3
# Copyright (C) 2005 Canonical Ltd
5
# This program is free software; you can redistribute it and/or modify
6
# it under the terms of the GNU General Public License as published by
7
# the Free Software Foundation; either version 2 of the License, or
8
# (at your option) any later version.
10
# This program is distributed in the hope that it will be useful,
11
# but WITHOUT ANY WARRANTY; without even the implied warranty of
12
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13
# GNU General Public License for more details.
15
# You should have received a copy of the GNU General Public License
16
# along with this program; if not, write to the Free Software
17
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
19
# Author: Martin Pool <mbp@canonical.com>
22
"""Weave - storage of related text file versions"""
25
# XXX: If we do weaves this way, will a merge still behave the same
26
# way if it's done in a different order? That's a pretty desirable
29
# TODO: Nothing here so far assumes the lines are really \n newlines,
30
# rather than being split up in some other way. We could accommodate
31
# binaries, perhaps by naively splitting on \n or perhaps using
32
# something like a rolling checksum.
34
# TODO: End marker for each version so we can stop reading?
36
# TODO: Check that no insertion occurs inside a deletion that was
37
# active in the version of the insertion.
39
# TODO: In addition to the SHA-1 check, perhaps have some code that
40
# checks structural constraints of the weave: ie that insertions are
41
# properly nested, that there is no text outside of an insertion, that
42
# insertions or deletions are not repeated, etc.
44
# TODO: Parallel-extract that passes back each line along with a
45
# description of which revisions include it. Nice for checking all
46
# shas or calculating stats in parallel.
48
# TODO: Using a single _extract routine and then processing the output
49
# is probably inefficient. It's simple enough that we can afford to
50
# have slight specializations for different ways its used: annotate,
51
# basis for add, get, etc.
53
# TODO: Probably the API should work only in names to hide the integer
54
# indexes from the user.
56
# TODO: Is there any potential performance win by having an add()
57
# variant that is passed a pre-cooked version of the single basis
60
# TODO: Reweave can possibly be made faster by remembering diffs
61
# where the basis and destination are unchanged.
63
# FIXME: Sometimes we will be given a parents list for a revision
64
# that includes some redundant parents (i.e. already a parent of
65
# something in the list.) We should eliminate them. This can
66
# be done fairly efficiently because the sequence numbers constrain
67
# the possible relationships.
69
# FIXME: the conflict markers should be *7* characters
72
from cStringIO import StringIO
81
from bzrlib.errors import (WeaveError, WeaveFormatError, WeaveParentMismatch,
82
RevisionAlreadyPresent,
84
UnavailableRepresentation,
85
WeaveRevisionAlreadyPresent,
86
WeaveRevisionNotPresent,
88
import bzrlib.errors as errors
89
from bzrlib.osutils import sha_strings, split_lines
90
import bzrlib.patiencediff
91
from bzrlib.trace import mutter
92
from bzrlib.tsort import topo_sort
93
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])[0]
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[0])
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']
218
def __init__(self, weave_name=None, access_mode='w', matcher=None, get_scope=None):
221
:param get_scope: A callable that returns an opaque object to be used
222
for detecting when this weave goes out of scope (should stop
223
answering requests or allowing mutation).
225
super(Weave, self).__init__(access_mode)
231
self._weave_name = weave_name
233
self._matcher = bzrlib.patiencediff.PatienceSequenceMatcher
235
self._matcher = matcher
236
if get_scope is None:
237
get_scope = lambda:None
238
self._get_scope = get_scope
239
self._scope = get_scope()
240
self._access_mode = access_mode
243
return "Weave(%r)" % self._weave_name
245
def _check_write_ok(self):
246
"""Is the versioned file marked as 'finished' ? Raise if it is."""
247
if self._get_scope() != self._scope:
248
raise errors.OutSideTransaction()
249
if self._access_mode != 'w':
250
raise errors.ReadOnlyObjectDirtiedError(self)
253
"""Return a deep copy of self.
255
The copy can be modified without affecting the original weave."""
