributeError:
1357 # Fall through to pure code
1357 # Fall through to pure code
1358 pass
1358 pass
1359
1359
1360 if node == wdirid:
1360 if node == wdirid:
1361 for length in range(minlength, 41):
1361 for length in range(minlength, 41):
1362 prefix = hexnode[:length]
1362 prefix = hexnode[:length]
1363 if isvalid(prefix):
1363 if isvalid(prefix):
1364 return prefix
1364 return prefix
1365
1365
1366 for length in range(minlength, 41):
1366 for length in range(minlength, 41):
1367 prefix = hexnode[:length]
1367 prefix = hexnode[:length]
1368 if isvalid(prefix):
1368 if isvalid(prefix):
1369 return disambiguate(hexnode, length)
1369 return disambiguate(hexnode, length)
1370
1370
1371 def cmp(self, node, text):
1371 def cmp(self, node, text):
1372 """compare text with a given file revision
1372 """compare text with a given file revision
1373
1373
1374 returns True if text is different than what is stored.
1374 returns True if text is different than what is stored.
1375 """
1375 """
1376 p1, p2 = self.parents(node)
1376 p1, p2 = self.parents(node)
1377 return hash(text, p1, p2) != node
1377 return hash(text, p1, p2) != node
1378
1378
1379 def _cachesegment(self, offset, data):
1379 def _cachesegment(self, offset, data):
1380 """Add a segment to the revlog cache.
1380 """Add a segment to the revlog cache.
1381
1381
1382 Accepts an absolute offset and the data that is at that location.
1382 Accepts an absolute offset and the data that is at that location.
1383 """
1383 """
1384 o, d = self._chunkcache
1384 o, d = self._chunkcache
1385 # try to add to existing cache
1385 # try to add to existing cache
1386 if o + len(d) == offset and len(d) + len(data) < _chunksize:
1386 if o + len(d) == offset and len(d) + len(data) < _chunksize:
1387 self._chunkcache = o, d + data
1387 self._chunkcache = o, d + data
1388 else:
1388 else:
1389 self._chunkcache = offset, data
1389 self._chunkcache = offset, data
1390
1390
1391 def _readsegment(self, offset, length, df=None):
1391 def _readsegment(self, offset, length, df=None):
1392 """Load a segment of raw data from the revlog.
1392 """Load a segment of raw data from the revlog.
1393
1393
1394 Accepts an absolute offset, length to read, and an optional existing
1394 Accepts an absolute offset, length to read, and an optional existing
1395 file handle to read from.
1395 file handle to read from.
1396
1396
1397 If an existing file handle is passed, it will be seeked and the
1397 If an existing file handle is passed, it will be seeked and the
1398 original seek position will NOT be restored.
1398 original seek position will NOT be restored.
1399
1399
1400 Returns a str or buffer of raw byte data.
1400 Returns a str or buffer of raw byte data.
1401 """
1401 """
1402 # Cache data both forward and backward around the requested
1402 # Cache data both forward and backward around the requested
1403 # data, in a fixed size window. This helps speed up operations
1403 # data, in a fixed size window. This helps speed up operations
1404 # involving reading the revlog backwards.
1404 # involving reading the revlog backwards.
1405 cachesize = self._chunkcachesize
1405 cachesize = self._chunkcachesize
1406 realoffset = offset & ~(cachesize - 1)
1406 realoffset = offset & ~(cachesize - 1)
1407 reallength = (((offset + length + cachesize) & ~(cachesize - 1))
1407 reallength = (((offset + length + cachesize) & ~(cachesize - 1))
1408 - realoffset)
1408 - realoffset)
1409 with self._datareadfp(df) as df:
1409 with self._datareadfp(df) as df:
1410 df.seek(realoffset)
1410 df.seek(realoffset)
1411 d = df.read(reallength)
1411 d = df.read(reallength)
1412 self._cachesegment(realoffset, d)
1412 self._cachesegment(realoffset, d)
1413 if offset != realoffset or reallength != length:
1413 if offset != realoffset or reallength != length:
1414 return util.buffer(d, offset - realoffset, length)
1414 return util.buffer(d, offset - realoffset, length)
1415 return d
1415 return d
1416
1416
1417 def _getsegment(self, offset, length, df=None):
1417 def _getsegment(self, offset, length, df=None):
1418 """Obtain a segment of raw data from the revlog.
1418 """Obtain a segment of raw data from the revlog.
1419
1419
1420 Accepts an absolute offset, length of bytes to obtain, and an
1420 Accepts an absolute offset, length of bytes to obtain, and an
1421 optional file handle to the already-opened revlog. If the file
1421 optional file handle to the already-opened revlog. If the file
1422 handle is used, it's original seek position will not be preserved.
1422 handle is used, it's original seek position will not be preserved.
1423
1423
1424 Requests for data may be returned from a cache.
1424 Requests for data may be returned from a cache.
1425
1425
1426 Returns a str or a buffer instance of raw byte data.
1426 Returns a str or a buffer instance of raw byte data.
1427 """
1427 """
1428 o, d = self._chunkcache
1428 o, d = self._chunkcache
1429 l = len(d)
1429 l = len(d)
1430
1430
1431 # is it in the cache?
1431 # is it in the cache?
1432 cachestart = offset - o
1432 cachestart = offset - o
1433 cacheend = cachestart + length
1433 cacheend = cachestart + length
1434 if cachestart >= 0 and cacheend <= l:
1434 if cachestart >= 0 and cacheend <= l:
1435 if cachestart == 0 and cacheend == l:
1435 if cachestart == 0 and cacheend == l:
1436 return d # avoid a copy
1436 return d # avoid a copy
1437 return util.buffer(d, cachestart, cacheend - cachestart)
1437 return util.buffer(d, cachestart, cacheend - cachestart)
1438
1438
1439 return self._readsegment(offset, length, df=df)
1439 return self._readsegment(offset, length, df=df)
1440
1440
1441 def _getsegmentforrevs(self, startrev, endrev, df=None):
1441 def _getsegmentforrevs(self, startrev, endrev, df=None):
1442 """Obtain a segment of raw data corresponding to a range of revisions.
1442 """Obtain a segment of raw data corresponding to a range of revisions.
1443
1443
1444 Accepts the start and end revisions and an optional already-open
1444 Accepts the start and end revisions and an optional already-open
1445 file handle to be used for reading. If the file handle is read, its
1445 file handle to be used for reading. If the file handle is read, its
1446 seek position will not be preserved.
1446 seek position will not be preserved.
1447
1447
1448 Requests for data may be satisfied by a cache.
1448 Requests for data may be satisfied by a cache.
1449
1449
1450 Returns a 2-tuple of (offset, data) for the requested range of
1450 Returns a 2-tuple of (offset, data) for the requested range of
1451 revisions. Offset is the integer offset from the beginning of the
1451 revisions. Offset is the integer offset from the beginning of the
1452 revlog and data is a str or buffer of the raw byte data.
1452 revlog and data is a str or buffer of the raw byte data.
1453
1453
1454 Callers will need to call ``self.start(rev)`` and ``self.length(rev)``
1454 Callers will need to call ``self.start(rev)`` and ``self.length(rev)``
1455 to determine where each revision's data begins and ends.
1455 to determine where each revision's data begins and ends.
1456 """
1456 """
1457 # Inlined self.start(startrev) & self.end(endrev) for perf reasons
1457 # Inlined self.start(startrev) & self.end(endrev) for perf reasons
1458 # (functions are expensive).
1458 # (functions are expensive).
1459 index = self.index
1459 index = self.index
1460 istart = index[startrev]
1460 istart = index[startrev]
1461 start = int(istart[0] >> 16)
1461 start = int(istart[0] >> 16)
1462 if startrev == endrev:
1462 if startrev == endrev:
1463 end = start + istart[1]
1463 end = start + istart[1]
1464 else:
1464 else:
1465 iend = index[endrev]
1465 iend = index[endrev]
1466 end = int(iend[0] >> 16) + iend[1]
1466 end = int(iend[0] >> 16) + iend[1]
1467
1467
1468 if self._inline:
1468 if self._inline:
1469 start += (startrev + 1) * self._io.size
1469 start += (startrev + 1) * self._io.size
1470 end += (endrev + 1) * self._io.size
1470 end += (endrev + 1) * self._io.size
1471 length = end - start
1471 length = end - start
1472
1472
1473 return start, self._getsegment(start, length, df=df)
1473 return start, self._getsegment(start, length, df=df)
1474
1474
1475 def _chunk(self, rev, df=None):
1475 def _chunk(self, rev, df=None):
1476 """Obtain a single decompressed chunk for a revision.
1476 """Obtain a single decompressed chunk for a revision.
1477
1477
1478 Accepts an integer revision and an optional already-open file handle
1478 Accepts an integer revision and an optional already-open file handle
1479 to be used for reading. If used, the seek position of the file will not
1479 to be used for reading. If used, the seek position of the file will not
1480 be preserved.
1480 be preserved.
1481
1481
1482 Returns a str holding uncompressed data for the requested revision.
1482 Returns a str holding uncompressed data for the requested revision.
1483 """
1483 """
1484 return self.decompress(self._getsegmentforrevs(rev, rev, df=df)[1])
1484 return self.decompress(self._getsegmentforrevs(rev, rev, df=df)[1])
1485
1485
1486 def _chunks(self, revs, df=None, targetsize=None):
1486 def _chunks(self, revs, df=None, targetsize=None):
1487 """Obtain decompressed chunks for the specified revisions.
1487 """Obtain decompressed chunks for the specified revisions.
1488
1488
1489 Accepts an iterable of numeric revisions that are assumed to be in
1489 Accepts an iterable of numeric revisions that are assumed to be in
1490 ascending order. Also accepts an optional already-open file handle
1490 ascending order. Also accepts an optional already-open file handle
1491 to be used for reading. If used, the seek position of the file will
1491 to be used for reading. If used, the seek position of the file will
1492 not be preserved.
1492 not be preserved.
1493
1493
1494 This function is similar to calling ``self._chunk()`` multiple times,
1494 This function is similar to calling ``self._chunk()`` multiple times,
1495 but is faster.
1495 but is faster.
