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
80
from bzrlib.errors import (WeaveError, WeaveFormatError, WeaveParentMismatch,
81
RevisionAlreadyPresent,
83
UnavailableRepresentation,
84
WeaveRevisionAlreadyPresent,
85
WeaveRevisionNotPresent,
87
import bzrlib.errors as errors
88
from bzrlib.osutils import dirname, sha, sha_strings, split_lines
89
import bzrlib.patiencediff
90
from bzrlib.revision import NULL_REVISION
91
from bzrlib.symbol_versioning import *
92
from bzrlib.trace import mutter
93
from bzrlib.tsort import topo_sort
94
from bzrlib.versionedfile import (
100
from bzrlib.weavefile import _read_weave_v5, write_weave_v5
103
class WeaveContentFactory(ContentFactory):
104
"""Content factory for streaming from weaves.
106
:seealso ContentFactory:
109
def __init__(self, version, weave):
110
"""Create a WeaveContentFactory for version from weave."""
111
ContentFactory.__init__(self)
112
self.sha1 = weave.get_sha1s([version])[version]
113
self.key = (version,)
114
parents = weave.get_parent_map([version])[version]
115
self.parents = tuple((parent,) for parent in parents)
116
self.storage_kind = 'fulltext'
119
def get_bytes_as(self, storage_kind):
120
if storage_kind == 'fulltext':
121
return self._weave.get_text(self.key[-1])
123
raise UnavailableRepresentation(self.key, storage_kind, 'fulltext')
126
class Weave(VersionedFile):
127
"""weave - versioned text file storage.
129
A Weave manages versions of line-based text files, keeping track
130
of the originating version for each line.
132
To clients the "lines" of the file are represented as a list of strings.
133
These strings will typically have terminal newline characters, but
134
this is not required. In particular files commonly do not have a newline
135
at the end of the file.
137
Texts can be identified in either of two ways:
139
* a nonnegative index number.
141
* a version-id string.
143
Typically the index number will be valid only inside this weave and
144
the version-id is used to reference it in the larger world.
146
The weave is represented as a list mixing edit instructions and
147
literal text. Each entry in _weave can be either a string (or
148
unicode), or a tuple. If a string, it means that the given line
149
should be output in the currently active revisions.
151
If a tuple, it gives a processing instruction saying in which
152
revisions the enclosed lines are active. The tuple has the form
153
(instruction, version).
155
The instruction can be '{' or '}' for an insertion block, and '['
156
and ']' for a deletion block respectively. The version is the
157
integer version index. There is no replace operator, only deletes
158
and inserts. For '}', the end of an insertion, there is no
159
version parameter because it always closes the most recently
164
* A later version can delete lines that were introduced by any
165
number of ancestor versions; this implies that deletion
166
instructions can span insertion blocks without regard to the
167
insertion block's nesting.
169
* Similarly, deletions need not be properly nested with regard to
170
each other, because they might have been generated by
171
independent revisions.
173
* Insertions are always made by inserting a new bracketed block
174
into a single point in the previous weave. This implies they
175
can nest but not overlap, and the nesting must always have later
176
insertions on the inside.
178
* It doesn't seem very useful to have an active insertion
179
inside an inactive insertion, but it might happen.
181
* Therefore, all instructions are always"considered"; that
182
is passed onto and off the stack. An outer inactive block
183
doesn't disable an inner block.
185
* Lines are enabled if the most recent enclosing insertion is
186
active and none of the enclosing deletions are active.
188
* There is no point having a deletion directly inside its own
189
insertion; you might as well just not write it. And there
190
should be no way to get an earlier version deleting a later
194
Text of the weave; list of control instruction tuples and strings.
197
List of parents, indexed by version number.
198
It is only necessary to store the minimal set of parents for
199
each version; the parent's parents are implied.
202
List of hex SHA-1 of each version.
205
List of symbolic names for each version. Each should be unique.
208
For each name, the version number.
211
Descriptive name of this weave; typically the filename if known.
