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., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 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
77
from bzrlib.lazy_import import lazy_import
78
lazy_import(globals(), """
79
from bzrlib import tsort
86
from bzrlib.errors import (WeaveError, WeaveFormatError, WeaveParentMismatch,
87
RevisionAlreadyPresent,
89
UnavailableRepresentation,
90
WeaveRevisionAlreadyPresent,
91
WeaveRevisionNotPresent,
93
from bzrlib.osutils import dirname, sha, sha_strings, split_lines
94
import bzrlib.patiencediff
95
from bzrlib.revision import NULL_REVISION
96
from bzrlib.symbol_versioning import *
97
from bzrlib.trace import mutter
98
from bzrlib.versionedfile import (
105
from bzrlib.weavefile import _read_weave_v5, write_weave_v5
108
class WeaveContentFactory(ContentFactory):
109
"""Content factory for streaming from weaves.
111
:seealso ContentFactory:
114
def __init__(self, version, weave):
115
"""Create a WeaveContentFactory for version from weave."""
116
ContentFactory.__init__(self)
117
self.sha1 = weave.get_sha1s([version])[version]
118
self.key = (version,)
119
parents = weave.get_parent_map([version])[version]
120
self.parents = tuple((parent,) for parent in parents)
121
self.storage_kind = 'fulltext'
124
def get_bytes_as(self, storage_kind):
125
if storage_kind == 'fulltext':
126
return self._weave.get_text(self.key[-1])
127
elif storage_kind == 'chunked':
128
return self._weave.get_lines(self.key[-1])
130
raise UnavailableRepresentation(self.key, storage_kind, 'fulltext')
133
class Weave(VersionedFile):
134
"""weave - versioned text file storage.
136
A Weave manages versions of line-based text files, keeping track
137
of the originating version for each line.
139
To clients the "lines" of the file are represented as a list of strings.
140
These strings will typically have terminal newline characters, but
141
this is not required. In particular files commonly do not have a newline
142
at the end of the file.
144
Texts can be identified in either of two ways:
146
* a nonnegative index number.
148
* a version-id string.
150
Typically the index number will be valid only inside this weave and
151
the version-id is used to reference it in the larger world.
153
The weave is represented as a list mixing edit instructions and
154
literal text. Each entry in _weave can be either a string (or
155
unicode), or a tuple. If a string, it means that the given line
156
should be output in the currently active revisions.
158
If a tuple, it gives a processing instruction saying in which
159
revisions the enclosed lines are active. The tuple has the form
160
(instruction, version).
162
The instruction can be '{' or '}' for an insertion block, and '['
163
and ']' for a deletion block respectively. The version is the
164
integer version index. There is no replace operator, only deletes
165
and inserts. For '}', the end of an insertion, there is no
166
version parameter because it always closes the most recently
171
* A later version can delete lines that were introduced by any
172
number of ancestor versions; this implies that deletion
173
instructions can span insertion blocks without regard to the
174
insertion block's nesting.
176
* Similarly, deletions need not be properly nested with regard to
177
each other, because they might have been generated by
178
independent revisions.
180
* Insertions are always made by inserting a new bracketed block
181
into a single point in the previous weave. This implies they
182
can nest but not overlap, and the nesting must always have later
183
insertions on the inside.
185
* It doesn't seem very useful to have an active insertion
186
inside an inactive insertion, but it might happen.
188
* Therefore, all instructions are always"considered"; that
189
is passed onto and off the stack. An outer inactive block
190
doesn't disable an inner block.
192
* Lines are enabled if the most recent enclosing insertion is
193
active and none of the enclosing deletions are active.
195
* There is no point having a deletion directly inside its own
196
insertion; you might as well just not write it. And there
197
should be no way to get an earlier version deleting a later
201
Text of the weave; list of control instruction tuples and strings.
204
List of parents, indexed by version number.
205
It is only necessary to store the minimal set of parents for
206
each version; the parent's parents are implied.
209
List of hex SHA-1 of each version.
