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.trace import mutter
82
from bzrlib.errors import (WeaveError, WeaveFormatError, WeaveParentMismatch,
83
RevisionAlreadyPresent,
85
WeaveRevisionAlreadyPresent,
86
WeaveRevisionNotPresent,
88
import bzrlib.errors as errors
89
from bzrlib.osutils import sha_strings
90
import bzrlib.patiencediff
91
from bzrlib.tsort import topo_sort
92
from bzrlib.versionedfile import VersionedFile, InterVersionedFile
93
from bzrlib.weavefile import _read_weave_v5, write_weave_v5
96
class Weave(VersionedFile):
97
"""weave - versioned text file storage.
99
A Weave manages versions of line-based text files, keeping track
100
of the originating version for each line.
102
To clients the "lines" of the file are represented as a list of strings.
103
These strings will typically have terminal newline characters, but
104
this is not required. In particular files commonly do not have a newline
105
at the end of the file.
107
Texts can be identified in either of two ways:
109
* a nonnegative index number.
111
* a version-id string.
113
Typically the index number will be valid only inside this weave and
114
the version-id is used to reference it in the larger world.
116
The weave is represented as a list mixing edit instructions and
117
literal text. Each entry in _weave can be either a string (or
118
unicode), or a tuple. If a string, it means that the given line
119
should be output in the currently active revisions.
121
If a tuple, it gives a processing instruction saying in which
122
revisions the enclosed lines are active. The tuple has the form
123
(instruction, version).
125
The instruction can be '{' or '}' for an insertion block, and '['
126
and ']' for a deletion block respectively. The version is the
127
integer version index. There is no replace operator, only deletes
128
and inserts. For '}', the end of an insertion, there is no
129
version parameter because it always closes the most recently
134
* A later version can delete lines that were introduced by any
135
number of ancestor versions; this implies that deletion
136
instructions can span insertion blocks without regard to the
137
insertion block's nesting.
139
* Similarly, deletions need not be properly nested with regard to
140
each other, because they might have been generated by
141
independent revisions.
143
* Insertions are always made by inserting a new bracketed block
144
into a single point in the previous weave. This implies they
145
can nest but not overlap, and the nesting must always have later
146
insertions on the inside.
148
* It doesn't seem very useful to have an active insertion
149
inside an inactive insertion, but it might happen.
151
* Therefore, all instructions are always"considered"; that
152
is passed onto and off the stack. An outer inactive block
153
doesn't disable an inner block.
155
* Lines are enabled if the most recent enclosing insertion is
156
active and none of the enclosing deletions are active.
158
* There is no point having a deletion directly inside its own
159
insertion; you might as well just not write it. And there
160
should be no way to get an earlier version deleting a later
164
Text of the weave; list of control instruction tuples and strings.
167
List of parents, indexed by version number.
168
It is only necessary to store the minimal set of parents for
169
each version; the parent's parents are implied.
172
List of hex SHA-1 of each version.
175
List of symbolic names for each version. Each should be unique.
178
For each name, the version number.
181
Descriptive name of this weave; typically the filename if known.
185
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
186
'_weave_name', '_matcher']
188
def __init__(self, weave_name=None, access_mode='w', matcher=None):
189
super(Weave, self).__init__(access_mode)
195
self._weave_name = weave_name
197
self._matcher = bzrlib.patiencediff.PatienceSequenceMatcher
199
self._matcher = matcher
202
return "Weave(%r)" % self._weave_name
205
"""Return a deep copy of self.
207
The copy can be modified without affecting the original weave."""
209
other._weave = self._weave[:]
210
other._parents = self._parents[:]
211
other._sha1s = self._sha1s[:]
212
other._names = self._names[:]
213
other._name_map = self._name_map.copy()
214
other._weave_name = self._weave_name
217
def __eq__(self, other):
218
if not isinstance(other, Weave):
220
return self._parents == other._parents \
221
and self._weave == other._weave \
222
and self._sha1s == other._sha1s
224
def __ne__(self, other):
225
return not self.__eq__(other)
227
def _idx_to_name(self, version):
228
return self._names[version]
230
def _lookup(self, name):
231
"""Convert symbolic version name to index."""
232
self.check_not_reserved_id(name)
234
return self._name_map[name]
236
raise RevisionNotPresent(name, self._weave_name)
239
"""See VersionedFile.versions."""
