1
# Copyright (C) 2005, 2009 Canonical Ltd
3
# This program is free software; you can redistribute it and/or modify
4
# it under the terms of the GNU General Public License as published by
5
# the Free Software Foundation; either version 2 of the License, or
6
# (at your option) any later version.
8
# This program is distributed in the hope that it will be useful,
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
# GNU General Public License for more details.
13
# You should have received a copy of the GNU General Public License
14
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
17
from __future__ import absolute_import
19
# Author: Martin Pool <mbp@canonical.com>
22
"""Weave - storage of related text file versions"""
25
# XXX: If we do weaves this way, will a merge still behave the same
26
# way if it's done in a different order? That's a pretty desirable
29
# TODO: Nothing here so far assumes the lines are really \n newlines,
30
# rather than being split up in some other way. We could accommodate
31
# binaries, perhaps by naively splitting on \n or perhaps using
32
# something like a rolling checksum.
34
# TODO: End marker for each version so we can stop reading?
36
# TODO: Check that no insertion occurs inside a deletion that was
37
# active in the version of the insertion.
39
# TODO: In addition to the SHA-1 check, perhaps have some code that
40
# checks structural constraints of the weave: ie that insertions are
41
# properly nested, that there is no text outside of an insertion, that
42
# insertions or deletions are not repeated, etc.
44
# TODO: Parallel-extract that passes back each line along with a
45
# description of which revisions include it. Nice for checking all
46
# shas or calculating stats in parallel.
48
# TODO: Using a single _extract routine and then processing the output
49
# is probably inefficient. It's simple enough that we can afford to
50
# have slight specializations for different ways its used: annotate,
51
# basis for add, get, etc.
53
# TODO: Probably the API should work only in names to hide the integer
54
# indexes from the user.
56
# TODO: Is there any potential performance win by having an add()
57
# variant that is passed a pre-cooked version of the single basis
60
# TODO: Reweave can possibly be made faster by remembering diffs
61
# where the basis and destination are unchanged.
63
# FIXME: Sometimes we will be given a parents list for a revision
64
# that includes some redundant parents (i.e. already a parent of
65
# something in the list.) We should eliminate them. This can
66
# be done fairly efficiently because the sequence numbers constrain
67
# the possible relationships.
69
# FIXME: the conflict markers should be *7* characters
72
from cStringIO import StringIO
75
from bzrlib.lazy_import import lazy_import
76
lazy_import(globals(), """
77
from bzrlib import tsort
83
from bzrlib.errors import (WeaveError, WeaveFormatError, WeaveParentMismatch,
84
RevisionAlreadyPresent,
86
UnavailableRepresentation,
88
from bzrlib.osutils import dirname, sha, sha_strings, split_lines
89
import bzrlib.patiencediff
90
from bzrlib.revision import NULL_REVISION
91
from bzrlib.symbol_versioning import *
92
from bzrlib.trace import mutter
93
from bzrlib.versionedfile import (
100
from bzrlib.weavefile import _read_weave_v5, write_weave_v5
103
class WeaveContentFactory(ContentFactory):
104
"""Content factory for streaming from weaves.
106
:seealso ContentFactory:
109
def __init__(self, version, weave):
110
"""Create a WeaveContentFactory for version from weave."""
111
ContentFactory.__init__(self)
112
self.sha1 = weave.get_sha1s([version])[version]
113
self.key = (version,)
114
parents = weave.get_parent_map([version])[version]
115
self.parents = tuple((parent,) for parent in parents)
116
self.storage_kind = 'fulltext'
119
def get_bytes_as(self, storage_kind):
120
if storage_kind == 'fulltext':
121
return self._weave.get_text(self.key[-1])
122
elif storage_kind == 'chunked':
123
return self._weave.get_lines(self.key[-1])
125
raise UnavailableRepresentation(self.key, storage_kind, 'fulltext')
128
class Weave(VersionedFile):
129
"""weave - versioned text file storage.
131
A Weave manages versions of line-based text files, keeping track
132
of the originating version for each line.
134
To clients the "lines" of the file are represented as a list of strings.
135
These strings will typically have terminal newline characters, but
136
this is not required. In particular files commonly do not have a newline
137
at the end of the file.
