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
78
from bzrlib.lazy_import import lazy_import
79
lazy_import(globals(), """
80
from bzrlib import tsort
85
from bzrlib.trace import mutter
86
from bzrlib.errors import (WeaveError, WeaveFormatError, WeaveParentMismatch,
87
RevisionAlreadyPresent,
89
WeaveRevisionAlreadyPresent,
90
WeaveRevisionNotPresent,
92
import bzrlib.errors as errors
93
from bzrlib.osutils import sha_strings
94
import bzrlib.patiencediff
95
from bzrlib.versionedfile import VersionedFile, InterVersionedFile
96
from bzrlib.weavefile import _read_weave_v5, write_weave_v5
99
class Weave(VersionedFile):
100
"""weave - versioned text file storage.
102
A Weave manages versions of line-based text files, keeping track
103
of the originating version for each line.
105
To clients the "lines" of the file are represented as a list of strings.
106
These strings will typically have terminal newline characters, but
107
this is not required. In particular files commonly do not have a newline
108
at the end of the file.
110
Texts can be identified in either of two ways:
112
* a nonnegative index number.
114
* a version-id string.
116
Typically the index number will be valid only inside this weave and
117
the version-id is used to reference it in the larger world.
119
The weave is represented as a list mixing edit instructions and
120
literal text. Each entry in _weave can be either a string (or
121
unicode), or a tuple. If a string, it means that the given line
122
should be output in the currently active revisions.
124
If a tuple, it gives a processing instruction saying in which
125
revisions the enclosed lines are active. The tuple has the form
126
(instruction, version).
128
The instruction can be '{' or '}' for an insertion block, and '['
129
and ']' for a deletion block respectively. The version is the
130
integer version index. There is no replace operator, only deletes
131
and inserts. For '}', the end of an insertion, there is no
132
version parameter because it always closes the most recently
137
* A later version can delete lines that were introduced by any
138
number of ancestor versions; this implies that deletion
139
instructions can span insertion blocks without regard to the
140
insertion block's nesting.
142
* Similarly, deletions need not be properly nested with regard to
143
each other, because they might have been generated by
144
independent revisions.
146
* Insertions are always made by inserting a new bracketed block
147
into a single point in the previous weave. This implies they
148
can nest but not overlap, and the nesting must always have later
149
insertions on the inside.
151
* It doesn't seem very useful to have an active insertion
152
inside an inactive insertion, but it might happen.
154
* Therefore, all instructions are always"considered"; that
155
is passed onto and off the stack. An outer inactive block
156
doesn't disable an inner block.
158
* Lines are enabled if the most recent enclosing insertion is
159
active and none of the enclosing deletions are active.
161
* There is no point having a deletion directly inside its own
162
insertion; you might as well just not write it. And there
163
should be no way to get an earlier version deleting a later
167
Text of the weave; list of control instruction tuples and strings.
170
List of parents, indexed by version number.
171
It is only necessary to store the minimal set of parents for
172
each version; the parent's parents are implied.
175
List of hex SHA-1 of each version.
178
List of symbolic names for each version. Each should be unique.
181
For each name, the version number.
184
Descriptive name of this weave; typically the filename if known.
188
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
189
'_weave_name', '_matcher']
191
def __init__(self, weave_name=None, access_mode='w', matcher=None):
192
super(Weave, self).__init__(access_mode)
198
self._weave_name = weave_name
200
self._matcher = bzrlib.patiencediff.PatienceSequenceMatcher
202
self._matcher = matcher
205
return "Weave(%r)" % self._weave_name
208
"""Return a deep copy of self.
210
The copy can be modified without affecting the original weave."""
212
other._weave = self._weave[:]
213
other._parents = self._parents[:]
214
other._sha1s = self._sha1s[:]
215
other._names = self._names[:]
216
other._name_map = self._name_map.copy()
217
other._weave_name = self._weave_name
220
def __eq__(self, other):
221
if not isinstance(other, Weave):
223
return self._parents == other._parents \
224
and self._weave == other._weave \
225
and self._sha1s == other._sha1s
227
def __ne__(self, other):
228
return not self.__eq__(other)
230
def _idx_to_name(self, version):
231
return self._names[version]
233
def _lookup(self, name):
234
"""Convert symbolic version name to index."""
