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.trace import mutter
82
from bzrlib.errors import (WeaveError, WeaveFormatError, WeaveParentMismatch,
83
RevisionAlreadyPresent,
85
UnavailableRepresentation,
86
WeaveRevisionAlreadyPresent,
87
WeaveRevisionNotPresent,
89
import bzrlib.errors as errors
90
from bzrlib.osutils import sha_strings
91
import bzrlib.patiencediff
92
from bzrlib.tsort import topo_sort
93
from bzrlib.versionedfile import (
98
from bzrlib.weavefile import _read_weave_v5, write_weave_v5
101
class WeaveContentFactory(ContentFactory):
102
"""Content factory for streaming from weaves.
104
:seealso ContentFactory:
107
def __init__(self, version, weave):
108
"""Create a WeaveContentFactory for version from weave."""
109
ContentFactory.__init__(self)
110
self.sha1 = weave.get_sha1s([version])[0]
111
self.key = (version,)
112
parents = weave.get_parent_map([version])[version]
113
self.parents = tuple((parent,) for parent in parents)
114
self.storage_kind = 'fulltext'
117
def get_bytes_as(self, storage_kind):
118
if storage_kind == 'fulltext':
119
return self._weave.get_text(self.key[0])
121
raise UnavailableRepresentation(self.key, storage_kind, 'fulltext')
124
class Weave(VersionedFile):
125
"""weave - versioned text file storage.
127
A Weave manages versions of line-based text files, keeping track
128
of the originating version for each line.
130
To clients the "lines" of the file are represented as a list of strings.
131
These strings will typically have terminal newline characters, but
132
this is not required. In particular files commonly do not have a newline
133
at the end of the file.
135
Texts can be identified in either of two ways:
137
* a nonnegative index number.
139
* a version-id string.
141
Typically the index number will be valid only inside this weave and
142
the version-id is used to reference it in the larger world.
144
The weave is represented as a list mixing edit instructions and
145
literal text. Each entry in _weave can be either a string (or
146
unicode), or a tuple. If a string, it means that the given line
147
should be output in the currently active revisions.
149
If a tuple, it gives a processing instruction saying in which
150
revisions the enclosed lines are active. The tuple has the form
151
(instruction, version).
153
The instruction can be '{' or '}' for an insertion block, and '['
154
and ']' for a deletion block respectively. The version is the
155
integer version index. There is no replace operator, only deletes
156
and inserts. For '}', the end of an insertion, there is no
157
version parameter because it always closes the most recently
162
* A later version can delete lines that were introduced by any
163
number of ancestor versions; this implies that deletion
164
instructions can span insertion blocks without regard to the
165
insertion block's nesting.
167
* Similarly, deletions need not be properly nested with regard to
168
each other, because they might have been generated by
169
independent revisions.
171
* Insertions are always made by inserting a new bracketed block
172
into a single point in the previous weave. This implies they
173
can nest but not overlap, and the nesting must always have later
174
insertions on the inside.
176
* It doesn't seem very useful to have an active insertion
177
inside an inactive insertion, but it might happen.
179
* Therefore, all instructions are always"considered"; that
180
is passed onto and off the stack. An outer inactive block
181
doesn't disable an inner block.
183
* Lines are enabled if the most recent enclosing insertion is
184
active and none of the enclosing deletions are active.
186
* There is no point having a deletion directly inside its own
187
insertion; you might as well just not write it. And there
188
should be no way to get an earlier version deleting a later
192
Text of the weave; list of control instruction tuples and strings.
195
List of parents, indexed by version number.
196
It is only necessary to store the minimal set of parents for
197
each version; the parent's parents are implied.
200
List of hex SHA-1 of each version.
203
List of symbolic names for each version. Each should be unique.
206
For each name, the version number.
209
Descriptive name of this weave; typically the filename if known.
213
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
214
'_weave_name', '_matcher']
216
def __init__(self, weave_name=None, access_mode='w', matcher=None, get_scope=None):
219
:param get_scope: A callable that returns an opaque object to be used
220
for detecting when this weave goes out of scope (should stop
221
answering requests or allowing mutation).