240
return self._names[:]
242
def has_version(self, version_id):
243
"""See VersionedFile.has_version."""
244
return (version_id in self._name_map)
246
__contains__ = has_version
248
def get_parents(self, version_id):
249
"""See VersionedFile.get_parent."""
250
return map(self._idx_to_name, self._parents[self._lookup(version_id)])
252
def _check_repeated_add(self, name, parents, text, sha1):
253
"""Check that a duplicated add is OK.
255
If it is, return the (old) index; otherwise raise an exception.
257
idx = self._lookup(name)
258
if sorted(self._parents[idx]) != sorted(parents) \
259
or sha1 != self._sha1s[idx]:
260
raise RevisionAlreadyPresent(name, self._weave_name)
263
def _add_lines(self, version_id, parents, lines, parent_texts,
264
left_matching_blocks, nostore_sha, random_id, check_content):
265
"""See VersionedFile.add_lines."""
266
idx = self._add(version_id, lines, map(self._lookup, parents),
267
nostore_sha=nostore_sha)
268
return sha_strings(lines), sum(map(len, lines)), idx
270
def _add(self, version_id, lines, parents, sha1=None, nostore_sha=None):
271
"""Add a single text on top of the weave.
273
Returns the index number of the newly added version.
276
Symbolic name for this version.
277
(Typically the revision-id of the revision that added it.)
280
List or set of direct parent version numbers.
283
Sequence of lines to be added in the new version.
285
:param nostore_sha: See VersionedFile.add_lines.
287
assert isinstance(version_id, basestring)
288
self._check_lines_not_unicode(lines)
289
self._check_lines_are_lines(lines)
291
sha1 = sha_strings(lines)
292
if sha1 == nostore_sha:
293
raise errors.ExistingContent
294
if version_id in self._name_map:
295
return self._check_repeated_add(version_id, parents, lines, sha1)
297
self._check_versions(parents)
298
## self._check_lines(lines)
299
new_version = len(self._parents)
301
# if we abort after here the (in-memory) weave will be corrupt because only
302
# some fields are updated
303
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
304
# - Robert Collins 20060226
305
self._parents.append(parents[:])
306
self._sha1s.append(sha1)
307
self._names.append(version_id)
308
self._name_map[version_id] = new_version
312
# special case; adding with no parents revision; can do
313
# this more quickly by just appending unconditionally.
314
# even more specially, if we're adding an empty text we
315
# need do nothing at all.
317
self._weave.append(('{', new_version))
318
self._weave.extend(lines)
319
self._weave.append(('}', None))
322
if len(parents) == 1:
323
pv = list(parents)[0]
324
if sha1 == self._sha1s[pv]:
325
# special case: same as the single parent
329
ancestors = self._inclusions(parents)
333
# basis a list of (origin, lineno, line)
336
for origin, lineno, line in self._extract(ancestors):
337
basis_lineno.append(lineno)
338
basis_lines.append(line)
340
# another small special case: a merge, producing the same text
342
if lines == basis_lines:
345
# add a sentinel, because we can also match against the final line
346
basis_lineno.append(len(self._weave))
348
# XXX: which line of the weave should we really consider
349
# matches the end of the file? the current code says it's the
350
# last line of the weave?
352
#print 'basis_lines:', basis_lines
353
#print 'new_lines: ', lines
355
s = self._matcher(None, basis_lines, lines)
357
# offset gives the number of lines that have been inserted
358
# into the weave up to the current point; if the original edit instruction
359
# says to change line A then we actually change (A+offset)
362
for tag, i1, i2, j1, j2 in s.get_opcodes():
363
# i1,i2 are given in offsets within basis_lines; we need to map them
364
# back to offsets within the entire weave
365
#print 'raw match', tag, i1, i2, j1, j2
369
i1 = basis_lineno[i1]
370
i2 = basis_lineno[i2]
372
assert 0 <= j1 <= j2 <= len(lines)
374
#print tag, i1, i2, j1, j2
376
# the deletion and insertion are handled separately.
377
# first delete the region.
379
self._weave.insert(i1+offset, ('[', new_version))
380
self._weave.insert(i2+offset+1, (']', new_version))
384
# there may have been a deletion spanning up to
385
# i2; we want to insert after this region to make sure
386
# we don't destroy ourselves
388
self._weave[i:i] = ([('{', new_version)]
391
offset += 2 + (j2 - j1)
394
def _clone_text(self, new_version_id, old_version_id, parents):
395
"""See VersionedFile.clone_text."""