215
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
216
'_weave_name', '_matcher', '_allow_reserved']
218
def __init__(self, weave_name=None, access_mode='w', matcher=None,
219
get_scope=None, allow_reserved=False):
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__()
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
242
self._allow_reserved = allow_reserved
245
return "Weave(%r)" % self._weave_name
247
def _check_write_ok(self):
248
"""Is the versioned file marked as 'finished' ? Raise if it is."""
249
if self._get_scope() != self._scope:
250
raise errors.OutSideTransaction()
251
if self._access_mode != 'w':
252
raise errors.ReadOnlyObjectDirtiedError(self)
255
"""Return a deep copy of self.
257
The copy can be modified without affecting the original weave."""
259
other._weave = self._weave[:]
260
other._parents = self._parents[:]
261
other._sha1s = self._sha1s[:]
262
other._names = self._names[:]
263
other._name_map = self._name_map.copy()
264
other._weave_name = self._weave_name
267
def __eq__(self, other):
268
if not isinstance(other, Weave):
270
return self._parents == other._parents \
271
and self._weave == other._weave \
272
and self._sha1s == other._sha1s
274
def __ne__(self, other):
275
return not self.__eq__(other)
277
def _idx_to_name(self, version):
278
return self._names[version]
280
def _lookup(self, name):
281
"""Convert symbolic version name to index."""
282
if not self._allow_reserved:
283
self.check_not_reserved_id(name)
285
return self._name_map[name]
287
raise RevisionNotPresent(name, self._weave_name)
290
"""See VersionedFile.versions."""
291
return self._names[:]
293
def has_version(self, version_id):
294
"""See VersionedFile.has_version."""
295
return (version_id in self._name_map)
297
__contains__ = has_version
299
def get_record_stream(self, versions, ordering, include_delta_closure):
300
"""Get a stream of records for versions.
302
:param versions: The versions to include. Each version is a tuple
304
:param ordering: Either 'unordered' or 'topological'. A topologically
305
sorted stream has compression parents strictly before their
307
:param include_delta_closure: If True then the closure across any
308
compression parents will be included (in the opaque data).
309
:return: An iterator of ContentFactory objects, each of which is only
310
valid until the iterator is advanced.
312
versions = [version[-1] for version in versions]
313
if ordering == 'topological':
314
parents = self.get_parent_map(versions)
315
new_versions = topo_sort(parents)
316
new_versions.extend(set(versions).difference(set(parents)))
317
versions = new_versions
318
for version in versions:
320
yield WeaveContentFactory(version, self)
322
yield AbsentContentFactory((version,))
324
def get_parent_map(self, version_ids):
325
"""See VersionedFile.get_parent_map."""
327
for version_id in version_ids:
328
if version_id == NULL_REVISION:
333
map(self._idx_to_name,
334
self._parents[self._lookup(version_id)]))
335
except RevisionNotPresent:
337
result[version_id] = parents
340
def get_parents_with_ghosts(self, version_id):
341
raise NotImplementedError(self.get_parents_with_ghosts)
343
def insert_record_stream(self, stream):
344
"""Insert a record stream into this versioned file.
346
:param stream: A stream of records to insert.
348
:seealso VersionedFile.get_record_stream:
351
for record in stream:
352
# Raise an error when a record is missing.
353
if record.storage_kind == 'absent':
354
raise RevisionNotPresent([record.key[0]], self)
355
# adapt to non-tuple interface
356
parents = [parent[0] for parent in record.parents]
357
if record.storage_kind == 'fulltext':
358
self.add_lines(record.key[0], parents,
359
split_lines(record.get_bytes_as('fulltext')))
361
adapter_key = record.storage_kind, 'fulltext'
363
adapter = adapters[adapter_key]
365
adapter_factory = adapter_registry.get(adapter_key)
366
adapter = adapter_factory(self)
367
adapters[adapter_key] = adapter
368
lines = split_lines(adapter.get_bytes(
369
record, record.get_bytes_as(record.storage_kind)))
371
self.add_lines(record.key[0], parents, lines)
372
except RevisionAlreadyPresent:
375
def _check_repeated_add(self, name, parents, text, sha1):
376
"""Check that a duplicated add is OK.