1496
1496
1497 Returns a list with decompressed data for each requested revision.
1497 Returns a list with decompressed data for each requested revision.
1498 """
1498 """
1499 if not revs:
1499 if not revs:
1500 return []
1500 return []
1501 start = self.start
1501 start = self.start
1502 length = self.length
1502 length = self.length
1503 inline = self._inline
1503 inline = self._inline
1504 iosize = self._io.size
1504 iosize = self._io.size
1505 buffer = util.buffer
1505 buffer = util.buffer
1506
1506
1507 l = []
1507 l = []
1508 ladd = l.append
1508 ladd = l.append
1509
1509
1510 if not self._withsparseread:
1510 if not self._withsparseread:
1511 slicedchunks = (revs,)
1511 slicedchunks = (revs,)
1512 else:
1512 else:
1513 slicedchunks = deltautil.slicechunk(self, revs,
1513 slicedchunks = deltautil.slicechunk(self, revs,
1514 targetsize=targetsize)
1514 targetsize=targetsize)
1515
1515
1516 for revschunk in slicedchunks:
1516 for revschunk in slicedchunks:
1517 firstrev = revschunk[0]
1517 firstrev = revschunk[0]
1518 # Skip trailing revisions with empty diff
1518 # Skip trailing revisions with empty diff
1519 for lastrev in revschunk[::-1]:
1519 for lastrev in revschunk[::-1]:
1520 if length(lastrev) != 0:
1520 if length(lastrev) != 0:
1521 break
1521 break
1522
1522
1523 try:
1523 try:
1524 offset, data = self._getsegmentforrevs(firstrev, lastrev, df=df)
1524 offset, data = self._getsegmentforrevs(firstrev, lastrev, df=df)
1525 except OverflowError:
1525 except OverflowError:
1526 # issue4215 - we can't cache a run of chunks greater than
1526 # issue4215 - we can't cache a run of chunks greater than
1527 # 2G on Windows
1527 # 2G on Windows
1528 return [self._chunk(rev, df=df) for rev in revschunk]
1528 return [self._chunk(rev, df=df) for rev in revschunk]
1529
1529
1530 decomp = self.decompress
1530 decomp = self.decompress
1531 for rev in revschunk:
1531 for rev in revschunk:
1532 chunkstart = start(rev)
1532 chunkstart = start(rev)
1533 if inline:
1533 if inline:
1534 chunkstart += (rev + 1) * iosize
1534 chunkstart += (rev + 1) * iosize
1535 chunklength = length(rev)
1535 chunklength = length(rev)
1536 ladd(decomp(buffer(data, chunkstart - offset, chunklength)))
1536 ladd(decomp(buffer(data, chunkstart - offset, chunklength)))
1537
1537
1538 return l
1538 return l
1539
1539
1540 def _chunkclear(self):
1540 def _chunkclear(self):
1541 """Clear the raw chunk cache."""
1541 """Clear the raw chunk cache."""
1542 self._chunkcache = (0, '')
1542 self._chunkcache = (0, '')
1543
1543
1544 def deltaparent(self, rev):
1544 def deltaparent(self, rev):
1545 """return deltaparent of the given revision"""
1545 """return deltaparent of the given revision"""
1546 base = self.index[rev][3]
1546 base = self.index[rev][3]
1547 if base == rev:
1547 if base == rev:
1548 return nullrev
1548 return nullrev
1549 elif self._generaldelta:
1549 elif self._generaldelta:
1550 return base
1550 return base
1551 else:
1551 else:
1552 return rev - 1
1552 return rev - 1
1553
1553
1554 def issnapshot(self, rev):
1554 def issnapshot(self, rev):
1555 """tells whether rev is a snapshot
1555 """tells whether rev is a snapshot
1556 """
1556 """
1557 if rev == nullrev:
1557 if rev == nullrev:
1558 return True
1558 return True
1559 deltap = self.deltaparent(rev)
1559 deltap = self.deltaparent(rev)
1560 if deltap == nullrev:
1560 if deltap == nullrev:
1561 return True
1561 return True
1562 p1, p2 = self.parentrevs(rev)
1562 p1, p2 = self.parentrevs(rev)
1563 if deltap in (p1, p2):
1563 if deltap in (p1, p2):
1564 return False
1564 return False
1565 return self.issnapshot(deltap)
1565 return self.issnapshot(deltap)
1566
1566
1567 def snapshotdepth(self, rev):
1567 def snapshotdepth(self, rev):
1568 """number of snapshot in the chain before this one"""
1568 """number of snapshot in the chain before this one"""
1569 if not self.issnapshot(rev):
1569 if not self.issnapshot(rev):
1570 raise error.ProgrammingError('revision %d not a snapshot')
1570 raise error.ProgrammingError('revision %d not a snapshot')
1571 return len(self._deltachain(rev)[0]) - 1
1571 return len(self._deltachain(rev)[0]) - 1
1572
1572
1573 def revdiff(self, rev1, rev2):
1573 def revdiff(self, rev1, rev2):
1574 """return or calculate a delta between two revisions
1574 """return or calculate a delta between two revisions
1575
1575
1576 The delta calculated is in binary form and is intended to be written to
1576 The delta calculated is in binary form and is intended to be written to
1577 revlog data directly. So this function needs raw revision data.
1577 revlog data directly. So this function needs raw revision data.
1578 """
1578 """
1579 if rev1 != nullrev and self.deltaparent(rev2) == rev1:
1579 if rev1 != nullrev and self.deltaparent(rev2) == rev1:
1580 return bytes(self._chunk(rev2))
1580 return bytes(self._chunk(rev2))
1581
1581
1582 return mdiff.textdiff(self.revision(rev1, raw=True),
1582 return mdiff.textdiff(self.revision(rev1, raw=True),
1583 self.revision(rev2, raw=True))
1583 self.revision(rev2, raw=True))
1584
1584
1585 def revision(self, nodeorrev, _df=None, raw=False):
1585 def revision(self, nodeorrev, _df=None, raw=False):
1586 """return an uncompressed revision of a given node or revision
1586 """return an uncompressed revision of a given node or revision
1587 number.
1587 number.
1588
1588
1589 _df - an existing file handle to read from. (internal-only)
1589 _df - an existing file handle to read from. (internal-only)
1590 raw - an optional argument specifying if the revision data is to be
1590 raw - an optional argument specifying if the revision data is to be
1591 treated as raw data when applying flag transforms. 'raw' should be set
1591 treated as raw data when applying flag transforms. 'raw' should be set
1592 to True when generating changegroups or in debug commands.
1592 to True when generating changegroups or in debug commands.
1593 """
1593 """
1594 if isinstance(nodeorrev, int):
1594 if isinstance(nodeorrev, int):
1595 rev = nodeorrev
1595 rev = nodeorrev
1596 node = self.node(rev)
1596 node = self.node(rev)
1597 else:
1597 else:
1598 node = nodeorrev
1598 node = nodeorrev
1599 rev = None
1599 rev = None
1600
1600
1601 cachedrev = None
1601 cachedrev = None
1602 flags = None
1602 flags = None
1603 rawtext = None
1603 rawtext = None
1604 if node == nullid:
1604 if node == nullid:
1605 return ""
1605 return ""
1606 if self._cache:
1606 if self._cache:
1607 if self._cache[0] == node:
1607 if self._cache[0] == node:
1608 # _cache only stores rawtext
1608 # _cache only stores rawtext
1609 if raw:
1609 if raw:
1610 return self._cache[2]
1610 return self._cache[2]
1611 # duplicated, but good for perf
1611 # duplicated, but good for perf
1612 if rev is None:
1612 if rev is None:
1613 rev = self.rev(node)
1613 rev = self.rev(node)
1614 if flags is None:
1614 if flags is None:
1615 flags = self.flags(rev)
1615 flags = self.flags(rev)
1616 # no extra flags set, no flag processor runs, text = rawtext
1616 # no extra flags set, no flag processor runs, text = rawtext
1617 if flags == REVIDX_DEFAULT_FLAGS:
1617 if flags == REVIDX_DEFAULT_FLAGS:
1618 return self._cache[2]
1618 return self._cache[2]
1619 # rawtext is reusable. need to run flag processor
1619 # rawtext is reusable. need to run flag processor
1620 rawtext = self._cache[2]
1620 rawtext = self._cache[2]
1621
1621
1622 cachedrev = self._cache[1]
1622 cachedrev = self._cache[1]
1623
1623
1624 # look up what we need to read
1624 # look up what we need to read
1625 if rawtext is None:
1625 if rawtext is None:
1626 if rev is None:
1626 if rev is None:
1627 rev = self.rev(node)
1627 rev = self.rev(node)
1628
1628
1629 chain, stopped = self._deltachain(rev, stoprev=cachedrev)
1629 chain, stopped = self._deltachain(rev, stoprev=cachedrev)
1630 if stopped:
1630 if stopped:
1631 rawtext = self._cache[2]
1631 rawtext = self._cache[2]
1632
1632
1633 # drop cache to save memory
1633 # drop cache to save memory
1634 self._cache = None
1634 self._cache = None
1635
1635
1636 targetsize = None
1636 targetsize = None
1637 rawsize = self.index[rev][2]
1637 rawsize = self.index[rev][2]
1638 if 0 <= rawsize:
1638 if 0 <= rawsize:
1639 targetsize = 4 * rawsize
1639 targetsize = 4 * rawsize
1640
1640
1641 bins = self._chunks(chain, df=_df, targetsize=targetsize)
1641 bins = self._chunks(chain, df=_df, targetsize=targetsize)
1642 if rawtext is None:
1642 if rawtext is None:
1643 rawtext = bytes(bins[0])
1643 rawtext = bytes(bins[0])
1644 bins = bins[1:]
1644 bins = bins[1:]
1645
1645
1646 rawtext = mdiff.patches(rawtext, bins)
1646 rawtext = mdiff.patches(rawtext, bins)
1647 self._cache = (node, rev, rawtext)
1647 self._cache = (node, rev, rawtext)
1648
1648
1649 if flags is None:
1649 if flags is None:
1650 if rev is None:
1650 if rev is None:
1651 rev = self.rev(node)
1651 rev = self.rev(node)
1652 flags = self.flags(rev)
1652 flags = self.flags(rev)
1653
1653
1654 text, validatehash = self._processflags(rawtext, flags, 'read', raw=raw)
1654 text, validatehash = self._processflags(rawtext, flags, 'read', raw=raw)
1655 if validatehash:
1655 if validatehash:
1656 self.checkhash(text, node, rev=rev)
1656 self.checkhash(text, node, rev=rev)
1657
1657
1658 return text
1658 return text
1659
1659
1660 def hash(self, text, p1, p2):
1660 def hash(self, text, p1, p2):
1661 """Compute a node hash.
1661 """Compute a node hash.
1662
1662
1663 Available as a function so that subclasses can replace the hash
1663 Available as a function so that subclasses can replace the hash
1664 as needed.
1664 as needed.
1665 """
1665 """
1666 return hash(text, p1, p2)
1666 return hash(text, p1, p2)
1667
1667
1668 def _processflags(self, text, flags, operation, raw=False):
1668 def _processflags(self, text, flags, operation, raw=False):
1669 """Inspect revision data flags and applies transforms defined by
1669 """Inspect revision data flags and applies transforms defined by
1670 registered flag processors.
1670 registered flag processors.
1671
1671
1672 ``text`` - the revision data to process
1672 ``text`` - the revision data to process
1673 ``flags`` - the revision flags
1673 ``flags`` - the revision flags
1674 ``operation`` - the operation being performed (read or write)
1674 ``operation`` - the operation being performed (read or write)
1675 ``raw`` - an optional argument describing if the raw transform should be
1675 ``raw`` - an optional argument describing if the raw transform should be
1676 applied.
1676 applied.
1677
1677
1678 This method processes the flags in the order (or reverse order if
1678 This method processes the flags in the order (or reverse order if
1679 ``operation`` is 'write') defined by REVIDX_FLAGS_ORDER, applying the
1679 ``operation`` is 'write') defined by REVIDX_FLAGS_ORDER, applying the
1680 flag processors registered for present flags. The order of flags defined
1680 flag processors registered for present flags. The order of flags defined
1681 in REVIDX_FLAGS_ORDER needs to be stable to allow non-commutativity.
