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 dirname, sha_strings, split_lines
90
import bzrlib.patiencediff
91
from bzrlib.revision import NULL_REVISION
92
from bzrlib.symbol_versioning import *
93
from bzrlib.trace import mutter
94
from bzrlib.tsort import topo_sort
95
from bzrlib.versionedfile import (
101
from bzrlib.weavefile import _read_weave_v5, write_weave_v5
104
class WeaveContentFactory(ContentFactory):
105
"""Content factory for streaming from weaves.
107
:seealso ContentFactory:
110
def __init__(self, version, weave):
111
"""Create a WeaveContentFactory for version from weave."""
112
ContentFactory.__init__(self)
113
self.sha1 = weave.get_sha1s([version])[0]
114
self.key = (version,)
115
parents = weave.get_parent_map([version])[version]
116
self.parents = tuple((parent,) for parent in parents)
117
self.storage_kind = 'fulltext'
120
def get_bytes_as(self, storage_kind):
121
if storage_kind == 'fulltext':
122
return self._weave.get_text(self.key[-1])
124
raise UnavailableRepresentation(self.key, storage_kind, 'fulltext')
127
class Weave(VersionedFile):
128
"""weave - versioned text file storage.
130
A Weave manages versions of line-based text files, keeping track
131
of the originating version for each line.
133
To clients the "lines" of the file are represented as a list of strings.
134
These strings will typically have terminal newline characters, but
135
this is not required. In particular files commonly do not have a newline
136
at the end of the file.
138
Texts can be identified in either of two ways:
140
* a nonnegative index number.
142
* a version-id string.
144
Typically the index number will be valid only inside this weave and
145
the version-id is used to reference it in the larger world.
147
The weave is represented as a list mixing edit instructions and
148
literal text. Each entry in _weave can be either a string (or
149
unicode), or a tuple. If a string, it means that the given line
150
should be output in the currently active revisions.
152
If a tuple, it gives a processing instruction saying in which
153
revisions the enclosed lines are active. The tuple has the form
154
(instruction, version).
156
The instruction can be '{' or '}' for an insertion block, and '['
157
and ']' for a deletion block respectively. The version is the
158
integer version index. There is no replace operator, only deletes
159
and inserts. For '}', the end of an insertion, there is no
160
version parameter because it always closes the most recently
165
* A later version can delete lines that were introduced by any
166
number of ancestor versions; this implies that deletion
167
instructions can span insertion blocks without regard to the
168
insertion block's nesting.
170
* Similarly, deletions need not be properly nested with regard to
171
each other, because they might have been generated by
172
independent revisions.
174
* Insertions are always made by inserting a new bracketed block
175
into a single point in the previous weave. This implies they
176
can nest but not overlap, and the nesting must always have later
177
insertions on the inside.
179
* It doesn't seem very useful to have an active insertion
180
inside an inactive insertion, but it might happen.
182
* Therefore, all instructions are always"considered"; that
183
is passed onto and off the stack. An outer inactive block
184
doesn't disable an inner block.
186
* Lines are enabled if the most recent enclosing insertion is
187
active and none of the enclosing deletions are active.
189
* There is no point having a deletion directly inside its own
190
insertion; you might as well just not write it. And there
191
should be no way to get an earlier version deleting a later
195
Text of the weave; list of control instruction tuples and strings.
198
List of parents, indexed by version number.
199
It is only necessary to store the minimal set of parents for
200
each version; the parent's parents are implied.
203
List of hex SHA-1 of each version.
206
List of symbolic names for each version. Each should be unique.
209
For each name, the version number.
212
Descriptive name of this weave; typically the filename if known.
216
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
217
'_weave_name', '_matcher']
219
def __init__(self, weave_name=None, access_mode='w', matcher=None, get_scope=None):
222
:param get_scope: A callable that returns an opaque object to be used
223
for detecting when this weave goes out of scope (should stop
224
answering requests or allowing mutation).
