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
77
from bzrlib.lazy_import import lazy_import
78
lazy_import(globals(), """
79
from bzrlib import tsort
84
from bzrlib.errors import (WeaveError, WeaveFormatError, WeaveParentMismatch,
85
RevisionAlreadyPresent,
87
UnavailableRepresentation,
88
WeaveRevisionAlreadyPresent,
89
WeaveRevisionNotPresent,
91
import bzrlib.errors as errors
92
from bzrlib.osutils import dirname, sha, sha_strings, split_lines
93
import bzrlib.patiencediff
94
from bzrlib.revision import NULL_REVISION
95
from bzrlib.symbol_versioning import *
96
from bzrlib.trace import mutter
97
from bzrlib.versionedfile import (
103
from bzrlib.weavefile import _read_weave_v5, write_weave_v5
106
class WeaveContentFactory(ContentFactory):
107
"""Content factory for streaming from weaves.
109
:seealso ContentFactory:
112
def __init__(self, version, weave):
113
"""Create a WeaveContentFactory for version from weave."""
114
ContentFactory.__init__(self)
115
self.sha1 = weave.get_sha1s([version])[version]
116
self.key = (version,)
117
parents = weave.get_parent_map([version])[version]
118
self.parents = tuple((parent,) for parent in parents)
119
self.storage_kind = 'fulltext'
122
def get_bytes_as(self, storage_kind):
123
if storage_kind == 'fulltext':
124
return self._weave.get_text(self.key[-1])
125
elif storage_kind == 'chunked':
126
return self._weave.get_lines(self.key[-1])
128
raise UnavailableRepresentation(self.key, storage_kind, 'fulltext')
131
class Weave(VersionedFile):
132
"""weave - versioned text file storage.
134
A Weave manages versions of line-based text files, keeping track
135
of the originating version for each line.
137
To clients the "lines" of the file are represented as a list of strings.
138
These strings will typically have terminal newline characters, but
139
this is not required. In particular files commonly do not have a newline
140
at the end of the file.
142
Texts can be identified in either of two ways:
144
* a nonnegative index number.
146
* a version-id string.
148
Typically the index number will be valid only inside this weave and
149
the version-id is used to reference it in the larger world.
151
The weave is represented as a list mixing edit instructions and
152
literal text. Each entry in _weave can be either a string (or
153
unicode), or a tuple. If a string, it means that the given line
154
should be output in the currently active revisions.
156
If a tuple, it gives a processing instruction saying in which
157
revisions the enclosed lines are active. The tuple has the form
158
(instruction, version).
160
The instruction can be '{' or '}' for an insertion block, and '['
161
and ']' for a deletion block respectively. The version is the
162
integer version index. There is no replace operator, only deletes
163
and inserts. For '}', the end of an insertion, there is no
164
version parameter because it always closes the most recently
169
* A later version can delete lines that were introduced by any
170
number of ancestor versions; this implies that deletion
171
instructions can span insertion blocks without regard to the
172
insertion block's nesting.
174
* Similarly, deletions need not be properly nested with regard to
175
each other, because they might have been generated by
176
independent revisions.
178
* Insertions are always made by inserting a new bracketed block
179
into a single point in the previous weave. This implies they
180
can nest but not overlap, and the nesting must always have later
181
insertions on the inside.
183
* It doesn't seem very useful to have an active insertion
184
inside an inactive insertion, but it might happen.
186
* Therefore, all instructions are always"considered"; that
187
is passed onto and off the stack. An outer inactive block
188
doesn't disable an inner block.
190
* Lines are enabled if the most recent enclosing insertion is
191
active and none of the enclosing deletions are active.
193
* There is no point having a deletion directly inside its own
194
insertion; you might as well just not write it. And there
195
should be no way to get an earlier version deleting a later
199
Text of the weave; list of control instruction tuples and strings.
202
List of parents, indexed by version number.
203
It is only necessary to store the minimal set of parents for
204
each version; the parent's parents are implied.
207
List of hex SHA-1 of each version.
