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
# Author: Martin Pool <mbp@canonical.com>
20
"""Weave - storage of related text file versions"""
23
# XXX: If we do weaves this way, will a merge still behave the same
24
# way if it's done in a different order? That's a pretty desirable
27
# TODO: Nothing here so far assumes the lines are really \n newlines,
28
# rather than being split up in some other way. We could accommodate
29
# binaries, perhaps by naively splitting on \n or perhaps using
30
# something like a rolling checksum.
32
# TODO: End marker for each version so we can stop reading?
34
# TODO: Check that no insertion occurs inside a deletion that was
35
# active in the version of the insertion.
37
# TODO: In addition to the SHA-1 check, perhaps have some code that
38
# checks structural constraints of the weave: ie that insertions are
39
# properly nested, that there is no text outside of an insertion, that
40
# insertions or deletions are not repeated, etc.
42
# TODO: Parallel-extract that passes back each line along with a
43
# description of which revisions include it. Nice for checking all
44
# shas or calculating stats in parallel.
46
# TODO: Using a single _extract routine and then processing the output
47
# is probably inefficient. It's simple enough that we can afford to
48
# have slight specializations for different ways its used: annotate,
49
# basis for add, get, etc.
51
# TODO: Probably the API should work only in names to hide the integer
52
# indexes from the user.
54
# TODO: Is there any potential performance win by having an add()
55
# variant that is passed a pre-cooked version of the single basis
58
# TODO: Reweave can possibly be made faster by remembering diffs
59
# where the basis and destination are unchanged.
61
# FIXME: Sometimes we will be given a parents list for a revision
62
# that includes some redundant parents (i.e. already a parent of
63
# something in the list.) We should eliminate them. This can
64
# be done fairly efficiently because the sequence numbers constrain
65
# the possible relationships.
67
# FIXME: the conflict markers should be *7* characters
70
from cStringIO import StringIO
75
from bzrlib.lazy_import import lazy_import
76
lazy_import(globals(), """
77
from bzrlib import tsort
84
from bzrlib.errors import (WeaveError, WeaveFormatError, WeaveParentMismatch,
85
RevisionAlreadyPresent,
87
UnavailableRepresentation,
88
WeaveRevisionAlreadyPresent,
89
WeaveRevisionNotPresent,
91
from bzrlib.osutils import dirname, sha, sha_strings, split_lines
92
import bzrlib.patiencediff
93
from bzrlib.revision import NULL_REVISION
94
from bzrlib.symbol_versioning import *
95
from bzrlib.trace import mutter
96
from bzrlib.versionedfile import (
103
from bzrlib.weavefile import _read_weave_v5, write_weave_v5
106
class WeaveContentFactory(ContentFactory):
107
"""Content factory for streaming from weaves.
109
:seealso ContentFactory:
112
def __init__(self, version, weave):
113
"""Create a WeaveContentFactory for version from weave."""
114
ContentFactory.__init__(self)
115
self.sha1 = weave.get_sha1s([version])[version]
116
self.key = (version,)
117
parents = weave.get_parent_map([version])[version]
118
self.parents = tuple((parent,) for parent in parents)
119
self.storage_kind = 'fulltext'
122
def get_bytes_as(self, storage_kind):
123
if storage_kind == 'fulltext':
124
return self._weave.get_text(self.key[-1])
125
elif storage_kind == 'chunked':
126
return self._weave.get_lines(self.key[-1])
128
raise UnavailableRepresentation(self.key, storage_kind, 'fulltext')
131
class Weave(VersionedFile):
132
"""weave - versioned text file storage.
134
A Weave manages versions of line-based text files, keeping track
135
of the originating version for each line.
137
To clients the "lines" of the file are represented as a list of strings.
138
These strings will typically have terminal newline characters, but
139
this is not required. In particular files commonly do not have a newline
140
at the end of the file.
142
Texts can be identified in either of two ways:
144
* a nonnegative index number.
146
* a version-id string.
148
Typically the index number will be valid only inside this weave and
149
the version-id is used to reference it in the larger world.
151
The weave is represented as a list mixing edit instructions and
152
literal text. Each entry in _weave can be either a string (or
153
unicode), or a tuple. If a string, it means that the given line
154
should be output in the currently active revisions.