235
self.check_not_reserved_id(name)
237
return self._name_map[name]
239
raise RevisionNotPresent(name, self._weave_name)
242
"""See VersionedFile.versions."""
243
return self._names[:]
245
def has_version(self, version_id):
246
"""See VersionedFile.has_version."""
247
return (version_id in self._name_map)
249
__contains__ = has_version
251
def get_parent_map(self, version_ids):
252
"""See VersionedFile.get_parent_map."""
254
for version_id in version_ids:
256
result[version_id] = tuple(
257
map(self._idx_to_name, self._parents[self._lookup(version_id)]))
258
except RevisionNotPresent:
262
def get_parents_with_ghosts(self, version_id):
263
raise NotImplementedError(self.get_parents_with_ghosts)
265
def _check_repeated_add(self, name, parents, text, sha1):
266
"""Check that a duplicated add is OK.
268
If it is, return the (old) index; otherwise raise an exception.
270
idx = self._lookup(name)
271
if sorted(self._parents[idx]) != sorted(parents) \
272
or sha1 != self._sha1s[idx]:
273
raise RevisionAlreadyPresent(name, self._weave_name)
276
def _add_lines(self, version_id, parents, lines, parent_texts,
277
left_matching_blocks, nostore_sha, random_id, check_content):
278
"""See VersionedFile.add_lines."""
279
idx = self._add(version_id, lines, map(self._lookup, parents),
280
nostore_sha=nostore_sha)
281
return sha_strings(lines), sum(map(len, lines)), idx
283
def _add(self, version_id, lines, parents, sha1=None, nostore_sha=None):
284
"""Add a single text on top of the weave.
286
Returns the index number of the newly added version.
289
Symbolic name for this version.
290
(Typically the revision-id of the revision that added it.)
293
List or set of direct parent version numbers.
296
Sequence of lines to be added in the new version.
298
:param nostore_sha: See VersionedFile.add_lines.
300
assert isinstance(version_id, basestring)
301
self._check_lines_not_unicode(lines)
302
self._check_lines_are_lines(lines)
304
sha1 = sha_strings(lines)
305
if sha1 == nostore_sha:
306
raise errors.ExistingContent
307
if version_id in self._name_map:
308
return self._check_repeated_add(version_id, parents, lines, sha1)
310
self._check_versions(parents)
311
## self._check_lines(lines)
312
new_version = len(self._parents)
314
# if we abort after here the (in-memory) weave will be corrupt because only
315
# some fields are updated
316
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
317
# - Robert Collins 20060226
318
self._parents.append(parents[:])
319
self._sha1s.append(sha1)
320
self._names.append(version_id)
321
self._name_map[version_id] = new_version
325
# special case; adding with no parents revision; can do
326
# this more quickly by just appending unconditionally.
327
# even more specially, if we're adding an empty text we
328
# need do nothing at all.
330
self._weave.append(('{', new_version))
331
self._weave.extend(lines)
332
self._weave.append(('}', None))
335
if len(parents) == 1:
336
pv = list(parents)[0]
337
if sha1 == self._sha1s[pv]:
338
# special case: same as the single parent
342
ancestors = self._inclusions(parents)
346
# basis a list of (origin, lineno, line)
349
for origin, lineno, line in self._extract(ancestors):
350
basis_lineno.append(lineno)
351
basis_lines.append(line)
353
# another small special case: a merge, producing the same text
355
if lines == basis_lines:
358
# add a sentinel, because we can also match against the final line
359
basis_lineno.append(len(self._weave))
361
# XXX: which line of the weave should we really consider
362
# matches the end of the file? the current code says it's the
363
# last line of the weave?
365
#print 'basis_lines:', basis_lines
366
#print 'new_lines: ', lines
368
s = self._matcher(None, basis_lines, lines)
370
# offset gives the number of lines that have been inserted
371
# into the weave up to the current point; if the original edit instruction
372
# says to change line A then we actually change (A+offset)
375
for tag, i1, i2, j1, j2 in s.get_opcodes():
376
# i1,i2 are given in offsets within basis_lines; we need to map them
377
# back to offsets within the entire weave
378
#print 'raw match', tag, i1, i2, j1, j2
382
i1 = basis_lineno[i1]
383
i2 = basis_lineno[i2]
385
assert 0 <= j1 <= j2 <= len(lines)