257
other._weave = self._weave[:]
258
other._parents = self._parents[:]
259
other._sha1s = self._sha1s[:]
260
other._names = self._names[:]
261
other._name_map = self._name_map.copy()
262
other._weave_name = self._weave_name
265
def __eq__(self, other):
266
if not isinstance(other, Weave):
268
return self._parents == other._parents \
269
and self._weave == other._weave \
270
and self._sha1s == other._sha1s
272
def __ne__(self, other):
273
return not self.__eq__(other)
275
def _idx_to_name(self, version):
276
return self._names[version]
278
def _lookup(self, name):
279
"""Convert symbolic version name to index."""
280
self.check_not_reserved_id(name)
282
return self._name_map[name]
284
raise RevisionNotPresent(name, self._weave_name)
287
"""See VersionedFile.versions."""
288
return self._names[:]
290
def has_version(self, version_id):
291
"""See VersionedFile.has_version."""
292
return (version_id in self._name_map)
294
__contains__ = has_version
296
def get_record_stream(self, versions, ordering, include_delta_closure):
297
"""Get a stream of records for versions.
299
:param versions: The versions to include. Each version is a tuple
301
:param ordering: Either 'unordered' or 'topological'. A topologically
302
sorted stream has compression parents strictly before their
304
:param include_delta_closure: If True then the closure across any
305
compression parents will be included (in the opaque data).
306
:return: An iterator of ContentFactory objects, each of which is only
307
valid until the iterator is advanced.
309
if ordering == 'topological':
310
parents = self.get_parent_map(versions)
311
versions = topo_sort(parents)
312
for version in versions:
314
yield WeaveContentFactory(version, self)
316
yield AbsentContentFactory((version,))
318
def get_parent_map(self, version_ids):
319
"""See VersionedFile.get_parent_map."""
321
for version_id in version_ids:
323
result[version_id] = tuple(
324
map(self._idx_to_name, self._parents[self._lookup(version_id)]))
325
except RevisionNotPresent:
329
def get_parents_with_ghosts(self, version_id):
330
raise NotImplementedError(self.get_parents_with_ghosts)
332
def insert_record_stream(self, stream):
333
"""Insert a record stream into this versioned file.
335
:param stream: A stream of records to insert.
337
:seealso VersionedFile.get_record_stream:
340
for record in stream:
341
# adapt to non-tuple interface
342
parents = [parent[0] for parent in record.parents]
343
if record.storage_kind == 'fulltext':
344
self.add_lines(record.key[0], parents,
345
split_lines(record.get_bytes_as('fulltext')))
347
adapter_key = record.storage_kind, 'fulltext'
349
adapter = adapters[adapter_key]
351
adapter_factory = adapter_registry.get(adapter_key)
352
adapter = adapter_factory(self)
353
adapters[adapter_key] = adapter
354
lines = split_lines(adapter.get_bytes(
355
record, record.get_bytes_as(record.storage_kind)))
357
self.add_lines(record.key[0], parents, lines)
358
except RevisionAlreadyPresent:
361
def _check_repeated_add(self, name, parents, text, sha1):
362
"""Check that a duplicated add is OK.
364
If it is, return the (old) index; otherwise raise an exception.
366
idx = self._lookup(name)
367
if sorted(self._parents[idx]) != sorted(parents) \
368
or sha1 != self._sha1s[idx]:
369
raise RevisionAlreadyPresent(name, self._weave_name)
372
def _add_lines(self, version_id, parents, lines, parent_texts,
373
left_matching_blocks, nostore_sha, random_id, check_content):
374
"""See VersionedFile.add_lines."""
375
idx = self._add(version_id, lines, map(self._lookup, parents),
376
nostore_sha=nostore_sha)
377
return sha_strings(lines), sum(map(len, lines)), idx
379
def _add(self, version_id, lines, parents, sha1=None, nostore_sha=None):
380
"""Add a single text on top of the weave.
382
Returns the index number of the newly added version.
385
Symbolic name for this version.
386
(Typically the revision-id of the revision that added it.)
389
List or set of direct parent version numbers.
392
Sequence of lines to be added in the new version.