226
super(Weave, self).__init__(access_mode)
232
self._weave_name = weave_name
234
self._matcher = bzrlib.patiencediff.PatienceSequenceMatcher
236
self._matcher = matcher
237
if get_scope is None:
238
get_scope = lambda:None
239
self._get_scope = get_scope
240
self._scope = get_scope()
241
self._access_mode = access_mode
244
return "Weave(%r)" % self._weave_name
246
def _check_write_ok(self):
247
"""Is the versioned file marked as 'finished' ? Raise if it is."""
248
if self._get_scope() != self._scope:
249
raise errors.OutSideTransaction()
250
if self._access_mode != 'w':
251
raise errors.ReadOnlyObjectDirtiedError(self)
254
"""Return a deep copy of self.
256
The copy can be modified without affecting the original weave."""
258
other._weave = self._weave[:]
259
other._parents = self._parents[:]
260
other._sha1s = self._sha1s[:]
261
other._names = self._names[:]
262
other._name_map = self._name_map.copy()
263
other._weave_name = self._weave_name
266
def __eq__(self, other):
267
if not isinstance(other, Weave):
269
return self._parents == other._parents \
270
and self._weave == other._weave \
271
and self._sha1s == other._sha1s
273
def __ne__(self, other):
274
return not self.__eq__(other)
276
def _idx_to_name(self, version):
277
return self._names[version]
279
def _lookup(self, name):
280
"""Convert symbolic version name to index."""
281
self.check_not_reserved_id(name)
283
return self._name_map[name]
285
raise RevisionNotPresent(name, self._weave_name)
288
"""See VersionedFile.versions."""
289
return self._names[:]
291
def has_version(self, version_id):
292
"""See VersionedFile.has_version."""
293
return (version_id in self._name_map)
295
__contains__ = has_version
297
def get_record_stream(self, versions, ordering, include_delta_closure):
298
"""Get a stream of records for versions.
300
:param versions: The versions to include. Each version is a tuple
302
:param ordering: Either 'unordered' or 'topological'. A topologically
303
sorted stream has compression parents strictly before their
305
:param include_delta_closure: If True then the closure across any
306
compression parents will be included (in the opaque data).
307
:return: An iterator of ContentFactory objects, each of which is only
308
valid until the iterator is advanced.
310
versions = [version[-1] for version in versions]
311
if ordering == 'topological':
312
parents = self.get_parent_map(versions)
313
new_versions = topo_sort(parents)
314
new_versions.extend(set(versions).difference(set(parents)))
315
versions = new_versions
316
for version in versions:
318
yield WeaveContentFactory(version, self)
320
yield AbsentContentFactory((version,))
322
def get_parent_map(self, version_ids):
323
"""See VersionedFile.get_parent_map."""
325
for version_id in version_ids:
326
if version_id == NULL_REVISION:
331
map(self._idx_to_name,
332
self._parents[self._lookup(version_id)]))
333
except RevisionNotPresent:
335
result[version_id] = parents
338
def get_parents_with_ghosts(self, version_id):
339
raise NotImplementedError(self.get_parents_with_ghosts)
341
def insert_record_stream(self, stream):
342
"""Insert a record stream into this versioned file.
344
:param stream: A stream of records to insert.
346
:seealso VersionedFile.get_record_stream:
349
for record in stream:
350
# Raise an error when a record is missing.
351
if record.storage_kind == 'absent':
352
raise RevisionNotPresent([record.key[0]], self)
353
# adapt to non-tuple interface
354
parents = [parent[0] for parent in record.parents]
355
if record.storage_kind == 'fulltext':
356
self.add_lines(record.key[0], parents,
357
split_lines(record.get_bytes_as('fulltext')))
359
adapter_key = record.storage_kind, 'fulltext'
361
adapter = adapters[adapter_key]
363
adapter_factory = adapter_registry.get(adapter_key)
364
adapter = adapter_factory(self)
365
adapters[adapter_key] = adapter
366
lines = split_lines(adapter.get_bytes(
367
record, record.get_bytes_as(record.storage_kind)))
369
self.add_lines(record.key[0], parents, lines)
370
except RevisionAlreadyPresent:
373
def _check_repeated_add(self, name, parents, text, sha1):
374
"""Check that a duplicated add is OK.
