3
# Copyright (C) 2005 Canonical Ltd
5
# This program is free software; you can redistribute it and/or modify
6
# it under the terms of the GNU General Public License as published by
7
# the Free Software Foundation; either version 2 of the License, or
8
# (at your option) any later version.
10
# This program is distributed in the hope that it will be useful,
11
# but WITHOUT ANY WARRANTY; without even the implied warranty of
12
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13
# GNU General Public License for more details.
15
# You should have received a copy of the GNU General Public License
16
# along with this program; if not, write to the Free Software
17
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
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
81
from bzrlib.errors import (WeaveError, WeaveFormatError, WeaveParentMismatch,
82
RevisionAlreadyPresent,
84
UnavailableRepresentation,
85
WeaveRevisionAlreadyPresent,
86
WeaveRevisionNotPresent,
88
import bzrlib.errors as errors
89
from bzrlib.osutils import sha_strings, split_lines
90
import bzrlib.patiencediff
91
from bzrlib.trace import mutter
92
from bzrlib.tsort import topo_sort
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])[0]
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[0])
123
raise UnavailableRepresentation(self.key, storage_kind, 'fulltext')
126
class Weave(VersionedFile):
127
"""weave - versioned text file storage.
129
A Weave manages versions of line-based text files, keeping track
130
of the originating version for each line.
132
To clients the "lines" of the file are represented as a list of strings.
133
These strings will typically have terminal newline characters, but
134
this is not required. In particular files commonly do not have a newline
135
at the end of the file.
137
Texts can be identified in either of two ways:
139
* a nonnegative index number.
141
* a version-id string.
143
Typically the index number will be valid only inside this weave and
144
the version-id is used to reference it in the larger world.
146
The weave is represented as a list mixing edit instructions and
147
literal text. Each entry in _weave can be either a string (or
148
unicode), or a tuple. If a string, it means that the given line
149
should be output in the currently active revisions.
151
If a tuple, it gives a processing instruction saying in which
152
revisions the enclosed lines are active. The tuple has the form
153
(instruction, version).
155
The instruction can be '{' or '}' for an insertion block, and '['
156
and ']' for a deletion block respectively. The version is the
157
integer version index. There is no replace operator, only deletes
158
and inserts. For '}', the end of an insertion, there is no
159
version parameter because it always closes the most recently
164
* A later version can delete lines that were introduced by any
165
number of ancestor versions; this implies that deletion
166
instructions can span insertion blocks without regard to the
167
insertion block's nesting.
169
* Similarly, deletions need not be properly nested with regard to
170
each other, because they might have been generated by
171
independent revisions.
173
* Insertions are always made by inserting a new bracketed block
174
into a single point in the previous weave. This implies they
175
can nest but not overlap, and the nesting must always have later
176
insertions on the inside.
178
* It doesn't seem very useful to have an active insertion
179
inside an inactive insertion, but it might happen.
181
* Therefore, all instructions are always"considered"; that
182
is passed onto and off the stack. An outer inactive block
183
doesn't disable an inner block.
185
* Lines are enabled if the most recent enclosing insertion is
186
active and none of the enclosing deletions are active.
188
* There is no point having a deletion directly inside its own
189
insertion; you might as well just not write it. And there
190
should be no way to get an earlier version deleting a later
194
Text of the weave; list of control instruction tuples and strings.
197
List of parents, indexed by version number.
198
It is only necessary to store the minimal set of parents for
199
each version; the parent's parents are implied.
202
List of hex SHA-1 of each version.
205
List of symbolic names for each version. Each should be unique.
208
For each name, the version number.
211
Descriptive name of this weave; typically the filename if known.
215
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
216
'_weave_name', '_matcher']
218
def __init__(self, weave_name=None, access_mode='w', matcher=None, get_scope=None):
221
:param get_scope: A callable that returns an opaque object to be used
222
for detecting when this weave goes out of scope (should stop
223
answering requests or allowing mutation).