212
List of symbolic names for each version. Each should be unique.
215
For each name, the version number.
218
Descriptive name of this weave; typically the filename if known.
222
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
223
'_weave_name', '_matcher', '_allow_reserved']
225
def __init__(self, weave_name=None, access_mode='w', matcher=None,
226
get_scope=None, allow_reserved=False):
229
:param get_scope: A callable that returns an opaque object to be used
230
for detecting when this weave goes out of scope (should stop
231
answering requests or allowing mutation).
233
super(Weave, self).__init__()
239
self._weave_name = weave_name
241
self._matcher = bzrlib.patiencediff.PatienceSequenceMatcher
243
self._matcher = matcher
244
if get_scope is None:
245
get_scope = lambda:None
246
self._get_scope = get_scope
247
self._scope = get_scope()
248
self._access_mode = access_mode
249
self._allow_reserved = allow_reserved
252
return "Weave(%r)" % self._weave_name
254
def _check_write_ok(self):
255
"""Is the versioned file marked as 'finished' ? Raise if it is."""
256
if self._get_scope() != self._scope:
257
raise errors.OutSideTransaction()
258
if self._access_mode != 'w':
259
raise errors.ReadOnlyObjectDirtiedError(self)
262
"""Return a deep copy of self.
264
The copy can be modified without affecting the original weave."""
266
other._weave = self._weave[:]
267
other._parents = self._parents[:]
268
other._sha1s = self._sha1s[:]
269
other._names = self._names[:]
270
other._name_map = self._name_map.copy()
271
other._weave_name = self._weave_name
274
def __eq__(self, other):
275
if not isinstance(other, Weave):
277
return self._parents == other._parents \
278
and self._weave == other._weave \
279
and self._sha1s == other._sha1s
281
def __ne__(self, other):
282
return not self.__eq__(other)
284
def _idx_to_name(self, version):
285
return self._names[version]
287
def _lookup(self, name):
288
"""Convert symbolic version name to index."""
289
if not self._allow_reserved:
290
self.check_not_reserved_id(name)
292
return self._name_map[name]
294
raise RevisionNotPresent(name, self._weave_name)
297
"""See VersionedFile.versions."""
298
return self._names[:]
300
def has_version(self, version_id):
301
"""See VersionedFile.has_version."""
302
return (version_id in self._name_map)
304
__contains__ = has_version
306
def get_record_stream(self, versions, ordering, include_delta_closure):
307
"""Get a stream of records for versions.
309
:param versions: The versions to include. Each version is a tuple
311
:param ordering: Either 'unordered' or 'topological'. A topologically
312
sorted stream has compression parents strictly before their
314
:param include_delta_closure: If True then the closure across any
315
compression parents will be included (in the opaque data).
316
:return: An iterator of ContentFactory objects, each of which is only
317
valid until the iterator is advanced.
319
versions = [version[-1] for version in versions]
320
if ordering == 'topological':
321
parents = self.get_parent_map(versions)
322
new_versions = tsort.topo_sort(parents)
323
new_versions.extend(set(versions).difference(set(parents)))
324
versions = new_versions
325
elif ordering == 'groupcompress':
326
parents = self.get_parent_map(versions)
327
new_versions = sort_groupcompress(parents)
328
new_versions.extend(set(versions).difference(set(parents)))
329
versions = new_versions
330
for version in versions:
332
yield WeaveContentFactory(version, self)
334
yield AbsentContentFactory((version,))
336
def get_parent_map(self, version_ids):
337
"""See VersionedFile.get_parent_map."""
339
for version_id in version_ids:
340
if version_id == NULL_REVISION:
345
map(self._idx_to_name,
346
self._parents[self._lookup(version_id)]))
347
except RevisionNotPresent:
349
result[version_id] = parents
352
def get_parents_with_ghosts(self, version_id):
353
raise NotImplementedError(self.get_parents_with_ghosts)
355
def insert_record_stream(self, stream):
356
"""Insert a record stream into this versioned file.