1681 in REVIDX_FLAGS_ORDER needs to be stable to allow non-commutativity.
1682
1682
1683 Returns a 2-tuple of ``(text, validatehash)`` where ``text`` is the
1683 Returns a 2-tuple of ``(text, validatehash)`` where ``text`` is the
1684 processed text and ``validatehash`` is a bool indicating whether the
1684 processed text and ``validatehash`` is a bool indicating whether the
1685 returned text should be checked for hash integrity.
1685 returned text should be checked for hash integrity.
1686
1686
1687 Note: If the ``raw`` argument is set, it has precedence over the
1687 Note: If the ``raw`` argument is set, it has precedence over the
1688 operation and will only update the value of ``validatehash``.
1688 operation and will only update the value of ``validatehash``.
1689 """
1689 """
1690 # fast path: no flag processors will run
1690 # fast path: no flag processors will run
1691 if flags == 0:
1691 if flags == 0:
1692 return text, True
1692 return text, True
1693 if not operation in ('read', 'write'):
1693 if not operation in ('read', 'write'):
1694 raise error.ProgrammingError(_("invalid '%s' operation") %
1694 raise error.ProgrammingError(_("invalid '%s' operation") %
1695 operation)
1695 operation)
1696 # Check all flags are known.
1696 # Check all flags are known.
1697 if flags & ~REVIDX_KNOWN_FLAGS:
1697 if flags & ~REVIDX_KNOWN_FLAGS:
1698 raise error.RevlogError(_("incompatible revision flag '%#x'") %
1698 raise error.RevlogError(_("incompatible revision flag '%#x'") %
1699 (flags & ~REVIDX_KNOWN_FLAGS))
1699 (flags & ~REVIDX_KNOWN_FLAGS))
1700 validatehash = True
1700 validatehash = True
1701 # Depending on the operation (read or write), the order might be
1701 # Depending on the operation (read or write), the order might be
1702 # reversed due to non-commutative transforms.
1702 # reversed due to non-commutative transforms.
1703 orderedflags = REVIDX_FLAGS_ORDER
1703 orderedflags = REVIDX_FLAGS_ORDER
1704 if operation == 'write':
1704 if operation == 'write':
1705 orderedflags = reversed(orderedflags)
1705 orderedflags = reversed(orderedflags)
1706
1706
1707 for flag in orderedflags:
1707 for flag in orderedflags:
1708 # If a flagprocessor has been registered for a known flag, apply the
1708 # If a flagprocessor has been registered for a known flag, apply the
1709 # related operation transform and update result tuple.
1709 # related operation transform and update result tuple.
1710 if flag & flags:
1710 if flag & flags:
1711 vhash = True
1711 vhash = True
1712
1712
1713 if flag not in self._flagprocessors:
1713 if flag not in self._flagprocessors:
1714 message = _("missing processor for flag '%#x'") % (flag)
1714 message = _("missing processor for flag '%#x'") % (flag)
1715 raise error.RevlogError(message)
1715 raise error.RevlogError(message)
1716
1716
1717 processor = self._flagprocessors[flag]
1717 processor = self._flagprocessors[flag]
1718 if processor is not None:
1718 if processor is not None:
1719 readtransform, writetransform, rawtransform = processor
1719 readtransform, writetransform, rawtransform = processor
1720
1720
1721 if raw:
1721 if raw:
1722 vhash = rawtransform(self, text)
1722 vhash = rawtransform(self, text)
1723 elif operation == 'read':
1723 elif operation == 'read':
1724 text, vhash = readtransform(self, text)
1724 text, vhash = readtransform(self, text)
1725 else: # write operation
1725 else: # write operation
1726 text, vhash = writetransform(self, text)
1726 text, vhash = writetransform(self, text)
1727 validatehash = validatehash and vhash
1727 validatehash = validatehash and vhash
1728
1728
1729 return text, validatehash
1729 return text, validatehash
1730
1730
1731 def checkhash(self, text, node, p1=None, p2=None, rev=None):
1731 def checkhash(self, text, node, p1=None, p2=None, rev=None):
1732 """Check node hash integrity.
1732 """Check node hash integrity.
1733
1733
1734 Available as a function so that subclasses can extend hash mismatch
1734 Available as a function so that subclasses can extend hash mismatch
1735 behaviors as needed.
1735 behaviors as needed.
1736 """
1736 """
1737 try:
1737 try:
1738 if p1 is None and p2 is None:
1738 if p1 is None and p2 is None:
1739 p1, p2 = self.parents(node)
1739 p1, p2 = self.parents(node)
1740 if node != self.hash(text, p1, p2):
1740 if node != self.hash(text, p1, p2):
1741 revornode = rev
1741 revornode = rev
1742 if revornode is None:
1742 if revornode is None:
1743 revornode = templatefilters.short(hex(node))
1743 revornode = templatefilters.short(hex(node))
1744 raise error.RevlogError(_("integrity check failed on %s:%s")
1744 raise error.RevlogError(_("integrity check failed on %s:%s")
1745 % (self.indexfile, pycompat.bytestr(revornode)))
1745 % (self.indexfile, pycompat.bytestr(revornode)))
1746 except error.RevlogError:
1746 except error.RevlogError:
1747 if self._censorable and _censoredtext(text):
1747 if self._censorable and _censoredtext(text):
1748 raise error.CensoredNodeError(self.indexfile, node, text)
1748 raise error.CensoredNodeError(self.indexfile, node, text)
1749 raise
1749 raise
1750
1750
1751 def _enforceinlinesize(self, tr, fp=None):
1751 def _enforceinlinesize(self, tr, fp=None):
1752 """Check if the revlog is too big for inline and convert if so.
1752 """Check if the revlog is too big for inline and convert if so.
1753
1753
1754 This should be called after revisions are added to the revlog. If the
1754 This should be called after revisions are added to the revlog. If the
1755 revlog has grown too large to be an inline revlog, it will convert it
1755 revlog has grown too large to be an inline revlog, it will convert it
1756 to use multiple index and data files.
1756 to use multiple index and data files.
1757 """
1757 """
1758 tiprev = len(self) - 1
1758 tiprev = len(self) - 1
1759 if (not self._inline or
1759 if (not self._inline or
1760 (self.start(tiprev) + self.length(tiprev)) < _maxinline):
1760 (self.start(tiprev) + self.length(tiprev)) < _maxinline):
1761 return
1761 return
1762
1762
1763 trinfo = tr.find(self.indexfile)
1763 trinfo = tr.find(self.indexfile)
1764 if trinfo is None:
1764 if trinfo is None:
1765 raise error.RevlogError(_("%s not found in the transaction")
1765 raise error.RevlogError(_("%s not found in the transaction")
1766 % self.indexfile)
1766 % self.indexfile)
1767
1767
1768 trindex = trinfo[2]
1768 trindex = trinfo[2]
1769 if trindex is not None:
1769 if trindex is not None:
1770 dataoff = self.start(trindex)
1770 dataoff = self.start(trindex)
1771 else:
1771 else:
1772 # revlog was stripped at start of transaction, use all leftover data
1772 # revlog was stripped at start of transaction, use all leftover data
1773 trindex = len(self) - 1
1773 trindex = len(self) - 1
1774 dataoff = self.end(tiprev)
1774 dataoff = self.end(tiprev)
1775
1775
1776 tr.add(self.datafile, dataoff)
1776 tr.add(self.datafile, dataoff)
1777
1777
1778 if fp:
1778 if fp:
1779 fp.flush()
1779 fp.flush()
1780 fp.close()
1780 fp.close()
1781
1781
1782 with self._datafp('w') as df:
1782 with self._datafp('w') as df:
1783 for r in self:
1783 for r in self:
1784 df.write(self._getsegmentforrevs(r, r)[1])
1784 df.write(self._getsegmentforrevs(r, r)[1])
1785
1785
1786 with self._indexfp('w') as fp:
1786 with self._indexfp('w') as fp:
1787 self.version &= ~FLAG_INLINE_DATA
1787 self.version &= ~FLAG_INLINE_DATA
1788 self._inline = False
1788 self._inline = False
1789 io = self._io
1789 io = self._io
1790 for i in self:
1790 for i in self:
1791 e = io.packentry(self.index[i], self.node, self.version, i)
1791 e = io.packentry(self.index[i], self.node, self.version, i)
1792 fp.write(e)
1792 fp.write(e)
1793
1793
1794 # the temp file replace the real index when we exit the context
1794 # the temp file replace the real index when we exit the context
1795 # manager
1795 # manager
1796
1796
1797 tr.replace(self.indexfile, trindex * self._io.size)
1797 tr.replace(self.indexfile, trindex * self._io.size)
1798 self._chunkclear()
1798 self._chunkclear()
1799
1799
1800 def addrevision(self, text, transaction, link, p1, p2, cachedelta=None,
1800 def addrevision(self, text, transaction, link, p1, p2, cachedelta=None,
1801 node=None, flags=REVIDX_DEFAULT_FLAGS, deltacomputer=None):
1801 node=None, flags=REVIDX_DEFAULT_FLAGS, deltacomputer=None):
1802 """add a revision to the log
1802 """add a revision to the log
1803
1803
1804 text - the revision data to add
1804 text - the revision data to add
1805 transaction - the transaction object used for rollback
1805 transaction - the transaction object used for rollback
1806 link - the linkrev data to add
1806 link - the linkrev data to add
1807 p1, p2 - the parent nodeids of the revision
1807 p1, p2 - the parent nodeids of the revision
1808 cachedelta - an optional precomputed delta
1808 cachedelta - an optional precomputed delta
1809 node - nodeid of revision; typically node is not specified, and it is
1809 node - nodeid of revision; typically node is not specified, and it is
1810 computed by default as hash(text, p1, p2), however subclasses might
1810 computed by default as hash(text, p1, p2), however subclasses might
1811 use different hashing method (and override checkhash() in such case)
1811 use different hashing method (and override checkhash() in such case)
1812 flags - the known flags to set on the revision
1812 flags - the known flags to set on the revision
1813 deltacomputer - an optional deltacomputer instance shared between
1813 deltacomputer - an optional deltacomputer instance shared between
1814 multiple calls
1814 multiple calls
1815 """
1815 """
1816 if link == nullrev:
1816 if link == nullrev:
1817 raise error.RevlogError(_("attempted to add linkrev -1 to %s")
1817 raise error.RevlogError(_("attempted to add linkrev -1 to %s")
1818 % self.indexfile)
1818 % self.indexfile)
1819
1819
1820 if flags:
1820 if flags:
1821 node = node or self.hash(text, p1, p2)
1821 node = node or self.hash(text, p1, p2)
1822
1822
1823 rawtext, validatehash = self._processflags(text, flags, 'write')
1823 rawtext, validatehash = self._processflags(text, flags, 'write')
1824
1824
1825 # If the flag processor modifies the revision data, ignore any provided
1825 # If the flag processor modifies the revision data, ignore any provided
1826 # cachedelta.
1826 # cachedelta.