225
super(Weave, self).__init__(access_mode)
231
self._weave_name = weave_name
233
self._matcher = bzrlib.patiencediff.PatienceSequenceMatcher
235
self._matcher = matcher
236
if get_scope is None:
237
get_scope = lambda:None
238
self._get_scope = get_scope
239
self._scope = get_scope()
240
self._access_mode = access_mode
243
return "Weave(%r)" % self._weave_name
245
def _check_write_ok(self):
246
"""Is the versioned file marked as 'finished' ? Raise if it is."""
247
if self._get_scope() != self._scope:
248
raise errors.OutSideTransaction()
249
if self._access_mode != 'w':
250
raise errors.ReadOnlyObjectDirtiedError(self)
253
"""Return a deep copy of self.
255
The copy can be modified without affecting the original weave."""
257
other._weave = self._weave[:]
258
other._parents = self._parents[:]
259
other._sha1s = self._sha1s[:]
260
other._names = self._names[:]
261
other._name_map = self._name_map.copy()
262
other._weave_name = self._weave_name
265
def __eq__(self, other):
266
if not isinstance(other, Weave):
268
return self._parents == other._parents \
269
and self._weave == other._weave \
270
and self._sha1s == other._sha1s
272
def __ne__(self, other):
273
return not self.__eq__(other)
275
def _idx_to_name(self, version):
276
return self._names[version]
278
def _lookup(self, name):
279
"""Convert symbolic version name to index."""
280
self.check_not_reserved_id(name)
282
return self._name_map[name]
284
raise RevisionNotPresent(name, self._weave_name)
287
"""See VersionedFile.versions."""
288
return self._names[:]
290
def has_version(self, version_id):
291
"""See VersionedFile.has_version."""
292
return (version_id in self._name_map)
294
__contains__ = has_version
296
def get_record_stream(self, versions, ordering, include_delta_closure):
297
"""Get a stream of records for versions.
299
:param versions: The versions to include. Each version is a tuple
301
:param ordering: Either 'unordered' or 'topological'. A topologically
302
sorted stream has compression parents strictly before their
304
:param include_delta_closure: If True then the closure across any
305
compression parents will be included (in the opaque data).
306
:return: An iterator of ContentFactory objects, each of which is only
307
valid until the iterator is advanced.
309
if ordering == 'topological':
310
parents = self.get_parent_map(versions)
311
new_versions = topo_sort(parents)
312
new_versions.extend(set(versions).difference(set(parents)))
313
versions = new_versions
314
for version in versions:
316
yield WeaveContentFactory(version, self)
318
yield AbsentContentFactory((version,))
320
def get_parent_map(self, version_ids):
321
"""See VersionedFile.get_parent_map."""
323
for version_id in version_ids:
325
result[version_id] = tuple(
326
map(self._idx_to_name, self._parents[self._lookup(version_id)]))
327
except RevisionNotPresent:
331
def get_parents_with_ghosts(self, version_id):
332
raise NotImplementedError(self.get_parents_with_ghosts)
334
def insert_record_stream(self, stream):
335
"""Insert a record stream into this versioned file.
337
:param stream: A stream of records to insert.
339
:seealso VersionedFile.get_record_stream:
342
for record in stream:
343
# Raise an error when a record is missing.
344
if record.storage_kind == 'absent':
345
raise RevisionNotPresent([record.key[0]], self)
346
# adapt to non-tuple interface
347
parents = [parent[0] for parent in record.parents]
348
if record.storage_kind == 'fulltext':
349
self.add_lines(record.key[0], parents,
350
split_lines(record.get_bytes_as('fulltext')))
352
adapter_key = record.storage_kind, 'fulltext'
354
adapter = adapters[adapter_key]
356
adapter_factory = adapter_registry.get(adapter_key)
357
adapter = adapter_factory(self)
358
adapters[adapter_key] = adapter
359
lines = split_lines(adapter.get_bytes(
360
record, record.get_bytes_as(record.storage_kind)))
362
self.add_lines(record.key[0], parents, lines)
363
except RevisionAlreadyPresent:
366
def _check_repeated_add(self, name, parents, text, sha1):
367
"""Check that a duplicated add is OK.
369
If it is, return the (old) index; otherwise raise an exception.
371
idx = self._lookup(name)
372
if sorted(self._parents[idx]) != sorted(parents) \
373
or sha1 != self._sha1s[idx]:
374
raise RevisionAlreadyPresent(name, self._weave_name)
377
def _add_lines(self, version_id, parents, lines, parent_texts,
378
left_matching_blocks, nostore_sha, random_id, check_content):
379
"""See VersionedFile.add_lines."""