223
super(Weave, self).__init__(access_mode)
229
self._weave_name = weave_name
231
self._matcher = bzrlib.patiencediff.PatienceSequenceMatcher
233
self._matcher = matcher
234
if get_scope is None:
235
get_scope = lambda:None
236
self._get_scope = get_scope
237
self._scope = get_scope()
238
self._access_mode = access_mode
241
return "Weave(%r)" % self._weave_name
243
def _check_write_ok(self):
244
"""Is the versioned file marked as 'finished' ? Raise if it is."""
245
if self._get_scope() != self._scope:
246
raise errors.OutSideTransaction()
247
if self._access_mode != 'w':
248
raise errors.ReadOnlyObjectDirtiedError(self)
251
"""Return a deep copy of self.
253
The copy can be modified without affecting the original weave."""
255
other._weave = self._weave[:]
256
other._parents = self._parents[:]
257
other._sha1s = self._sha1s[:]
258
other._names = self._names[:]
259
other._name_map = self._name_map.copy()
260
other._weave_name = self._weave_name
263
def __eq__(self, other):
264
if not isinstance(other, Weave):
266
return self._parents == other._parents \
267
and self._weave == other._weave \
268
and self._sha1s == other._sha1s
270
def __ne__(self, other):
271
return not self.__eq__(other)
273
def _idx_to_name(self, version):
274
return self._names[version]
276
def _lookup(self, name):
277
"""Convert symbolic version name to index."""
278
self.check_not_reserved_id(name)
280
return self._name_map[name]
282
raise RevisionNotPresent(name, self._weave_name)
285
"""See VersionedFile.versions."""
286
return self._names[:]
288
def has_version(self, version_id):
289
"""See VersionedFile.has_version."""
290
return (version_id in self._name_map)
292
__contains__ = has_version
294
def get_record_stream(self, versions, ordering, include_delta_closure):
295
"""Get a stream of records for versions.
297
:param versions: The versions to include. Each version is a tuple
299
:param ordering: Either 'unordered' or 'topological'. A topologically
300
sorted stream has compression parents strictly before their
302
:param include_delta_closure: If True then the closure across any
303
compression parents will be included (in the opaque data).
304
:return: An iterator of ContentFactory objects, each of which is only
305
valid until the iterator is advanced.
307
if ordering == 'topological':
308
parents = self.get_parent_map(versions)
309
versions = topo_sort(parents)
310
for version in versions:
311
yield WeaveContentFactory(version, self)
313
def get_parent_map(self, version_ids):
314
"""See VersionedFile.get_parent_map."""
316
for version_id in version_ids:
318
result[version_id] = tuple(
319
map(self._idx_to_name, self._parents[self._lookup(version_id)]))
320
except RevisionNotPresent:
324
def get_parents_with_ghosts(self, version_id):
325
raise NotImplementedError(self.get_parents_with_ghosts)
327
def _check_repeated_add(self, name, parents, text, sha1):
328
"""Check that a duplicated add is OK.
330
If it is, return the (old) index; otherwise raise an exception.
332
idx = self._lookup(name)
333
if sorted(self._parents[idx]) != sorted(parents) \
334
or sha1 != self._sha1s[idx]:
335
raise RevisionAlreadyPresent(name, self._weave_name)
338
def _add_lines(self, version_id, parents, lines, parent_texts,
339
left_matching_blocks, nostore_sha, random_id, check_content):
340
"""See VersionedFile.add_lines."""
341
idx = self._add(version_id, lines, map(self._lookup, parents),
342
nostore_sha=nostore_sha)
343
return sha_strings(lines), sum(map(len, lines)), idx
345
def _add(self, version_id, lines, parents, sha1=None, nostore_sha=None):
346
"""Add a single text on top of the weave.
348
Returns the index number of the newly added version.
351
Symbolic name for this version.
352
(Typically the revision-id of the revision that added it.)
355
List or set of direct parent version numbers.
358
Sequence of lines to be added in the new version.
360
:param nostore_sha: See VersionedFile.add_lines.