210
List of symbolic names for each version. Each should be unique.
213
For each name, the version number.
216
Descriptive name of this weave; typically the filename if known.
220
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
221
'_weave_name', '_matcher', '_allow_reserved']
223
def __init__(self, weave_name=None, access_mode='w', matcher=None,
224
get_scope=None, allow_reserved=False):
227
:param get_scope: A callable that returns an opaque object to be used
228
for detecting when this weave goes out of scope (should stop
229
answering requests or allowing mutation).
231
super(Weave, self).__init__()
237
self._weave_name = weave_name
239
self._matcher = bzrlib.patiencediff.PatienceSequenceMatcher
241
self._matcher = matcher
242
if get_scope is None:
243
get_scope = lambda:None
244
self._get_scope = get_scope
245
self._scope = get_scope()
246
self._access_mode = access_mode
247
self._allow_reserved = allow_reserved
250
return "Weave(%r)" % self._weave_name
252
def _check_write_ok(self):
253
"""Is the versioned file marked as 'finished' ? Raise if it is."""
254
if self._get_scope() != self._scope:
255
raise errors.OutSideTransaction()
256
if self._access_mode != 'w':
257
raise errors.ReadOnlyObjectDirtiedError(self)
260
"""Return a deep copy of self.
262
The copy can be modified without affecting the original weave."""
264
other._weave = self._weave[:]
265
other._parents = self._parents[:]
266
other._sha1s = self._sha1s[:]
267
other._names = self._names[:]
268
other._name_map = self._name_map.copy()
269
other._weave_name = self._weave_name
272
def __eq__(self, other):
273
if not isinstance(other, Weave):
275
return self._parents == other._parents \
276
and self._weave == other._weave \
277
and self._sha1s == other._sha1s
279
def __ne__(self, other):
280
return not self.__eq__(other)
282
def _idx_to_name(self, version):
283
return self._names[version]
285
def _lookup(self, name):
286
"""Convert symbolic version name to index."""
287
if not self._allow_reserved:
288
self.check_not_reserved_id(name)
290
return self._name_map[name]
292
raise RevisionNotPresent(name, self._weave_name)
295
"""See VersionedFile.versions."""
296
return self._names[:]
298
def has_version(self, version_id):
299
"""See VersionedFile.has_version."""
300
return (version_id in self._name_map)
302
__contains__ = has_version
304
def get_record_stream(self, versions, ordering, include_delta_closure):
305
"""Get a stream of records for versions.
307
:param versions: The versions to include. Each version is a tuple
309
:param ordering: Either 'unordered' or 'topological'. A topologically
310
sorted stream has compression parents strictly before their
312
:param include_delta_closure: If True then the closure across any
313
compression parents will be included (in the opaque data).
314
:return: An iterator of ContentFactory objects, each of which is only
315
valid until the iterator is advanced.
317
versions = [version[-1] for version in versions]
318
if ordering == 'topological':
319
parents = self.get_parent_map(versions)
320
new_versions = tsort.topo_sort(parents)
321
new_versions.extend(set(versions).difference(set(parents)))
322
versions = new_versions
323
for version in versions:
325
yield WeaveContentFactory(version, self)
327
yield AbsentContentFactory((version,))
329
def get_parent_map(self, version_ids):
330
"""See VersionedFile.get_parent_map."""
332
for version_id in version_ids:
333
if version_id == NULL_REVISION:
338
map(self._idx_to_name,
339
self._parents[self._lookup(version_id)]))
340
except RevisionNotPresent:
342
result[version_id] = parents
345
def get_parents_with_ghosts(self, version_id):
346
raise NotImplementedError(self.get_parents_with_ghosts)
348
def insert_record_stream(self, stream):
349
"""Insert a record stream into this versioned file.
351
:param stream: A stream of records to insert.