380
idx = self._add(version_id, lines, map(self._lookup, parents),
381
nostore_sha=nostore_sha)
382
return sha_strings(lines), sum(map(len, lines)), idx
384
def _add(self, version_id, lines, parents, sha1=None, nostore_sha=None):
385
"""Add a single text on top of the weave.
387
Returns the index number of the newly added version.
390
Symbolic name for this version.
391
(Typically the revision-id of the revision that added it.)
394
List or set of direct parent version numbers.
397
Sequence of lines to be added in the new version.
399
:param nostore_sha: See VersionedFile.add_lines.
401
assert isinstance(version_id, basestring)
402
self._check_lines_not_unicode(lines)
403
self._check_lines_are_lines(lines)
405
sha1 = sha_strings(lines)
406
if sha1 == nostore_sha:
407
raise errors.ExistingContent
408
if version_id in self._name_map:
409
return self._check_repeated_add(version_id, parents, lines, sha1)
411
self._check_versions(parents)
412
## self._check_lines(lines)
413
new_version = len(self._parents)
415
# if we abort after here the (in-memory) weave will be corrupt because only
416
# some fields are updated
417
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
418
# - Robert Collins 20060226
419
self._parents.append(parents[:])
420
self._sha1s.append(sha1)
421
self._names.append(version_id)
422
self._name_map[version_id] = new_version
426
# special case; adding with no parents revision; can do
427
# this more quickly by just appending unconditionally.
428
# even more specially, if we're adding an empty text we
429
# need do nothing at all.
431
self._weave.append(('{', new_version))
432
self._weave.extend(lines)
433
self._weave.append(('}', None))
436
if len(parents) == 1:
437
pv = list(parents)[0]
438
if sha1 == self._sha1s[pv]:
439
# special case: same as the single parent
443
ancestors = self._inclusions(parents)
447
# basis a list of (origin, lineno, line)
450
for origin, lineno, line in self._extract(ancestors):
451
basis_lineno.append(lineno)
452
basis_lines.append(line)
454
# another small special case: a merge, producing the same text
456
if lines == basis_lines:
459
# add a sentinel, because we can also match against the final line
460
basis_lineno.append(len(self._weave))
462
# XXX: which line of the weave should we really consider
463
# matches the end of the file? the current code says it's the
464
# last line of the weave?
466
#print 'basis_lines:', basis_lines
467
#print 'new_lines: ', lines
469
s = self._matcher(None, basis_lines, lines)
471
# offset gives the number of lines that have been inserted
472
# into the weave up to the current point; if the original edit instruction
473
# says to change line A then we actually change (A+offset)
476
for tag, i1, i2, j1, j2 in s.get_opcodes():
477
# i1,i2 are given in offsets within basis_lines; we need to map them
478
# back to offsets within the entire weave
479
#print 'raw match', tag, i1, i2, j1, j2
483
i1 = basis_lineno[i1]
484
i2 = basis_lineno[i2]
486
assert 0 <= j1 <= j2 <= len(lines)
488
#print tag, i1, i2, j1, j2
490
# the deletion and insertion are handled separately.
491
# first delete the region.
493
self._weave.insert(i1+offset, ('[', new_version))
494
self._weave.insert(i2+offset+1, (']', new_version))
498
# there may have been a deletion spanning up to
499
# i2; we want to insert after this region to make sure
500
# we don't destroy ourselves
502
self._weave[i:i] = ([('{', new_version)]
505
offset += 2 + (j2 - j1)
508
def _inclusions(self, versions):
509
"""Return set of all ancestors of given version(s)."""
510
if not len(versions):
513
for v in xrange(max(versions), 0, -1):
515
# include all its parents
516
i.update(self._parents[v])
518
## except IndexError:
519
## raise ValueError("version %d not present in weave" % v)
521
def get_ancestry(self, version_ids, topo_sorted=True):
522
"""See VersionedFile.get_ancestry."""