376
If it is, return the (old) index; otherwise raise an exception.
378
idx = self._lookup(name)
379
if sorted(self._parents[idx]) != sorted(parents) \
380
or sha1 != self._sha1s[idx]:
381
raise RevisionAlreadyPresent(name, self._weave_name)
384
def _add_lines(self, version_id, parents, lines, parent_texts,
385
left_matching_blocks, nostore_sha, random_id, check_content):
386
"""See VersionedFile.add_lines."""
387
idx = self._add(version_id, lines, map(self._lookup, parents),
388
nostore_sha=nostore_sha)
389
return sha_strings(lines), sum(map(len, lines)), idx
391
def _add(self, version_id, lines, parents, sha1=None, nostore_sha=None):
392
"""Add a single text on top of the weave.
394
Returns the index number of the newly added version.
397
Symbolic name for this version.
398
(Typically the revision-id of the revision that added it.)
401
List or set of direct parent version numbers.
404
Sequence of lines to be added in the new version.
406
:param nostore_sha: See VersionedFile.add_lines.
408
self._check_lines_not_unicode(lines)
409
self._check_lines_are_lines(lines)
411
sha1 = sha_strings(lines)
412
if sha1 == nostore_sha:
413
raise errors.ExistingContent
414
if version_id in self._name_map:
415
return self._check_repeated_add(version_id, parents, lines, sha1)
417
self._check_versions(parents)
418
## self._check_lines(lines)
419
new_version = len(self._parents)
421
# if we abort after here the (in-memory) weave will be corrupt because only
422
# some fields are updated
423
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
424
# - Robert Collins 20060226
425
self._parents.append(parents[:])
426
self._sha1s.append(sha1)
427
self._names.append(version_id)
428
self._name_map[version_id] = new_version
432
# special case; adding with no parents revision; can do
433
# this more quickly by just appending unconditionally.
434
# even more specially, if we're adding an empty text we
435
# need do nothing at all.
437
self._weave.append(('{', new_version))
438
self._weave.extend(lines)
439
self._weave.append(('}', None))
442
if len(parents) == 1:
443
pv = list(parents)[0]
444
if sha1 == self._sha1s[pv]:
445
# special case: same as the single parent
449
ancestors = self._inclusions(parents)
453
# basis a list of (origin, lineno, line)
456
for origin, lineno, line in self._extract(ancestors):
457
basis_lineno.append(lineno)
458
basis_lines.append(line)
460
# another small special case: a merge, producing the same text
462
if lines == basis_lines:
465
# add a sentinel, because we can also match against the final line
466
basis_lineno.append(len(self._weave))
468
# XXX: which line of the weave should we really consider
469
# matches the end of the file? the current code says it's the
470
# last line of the weave?
472
#print 'basis_lines:', basis_lines
473
#print 'new_lines: ', lines
475
s = self._matcher(None, basis_lines, lines)
477
# offset gives the number of lines that have been inserted
478
# into the weave up to the current point; if the original edit instruction
479
# says to change line A then we actually change (A+offset)
482
for tag, i1, i2, j1, j2 in s.get_opcodes():
483
# i1,i2 are given in offsets within basis_lines; we need to map them
484
# back to offsets within the entire weave
485
#print 'raw match', tag, i1, i2, j1, j2
488
i1 = basis_lineno[i1]
489
i2 = basis_lineno[i2]
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
dset.add(self._names[v])
598
dset.remove(self._names[v])
600
raise WeaveFormatError('unexpected instruction %r' % v)
602
yield lineno, istack[-1], frozenset(dset), l
606
raise WeaveFormatError("unclosed insertion blocks "
607
"at end of weave: %s" % istack)
609
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
612
def plan_merge(self, ver_a, ver_b):
613
"""Return pseudo-annotation indicating how the two versions merge.
615
This is computed between versions a and b and their common
618
Weave lines present in none of them are skipped entirely.