139
Texts can be identified in either of two ways:
141
* a nonnegative index number.
143
* a version-id string.
145
Typically the index number will be valid only inside this weave and
146
the version-id is used to reference it in the larger world.
148
The weave is represented as a list mixing edit instructions and
149
literal text. Each entry in _weave can be either a string (or
150
unicode), or a tuple. If a string, it means that the given line
151
should be output in the currently active revisions.
153
If a tuple, it gives a processing instruction saying in which
154
revisions the enclosed lines are active. The tuple has the form
155
(instruction, version).
157
The instruction can be '{' or '}' for an insertion block, and '['
158
and ']' for a deletion block respectively. The version is the
159
integer version index. There is no replace operator, only deletes
160
and inserts. For '}', the end of an insertion, there is no
161
version parameter because it always closes the most recently
166
* A later version can delete lines that were introduced by any
167
number of ancestor versions; this implies that deletion
168
instructions can span insertion blocks without regard to the
169
insertion block's nesting.
171
* Similarly, deletions need not be properly nested with regard to
172
each other, because they might have been generated by
173
independent revisions.
175
* Insertions are always made by inserting a new bracketed block
176
into a single point in the previous weave. This implies they
177
can nest but not overlap, and the nesting must always have later
178
insertions on the inside.
180
* It doesn't seem very useful to have an active insertion
181
inside an inactive insertion, but it might happen.
183
* Therefore, all instructions are always"considered"; that
184
is passed onto and off the stack. An outer inactive block
185
doesn't disable an inner block.
187
* Lines are enabled if the most recent enclosing insertion is
188
active and none of the enclosing deletions are active.
190
* There is no point having a deletion directly inside its own
191
insertion; you might as well just not write it. And there
192
should be no way to get an earlier version deleting a later
196
Text of the weave; list of control instruction tuples and strings.
199
List of parents, indexed by version number.
200
It is only necessary to store the minimal set of parents for
201
each version; the parent's parents are implied.
204
List of hex SHA-1 of each version.
207
List of symbolic names for each version. Each should be unique.
210
For each name, the version number.
213
Descriptive name of this weave; typically the filename if known.
217
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
218
'_weave_name', '_matcher', '_allow_reserved']
220
def __init__(self, weave_name=None, access_mode='w', matcher=None,
221
get_scope=None, allow_reserved=False):
224
:param get_scope: A callable that returns an opaque object to be used
225
for detecting when this weave goes out of scope (should stop
226
answering requests or allowing mutation).
228
super(Weave, self).__init__()
234
self._weave_name = weave_name
236
self._matcher = bzrlib.patiencediff.PatienceSequenceMatcher
238
self._matcher = matcher
239
if get_scope is None:
240
get_scope = lambda:None
241
self._get_scope = get_scope
242
self._scope = get_scope()
243
self._access_mode = access_mode
244
self._allow_reserved = allow_reserved
247
return "Weave(%r)" % self._weave_name
249
def _check_write_ok(self):
250
"""Is the versioned file marked as 'finished' ? Raise if it is."""
251
if self._get_scope() != self._scope:
252
raise errors.OutSideTransaction()
253
if self._access_mode != 'w':
254
raise errors.ReadOnlyObjectDirtiedError(self)
257
"""Return a deep copy of self.
259
The copy can be modified without affecting the original weave."""
261
other._weave = self._weave[:]
262
other._parents = self._parents[:]
263
other._sha1s = self._sha1s[:]
264
other._names = self._names[:]
265
other._name_map = self._name_map.copy()
266
other._weave_name = self._weave_name
269
def __eq__(self, other):
270
if not isinstance(other, Weave):
272
return self._parents == other._parents \
273
and self._weave == other._weave \
274
and self._sha1s == other._sha1s
276
def __ne__(self, other):
277
return not self.__eq__(other)
279
def _idx_to_name(self, version):
280
return self._names[version]
282
def _lookup(self, name):
283
"""Convert symbolic version name to index."""
284
if not self._allow_reserved:
285
self.check_not_reserved_id(name)
287
return self._name_map[name]
289
raise RevisionNotPresent(name, self._weave_name)
292
"""See VersionedFile.versions."""
293
return self._names[:]
295
def has_version(self, version_id):
296
"""See VersionedFile.has_version."""