523
if isinstance(version_ids, basestring):
524
version_ids = [version_ids]
525
i = self._inclusions([self._lookup(v) for v in version_ids])
526
return [self._idx_to_name(v) for v in i]
528
def _check_lines(self, text):
529
if not isinstance(text, list):
530
raise ValueError("text should be a list, not %s" % type(text))
533
if not isinstance(l, basestring):
534
raise ValueError("text line should be a string or unicode, not %s"
539
def _check_versions(self, indexes):
540
"""Check everything in the sequence of indexes is valid"""
545
raise IndexError("invalid version number %r" % i)
547
def _compatible_parents(self, my_parents, other_parents):
548
"""During join check that other_parents are joinable with my_parents.
550
Joinable is defined as 'is a subset of' - supersets may require
551
regeneration of diffs, but subsets do not.
553
return len(other_parents.difference(my_parents)) == 0
555
def annotate(self, version_id):
556
"""Return a list of (version-id, line) tuples for version_id.
558
The index indicates when the line originated in the weave."""
559
incls = [self._lookup(version_id)]
560
return [(self._idx_to_name(origin), text) for origin, lineno, text in
561
self._extract(incls)]
563
def iter_lines_added_or_present_in_versions(self, version_ids=None,
565
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
566
if version_ids is None:
567
version_ids = self.versions()
568
version_ids = set(version_ids)
569
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
570
# if inserted not in version_ids then it was inserted before the
571
# versions we care about, but because weaves cannot represent ghosts
572
# properly, we do not filter down to that
573
# if inserted not in version_ids: continue
575
yield line + '\n', inserted
579
def _walk_internal(self, version_ids=None):
580
"""Helper method for weave actions."""
585
lineno = 0 # line of weave, 0-based
587
for l in self._weave:
588
if l.__class__ == tuple:
592
istack.append(self._names[v])
596
assert self._names[v] not in dset
597
dset.add(self._names[v])
599
dset.remove(self._names[v])
601
raise WeaveFormatError('unexpected instruction %r' % v)
603
assert l.__class__ in (str, unicode)
605
yield lineno, istack[-1], frozenset(dset), l
609
raise WeaveFormatError("unclosed insertion blocks "
610
"at end of weave: %s" % istack)
612
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
615
def plan_merge(self, ver_a, ver_b):
616
"""Return pseudo-annotation indicating how the two versions merge.
618
This is computed between versions a and b and their common
621
Weave lines present in none of them are skipped entirely.
623
inc_a = set(self.get_ancestry([ver_a]))
624
inc_b = set(self.get_ancestry([ver_b]))
625
inc_c = inc_a & inc_b
627
for lineno, insert, deleteset, line in self._walk_internal([ver_a, ver_b]):
628
if deleteset & inc_c:
629
# killed in parent; can't be in either a or b
630
# not relevant to our work
631
yield 'killed-base', line
632
elif insert in inc_c:
633
# was inserted in base
634
killed_a = bool(deleteset & inc_a)
635
killed_b = bool(deleteset & inc_b)
636
if killed_a and killed_b:
637
yield 'killed-both', line
639
yield 'killed-a', line
641
yield 'killed-b', line
643
yield 'unchanged', line
644
elif insert in inc_a:
645
if deleteset & inc_a:
646
yield 'ghost-a', line
650
elif insert in inc_b:
651
if deleteset & inc_b:
652
yield 'ghost-b', line
656
# not in either revision
657
yield 'irrelevant', line
659
yield 'unchanged', '' # terminator
661
def _extract(self, versions):
662
"""Yield annotation of lines in included set.
664
Yields a sequence of tuples (origin, lineno, text), where
665
origin is the origin version, lineno the index in the weave,
666
and text the text of the line.
668
The set typically but not necessarily corresponds to a version.