358
:param stream: A stream of records to insert.
360
:seealso VersionedFile.get_record_stream:
363
for record in stream:
364
# Raise an error when a record is missing.
365
if record.storage_kind == 'absent':
366
raise RevisionNotPresent([record.key[0]], self)
367
# adapt to non-tuple interface
368
parents = [parent[0] for parent in record.parents]
369
if (record.storage_kind == 'fulltext'
370
or record.storage_kind == 'chunked'):
371
self.add_lines(record.key[0], parents,
372
osutils.chunks_to_lines(record.get_bytes_as('chunked')))
374
adapter_key = record.storage_kind, 'fulltext'
376
adapter = adapters[adapter_key]
378
adapter_factory = adapter_registry.get(adapter_key)
379
adapter = adapter_factory(self)
380
adapters[adapter_key] = adapter
381
lines = split_lines(adapter.get_bytes(record))
383
self.add_lines(record.key[0], parents, lines)
384
except RevisionAlreadyPresent:
387
def _check_repeated_add(self, name, parents, text, sha1):
388
"""Check that a duplicated add is OK.
390
If it is, return the (old) index; otherwise raise an exception.
392
idx = self._lookup(name)
393
if sorted(self._parents[idx]) != sorted(parents) \
394
or sha1 != self._sha1s[idx]:
395
raise RevisionAlreadyPresent(name, self._weave_name)
398
def _add_lines(self, version_id, parents, lines, parent_texts,
399
left_matching_blocks, nostore_sha, random_id, check_content):
400
"""See VersionedFile.add_lines."""
401
idx = self._add(version_id, lines, map(self._lookup, parents),
402
nostore_sha=nostore_sha)
403
return sha_strings(lines), sum(map(len, lines)), idx
405
def _add(self, version_id, lines, parents, sha1=None, nostore_sha=None):
406
"""Add a single text on top of the weave.
408
Returns the index number of the newly added version.
411
Symbolic name for this version.
412
(Typically the revision-id of the revision that added it.)
413
If None, a name will be allocated based on the hash. (sha1:SHAHASH)
416
List or set of direct parent version numbers.
419
Sequence of lines to be added in the new version.
421
:param nostore_sha: See VersionedFile.add_lines.
423
self._check_lines_not_unicode(lines)
424
self._check_lines_are_lines(lines)
426
sha1 = sha_strings(lines)
427
if sha1 == nostore_sha:
428
raise errors.ExistingContent
429
if version_id is None:
430
version_id = "sha1:" + sha1
431
if version_id in self._name_map:
432
return self._check_repeated_add(version_id, parents, lines, sha1)
434
self._check_versions(parents)
435
## self._check_lines(lines)
436
new_version = len(self._parents)
438
# if we abort after here the (in-memory) weave will be corrupt because only
439
# some fields are updated
440
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
441
# - Robert Collins 20060226
442
self._parents.append(parents[:])
443
self._sha1s.append(sha1)
444
self._names.append(version_id)
445
self._name_map[version_id] = new_version
449
# special case; adding with no parents revision; can do
450
# this more quickly by just appending unconditionally.
451
# even more specially, if we're adding an empty text we
452
# need do nothing at all.
454
self._weave.append(('{', new_version))
455
self._weave.extend(lines)
456
self._weave.append(('}', None))
459
if len(parents) == 1:
460
pv = list(parents)[0]
461
if sha1 == self._sha1s[pv]:
462
# special case: same as the single parent
466
ancestors = self._inclusions(parents)
470
# basis a list of (origin, lineno, line)
473
for origin, lineno, line in self._extract(ancestors):
474
basis_lineno.append(lineno)
475
basis_lines.append(line)
477
# another small special case: a merge, producing the same text
479
if lines == basis_lines:
482
# add a sentinel, because we can also match against the final line
483
basis_lineno.append(len(self._weave))