1827 if rawtext != text:
1827 if rawtext != text:
1828 cachedelta = None
1828 cachedelta = None
1829
1829
1830 if len(rawtext) > _maxentrysize:
1830 if len(rawtext) > _maxentrysize:
1831 raise error.RevlogError(
1831 raise error.RevlogError(
1832 _("%s: size of %d bytes exceeds maximum revlog storage of 2GiB")
1832 _("%s: size of %d bytes exceeds maximum revlog storage of 2GiB")
1833 % (self.indexfile, len(rawtext)))
1833 % (self.indexfile, len(rawtext)))
1834
1834
1835 node = node or self.hash(rawtext, p1, p2)
1835 node = node or self.hash(rawtext, p1, p2)
1836 if node in self.nodemap:
1836 if node in self.nodemap:
1837 return node
1837 return node
1838
1838
1839 if validatehash:
1839 if validatehash:
1840 self.checkhash(rawtext, node, p1=p1, p2=p2)
1840 self.checkhash(rawtext, node, p1=p1, p2=p2)
1841
1841
1842 return self.addrawrevision(rawtext, transaction, link, p1, p2, node,
1842 return self.addrawrevision(rawtext, transaction, link, p1, p2, node,
1843 flags, cachedelta=cachedelta,
1843 flags, cachedelta=cachedelta,
1844 deltacomputer=deltacomputer)
1844 deltacomputer=deltacomputer)
1845
1845
1846 def addrawrevision(self, rawtext, transaction, link, p1, p2, node, flags,
1846 def addrawrevision(self, rawtext, transaction, link, p1, p2, node, flags,
1847 cachedelta=None, deltacomputer=None):
1847 cachedelta=None, deltacomputer=None):
1848 """add a raw revision with known flags, node and parents
1848 """add a raw revision with known flags, node and parents
1849 useful when reusing a revision not stored in this revlog (ex: received
1849 useful when reusing a revision not stored in this revlog (ex: received
1850 over wire, or read from an external bundle).
1850 over wire, or read from an external bundle).
1851 """
1851 """
1852 dfh = None
1852 dfh = None
1853 if not self._inline:
1853 if not self._inline:
1854 dfh = self._datafp("a+")
1854 dfh = self._datafp("a+")
1855 ifh = self._indexfp("a+")
1855 ifh = self._indexfp("a+")
1856 try:
1856 try:
1857 return self._addrevision(node, rawtext, transaction, link, p1, p2,
1857 return self._addrevision(node, rawtext, transaction, link, p1, p2,
1858 flags, cachedelta, ifh, dfh,
1858 flags, cachedelta, ifh, dfh,
1859 deltacomputer=deltacomputer)
1859 deltacomputer=deltacomputer)
1860 finally:
1860 finally:
1861 if dfh:
1861 if dfh:
1862 dfh.close()
1862 dfh.close()
1863 ifh.close()
1863 ifh.close()
1864
1864
1865 def compress(self, data):
1865 def compress(self, data):
1866 """Generate a possibly-compressed representation of data."""
1866 """Generate a possibly-compressed representation of data."""
1867 if not data:
1867 if not data:
1868 return '', data
1868 return '', data
1869
1869
1870 compressed = self._compressor.compress(data)
1870 compressed = self._compressor.compress(data)
1871
1871
1872 if compressed:
1872 if compressed:
1873 # The revlog compressor added the header in the returned data.
1873 # The revlog compressor added the header in the returned data.
1874 return '', compressed
1874 return '', compressed
1875
1875
1876 if data[0:1] == '\0':
1876 if data[0:1] == '\0':
1877 return '', data
1877 return '', data
1878 return 'u', data
1878 return 'u', data
1879
1879
1880 def decompress(self, data):
1880 def decompress(self, data):
1881 """Decompress a revlog chunk.
1881 """Decompress a revlog chunk.
1882
1882
1883 The chunk is expected to begin with a header identifying the
1883 The chunk is expected to begin with a header identifying the
1884 format type so it can be routed to an appropriate decompressor.
1884 format type so it can be routed to an appropriate decompressor.
1885 """
1885 """
1886 if not data:
1886 if not data:
1887 return data
1887 return data
1888
1888
1889 # Revlogs are read much more frequently than they are written and many
1889 # Revlogs are read much more frequently than they are written and many
1890 # chunks only take microseconds to decompress, so performance is
1890 # chunks only take microseconds to decompress, so performance is
1891 # important here.
1891 # important here.
1892 #
1892 #
1893 # We can make a few assumptions about revlogs:
1893 # We can make a few assumptions about revlogs:
1894 #
1894 #
1895 # 1) the majority of chunks will be compressed (as opposed to inline
1895 # 1) the majority of chunks will be compressed (as opposed to inline
1896 # raw data).
1896 # raw data).
1897 # 2) decompressing *any* data will likely by at least 10x slower than
1897 # 2) decompressing *any* data will likely by at least 10x slower than
1898 # returning raw inline data.
1898 # returning raw inline data.
1899 # 3) we want to prioritize common and officially supported compression
1899 # 3) we want to prioritize common and officially supported compression
1900 # engines
1900 # engines
1901 #
1901 #
1902 # It follows that we want to optimize for "decompress compressed data
1902 # It follows that we want to optimize for "decompress compressed data
1903 # when encoded with common and officially supported compression engines"
1903 # when encoded with common and officially supported compression engines"
1904 # case over "raw data" and "data encoded by less common or non-official
1904 # case over "raw data" and "data encoded by less common or non-official
1905 # compression engines." That is why we have the inline lookup first
1905 # compression engines." That is why we have the inline lookup first
1906 # followed by the compengines lookup.
1906 # followed by the compengines lookup.
1907 #
1907 #
1908 # According to `hg perfrevlogchunks`, this is ~0.5% faster for zlib
1908 # According to `hg perfrevlogchunks`, this is ~0.5% faster for zlib
1909 # compressed chunks. And this matters for changelog and manifest reads.
1909 # compressed chunks. And this matters for changelog and manifest reads.
1910 t = data[0:1]
1910 t = data[0:1]
1911
1911
1912 if t == 'x':
1912 if t == 'x':
1913 try:
1913 try:
1914 return _zlibdecompress(data)
1914 return _zlibdecompress(data)
1915 except zlib.error as e:
1915 except zlib.error as e:
1916 raise error.RevlogError(_('revlog decompress error: %s') %
1916 raise error.RevlogError(_('revlog decompress error: %s') %
1917 stringutil.forcebytestr(e))
1917 stringutil.forcebytestr(e))
1918 # '\0' is more common than 'u' so it goes first.
1918 # '\0' is more common than 'u' so it goes first.
1919 elif t == '\0':
1919 elif t == '\0':
1920 return data
1920 return data
1921 elif t == 'u':
1921 elif t == 'u':
1922 return util.buffer(data, 1)
1922 return util.buffer(data, 1)
1923
1923
1924 try:
1924 try:
1925 compressor = self._decompressors[t]
1925 compressor = self._decompressors[t]
1926 except KeyError:
1926 except KeyError:
1927 try:
1927 try:
1928 engine = util.compengines.forrevlogheader(t)
1928 engine = util.compengines.forrevlogheader(t)
1929 compressor = engine.revlogcompressor()
1929 compressor = engine.revlogcompressor()
1930 self._decompressors[t] = compressor
1930 self._decompressors[t] = compressor
1931 except KeyError:
1931 except KeyError:
1932 raise error.RevlogError(_('unknown compression type %r') % t)
1932 raise error.RevlogError(_('unknown compression type %r') % t)
1933
1933
1934 return compressor.decompress(data)
1934 return compressor.decompress(data)
1935
1935
1936 def _addrevision(self, node, rawtext, transaction, link, p1, p2, flags,
1936 def _addrevision(self, node, rawtext, transaction, link, p1, p2, flags,
1937 cachedelta, ifh, dfh, alwayscache=False,
1937 cachedelta, ifh, dfh, alwayscache=False,
1938 deltacomputer=None):
1938 deltacomputer=None):
1939 """internal function to add revisions to the log
1939 """internal function to add revisions to the log
1940
1940
1941 see addrevision for argument descriptions.
1941 see addrevision for argument descriptions.
1942
1942
1943 note: "addrevision" takes non-raw text, "_addrevision" takes raw text.
1943 note: "addrevision" takes non-raw text, "_addrevision" takes raw text.
1944
1944
1945 if "deltacomputer" is not provided or None, a defaultdeltacomputer will
1945 if "deltacomputer" is not provided or None, a defaultdeltacomputer will
1946 be used.
1946 be used.
1947
1947
1948 invariants:
1948 invariants:
1949 - rawtext is optional (can be None); if not set, cachedelta must be set.
1949 - rawtext is optional (can be None); if not set, cachedelta must be set.
1950 if both are set, they must correspond to each other.
1950 if both are set, they must correspond to each other.
1951 """
1951 """
1952 if node == nullid:
1952 if node == nullid:
1953 raise error.RevlogError(_("%s: attempt to add null revision") %
1953 raise error.RevlogError(_("%s: attempt to add null revision") %
1954 self.indexfile)
1954 self.indexfile)
1955 if node == wdirid or node in wdirfilenodeids:
1955 if node == wdirid or node in wdirfilenodeids:
1956 raise error.RevlogError(_("%s: attempt to add wdir revision") %
1956 raise error.RevlogError(_("%s: attempt to add wdir revision") %
1957 self.indexfile)
1957 self.indexfile)
1958
1958
1959 if self._inline:
1959 if self._inline:
1960 fh = ifh
1960 fh = ifh
1961 else:
1961 else:
1962 fh = dfh
1962 fh = dfh
1963
1963
1964 btext = [rawtext]
1964 btext = [rawtext]
1965
1965
1966 curr = len(self)
1966 curr = len(self)
1967 prev = curr - 1
1967 prev = curr - 1
1968 offset = self.end(prev)
1968 offset = self.end(prev)
1969 p1r, p2r = self.rev(p1), self.rev(p2)
1969 p1r, p2r = self.rev(p1), self.rev(p2)
1970
1970
1971 # full versions are inserted when the needed deltas
1971 # full versions are inserted when the needed deltas
1972 # become comparable to the uncompressed text
1972 # become comparable to the uncompressed text
1973 if rawtext is None:
1973 if rawtext is None:
1974 # need rawtext size, before changed by flag processors, which is
1974 # need rawtext size, before changed by flag processors, which is
1975 # the non-raw size. use revlog explicitly to avoid filelog's extra
1975 # the non-raw size. use revlog explicitly to avoid filelog's extra
1976 # logic that might remove metadata size.
1976 # logic that might remove metadata size.