671
if not isinstance(i, int):
674
included = self._inclusions(versions)
680
lineno = 0 # line of weave, 0-based
686
WFE = WeaveFormatError
689
# 449 0 4474.6820 2356.5590 bzrlib.weave:556(_extract)
690
# +285282 0 1676.8040 1676.8040 +<isinstance>
691
# 1.6 seconds in 'isinstance'.
692
# changing the first isinstance:
693
# 449 0 2814.2660 1577.1760 bzrlib.weave:556(_extract)
694
# +140414 0 762.8050 762.8050 +<isinstance>
695
# note that the inline time actually dropped (less function calls)
696
# and total processing time was halved.
697
# we're still spending ~1/4 of the method in isinstance though.
698
# so lets hard code the acceptable string classes we expect:
699
# 449 0 1202.9420 786.2930 bzrlib.weave:556(_extract)
700
# +71352 0 377.5560 377.5560 +<method 'append' of 'list'
702
# yay, down to ~1/4 the initial extract time, and our inline time
703
# has shrunk again, with isinstance no longer dominating.
704
# tweaking the stack inclusion test to use a set gives:
705
# 449 0 1122.8030 713.0080 bzrlib.weave:556(_extract)
706
# +71352 0 354.9980 354.9980 +<method 'append' of 'list'
708
# - a 5% win, or possibly just noise. However with large istacks that
709
# 'in' test could dominate, so I'm leaving this change in place -
710
# when its fast enough to consider profiling big datasets we can review.
715
for l in self._weave:
716
if l.__class__ == tuple:
724
iset.remove(istack.pop())
735
assert l.__class__ in (str, unicode)
737
isactive = (not dset) and istack and (istack[-1] in included)
739
result.append((istack[-1], lineno, l))
742
raise WeaveFormatError("unclosed insertion blocks "
743
"at end of weave: %s" % istack)
745
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
749
def _maybe_lookup(self, name_or_index):
750
"""Convert possible symbolic name to index, or pass through indexes.
754
if isinstance(name_or_index, (int, long)):
757
return self._lookup(name_or_index)
759
def get_lines(self, version_id):
760
"""See VersionedFile.get_lines()."""
761
int_index = self._maybe_lookup(version_id)
762
result = [line for (origin, lineno, line) in self._extract([int_index])]
763
expected_sha1 = self._sha1s[int_index]
764
measured_sha1 = sha_strings(result)
765
if measured_sha1 != expected_sha1:
766
raise errors.WeaveInvalidChecksum(
767
'file %s, revision %s, expected: %s, measured %s'
768
% (self._weave_name, version_id,
769
expected_sha1, measured_sha1))
772
def get_sha1s(self, version_ids):
773
"""See VersionedFile.get_sha1s()."""
774
return [self._sha1s[self._lookup(v)] for v in version_ids]
776
def num_versions(self):
777
"""How many versions are in this weave?"""
778
l = len(self._parents)
779
assert l == len(self._sha1s)
782
__len__ = num_versions
784
def check(self, progress_bar=None):
785
# TODO evaluate performance hit of using string sets in this routine.
786
# TODO: check no circular inclusions
787
# TODO: create a nested progress bar
788
for version in range(self.num_versions()):
789
inclusions = list(self._parents[version])
792
if inclusions[-1] >= version:
793
raise WeaveFormatError("invalid included version %d for index %d"
794
% (inclusions[-1], version))
796
# try extracting all versions; parallel extraction is used
797
nv = self.num_versions()
802
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
803
# The problem is that set membership is much more expensive
804
name = self._idx_to_name(i)
805
sha1s[name] = sha.new()
807
new_inc = set([name])
808
for p in self._parents[i]:
809
new_inc.update(inclusions[self._idx_to_name(p)])
811
assert set(new_inc) == set(self.get_ancestry(name)), \
812
'failed %s != %s' % (set(new_inc), set(self.get_ancestry(name)))
813
inclusions[name] = new_inc
815
nlines = len(self._weave)
817
update_text = 'checking weave'
819
short_name = os.path.basename(self._weave_name)
820
update_text = 'checking %s' % (short_name,)
821
update_text = update_text[:25]
823
for lineno, insert, deleteset, line in self._walk_internal():
825
progress_bar.update(update_text, lineno, nlines)
827
for name, name_inclusions in inclusions.items():
828
# The active inclusion must be an ancestor,
829
# and no ancestors must have deleted this line,
830
# because we don't support resurrection.
831
if (insert in name_inclusions) and not (deleteset & name_inclusions):
832
sha1s[name].update(line)
835
version = self._idx_to_name(i)
836
hd = sha1s[version].hexdigest()
837
expected = self._sha1s[i]
839
raise errors.WeaveInvalidChecksum(
840
"mismatched sha1 for version %s: "
841
"got %s, expected %s"
842
% (version, hd, expected))
844
# TODO: check insertions are properly nested, that there are
845
# no lines outside of insertion blocks, that deletions are
846
# properly paired, etc.
848
def _join(self, other, pb, msg, version_ids, ignore_missing):
849
"""Worker routine for join()."""