297
return (version_id in self._name_map)
299
__contains__ = has_version
301
def get_record_stream(self, versions, ordering, include_delta_closure):
302
"""Get a stream of records for versions.
304
:param versions: The versions to include. Each version is a tuple
306
:param ordering: Either 'unordered' or 'topological'. A topologically
307
sorted stream has compression parents strictly before their
309
:param include_delta_closure: If True then the closure across any
310
compression parents will be included (in the opaque data).
311
:return: An iterator of ContentFactory objects, each of which is only
312
valid until the iterator is advanced.
314
versions = [version[-1] for version in versions]
315
if ordering == 'topological':
316
parents = self.get_parent_map(versions)
317
new_versions = tsort.topo_sort(parents)
318
new_versions.extend(set(versions).difference(set(parents)))
319
versions = new_versions
320
elif ordering == 'groupcompress':
321
parents = self.get_parent_map(versions)
322
new_versions = sort_groupcompress(parents)
323
new_versions.extend(set(versions).difference(set(parents)))
324
versions = new_versions
325
for version in versions:
327
yield WeaveContentFactory(version, self)
329
yield AbsentContentFactory((version,))
331
def get_parent_map(self, version_ids):
332
"""See VersionedFile.get_parent_map."""
334
for version_id in version_ids:
335
if version_id == NULL_REVISION:
340
map(self._idx_to_name,
341
self._parents[self._lookup(version_id)]))
342
except RevisionNotPresent:
344
result[version_id] = parents
347
def get_parents_with_ghosts(self, version_id):
348
raise NotImplementedError(self.get_parents_with_ghosts)
350
def insert_record_stream(self, stream):
351
"""Insert a record stream into this versioned file.
353
:param stream: A stream of records to insert.
355
:seealso VersionedFile.get_record_stream:
358
for record in stream:
359
# Raise an error when a record is missing.
360
if record.storage_kind == 'absent':
361
raise RevisionNotPresent([record.key[0]], self)
362
# adapt to non-tuple interface
363
parents = [parent[0] for parent in record.parents]
364
if (record.storage_kind == 'fulltext'
365
or record.storage_kind == 'chunked'):
366
self.add_lines(record.key[0], parents,
367
osutils.chunks_to_lines(record.get_bytes_as('chunked')))
369
adapter_key = record.storage_kind, 'fulltext'
371
adapter = adapters[adapter_key]
373
adapter_factory = adapter_registry.get(adapter_key)
374
adapter = adapter_factory(self)
375
adapters[adapter_key] = adapter
376
lines = split_lines(adapter.get_bytes(record))
378
self.add_lines(record.key[0], parents, lines)
379
except RevisionAlreadyPresent:
382
def _check_repeated_add(self, name, parents, text, sha1):
383
"""Check that a duplicated add is OK.
385
If it is, return the (old) index; otherwise raise an exception.
387
idx = self._lookup(name)
388
if sorted(self._parents[idx]) != sorted(parents) \
389
or sha1 != self._sha1s[idx]:
390
raise RevisionAlreadyPresent(name, self._weave_name)
393
def _add_lines(self, version_id, parents, lines, parent_texts,
394
left_matching_blocks, nostore_sha, random_id, check_content):
395
"""See VersionedFile.add_lines."""
396
idx = self._add(version_id, lines, map(self._lookup, parents),
397
nostore_sha=nostore_sha)
398
return sha_strings(lines), sum(map(len, lines)), idx
400
def _add(self, version_id, lines, parents, sha1=None, nostore_sha=None):
401
"""Add a single text on top of the weave.
403
Returns the index number of the newly added version.
406
Symbolic name for this version.
407
(Typically the revision-id of the revision that added it.)
408
If None, a name will be allocated based on the hash. (sha1:SHAHASH)
411
List or set of direct parent version numbers.
414
Sequence of lines to be added in the new version.
416
:param nostore_sha: See VersionedFile.add_lines.