353
:seealso VersionedFile.get_record_stream:
356
for record in stream:
357
# Raise an error when a record is missing.
358
if record.storage_kind == 'absent':
359
raise RevisionNotPresent([record.key[0]], self)
360
# adapt to non-tuple interface
361
parents = [parent[0] for parent in record.parents]
362
if (record.storage_kind == 'fulltext'
363
or record.storage_kind == 'chunked'):
364
self.add_lines(record.key[0], parents,
365
split_lines(record.get_bytes_as('fulltext')))
367
adapter_key = record.storage_kind, 'fulltext'
369
adapter = adapters[adapter_key]
371
adapter_factory = adapter_registry.get(adapter_key)
372
adapter = adapter_factory(self)
373
adapters[adapter_key] = adapter
374
lines = split_lines(adapter.get_bytes(
375
record, record.get_bytes_as(record.storage_kind)))
377
self.add_lines(record.key[0], parents, lines)
378
except RevisionAlreadyPresent:
381
def _check_repeated_add(self, name, parents, text, sha1):
382
"""Check that a duplicated add is OK.
384
If it is, return the (old) index; otherwise raise an exception.
386
idx = self._lookup(name)
387
if sorted(self._parents[idx]) != sorted(parents) \
388
or sha1 != self._sha1s[idx]:
389
raise RevisionAlreadyPresent(name, self._weave_name)
392
def _add_lines(self, version_id, parents, lines, parent_texts,
393
left_matching_blocks, nostore_sha, random_id, check_content):
394
"""See VersionedFile.add_lines."""
395
idx = self._add(version_id, lines, map(self._lookup, parents),
396
nostore_sha=nostore_sha)
397
return sha_strings(lines), sum(map(len, lines)), idx
399
def _add(self, version_id, lines, parents, sha1=None, nostore_sha=None):
400
"""Add a single text on top of the weave.
402
Returns the index number of the newly added version.
405
Symbolic name for this version.
406
(Typically the revision-id of the revision that added it.)
409
List or set of direct parent version numbers.
412
Sequence of lines to be added in the new version.
414
:param nostore_sha: See VersionedFile.add_lines.
416
self._check_lines_not_unicode(lines)
417
self._check_lines_are_lines(lines)
419
sha1 = sha_strings(lines)
420
if sha1 == nostore_sha:
421
raise errors.ExistingContent
422
if version_id in self._name_map:
423
return self._check_repeated_add(version_id, parents, lines, sha1)
425
self._check_versions(parents)
426
## self._check_lines(lines)
427
new_version = len(self._parents)
429
# if we abort after here the (in-memory) weave will be corrupt because only
430
# some fields are updated
431
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
432
# - Robert Collins 20060226
433
self._parents.append(parents[:])
434
self._sha1s.append(sha1)
435
self._names.append(version_id)
436
self._name_map[version_id] = new_version
440
# special case; adding with no parents revision; can do
441
# this more quickly by just appending unconditionally.
442
# even more specially, if we're adding an empty text we
443
# need do nothing at all.
445
self._weave.append(('{', new_version))
446
self._weave.extend(lines)
447
self._weave.append(('}', None))
450
if len(parents) == 1:
451
pv = list(parents)[0]
452
if sha1 == self._sha1s[pv]:
453
# special case: same as the single parent
457
ancestors = self._inclusions(parents)
461
# basis a list of (origin, lineno, line)
464
for origin, lineno, line in self._extract(ancestors):
465
basis_lineno.append(lineno)
466
basis_lines.append(line)
468
# another small special case: a merge, producing the same text
470
if lines == basis_lines:
473
# add a sentinel, because we can also match against the final line
474
basis_lineno.append(len(self._weave))
476
# XXX: which line of the weave should we really consider
477
# matches the end of the file? the current code says it's the
478
# last line of the weave?
480
#print 'basis_lines:', basis_lines
481
#print 'new_lines: ', lines
483
s = self._matcher(None, basis_lines, lines)
485
# offset gives the number of lines that have been inserted
486
# into the weave up to the current point; if the original edit instruction
487
# says to change line A then we actually change (A+offset)
490
for tag, i1, i2, j1, j2 in s.get_opcodes():
491
# i1,i2 are given in offsets within basis_lines; we need to map them
492
# back to offsets within the entire weave
493
#print 'raw match', tag, i1, i2, j1, j2
496
i1 = basis_lineno[i1]
497
i2 = basis_lineno[i2]