396
old_lines = self.get_text(old_version_id)
397
self.add_lines(new_version_id, parents, old_lines)
399
def _inclusions(self, versions):
400
"""Return set of all ancestors of given version(s)."""
401
if not len(versions):
404
for v in xrange(max(versions), 0, -1):
406
# include all its parents
407
i.update(self._parents[v])
409
## except IndexError:
410
## raise ValueError("version %d not present in weave" % v)
412
def get_ancestry(self, version_ids, topo_sorted=True):
413
"""See VersionedFile.get_ancestry."""
414
if isinstance(version_ids, basestring):
415
version_ids = [version_ids]
416
i = self._inclusions([self._lookup(v) for v in version_ids])
417
return [self._idx_to_name(v) for v in i]
419
def _check_lines(self, text):
420
if not isinstance(text, list):
421
raise ValueError("text should be a list, not %s" % type(text))
424
if not isinstance(l, basestring):
425
raise ValueError("text line should be a string or unicode, not %s"
430
def _check_versions(self, indexes):
431
"""Check everything in the sequence of indexes is valid"""
436
raise IndexError("invalid version number %r" % i)
438
def _compatible_parents(self, my_parents, other_parents):
439
"""During join check that other_parents are joinable with my_parents.
441
Joinable is defined as 'is a subset of' - supersets may require
442
regeneration of diffs, but subsets do not.
444
return len(other_parents.difference(my_parents)) == 0
446
def annotate_iter(self, version_id):
447
"""Yield list of (version-id, line) pairs for the specified version.
449
The index indicates when the line originated in the weave."""
450
incls = [self._lookup(version_id)]
451
for origin, lineno, text in self._extract(incls):
452
yield self._idx_to_name(origin), text
454
def iter_lines_added_or_present_in_versions(self, version_ids=None,
456
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
457
if version_ids is None:
458
version_ids = self.versions()
459
version_ids = set(version_ids)
460
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
461
# if inserted not in version_ids then it was inserted before the
462
# versions we care about, but because weaves cannot represent ghosts
463
# properly, we do not filter down to that
464
# if inserted not in version_ids: continue
470
def _walk_internal(self, version_ids=None):
471
"""Helper method for weave actions."""
476
lineno = 0 # line of weave, 0-based
478
for l in self._weave:
479
if l.__class__ == tuple:
483
istack.append(self._names[v])
487
assert self._names[v] not in dset
488
dset.add(self._names[v])
490
dset.remove(self._names[v])
492
raise WeaveFormatError('unexpected instruction %r' % v)
494
assert l.__class__ in (str, unicode)
496
yield lineno, istack[-1], frozenset(dset), l
500
raise WeaveFormatError("unclosed insertion blocks "
501
"at end of weave: %s" % istack)
503
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
506
def plan_merge(self, ver_a, ver_b):
507
"""Return pseudo-annotation indicating how the two versions merge.
509
This is computed between versions a and b and their common
512
Weave lines present in none of them are skipped entirely.
514
inc_a = set(self.get_ancestry([ver_a]))
515
inc_b = set(self.get_ancestry([ver_b]))
516
inc_c = inc_a & inc_b
518
for lineno, insert, deleteset, line in self._walk_internal([ver_a, ver_b]):
519
if deleteset & inc_c:
520
# killed in parent; can't be in either a or b
521
# not relevant to our work
522
yield 'killed-base', line
523
elif insert in inc_c:
524
# was inserted in base
525
killed_a = bool(deleteset & inc_a)
526
killed_b = bool(deleteset & inc_b)
527
if killed_a and killed_b:
528
yield 'killed-both', line
530
yield 'killed-a', line
532
yield 'killed-b', line
534
yield 'unchanged', line
535
elif insert in inc_a:
536
if deleteset & inc_a:
537
yield 'ghost-a', line
541
elif insert in inc_b:
542
if deleteset & inc_b:
543
yield 'ghost-b', line
547
# not in either revision
548
yield 'irrelevant', line
550
yield 'unchanged', '' # terminator
552
def _extract(self, versions):
553
"""Yield annotation of lines in included set.
555
Yields a sequence of tuples (origin, lineno, text), where
556
origin is the origin version, lineno the index in the weave,
557
and text the text of the line.