418
self._check_lines_not_unicode(lines)
419
self._check_lines_are_lines(lines)
421
sha1 = sha_strings(lines)
422
if sha1 == nostore_sha:
423
raise errors.ExistingContent
424
if version_id is None:
425
version_id = "sha1:" + sha1
426
if version_id in self._name_map:
427
return self._check_repeated_add(version_id, parents, lines, sha1)
429
self._check_versions(parents)
430
## self._check_lines(lines)
431
new_version = len(self._parents)
433
# if we abort after here the (in-memory) weave will be corrupt because only
434
# some fields are updated
435
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
436
# - Robert Collins 20060226
437
self._parents.append(parents[:])
438
self._sha1s.append(sha1)
439
self._names.append(version_id)
440
self._name_map[version_id] = new_version
444
# special case; adding with no parents revision; can do
445
# this more quickly by just appending unconditionally.
446
# even more specially, if we're adding an empty text we
447
# need do nothing at all.
449
self._weave.append(('{', new_version))
450
self._weave.extend(lines)
451
self._weave.append(('}', None))
454
if len(parents) == 1:
455
pv = list(parents)[0]
456
if sha1 == self._sha1s[pv]:
457
# special case: same as the single parent
461
ancestors = self._inclusions(parents)
465
# basis a list of (origin, lineno, line)
468
for origin, lineno, line in self._extract(ancestors):
469
basis_lineno.append(lineno)
470
basis_lines.append(line)
472
# another small special case: a merge, producing the same text
474
if lines == basis_lines:
477
# add a sentinel, because we can also match against the final line
478
basis_lineno.append(len(self._weave))
480
# XXX: which line of the weave should we really consider
481
# matches the end of the file? the current code says it's the
482
# last line of the weave?
484
#print 'basis_lines:', basis_lines
485
#print 'new_lines: ', lines
487
s = self._matcher(None, basis_lines, lines)
489
# offset gives the number of lines that have been inserted
490
# into the weave up to the current point; if the original edit instruction
491
# says to change line A then we actually change (A+offset)
494
for tag, i1, i2, j1, j2 in s.get_opcodes():
495
# i1,i2 are given in offsets within basis_lines; we need to map them
496
# back to offsets within the entire weave
497
#print 'raw match', tag, i1, i2, j1, j2
500
i1 = basis_lineno[i1]
501
i2 = basis_lineno[i2]
502
# the deletion and insertion are handled separately.
503
# first delete the region.
505
self._weave.insert(i1+offset, ('[', new_version))
506
self._weave.insert(i2+offset+1, (']', new_version))
510
# there may have been a deletion spanning up to
511
# i2; we want to insert after this region to make sure
512
# we don't destroy ourselves
514
self._weave[i:i] = ([('{', new_version)]
517
offset += 2 + (j2 - j1)
520
def _inclusions(self, versions):
521
"""Return set of all ancestors of given version(s)."""
522
if not len(versions):
525
for v in xrange(max(versions), 0, -1):
527
# include all its parents
528
i.update(self._parents[v])
530
## except IndexError:
531
## raise ValueError("version %d not present in weave" % v)
533
def get_ancestry(self, version_ids, topo_sorted=True):
534
"""See VersionedFile.get_ancestry."""
535
if isinstance(version_ids, basestring):
536
version_ids = [version_ids]
537
i = self._inclusions([self._lookup(v) for v in version_ids])
538
return [self._idx_to_name(v) for v in i]
540
def _check_lines(self, text):
541
if not isinstance(text, list):
542
raise ValueError("text should be a list, not %s" % type(text))
545
if not isinstance(l, basestring):
546
raise ValueError("text line should be a string or unicode, not %s"
551
def _check_versions(self, indexes):
552
"""Check everything in the sequence of indexes is valid"""
557
raise IndexError("invalid version number %r" % i)
559
def _compatible_parents(self, my_parents, other_parents):
560
"""During join check that other_parents are joinable with my_parents.
562
Joinable is defined as 'is a subset of' - supersets may require
563
regeneration of diffs, but subsets do not.
565
return len(other_parents.difference(my_parents)) == 0
567
def annotate(self, version_id):
568
"""Return a list of (version-id, line) tuples for version_id.
570
The index indicates when the line originated in the weave."""
571
incls = [self._lookup(version_id)]
572
return [(self._idx_to_name(origin), text) for origin, lineno, text in
573
self._extract(incls)]
575
def iter_lines_added_or_present_in_versions(self, version_ids=None,
577
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
578
if version_ids is None:
579
version_ids = self.versions()
580
version_ids = set(version_ids)
581
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
582
if inserted not in version_ids: continue
584
yield line + '\n', inserted
588
def _walk_internal(self, version_ids=None):
589
"""Helper method for weave actions."""