498
# the deletion and insertion are handled separately.
499
# first delete the region.
501
self._weave.insert(i1+offset, ('[', new_version))
502
self._weave.insert(i2+offset+1, (']', new_version))
506
# there may have been a deletion spanning up to
507
# i2; we want to insert after this region to make sure
508
# we don't destroy ourselves
510
self._weave[i:i] = ([('{', new_version)]
513
offset += 2 + (j2 - j1)
516
def _inclusions(self, versions):
517
"""Return set of all ancestors of given version(s)."""
518
if not len(versions):
521
for v in xrange(max(versions), 0, -1):
523
# include all its parents
524
i.update(self._parents[v])
526
## except IndexError:
527
## raise ValueError("version %d not present in weave" % v)
529
def get_ancestry(self, version_ids, topo_sorted=True):
530
"""See VersionedFile.get_ancestry."""
531
if isinstance(version_ids, basestring):
532
version_ids = [version_ids]
533
i = self._inclusions([self._lookup(v) for v in version_ids])
534
return [self._idx_to_name(v) for v in i]
536
def _check_lines(self, text):
537
if not isinstance(text, list):
538
raise ValueError("text should be a list, not %s" % type(text))
541
if not isinstance(l, basestring):
542
raise ValueError("text line should be a string or unicode, not %s"
547
def _check_versions(self, indexes):
548
"""Check everything in the sequence of indexes is valid"""
553
raise IndexError("invalid version number %r" % i)
555
def _compatible_parents(self, my_parents, other_parents):
556
"""During join check that other_parents are joinable with my_parents.
558
Joinable is defined as 'is a subset of' - supersets may require
559
regeneration of diffs, but subsets do not.
561
return len(other_parents.difference(my_parents)) == 0
563
def annotate(self, version_id):
564
"""Return a list of (version-id, line) tuples for version_id.
566
The index indicates when the line originated in the weave."""
567
incls = [self._lookup(version_id)]
568
return [(self._idx_to_name(origin), text) for origin, lineno, text in
569
self._extract(incls)]
571
def iter_lines_added_or_present_in_versions(self, version_ids=None,
573
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
574
if version_ids is None:
575
version_ids = self.versions()
576
version_ids = set(version_ids)
577
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
578
# if inserted not in version_ids then it was inserted before the
579
# versions we care about, but because weaves cannot represent ghosts
580
# properly, we do not filter down to that
581
# if inserted not in version_ids: continue
583
yield line + '\n', inserted
587
def _walk_internal(self, version_ids=None):
588
"""Helper method for weave actions."""
593
lineno = 0 # line of weave, 0-based
595
for l in self._weave:
596
if l.__class__ == tuple:
600
istack.append(self._names[v])
604
dset.add(self._names[v])
606
dset.remove(self._names[v])
608
raise WeaveFormatError('unexpected instruction %r' % v)
610
yield lineno, istack[-1], frozenset(dset), l
614
raise WeaveFormatError("unclosed insertion blocks "
615
"at end of weave: %s" % istack)
617
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
620
def plan_merge(self, ver_a, ver_b):
621
"""Return pseudo-annotation indicating how the two versions merge.
623
This is computed between versions a and b and their common
626
Weave lines present in none of them are skipped entirely.