394
:param nostore_sha: See VersionedFile.add_lines.
396
assert isinstance(version_id, basestring)
397
self._check_lines_not_unicode(lines)
398
self._check_lines_are_lines(lines)
400
sha1 = sha_strings(lines)
401
if sha1 == nostore_sha:
402
raise errors.ExistingContent
403
if version_id in self._name_map:
404
return self._check_repeated_add(version_id, parents, lines, sha1)
406
self._check_versions(parents)
407
## self._check_lines(lines)
408
new_version = len(self._parents)
410
# if we abort after here the (in-memory) weave will be corrupt because only
411
# some fields are updated
412
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
413
# - Robert Collins 20060226
414
self._parents.append(parents[:])
415
self._sha1s.append(sha1)
416
self._names.append(version_id)
417
self._name_map[version_id] = new_version
421
# special case; adding with no parents revision; can do
422
# this more quickly by just appending unconditionally.
423
# even more specially, if we're adding an empty text we
424
# need do nothing at all.
426
self._weave.append(('{', new_version))
427
self._weave.extend(lines)
428
self._weave.append(('}', None))
431
if len(parents) == 1:
432
pv = list(parents)[0]
433
if sha1 == self._sha1s[pv]:
434
# special case: same as the single parent
438
ancestors = self._inclusions(parents)
442
# basis a list of (origin, lineno, line)
445
for origin, lineno, line in self._extract(ancestors):
446
basis_lineno.append(lineno)
447
basis_lines.append(line)
449
# another small special case: a merge, producing the same text
451
if lines == basis_lines:
454
# add a sentinel, because we can also match against the final line
455
basis_lineno.append(len(self._weave))
457
# XXX: which line of the weave should we really consider
458
# matches the end of the file? the current code says it's the
459
# last line of the weave?
461
#print 'basis_lines:', basis_lines
462
#print 'new_lines: ', lines
464
s = self._matcher(None, basis_lines, lines)
466
# offset gives the number of lines that have been inserted
467
# into the weave up to the current point; if the original edit instruction
468
# says to change line A then we actually change (A+offset)
471
for tag, i1, i2, j1, j2 in s.get_opcodes():
472
# i1,i2 are given in offsets within basis_lines; we need to map them
473
# back to offsets within the entire weave
474
#print 'raw match', tag, i1, i2, j1, j2
478
i1 = basis_lineno[i1]
479
i2 = basis_lineno[i2]
481
assert 0 <= j1 <= j2 <= len(lines)
483
#print tag, i1, i2, j1, j2
485
# the deletion and insertion are handled separately.
486
# first delete the region.
488
self._weave.insert(i1+offset, ('[', new_version))
489
self._weave.insert(i2+offset+1, (']', new_version))
493
# there may have been a deletion spanning up to
494
# i2; we want to insert after this region to make sure
495
# we don't destroy ourselves
497
self._weave[i:i] = ([('{', new_version)]
500
offset += 2 + (j2 - j1)
503
def _inclusions(self, versions):
504
"""Return set of all ancestors of given version(s)."""
505
if not len(versions):
508
for v in xrange(max(versions), 0, -1):
510
# include all its parents
511
i.update(self._parents[v])
513
## except IndexError:
514
## raise ValueError("version %d not present in weave" % v)
516
def get_ancestry(self, version_ids, topo_sorted=True):
517
"""See VersionedFile.get_ancestry."""
518
if isinstance(version_ids, basestring):
519
version_ids = [version_ids]
520
i = self._inclusions([self._lookup(v) for v in version_ids])
521
return [self._idx_to_name(v) for v in i]
523
def _check_lines(self, text):
524
if not isinstance(text, list):
525
raise ValueError("text should be a list, not %s" % type(text))
528
if not isinstance(l, basestring):
529
raise ValueError("text line should be a string or unicode, not %s"
534
def _check_versions(self, indexes):
535
"""Check everything in the sequence of indexes is valid"""
540
raise IndexError("invalid version number %r" % i)
542
def _compatible_parents(self, my_parents, other_parents):
543
"""During join check that other_parents are joinable with my_parents.