850
if not other.versions():
851
return # nothing to update, easy
854
# versions is never none, InterWeave checks this.
857
# two loops so that we do not change ourselves before verifying it
859
# work through in index order to make sure we get all dependencies
862
# get the selected versions only that are in other.versions.
863
version_ids = set(other.versions()).intersection(set(version_ids))
864
# pull in the referenced graph.
865
version_ids = other.get_ancestry(version_ids)
866
pending_parents = other.get_parent_map(version_ids)
867
pending_graph = pending_parents.items()
868
if len(pending_graph) != len(version_ids):
869
raise RevisionNotPresent(
870
set(version_ids) - set(pending_parents.keys()), self)
871
for name in topo_sort(pending_graph):
872
other_idx = other._name_map[name]
873
# returns True if we have it, False if we need it.
874
if not self._check_version_consistent(other, other_idx, name):
875
names_to_join.append((other_idx, name))
883
for other_idx, name in names_to_join:
884
# TODO: If all the parents of the other version are already
885
# present then we can avoid some work by just taking the delta
886
# and adjusting the offsets.
887
new_parents = self._imported_parents(other, other_idx)
888
sha1 = other._sha1s[other_idx]
893
pb.update(msg, merged, len(names_to_join))
895
lines = other.get_lines(other_idx)
896
self._add(name, lines, new_parents, sha1)
898
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
899
merged, processed, self._weave_name, time.time()-time0))
901
def _imported_parents(self, other, other_idx):
902
"""Return list of parents in self corresponding to indexes in other."""
904
for parent_idx in other._parents[other_idx]:
905
parent_name = other._names[parent_idx]
906
if parent_name not in self._name_map:
907
# should not be possible
908
raise WeaveError("missing parent {%s} of {%s} in %r"
909
% (parent_name, other._name_map[other_idx], self))
910
new_parents.append(self._name_map[parent_name])
913
def _check_version_consistent(self, other, other_idx, name):
914
"""Check if a version in consistent in this and other.
916
To be consistent it must have:
919
* the same direct parents (by name, not index, and disregarding
922
If present & correct return True;
923
if not present in self return False;
924
if inconsistent raise error."""
925
this_idx = self._name_map.get(name, -1)
927
if self._sha1s[this_idx] != other._sha1s[other_idx]:
928
raise errors.WeaveTextDiffers(name, self, other)
929
self_parents = self._parents[this_idx]
930
other_parents = other._parents[other_idx]
931
n1 = set([self._names[i] for i in self_parents])
932
n2 = set([other._names[i] for i in other_parents])
933
if not self._compatible_parents(n1, n2):
934
raise WeaveParentMismatch("inconsistent parents "
935
"for version {%s}: %s vs %s" % (name, n1, n2))
941
def _reweave(self, other, pb, msg):
942
"""Reweave self with other - internal helper for join().
944
:param other: The other weave to merge
945
:param pb: An optional progress bar, indicating how far done we are
946
:param msg: An optional message for the progress
948
new_weave = _reweave(self, other, pb=pb, msg=msg)
949
self._copy_weave_content(new_weave)
951
def _copy_weave_content(self, otherweave):
952
"""adsorb the content from otherweave."""
953
for attr in self.__slots__:
954
if attr != '_weave_name':
955
setattr(self, attr, copy(getattr(otherweave, attr)))
958
class WeaveFile(Weave):
959
"""A WeaveFile represents a Weave on disk and writes on change."""
961
WEAVE_SUFFIX = '.weave'
963
def __init__(self, name, transport, filemode=None, create=False, access_mode='w', get_scope=None):
964
"""Create a WeaveFile.
966
:param create: If not True, only open an existing knit.