362
assert isinstance(version_id, basestring)
363
self._check_lines_not_unicode(lines)
364
self._check_lines_are_lines(lines)
366
sha1 = sha_strings(lines)
367
if sha1 == nostore_sha:
368
raise errors.ExistingContent
369
if version_id in self._name_map:
370
return self._check_repeated_add(version_id, parents, lines, sha1)
372
self._check_versions(parents)
373
## self._check_lines(lines)
374
new_version = len(self._parents)
376
# if we abort after here the (in-memory) weave will be corrupt because only
377
# some fields are updated
378
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
379
# - Robert Collins 20060226
380
self._parents.append(parents[:])
381
self._sha1s.append(sha1)
382
self._names.append(version_id)
383
self._name_map[version_id] = new_version
387
# special case; adding with no parents revision; can do
388
# this more quickly by just appending unconditionally.
389
# even more specially, if we're adding an empty text we
390
# need do nothing at all.
392
self._weave.append(('{', new_version))
393
self._weave.extend(lines)
394
self._weave.append(('}', None))
397
if len(parents) == 1:
398
pv = list(parents)[0]
399
if sha1 == self._sha1s[pv]:
400
# special case: same as the single parent
404
ancestors = self._inclusions(parents)
408
# basis a list of (origin, lineno, line)
411
for origin, lineno, line in self._extract(ancestors):
412
basis_lineno.append(lineno)
413
basis_lines.append(line)
415
# another small special case: a merge, producing the same text
417
if lines == basis_lines:
420
# add a sentinel, because we can also match against the final line
421
basis_lineno.append(len(self._weave))
423
# XXX: which line of the weave should we really consider
424
# matches the end of the file? the current code says it's the
425
# last line of the weave?
427
#print 'basis_lines:', basis_lines
428
#print 'new_lines: ', lines
430
s = self._matcher(None, basis_lines, lines)
432
# offset gives the number of lines that have been inserted
433
# into the weave up to the current point; if the original edit instruction
434
# says to change line A then we actually change (A+offset)
437
for tag, i1, i2, j1, j2 in s.get_opcodes():
438
# i1,i2 are given in offsets within basis_lines; we need to map them
439
# back to offsets within the entire weave
440
#print 'raw match', tag, i1, i2, j1, j2
444
i1 = basis_lineno[i1]
445
i2 = basis_lineno[i2]
447
assert 0 <= j1 <= j2 <= len(lines)
449
#print tag, i1, i2, j1, j2
451
# the deletion and insertion are handled separately.
452
# first delete the region.
454
self._weave.insert(i1+offset, ('[', new_version))
455
self._weave.insert(i2+offset+1, (']', new_version))
459
# there may have been a deletion spanning up to
460
# i2; we want to insert after this region to make sure
461
# we don't destroy ourselves
463
self._weave[i:i] = ([('{', new_version)]
466
offset += 2 + (j2 - j1)
469
def _inclusions(self, versions):
470
"""Return set of all ancestors of given version(s)."""
471
if not len(versions):
474
for v in xrange(max(versions), 0, -1):
476
# include all its parents
477
i.update(self._parents[v])
479
## except IndexError:
480
## raise ValueError("version %d not present in weave" % v)
482
def get_ancestry(self, version_ids, topo_sorted=True):
483
"""See VersionedFile.get_ancestry."""
484
if isinstance(version_ids, basestring):
485
version_ids = [version_ids]
486
i = self._inclusions([self._lookup(v) for v in version_ids])
487
return [self._idx_to_name(v) for v in i]
489
def _check_lines(self, text):
490
if not isinstance(text, list):
491
raise ValueError("text should be a list, not %s" % type(text))
494
if not isinstance(l, basestring):
495
raise ValueError("text line should be a string or unicode, not %s"
500
def _check_versions(self, indexes):
501
"""Check everything in the sequence of indexes is valid"""
506
raise IndexError("invalid version number %r" % i)
508
def _compatible_parents(self, my_parents, other_parents):
509
"""During join check that other_parents are joinable with my_parents.