156
If a tuple, it gives a processing instruction saying in which
157
revisions the enclosed lines are active. The tuple has the form
158
(instruction, version).
160
The instruction can be '{' or '}' for an insertion block, and '['
161
and ']' for a deletion block respectively. The version is the
162
integer version index. There is no replace operator, only deletes
163
and inserts. For '}', the end of an insertion, there is no
164
version parameter because it always closes the most recently
169
* A later version can delete lines that were introduced by any
170
number of ancestor versions; this implies that deletion
171
instructions can span insertion blocks without regard to the
172
insertion block's nesting.
174
* Similarly, deletions need not be properly nested with regard to
175
each other, because they might have been generated by
176
independent revisions.
178
* Insertions are always made by inserting a new bracketed block
179
into a single point in the previous weave. This implies they
180
can nest but not overlap, and the nesting must always have later
181
insertions on the inside.
183
* It doesn't seem very useful to have an active insertion
184
inside an inactive insertion, but it might happen.
186
* Therefore, all instructions are always"considered"; that
187
is passed onto and off the stack. An outer inactive block
188
doesn't disable an inner block.
190
* Lines are enabled if the most recent enclosing insertion is
191
active and none of the enclosing deletions are active.
193
* There is no point having a deletion directly inside its own
194
insertion; you might as well just not write it. And there
195
should be no way to get an earlier version deleting a later
199
Text of the weave; list of control instruction tuples and strings.
202
List of parents, indexed by version number.
203
It is only necessary to store the minimal set of parents for
204
each version; the parent's parents are implied.
207
List of hex SHA-1 of each version.
210
List of symbolic names for each version. Each should be unique.
213
For each name, the version number.
216
Descriptive name of this weave; typically the filename if known.
220
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
221
'_weave_name', '_matcher', '_allow_reserved']
223
def __init__(self, weave_name=None, access_mode='w', matcher=None,
224
get_scope=None, allow_reserved=False):
227
:param get_scope: A callable that returns an opaque object to be used
228
for detecting when this weave goes out of scope (should stop
229
answering requests or allowing mutation).
231
super(Weave, self).__init__()
237
self._weave_name = weave_name
239
self._matcher = bzrlib.patiencediff.PatienceSequenceMatcher
241
self._matcher = matcher
242
if get_scope is None:
243
get_scope = lambda:None
244
self._get_scope = get_scope
245
self._scope = get_scope()
246
self._access_mode = access_mode
247
self._allow_reserved = allow_reserved
250
return "Weave(%r)" % self._weave_name
252
def _check_write_ok(self):
253
"""Is the versioned file marked as 'finished' ? Raise if it is."""
254
if self._get_scope() != self._scope:
255
raise errors.OutSideTransaction()
256
if self._access_mode != 'w':
257
raise errors.ReadOnlyObjectDirtiedError(self)
260
"""Return a deep copy of self.
262
The copy can be modified without affecting the original weave."""
264
other._weave = self._weave[:]
265
other._parents = self._parents[:]
266
other._sha1s = self._sha1s[:]
267
other._names = self._names[:]
268
other._name_map = self._name_map.copy()
269
other._weave_name = self._weave_name
272
def __eq__(self, other):
273
if not isinstance(other, Weave):
275
return self._parents == other._parents \
276
and self._weave == other._weave \
277
and self._sha1s == other._sha1s
279
def __ne__(self, other):
280
return not self.__eq__(other)
282
def _idx_to_name(self, version):
283
return self._names[version]
285
def _lookup(self, name):
286
"""Convert symbolic version name to index."""
287
if not self._allow_reserved:
288
self.check_not_reserved_id(name)
290
return self._name_map[name]
292
raise RevisionNotPresent(name, self._weave_name)
295
"""See VersionedFile.versions."""
296
return self._names[:]
298
def has_version(self, version_id):
299
"""See VersionedFile.has_version."""
300
return (version_id in self._name_map)
302
__contains__ = has_version
304
def get_record_stream(self, versions, ordering, include_delta_closure):
305
"""Get a stream of records for versions.