628
inc_a = set(self.get_ancestry([ver_a]))
629
inc_b = set(self.get_ancestry([ver_b]))
630
inc_c = inc_a & inc_b
632
for lineno, insert, deleteset, line in self._walk_internal([ver_a, ver_b]):
633
if deleteset & inc_c:
634
# killed in parent; can't be in either a or b
635
# not relevant to our work
636
yield 'killed-base', line
637
elif insert in inc_c:
638
# was inserted in base
639
killed_a = bool(deleteset & inc_a)
640
killed_b = bool(deleteset & inc_b)
641
if killed_a and killed_b:
642
yield 'killed-both', line
644
yield 'killed-a', line
646
yield 'killed-b', line
648
yield 'unchanged', line
649
elif insert in inc_a:
650
if deleteset & inc_a:
651
yield 'ghost-a', line
655
elif insert in inc_b:
656
if deleteset & inc_b:
657
yield 'ghost-b', line
661
# not in either revision
662
yield 'irrelevant', line
664
def _extract(self, versions):
665
"""Yield annotation of lines in included set.
667
Yields a sequence of tuples (origin, lineno, text), where
668
origin is the origin version, lineno the index in the weave,
669
and text the text of the line.
671
The set typically but not necessarily corresponds to a version.
674
if not isinstance(i, int):
677
included = self._inclusions(versions)
683
lineno = 0 # line of weave, 0-based
689
WFE = WeaveFormatError
692
# 449 0 4474.6820 2356.5590 bzrlib.weave:556(_extract)
693
# +285282 0 1676.8040 1676.8040 +<isinstance>
694
# 1.6 seconds in 'isinstance'.
695
# changing the first isinstance:
696
# 449 0 2814.2660 1577.1760 bzrlib.weave:556(_extract)
697
# +140414 0 762.8050 762.8050 +<isinstance>
698
# note that the inline time actually dropped (less function calls)
699
# and total processing time was halved.
700
# we're still spending ~1/4 of the method in isinstance though.
701
# so lets hard code the acceptable string classes we expect:
702
# 449 0 1202.9420 786.2930 bzrlib.weave:556(_extract)
703
# +71352 0 377.5560 377.5560 +<method 'append' of 'list'
705
# yay, down to ~1/4 the initial extract time, and our inline time
706
# has shrunk again, with isinstance no longer dominating.
707
# tweaking the stack inclusion test to use a set gives:
708
# 449 0 1122.8030 713.0080 bzrlib.weave:556(_extract)
709
# +71352 0 354.9980 354.9980 +<method 'append' of 'list'
711
# - a 5% win, or possibly just noise. However with large istacks that
712
# 'in' test could dominate, so I'm leaving this change in place -
713
# when its fast enough to consider profiling big datasets we can review.
718
for l in self._weave:
719
if l.__class__ == tuple:
726
iset.remove(istack.pop())
734
raise AssertionError()
737
isactive = (not dset) and istack and (istack[-1] in included)
739
result.append((istack[-1], lineno, l))
742
raise WeaveFormatError("unclosed insertion blocks "
743
"at end of weave: %s" % istack)
745
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
749
def _maybe_lookup(self, name_or_index):
750
"""Convert possible symbolic name to index, or pass through indexes.
754
if isinstance(name_or_index, (int, long)):
757
return self._lookup(name_or_index)
759
def get_lines(self, version_id):
760
"""See VersionedFile.get_lines()."""
761
int_index = self._maybe_lookup(version_id)
762
result = [line for (origin, lineno, line) in self._extract([int_index])]
763
expected_sha1 = self._sha1s[int_index]
764
measured_sha1 = sha_strings(result)
765
if measured_sha1 != expected_sha1:
766
raise errors.WeaveInvalidChecksum(
767
'file %s, revision %s, expected: %s, measured %s'
768
% (self._weave_name, version_id,
769
expected_sha1, measured_sha1))
772
def get_sha1s(self, version_ids):
773
"""See VersionedFile.get_sha1s()."""
775
for v in version_ids:
776
result[v] = self._sha1s[self._lookup(v)]
779
def num_versions(self):
780
"""How many versions are in this weave?"""