1977 textlen = mdiff.patchedsize(revlog.size(self, cachedelta[0]),
1977 textlen = mdiff.patchedsize(revlog.size(self, cachedelta[0]),
1978 cachedelta[1])
1978 cachedelta[1])
1979 else:
1979 else:
1980 textlen = len(rawtext)
1980 textlen = len(rawtext)
1981
1981
1982 if deltacomputer is None:
1982 if deltacomputer is None:
1983 deltacomputer = deltautil.deltacomputer(self)
1983 deltacomputer = deltautil.deltacomputer(self)
1984
1984
1985 revinfo = _revisioninfo(node, p1, p2, btext, textlen, cachedelta, flags)
1985 revinfo = _revisioninfo(node, p1, p2, btext, textlen, cachedelta, flags)
1986
1986
1987 deltainfo = deltacomputer.finddeltainfo(revinfo, fh)
1987 deltainfo = deltacomputer.finddeltainfo(revinfo, fh)
1988
1988
1989 e = (offset_type(offset, flags), deltainfo.deltalen, textlen,
1989 e = (offset_type(offset, flags), deltainfo.deltalen, textlen,
1990 deltainfo.base, link, p1r, p2r, node)
1990 deltainfo.base, link, p1r, p2r, node)
1991 self.index.append(e)
1991 self.index.append(e)
1992 self.nodemap[node] = curr
1992 self.nodemap[node] = curr
1993
1993
1994 entry = self._io.packentry(e, self.node, self.version, curr)
1994 entry = self._io.packentry(e, self.node, self.version, curr)
1995 self._writeentry(transaction, ifh, dfh, entry, deltainfo.data,
1995 self._writeentry(transaction, ifh, dfh, entry, deltainfo.data,
1996 link, offset)
1996 link, offset)
1997
1997
1998 rawtext = btext[0]
1998 rawtext = btext[0]
1999
1999
2000 if alwayscache and rawtext is None:
2000 if alwayscache and rawtext is None:
2001 rawtext = deltacomputer.buildtext(revinfo, fh)
2001 rawtext = deltacomputer.buildtext(revinfo, fh)
2002
2002
2003 if type(rawtext) == bytes: # only accept immutable objects
2003 if type(rawtext) == bytes: # only accept immutable objects
2004 self._cache = (node, curr, rawtext)
2004 self._cache = (node, curr, rawtext)
2005 self._chainbasecache[curr] = deltainfo.chainbase
2005 self._chainbasecache[curr] = deltainfo.chainbase
2006 return node
2006 return node
2007
2007
2008 def _writeentry(self, transaction, ifh, dfh, entry, data, link, offset):
2008 def _writeentry(self, transaction, ifh, dfh, entry, data, link, offset):
2009 # Files opened in a+ mode have inconsistent behavior on various
2009 # Files opened in a+ mode have inconsistent behavior on various
2010 # platforms. Windows requires that a file positioning call be made
2010 # platforms. Windows requires that a file positioning call be made
2011 # when the file handle transitions between reads and writes. See
2011 # when the file handle transitions between reads and writes. See
2012 # 3686fa2b8eee and the mixedfilemodewrapper in windows.py. On other
2012 # 3686fa2b8eee and the mixedfilemodewrapper in windows.py. On other
2013 # platforms, Python or the platform itself can be buggy. Some versions
2013 # platforms, Python or the platform itself can be buggy. Some versions
2014 # of Solaris have been observed to not append at the end of the file
2014 # of Solaris have been observed to not append at the end of the file
2015 # if the file was seeked to before the end. See issue4943 for more.
2015 # if the file was seeked to before the end. See issue4943 for more.
2016 #
2016 #
2017 # We work around this issue by inserting a seek() before writing.
2017 # We work around this issue by inserting a seek() before writing.
2018 # Note: This is likely not necessary on Python 3.
2018 # Note: This is likely not necessary on Python 3.
2019 ifh.seek(0, os.SEEK_END)
2019 ifh.seek(0, os.SEEK_END)
2020 if dfh:
2020 if dfh:
2021 dfh.seek(0, os.SEEK_END)
2021 dfh.seek(0, os.SEEK_END)
2022
2022
2023 curr = len(self) - 1
2023 curr = len(self) - 1
2024 if not self._inline:
2024 if not self._inline:
2025 transaction.add(self.datafile, offset)
2025 transaction.add(self.datafile, offset)
2026 transaction.add(self.indexfile, curr * len(entry))
2026 transaction.add(self.indexfile, curr * len(entry))
2027 if data[0]:
2027 if data[0]:
2028 dfh.write(data[0])
2028 dfh.write(data[0])
2029 dfh.write(data[1])
2029 dfh.write(data[1])
2030 ifh.write(entry)
2030 ifh.write(entry)
2031 else:
2031 else:
2032 offset += curr * self._io.size
2032 offset += curr * self._io.size
2033 transaction.add(self.indexfile, offset, curr)
2033 transaction.add(self.indexfile, offset, curr)
2034 ifh.write(entry)
2034 ifh.write(entry)
2035 ifh.write(data[0])
2035 ifh.write(data[0])
2036 ifh.write(data[1])
2036 ifh.write(data[1])
2037 self._enforceinlinesize(transaction, ifh)
2037 self._enforceinlinesize(transaction, ifh)
2038
2038
2039 def addgroup(self, deltas, linkmapper, transaction, addrevisioncb=None):
2039 def addgroup(self, deltas, linkmapper, transaction, addrevisioncb=None):
2040 """
2040 """
2041 add a delta group
2041 add a delta group
2042
2042
2043 given a set of deltas, add them to the revision log. the
2043 given a set of deltas, add them to the revision log. the
2044 first delta is against its parent, which should be in our
2044 first delta is against its parent, which should be in our
2045 log, the rest are against the previous delta.
2045 log, the rest are against the previous delta.
2046
2046
2047 If ``addrevisioncb`` is defined, it will be called with arguments of
2047 If ``addrevisioncb`` is defined, it will be called with arguments of
2048 this revlog and the node that was added.
2048 this revlog and the node that was added.
2049 """
2049 """
2050
2050
2051 nodes = []
2051 nodes = []
2052
2052
2053 r = len(self)
2053 r = len(self)
2054 end = 0
2054 end = 0
2055 if r:
2055 if r:
2056 end = self.end(r - 1)
2056 end = self.end(r - 1)
2057 ifh = self._indexfp("a+")
2057 ifh = self._indexfp("a+")
2058 isize = r * self._io.size
2058 isize = r * self._io.size
2059 if self._inline:
2059 if self._inline:
2060 transaction.add(self.indexfile, end + isize, r)
2060 transaction.add(self.indexfile, end + isize, r)
2061 dfh = None
2061 dfh = None
2062 else:
2062 else:
2063 transaction.add(self.indexfile, isize, r)
2063 transaction.add(self.indexfile, isize, r)
2064 transaction.add(self.datafile, end)
2064 transaction.add(self.datafile, end)
2065 dfh = self._datafp("a+")
2065 dfh = self._datafp("a+")
2066 def flush():
2066 def flush():
2067 if dfh:
2067 if dfh:
2068 dfh.flush()
2068 dfh.flush()
2069 ifh.flush()
2069 ifh.flush()
2070 try:
2070 try:
2071 deltacomputer = deltautil.deltacomputer(self)
2071 deltacomputer = deltautil.deltacomputer(self)
2072 # loop through our set of deltas
2072 # loop through our set of deltas
2073 for data in deltas:
2073 for data in deltas:
2074 node, p1, p2, linknode, deltabase, delta, flags = data
2074 node, p1, p2, linknode, deltabase, delta, flags = data
2075 link = linkmapper(linknode)
2075 link = linkmapper(linknode)
2076 flags = flags or REVIDX_DEFAULT_FLAGS
2076 flags = flags or REVIDX_DEFAULT_FLAGS
2077
2077
2078 nodes.append(node)
2078 nodes.append(node)
2079
2079
2080 if node in self.nodemap:
2080 if node in self.nodemap:
2081 # this can happen if two branches make the same change
2081 # this can happen if two branches make the same change
2082 continue
2082 continue
2083
2083
2084 for p in (p1, p2):
2084 for p in (p1, p2):
2085 if p not in self.nodemap:
2085 if p not in self.nodemap:
2086 raise error.LookupError(p, self.indexfile,
2086 raise error.LookupError(p, self.indexfile,
2087 _('unknown parent'))
2087 _('unknown parent'))
2088
2088
2089 if deltabase not in self.nodemap:
2089 if deltabase not in self.nodemap:
2090 raise error.LookupError(deltabase, self.indexfile,
2090 raise error.LookupError(deltabase, self.indexfile,
2091 _('unknown delta base'))
2091 _('unknown delta base'))
2092
2092
2093 baserev = self.rev(deltabase)
2093 baserev = self.rev(deltabase)
2094
2094
2095 if baserev != nullrev and self.iscensored(baserev):
2095 if baserev != nullrev and self.iscensored(baserev):
2096 # if base is censored, delta must be full replacement in a
2096 # if base is censored, delta must be full replacement in a
2097 # single patch operation
2097 # single patch operation
2098 hlen = struct.calcsize(">lll")
2098 hlen = struct.calcsize(">lll")
2099 oldlen = self.rawsize(baserev)
2099 oldlen = self.rawsize(baserev)
2100 newlen = len(delta) - hlen
2100 newlen = len(delta) - hlen
2101 if delta[:hlen] != mdiff.replacediffheader(oldlen, newlen):
2101 if delta[:hlen] != mdiff.replacediffheader(oldlen, newlen):
2102 raise error.CensoredBaseError(self.indexfile,
2102 raise error.CensoredBaseError(self.indexfile,
2103 self.node(baserev))
2103 self.node(baserev))
2104
2104
2105 if not flags and self._peek_iscensored(baserev, delta, flush):
2105 if not flags and self._peek_iscensored(baserev, delta, flush):
2106 flags |= REVIDX_ISCENSORED
2106 flags |= REVIDX_ISCENSORED
2107
2107
2108 # We assume consumers of addrevisioncb will want to retrieve
2108 # We assume consumers of addrevisioncb will want to retrieve
2109 # the added revision, which will require a call to
2109 # the added revision, which will require a call to
2110 # revision(). revision() will fast path if there is a cache
2110 # revision(). revision() will fast path if there is a cache
2111 # hit. So, we tell _addrevision() to always cache in this case.
2111 # hit. So, we tell _addrevision() to always cache in this case.
2112 # We're only using addgroup() in the context of changegroup
2112 # We're only using addgroup() in the context of changegroup
2113 # generation so the revision data can always be handled as raw
2113 # generation so the revision data can always be handled as raw
2114 # by the flagprocessor.
2114 # by the flagprocessor.
2115 self._addrevision(node, None, transaction, link,
2115 self._addrevision(node, None, transaction, link,
2116 p1, p2, flags, (baserev, delta),
2116 p1, p2, flags, (baserev, delta),
2117 ifh, dfh,
2117 ifh, dfh,
2118 alwayscache=bool(addrevisioncb),
2118 alwayscache=bool(addrevisioncb),
2119 deltacomputer=deltacomputer)
2119 deltacomputer=deltacomputer)
2120
2120
2121 if addrevisioncb:
2121 if addrevisioncb:
2122 addrevisioncb(self, node)
2122 addrevisioncb(self, node)
2123
2123
2124 if not dfh and not self._inline:
2124 if not dfh and not self._inline:
2125 # addrevision switched from inline to conventional
2125 # addrevision switched from inline to conventional
2126 # reopen the index
2126 # reopen the index
2127 ifh.close()
2127 ifh.close()
2128 dfh = self._datafp("a+")
2128 dfh = self._datafp("a+")
2129 ifh = self._indexfp("a+")
2129 ifh = self._indexfp("a+")
2130 finally:
2130 finally:
2131 if dfh:
2131 if dfh:
2132 dfh.close()
2132 dfh.close()
2133 ifh.close()
2133 ifh.close()
2134
2134
2135 return nodes
2135 return nodes
2136
2136
2137 def iscensored(self, rev):
2137 def iscensored(self, rev):
2138 """Check if a file revision is censored."""