378
If it is, return the (old) index; otherwise raise an exception.
380
idx = self._lookup(name)
381
if sorted(self._parents[idx]) != sorted(parents) \
382
or sha1 != self._sha1s[idx]:
383
raise RevisionAlreadyPresent(name, self._weave_name)
386
def _add_lines(self, version_id, parents, lines, parent_texts,
387
left_matching_blocks, nostore_sha, random_id, check_content):
388
"""See VersionedFile.add_lines."""
389
idx = self._add(version_id, lines, map(self._lookup, parents),
390
nostore_sha=nostore_sha)
391
return sha_strings(lines), sum(map(len, lines)), idx
393
def _add(self, version_id, lines, parents, sha1=None, nostore_sha=None):
394
"""Add a single text on top of the weave.
396
Returns the index number of the newly added version.
399
Symbolic name for this version.
400
(Typically the revision-id of the revision that added it.)
403
List or set of direct parent version numbers.
406
Sequence of lines to be added in the new version.
408
:param nostore_sha: See VersionedFile.add_lines.
410
self._check_lines_not_unicode(lines)
411
self._check_lines_are_lines(lines)
413
sha1 = sha_strings(lines)
414
if sha1 == nostore_sha:
415
raise errors.ExistingContent
416
if version_id in self._name_map:
417
return self._check_repeated_add(version_id, parents, lines, sha1)
419
self._check_versions(parents)
420
## self._check_lines(lines)
421
new_version = len(self._parents)
423
# if we abort after here the (in-memory) weave will be corrupt because only
424
# some fields are updated
425
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
426
# - Robert Collins 20060226
427
self._parents.append(parents[:])
428
self._sha1s.append(sha1)
429
self._names.append(version_id)
430
self._name_map[version_id] = new_version
434
# special case; adding with no parents revision; can do
435
# this more quickly by just appending unconditionally.
436
# even more specially, if we're adding an empty text we
437
# need do nothing at all.
439
self._weave.append(('{', new_version))
440
self._weave.extend(lines)
441
self._weave.append(('}', None))
444
if len(parents) == 1:
445
pv = list(parents)[0]
446
if sha1 == self._sha1s[pv]:
447
# special case: same as the single parent
451
ancestors = self._inclusions(parents)
455
# basis a list of (origin, lineno, line)
458
for origin, lineno, line in self._extract(ancestors):
459
basis_lineno.append(lineno)
460
basis_lines.append(line)
462
# another small special case: a merge, producing the same text
464
if lines == basis_lines:
467
# add a sentinel, because we can also match against the final line
468
basis_lineno.append(len(self._weave))
470
# XXX: which line of the weave should we really consider
471
# matches the end of the file? the current code says it's the
472
# last line of the weave?
474
#print 'basis_lines:', basis_lines
475
#print 'new_lines: ', lines
477
s = self._matcher(None, basis_lines, lines)
479
# offset gives the number of lines that have been inserted
480
# into the weave up to the current point; if the original edit instruction
481
# says to change line A then we actually change (A+offset)
484
for tag, i1, i2, j1, j2 in s.get_opcodes():
485
# i1,i2 are given in offsets within basis_lines; we need to map them
486
# back to offsets within the entire weave
487
#print 'raw match', tag, i1, i2, j1, j2
490
i1 = basis_lineno[i1]
491
i2 = basis_lineno[i2]
492
# the deletion and insertion are handled separately.
493
# first delete the region.
495
self._weave.insert(i1+offset, ('[', new_version))
496
self._weave.insert(i2+offset+1, (']', new_version))
500
# there may have been a deletion spanning up to
501
# i2; we want to insert after this region to make sure
502
# we don't destroy ourselves
504
self._weave[i:i] = ([('{', new_version)]
507
offset += 2 + (j2 - j1)
510
def _inclusions(self, versions):
511
"""Return set of all ancestors of given version(s)."""