781
l = len(self._parents)
784
__len__ = num_versions
786
def check(self, progress_bar=None):
787
# TODO evaluate performance hit of using string sets in this routine.
788
# TODO: check no circular inclusions
789
# TODO: create a nested progress bar
790
for version in range(self.num_versions()):
791
inclusions = list(self._parents[version])
794
if inclusions[-1] >= version:
795
raise WeaveFormatError("invalid included version %d for index %d"
796
% (inclusions[-1], version))
798
# try extracting all versions; parallel extraction is used
799
nv = self.num_versions()
804
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
805
# The problem is that set membership is much more expensive
806
name = self._idx_to_name(i)
809
new_inc = set([name])
810
for p in self._parents[i]:
811
new_inc.update(inclusions[self._idx_to_name(p)])
813
if set(new_inc) != set(self.get_ancestry(name)):
814
raise AssertionError(
816
% (set(new_inc), set(self.get_ancestry(name))))
817
inclusions[name] = new_inc
819
nlines = len(self._weave)
821
update_text = 'checking weave'
823
short_name = os.path.basename(self._weave_name)
824
update_text = 'checking %s' % (short_name,)
825
update_text = update_text[:25]
827
for lineno, insert, deleteset, line in self._walk_internal():
829
progress_bar.update(update_text, lineno, nlines)
831
for name, name_inclusions in inclusions.items():
832
# The active inclusion must be an ancestor,
833
# and no ancestors must have deleted this line,
834
# because we don't support resurrection.
835
if (insert in name_inclusions) and not (deleteset & name_inclusions):
836
sha1s[name].update(line)
839
version = self._idx_to_name(i)
840
hd = sha1s[version].hexdigest()
841
expected = self._sha1s[i]
843
raise errors.WeaveInvalidChecksum(
844
"mismatched sha1 for version %s: "
845
"got %s, expected %s"
846
% (version, hd, expected))
848
# TODO: check insertions are properly nested, that there are
849
# no lines outside of insertion blocks, that deletions are
850
# properly paired, etc.
852
def _imported_parents(self, other, other_idx):
853
"""Return list of parents in self corresponding to indexes in other."""
855
for parent_idx in other._parents[other_idx]:
856
parent_name = other._names[parent_idx]
857
if parent_name not in self._name_map:
858
# should not be possible
859
raise WeaveError("missing parent {%s} of {%s} in %r"
860
% (parent_name, other._name_map[other_idx], self))
861
new_parents.append(self._name_map[parent_name])
864
def _check_version_consistent(self, other, other_idx, name):
865
"""Check if a version in consistent in this and other.
867
To be consistent it must have:
870
* the same direct parents (by name, not index, and disregarding
873
If present & correct return True;
874
if not present in self return False;
875
if inconsistent raise error."""
876
this_idx = self._name_map.get(name, -1)
878
if self._sha1s[this_idx] != other._sha1s[other_idx]:
879
raise errors.WeaveTextDiffers(name, self, other)
880
self_parents = self._parents[this_idx]
881
other_parents = other._parents[other_idx]
882
n1 = set([self._names[i] for i in self_parents])
883
n2 = set([other._names[i] for i in other_parents])
884
if not self._compatible_parents(n1, n2):
885
raise WeaveParentMismatch("inconsistent parents "
886
"for version {%s}: %s vs %s" % (name, n1, n2))
892
def _reweave(self, other, pb, msg):
893
"""Reweave self with other - internal helper for join().
895
:param other: The other weave to merge
896
:param pb: An optional progress bar, indicating how far done we are
897
:param msg: An optional message for the progress
899
new_weave = _reweave(self, other, pb=pb, msg=msg)
900
self._copy_weave_content(new_weave)
902
def _copy_weave_content(self, otherweave):
903
"""adsorb the content from otherweave."""
904
for attr in self.__slots__:
905
if attr != '_weave_name':
906
setattr(self, attr, copy(getattr(otherweave, attr)))
909
class WeaveFile(Weave):
910
"""A WeaveFile represents a Weave on disk and writes on change."""
912
WEAVE_SUFFIX = '.weave'
914
def __init__(self, name, transport, filemode=None, create=False, access_mode='w', get_scope=None):
915
"""Create a WeaveFile.
917
:param create: If not True, only open an existing knit.