594
lineno = 0 # line of weave, 0-based
596
for l in self._weave:
597
if l.__class__ == tuple:
601
istack.append(self._names[v])
605
dset.add(self._names[v])
607
dset.remove(self._names[v])
609
raise WeaveFormatError('unexpected instruction %r' % v)
611
yield lineno, istack[-1], frozenset(dset), l
615
raise WeaveFormatError("unclosed insertion blocks "
616
"at end of weave: %s" % istack)
618
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
621
def plan_merge(self, ver_a, ver_b):
622
"""Return pseudo-annotation indicating how the two versions merge.
624
This is computed between versions a and b and their common
627
Weave lines present in none of them are skipped entirely.
629
inc_a = set(self.get_ancestry([ver_a]))
630
inc_b = set(self.get_ancestry([ver_b]))
631
inc_c = inc_a & inc_b
633
for lineno, insert, deleteset, line in self._walk_internal([ver_a, ver_b]):
634
if deleteset & inc_c:
635
# killed in parent; can't be in either a or b
636
# not relevant to our work
637
yield 'killed-base', line
638
elif insert in inc_c:
639
# was inserted in base
640
killed_a = bool(deleteset & inc_a)
641
killed_b = bool(deleteset & inc_b)
642
if killed_a and killed_b:
643
yield 'killed-both', line
645
yield 'killed-a', line
647
yield 'killed-b', line
649
yield 'unchanged', line
650
elif insert in inc_a:
651
if deleteset & inc_a:
652
yield 'ghost-a', line
656
elif insert in inc_b:
657
if deleteset & inc_b:
658
yield 'ghost-b', line
662
# not in either revision
663
yield 'irrelevant', line
665
def _extract(self, versions):
666
"""Yield annotation of lines in included set.
668
Yields a sequence of tuples (origin, lineno, text), where
669
origin is the origin version, lineno the index in the weave,
670
and text the text of the line.
672
The set typically but not necessarily corresponds to a version.
675
if not isinstance(i, int):
678
included = self._inclusions(versions)
684
lineno = 0 # line of weave, 0-based
690
WFE = WeaveFormatError
693
# 449 0 4474.6820 2356.5590 bzrlib.weave:556(_extract)
694
# +285282 0 1676.8040 1676.8040 +<isinstance>
695
# 1.6 seconds in 'isinstance'.
696
# changing the first isinstance:
697
# 449 0 2814.2660 1577.1760 bzrlib.weave:556(_extract)
698
# +140414 0 762.8050 762.8050 +<isinstance>
699
# note that the inline time actually dropped (less function calls)
700
# and total processing time was halved.
701
# we're still spending ~1/4 of the method in isinstance though.
702
# so lets hard code the acceptable string classes we expect:
703
# 449 0 1202.9420 786.2930 bzrlib.weave:556(_extract)
704
# +71352 0 377.5560 377.5560 +<method 'append' of 'list'
706
# yay, down to ~1/4 the initial extract time, and our inline time
707
# has shrunk again, with isinstance no longer dominating.
708
# tweaking the stack inclusion test to use a set gives:
709
# 449 0 1122.8030 713.0080 bzrlib.weave:556(_extract)
710
# +71352 0 354.9980 354.9980 +<method 'append' of 'list'
712
# - a 5% win, or possibly just noise. However with large istacks that
713
# 'in' test could dominate, so I'm leaving this change in place -
714
# when its fast enough to consider profiling big datasets we can review.
719
for l in self._weave:
720
if l.__class__ == tuple:
727
iset.remove(istack.pop())
735
raise AssertionError()
738
isactive = (not dset) and istack and (istack[-1] in included)
740
result.append((istack[-1], lineno, l))
743
raise WeaveFormatError("unclosed insertion blocks "
744
"at end of weave: %s" % istack)
746
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
750
def _maybe_lookup(self, name_or_index):
751
"""Convert possible symbolic name to index, or pass through indexes.
755
if isinstance(name_or_index, (int, long)):
758
return self._lookup(name_or_index)
760
def get_lines(self, version_id):
761
"""See VersionedFile.get_lines()."""