512
if not len(versions):
515
for v in xrange(max(versions), 0, -1):
517
# include all its parents
518
i.update(self._parents[v])
520
## except IndexError:
521
## raise ValueError("version %d not present in weave" % v)
523
def get_ancestry(self, version_ids, topo_sorted=True):
524
"""See VersionedFile.get_ancestry."""
525
if isinstance(version_ids, basestring):
526
version_ids = [version_ids]
527
i = self._inclusions([self._lookup(v) for v in version_ids])
528
return [self._idx_to_name(v) for v in i]
530
def _check_lines(self, text):
531
if not isinstance(text, list):
532
raise ValueError("text should be a list, not %s" % type(text))
535
if not isinstance(l, basestring):
536
raise ValueError("text line should be a string or unicode, not %s"
541
def _check_versions(self, indexes):
542
"""Check everything in the sequence of indexes is valid"""
547
raise IndexError("invalid version number %r" % i)
549
def _compatible_parents(self, my_parents, other_parents):
550
"""During join check that other_parents are joinable with my_parents.
552
Joinable is defined as 'is a subset of' - supersets may require
553
regeneration of diffs, but subsets do not.
555
return len(other_parents.difference(my_parents)) == 0
557
def annotate(self, version_id):
558
"""Return a list of (version-id, line) tuples for version_id.
560
The index indicates when the line originated in the weave."""
561
incls = [self._lookup(version_id)]
562
return [(self._idx_to_name(origin), text) for origin, lineno, text in
563
self._extract(incls)]
565
def iter_lines_added_or_present_in_versions(self, version_ids=None,
567
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
568
if version_ids is None:
569
version_ids = self.versions()
570
version_ids = set(version_ids)
571
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
572
# if inserted not in version_ids then it was inserted before the
573
# versions we care about, but because weaves cannot represent ghosts
574
# properly, we do not filter down to that
575
# if inserted not in version_ids: continue
577
yield line + '\n', inserted
581
def _walk_internal(self, version_ids=None):
582
"""Helper method for weave actions."""
587
lineno = 0 # line of weave, 0-based
589
for l in self._weave:
590
if l.__class__ == tuple:
594
istack.append(self._names[v])
598
dset.add(self._names[v])
600
dset.remove(self._names[v])
602
raise WeaveFormatError('unexpected instruction %r' % v)
604
yield lineno, istack[-1], frozenset(dset), l
608
raise WeaveFormatError("unclosed insertion blocks "
609
"at end of weave: %s" % istack)
611
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
614
def plan_merge(self, ver_a, ver_b):
615
"""Return pseudo-annotation indicating how the two versions merge.
617
This is computed between versions a and b and their common
620
Weave lines present in none of them are skipped entirely.
622
inc_a = set(self.get_ancestry([ver_a]))
623
inc_b = set(self.get_ancestry([ver_b]))
624
inc_c = inc_a & inc_b
626
for lineno, insert, deleteset, line in self._walk_internal([ver_a, ver_b]):
627
if deleteset & inc_c:
628
# killed in parent; can't be in either a or b
629
# not relevant to our work
630
yield 'killed-base', line
631
elif insert in inc_c:
632
# was inserted in base
633
killed_a = bool(deleteset & inc_a)
634
killed_b = bool(deleteset & inc_b)
635
if killed_a and killed_b:
636
yield 'killed-both', line
638
yield 'killed-a', line
640
yield 'killed-b', line
642
yield 'unchanged', line
643
elif insert in inc_a:
644
if deleteset & inc_a:
645
yield 'ghost-a', line
649
elif insert in inc_b:
650
if deleteset & inc_b:
651
yield 'ghost-b', line
655
# not in either revision
656
yield 'irrelevant', line
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()."""
769
for v in version_ids:
770
result[v] = self._sha1s[self._lookup(v)]
773
def num_versions(self):
774
"""How many versions are in this weave?"""