620
inc_a = set(self.get_ancestry([ver_a]))
621
inc_b = set(self.get_ancestry([ver_b]))
622
inc_c = inc_a & inc_b
624
for lineno, insert, deleteset, line in self._walk_internal([ver_a, ver_b]):
625
if deleteset & inc_c:
626
# killed in parent; can't be in either a or b
627
# not relevant to our work
628
yield 'killed-base', line
629
elif insert in inc_c:
630
# was inserted in base
631
killed_a = bool(deleteset & inc_a)
632
killed_b = bool(deleteset & inc_b)
633
if killed_a and killed_b:
634
yield 'killed-both', line
636
yield 'killed-a', line
638
yield 'killed-b', line
640
yield 'unchanged', line
641
elif insert in inc_a:
642
if deleteset & inc_a:
643
yield 'ghost-a', line
647
elif insert in inc_b:
648
if deleteset & inc_b:
649
yield 'ghost-b', line
653
# not in either revision
654
yield 'irrelevant', line
656
yield 'unchanged', '' # terminator
658
def _extract(self, versions):
659
"""Yield annotation of lines in included set.
661
Yields a sequence of tuples (origin, lineno, text), where
662
origin is the origin version, lineno the index in the weave,
663
and text the text of the line.
665
The set typically but not necessarily corresponds to a version.
668
if not isinstance(i, int):
671
included = self._inclusions(versions)
677
lineno = 0 # line of weave, 0-based
683
WFE = WeaveFormatError
686
# 449 0 4474.6820 2356.5590 bzrlib.weave:556(_extract)
687
# +285282 0 1676.8040 1676.8040 +<isinstance>
688
# 1.6 seconds in 'isinstance'.
689
# changing the first isinstance:
690
# 449 0 2814.2660 1577.1760 bzrlib.weave:556(_extract)
691
# +140414 0 762.8050 762.8050 +<isinstance>
692
# note that the inline time actually dropped (less function calls)
693
# and total processing time was halved.
694
# we're still spending ~1/4 of the method in isinstance though.
695
# so lets hard code the acceptable string classes we expect:
696
# 449 0 1202.9420 786.2930 bzrlib.weave:556(_extract)
697
# +71352 0 377.5560 377.5560 +<method 'append' of 'list'
699
# yay, down to ~1/4 the initial extract time, and our inline time
700
# has shrunk again, with isinstance no longer dominating.
701
# tweaking the stack inclusion test to use a set gives:
702
# 449 0 1122.8030 713.0080 bzrlib.weave:556(_extract)
703
# +71352 0 354.9980 354.9980 +<method 'append' of 'list'
705
# - a 5% win, or possibly just noise. However with large istacks that
706
# 'in' test could dominate, so I'm leaving this change in place -
707
# when its fast enough to consider profiling big datasets we can review.
712
for l in self._weave:
713
if l.__class__ == tuple:
720
iset.remove(istack.pop())
728
raise AssertionError()
731
isactive = (not dset) and istack and (istack[-1] in included)
733
result.append((istack[-1], lineno, l))
736
raise WeaveFormatError("unclosed insertion blocks "
737
"at end of weave: %s" % istack)
739
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
743
def _maybe_lookup(self, name_or_index):
744
"""Convert possible symbolic name to index, or pass through indexes.
748
if isinstance(name_or_index, (int, long)):
751
return self._lookup(name_or_index)
753
def get_lines(self, version_id):
754
"""See VersionedFile.get_lines()."""
755
int_index = self._maybe_lookup(version_id)
756
result = [line for (origin, lineno, line) in self._extract([int_index])]
757
expected_sha1 = self._sha1s[int_index]
758
measured_sha1 = sha_strings(result)
759
if measured_sha1 != expected_sha1:
760
raise errors.WeaveInvalidChecksum(
761
'file %s, revision %s, expected: %s, measured %s'
762
% (self._weave_name, version_id,
763
expected_sha1, measured_sha1))
766
def get_sha1s(self, version_ids):
767
"""See VersionedFile.get_sha1s()."""
768
return [self._sha1s[self._lookup(v)] for v in version_ids]
770
def num_versions(self):
771
"""How many versions are in this weave?"""