559
The set typically but not necessarily corresponds to a version.
562
if not isinstance(i, int):
565
included = self._inclusions(versions)
571
lineno = 0 # line of weave, 0-based
577
WFE = WeaveFormatError
580
# 449 0 4474.6820 2356.5590 bzrlib.weave:556(_extract)
581
# +285282 0 1676.8040 1676.8040 +<isinstance>
582
# 1.6 seconds in 'isinstance'.
583
# changing the first isinstance:
584
# 449 0 2814.2660 1577.1760 bzrlib.weave:556(_extract)
585
# +140414 0 762.8050 762.8050 +<isinstance>
586
# note that the inline time actually dropped (less function calls)
587
# and total processing time was halved.
588
# we're still spending ~1/4 of the method in isinstance though.
589
# so lets hard code the acceptable string classes we expect:
590
# 449 0 1202.9420 786.2930 bzrlib.weave:556(_extract)
591
# +71352 0 377.5560 377.5560 +<method 'append' of 'list'
593
# yay, down to ~1/4 the initial extract time, and our inline time
594
# has shrunk again, with isinstance no longer dominating.
595
# tweaking the stack inclusion test to use a set gives:
596
# 449 0 1122.8030 713.0080 bzrlib.weave:556(_extract)
597
# +71352 0 354.9980 354.9980 +<method 'append' of 'list'
599
# - a 5% win, or possibly just noise. However with large istacks that
600
# 'in' test could dominate, so I'm leaving this change in place -
601
# when its fast enough to consider profiling big datasets we can review.
606
for l in self._weave:
607
if l.__class__ == tuple:
615
iset.remove(istack.pop())
626
assert l.__class__ in (str, unicode)
628
isactive = (not dset) and istack and (istack[-1] in included)
630
result.append((istack[-1], lineno, l))
633
raise WeaveFormatError("unclosed insertion blocks "
634
"at end of weave: %s" % istack)
636
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
640
def _maybe_lookup(self, name_or_index):
641
"""Convert possible symbolic name to index, or pass through indexes.
645
if isinstance(name_or_index, (int, long)):
648
return self._lookup(name_or_index)
650
def get_lines(self, version_id):
651
"""See VersionedFile.get_lines()."""
652
int_index = self._maybe_lookup(version_id)
653
result = [line for (origin, lineno, line) in self._extract([int_index])]
654
expected_sha1 = self._sha1s[int_index]
655
measured_sha1 = sha_strings(result)
656
if measured_sha1 != expected_sha1:
657
raise errors.WeaveInvalidChecksum(
658
'file %s, revision %s, expected: %s, measured %s'
659
% (self._weave_name, version_id,
660
expected_sha1, measured_sha1))
663
def get_sha1(self, version_id):
664
"""See VersionedFile.get_sha1()."""
665
return self._sha1s[self._lookup(version_id)]
667
def get_sha1s(self, version_ids):
668
"""See VersionedFile.get_sha1s()."""
669
return [self._sha1s[self._lookup(v)] for v in version_ids]
671
def num_versions(self):
672
"""How many versions are in this weave?"""
673
l = len(self._parents)
674
assert l == len(self._sha1s)
677
__len__ = num_versions
679
def check(self, progress_bar=None):
680
# TODO evaluate performance hit of using string sets in this routine.
681
# TODO: check no circular inclusions
682
# TODO: create a nested progress bar
683
for version in range(self.num_versions()):
684
inclusions = list(self._parents[version])
687
if inclusions[-1] >= version:
688
raise WeaveFormatError("invalid included version %d for index %d"
689
% (inclusions[-1], version))
691
# try extracting all versions; parallel extraction is used
692
nv = self.num_versions()
697
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
698
# The problem is that set membership is much more expensive
699
name = self._idx_to_name(i)
700
sha1s[name] = sha.new()
702
new_inc = set([name])
703
for p in self._parents[i]:
704
new_inc.update(inclusions[self._idx_to_name(p)])
706
assert set(new_inc) == set(self.get_ancestry(name)), \
707
'failed %s != %s' % (set(new_inc), set(self.get_ancestry(name)))
708
inclusions[name] = new_inc
710
nlines = len(self._weave)
712
update_text = 'checking weave'
714
short_name = os.path.basename(self._weave_name)
715
update_text = 'checking %s' % (short_name,)
716
update_text = update_text[:25]
718
for lineno, insert, deleteset, line in self._walk_internal():
720
progress_bar.update(update_text, lineno, nlines)
722
for name, name_inclusions in inclusions.items():