968
super(WeaveFile, self).__init__(name, access_mode, get_scope=get_scope)
969
self._transport = transport
970
self._filemode = filemode
972
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
973
except errors.NoSuchFile:
979
def _add_lines(self, version_id, parents, lines, parent_texts,
980
left_matching_blocks, nostore_sha, random_id, check_content):
981
"""Add a version and save the weave."""
982
self.check_not_reserved_id(version_id)
983
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
984
parent_texts, left_matching_blocks, nostore_sha, random_id,
989
def copy_to(self, name, transport):
990
"""See VersionedFile.copy_to()."""
991
# as we are all in memory always, just serialise to the new place.
993
write_weave_v5(self, sio)
995
transport.put_file(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
998
"""Save the weave."""
999
self._check_write_ok()
1001
write_weave_v5(self, sio)
1003
self._transport.put_file(self._weave_name + WeaveFile.WEAVE_SUFFIX,
1009
"""See VersionedFile.get_suffixes()."""
1010
return [WeaveFile.WEAVE_SUFFIX]
1012
def join(self, other, pb=None, msg=None, version_ids=None,
1013
ignore_missing=False):
1014
"""Join other into self and save."""
1015
super(WeaveFile, self).join(other, pb, msg, version_ids, ignore_missing)
1019
def _reweave(wa, wb, pb=None, msg=None):
1020
"""Combine two weaves and return the result.
1022
This works even if a revision R has different parents in
1023
wa and wb. In the resulting weave all the parents are given.
1025
This is done by just building up a new weave, maintaining ordering
1026
of the versions in the two inputs. More efficient approaches
1027
might be possible but it should only be necessary to do
1028
this operation rarely, when a new previously ghost version is
1031
:param pb: An optional progress bar, indicating how far done we are
1032
:param msg: An optional message for the progress
1036
queue_a = range(wa.num_versions())
1037
queue_b = range(wb.num_versions())
1038
# first determine combined parents of all versions
1039
# map from version name -> all parent names
1040
combined_parents = _reweave_parent_graphs(wa, wb)
1041
mutter("combined parents: %r", combined_parents)
1042
order = topo_sort(combined_parents.iteritems())
1043
mutter("order to reweave: %r", order)
1048
for idx, name in enumerate(order):
1050
pb.update(msg, idx, len(order))
1051
if name in wa._name_map:
1052
lines = wa.get_lines(name)
1053
if name in wb._name_map:
1054
lines_b = wb.get_lines(name)
1055
if lines != lines_b:
1056
mutter('Weaves differ on content. rev_id {%s}', name)
1057
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
1059
lines = list(difflib.unified_diff(lines, lines_b,
1060
wa._weave_name, wb._weave_name))
1061
mutter('lines:\n%s', ''.join(lines))
1062
raise errors.WeaveTextDiffers(name, wa, wb)
1064
lines = wb.get_lines(name)
1065
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
1068
def _reweave_parent_graphs(wa, wb):
1069
"""Return combined parent ancestry for two weaves.
1071
Returned as a list of (version_name, set(parent_names))"""
1073
for weave in [wa, wb]:
1074
for idx, name in enumerate(weave._names):
1075
p = combined.setdefault(name, set())
1076
p.update(map(weave._idx_to_name, weave._parents[idx]))
1081
"""Show the weave's table-of-contents"""
1082
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
1083
for i in (6, 50, 10, 10):
1086
for i in range(w.num_versions()):
1089
parent_str = ' '.join(map(str, w._parents[i]))
1090
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
1094
def weave_stats(weave_file, pb):
1095
from bzrlib.weavefile import read_weave
1097
wf = file(weave_file, 'rb')
1099
# FIXME: doesn't work on pipes
1100
weave_size = wf.tell()
1104
for i in range(vers):
1105
pb.update('checking sizes', i, vers)
1106
for origin, lineno, line in w._extract([i]):
1111
print 'versions %9d' % vers
1112
print 'weave file %9d bytes' % weave_size
1113
print 'total contents %9d bytes' % total
1114
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1117
print 'average size %9d bytes' % avg
1118
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1122
print """bzr weave tool
1124
Experimental tool for weave algorithm.
1127
weave init WEAVEFILE
1128
Create an empty weave file
1129
weave get WEAVEFILE VERSION
1130
Write out specified version.