775
l = len(self._parents)
778
__len__ = num_versions
780
def check(self, progress_bar=None):
781
# TODO evaluate performance hit of using string sets in this routine.
782
# TODO: check no circular inclusions
783
# TODO: create a nested progress bar
784
for version in range(self.num_versions()):
785
inclusions = list(self._parents[version])
788
if inclusions[-1] >= version:
789
raise WeaveFormatError("invalid included version %d for index %d"
790
% (inclusions[-1], version))
792
# try extracting all versions; parallel extraction is used
793
nv = self.num_versions()
798
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
799
# The problem is that set membership is much more expensive
800
name = self._idx_to_name(i)
803
new_inc = set([name])
804
for p in self._parents[i]:
805
new_inc.update(inclusions[self._idx_to_name(p)])
807
if set(new_inc) != set(self.get_ancestry(name)):
808
raise AssertionError(
810
% (set(new_inc), set(self.get_ancestry(name))))
811
inclusions[name] = new_inc
813
nlines = len(self._weave)
815
update_text = 'checking weave'
817
short_name = os.path.basename(self._weave_name)
818
update_text = 'checking %s' % (short_name,)
819
update_text = update_text[:25]
821
for lineno, insert, deleteset, line in self._walk_internal():
823
progress_bar.update(update_text, lineno, nlines)
825
for name, name_inclusions in inclusions.items():
826
# The active inclusion must be an ancestor,
827
# and no ancestors must have deleted this line,
828
# because we don't support resurrection.
829
if (insert in name_inclusions) and not (deleteset & name_inclusions):
830
sha1s[name].update(line)
833
version = self._idx_to_name(i)
834
hd = sha1s[version].hexdigest()
835
expected = self._sha1s[i]
837
raise errors.WeaveInvalidChecksum(
838
"mismatched sha1 for version %s: "
839
"got %s, expected %s"
840
% (version, hd, expected))
842
# TODO: check insertions are properly nested, that there are
843
# no lines outside of insertion blocks, that deletions are
844
# properly paired, etc.
846
def _imported_parents(self, other, other_idx):
847
"""Return list of parents in self corresponding to indexes in other."""
849
for parent_idx in other._parents[other_idx]:
850
parent_name = other._names[parent_idx]
851
if parent_name not in self._name_map:
852
# should not be possible
853
raise WeaveError("missing parent {%s} of {%s} in %r"
854
% (parent_name, other._name_map[other_idx], self))
855
new_parents.append(self._name_map[parent_name])
858
def _check_version_consistent(self, other, other_idx, name):
859
"""Check if a version in consistent in this and other.
861
To be consistent it must have:
864
* the same direct parents (by name, not index, and disregarding
867
If present & correct return True;
868
if not present in self return False;
869
if inconsistent raise error."""
870
this_idx = self._name_map.get(name, -1)
872
if self._sha1s[this_idx] != other._sha1s[other_idx]:
873
raise errors.WeaveTextDiffers(name, self, other)
874
self_parents = self._parents[this_idx]
875
other_parents = other._parents[other_idx]
876
n1 = set([self._names[i] for i in self_parents])
877
n2 = set([other._names[i] for i in other_parents])
878
if not self._compatible_parents(n1, n2):
879
raise WeaveParentMismatch("inconsistent parents "
880
"for version {%s}: %s vs %s" % (name, n1, n2))
886
def _reweave(self, other, pb, msg):
887
"""Reweave self with other - internal helper for join().
889
:param other: The other weave to merge
890
:param pb: An optional progress bar, indicating how far done we are
891
:param msg: An optional message for the progress
893
new_weave = _reweave(self, other, pb=pb, msg=msg)
894
self._copy_weave_content(new_weave)
896
def _copy_weave_content(self, otherweave):
897
"""adsorb the content from otherweave."""
898
for attr in self.__slots__:
899
if attr != '_weave_name':
900
setattr(self, attr, copy(getattr(otherweave, attr)))
903
class WeaveFile(Weave):
904
"""A WeaveFile represents a Weave on disk and writes on change."""
906
WEAVE_SUFFIX = '.weave'
908
def __init__(self, name, transport, filemode=None, create=False, access_mode='w', get_scope=None):
909
"""Create a WeaveFile.