307
:param versions: The versions to include. Each version is a tuple
309
:param ordering: Either 'unordered' or 'topological'. A topologically
310
sorted stream has compression parents strictly before their
312
:param include_delta_closure: If True then the closure across any
313
compression parents will be included (in the opaque data).
314
:return: An iterator of ContentFactory objects, each of which is only
315
valid until the iterator is advanced.
317
versions = [version[-1] for version in versions]
318
if ordering == 'topological':
319
parents = self.get_parent_map(versions)
320
new_versions = tsort.topo_sort(parents)
321
new_versions.extend(set(versions).difference(set(parents)))
322
versions = new_versions
323
elif ordering == 'groupcompress':
324
parents = self.get_parent_map(versions)
325
new_versions = sort_groupcompress(parents)
326
new_versions.extend(set(versions).difference(set(parents)))
327
versions = new_versions
328
for version in versions:
330
yield WeaveContentFactory(version, self)
332
yield AbsentContentFactory((version,))
334
def get_parent_map(self, version_ids):
335
"""See VersionedFile.get_parent_map."""
337
for version_id in version_ids:
338
if version_id == NULL_REVISION:
343
map(self._idx_to_name,
344
self._parents[self._lookup(version_id)]))
345
except RevisionNotPresent:
347
result[version_id] = parents
350
def get_parents_with_ghosts(self, version_id):
351
raise NotImplementedError(self.get_parents_with_ghosts)
353
def insert_record_stream(self, stream):
354
"""Insert a record stream into this versioned file.
356
:param stream: A stream of records to insert.
358
:seealso VersionedFile.get_record_stream:
361
for record in stream:
362
# Raise an error when a record is missing.
363
if record.storage_kind == 'absent':
364
raise RevisionNotPresent([record.key[0]], self)
365
# adapt to non-tuple interface
366
parents = [parent[0] for parent in record.parents]
367
if (record.storage_kind == 'fulltext'
368
or record.storage_kind == 'chunked'):
369
self.add_lines(record.key[0], parents,
370
osutils.chunks_to_lines(record.get_bytes_as('chunked')))
372
adapter_key = record.storage_kind, 'fulltext'
374
adapter = adapters[adapter_key]
376
adapter_factory = adapter_registry.get(adapter_key)
377
adapter = adapter_factory(self)
378
adapters[adapter_key] = adapter
379
lines = split_lines(adapter.get_bytes(record))
381
self.add_lines(record.key[0], parents, lines)
382
except RevisionAlreadyPresent:
385
def _check_repeated_add(self, name, parents, text, sha1):
386
"""Check that a duplicated add is OK.
388
If it is, return the (old) index; otherwise raise an exception.
390
idx = self._lookup(name)
391
if sorted(self._parents[idx]) != sorted(parents) \
392
or sha1 != self._sha1s[idx]:
393
raise RevisionAlreadyPresent(name, self._weave_name)
396
def _add_lines(self, version_id, parents, lines, parent_texts,
397
left_matching_blocks, nostore_sha, random_id, check_content):
398
"""See VersionedFile.add_lines."""
399
idx = self._add(version_id, lines, map(self._lookup, parents),
400
nostore_sha=nostore_sha)
401
return sha_strings(lines), sum(map(len, lines)), idx
403
def _add(self, version_id, lines, parents, sha1=None, nostore_sha=None):
404
"""Add a single text on top of the weave.
406
Returns the index number of the newly added version.
409
Symbolic name for this version.
410
(Typically the revision-id of the revision that added it.)
411
If None, a name will be allocated based on the hash. (sha1:SHAHASH)
414
List or set of direct parent version numbers.
417
Sequence of lines to be added in the new version.
419
:param nostore_sha: See VersionedFile.add_lines.