511
Joinable is defined as 'is a subset of' - supersets may require
512
regeneration of diffs, but subsets do not.
514
return len(other_parents.difference(my_parents)) == 0
516
def annotate(self, version_id):
517
"""Return a list of (version-id, line) tuples for version_id.
519
The index indicates when the line originated in the weave."""
520
incls = [self._lookup(version_id)]
521
return [(self._idx_to_name(origin), text) for origin, lineno, text in
522
self._extract(incls)]
524
def iter_lines_added_or_present_in_versions(self, version_ids=None,
526
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
527
if version_ids is None:
528
version_ids = self.versions()
529
version_ids = set(version_ids)
530
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
531
# if inserted not in version_ids then it was inserted before the
532
# versions we care about, but because weaves cannot represent ghosts
533
# properly, we do not filter down to that
534
# if inserted not in version_ids: continue
536
yield line + '\n', inserted
540
def _walk_internal(self, version_ids=None):
541
"""Helper method for weave actions."""
546
lineno = 0 # line of weave, 0-based
548
for l in self._weave:
549
if l.__class__ == tuple:
553
istack.append(self._names[v])
557
assert self._names[v] not in dset
558
dset.add(self._names[v])
560
dset.remove(self._names[v])
562
raise WeaveFormatError('unexpected instruction %r' % v)
564
assert l.__class__ in (str, unicode)
566
yield lineno, istack[-1], frozenset(dset), l
570
raise WeaveFormatError("unclosed insertion blocks "
571
"at end of weave: %s" % istack)
573
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
576
def plan_merge(self, ver_a, ver_b):
577
"""Return pseudo-annotation indicating how the two versions merge.
579
This is computed between versions a and b and their common
582
Weave lines present in none of them are skipped entirely.
584
inc_a = set(self.get_ancestry([ver_a]))
585
inc_b = set(self.get_ancestry([ver_b]))
586
inc_c = inc_a & inc_b
588
for lineno, insert, deleteset, line in self._walk_internal([ver_a, ver_b]):
589
if deleteset & inc_c:
590
# killed in parent; can't be in either a or b
591
# not relevant to our work
592
yield 'killed-base', line
593
elif insert in inc_c:
594
# was inserted in base
595
killed_a = bool(deleteset & inc_a)
596
killed_b = bool(deleteset & inc_b)
597
if killed_a and killed_b:
598
yield 'killed-both', line
600
yield 'killed-a', line
602
yield 'killed-b', line
604
yield 'unchanged', line
605
elif insert in inc_a:
606
if deleteset & inc_a:
607
yield 'ghost-a', line
611
elif insert in inc_b:
612
if deleteset & inc_b:
613
yield 'ghost-b', line
617
# not in either revision
618
yield 'irrelevant', line
620
yield 'unchanged', '' # terminator
622
def _extract(self, versions):
623
"""Yield annotation of lines in included set.
625
Yields a sequence of tuples (origin, lineno, text), where
626
origin is the origin version, lineno the index in the weave,
627
and text the text of the line.
629
The set typically but not necessarily corresponds to a version.
632
if not isinstance(i, int):
635
included = self._inclusions(versions)