485
# XXX: which line of the weave should we really consider
486
# matches the end of the file? the current code says it's the
487
# last line of the weave?
489
#print 'basis_lines:', basis_lines
490
#print 'new_lines: ', lines
492
s = self._matcher(None, basis_lines, lines)
494
# offset gives the number of lines that have been inserted
495
# into the weave up to the current point; if the original edit instruction
496
# says to change line A then we actually change (A+offset)
499
for tag, i1, i2, j1, j2 in s.get_opcodes():
500
# i1,i2 are given in offsets within basis_lines; we need to map them
501
# back to offsets within the entire weave
502
#print 'raw match', tag, i1, i2, j1, j2
505
i1 = basis_lineno[i1]
506
i2 = basis_lineno[i2]
507
# the deletion and insertion are handled separately.
508
# first delete the region.
510
self._weave.insert(i1+offset, ('[', new_version))
511
self._weave.insert(i2+offset+1, (']', new_version))
515
# there may have been a deletion spanning up to
516
# i2; we want to insert after this region to make sure
517
# we don't destroy ourselves
519
self._weave[i:i] = ([('{', new_version)]
522
offset += 2 + (j2 - j1)
525
def _inclusions(self, versions):
526
"""Return set of all ancestors of given version(s)."""
527
if not len(versions):
530
for v in xrange(max(versions), 0, -1):
532
# include all its parents
533
i.update(self._parents[v])
535
## except IndexError:
536
## raise ValueError("version %d not present in weave" % v)
538
def get_ancestry(self, version_ids, topo_sorted=True):
539
"""See VersionedFile.get_ancestry."""
540
if isinstance(version_ids, basestring):
541
version_ids = [version_ids]
542
i = self._inclusions([self._lookup(v) for v in version_ids])
543
return [self._idx_to_name(v) for v in i]
545
def _check_lines(self, text):
546
if not isinstance(text, list):
547
raise ValueError("text should be a list, not %s" % type(text))
550
if not isinstance(l, basestring):
551
raise ValueError("text line should be a string or unicode, not %s"
556
def _check_versions(self, indexes):
557
"""Check everything in the sequence of indexes is valid"""
562
raise IndexError("invalid version number %r" % i)
564
def _compatible_parents(self, my_parents, other_parents):
565
"""During join check that other_parents are joinable with my_parents.
567
Joinable is defined as 'is a subset of' - supersets may require
568
regeneration of diffs, but subsets do not.
570
return len(other_parents.difference(my_parents)) == 0
572
def annotate(self, version_id):
573
"""Return a list of (version-id, line) tuples for version_id.
575
The index indicates when the line originated in the weave."""
576
incls = [self._lookup(version_id)]
577
return [(self._idx_to_name(origin), text) for origin, lineno, text in
578
self._extract(incls)]
580
def iter_lines_added_or_present_in_versions(self, version_ids=None,
582
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
583
if version_ids is None:
584
version_ids = self.versions()
585
version_ids = set(version_ids)
586
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
587
if inserted not in version_ids: continue
589
yield line + '\n', inserted
593
def _walk_internal(self, version_ids=None):
594
"""Helper method for weave actions."""
599
lineno = 0 # line of weave, 0-based
601
for l in self._weave:
602
if l.__class__ == tuple:
606
istack.append(self._names[v])
610
dset.add(self._names[v])
612
dset.remove(self._names[v])
614
raise WeaveFormatError('unexpected instruction %r' % v)
616
yield lineno, istack[-1], frozenset(dset), l
620
raise WeaveFormatError("unclosed insertion blocks "
621
"at end of weave: %s" % istack)
623
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
626
def plan_merge(self, ver_a, ver_b):
627
"""Return pseudo-annotation indicating how the two versions merge.
629
This is computed between versions a and b and their common
632
Weave lines present in none of them are skipped entirely.