545
Joinable is defined as 'is a subset of' - supersets may require
546
regeneration of diffs, but subsets do not.
548
return len(other_parents.difference(my_parents)) == 0
550
def annotate(self, version_id):
551
"""Return a list of (version-id, line) tuples for version_id.
553
The index indicates when the line originated in the weave."""
554
incls = [self._lookup(version_id)]
555
return [(self._idx_to_name(origin), text) for origin, lineno, text in
556
self._extract(incls)]
558
def iter_lines_added_or_present_in_versions(self, version_ids=None,
560
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
561
if version_ids is None:
562
version_ids = self.versions()
563
version_ids = set(version_ids)
564
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
565
# if inserted not in version_ids then it was inserted before the
566
# versions we care about, but because weaves cannot represent ghosts
567
# properly, we do not filter down to that
568
# if inserted not in version_ids: continue
570
yield line + '\n', inserted
574
def _walk_internal(self, version_ids=None):
575
"""Helper method for weave actions."""
580
lineno = 0 # line of weave, 0-based
582
for l in self._weave:
583
if l.__class__ == tuple:
587
istack.append(self._names[v])
591
assert self._names[v] not in dset
592
dset.add(self._names[v])
594
dset.remove(self._names[v])
596
raise WeaveFormatError('unexpected instruction %r' % v)
598
assert l.__class__ in (str, unicode)
600
yield lineno, istack[-1], frozenset(dset), l
604
raise WeaveFormatError("unclosed insertion blocks "
605
"at end of weave: %s" % istack)
607
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
610
def plan_merge(self, ver_a, ver_b):
611
"""Return pseudo-annotation indicating how the two versions merge.
613
This is computed between versions a and b and their common
616
Weave lines present in none of them are skipped entirely.
618
inc_a = set(self.get_ancestry([ver_a]))
619
inc_b = set(self.get_ancestry([ver_b]))
620
inc_c = inc_a & inc_b
622
for lineno, insert, deleteset, line in self._walk_internal([ver_a, ver_b]):
623
if deleteset & inc_c:
624
# killed in parent; can't be in either a or b
625
# not relevant to our work
626
yield 'killed-base', line
627
elif insert in inc_c:
628
# was inserted in base
629
killed_a = bool(deleteset & inc_a)
630
killed_b = bool(deleteset & inc_b)
631
if killed_a and killed_b:
632
yield 'killed-both', line
634
yield 'killed-a', line
636
yield 'killed-b', line
638
yield 'unchanged', line
639
elif insert in inc_a:
640
if deleteset & inc_a:
641
yield 'ghost-a', line
645
elif insert in inc_b:
646
if deleteset & inc_b:
647
yield 'ghost-b', line
651
# not in either revision
652
yield 'irrelevant', line
654
yield 'unchanged', '' # terminator
656
def _extract(self, versions):
657
"""Yield annotation of lines in included set.
659
Yields a sequence of tuples (origin, lineno, text), where
660
origin is the origin version, lineno the index in the weave,
661
and text the text of the line.
663
The set typically but not necessarily corresponds to a version.
666
if not isinstance(i, int):
669
included = self._inclusions(versions)