723
# The active inclusion must be an ancestor,
724
# and no ancestors must have deleted this line,
725
# because we don't support resurrection.
726
if (insert in name_inclusions) and not (deleteset & name_inclusions):
727
sha1s[name].update(line)
730
version = self._idx_to_name(i)
731
hd = sha1s[version].hexdigest()
732
expected = self._sha1s[i]
734
raise errors.WeaveInvalidChecksum(
735
"mismatched sha1 for version %s: "
736
"got %s, expected %s"
737
% (version, hd, expected))
739
# TODO: check insertions are properly nested, that there are
740
# no lines outside of insertion blocks, that deletions are
741
# properly paired, etc.
743
def _join(self, other, pb, msg, version_ids, ignore_missing):
744
"""Worker routine for join()."""
745
if not other.versions():
746
return # nothing to update, easy
749
# versions is never none, InterWeave checks this.
752
# two loops so that we do not change ourselves before verifying it
754
# work through in index order to make sure we get all dependencies
757
# get the selected versions only that are in other.versions.
758
version_ids = set(other.versions()).intersection(set(version_ids))
759
# pull in the referenced graph.
760
version_ids = other.get_ancestry(version_ids)
761
pending_graph = [(version, other.get_parents(version)) for
762
version in version_ids]
763
for name in topo_sort(pending_graph):
764
other_idx = other._name_map[name]
765
# returns True if we have it, False if we need it.
766
if not self._check_version_consistent(other, other_idx, name):
767
names_to_join.append((other_idx, name))
776
for other_idx, name in names_to_join:
777
# TODO: If all the parents of the other version are already
778
# present then we can avoid some work by just taking the delta
779
# and adjusting the offsets.
780
new_parents = self._imported_parents(other, other_idx)
781
sha1 = other._sha1s[other_idx]
786
pb.update(msg, merged, len(names_to_join))
788
lines = other.get_lines(other_idx)
789
self._add(name, lines, new_parents, sha1)
791
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
792
merged, processed, self._weave_name, time.time()-time0))
794
def _imported_parents(self, other, other_idx):
795
"""Return list of parents in self corresponding to indexes in other."""
797
for parent_idx in other._parents[other_idx]:
798
parent_name = other._names[parent_idx]
799
if parent_name not in self._name_map:
800
# should not be possible
801
raise WeaveError("missing parent {%s} of {%s} in %r"
802
% (parent_name, other._name_map[other_idx], self))
803
new_parents.append(self._name_map[parent_name])
806
def _check_version_consistent(self, other, other_idx, name):
807
"""Check if a version in consistent in this and other.
809
To be consistent it must have:
812
* the same direct parents (by name, not index, and disregarding
815
If present & correct return True;
816
if not present in self return False;
817
if inconsistent raise error."""
818
this_idx = self._name_map.get(name, -1)
820
if self._sha1s[this_idx] != other._sha1s[other_idx]:
821
raise errors.WeaveTextDiffers(name, self, other)
822
self_parents = self._parents[this_idx]
823
other_parents = other._parents[other_idx]
824
n1 = set([self._names[i] for i in self_parents])
825
n2 = set([other._names[i] for i in other_parents])
826
if not self._compatible_parents(n1, n2):
827
raise WeaveParentMismatch("inconsistent parents "
828
"for version {%s}: %s vs %s" % (name, n1, n2))
834
def _reweave(self, other, pb, msg):
835
"""Reweave self with other - internal helper for join().
837
:param other: The other weave to merge
838
:param pb: An optional progress bar, indicating how far done we are
839
:param msg: An optional message for the progress
841
new_weave = _reweave(self, other, pb=pb, msg=msg)
842
self._copy_weave_content(new_weave)
844
def _copy_weave_content(self, otherweave):
845
"""adsorb the content from otherweave."""
846
for attr in self.__slots__:
847
if attr != '_weave_name':
848
setattr(self, attr, copy(getattr(otherweave, attr)))
851
class WeaveFile(Weave):
852
"""A WeaveFile represents a Weave on disk and writes on change."""
854
WEAVE_SUFFIX = '.weave'
856
def __init__(self, name, transport, filemode=None, create=False, access_mode='w'):
857
"""Create a WeaveFile.