634
inc_a = set(self.get_ancestry([ver_a]))
635
inc_b = set(self.get_ancestry([ver_b]))
636
inc_c = inc_a & inc_b
638
for lineno, insert, deleteset, line in self._walk_internal([ver_a, ver_b]):
639
if deleteset & inc_c:
640
# killed in parent; can't be in either a or b
641
# not relevant to our work
642
yield 'killed-base', line
643
elif insert in inc_c:
644
# was inserted in base
645
killed_a = bool(deleteset & inc_a)
646
killed_b = bool(deleteset & inc_b)
647
if killed_a and killed_b:
648
yield 'killed-both', line
650
yield 'killed-a', line
652
yield 'killed-b', line
654
yield 'unchanged', line
655
elif insert in inc_a:
656
if deleteset & inc_a:
657
yield 'ghost-a', line
661
elif insert in inc_b:
662
if deleteset & inc_b:
663
yield 'ghost-b', line
667
# not in either revision
668
yield 'irrelevant', line
670
def _extract(self, versions):
671
"""Yield annotation of lines in included set.
673
Yields a sequence of tuples (origin, lineno, text), where
674
origin is the origin version, lineno the index in the weave,
675
and text the text of the line.
677
The set typically but not necessarily corresponds to a version.
680
if not isinstance(i, int):
683
included = self._inclusions(versions)
689
lineno = 0 # line of weave, 0-based
695
WFE = WeaveFormatError
698
# 449 0 4474.6820 2356.5590 bzrlib.weave:556(_extract)
699
# +285282 0 1676.8040 1676.8040 +<isinstance>
700
# 1.6 seconds in 'isinstance'.
701
# changing the first isinstance:
702
# 449 0 2814.2660 1577.1760 bzrlib.weave:556(_extract)
703
# +140414 0 762.8050 762.8050 +<isinstance>
704
# note that the inline time actually dropped (less function calls)
705
# and total processing time was halved.
706
# we're still spending ~1/4 of the method in isinstance though.
707
# so lets hard code the acceptable string classes we expect:
708
# 449 0 1202.9420 786.2930 bzrlib.weave:556(_extract)
709
# +71352 0 377.5560 377.5560 +<method 'append' of 'list'
711
# yay, down to ~1/4 the initial extract time, and our inline time
712
# has shrunk again, with isinstance no longer dominating.
713
# tweaking the stack inclusion test to use a set gives:
714
# 449 0 1122.8030 713.0080 bzrlib.weave:556(_extract)
715
# +71352 0 354.9980 354.9980 +<method 'append' of 'list'
717
# - a 5% win, or possibly just noise. However with large istacks that
718
# 'in' test could dominate, so I'm leaving this change in place -
719
# when its fast enough to consider profiling big datasets we can review.
724
for l in self._weave:
725
if l.__class__ == tuple:
732
iset.remove(istack.pop())
740
raise AssertionError()
743
isactive = (not dset) and istack and (istack[-1] in included)
745
result.append((istack[-1], lineno, l))
748
raise WeaveFormatError("unclosed insertion blocks "
749
"at end of weave: %s" % istack)
751
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
755
def _maybe_lookup(self, name_or_index):
756
"""Convert possible symbolic name to index, or pass through indexes.
760
if isinstance(name_or_index, (int, long)):
763
return self._lookup(name_or_index)
765
def get_lines(self, version_id):
766
"""See VersionedFile.get_lines()."""
767
int_index = self._maybe_lookup(version_id)
768
result = [line for (origin, lineno, line) in self._extract([int_index])]
769
expected_sha1 = self._sha1s[int_index]
770
measured_sha1 = sha_strings(result)
771
if measured_sha1 != expected_sha1:
772
raise errors.WeaveInvalidChecksum(
773
'file %s, revision %s, expected: %s, measured %s'
774
% (self._weave_name, version_id,
775
expected_sha1, measured_sha1))
778
def get_sha1s(self, version_ids):
779
"""See VersionedFile.get_sha1s()."""
781
for v in version_ids:
782
result[v] = self._sha1s[self._lookup(v)]
785
def num_versions(self):
786
"""How many versions are in this weave?"""