641
lineno = 0 # line of weave, 0-based
647
WFE = WeaveFormatError
650
# 449 0 4474.6820 2356.5590 bzrlib.weave:556(_extract)
651
# +285282 0 1676.8040 1676.8040 +<isinstance>
652
# 1.6 seconds in 'isinstance'.
653
# changing the first isinstance:
654
# 449 0 2814.2660 1577.1760 bzrlib.weave:556(_extract)
655
# +140414 0 762.8050 762.8050 +<isinstance>
656
# note that the inline time actually dropped (less function calls)
657
# and total processing time was halved.
658
# we're still spending ~1/4 of the method in isinstance though.
659
# so lets hard code the acceptable string classes we expect:
660
# 449 0 1202.9420 786.2930 bzrlib.weave:556(_extract)
661
# +71352 0 377.5560 377.5560 +<method 'append' of 'list'
663
# yay, down to ~1/4 the initial extract time, and our inline time
664
# has shrunk again, with isinstance no longer dominating.
665
# tweaking the stack inclusion test to use a set gives:
666
# 449 0 1122.8030 713.0080 bzrlib.weave:556(_extract)
667
# +71352 0 354.9980 354.9980 +<method 'append' of 'list'
669
# - a 5% win, or possibly just noise. However with large istacks that
670
# 'in' test could dominate, so I'm leaving this change in place -
671
# when its fast enough to consider profiling big datasets we can review.
676
for l in self._weave:
677
if l.__class__ == tuple:
685
iset.remove(istack.pop())
696
assert l.__class__ in (str, unicode)
698
isactive = (not dset) and istack and (istack[-1] in included)
700
result.append((istack[-1], lineno, l))
703
raise WeaveFormatError("unclosed insertion blocks "
704
"at end of weave: %s" % istack)
706
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
710
def _maybe_lookup(self, name_or_index):
711
"""Convert possible symbolic name to index, or pass through indexes.
715
if isinstance(name_or_index, (int, long)):
718
return self._lookup(name_or_index)
720
def get_lines(self, version_id):
721
"""See VersionedFile.get_lines()."""
722
int_index = self._maybe_lookup(version_id)
723
result = [line for (origin, lineno, line) in self._extract([int_index])]
724
expected_sha1 = self._sha1s[int_index]
725
measured_sha1 = sha_strings(result)
726
if measured_sha1 != expected_sha1:
727
raise errors.WeaveInvalidChecksum(
728
'file %s, revision %s, expected: %s, measured %s'
729
% (self._weave_name, version_id,
730
expected_sha1, measured_sha1))
733
def get_sha1s(self, version_ids):
734
"""See VersionedFile.get_sha1s()."""
735
return [self._sha1s[self._lookup(v)] for v in version_ids]
737
def num_versions(self):
738
"""How many versions are in this weave?"""
739
l = len(self._parents)
740
assert l == len(self._sha1s)
743
__len__ = num_versions
745
def check(self, progress_bar=None):
746
# TODO evaluate performance hit of using string sets in this routine.
747
# TODO: check no circular inclusions
748
# TODO: create a nested progress bar
749
for version in range(self.num_versions()):
750
inclusions = list(self._parents[version])
753
if inclusions[-1] >= version:
754
raise WeaveFormatError("invalid included version %d for index %d"
755
% (inclusions[-1], version))
757
# try extracting all versions; parallel extraction is used
758
nv = self.num_versions()
763
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
764
# The problem is that set membership is much more expensive
765
name = self._idx_to_name(i)
766
sha1s[name] = sha.new()
768
new_inc = set([name])
769
for p in self._parents[i]:
770
new_inc.update(inclusions[self._idx_to_name(p)])
772
assert set(new_inc) == set(self.get_ancestry(name)), \
773
'failed %s != %s' % (set(new_inc), set(self.get_ancestry(name)))
774
inclusions[name] = new_inc
776
nlines = len(self._weave)
778
update_text = 'checking weave'
780
short_name = os.path.basename(self._weave_name)
781
update_text = 'checking %s' % (short_name,)
782
update_text = update_text[:25]
784
for lineno, insert, deleteset, line in self._walk_internal():
786
progress_bar.update(update_text, lineno, nlines)
788
for name, name_inclusions in inclusions.items():
789
# The active inclusion must be an ancestor,
790
# and no ancestors must have deleted this line,
791
# because we don't support resurrection.
792
if (insert in name_inclusions) and not (deleteset & name_inclusions):
793
sha1s[name].update(line)
796
version = self._idx_to_name(i)
797
hd = sha1s[version].hexdigest()
798
expected = self._sha1s[i]
800
raise errors.WeaveInvalidChecksum(
801
"mismatched sha1 for version %s: "
802
"got %s, expected %s"
803
% (version, hd, expected))
805
# TODO: check insertions are properly nested, that there are
806
# no lines outside of insertion blocks, that deletions are
807
# properly paired, etc.
809
def _join(self, other, pb, msg, version_ids, ignore_missing):
810
"""Worker routine for join()."""