919
super(WeaveFile, self).__init__(name, access_mode, get_scope=get_scope,
920
allow_reserved=False)
921
self._transport = transport
922
self._filemode = filemode
924
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
925
except errors.NoSuchFile:
931
def _add_lines(self, version_id, parents, lines, parent_texts,
932
left_matching_blocks, nostore_sha, random_id, check_content):
933
"""Add a version and save the weave."""
934
self.check_not_reserved_id(version_id)
935
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
936
parent_texts, left_matching_blocks, nostore_sha, random_id,
941
def copy_to(self, name, transport):
942
"""See VersionedFile.copy_to()."""
943
# as we are all in memory always, just serialise to the new place.
945
write_weave_v5(self, sio)
947
transport.put_file(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
950
"""Save the weave."""
951
self._check_write_ok()
953
write_weave_v5(self, sio)
955
bytes = sio.getvalue()
956
path = self._weave_name + WeaveFile.WEAVE_SUFFIX
958
self._transport.put_bytes(path, bytes, self._filemode)
959
except errors.NoSuchFile:
960
self._transport.mkdir(dirname(path))
961
self._transport.put_bytes(path, bytes, self._filemode)
965
"""See VersionedFile.get_suffixes()."""
966
return [WeaveFile.WEAVE_SUFFIX]
968
def insert_record_stream(self, stream):
969
super(WeaveFile, self).insert_record_stream(stream)
972
@deprecated_method(one_five)
973
def join(self, other, pb=None, msg=None, version_ids=None,
974
ignore_missing=False):
975
"""Join other into self and save."""
976
super(WeaveFile, self).join(other, pb, msg, version_ids, ignore_missing)
980
def _reweave(wa, wb, pb=None, msg=None):
981
"""Combine two weaves and return the result.
983
This works even if a revision R has different parents in
984
wa and wb. In the resulting weave all the parents are given.
986
This is done by just building up a new weave, maintaining ordering
987
of the versions in the two inputs. More efficient approaches
988
might be possible but it should only be necessary to do
989
this operation rarely, when a new previously ghost version is
992
:param pb: An optional progress bar, indicating how far done we are
993
:param msg: An optional message for the progress
997
queue_a = range(wa.num_versions())
998
queue_b = range(wb.num_versions())
999
# first determine combined parents of all versions
1000
# map from version name -> all parent names
1001
combined_parents = _reweave_parent_graphs(wa, wb)
1002
mutter("combined parents: %r", combined_parents)
1003
order = tsort.topo_sort(combined_parents.iteritems())
1004
mutter("order to reweave: %r", order)
1009
for idx, name in enumerate(order):
1011
pb.update(msg, idx, len(order))
1012
if name in wa._name_map:
1013
lines = wa.get_lines(name)
1014
if name in wb._name_map:
1015
lines_b = wb.get_lines(name)
1016
if lines != lines_b:
1017
mutter('Weaves differ on content. rev_id {%s}', name)
1018
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
1020
lines = list(difflib.unified_diff(lines, lines_b,
1021
wa._weave_name, wb._weave_name))
1022
mutter('lines:\n%s', ''.join(lines))
1023
raise errors.WeaveTextDiffers(name, wa, wb)
1025
lines = wb.get_lines(name)
1026
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
1029
def _reweave_parent_graphs(wa, wb):
1030
"""Return combined parent ancestry for two weaves.
1032
Returned as a list of (version_name, set(parent_names))"""
1034
for weave in [wa, wb]:
1035
for idx, name in enumerate(weave._names):
1036
p = combined.setdefault(name, set())
1037
p.update(map(weave._idx_to_name, weave._parents[idx]))
1042
"""Show the weave's table-of-contents"""
1043
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
1044
for i in (6, 50, 10, 10):
1047
for i in range(w.num_versions()):
1050
parent_str = ' '.join(map(str, w._parents[i]))
1051
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
1055
def weave_stats(weave_file, pb):
1056
from bzrlib.weavefile import read_weave
1058
wf = file(weave_file, 'rb')
1060
# FIXME: doesn't work on pipes
1061
weave_size = wf.tell()
1065
for i in range(vers):
1066
pb.update('checking sizes', i, vers)
1067
for origin, lineno, line in w._extract([i]):
1072
print 'versions %9d' % vers
1073
print 'weave file %9d bytes' % weave_size
1074
print 'total contents %9d bytes' % total
1075
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1078
print 'average size %9d bytes' % avg
1079
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1083
print """bzr weave tool
1085
Experimental tool for weave algorithm.