859
:param create: If not True, only open an existing knit.
861
super(WeaveFile, self).__init__(name, access_mode)
862
self._transport = transport
863
self._filemode = filemode
865
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
866
except errors.NoSuchFile:
872
def _add_lines(self, version_id, parents, lines, parent_texts,
873
left_matching_blocks, nostore_sha, random_id, check_content):
874
"""Add a version and save the weave."""
875
self.check_not_reserved_id(version_id)
876
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
877
parent_texts, left_matching_blocks, nostore_sha, random_id,
882
def _clone_text(self, new_version_id, old_version_id, parents):
883
"""See VersionedFile.clone_text."""
884
super(WeaveFile, self)._clone_text(new_version_id, old_version_id, parents)
887
def copy_to(self, name, transport):
888
"""See VersionedFile.copy_to()."""
889
# as we are all in memory always, just serialise to the new place.
891
write_weave_v5(self, sio)
893
transport.put_file(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
895
def create_empty(self, name, transport, filemode=None):
896
return WeaveFile(name, transport, filemode, create=True)
899
"""Save the weave."""
900
self._check_write_ok()
902
write_weave_v5(self, sio)
904
self._transport.put_file(self._weave_name + WeaveFile.WEAVE_SUFFIX,
910
"""See VersionedFile.get_suffixes()."""
911
return [WeaveFile.WEAVE_SUFFIX]
913
def join(self, other, pb=None, msg=None, version_ids=None,
914
ignore_missing=False):
915
"""Join other into self and save."""
916
super(WeaveFile, self).join(other, pb, msg, version_ids, ignore_missing)
920
def _reweave(wa, wb, pb=None, msg=None):
921
"""Combine two weaves and return the result.
923
This works even if a revision R has different parents in
924
wa and wb. In the resulting weave all the parents are given.
926
This is done by just building up a new weave, maintaining ordering
927
of the versions in the two inputs. More efficient approaches
928
might be possible but it should only be necessary to do
929
this operation rarely, when a new previously ghost version is
932
:param pb: An optional progress bar, indicating how far done we are
933
:param msg: An optional message for the progress
937
queue_a = range(wa.num_versions())
938
queue_b = range(wb.num_versions())
939
# first determine combined parents of all versions
940
# map from version name -> all parent names
941
combined_parents = _reweave_parent_graphs(wa, wb)
942
mutter("combined parents: %r", combined_parents)
943
order = topo_sort(combined_parents.iteritems())
944
mutter("order to reweave: %r", order)
949
for idx, name in enumerate(order):
951
pb.update(msg, idx, len(order))
952
if name in wa._name_map:
953
lines = wa.get_lines(name)
954
if name in wb._name_map:
955
lines_b = wb.get_lines(name)
957
mutter('Weaves differ on content. rev_id {%s}', name)
958
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
960
lines = list(difflib.unified_diff(lines, lines_b,
961
wa._weave_name, wb._weave_name))
962
mutter('lines:\n%s', ''.join(lines))
963
raise errors.WeaveTextDiffers(name, wa, wb)
965
lines = wb.get_lines(name)
966
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
969
def _reweave_parent_graphs(wa, wb):
970
"""Return combined parent ancestry for two weaves.
972
Returned as a list of (version_name, set(parent_names))"""
974
for weave in [wa, wb]:
975
for idx, name in enumerate(weave._names):
976
p = combined.setdefault(name, set())
977
p.update(map(weave._idx_to_name, weave._parents[idx]))
982
"""Show the weave's table-of-contents"""
983
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
984
for i in (6, 50, 10, 10):
987
for i in range(w.num_versions()):
990
parent_str = ' '.join(map(str, w._parents[i]))
991
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
995
def weave_stats(weave_file, pb):
996
from bzrlib.weavefile import read_weave
998
wf = file(weave_file, 'rb')
1000
# FIXME: doesn't work on pipes
1001
weave_size = wf.tell()
1005
for i in range(vers):
1006
pb.update('checking sizes', i, vers)
1007
for origin, lineno, line in w._extract([i]):
1012
print 'versions %9d' % vers
1013
print 'weave file %9d bytes' % weave_size
1014
print 'total contents %9d bytes' % total
1015
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1018
print 'average size %9d bytes' % avg
1019
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1023
print """bzr weave tool
1025
Experimental tool for weave algorithm.