675
lineno = 0 # line of weave, 0-based
681
WFE = WeaveFormatError
684
# 449 0 4474.6820 2356.5590 bzrlib.weave:556(_extract)
685
# +285282 0 1676.8040 1676.8040 +<isinstance>
686
# 1.6 seconds in 'isinstance'.
687
# changing the first isinstance:
688
# 449 0 2814.2660 1577.1760 bzrlib.weave:556(_extract)
689
# +140414 0 762.8050 762.8050 +<isinstance>
690
# note that the inline time actually dropped (less function calls)
691
# and total processing time was halved.
692
# we're still spending ~1/4 of the method in isinstance though.
693
# so lets hard code the acceptable string classes we expect:
694
# 449 0 1202.9420 786.2930 bzrlib.weave:556(_extract)
695
# +71352 0 377.5560 377.5560 +<method 'append' of 'list'
697
# yay, down to ~1/4 the initial extract time, and our inline time
698
# has shrunk again, with isinstance no longer dominating.
699
# tweaking the stack inclusion test to use a set gives:
700
# 449 0 1122.8030 713.0080 bzrlib.weave:556(_extract)
701
# +71352 0 354.9980 354.9980 +<method 'append' of 'list'
703
# - a 5% win, or possibly just noise. However with large istacks that
704
# 'in' test could dominate, so I'm leaving this change in place -
705
# when its fast enough to consider profiling big datasets we can review.
710
for l in self._weave:
711
if l.__class__ == tuple:
719
iset.remove(istack.pop())
730
assert l.__class__ in (str, unicode)
732
isactive = (not dset) and istack and (istack[-1] in included)
734
result.append((istack[-1], lineno, l))
737
raise WeaveFormatError("unclosed insertion blocks "
738
"at end of weave: %s" % istack)
740
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
744
def _maybe_lookup(self, name_or_index):
745
"""Convert possible symbolic name to index, or pass through indexes.
749
if isinstance(name_or_index, (int, long)):
752
return self._lookup(name_or_index)
754
def get_lines(self, version_id):
755
"""See VersionedFile.get_lines()."""
756
int_index = self._maybe_lookup(version_id)
757
result = [line for (origin, lineno, line) in self._extract([int_index])]
758
expected_sha1 = self._sha1s[int_index]
759
measured_sha1 = sha_strings(result)
760
if measured_sha1 != expected_sha1:
761
raise errors.WeaveInvalidChecksum(
762
'file %s, revision %s, expected: %s, measured %s'
763
% (self._weave_name, version_id,
764
expected_sha1, measured_sha1))
767
def get_sha1s(self, version_ids):
768
"""See VersionedFile.get_sha1s()."""
769
return [self._sha1s[self._lookup(v)] for v in version_ids]
771
def num_versions(self):
772
"""How many versions are in this weave?"""
773
l = len(self._parents)
774
assert l == len(self._sha1s)
777
__len__ = num_versions
779
def check(self, progress_bar=None):
780
# TODO evaluate performance hit of using string sets in this routine.
781
# TODO: check no circular inclusions
782
# TODO: create a nested progress bar
783
for version in range(self.num_versions()):
784
inclusions = list(self._parents[version])
787
if inclusions[-1] >= version:
788
raise WeaveFormatError("invalid included version %d for index %d"
789
% (inclusions[-1], version))
791
# try extracting all versions; parallel extraction is used
792
nv = self.num_versions()
797
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
798
# The problem is that set membership is much more expensive
799
name = self._idx_to_name(i)
800
sha1s[name] = sha.new()
802
new_inc = set([name])
803
for p in self._parents[i]:
804
new_inc.update(inclusions[self._idx_to_name(p)])
806
assert set(new_inc) == set(self.get_ancestry(name)), \
807
'failed %s != %s' % (set(new_inc), set(self.get_ancestry(name)))
808
inclusions[name] = new_inc
810
nlines = len(self._weave)
812
update_text = 'checking weave'
814
short_name = os.path.basename(self._weave_name)
815
update_text = 'checking %s' % (short_name,)
816
update_text = update_text[:25]
818
for lineno, insert, deleteset, line in self._walk_internal():
820
progress_bar.update(update_text, lineno, nlines)
822
for name, name_inclusions in inclusions.items():
823
# The active inclusion must be an ancestor,
824
# and no ancestors must have deleted this line,
825
# because we don't support resurrection.
826
if (insert in name_inclusions) and not (deleteset & name_inclusions):
827
sha1s[name].update(line)
830
version = self._idx_to_name(i)
831
hd = sha1s[version].hexdigest()
832
expected = self._sha1s[i]
834
raise errors.WeaveInvalidChecksum(
835
"mismatched sha1 for version %s: "
836
"got %s, expected %s"
837
% (version, hd, expected))
839
# TODO: check insertions are properly nested, that there are
840
# no lines outside of insertion blocks, that deletions are
841
# properly paired, etc.
843
def _join(self, other, pb, msg, version_ids, ignore_missing):
844
"""Worker routine for join()."""