911
:param create: If not True, only open an existing knit.
913
super(WeaveFile, self).__init__(name, access_mode, get_scope=get_scope,
914
allow_reserved=False)
915
self._transport = transport
916
self._filemode = filemode
918
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
919
except errors.NoSuchFile:
925
def _add_lines(self, version_id, parents, lines, parent_texts,
926
left_matching_blocks, nostore_sha, random_id, check_content):
927
"""Add a version and save the weave."""
928
self.check_not_reserved_id(version_id)
929
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
930
parent_texts, left_matching_blocks, nostore_sha, random_id,
935
def copy_to(self, name, transport):
936
"""See VersionedFile.copy_to()."""
937
# as we are all in memory always, just serialise to the new place.
939
write_weave_v5(self, sio)
941
transport.put_file(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
944
"""Save the weave."""
945
self._check_write_ok()
947
write_weave_v5(self, sio)
949
bytes = sio.getvalue()
950
path = self._weave_name + WeaveFile.WEAVE_SUFFIX
952
self._transport.put_bytes(path, bytes, self._filemode)
953
except errors.NoSuchFile:
954
self._transport.mkdir(dirname(path))
955
self._transport.put_bytes(path, bytes, self._filemode)
959
"""See VersionedFile.get_suffixes()."""
960
return [WeaveFile.WEAVE_SUFFIX]
962
def insert_record_stream(self, stream):
963
super(WeaveFile, self).insert_record_stream(stream)
966
@deprecated_method(one_five)
967
def join(self, other, pb=None, msg=None, version_ids=None,
968
ignore_missing=False):
969
"""Join other into self and save."""
970
super(WeaveFile, self).join(other, pb, msg, version_ids, ignore_missing)
974
def _reweave(wa, wb, pb=None, msg=None):
975
"""Combine two weaves and return the result.
977
This works even if a revision R has different parents in
978
wa and wb. In the resulting weave all the parents are given.
980
This is done by just building up a new weave, maintaining ordering
981
of the versions in the two inputs. More efficient approaches
982
might be possible but it should only be necessary to do
983
this operation rarely, when a new previously ghost version is
986
:param pb: An optional progress bar, indicating how far done we are
987
:param msg: An optional message for the progress
991
queue_a = range(wa.num_versions())
992
queue_b = range(wb.num_versions())
993
# first determine combined parents of all versions
994
# map from version name -> all parent names
995
combined_parents = _reweave_parent_graphs(wa, wb)
996
mutter("combined parents: %r", combined_parents)
997
order = topo_sort(combined_parents.iteritems())
998
mutter("order to reweave: %r", order)
1003
for idx, name in enumerate(order):
1005
pb.update(msg, idx, len(order))
1006
if name in wa._name_map:
1007
lines = wa.get_lines(name)
1008
if name in wb._name_map:
1009
lines_b = wb.get_lines(name)
1010
if lines != lines_b:
1011
mutter('Weaves differ on content. rev_id {%s}', name)
1012
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
1014
lines = list(difflib.unified_diff(lines, lines_b,
1015
wa._weave_name, wb._weave_name))
1016
mutter('lines:\n%s', ''.join(lines))
1017
raise errors.WeaveTextDiffers(name, wa, wb)
1019
lines = wb.get_lines(name)
1020
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
1023
def _reweave_parent_graphs(wa, wb):
1024
"""Return combined parent ancestry for two weaves.