787
l = len(self._parents)
790
__len__ = num_versions
792
def check(self, progress_bar=None):
793
# TODO evaluate performance hit of using string sets in this routine.
794
# TODO: check no circular inclusions
795
# TODO: create a nested progress bar
796
for version in range(self.num_versions()):
797
inclusions = list(self._parents[version])
800
if inclusions[-1] >= version:
801
raise WeaveFormatError("invalid included version %d for index %d"
802
% (inclusions[-1], version))
804
# try extracting all versions; parallel extraction is used
805
nv = self.num_versions()
810
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
811
# The problem is that set membership is much more expensive
812
name = self._idx_to_name(i)
815
new_inc = set([name])
816
for p in self._parents[i]:
817
new_inc.update(inclusions[self._idx_to_name(p)])
819
if set(new_inc) != set(self.get_ancestry(name)):
820
raise AssertionError(
822
% (set(new_inc), set(self.get_ancestry(name))))
823
inclusions[name] = new_inc
825
nlines = len(self._weave)
827
update_text = 'checking weave'
829
short_name = os.path.basename(self._weave_name)
830
update_text = 'checking %s' % (short_name,)
831
update_text = update_text[:25]
833
for lineno, insert, deleteset, line in self._walk_internal():
835
progress_bar.update(update_text, lineno, nlines)
837
for name, name_inclusions in inclusions.items():
838
# The active inclusion must be an ancestor,
839
# and no ancestors must have deleted this line,
840
# because we don't support resurrection.
841
if (insert in name_inclusions) and not (deleteset & name_inclusions):
842
sha1s[name].update(line)
845
version = self._idx_to_name(i)
846
hd = sha1s[version].hexdigest()
847
expected = self._sha1s[i]
849
raise errors.WeaveInvalidChecksum(
850
"mismatched sha1 for version %s: "
851
"got %s, expected %s"
852
% (version, hd, expected))
854
# TODO: check insertions are properly nested, that there are
855
# no lines outside of insertion blocks, that deletions are
856
# properly paired, etc.
858
def _imported_parents(self, other, other_idx):
859
"""Return list of parents in self corresponding to indexes in other."""
861
for parent_idx in other._parents[other_idx]:
862
parent_name = other._names[parent_idx]
863
if parent_name not in self._name_map:
864
# should not be possible
865
raise WeaveError("missing parent {%s} of {%s} in %r"
866
% (parent_name, other._name_map[other_idx], self))
867
new_parents.append(self._name_map[parent_name])
870
def _check_version_consistent(self, other, other_idx, name):
871
"""Check if a version in consistent in this and other.
873
To be consistent it must have:
876
* the same direct parents (by name, not index, and disregarding
879
If present & correct return True;
880
if not present in self return False;
881
if inconsistent raise error."""
882
this_idx = self._name_map.get(name, -1)
884
if self._sha1s[this_idx] != other._sha1s[other_idx]:
885
raise errors.WeaveTextDiffers(name, self, other)
886
self_parents = self._parents[this_idx]
887
other_parents = other._parents[other_idx]
888
n1 = set([self._names[i] for i in self_parents])
889
n2 = set([other._names[i] for i in other_parents])
890
if not self._compatible_parents(n1, n2):
891
raise WeaveParentMismatch("inconsistent parents "
892
"for version {%s}: %s vs %s" % (name, n1, n2))
898
def _reweave(self, other, pb, msg):
899
"""Reweave self with other - internal helper for join().
901
:param other: The other weave to merge
902
:param pb: An optional progress bar, indicating how far done we are
903
:param msg: An optional message for the progress
905
new_weave = _reweave(self, other, pb=pb, msg=msg)
906
self._copy_weave_content(new_weave)
908
def _copy_weave_content(self, otherweave):
909
"""adsorb the content from otherweave."""
910
for attr in self.__slots__:
911
if attr != '_weave_name':
912
setattr(self, attr, copy(getattr(otherweave, attr)))
915
class WeaveFile(Weave):
916
"""A WeaveFile represents a Weave on disk and writes on change."""