845
if not other.versions():
846
return # nothing to update, easy
849
# versions is never none, InterWeave checks this.
852
# two loops so that we do not change ourselves before verifying it
854
# work through in index order to make sure we get all dependencies
857
# get the selected versions only that are in other.versions.
858
version_ids = set(other.versions()).intersection(set(version_ids))
859
# pull in the referenced graph.
860
version_ids = other.get_ancestry(version_ids)
861
pending_parents = other.get_parent_map(version_ids)
862
pending_graph = pending_parents.items()
863
if len(pending_graph) != len(version_ids):
864
raise RevisionNotPresent(
865
set(version_ids) - set(pending_parents.keys()), self)
866
for name in topo_sort(pending_graph):
867
other_idx = other._name_map[name]
868
# returns True if we have it, False if we need it.
869
if not self._check_version_consistent(other, other_idx, name):
870
names_to_join.append((other_idx, name))
878
for other_idx, name in names_to_join:
879
# TODO: If all the parents of the other version are already
880
# present then we can avoid some work by just taking the delta
881
# and adjusting the offsets.
882
new_parents = self._imported_parents(other, other_idx)
883
sha1 = other._sha1s[other_idx]
888
pb.update(msg, merged, len(names_to_join))
890
lines = other.get_lines(other_idx)
891
self._add(name, lines, new_parents, sha1)
893
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
894
merged, processed, self._weave_name, time.time()-time0))
896
def _imported_parents(self, other, other_idx):
897
"""Return list of parents in self corresponding to indexes in other."""
899
for parent_idx in other._parents[other_idx]:
900
parent_name = other._names[parent_idx]
901
if parent_name not in self._name_map:
902
# should not be possible
903
raise WeaveError("missing parent {%s} of {%s} in %r"
904
% (parent_name, other._name_map[other_idx], self))
905
new_parents.append(self._name_map[parent_name])
908
def _check_version_consistent(self, other, other_idx, name):
909
"""Check if a version in consistent in this and other.
911
To be consistent it must have:
914
* the same direct parents (by name, not index, and disregarding
917
If present & correct return True;
918
if not present in self return False;
919
if inconsistent raise error."""
920
this_idx = self._name_map.get(name, -1)
922
if self._sha1s[this_idx] != other._sha1s[other_idx]:
923
raise errors.WeaveTextDiffers(name, self, other)
924
self_parents = self._parents[this_idx]
925
other_parents = other._parents[other_idx]
926
n1 = set([self._names[i] for i in self_parents])
927
n2 = set([other._names[i] for i in other_parents])
928
if not self._compatible_parents(n1, n2):
929
raise WeaveParentMismatch("inconsistent parents "
930
"for version {%s}: %s vs %s" % (name, n1, n2))
936
def _reweave(self, other, pb, msg):
937
"""Reweave self with other - internal helper for join().
939
:param other: The other weave to merge
940
:param pb: An optional progress bar, indicating how far done we are
941
:param msg: An optional message for the progress
943
new_weave = _reweave(self, other, pb=pb, msg=msg)
944
self._copy_weave_content(new_weave)
946
def _copy_weave_content(self, otherweave):
947
"""adsorb the content from otherweave."""
948
for attr in self.__slots__:
949
if attr != '_weave_name':
950
setattr(self, attr, copy(getattr(otherweave, attr)))
953
class WeaveFile(Weave):
954
"""A WeaveFile represents a Weave on disk and writes on change."""
956
WEAVE_SUFFIX = '.weave'
958
def __init__(self, name, transport, filemode=None, create=False, access_mode='w', get_scope=None):
959
"""Create a WeaveFile.
961
:param create: If not True, only open an existing knit.
963
super(WeaveFile, self).__init__(name, access_mode, get_scope=get_scope)
964
self._transport = transport
965
self._filemode = filemode
967
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
968
except errors.NoSuchFile:
974
def _add_lines(self, version_id, parents, lines, parent_texts,
975
left_matching_blocks, nostore_sha, random_id, check_content):
976
"""Add a version and save the weave."""