421
self._check_lines_not_unicode(lines)
422
self._check_lines_are_lines(lines)
424
sha1 = sha_strings(lines)
425
if sha1 == nostore_sha:
426
raise errors.ExistingContent
427
if version_id is None:
428
version_id = "sha1:" + sha1
429
if version_id in self._name_map:
430
return self._check_repeated_add(version_id, parents, lines, sha1)
432
self._check_versions(parents)
433
## self._check_lines(lines)
434
new_version = len(self._parents)
436
# if we abort after here the (in-memory) weave will be corrupt because only
437
# some fields are updated
438
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
439
# - Robert Collins 20060226
440
self._parents.append(parents[:])
441
self._sha1s.append(sha1)
442
self._names.append(version_id)
443
self._name_map[version_id] = new_version
447
# special case; adding with no parents revision; can do
448
# this more quickly by just appending unconditionally.
449
# even more specially, if we're adding an empty text we
450
# need do nothing at all.
452
self._weave.append(('{', new_version))
453
self._weave.extend(lines)
454
self._weave.append(('}', None))
457
if len(parents) == 1:
458
pv = list(parents)[0]
459
if sha1 == self._sha1s[pv]:
460
# special case: same as the single parent
464
ancestors = self._inclusions(parents)
468
# basis a list of (origin, lineno, line)
471
for origin, lineno, line in self._extract(ancestors):
472
basis_lineno.append(lineno)
473
basis_lines.append(line)
475
# another small special case: a merge, producing the same text
477
if lines == basis_lines:
480
# add a sentinel, because we can also match against the final line
481
basis_lineno.append(len(self._weave))
483
# XXX: which line of the weave should we really consider
484
# matches the end of the file? the current code says it's the
485
# last line of the weave?
487
#print 'basis_lines:', basis_lines
488
#print 'new_lines: ', lines
490
s = self._matcher(None, basis_lines, lines)
492
# offset gives the number of lines that have been inserted
493
# into the weave up to the current point; if the original edit instruction
494
# says to change line A then we actually change (A+offset)
497
for tag, i1, i2, j1, j2 in s.get_opcodes():
498
# i1,i2 are given in offsets within basis_lines; we need to map them
499
# back to offsets within the entire weave
500
#print 'raw match', tag, i1, i2, j1, j2
503
i1 = basis_lineno[i1]
504
i2 = basis_lineno[i2]
505
# the deletion and insertion are handled separately.
506
# first delete the region.
508
self._weave.insert(i1+offset, ('[', new_version))
509
self._weave.insert(i2+offset+1, (']', new_version))
513
# there may have been a deletion spanning up to
514
# i2; we want to insert after this region to make sure
515
# we don't destroy ourselves
517
self._weave[i:i] = ([('{', new_version)]
520
offset += 2 + (j2 - j1)
523
def _inclusions(self, versions):
524
"""Return set of all ancestors of given version(s)."""
525
if not len(versions):
528
for v in xrange(max(versions), 0, -1):
530
# include all its parents
531
i.update(self._parents[v])
533
## except IndexError:
534
## raise ValueError("version %d not present in weave" % v)
536
def get_ancestry(self, version_ids, topo_sorted=True):
537
"""See VersionedFile.get_ancestry."""
538
if isinstance(version_ids, basestring):
539
version_ids = [version_ids]
540
i = self._inclusions([self._lookup(v) for v in version_ids])
541
return [self._idx_to_name(v) for v in i]
543
def _check_lines(self, text):
544
if not isinstance(text, list):
545
raise ValueError("text should be a list, not %s" % type(text))
548
if not isinstance(l, basestring):
549
raise ValueError("text line should be a string or unicode, not %s"
554
def _check_versions(self, indexes):
555
"""Check everything in the sequence of indexes is valid"""
560
raise IndexError("invalid version number %r" % i)
562
def _compatible_parents(self, my_parents, other_parents):
563
"""During join check that other_parents are joinable with my_parents.
565
Joinable is defined as 'is a subset of' - supersets may require
566
regeneration of diffs, but subsets do not.
568
return len(other_parents.difference(my_parents)) == 0
570
def annotate(self, version_id):
571
"""Return a list of (version-id, line) tuples for version_id.
573
The index indicates when the line originated in the weave."""
574
incls = [self._lookup(version_id)]
575
return [(self._idx_to_name(origin), text) for origin, lineno, text in
576
self._extract(incls)]
578
def iter_lines_added_or_present_in_versions(self, version_ids=None,
580
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
581
if version_ids is None:
582
version_ids = self.versions()
583
version_ids = set(version_ids)
584
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
585
if inserted not in version_ids: continue
587
yield line + '\n', inserted
591
def _walk_internal(self, version_ids=None):
592
"""Helper method for weave actions."""