918
WEAVE_SUFFIX = '.weave'
920
def __init__(self, name, transport, filemode=None, create=False, access_mode='w', get_scope=None):
921
"""Create a WeaveFile.
923
:param create: If not True, only open an existing knit.
925
super(WeaveFile, self).__init__(name, access_mode, get_scope=get_scope,
926
allow_reserved=False)
927
self._transport = transport
928
self._filemode = filemode
930
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
931
except errors.NoSuchFile:
937
def _add_lines(self, version_id, parents, lines, parent_texts,
938
left_matching_blocks, nostore_sha, random_id, check_content):
939
"""Add a version and save the weave."""
940
self.check_not_reserved_id(version_id)
941
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
942
parent_texts, left_matching_blocks, nostore_sha, random_id,
947
def copy_to(self, name, transport):
948
"""See VersionedFile.copy_to()."""
949
# as we are all in memory always, just serialise to the new place.
951
write_weave_v5(self, sio)
953
transport.put_file(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
956
"""Save the weave."""
957
self._check_write_ok()
959
write_weave_v5(self, sio)
961
bytes = sio.getvalue()
962
path = self._weave_name + WeaveFile.WEAVE_SUFFIX
964
self._transport.put_bytes(path, bytes, self._filemode)
965
except errors.NoSuchFile:
966
self._transport.mkdir(dirname(path))
967
self._transport.put_bytes(path, bytes, self._filemode)
971
"""See VersionedFile.get_suffixes()."""
972
return [WeaveFile.WEAVE_SUFFIX]
974
def insert_record_stream(self, stream):
975
super(WeaveFile, self).insert_record_stream(stream)
979
def _reweave(wa, wb, pb=None, msg=None):
980
"""Combine two weaves and return the result.
982
This works even if a revision R has different parents in
983
wa and wb. In the resulting weave all the parents are given.
985
This is done by just building up a new weave, maintaining ordering
986
of the versions in the two inputs. More efficient approaches
987
might be possible but it should only be necessary to do
988
this operation rarely, when a new previously ghost version is
991
:param pb: An optional progress bar, indicating how far done we are
992
:param msg: An optional message for the progress
996
queue_a = range(wa.num_versions())
997
queue_b = range(wb.num_versions())
998
# first determine combined parents of all versions
999
# map from version name -> all parent names
1000
combined_parents = _reweave_parent_graphs(wa, wb)
1001
mutter("combined parents: %r", combined_parents)
1002
order = tsort.topo_sort(combined_parents.iteritems())
1003
mutter("order to reweave: %r", order)
1008
for idx, name in enumerate(order):
1010
pb.update(msg, idx, len(order))
1011
if name in wa._name_map:
1012
lines = wa.get_lines(name)
1013
if name in wb._name_map:
1014
lines_b = wb.get_lines(name)
1015
if lines != lines_b:
1016
mutter('Weaves differ on content. rev_id {%s}', name)
1017
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
1019
lines = list(difflib.unified_diff(lines, lines_b,
1020
wa._weave_name, wb._weave_name))
1021
mutter('lines:\n%s', ''.join(lines))
1022
raise errors.WeaveTextDiffers(name, wa, wb)
1024
lines = wb.get_lines(name)
1025
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
1028
def _reweave_parent_graphs(wa, wb):
1029
"""Return combined parent ancestry for two weaves.
1031
Returned as a list of (version_name, set(parent_names))"""
1033
for weave in [wa, wb]:
1034
for idx, name in enumerate(weave._names):
1035
p = combined.setdefault(name, set())
1036
p.update(map(weave._idx_to_name, weave._parents[idx]))
1041
"""Show the weave's table-of-contents"""
1042
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
1043
for i in (6, 50, 10, 10):
1046
for i in range(w.num_versions()):
1049
parent_str = ' '.join(map(str, w._parents[i]))
1050
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
1054
def weave_stats(weave_file, pb):
1055
from bzrlib.weavefile import read_weave
1057
wf = file(weave_file, 'rb')
1059
# FIXME: doesn't work on pipes
1060
weave_size = wf.tell()
1064
for i in range(vers):
1065
pb.update('checking sizes', i, vers)
1066
for origin, lineno, line in w._extract([i]):
1071
print 'versions %9d' % vers
1072
print 'weave file %9d bytes' % weave_size
1073
print 'total contents %9d bytes' % total
1074
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1077
print 'average size %9d bytes' % avg
1078
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1082
print """bzr weave tool
1084
Experimental tool for weave algorithm.