762
int_index = self._maybe_lookup(version_id)
763
result = [line for (origin, lineno, line) in self._extract([int_index])]
764
expected_sha1 = self._sha1s[int_index]
765
measured_sha1 = sha_strings(result)
766
if measured_sha1 != expected_sha1:
767
raise errors.WeaveInvalidChecksum(
768
'file %s, revision %s, expected: %s, measured %s'
769
% (self._weave_name, version_id,
770
expected_sha1, measured_sha1))
773
def get_sha1s(self, version_ids):
774
"""See VersionedFile.get_sha1s()."""
776
for v in version_ids:
777
result[v] = self._sha1s[self._lookup(v)]
780
def num_versions(self):
781
"""How many versions are in this weave?"""
782
l = len(self._parents)
785
__len__ = num_versions
787
def check(self, progress_bar=None):
788
# TODO evaluate performance hit of using string sets in this routine.
789
# TODO: check no circular inclusions
790
# TODO: create a nested progress bar
791
for version in range(self.num_versions()):
792
inclusions = list(self._parents[version])
795
if inclusions[-1] >= version:
796
raise WeaveFormatError("invalid included version %d for index %d"
797
% (inclusions[-1], version))
799
# try extracting all versions; parallel extraction is used
800
nv = self.num_versions()
805
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
806
# The problem is that set membership is much more expensive
807
name = self._idx_to_name(i)
810
new_inc = set([name])
811
for p in self._parents[i]:
812
new_inc.update(inclusions[self._idx_to_name(p)])
814
if set(new_inc) != set(self.get_ancestry(name)):
815
raise AssertionError(
817
% (set(new_inc), set(self.get_ancestry(name))))
818
inclusions[name] = new_inc
820
nlines = len(self._weave)
822
update_text = 'checking weave'
824
short_name = os.path.basename(self._weave_name)
825
update_text = 'checking %s' % (short_name,)
826
update_text = update_text[:25]
828
for lineno, insert, deleteset, line in self._walk_internal():
830
progress_bar.update(update_text, lineno, nlines)
832
for name, name_inclusions in inclusions.items():
833
# The active inclusion must be an ancestor,
834
# and no ancestors must have deleted this line,
835
# because we don't support resurrection.
836
if (insert in name_inclusions) and not (deleteset & name_inclusions):
837
sha1s[name].update(line)
840
version = self._idx_to_name(i)
841
hd = sha1s[version].hexdigest()
842
expected = self._sha1s[i]
844
raise errors.WeaveInvalidChecksum(
845
"mismatched sha1 for version %s: "
846
"got %s, expected %s"
847
% (version, hd, expected))
849
# TODO: check insertions are properly nested, that there are
850
# no lines outside of insertion blocks, that deletions are
851
# properly paired, etc.
853
def _imported_parents(self, other, other_idx):
854
"""Return list of parents in self corresponding to indexes in other."""
856
for parent_idx in other._parents[other_idx]:
857
parent_name = other._names[parent_idx]
858
if parent_name not in self._name_map:
859
# should not be possible
860
raise WeaveError("missing parent {%s} of {%s} in %r"
861
% (parent_name, other._name_map[other_idx], self))
862
new_parents.append(self._name_map[parent_name])
865
def _check_version_consistent(self, other, other_idx, name):
866
"""Check if a version in consistent in this and other.
868
To be consistent it must have:
871
* the same direct parents (by name, not index, and disregarding
874
If present & correct return True;
875
if not present in self return False;
876
if inconsistent raise error."""
877
this_idx = self._name_map.get(name, -1)
879
if self._sha1s[this_idx] != other._sha1s[other_idx]:
880
raise errors.WeaveTextDiffers(name, self, other)
881
self_parents = self._parents[this_idx]
882
other_parents = other._parents[other_idx]
883
n1 = set([self._names[i] for i in self_parents])
884
n2 = set([other._names[i] for i in other_parents])
885
if not self._compatible_parents(n1, n2):
886
raise WeaveParentMismatch("inconsistent parents "
887
"for version {%s}: %s vs %s" % (name, n1, n2))
893
def _reweave(self, other, pb, msg):
894
"""Reweave self with other - internal helper for join().