811
if not other.versions():
812
return # nothing to update, easy
815
# versions is never none, InterWeave checks this.
818
# two loops so that we do not change ourselves before verifying it
820
# work through in index order to make sure we get all dependencies
823
# get the selected versions only that are in other.versions.
824
version_ids = set(other.versions()).intersection(set(version_ids))
825
# pull in the referenced graph.
826
version_ids = other.get_ancestry(version_ids)
827
pending_parents = other.get_parent_map(version_ids)
828
pending_graph = pending_parents.items()
829
if len(pending_graph) != len(version_ids):
830
raise RevisionNotPresent(
831
set(version_ids) - set(pending_parents.keys()), self)
832
for name in topo_sort(pending_graph):
833
other_idx = other._name_map[name]
834
# returns True if we have it, False if we need it.
835
if not self._check_version_consistent(other, other_idx, name):
836
names_to_join.append((other_idx, name))
844
for other_idx, name in names_to_join:
845
# TODO: If all the parents of the other version are already
846
# present then we can avoid some work by just taking the delta
847
# and adjusting the offsets.
848
new_parents = self._imported_parents(other, other_idx)
849
sha1 = other._sha1s[other_idx]
854
pb.update(msg, merged, len(names_to_join))
856
lines = other.get_lines(other_idx)
857
self._add(name, lines, new_parents, sha1)
859
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
860
merged, processed, self._weave_name, time.time()-time0))
862
def _imported_parents(self, other, other_idx):
863
"""Return list of parents in self corresponding to indexes in other."""
865
for parent_idx in other._parents[other_idx]:
866
parent_name = other._names[parent_idx]
867
if parent_name not in self._name_map:
868
# should not be possible
869
raise WeaveError("missing parent {%s} of {%s} in %r"
870
% (parent_name, other._name_map[other_idx], self))
871
new_parents.append(self._name_map[parent_name])
874
def _check_version_consistent(self, other, other_idx, name):
875
"""Check if a version in consistent in this and other.
877
To be consistent it must have:
880
* the same direct parents (by name, not index, and disregarding
883
If present & correct return True;
884
if not present in self return False;
885
if inconsistent raise error."""
886
this_idx = self._name_map.get(name, -1)
888
if self._sha1s[this_idx] != other._sha1s[other_idx]:
889
raise errors.WeaveTextDiffers(name, self, other)
890
self_parents = self._parents[this_idx]
891
other_parents = other._parents[other_idx]
892
n1 = set([self._names[i] for i in self_parents])
893
n2 = set([other._names[i] for i in other_parents])
894
if not self._compatible_parents(n1, n2):
895
raise WeaveParentMismatch("inconsistent parents "
896
"for version {%s}: %s vs %s" % (name, n1, n2))
902
def _reweave(self, other, pb, msg):
903
"""Reweave self with other - internal helper for join().
905
:param other: The other weave to merge
906
:param pb: An optional progress bar, indicating how far done we are
907
:param msg: An optional message for the progress
909
new_weave = _reweave(self, other, pb=pb, msg=msg)
910
self._copy_weave_content(new_weave)
912
def _copy_weave_content(self, otherweave):
913
"""adsorb the content from otherweave."""
914
for attr in self.__slots__:
915
if attr != '_weave_name':
916
setattr(self, attr, copy(getattr(otherweave, attr)))
919
class WeaveFile(Weave):
920
"""A WeaveFile represents a Weave on disk and writes on change."""
922
WEAVE_SUFFIX = '.weave'
924
def __init__(self, name, transport, filemode=None, create=False, access_mode='w', get_scope=None):
925
"""Create a WeaveFile.
927
:param create: If not True, only open an existing knit.
929
super(WeaveFile, self).__init__(name, access_mode, get_scope=get_scope)
930
self._transport = transport
931
self._filemode = filemode
933
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
934
except errors.NoSuchFile:
940
def _add_lines(self, version_id, parents, lines, parent_texts,
941
left_matching_blocks, nostore_sha, random_id, check_content):
942
"""Add a version and save the weave."""