1131
weave check WEAVEFILE
1132
Check consistency of all versions.
1134
Display table of contents.
1135
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1136
Add NEWTEXT, with specified parent versions.
1137
weave annotate WEAVEFILE VERSION
1138
Display origin of each line.
1139
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1140
Auto-merge two versions and display conflicts.
1141
weave diff WEAVEFILE VERSION1 VERSION2
1142
Show differences between two versions.
1146
% weave init foo.weave
1148
% weave add foo.weave ver0 < foo.txt
1151
(create updated version)
1153
% weave get foo.weave 0 | diff -u - foo.txt
1154
% weave add foo.weave ver1 0 < foo.txt
1157
% weave get foo.weave 0 > foo.txt (create forked version)
1159
% weave add foo.weave ver2 0 < foo.txt
1162
% weave merge foo.weave 1 2 > foo.txt (merge them)
1163
% vi foo.txt (resolve conflicts)
1164
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1176
# in case we're run directly from the subdirectory
1177
sys.path.append('..')
1179
from bzrlib.weavefile import write_weave, read_weave
1180
from bzrlib.progress import ProgressBar
1195
return read_weave(file(argv[2], 'rb'))
1201
# at the moment, based on everything in the file
1203
parents = map(int, argv[4:])
1204
lines = sys.stdin.readlines()
1205
ver = w.add(name, parents, lines)
1206
write_weave(w, file(argv[2], 'wb'))
1207
print 'added version %r %d' % (name, ver)
1210
if os.path.exists(fn):
1211
raise IOError("file exists")
1213
write_weave(w, file(fn, 'wb'))
1214
elif cmd == 'get': # get one version
1216
sys.stdout.writelines(w.get_iter(int(argv[3])))
1221
v1, v2 = map(int, argv[3:5])
1224
diff_gen = bzrlib.patiencediff.unified_diff(lines1, lines2,
1225
'%s version %d' % (fn, v1),
1226
'%s version %d' % (fn, v2))
1227
sys.stdout.writelines(diff_gen)
1229
elif cmd == 'annotate':
1231
# newline is added to all lines regardless; too hard to get
1232
# reasonable formatting otherwise
1234
for origin, text in w.annotate(int(argv[3])):
1235
text = text.rstrip('\r\n')
1237
print ' | %s' % (text)
1239
print '%5d | %s' % (origin, text)
1245
elif cmd == 'stats':
1246
weave_stats(argv[2], ProgressBar())
1248
elif cmd == 'check':
1253
print '%d versions ok' % w.num_versions()
1255
elif cmd == 'inclusions':
1257
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1259
elif cmd == 'parents':
1261
print ' '.join(map(str, w._parents[int(argv[3])]))
1263
elif cmd == 'plan-merge':
1264
# replaced by 'bzr weave-plan-merge'
1266
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1268
print '%14s | %s' % (state, line),
1269
elif cmd == 'merge':
1270
# replaced by 'bzr weave-merge-text'
1272
p = w.plan_merge(int(argv[3]), int(argv[4]))
1273
sys.stdout.writelines(w.weave_merge(p))
1275
raise ValueError('unknown command %r' % cmd)
1278
if __name__ == '__main__':
1280
sys.exit(main(sys.argv))
1283
class InterWeave(InterVersionedFile):
1284
"""Optimised code paths for weave to weave operations."""
1286
_matching_file_from_factory = staticmethod(WeaveFile)
1287
_matching_file_to_factory = staticmethod(WeaveFile)
1290
def is_compatible(source, target):
1291
"""Be compatible with weaves."""
1293
return (isinstance(source, Weave) and
1294
isinstance(target, Weave))
1295
except AttributeError:
1298
def join(self, pb=None, msg=None, version_ids=None, ignore_missing=False):
1299
"""See InterVersionedFile.join."""
1300
version_ids = self._get_source_version_ids(version_ids, ignore_missing)
1301
if self.target.versions() == [] and version_ids is None:
1302
self.target._copy_weave_content(self.source)
1304
self.target._join(self.source, pb, msg, version_ids, ignore_missing)
1307
InterVersionedFile.register_optimiser(InterWeave)
added
removed
bzrlib/weave.py