896
:param other: The other weave to merge
897
:param pb: An optional progress bar, indicating how far done we are
898
:param msg: An optional message for the progress
900
new_weave = _reweave(self, other, pb=pb, msg=msg)
901
self._copy_weave_content(new_weave)
903
def _copy_weave_content(self, otherweave):
904
"""adsorb the content from otherweave."""
905
for attr in self.__slots__:
906
if attr != '_weave_name':
907
setattr(self, attr, copy(getattr(otherweave, attr)))
910
class WeaveFile(Weave):
911
"""A WeaveFile represents a Weave on disk and writes on change."""
913
WEAVE_SUFFIX = '.weave'
915
def __init__(self, name, transport, filemode=None, create=False, access_mode='w', get_scope=None):
916
"""Create a WeaveFile.
918
:param create: If not True, only open an existing knit.
920
super(WeaveFile, self).__init__(name, access_mode, get_scope=get_scope,
921
allow_reserved=False)
922
self._transport = transport
923
self._filemode = filemode
925
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
926
except errors.NoSuchFile:
932
def _add_lines(self, version_id, parents, lines, parent_texts,
933
left_matching_blocks, nostore_sha, random_id, check_content):
934
"""Add a version and save the weave."""
935
self.check_not_reserved_id(version_id)
936
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
937
parent_texts, left_matching_blocks, nostore_sha, random_id,
942
def copy_to(self, name, transport):
943
"""See VersionedFile.copy_to()."""
944
# as we are all in memory always, just serialise to the new place.
946
write_weave_v5(self, sio)
948
transport.put_file(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
951
"""Save the weave."""
952
self._check_write_ok()
954
write_weave_v5(self, sio)
956
bytes = sio.getvalue()
957
path = self._weave_name + WeaveFile.WEAVE_SUFFIX
959
self._transport.put_bytes(path, bytes, self._filemode)
960
except errors.NoSuchFile:
961
self._transport.mkdir(dirname(path))
962
self._transport.put_bytes(path, bytes, self._filemode)
966
"""See VersionedFile.get_suffixes()."""
967
return [WeaveFile.WEAVE_SUFFIX]
969
def insert_record_stream(self, stream):
970
super(WeaveFile, self).insert_record_stream(stream)
974
def _reweave(wa, wb, pb=None, msg=None):
975
"""Combine two weaves and return the result.
977
This works even if a revision R has different parents in
978
wa and wb. In the resulting weave all the parents are given.
980
This is done by just building up a new weave, maintaining ordering
981
of the versions in the two inputs. More efficient approaches
982
might be possible but it should only be necessary to do
983
this operation rarely, when a new previously ghost version is
986
:param pb: An optional progress bar, indicating how far done we are
987
:param msg: An optional message for the progress
991
queue_a = range(wa.num_versions())
992
queue_b = range(wb.num_versions())
993
# first determine combined parents of all versions
994
# map from version name -> all parent names
995
combined_parents = _reweave_parent_graphs(wa, wb)
996
mutter("combined parents: %r", combined_parents)
997
order = tsort.topo_sort(combined_parents.iteritems())
998
mutter("order to reweave: %r", order)
1003
for idx, name in enumerate(order):
1005
pb.update(msg, idx, len(order))
1006
if name in wa._name_map:
1007
lines = wa.get_lines(name)
1008
if name in wb._name_map:
1009
lines_b = wb.get_lines(name)
1010
if lines != lines_b:
1011
mutter('Weaves differ on content. rev_id {%s}', name)
1012
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
1014
lines = list(difflib.unified_diff(lines, lines_b,
1015
wa._weave_name, wb._weave_name))
1016
mutter('lines:\n%s', ''.join(lines))
1017
raise errors.WeaveTextDiffers(name, wa, wb)
1019
lines = wb.get_lines(name)
1020
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
1024
def _reweave_parent_graphs(wa, wb):
1025
"""Return combined parent ancestry for two weaves.
1027
Returned as a list of (version_name, set(parent_names))"""
1029
for weave in [wa, wb]:
1030
for idx, name in enumerate(weave._names):
1031
p = combined.setdefault(name, set())
1032
p.update(map(weave._idx_to_name, weave._parents[idx]))
added
removed
bzrlib/weave.py