597
lineno = 0 # line of weave, 0-based
599
for l in self._weave:
600
if l.__class__ == tuple:
604
istack.append(self._names[v])
608
dset.add(self._names[v])
610
dset.remove(self._names[v])
612
raise WeaveFormatError('unexpected instruction %r' % v)
614
yield lineno, istack[-1], frozenset(dset), l
618
raise WeaveFormatError("unclosed insertion blocks "
619
"at end of weave: %s" % istack)
621
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
624
def plan_merge(self, ver_a, ver_b):
625
"""Return pseudo-annotation indicating how the two versions merge.
627
This is computed between versions a and b and their common
630
Weave lines present in none of them are skipped entirely.
632
inc_a = set(self.get_ancestry([ver_a]))
633
inc_b = set(self.get_ancestry([ver_b]))
634
inc_c = inc_a & inc_b
636
for lineno, insert, deleteset, line in self._walk_internal([ver_a, ver_b]):
637
if deleteset & inc_c:
638
# killed in parent; can't be in either a or b
639
# not relevant to our work
640
yield 'killed-base', line
641
elif insert in inc_c:
642
# was inserted in base
643
killed_a = bool(deleteset & inc_a)
644
killed_b = bool(deleteset & inc_b)
645
if killed_a and killed_b:
646
yield 'killed-both', line
648
yield 'killed-a', line
650
yield 'killed-b', line
652
yield 'unchanged', line
653
elif insert in inc_a:
654
if deleteset & inc_a:
655
yield 'ghost-a', line
659
elif insert in inc_b:
660
if deleteset & inc_b:
661
yield 'ghost-b', line
665
# not in either revision
666
yield 'irrelevant', line
668
def _extract(self, versions):
669
"""Yield annotation of lines in included set.
671
Yields a sequence of tuples (origin, lineno, text), where
672
origin is the origin version, lineno the index in the weave,
673
and text the text of the line.
675
The set typically but not necessarily corresponds to a version.
678
if not isinstance(i, int):
681
included = self._inclusions(versions)
687
lineno = 0 # line of weave, 0-based
693
WFE = WeaveFormatError
696
# 449 0 4474.6820 2356.5590 bzrlib.weave:556(_extract)
697
# +285282 0 1676.8040 1676.8040 +<isinstance>
698
# 1.6 seconds in 'isinstance'.
699
# changing the first isinstance:
700
# 449 0 2814.2660 1577.1760 bzrlib.weave:556(_extract)
701
# +140414 0 762.8050 762.8050 +<isinstance>
702
# note that the inline time actually dropped (less function calls)
703
# and total processing time was halved.
704
# we're still spending ~1/4 of the method in isinstance though.
705
# so lets hard code the acceptable string classes we expect:
706
# 449 0 1202.9420 786.2930 bzrlib.weave:556(_extract)
707
# +71352 0 377.5560 377.5560 +<method 'append' of 'list'
709
# yay, down to ~1/4 the initial extract time, and our inline time
710
# has shrunk again, with isinstance no longer dominating.
711
# tweaking the stack inclusion test to use a set gives:
712
# 449 0 1122.8030 713.0080 bzrlib.weave:556(_extract)
713
# +71352 0 354.9980 354.9980 +<method 'append' of 'list'
715
# - a 5% win, or possibly just noise. However with large istacks that
716
# 'in' test could dominate, so I'm leaving this change in place -
717
# when its fast enough to consider profiling big datasets we can review.
722
for l in self._weave:
723
if l.__class__ == tuple:
730
iset.remove(istack.pop())
738
raise AssertionError()
741
isactive = (not dset) and istack and (istack[-1] in included)
743
result.append((istack[-1], lineno, l))
746
raise WeaveFormatError("unclosed insertion blocks "
747
"at end of weave: %s" % istack)
749
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
753
def _maybe_lookup(self, name_or_index):
754
"""Convert possible symbolic name to index, or pass through indexes.
758
if isinstance(name_or_index, (int, long)):
761
return self._lookup(name_or_index)
763
def get_lines(self, version_id):
764
"""See VersionedFile.get_lines()."""