977
self.check_not_reserved_id(version_id)
978
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
979
parent_texts, left_matching_blocks, nostore_sha, random_id,
984
def copy_to(self, name, transport):
985
"""See VersionedFile.copy_to()."""
986
# as we are all in memory always, just serialise to the new place.
988
write_weave_v5(self, sio)
990
transport.put_file(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
993
"""Save the weave."""
994
self._check_write_ok()
996
write_weave_v5(self, sio)
998
self._transport.put_file(self._weave_name + WeaveFile.WEAVE_SUFFIX,
1004
"""See VersionedFile.get_suffixes()."""
1005
return [WeaveFile.WEAVE_SUFFIX]
1007
def join(self, other, pb=None, msg=None, version_ids=None,
1008
ignore_missing=False):
1009
"""Join other into self and save."""
1010
super(WeaveFile, self).join(other, pb, msg, version_ids, ignore_missing)
1014
def _reweave(wa, wb, pb=None, msg=None):
1015
"""Combine two weaves and return the result.
1017
This works even if a revision R has different parents in
1018
wa and wb. In the resulting weave all the parents are given.
1020
This is done by just building up a new weave, maintaining ordering
1021
of the versions in the two inputs. More efficient approaches
1022
might be possible but it should only be necessary to do
1023
this operation rarely, when a new previously ghost version is
1026
:param pb: An optional progress bar, indicating how far done we are
1027
:param msg: An optional message for the progress
1031
queue_a = range(wa.num_versions())
1032
queue_b = range(wb.num_versions())
1033
# first determine combined parents of all versions
1034
# map from version name -> all parent names
1035
combined_parents = _reweave_parent_graphs(wa, wb)
1036
mutter("combined parents: %r", combined_parents)
1037
order = topo_sort(combined_parents.iteritems())
1038
mutter("order to reweave: %r", order)
1043
for idx, name in enumerate(order):
1045
pb.update(msg, idx, len(order))
1046
if name in wa._name_map:
1047
lines = wa.get_lines(name)
1048
if name in wb._name_map:
1049
lines_b = wb.get_lines(name)
1050
if lines != lines_b:
1051
mutter('Weaves differ on content. rev_id {%s}', name)
1052
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
1054
lines = list(difflib.unified_diff(lines, lines_b,
1055
wa._weave_name, wb._weave_name))
1056
mutter('lines:\n%s', ''.join(lines))
1057
raise errors.WeaveTextDiffers(name, wa, wb)
1059
lines = wb.get_lines(name)
1060
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
1063
def _reweave_parent_graphs(wa, wb):
1064
"""Return combined parent ancestry for two weaves.
1066
Returned as a list of (version_name, set(parent_names))"""
1068
for weave in [wa, wb]:
1069
for idx, name in enumerate(weave._names):
1070
p = combined.setdefault(name, set())
1071
p.update(map(weave._idx_to_name, weave._parents[idx]))
1076
"""Show the weave's table-of-contents"""
1077
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
1078
for i in (6, 50, 10, 10):
1081
for i in range(w.num_versions()):
1084
parent_str = ' '.join(map(str, w._parents[i]))
1085
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
1089
def weave_stats(weave_file, pb):
1090
from bzrlib.weavefile import read_weave
1092
wf = file(weave_file, 'rb')
1094
# FIXME: doesn't work on pipes
1095
weave_size = wf.tell()
1099
for i in range(vers):
1100
pb.update('checking sizes', i, vers)
1101
for origin, lineno, line in w._extract([i]):
1106
print 'versions %9d' % vers
1107
print 'weave file %9d bytes' % weave_size
1108
print 'total contents %9d bytes' % total
1109
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1112
print 'average size %9d bytes' % avg
1113
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1117
print """bzr weave tool
1119
Experimental tool for weave algorithm.
1122
weave init WEAVEFILE
1123
Create an empty weave file
1124
weave get WEAVEFILE VERSION
1125
Write out specified version.
1126
weave check WEAVEFILE
1127
Check consistency of all versions.
1129
Display table of contents.