1028
weave init WEAVEFILE
1029
Create an empty weave file
1030
weave get WEAVEFILE VERSION
1031
Write out specified version.
1032
weave check WEAVEFILE
1033
Check consistency of all versions.
1035
Display table of contents.
1036
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1037
Add NEWTEXT, with specified parent versions.
1038
weave annotate WEAVEFILE VERSION
1039
Display origin of each line.
1040
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1041
Auto-merge two versions and display conflicts.
1042
weave diff WEAVEFILE VERSION1 VERSION2
1043
Show differences between two versions.
1047
% weave init foo.weave
1049
% weave add foo.weave ver0 < foo.txt
1052
(create updated version)
1054
% weave get foo.weave 0 | diff -u - foo.txt
1055
% weave add foo.weave ver1 0 < foo.txt
1058
% weave get foo.weave 0 > foo.txt (create forked version)
1060
% weave add foo.weave ver2 0 < foo.txt
1063
% weave merge foo.weave 1 2 > foo.txt (merge them)
1064
% vi foo.txt (resolve conflicts)
1065
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1077
# in case we're run directly from the subdirectory
1078
sys.path.append('..')
1080
from bzrlib.weavefile import write_weave, read_weave
1081
from bzrlib.progress import ProgressBar
1096
return read_weave(file(argv[2], 'rb'))
1102
# at the moment, based on everything in the file
1104
parents = map(int, argv[4:])
1105
lines = sys.stdin.readlines()
1106
ver = w.add(name, parents, lines)
1107
write_weave(w, file(argv[2], 'wb'))
1108
print 'added version %r %d' % (name, ver)
1111
if os.path.exists(fn):
1112
raise IOError("file exists")
1114
write_weave(w, file(fn, 'wb'))
1115
elif cmd == 'get': # get one version
1117
sys.stdout.writelines(w.get_iter(int(argv[3])))
1122
v1, v2 = map(int, argv[3:5])
1125
diff_gen = bzrlib.patiencediff.unified_diff(lines1, lines2,
1126
'%s version %d' % (fn, v1),
1127
'%s version %d' % (fn, v2))
1128
sys.stdout.writelines(diff_gen)
1130
elif cmd == 'annotate':
1132
# newline is added to all lines regardless; too hard to get
1133
# reasonable formatting otherwise
1135
for origin, text in w.annotate(int(argv[3])):
1136
text = text.rstrip('\r\n')
1138
print ' | %s' % (text)
1140
print '%5d | %s' % (origin, text)
1146
elif cmd == 'stats':
1147
weave_stats(argv[2], ProgressBar())
1149
elif cmd == 'check':
1154
print '%d versions ok' % w.num_versions()
1156
elif cmd == 'inclusions':
1158
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1160
elif cmd == 'parents':
1162
print ' '.join(map(str, w._parents[int(argv[3])]))
1164
elif cmd == 'plan-merge':
1165
# replaced by 'bzr weave-plan-merge'
1167
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1169
print '%14s | %s' % (state, line),
1170
elif cmd == 'merge':
1171
# replaced by 'bzr weave-merge-text'
1173
p = w.plan_merge(int(argv[3]), int(argv[4]))
1174
sys.stdout.writelines(w.weave_merge(p))
1176
raise ValueError('unknown command %r' % cmd)
1179
if __name__ == '__main__':
1181
sys.exit(main(sys.argv))
1184
class InterWeave(InterVersionedFile):
1185
"""Optimised code paths for weave to weave operations."""
1187
_matching_file_from_factory = staticmethod(WeaveFile)
1188
_matching_file_to_factory = staticmethod(WeaveFile)
1191
def is_compatible(source, target):
1192
"""Be compatible with weaves."""
1194
return (isinstance(source, Weave) and
1195
isinstance(target, Weave))
1196
except AttributeError:
1199
def join(self, pb=None, msg=None, version_ids=None, ignore_missing=False):
1200
"""See InterVersionedFile.join."""
1201
version_ids = self._get_source_version_ids(version_ids, ignore_missing)
1202
if self.target.versions() == [] and version_ids is None:
1203
self.target._copy_weave_content(self.source)
1205
self.target._join(self.source, pb, msg, version_ids, ignore_missing)
1208
InterVersionedFile.register_optimiser(InterWeave)
added
removed
bzrlib/weave.py