765
int_index = self._maybe_lookup(version_id)
766
result = [line for (origin, lineno, line) in self._extract([int_index])]
767
expected_sha1 = self._sha1s[int_index]
768
measured_sha1 = sha_strings(result)
769
if measured_sha1 != expected_sha1:
770
raise errors.WeaveInvalidChecksum(
771
'file %s, revision %s, expected: %s, measured %s'
772
% (self._weave_name, version_id,
773
expected_sha1, measured_sha1))
776
def get_sha1s(self, version_ids):
777
"""See VersionedFile.get_sha1s()."""
779
for v in version_ids:
780
result[v] = self._sha1s[self._lookup(v)]
783
def num_versions(self):
784
"""How many versions are in this weave?"""
785
l = len(self._parents)
788
__len__ = num_versions
790
def check(self, progress_bar=None):
791
# TODO evaluate performance hit of using string sets in this routine.
792
# TODO: check no circular inclusions
793
# TODO: create a nested progress bar
794
for version in range(self.num_versions()):
795
inclusions = list(self._parents[version])
798
if inclusions[-1] >= version:
799
raise WeaveFormatError("invalid included version %d for index %d"
800
% (inclusions[-1], version))
802
# try extracting all versions; parallel extraction is used
803
nv = self.num_versions()
808
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
809
# The problem is that set membership is much more expensive
810
name = self._idx_to_name(i)
813
new_inc = set([name])
814
for p in self._parents[i]:
815
new_inc.update(inclusions[self._idx_to_name(p)])
817
if set(new_inc) != set(self.get_ancestry(name)):
818
raise AssertionError(
820
% (set(new_inc), set(self.get_ancestry(name))))
821
inclusions[name] = new_inc
823
nlines = len(self._weave)
825
update_text = 'checking weave'
827
short_name = os.path.basename(self._weave_name)
828
update_text = 'checking %s' % (short_name,)
829
update_text = update_text[:25]
831
for lineno, insert, deleteset, line in self._walk_internal():
833
progress_bar.update(update_text, lineno, nlines)
835
for name, name_inclusions in inclusions.items():
836
# The active inclusion must be an ancestor,
837
# and no ancestors must have deleted this line,
838
# because we don't support resurrection.
839
if (insert in name_inclusions) and not (deleteset & name_inclusions):
840
sha1s[name].update(line)
843
version = self._idx_to_name(i)
844
hd = sha1s[version].hexdigest()
845
expected = self._sha1s[i]
847
raise errors.WeaveInvalidChecksum(
848
"mismatched sha1 for version %s: "
849
"got %s, expected %s"
850
% (version, hd, expected))
852
# TODO: check insertions are properly nested, that there are
853
# no lines outside of insertion blocks, that deletions are
854
# properly paired, etc.
856
def _imported_parents(self, other, other_idx):
857
"""Return list of parents in self corresponding to indexes in other."""
859
for parent_idx in other._parents[other_idx]:
860
parent_name = other._names[parent_idx]
861
if parent_name not in self._name_map:
862
# should not be possible
863
raise WeaveError("missing parent {%s} of {%s} in %r"
864
% (parent_name, other._name_map[other_idx], self))
865
new_parents.append(self._name_map[parent_name])
868
def _check_version_consistent(self, other, other_idx, name):
869
"""Check if a version in consistent in this and other.
871
To be consistent it must have:
874
* the same direct parents (by name, not index, and disregarding
877
If present & correct return True;
878
if not present in self return False;
879
if inconsistent raise error."""
880
this_idx = self._name_map.get(name, -1)
882
if self._sha1s[this_idx] != other._sha1s[other_idx]:
883
raise errors.WeaveTextDiffers(name, self, other)
884
self_parents = self._parents[this_idx]
885
other_parents = other._parents[other_idx]
886
n1 = set([self._names[i] for i in self_parents])
887
n2 = set([other._names[i] for i in other_parents])
888
if not self._compatible_parents(n1, n2):
889
raise WeaveParentMismatch("inconsistent parents "
890
"for version {%s}: %s vs %s" % (name, n1, n2))
896
def _reweave(self, other, pb, msg):
897
"""Reweave self with other - internal helper for join().