1088
weave init WEAVEFILE
1089
Create an empty weave file
1090
weave get WEAVEFILE VERSION
1091
Write out specified version.
1092
weave check WEAVEFILE
1093
Check consistency of all versions.
1095
Display table of contents.
1096
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1097
Add NEWTEXT, with specified parent versions.
1098
weave annotate WEAVEFILE VERSION
1099
Display origin of each line.
1100
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1101
Auto-merge two versions and display conflicts.
1102
weave diff WEAVEFILE VERSION1 VERSION2
1103
Show differences between two versions.
1107
% weave init foo.weave
1109
% weave add foo.weave ver0 < foo.txt
1112
(create updated version)
1114
% weave get foo.weave 0 | diff -u - foo.txt
1115
% weave add foo.weave ver1 0 < foo.txt
1118
% weave get foo.weave 0 > foo.txt (create forked version)
1120
% weave add foo.weave ver2 0 < foo.txt
1123
% weave merge foo.weave 1 2 > foo.txt (merge them)
1124
% vi foo.txt (resolve conflicts)
1125
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1137
# in case we're run directly from the subdirectory
1138
sys.path.append('..')
1140
from bzrlib.weavefile import write_weave, read_weave
1141
from bzrlib.progress import ProgressBar
1156
return read_weave(file(argv[2], 'rb'))
1162
# at the moment, based on everything in the file
1164
parents = map(int, argv[4:])
1165
lines = sys.stdin.readlines()
1166
ver = w.add(name, parents, lines)
1167
write_weave(w, file(argv[2], 'wb'))
1168
print 'added version %r %d' % (name, ver)
1171
if os.path.exists(fn):
1172
raise IOError("file exists")
1174
write_weave(w, file(fn, 'wb'))
1175
elif cmd == 'get': # get one version
1177
sys.stdout.writelines(w.get_iter(int(argv[3])))
1182
v1, v2 = map(int, argv[3:5])
1185
diff_gen = bzrlib.patiencediff.unified_diff(lines1, lines2,
1186
'%s version %d' % (fn, v1),
1187
'%s version %d' % (fn, v2))
1188
sys.stdout.writelines(diff_gen)
1190
elif cmd == 'annotate':
1192
# newline is added to all lines regardless; too hard to get
1193
# reasonable formatting otherwise
1195
for origin, text in w.annotate(int(argv[3])):
1196
text = text.rstrip('\r\n')
1198
print ' | %s' % (text)
1200
print '%5d | %s' % (origin, text)
1206
elif cmd == 'stats':
1207
weave_stats(argv[2], ProgressBar())
1209
elif cmd == 'check':
1214
print '%d versions ok' % w.num_versions()
1216
elif cmd == 'inclusions':
1218
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1220
elif cmd == 'parents':
1222
print ' '.join(map(str, w._parents[int(argv[3])]))
1224
elif cmd == 'plan-merge':
1225
# replaced by 'bzr weave-plan-merge'
1227
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1229
print '%14s | %s' % (state, line),
1230
elif cmd == 'merge':
1231
# replaced by 'bzr weave-merge-text'
1233
p = w.plan_merge(int(argv[3]), int(argv[4]))
1234
sys.stdout.writelines(w.weave_merge(p))
1236
raise ValueError('unknown command %r' % cmd)
1239
if __name__ == '__main__':
1241
sys.exit(main(sys.argv))
added
removed
bzrlib/weave.py