772
l = len(self._parents)
775
__len__ = num_versions
777
def check(self, progress_bar=None):
778
# TODO evaluate performance hit of using string sets in this routine.
779
# TODO: check no circular inclusions
780
# TODO: create a nested progress bar
781
for version in range(self.num_versions()):
782
inclusions = list(self._parents[version])
785
if inclusions[-1] >= version:
786
raise WeaveFormatError("invalid included version %d for index %d"
787
% (inclusions[-1], version))
789
# try extracting all versions; parallel extraction is used
790
nv = self.num_versions()
795
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
796
# The problem is that set membership is much more expensive
797
name = self._idx_to_name(i)
798
sha1s[name] = sha.new()
800
new_inc = set([name])
801
for p in self._parents[i]:
802
new_inc.update(inclusions[self._idx_to_name(p)])
804
if set(new_inc) != set(self.get_ancestry(name)):
805
raise AssertionError(
807
% (set(new_inc), set(self.get_ancestry(name))))
808
inclusions[name] = new_inc
810
nlines = len(self._weave)
812
update_text = 'checking weave'
814
short_name = os.path.basename(self._weave_name)
815
update_text = 'checking %s' % (short_name,)
816
update_text = update_text[:25]
818
for lineno, insert, deleteset, line in self._walk_internal():
820
progress_bar.update(update_text, lineno, nlines)
822
for name, name_inclusions in inclusions.items():
823
# The active inclusion must be an ancestor,
824
# and no ancestors must have deleted this line,
825
# because we don't support resurrection.
826
if (insert in name_inclusions) and not (deleteset & name_inclusions):
827
sha1s[name].update(line)
830
version = self._idx_to_name(i)
831
hd = sha1s[version].hexdigest()
832
expected = self._sha1s[i]
834
raise errors.WeaveInvalidChecksum(
835
"mismatched sha1 for version %s: "
836
"got %s, expected %s"
837
% (version, hd, expected))
839
# TODO: check insertions are properly nested, that there are
840
# no lines outside of insertion blocks, that deletions are
841
# properly paired, etc.
843
def _imported_parents(self, other, other_idx):
844
"""Return list of parents in self corresponding to indexes in other."""
846
for parent_idx in other._parents[other_idx]:
847
parent_name = other._names[parent_idx]
848
if parent_name not in self._name_map:
849
# should not be possible
850
raise WeaveError("missing parent {%s} of {%s} in %r"
851
% (parent_name, other._name_map[other_idx], self))
852
new_parents.append(self._name_map[parent_name])
855
def _check_version_consistent(self, other, other_idx, name):
856
"""Check if a version in consistent in this and other.
858
To be consistent it must have:
861
* the same direct parents (by name, not index, and disregarding
864
If present & correct return True;
865
if not present in self return False;
866
if inconsistent raise error."""
867
this_idx = self._name_map.get(name, -1)
869
if self._sha1s[this_idx] != other._sha1s[other_idx]:
870
raise errors.WeaveTextDiffers(name, self, other)
871
self_parents = self._parents[this_idx]
872
other_parents = other._parents[other_idx]
873
n1 = set([self._names[i] for i in self_parents])
874
n2 = set([other._names[i] for i in other_parents])
875
if not self._compatible_parents(n1, n2):
876
raise WeaveParentMismatch("inconsistent parents "
877
"for version {%s}: %s vs %s" % (name, n1, n2))
883
def _reweave(self, other, pb, msg):
884
"""Reweave self with other - internal helper for join().
886
:param other: The other weave to merge
887
:param pb: An optional progress bar, indicating how far done we are
888
:param msg: An optional message for the progress
890
new_weave = _reweave(self, other, pb=pb, msg=msg)
891
self._copy_weave_content(new_weave)
893
def _copy_weave_content(self, otherweave):
894
"""adsorb the content from otherweave."""
895
for attr in self.__slots__:
896
if attr != '_weave_name':
897
setattr(self, attr, copy(getattr(otherweave, attr)))
900
class WeaveFile(Weave):
901
"""A WeaveFile represents a Weave on disk and writes on change."""