943
self.check_not_reserved_id(version_id)
944
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
945
parent_texts, left_matching_blocks, nostore_sha, random_id,
950
def copy_to(self, name, transport):
951
"""See VersionedFile.copy_to()."""
952
# as we are all in memory always, just serialise to the new place.
954
write_weave_v5(self, sio)
956
transport.put_file(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
959
"""Save the weave."""
960
self._check_write_ok()
962
write_weave_v5(self, sio)
964
self._transport.put_file(self._weave_name + WeaveFile.WEAVE_SUFFIX,
970
"""See VersionedFile.get_suffixes()."""
971
return [WeaveFile.WEAVE_SUFFIX]
973
def join(self, other, pb=None, msg=None, version_ids=None,
974
ignore_missing=False):
975
"""Join other into self and save."""
976
super(WeaveFile, self).join(other, pb, msg, version_ids, ignore_missing)
980
def _reweave(wa, wb, pb=None, msg=None):
981
"""Combine two weaves and return the result.
983
This works even if a revision R has different parents in
984
wa and wb. In the resulting weave all the parents are given.
986
This is done by just building up a new weave, maintaining ordering
987
of the versions in the two inputs. More efficient approaches
988
might be possible but it should only be necessary to do
989
this operation rarely, when a new previously ghost version is
992
:param pb: An optional progress bar, indicating how far done we are
993
:param msg: An optional message for the progress
997
queue_a = range(wa.num_versions())
998
queue_b = range(wb.num_versions())
999
# first determine combined parents of all versions
1000
# map from version name -> all parent names
1001
combined_parents = _reweave_parent_graphs(wa, wb)
1002
mutter("combined parents: %r", combined_parents)
1003
order = topo_sort(combined_parents.iteritems())
1004
mutter("order to reweave: %r", order)
1009
for idx, name in enumerate(order):
1011
pb.update(msg, idx, len(order))
1012
if name in wa._name_map:
1013
lines = wa.get_lines(name)
1014
if name in wb._name_map:
1015
lines_b = wb.get_lines(name)
1016
if lines != lines_b:
1017
mutter('Weaves differ on content. rev_id {%s}', name)
1018
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
1020
lines = list(difflib.unified_diff(lines, lines_b,
1021
wa._weave_name, wb._weave_name))
1022
mutter('lines:\n%s', ''.join(lines))
1023
raise errors.WeaveTextDiffers(name, wa, wb)
1025
lines = wb.get_lines(name)
1026
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
1029
def _reweave_parent_graphs(wa, wb):
1030
"""Return combined parent ancestry for two weaves.
1032
Returned as a list of (version_name, set(parent_names))"""
1034
for weave in [wa, wb]:
1035
for idx, name in enumerate(weave._names):
1036
p = combined.setdefault(name, set())
1037
p.update(map(weave._idx_to_name, weave._parents[idx]))
1042
"""Show the weave's table-of-contents"""
1043
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
1044
for i in (6, 50, 10, 10):
1047
for i in range(w.num_versions()):
1050
parent_str = ' '.join(map(str, w._parents[i]))
1051
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
1055
def weave_stats(weave_file, pb):
1056
from bzrlib.weavefile import read_weave
1058
wf = file(weave_file, 'rb')
1060
# FIXME: doesn't work on pipes
1061
weave_size = wf.tell()
1065
for i in range(vers):
1066
pb.update('checking sizes', i, vers)
1067
for origin, lineno, line in w._extract([i]):
1072
print 'versions %9d' % vers
1073
print 'weave file %9d bytes' % weave_size
1074
print 'total contents %9d bytes' % total
1075
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1078
print 'average size %9d bytes' % avg
1079
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1083
print """bzr weave tool
1085
Experimental tool for weave algorithm.
1088
weave init WEAVEFILE
1089
Create an empty weave file
1090
weave get WEAVEFILE VERSION
1091
Write out specified version.
1092
weave check WEAVEFILE
1093
Check consistency of all versions.
1095
Display table of contents.