1087
weave init WEAVEFILE
1088
Create an empty weave file
1089
weave get WEAVEFILE VERSION
1090
Write out specified version.
1091
weave check WEAVEFILE
1092
Check consistency of all versions.
1094
Display table of contents.
1095
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1096
Add NEWTEXT, with specified parent versions.
1097
weave annotate WEAVEFILE VERSION
1098
Display origin of each line.
1099
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1100
Auto-merge two versions and display conflicts.
1101
weave diff WEAVEFILE VERSION1 VERSION2
1102
Show differences between two versions.
1106
% weave init foo.weave
1108
% weave add foo.weave ver0 < foo.txt
1111
(create updated version)
1113
% weave get foo.weave 0 | diff -u - foo.txt
1114
% weave add foo.weave ver1 0 < foo.txt
1117
% weave get foo.weave 0 > foo.txt (create forked version)
1119
% weave add foo.weave ver2 0 < foo.txt
1122
% weave merge foo.weave 1 2 > foo.txt (merge them)
1123
% vi foo.txt (resolve conflicts)
1124
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1136
# in case we're run directly from the subdirectory
1137
sys.path.append('..')
1139
from bzrlib.weavefile import write_weave, read_weave
1140
from bzrlib.progress import ProgressBar
1155
return read_weave(file(argv[2], 'rb'))
1161
# at the moment, based on everything in the file
1163
parents = map(int, argv[4:])
1164
lines = sys.stdin.readlines()
1165
ver = w.add(name, parents, lines)
1166
write_weave(w, file(argv[2], 'wb'))
1167
print 'added version %r %d' % (name, ver)
1170
if os.path.exists(fn):
1171
raise IOError("file exists")
1173
write_weave(w, file(fn, 'wb'))
1174
elif cmd == 'get': # get one version
1176
sys.stdout.writelines(w.get_iter(int(argv[3])))
1181
v1, v2 = map(int, argv[3:5])
1184
diff_gen = bzrlib.patiencediff.unified_diff(lines1, lines2,
1185
'%s version %d' % (fn, v1),
1186
'%s version %d' % (fn, v2))
1187
sys.stdout.writelines(diff_gen)
1189
elif cmd == 'annotate':
1191
# newline is added to all lines regardless; too hard to get
1192
# reasonable formatting otherwise
1194
for origin, text in w.annotate(int(argv[3])):
1195
text = text.rstrip('\r\n')
1197
print ' | %s' % (text)
1199
print '%5d | %s' % (origin, text)
1205
elif cmd == 'stats':
1206
weave_stats(argv[2], ProgressBar())
1208
elif cmd == 'check':
1213
print '%d versions ok' % w.num_versions()
1215
elif cmd == 'inclusions':
1217
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1219
elif cmd == 'parents':
1221
print ' '.join(map(str, w._parents[int(argv[3])]))
1223
elif cmd == 'plan-merge':
1224
# replaced by 'bzr weave-plan-merge'
1226
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1228
print '%14s | %s' % (state, line),
1229
elif cmd == 'merge':
1230
# replaced by 'bzr weave-merge-text'
1232
p = w.plan_merge(int(argv[3]), int(argv[4]))
1233
sys.stdout.writelines(w.weave_merge(p))
1235
raise ValueError('unknown command %r' % cmd)
1238
if __name__ == '__main__':
1240
sys.exit(main(sys.argv))
added
removed
bzrlib/weave.py