1130
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1131
Add NEWTEXT, with specified parent versions.
1132
weave annotate WEAVEFILE VERSION
1133
Display origin of each line.
1134
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1135
Auto-merge two versions and display conflicts.
1136
weave diff WEAVEFILE VERSION1 VERSION2
1137
Show differences between two versions.
1141
% weave init foo.weave
1143
% weave add foo.weave ver0 < foo.txt
1146
(create updated version)
1148
% weave get foo.weave 0 | diff -u - foo.txt
1149
% weave add foo.weave ver1 0 < foo.txt
1152
% weave get foo.weave 0 > foo.txt (create forked version)
1154
% weave add foo.weave ver2 0 < foo.txt
1157
% weave merge foo.weave 1 2 > foo.txt (merge them)
1158
% vi foo.txt (resolve conflicts)
1159
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1171
# in case we're run directly from the subdirectory
1172
sys.path.append('..')
1174
from bzrlib.weavefile import write_weave, read_weave
1175
from bzrlib.progress import ProgressBar
1190
return read_weave(file(argv[2], 'rb'))
1196
# at the moment, based on everything in the file
1198
parents = map(int, argv[4:])
1199
lines = sys.stdin.readlines()
1200
ver = w.add(name, parents, lines)
1201
write_weave(w, file(argv[2], 'wb'))
1202
print 'added version %r %d' % (name, ver)
1205
if os.path.exists(fn):
1206
raise IOError("file exists")
1208
write_weave(w, file(fn, 'wb'))
1209
elif cmd == 'get': # get one version
1211
sys.stdout.writelines(w.get_iter(int(argv[3])))
1216
v1, v2 = map(int, argv[3:5])
1219
diff_gen = bzrlib.patiencediff.unified_diff(lines1, lines2,
1220
'%s version %d' % (fn, v1),
1221
'%s version %d' % (fn, v2))
1222
sys.stdout.writelines(diff_gen)
1224
elif cmd == 'annotate':
1226
# newline is added to all lines regardless; too hard to get
1227
# reasonable formatting otherwise
1229
for origin, text in w.annotate(int(argv[3])):
1230
text = text.rstrip('\r\n')
1232
print ' | %s' % (text)
1234
print '%5d | %s' % (origin, text)
1240
elif cmd == 'stats':
1241
weave_stats(argv[2], ProgressBar())
1243
elif cmd == 'check':
1248
print '%d versions ok' % w.num_versions()
1250
elif cmd == 'inclusions':
1252
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1254
elif cmd == 'parents':
1256
print ' '.join(map(str, w._parents[int(argv[3])]))
1258
elif cmd == 'plan-merge':
1259
# replaced by 'bzr weave-plan-merge'
1261
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1263
print '%14s | %s' % (state, line),
1264
elif cmd == 'merge':
1265
# replaced by 'bzr weave-merge-text'
1267
p = w.plan_merge(int(argv[3]), int(argv[4]))
1268
sys.stdout.writelines(w.weave_merge(p))
1270
raise ValueError('unknown command %r' % cmd)
1273
if __name__ == '__main__':
1275
sys.exit(main(sys.argv))
1278
class InterWeave(InterVersionedFile):
1279
"""Optimised code paths for weave to weave operations."""
1281
_matching_file_from_factory = staticmethod(WeaveFile)
1282
_matching_file_to_factory = staticmethod(WeaveFile)
1285
def is_compatible(source, target):
1286
"""Be compatible with weaves."""
1288
return (isinstance(source, Weave) and
1289
isinstance(target, Weave))
1290
except AttributeError:
1293
def join(self, pb=None, msg=None, version_ids=None, ignore_missing=False):
1294
"""See InterVersionedFile.join."""
1295
version_ids = self._get_source_version_ids(version_ids, ignore_missing)
1296
if self.target.versions() == [] and version_ids is None:
1297
self.target._copy_weave_content(self.source)
1299
self.target._join(self.source, pb, msg, version_ids, ignore_missing)
1302
InterVersionedFile.register_optimiser(InterWeave)
added
removed
bzrlib/weave.py