1026
Returned as a list of (version_name, set(parent_names))"""
1028
for weave in [wa, wb]:
1029
for idx, name in enumerate(weave._names):
1030
p = combined.setdefault(name, set())
1031
p.update(map(weave._idx_to_name, weave._parents[idx]))
1036
"""Show the weave's table-of-contents"""
1037
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
1038
for i in (6, 50, 10, 10):
1041
for i in range(w.num_versions()):
1044
parent_str = ' '.join(map(str, w._parents[i]))
1045
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
1049
def weave_stats(weave_file, pb):
1050
from bzrlib.weavefile import read_weave
1052
wf = file(weave_file, 'rb')
1054
# FIXME: doesn't work on pipes
1055
weave_size = wf.tell()
1059
for i in range(vers):
1060
pb.update('checking sizes', i, vers)
1061
for origin, lineno, line in w._extract([i]):
1066
print 'versions %9d' % vers
1067
print 'weave file %9d bytes' % weave_size
1068
print 'total contents %9d bytes' % total
1069
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1072
print 'average size %9d bytes' % avg
1073
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1077
print """bzr weave tool
1079
Experimental tool for weave algorithm.
1082
weave init WEAVEFILE
1083
Create an empty weave file
1084
weave get WEAVEFILE VERSION
1085
Write out specified version.
1086
weave check WEAVEFILE
1087
Check consistency of all versions.
1089
Display table of contents.
1090
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1091
Add NEWTEXT, with specified parent versions.
1092
weave annotate WEAVEFILE VERSION
1093
Display origin of each line.
1094
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1095
Auto-merge two versions and display conflicts.
1096
weave diff WEAVEFILE VERSION1 VERSION2
1097
Show differences between two versions.
1101
% weave init foo.weave
1103
% weave add foo.weave ver0 < foo.txt
1106
(create updated version)
1108
% weave get foo.weave 0 | diff -u - foo.txt
1109
% weave add foo.weave ver1 0 < foo.txt
1112
% weave get foo.weave 0 > foo.txt (create forked version)
1114
% weave add foo.weave ver2 0 < foo.txt
1117
% weave merge foo.weave 1 2 > foo.txt (merge them)
1118
% vi foo.txt (resolve conflicts)
1119
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1131
# in case we're run directly from the subdirectory
1132
sys.path.append('..')
1134
from bzrlib.weavefile import write_weave, read_weave
1135
from bzrlib.progress import ProgressBar
1150
return read_weave(file(argv[2], 'rb'))
1156
# at the moment, based on everything in the file
1158
parents = map(int, argv[4:])
1159
lines = sys.stdin.readlines()
1160
ver = w.add(name, parents, lines)
1161
write_weave(w, file(argv[2], 'wb'))
1162
print 'added version %r %d' % (name, ver)
1165
if os.path.exists(fn):
1166
raise IOError("file exists")
1168
write_weave(w, file(fn, 'wb'))
1169
elif cmd == 'get': # get one version
1171
sys.stdout.writelines(w.get_iter(int(argv[3])))
1176
v1, v2 = map(int, argv[3:5])
1179
diff_gen = bzrlib.patiencediff.unified_diff(lines1, lines2,
1180
'%s version %d' % (fn, v1),
1181
'%s version %d' % (fn, v2))
1182
sys.stdout.writelines(diff_gen)
1184
elif cmd == 'annotate':
1186
# newline is added to all lines regardless; too hard to get
1187
# reasonable formatting otherwise
1189
for origin, text in w.annotate(int(argv[3])):
1190
text = text.rstrip('\r\n')
1192
print ' | %s' % (text)
1194
print '%5d | %s' % (origin, text)
1200
elif cmd == 'stats':
1201
weave_stats(argv[2], ProgressBar())
1203
elif cmd == 'check':
1208
print '%d versions ok' % w.num_versions()
1210
elif cmd == 'inclusions':
1212
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1214
elif cmd == 'parents':
1216
print ' '.join(map(str, w._parents[int(argv[3])]))
1218
elif cmd == 'plan-merge':
1219
# replaced by 'bzr weave-plan-merge'
1221
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1223
print '%14s | %s' % (state, line),
1224
elif cmd == 'merge':
1225
# replaced by 'bzr weave-merge-text'
1227
p = w.plan_merge(int(argv[3]), int(argv[4]))
1228
sys.stdout.writelines(w.weave_merge(p))
1230
raise ValueError('unknown command %r' % cmd)
1233
if __name__ == '__main__':
1235
sys.exit(main(sys.argv))
added
removed
bzrlib/weave.py