1096
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1097
Add NEWTEXT, with specified parent versions.
1098
weave annotate WEAVEFILE VERSION
1099
Display origin of each line.
1100
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1101
Auto-merge two versions and display conflicts.
1102
weave diff WEAVEFILE VERSION1 VERSION2
1103
Show differences between two versions.
1107
% weave init foo.weave
1109
% weave add foo.weave ver0 < foo.txt
1112
(create updated version)
1114
% weave get foo.weave 0 | diff -u - foo.txt
1115
% weave add foo.weave ver1 0 < foo.txt
1118
% weave get foo.weave 0 > foo.txt (create forked version)
1120
% weave add foo.weave ver2 0 < foo.txt
1123
% weave merge foo.weave 1 2 > foo.txt (merge them)
1124
% vi foo.txt (resolve conflicts)
1125
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1137
# in case we're run directly from the subdirectory
1138
sys.path.append('..')
1140
from bzrlib.weavefile import write_weave, read_weave
1141
from bzrlib.progress import ProgressBar
1156
return read_weave(file(argv[2], 'rb'))
1162
# at the moment, based on everything in the file
1164
parents = map(int, argv[4:])
1165
lines = sys.stdin.readlines()
1166
ver = w.add(name, parents, lines)
1167
write_weave(w, file(argv[2], 'wb'))
1168
print 'added version %r %d' % (name, ver)
1171
if os.path.exists(fn):
1172
raise IOError("file exists")
1174
write_weave(w, file(fn, 'wb'))
1175
elif cmd == 'get': # get one version
1177
sys.stdout.writelines(w.get_iter(int(argv[3])))
1182
v1, v2 = map(int, argv[3:5])
1185
diff_gen = bzrlib.patiencediff.unified_diff(lines1, lines2,
1186
'%s version %d' % (fn, v1),
1187
'%s version %d' % (fn, v2))
1188
sys.stdout.writelines(diff_gen)
1190
elif cmd == 'annotate':
1192
# newline is added to all lines regardless; too hard to get
1193
# reasonable formatting otherwise
1195
for origin, text in w.annotate(int(argv[3])):
1196
text = text.rstrip('\r\n')
1198
print ' | %s' % (text)
1200
print '%5d | %s' % (origin, text)
1206
elif cmd == 'stats':
1207
weave_stats(argv[2], ProgressBar())
1209
elif cmd == 'check':
1214
print '%d versions ok' % w.num_versions()
1216
elif cmd == 'inclusions':
1218
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1220
elif cmd == 'parents':
1222
print ' '.join(map(str, w._parents[int(argv[3])]))
1224
elif cmd == 'plan-merge':
1225
# replaced by 'bzr weave-plan-merge'
1227
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1229
print '%14s | %s' % (state, line),
1230
elif cmd == 'merge':
1231
# replaced by 'bzr weave-merge-text'
1233
p = w.plan_merge(int(argv[3]), int(argv[4]))
1234
sys.stdout.writelines(w.weave_merge(p))
1236
raise ValueError('unknown command %r' % cmd)
1239
if __name__ == '__main__':
1241
sys.exit(main(sys.argv))
1244
class InterWeave(InterVersionedFile):
1245
"""Optimised code paths for weave to weave operations."""
1247
_matching_file_from_factory = staticmethod(WeaveFile)
1248
_matching_file_to_factory = staticmethod(WeaveFile)
1251
def is_compatible(source, target):
1252
"""Be compatible with weaves."""
1254
return (isinstance(source, Weave) and
1255
isinstance(target, Weave))
1256
except AttributeError:
1259
def join(self, pb=None, msg=None, version_ids=None, ignore_missing=False):
1260
"""See InterVersionedFile.join."""
1261
version_ids = self._get_source_version_ids(version_ids, ignore_missing)
1262
if self.target.versions() == [] and version_ids is None:
1263
self.target._copy_weave_content(self.source)
1265
self.target._join(self.source, pb, msg, version_ids, ignore_missing)
1268
InterVersionedFile.register_optimiser(InterWeave)
added
removed
bzrlib/weave.py