903
WEAVE_SUFFIX = '.weave'
905
def __init__(self, name, transport, filemode=None, create=False, access_mode='w', get_scope=None):
906
"""Create a WeaveFile.
908
:param create: If not True, only open an existing knit.
910
super(WeaveFile, self).__init__(name, access_mode, get_scope=get_scope)
911
self._transport = transport
912
self._filemode = filemode
914
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
915
except errors.NoSuchFile:
921
def _add_lines(self, version_id, parents, lines, parent_texts,
922
left_matching_blocks, nostore_sha, random_id, check_content):
923
"""Add a version and save the weave."""
924
self.check_not_reserved_id(version_id)
925
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
926
parent_texts, left_matching_blocks, nostore_sha, random_id,
931
def copy_to(self, name, transport):
932
"""See VersionedFile.copy_to()."""
933
# as we are all in memory always, just serialise to the new place.
935
write_weave_v5(self, sio)
937
transport.put_file(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
940
"""Save the weave."""
941
self._check_write_ok()
943
write_weave_v5(self, sio)
945
bytes = sio.getvalue()
946
path = self._weave_name + WeaveFile.WEAVE_SUFFIX
948
self._transport.put_bytes(path, bytes, self._filemode)
949
except errors.NoSuchFile:
950
self._transport.mkdir(dirname(path))
951
self._transport.put_bytes(path, bytes, self._filemode)
955
"""See VersionedFile.get_suffixes()."""
956
return [WeaveFile.WEAVE_SUFFIX]
958
def insert_record_stream(self, stream):
959
super(WeaveFile, self).insert_record_stream(stream)
962
@deprecated_method(one_five)
963
def join(self, other, pb=None, msg=None, version_ids=None,
964
ignore_missing=False):
965
"""Join other into self and save."""
966
super(WeaveFile, self).join(other, pb, msg, version_ids, ignore_missing)
970
def _reweave(wa, wb, pb=None, msg=None):
971
"""Combine two weaves and return the result.
973
This works even if a revision R has different parents in
974
wa and wb. In the resulting weave all the parents are given.
976
This is done by just building up a new weave, maintaining ordering
977
of the versions in the two inputs. More efficient approaches
978
might be possible but it should only be necessary to do
979
this operation rarely, when a new previously ghost version is
982
:param pb: An optional progress bar, indicating how far done we are
983
:param msg: An optional message for the progress
987
queue_a = range(wa.num_versions())
988
queue_b = range(wb.num_versions())
989
# first determine combined parents of all versions
990
# map from version name -> all parent names
991
combined_parents = _reweave_parent_graphs(wa, wb)
992
mutter("combined parents: %r", combined_parents)
993
order = topo_sort(combined_parents.iteritems())
994
mutter("order to reweave: %r", order)
999
for idx, name in enumerate(order):
1001
pb.update(msg, idx, len(order))
1002
if name in wa._name_map:
1003
lines = wa.get_lines(name)
1004
if name in wb._name_map:
1005
lines_b = wb.get_lines(name)
1006
if lines != lines_b:
1007
mutter('Weaves differ on content. rev_id {%s}', name)
1008
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
1010
lines = list(difflib.unified_diff(lines, lines_b,
1011
wa._weave_name, wb._weave_name))
1012
mutter('lines:\n%s', ''.join(lines))
1013
raise errors.WeaveTextDiffers(name, wa, wb)
1015
lines = wb.get_lines(name)
1016
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
1019
def _reweave_parent_graphs(wa, wb):
1020
"""Return combined parent ancestry for two weaves.