2138 """Check if a file revision is censored."""
2139 if not self._censorable:
2139 if not self._censorable:
2140 return False
2140 return False
2141
2141
2142 return self.flags(rev) & REVIDX_ISCENSORED
2142 return self.flags(rev) & REVIDX_ISCENSORED
2143
2143
2144 def _peek_iscensored(self, baserev, delta, flush):
2144 def _peek_iscensored(self, baserev, delta, flush):
2145 """Quickly check if a delta produces a censored revision."""
2145 """Quickly check if a delta produces a censored revision."""
2146 if not self._censorable:
2146 if not self._censorable:
2147 return False
2147 return False
2148
2148
2149 # Fragile heuristic: unless new file meta keys are added alphabetically
2149 # Fragile heuristic: unless new file meta keys are added alphabetically
2150 # preceding "censored", all censored revisions are prefixed by
2150 # preceding "censored", all censored revisions are prefixed by
2151 # "\1\ncensored:". A delta producing such a censored revision must be a
2151 # "\1\ncensored:". A delta producing such a censored revision must be a
2152 # full-replacement delta, so we inspect the first and only patch in the
2152 # full-replacement delta, so we inspect the first and only patch in the
2153 # delta for this prefix.
2153 # delta for this prefix.
2154 hlen = struct.calcsize(">lll")
2154 hlen = struct.calcsize(">lll")
2155 if len(delta) <= hlen:
2155 if len(delta) <= hlen:
2156 return False
2156 return False
2157
2157
2158 oldlen = self.rawsize(baserev)
2158 oldlen = self.rawsize(baserev)
2159 newlen = len(delta) - hlen
2159 newlen = len(delta) - hlen
2160 if delta[:hlen] != mdiff.replacediffheader(oldlen, newlen):
2160 if delta[:hlen] != mdiff.replacediffheader(oldlen, newlen):
2161 return False
2161 return False
2162
2162
2163 add = "\1\ncensored:"
2163 add = "\1\ncensored:"
2164 addlen = len(add)
2164 addlen = len(add)
2165 return newlen >= addlen and delta[hlen:hlen + addlen] == add
2165 return newlen >= addlen and delta[hlen:hlen + addlen] == add
2166
2166
2167 def getstrippoint(self, minlink):
2167 def getstrippoint(self, minlink):
2168 """find the minimum rev that must be stripped to strip the linkrev
2168 """find the minimum rev that must be stripped to strip the linkrev
2169
2169
2170 Returns a tuple containing the minimum rev and a set of all revs that
2170 Returns a tuple containing the minimum rev and a set of all revs that
2171 have linkrevs that will be broken by this strip.
2171 have linkrevs that will be broken by this strip.
2172 """
2172 """
2173 brokenrevs = set()
2173 brokenrevs = set()
2174 strippoint = len(self)
2174 strippoint = len(self)
2175
2175
2176 heads = {}
2176 heads = {}
2177 futurelargelinkrevs = set()
2177 futurelargelinkrevs = set()
2178 for head in self.headrevs():
2178 for head in self.headrevs():
2179 headlinkrev = self.linkrev(head)
2179 headlinkrev = self.linkrev(head)
2180 heads[head] = headlinkrev
2180 heads[head] = headlinkrev
2181 if headlinkrev >= minlink:
2181 if headlinkrev >= minlink:
2182 futurelargelinkrevs.add(headlinkrev)
2182 futurelargelinkrevs.add(headlinkrev)
2183
2183
2184 # This algorithm involves walking down the rev graph, starting at the
2184 # This algorithm involves walking down the rev graph, starting at the
2185 # heads. Since the revs are topologically sorted according to linkrev,
2185 # heads. Since the revs are topologically sorted according to linkrev,
2186 # once all head linkrevs are below the minlink, we know there are
2186 # once all head linkrevs are below the minlink, we know there are
2187 # no more revs that could have a linkrev greater than minlink.
2187 # no more revs that could have a linkrev greater than minlink.
2188 # So we can stop walking.
2188 # So we can stop walking.
2189 while futurelargelinkrevs:
2189 while futurelargelinkrevs:
2190 strippoint -= 1
2190 strippoint -= 1
2191 linkrev = heads.pop(strippoint)
2191 linkrev = heads.pop(strippoint)
2192
2192
2193 if linkrev < minlink:
2193 if linkrev < minlink:
2194 brokenrevs.add(strippoint)
2194 brokenrevs.add(strippoint)
2195 else:
2195 else:
2196 futurelargelinkrevs.remove(linkrev)
2196 futurelargelinkrevs.remove(linkrev)
2197
2197
2198 for p in self.parentrevs(strippoint):
2198 for p in self.parentrevs(strippoint):
2199 if p != nullrev:
2199 if p != nullrev:
2200 plinkrev = self.linkrev(p)
2200 plinkrev = self.linkrev(p)
2201 heads[p] = plinkrev
2201 heads[p] = plinkrev
2202 if plinkrev >= minlink:
2202 if plinkrev >= minlink:
2203 futurelargelinkrevs.add(plinkrev)
2203 futurelargelinkrevs.add(plinkrev)
2204
2204
2205 return strippoint, brokenrevs
2205 return strippoint, brokenrevs
2206
2206
2207 def strip(self, minlink, transaction):
2207 def strip(self, minlink, transaction):
2208 """truncate the revlog on the first revision with a linkrev >= minlink
2208 """truncate the revlog on the first revision with a linkrev >= minlink
2209
2209
2210 This function is called when we're stripping revision minlink and
2210 This function is called when we're stripping revision minlink and
2211 its descendants from the repository.
2211 its descendants from the repository.
2212
2212
2213 We have to remove all revisions with linkrev >= minlink, because
2213 We have to remove all revisions with linkrev >= minlink, because
2214 the equivalent changelog revisions will be renumbered after the
2214 the equivalent changelog revisions will be renumbered after the
2215 strip.
2215 strip.
2216
2216
2217 So we truncate the revlog on the first of these revisions, and
2217 So we truncate the revlog on the first of these revisions, and
2218 trust that the caller has saved the revisions that shouldn't be
2218 trust that the caller has saved the revisions that shouldn't be
2219 removed and that it'll re-add them after this truncation.
2219 removed and that it'll re-add them after this truncation.
2220 """
2220 """
2221 if len(self) == 0:
2221 if len(self) == 0:
2222 return
2222 return
2223
2223
2224 rev, _ = self.getstrippoint(minlink)
2224 rev, _ = self.getstrippoint(minlink)
2225 if rev == len(self):
2225 if rev == len(self):
2226 return
2226 return
2227
2227
2228 # first truncate the files on disk
2228 # first truncate the files on disk
2229 end = self.start(rev)
2229 end = self.start(rev)
2230 if not self._inline:
2230 if not self._inline:
2231 transaction.add(self.datafile, end)
2231 transaction.add(self.datafile, end)
2232 end = rev * self._io.size
2232 end = rev * self._io.size
2233 else:
2233 else:
2234 end += rev * self._io.size
2234 end += rev * self._io.size
2235
2235
2236 transaction.add(self.indexfile, end)
2236 transaction.add(self.indexfile, end)
2237
2237
2238 # then reset internal state in memory to forget those revisions
2238 # then reset internal state in memory to forget those revisions
2239 self._cache = None
2239 self._cache = None
2240 self._chaininfocache = {}
2240 self._chaininfocache = {}
2241 self._chunkclear()
2241 self._chunkclear()
2242 for x in pycompat.xrange(rev, len(self)):
2242 for x in pycompat.xrange(rev, len(self)):
2243 del self.nodemap[self.node(x)]
2243 del self.nodemap[self.node(x)]
2244
2244
2245 del self.index[rev:-1]
2245 del self.index[rev:-1]
2246 self._nodepos = None
2246 self._nodepos = None
2247
2247
2248 def checksize(self):
2248 def checksize(self):
2249 expected = 0
2249 expected = 0
2250 if len(self):
2250 if len(self):
2251 expected = max(0, self.end(len(self) - 1))
2251 expected = max(0, self.end(len(self) - 1))
2252
2252
2253 try:
2253 try:
2254 with self._datafp() as f:
2254 with self._datafp() as f:
2255 f.seek(0, 2)
2255 f.seek(0, 2)
2256 actual = f.tell()
2256 actual = f.tell()
2257 dd = actual - expected
2257 dd = actual - expected
2258 except IOError as inst:
2258 except IOError as inst:
2259 if inst.errno != errno.ENOENT:
2259 if inst.errno != errno.ENOENT:
2260 raise
2260 raise
2261 dd = 0
2261 dd = 0
2262
2262
2263 try:
2263 try:
2264 f = self.opener(self.indexfile)
2264 f = self.opener(self.indexfile)
2265 f.seek(0, 2)
2265 f.seek(0, 2)
2266 actual = f.tell()
2266 actual = f.tell()
2267 f.close()
2267 f.close()
2268 s = self._io.size
2268 s = self._io.size
2269 i = max(0, actual // s)
2269 i = max(0, actual // s)
2270 di = actual - (i * s)
2270 di = actual - (i * s)
2271 if self._inline:
2271 if self._inline:
2272 databytes = 0
2272 databytes = 0
2273 for r in self:
2273 for r in self:
2274 databytes += max(0, self.length(r))
2274 databytes += max(0, self.length(r))
2275 dd = 0
2275 dd = 0
2276 di = actual - len(self) * s - databytes
2276 di = actual - len(self) * s - databytes
2277 except IOError as inst:
2277 except IOError as inst:
2278 if inst.errno != errno.ENOENT:
2278 if inst.errno != errno.ENOENT:
2279 raise
2279 raise
2280 di = 0
2280 di = 0
2281
2281
2282 return (dd, di)
2282 return (dd, di)
2283
2283
2284 def files(self):
2284 def files(self):
2285 res = [self.indexfile]
2285 res = [self.indexfile]
2286 if not self._inline:
2286 if not self._inline:
2287 res.append(self.datafile)
2287 res.append(self.datafile)
2288 return res
2288 return res
2289
2289
2290 def emitrevisiondeltas(self, requests):
2290 def emitrevisiondeltas(self, requests):
2291 frev = self.rev
2291 frev = self.rev
2292
2292
2293 prevrev = None
2293 prevrev = None
2294 for request in requests:
2294 for request in requests:
2295 node = request.node
2295 node = request.node
2296 rev = frev(node)
2296 rev = frev(node)
2297
2297
2298 if prevrev is None:
2298 if prevrev is None:
2299 prevrev = self.index[rev][5]
2299 prevrev = self.index[rev][5]
2300
2300
2301 # Requesting a full revision.
2301 # Requesting a full revision.
2302 if request.basenode == nullid:
2302 if request.basenode == nullid:
2303 baserev = nullrev
2303 baserev = nullrev
2304 # Requesting an explicit revision.
2304 # Requesting an explicit revision.
2305 elif request.basenode is not None:
2305 elif request.basenode is not None:
2306 baserev = frev(request.basenode)
2306 baserev = frev(request.basenode)
2307 # Allowing us to choose.
2307 # Allowing us to choose.