899
:param other: The other weave to merge
900
:param pb: An optional progress bar, indicating how far done we are
901
:param msg: An optional message for the progress
903
new_weave = _reweave(self, other, pb=pb, msg=msg)
904
self._copy_weave_content(new_weave)
906
def _copy_weave_content(self, otherweave):
907
"""adsorb the content from otherweave."""
908
for attr in self.__slots__:
909
if attr != '_weave_name':
910
setattr(self, attr, copy(getattr(otherweave, attr)))
913
class WeaveFile(Weave):
914
"""A WeaveFile represents a Weave on disk and writes on change."""
916
WEAVE_SUFFIX = '.weave'
918
def __init__(self, name, transport, filemode=None, create=False, access_mode='w', get_scope=None):
919
"""Create a WeaveFile.
921
:param create: If not True, only open an existing knit.
923
super(WeaveFile, self).__init__(name, access_mode, get_scope=get_scope,
924
allow_reserved=False)
925
self._transport = transport
926
self._filemode = filemode
928
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
929
except errors.NoSuchFile:
935
def _add_lines(self, version_id, parents, lines, parent_texts,
936
left_matching_blocks, nostore_sha, random_id, check_content):
937
"""Add a version and save the weave."""
938
self.check_not_reserved_id(version_id)
939
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
940
parent_texts, left_matching_blocks, nostore_sha, random_id,
945
def copy_to(self, name, transport):
946
"""See VersionedFile.copy_to()."""
947
# as we are all in memory always, just serialise to the new place.
949
write_weave_v5(self, sio)
951
transport.put_file(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
954
"""Save the weave."""
955
self._check_write_ok()
957
write_weave_v5(self, sio)
959
bytes = sio.getvalue()
960
path = self._weave_name + WeaveFile.WEAVE_SUFFIX
962
self._transport.put_bytes(path, bytes, self._filemode)
963
except errors.NoSuchFile:
964
self._transport.mkdir(dirname(path))
965
self._transport.put_bytes(path, bytes, self._filemode)
969
"""See VersionedFile.get_suffixes()."""
970
return [WeaveFile.WEAVE_SUFFIX]
972
def insert_record_stream(self, stream):
973
super(WeaveFile, self).insert_record_stream(stream)
977
def _reweave(wa, wb, pb=None, msg=None):
978
"""Combine two weaves and return the result.
980
This works even if a revision R has different parents in
981
wa and wb. In the resulting weave all the parents are given.
983
This is done by just building up a new weave, maintaining ordering
984
of the versions in the two inputs. More efficient approaches
985
might be possible but it should only be necessary to do
986
this operation rarely, when a new previously ghost version is
989
:param pb: An optional progress bar, indicating how far done we are
990
:param msg: An optional message for the progress
994
queue_a = range(wa.num_versions())
995
queue_b = range(wb.num_versions())
996
# first determine combined parents of all versions
997
# map from version name -> all parent names
998
combined_parents = _reweave_parent_graphs(wa, wb)
999
mutter("combined parents: %r", combined_parents)
1000
order = tsort.topo_sort(combined_parents.iteritems())
1001
mutter("order to reweave: %r", order)
1006
for idx, name in enumerate(order):
1008
pb.update(msg, idx, len(order))
1009
if name in wa._name_map:
1010
lines = wa.get_lines(name)
1011
if name in wb._name_map:
1012
lines_b = wb.get_lines(name)
1013
if lines != lines_b:
1014
mutter('Weaves differ on content. rev_id {%s}', name)
1015
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
1017
lines = list(difflib.unified_diff(lines, lines_b,
1018
wa._weave_name, wb._weave_name))
1019
mutter('lines:\n%s', ''.join(lines))
1020
raise errors.WeaveTextDiffers(name, wa, wb)
1022
lines = wb.get_lines(name)
1023
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
1027
def _reweave_parent_graphs(wa, wb):
1028
"""Return combined parent ancestry for two weaves.
1030
Returned as a list of (version_name, set(parent_names))"""
1032
for weave in [wa, wb]:
1033
for idx, name in enumerate(weave._names):
1034
p = combined.setdefault(name, set())
1035
p.update(map(weave._idx_to_name, weave._parents[idx]))
added
removed
bzrlib/weave.py