1022
Returned as a list of (version_name, set(parent_names))"""
1024
for weave in [wa, wb]:
1025
for idx, name in enumerate(weave._names):
1026
p = combined.setdefault(name, set())
1027
p.update(map(weave._idx_to_name, weave._parents[idx]))
1032
"""Show the weave's table-of-contents"""
1033
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
1034
for i in (6, 50, 10, 10):
1037
for i in range(w.num_versions()):
1040
parent_str = ' '.join(map(str, w._parents[i]))
1041
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
1045
def weave_stats(weave_file, pb):
1046
from bzrlib.weavefile import read_weave
1048
wf = file(weave_file, 'rb')
1050
# FIXME: doesn't work on pipes
1051
weave_size = wf.tell()
1055
for i in range(vers):
1056
pb.update('checking sizes', i, vers)
1057
for origin, lineno, line in w._extract([i]):
1062
print 'versions %9d' % vers
1063
print 'weave file %9d bytes' % weave_size
1064
print 'total contents %9d bytes' % total
1065
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1068
print 'average size %9d bytes' % avg
1069
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1073
print """bzr weave tool
1075
Experimental tool for weave algorithm.
1078
weave init WEAVEFILE
1079
Create an empty weave file
1080
weave get WEAVEFILE VERSION
1081
Write out specified version.
1082
weave check WEAVEFILE
1083
Check consistency of all versions.
1085
Display table of contents.
1086
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1087
Add NEWTEXT, with specified parent versions.
1088
weave annotate WEAVEFILE VERSION
1089
Display origin of each line.
1090
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1091
Auto-merge two versions and display conflicts.
1092
weave diff WEAVEFILE VERSION1 VERSION2
1093
Show differences between two versions.
1097
% weave init foo.weave
1099
% weave add foo.weave ver0 < foo.txt
1102
(create updated version)
1104
% weave get foo.weave 0 | diff -u - foo.txt
1105
% weave add foo.weave ver1 0 < foo.txt
1108
% weave get foo.weave 0 > foo.txt (create forked version)
1110
% weave add foo.weave ver2 0 < foo.txt
1113
% weave merge foo.weave 1 2 > foo.txt (merge them)
1114
% vi foo.txt (resolve conflicts)
1115
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1127
# in case we're run directly from the subdirectory
1128
sys.path.append('..')
1130
from bzrlib.weavefile import write_weave, read_weave
1131
from bzrlib.progress import ProgressBar
1146
return read_weave(file(argv[2], 'rb'))
1152
# at the moment, based on everything in the file
1154
parents = map(int, argv[4:])
1155
lines = sys.stdin.readlines()
1156
ver = w.add(name, parents, lines)
1157
write_weave(w, file(argv[2], 'wb'))
1158
print 'added version %r %d' % (name, ver)
1161
if os.path.exists(fn):
1162
raise IOError("file exists")
1164
write_weave(w, file(fn, 'wb'))
1165
elif cmd == 'get': # get one version
1167
sys.stdout.writelines(w.get_iter(int(argv[3])))
1172
v1, v2 = map(int, argv[3:5])
1175
diff_gen = bzrlib.patiencediff.unified_diff(lines1, lines2,
1176
'%s version %d' % (fn, v1),
1177
'%s version %d' % (fn, v2))
1178
sys.stdout.writelines(diff_gen)
1180
elif cmd == 'annotate':
1182
# newline is added to all lines regardless; too hard to get
1183
# reasonable formatting otherwise
1185
for origin, text in w.annotate(int(argv[3])):
1186
text = text.rstrip('\r\n')
1188
print ' | %s' % (text)
1190
print '%5d | %s' % (origin, text)
1196
elif cmd == 'stats':
1197
weave_stats(argv[2], ProgressBar())
1199
elif cmd == 'check':
1204
print '%d versions ok' % w.num_versions()
1206
elif cmd == 'inclusions':
1208
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1210
elif cmd == 'parents':
1212
print ' '.join(map(str, w._parents[int(argv[3])]))
1214
elif cmd == 'plan-merge':
1215
# replaced by 'bzr weave-plan-merge'
1217
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1219
print '%14s | %s' % (state, line),
1220
elif cmd == 'merge':
1221
# replaced by 'bzr weave-merge-text'
1223
p = w.plan_merge(int(argv[3]), int(argv[4]))
1224
sys.stdout.writelines(w.weave_merge(p))
1226
raise ValueError('unknown command %r' % cmd)
1229
if __name__ == '__main__':
1231
sys.exit(main(sys.argv))
added
removed
bzrlib/weave.py