2308 else:
2308 else:
2309 p1rev, p2rev = self.parentrevs(rev)
2309 p1rev, p2rev = self.parentrevs(rev)
2310 deltaparentrev = self.deltaparent(rev)
2310 deltaparentrev = self.deltaparent(rev)
2311
2311
2312 # Avoid sending full revisions when delta parent is null. Pick
2312 # Avoid sending full revisions when delta parent is null. Pick
2313 # prev in that case. It's tempting to pick p1 in this case, as
2313 # prev in that case. It's tempting to pick p1 in this case, as
2314 # p1 will be smaller in the common case. However, computing a
2314 # p1 will be smaller in the common case. However, computing a
2315 # delta against p1 may require resolving the raw text of p1,
2315 # delta against p1 may require resolving the raw text of p1,
2316 # which could be expensive. The revlog caches should have prev
2316 # which could be expensive. The revlog caches should have prev
2317 # cached, meaning less CPU for delta generation. There is
2317 # cached, meaning less CPU for delta generation. There is
2318 # likely room to add a flag and/or config option to control this
2318 # likely room to add a flag and/or config option to control this
2319 # behavior.
2319 # behavior.
2320 if deltaparentrev == nullrev and self._storedeltachains:
2320 if deltaparentrev == nullrev and self._storedeltachains:
2321 baserev = prevrev
2321 baserev = prevrev
2322
2322
2323 # Revlog is configured to use full snapshot for a reason.
2323 # Revlog is configured to use full snapshot for a reason.
2324 # Stick to full snapshot.
2324 # Stick to full snapshot.
2325 elif deltaparentrev == nullrev:
2325 elif deltaparentrev == nullrev:
2326 baserev = nullrev
2326 baserev = nullrev
2327
2327
2328 # Pick previous when we can't be sure the base is available
2328 # Pick previous when we can't be sure the base is available
2329 # on consumer.
2329 # on consumer.
2330 elif deltaparentrev not in (p1rev, p2rev, prevrev):
2330 elif deltaparentrev not in (p1rev, p2rev, prevrev):
2331 baserev = prevrev
2331 baserev = prevrev
2332 else:
2332 else:
2333 baserev = deltaparentrev
2333 baserev = deltaparentrev
2334
2334
2335 if baserev != nullrev and not self.candelta(baserev, rev):
2335 if baserev != nullrev and not self.candelta(baserev, rev):
2336 baserev = nullrev
2336 baserev = nullrev
2337
2337
2338 revision = None
2338 revision = None
2339 delta = None
2339 delta = None
2340 baserevisionsize = None
2340 baserevisionsize = None
2341
2341
2342 if self.iscensored(baserev) or self.iscensored(rev):
2342 if self.iscensored(baserev) or self.iscensored(rev):
2343 try:
2343 try:
2344 revision = self.revision(node, raw=True)
2344 revision = self.revision(node, raw=True)
2345 except error.CensoredNodeError as e:
2345 except error.CensoredNodeError as e:
2346 revision = e.tombstone
2346 revision = e.tombstone
2347
2347
2348 if baserev != nullrev:
2348 if baserev != nullrev:
2349 baserevisionsize = self.rawsize(baserev)
2349 baserevisionsize = self.rawsize(baserev)
2350
2350
2351 elif baserev == nullrev:
2351 elif baserev == nullrev:
2352 revision = self.revision(node, raw=True)
2352 revision = self.revision(node, raw=True)
2353 else:
2353 else:
2354 delta = self.revdiff(baserev, rev)
2354 delta = self.revdiff(baserev, rev)
2355
2355
2356 extraflags = REVIDX_ELLIPSIS if request.ellipsis else 0
2356 extraflags = REVIDX_ELLIPSIS if request.ellipsis else 0
2357
2357
2358 yield revlogrevisiondelta(
2358 yield revlogrevisiondelta(
2359 node=node,
2359 node=node,
2360 p1node=request.p1node,
2360 p1node=request.p1node,
2361 p2node=request.p2node,
2361 p2node=request.p2node,
2362 linknode=request.linknode,
2362 linknode=request.linknode,
2363 basenode=self.node(baserev),
2363 basenode=self.node(baserev),
2364 flags=self.flags(rev) | extraflags,
2364 flags=self.flags(rev) | extraflags,
2365 baserevisionsize=baserevisionsize,
2365 baserevisionsize=baserevisionsize,
2366 revision=revision,
2366 revision=revision,
2367 delta=delta)
2367 delta=delta)
2368
2368
2369 prevrev = rev
2369 prevrev = rev
2370
2370
2371 DELTAREUSEALWAYS = 'always'
2371 DELTAREUSEALWAYS = 'always'
2372 DELTAREUSESAMEREVS = 'samerevs'
2372 DELTAREUSESAMEREVS = 'samerevs'
2373 DELTAREUSENEVER = 'never'
2373 DELTAREUSENEVER = 'never'
2374
2374
2375 DELTAREUSEFULLADD = 'fulladd'
2375 DELTAREUSEFULLADD = 'fulladd'
2376
2376
2377 DELTAREUSEALL = {'always', 'samerevs', 'never', 'fulladd'}
2377 DELTAREUSEALL = {'always', 'samerevs', 'never', 'fulladd'}
2378
2378
2379 def clone(self, tr, destrevlog, addrevisioncb=None,
2379 def clone(self, tr, destrevlog, addrevisioncb=None,
2380 deltareuse=DELTAREUSESAMEREVS, deltabothparents=None):
2380 deltareuse=DELTAREUSESAMEREVS, deltabothparents=None):
2381 """Copy this revlog to another, possibly with format changes.
2381 """Copy this revlog to another, possibly with format changes.
2382
2382
2383 The destination revlog will contain the same revisions and nodes.
2383 The destination revlog will contain the same revisions and nodes.
2384 However, it may not be bit-for-bit identical due to e.g. delta encoding
2384 However, it may not be bit-for-bit identical due to e.g. delta encoding
2385 differences.
2385 differences.
2386
2386
2387 The ``deltareuse`` argument control how deltas from the existing revlog
2387 The ``deltareuse`` argument control how deltas from the existing revlog
2388 are preserved in the destination revlog. The argument can have the
2388 are preserved in the destination revlog. The argument can have the
2389 following values:
2389 following values:
2390
2390
2391 DELTAREUSEALWAYS
2391 DELTAREUSEALWAYS
2392 Deltas will always be reused (if possible), even if the destination
2392 Deltas will always be reused (if possible), even if the destination
2393 revlog would not select the same revisions for the delta. This is the
2393 revlog would not select the same revisions for the delta. This is the
2394 fastest mode of operation.
2394 fastest mode of operation.
2395 DELTAREUSESAMEREVS
2395 DELTAREUSESAMEREVS
2396 Deltas will be reused if the destination revlog would pick the same
2396 Deltas will be reused if the destination revlog would pick the same
2397 revisions for the delta. This mode strikes a balance between speed
2397 revisions for the delta. This mode strikes a balance between speed
2398 and optimization.
2398 and optimization.
2399 DELTAREUSENEVER
2399 DELTAREUSENEVER
2400 Deltas will never be reused. This is the slowest mode of execution.
2400 Deltas will never be reused. This is the slowest mode of execution.
2401 This mode can be used to recompute deltas (e.g. if the diff/delta
2401 This mode can be used to recompute deltas (e.g. if the diff/delta
2402 algorithm changes).
2402 algorithm changes).
2403
2403
2404 Delta computation can be slow, so the choice of delta reuse policy can
2404 Delta computation can be slow, so the choice of delta reuse policy can
2405 significantly affect run time.
2405 significantly affect run time.
2406
2406
2407 The default policy (``DELTAREUSESAMEREVS``) strikes a balance between
2407 The default policy (``DELTAREUSESAMEREVS``) strikes a balance between
2408 two extremes. Deltas will be reused if they are appropriate. But if the
2408 two extremes. Deltas will be reused if they are appropriate. But if the
2409 delta could choose a better revision, it will do so. This means if you
2409 delta could choose a better revision, it will do so. This means if you
2410 are converting a non-generaldelta revlog to a generaldelta revlog,
2410 are converting a non-generaldelta revlog to a generaldelta revlog,
2411 deltas will be recomputed if the delta's parent isn't a parent of the
2411 deltas will be recomputed if the delta's parent isn't a parent of the
2412 revision.
2412 revision.
2413
2413
2414 In addition to the delta policy, the ``deltabothparents`` argument
2414 In addition to the delta policy, the ``deltabothparents`` argument
2415 controls whether to compute deltas against both parents for merges.
2415 controls whether to compute deltas against both parents for merges.
2416 By default, the current default is used.
2416 By default, the current default is used.
2417 """
2417 """
2418 if deltareuse not in self.DELTAREUSEALL:
2418 if deltareuse not in self.DELTAREUSEALL:
2419 raise ValueError(_('value for deltareuse invalid: %s') % deltareuse)
2419 raise ValueError(_('value for deltareuse invalid: %s') % deltareuse)
2420
2420
2421 if len(destrevlog):
2421 if len(destrevlog):
2422 raise ValueError(_('destination revlog is not empty'))
2422 raise ValueError(_('destination revlog is not empty'))
2423
2423
2424 if getattr(self, 'filteredrevs', None):
2424 if getattr(self, 'filteredrevs', None):
2425 raise ValueError(_('source revlog has filtered revisions'))
2425 raise ValueError(_('source revlog has filtered revisions'))
2426 if getattr(destrevlog, 'filteredrevs', None):
2426 if getattr(destrevlog, 'filteredrevs', None):
2427 raise ValueError(_('destination revlog has filtered revisions'))
2427 raise ValueError(_('destination revlog has filtered revisions'))
2428
2428
2429 # lazydeltabase controls whether to reuse a cached delta, if possible.
2429 # lazydeltabase controls whether to reuse a cached delta, if possible.
2430 oldlazydeltabase = destrevlog._lazydeltabase
2430 oldlazydeltabase = destrevlog._lazydeltabase
2431 oldamd = destrevlog._deltabothparents
2431 oldamd = destrevlog._deltabothparents
2432
2432
2433 try:
2433 try:
2434 if deltareuse == self.DELTAREUSEALWAYS:
2434 if deltareuse == self.DELTAREUSEALWAYS:
2435 destrevlog._lazydeltabase = True
2435 destrevlog._lazydeltabase = True
2436 elif deltareuse == self.DELTAREUSESAMEREVS:
2436 elif deltareuse == self.DELTAREUSESAMEREVS:
2437 destrevlog._lazydeltabase = False
2437 destrevlog._lazydeltabase = False
2438
2438
2439 destrevlog._deltabothparents = deltabothparents or oldamd
2439 destrevlog._deltabothparents = deltabothparents or oldamd
2440
2440
2441 populatecachedelta = deltareuse in (self.DELTAREUSEALWAYS,
2441 populatecachedelta = deltareuse in (self.DELTAREUSEALWAYS,
2442 self.DELTAREUSESAMEREVS)
2442 self.DELTAREUSESAMEREVS)
2443
2443
2444 deltacomputer = deltautil.deltacomputer(destrevlog)
2444 deltacomputer = deltautil.deltacomputer(destrevlog)
2445 index = self.index
2445 index = self.index
2446 for rev in self:
2446 for rev in self:
2447 entry = index[rev]
2447 entry = index[rev]
2448
2448
2449 # Some classes override linkrev to take filtered revs into
2449 # Some classes override linkrev to take filtered revs into
2450 # account. Use raw entry from index.
2450 # account. Use raw entry from index.
2451 flags = entry[0] & 0xffff
2451 flags = entry[0] & 0xffff
2452 linkrev = entry[4]
2452 linkrev = entry[4]
2453 p1 = index[entry[5]][7]
2453 p1 = index[entry[5]][7]
2454 p2 = index[entry[6]][7]
2454 p2 = index[entry[6]][7]
2455 node = entry[7]
2455 node = entry[7]
2456
2456
2457 # (Possibly) reuse the delta from the revlog if allowed and
2457 # (Possibly) reuse the delta from the revlog if allowed and
2458 # the revlog chunk is a delta.
2458 # the revlog chunk is a delta.
2459 cachedelta = None
2459 cachedelta = None
2460 rawtext = None
2460 rawtext = None
2461 if populatecachedelta:
2461 if populatecachedelta:
2462 dp = self.deltaparent(rev)
2462 dp = self.deltaparent(rev)
2463 if dp != nullrev:
2463 if dp != nullrev:
2464 cachedelta = (dp, bytes(self._chunk(rev)))
2464 cachedelta = (dp, bytes(self._chunk(rev)))
2465
2465
2466 if not cachedelta:
2466 if not cachedelta:
2467 rawtext = self.revision(rev, raw=True)
2467 rawtext = self.revision(rev, raw=True)
2468
2468
2469
2469
2470 if deltareuse == self.DELTAREUSEFULLADD:
2470 if deltareuse == self.DELTAREUSEFULLADD:
2471 destrevlog.addrevision(rawtext, tr, linkrev, p1, p2,
2471 destrevlog.addrevision(rawtext, tr, linkrev, p1, p2,
2472 cachedelta=cachedelta,
2472 cachedelta=cachedelta,
2473 node=node, flags=flags,
2473 node=node, flags=flags,
2474 deltacomputer=deltacomputer)
2474 deltacomputer=deltacomputer)
2475 else:
2475 else:
2476 ifh = destrevlog.opener(destrevlog.indexfile, 'a+',
2476 ifh = destrevlog.opener(destrevlog.indexfile, 'a+',
2477 checkambig=False)
2477 checkambig=False)
2478 dfh = None
2478 dfh = None
2479 if not destrevlog._inline:
2479 if not destrevlog._inline:
2480 dfh = destrevlog.opener(destrevlog.datafile, 'a+')
2480 dfh = destrevlog.opener(destrevlog.datafile, 'a+')
2481 try:
2481 try:
2482 destrevlog._addrevision(node, rawtext, tr, linkrev, p1,
2482 destrevlog._addrevision(node, rawtext, tr, linkrev, p1,
2483 p2, flags, cachedelta, ifh, dfh,
2483 p2, flags, cachedelta, ifh, dfh,
2484 deltacomputer=deltacomputer)
2484 deltacomputer=deltacomputer)
2485 finally:
2485 finally:
2486 if dfh:
2486 if dfh:
2487 dfh.close()
2487 dfh.close()
2488 ifh.close()
2488 ifh.close()
2489
2489
2490 if addrevisioncb:
2490 if addrevisioncb:
2491 addrevisioncb(self, rev, node)
2491 addrevisioncb(self, rev, node)
2492 finally:
2492 finally:
2493 destrevlog._lazydeltabase = oldlazydeltabase
2493 destrevlog._lazydeltabase = oldlazydeltabase
2494 destrevlog._deltabothparents = oldamd
2494 destrevlog._deltabothparents = oldamd
2495
2495
2496 def censorrevision(self, node, tombstone=b''):
2496 def censorrevision(self, node, tombstone=b''):
2497 if (self.version & 0xFFFF) == REVLOGV0:
2497 if (self.version & 0xFFFF) == REVLOGV0:
2498 raise error.RevlogError(_('cannot censor with version %d revlogs') %
2498 raise error.RevlogError(_('cannot censor with version %d revlogs') %
2499 self.version)
2499 self.version)
2500
2500
2501 rev = self.rev(node)
2501 rev = self.rev(node)
2502 tombstone = packmeta({b'censored': tombstone}, b'')
2502 tombstone = packmeta({b'censored': tombstone}, b'')
2503
2503
2504 if len(tombstone) > self.rawsize(rev):
2504 if len(tombstone) > self.rawsize(rev):
2505 raise error.Abort(_('censor tombstone must be no longer than '
2505 raise error.Abort(_('censor tombstone must be no longer than '
2506 'censored data'))
2506 'censored data'))
2507
2507
2508 # Using two files instead of one makes it easy to rewrite entry-by-entry
2508 # Using two files instead of one makes it easy to rewrite entry-by-entry
2509 idxread = self.opener(self.indexfile, 'r')
2509 idxread = self.opener(self.indexfile, 'r')
2510 idxwrite = self.opener(self.indexfile, 'wb', atomictemp=True)
2510 idxwrite = self.opener(self.indexfile, 'wb', atomictemp=True)
2511 if self.version & FLAG_INLINE_DATA:
2511 if self.version & FLAG_INLINE_DATA:
2512 dataread, datawrite = idxread, idxwrite
2512 dataread, datawrite = idxread, idxwrite
2513 else:
2513 else:
2514 dataread = self.opener(self.datafile, 'r')
2514 dataread = self.opener(self.datafile, 'r')
2515 datawrite = self.opener(self.datafile, 'wb', atomictemp=True)
2515 datawrite = self.opener(self.datafile, 'wb', atomictemp=True)
2516
2516
2517 # Copy all revlog data up to the entry to be censored.
2517 # Copy all revlog data up to the entry to be censored.
2518 offset = self.start(rev)
2518 offset = self.start(rev)
2519
2519
2520 for chunk in util.filechunkiter(idxread, limit=rev * self._io.size):
2520 for chunk in util.filechunkiter(idxread, limit=rev * self._io.size):
2521 idxwrite.write(chunk)
2521 idxwrite.write(chunk)
2522 for chunk in util.filechunkiter(dataread, limit=offset):
2522 for chunk in util.filechunkiter(dataread, limit=offset):
2523 datawrite.write(chunk)
2523 datawrite.write(chunk)
2524
2524
2525 def rewriteindex(r, newoffs, newdata=None):
2525 def rewriteindex(r, newoffs, newdata=None):
2526 """Rewrite the index entry with a new data offset and new data.
2526 """Rewrite the index entry with a new data offset and new data.
2527
2527
2528 The newdata argument, if given, is a tuple of three positive
2528 The newdata argument, if given, is a tuple of three positive
2529 integers: (new compressed, new uncompressed, added flag bits).
2529 integers: (new compressed, new uncompressed, added flag bits).
2530 """
2530 """
2531 offlags, comp, uncomp, base, link, p1, p2, nodeid = self.index[r]
2531 offlags, comp, uncomp, base, link, p1, p2, nodeid = self.index[r]
2532 flags = gettype(offlags)
2532 flags = gettype(offlags)
2533 if newdata:
2533 if newdata:
2534 comp, uncomp, nflags = newdata
2534 comp, uncomp, nflags = newdata
2535 flags |= nflags
2535 flags |= nflags
2536 offlags = offset_type(newoffs, flags)
2536 offlags = offset_type(newoffs, flags)
2537 e = (offlags, comp, uncomp, r, link, p1, p2, nodeid)
2537 e = (offlags, comp, uncomp, r, link, p1, p2, nodeid)
2538 idxwrite.write(self._io.packentry(e, None, self.version, r))
2538 idxwrite.write(self._io.packentry(e, None, self.version, r))
2539 idxread.seek(self._io.size, 1)
2539 idxread.seek(self._io.size, 1)
2540
2540
2541 def rewrite(r, offs, data, nflags=REVIDX_DEFAULT_FLAGS):
2541 def rewrite(r, offs, data, nflags=REVIDX_DEFAULT_FLAGS):
2542 """Write the given fulltext with the given data offset.
2542 """Write the given fulltext with the given data offset.
2543
2543
2544 Returns:
2544 Returns:
2545 The integer number of data bytes written, for tracking data
2545 The integer number of data bytes written, for tracking data
2546 offsets.
2546 offsets.
2547 """
2547 """
2548 flag, compdata = self.compress(data)
2548 flag, compdata = self.compress(data)
2549 newcomp = len(flag) + len(compdata)
2549 newcomp = len(flag) + len(compdata)
2550 rewriteindex(r, offs, (newcomp, len(data), nflags))
2550 rewriteindex(r, offs, (newcomp, len(data), nflags))
2551 datawrite.write(flag)
2551 datawrite.write(flag)
2552 datawrite.write(compdata)
2552 datawrite.write(compdata)
2553 dataread.seek(self.length(r), 1)
2553 dataread.seek(self.length(r), 1)
2554 return newcomp
2554 return newcomp
2555
2555
2556 # Rewrite censored entry with (padded) tombstone data.
2556 # Rewrite censored entry with (padded) tombstone data.
2557 pad = ' ' * (self.rawsize(rev) - len(tombstone))
2557 pad = ' ' * (self.rawsize(rev) - len(tombstone))
2558 offset += rewrite(rev, offset, tombstone + pad, REVIDX_ISCENSORED)
2558 offset += rewrite(rev, offset, tombstone + pad, REVIDX_ISCENSORED)
2559
2559
2560 # Rewrite all following filelog revisions fixing up offsets and deltas.
2560 # Rewrite all following filelog revisions fixing up offsets and deltas.
2561 for srev in pycompat.xrange(rev + 1, len(self)):
2561 for srev in pycompat.xrange(rev + 1, len(self)):
2562 if rev in self.parentrevs(srev):
2562 if rev in self.parentrevs(srev):
2563 # Immediate children of censored node must be re-added as
2563 # Immediate children of censored node must be re-added as
2564 # fulltext.
2564 # fulltext.
2565 try:
2565 try:
2566 revdata = self.revision(srev)
2566 revdata = self.revision(srev)
2567 except error.CensoredNodeError as e:
2567 except error.CensoredNodeError as e:
2568 revdata = e.tombstone
2568 revdata = e.tombstone
2569 dlen = rewrite(srev, offset, revdata)
2569 dlen = rewrite(srev, offset, revdata)
2570 else:
2570 else:
2571 # Copy any other revision data verbatim after fixing up the
2571 # Copy any other revision data verbatim after fixing up the
2572 # offset.
2572 # offset.
2573 rewriteindex(srev, offset)
2573 rewriteindex(srev, offset)
2574 dlen = self.length(srev)
2574 dlen = self.length(srev)
2575 for chunk in util.filechunkiter(dataread, limit=dlen):
2575 for chunk in util.filechunkiter(dataread, limit=dlen):
2576 datawrite.write(chunk)
2576 datawrite.write(chunk)
2577 offset += dlen
2577 offset += dlen
2578
2578
2579 idxread.close()
2579 idxread.close()
2580 idxwrite.close()
2580 idxwrite.close()
2581 if dataread is not idxread:
2581 if dataread is not idxread:
2582 dataread.close()
2582 dataread.close()
2583 datawrite.close()
2583 datawrite.close()
General Comments 0
You need to be logged in to leave comments. Login now