387
#print tag, i1, i2, j1, j2
389
# the deletion and insertion are handled separately.
390
# first delete the region.
392
self._weave.insert(i1+offset, ('[', new_version))
393
self._weave.insert(i2+offset+1, (']', new_version))
397
# there may have been a deletion spanning up to
398
# i2; we want to insert after this region to make sure
399
# we don't destroy ourselves
401
self._weave[i:i] = ([('{', new_version)]
404
offset += 2 + (j2 - j1)
407
def _clone_text(self, new_version_id, old_version_id, parents):
408
"""See VersionedFile.clone_text."""
409
old_lines = self.get_text(old_version_id)
410
self.add_lines(new_version_id, parents, old_lines)
412
def _inclusions(self, versions):
413
"""Return set of all ancestors of given version(s)."""
414
if not len(versions):
417
for v in xrange(max(versions), 0, -1):
419
# include all its parents
420
i.update(self._parents[v])
422
## except IndexError:
423
## raise ValueError("version %d not present in weave" % v)
425
def get_ancestry(self, version_ids, topo_sorted=True):
426
"""See VersionedFile.get_ancestry."""
427
if isinstance(version_ids, basestring):
428
version_ids = [version_ids]
429
i = self._inclusions([self._lookup(v) for v in version_ids])
430
return [self._idx_to_name(v) for v in i]
432
def _check_lines(self, text):
433
if not isinstance(text, list):
434
raise ValueError("text should be a list, not %s" % type(text))
437
if not isinstance(l, basestring):
438
raise ValueError("text line should be a string or unicode, not %s"
443
def _check_versions(self, indexes):
444
"""Check everything in the sequence of indexes is valid"""
449
raise IndexError("invalid version number %r" % i)
451
def _compatible_parents(self, my_parents, other_parents):
452
"""During join check that other_parents are joinable with my_parents.
454
Joinable is defined as 'is a subset of' - supersets may require
455
regeneration of diffs, but subsets do not.
457
return len(other_parents.difference(my_parents)) == 0
459
def annotate_iter(self, version_id):
460
"""Yield list of (version-id, line) pairs for the specified version.
462
The index indicates when the line originated in the weave."""
463
incls = [self._lookup(version_id)]
464
for origin, lineno, text in self._extract(incls):
465
yield self._idx_to_name(origin), text
467
def iter_lines_added_or_present_in_versions(self, version_ids=None,
469
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
470
if version_ids is None:
471
version_ids = self.versions()
472
version_ids = set(version_ids)
473
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
474
# if inserted not in version_ids then it was inserted before the
475
# versions we care about, but because weaves cannot represent ghosts
476
# properly, we do not filter down to that
477
# if inserted not in version_ids: continue
479
yield line + '\n', inserted
483
def _walk_internal(self, version_ids=None):
484
"""Helper method for weave actions."""
489
lineno = 0 # line of weave, 0-based
491
for l in self._weave:
492
if l.__class__ == tuple:
496
istack.append(self._names[v])
500
assert self._names[v] not in dset
501
dset.add(self._names[v])
503
dset.remove(self._names[v])
505
raise WeaveFormatError('unexpected instruction %r' % v)
507
assert l.__class__ in (str, unicode)
509
yield lineno, istack[-1], frozenset(dset), l
513
raise WeaveFormatError("unclosed insertion blocks "
514
"at end of weave: %s" % istack)
516
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
519
def plan_merge(self, ver_a, ver_b):
520
"""Return pseudo-annotation indicating how the two versions merge.
522
This is computed between versions a and b and their common
525
Weave lines present in none of them are skipped entirely.
527
inc_a = set(self.get_ancestry([ver_a]))
528
inc_b = set(self.get_ancestry([ver_b]))
529
inc_c = inc_a & inc_b
531
for lineno, insert, deleteset, line in self._walk_internal([ver_a, ver_b]):
532
if deleteset & inc_c:
533
# killed in parent; can't be in either a or b
534
# not relevant to our work
535
yield 'killed-base', line
536
elif insert in inc_c:
537
# was inserted in base
538
killed_a = bool(deleteset & inc_a)
539
killed_b = bool(deleteset & inc_b)
540
if killed_a and killed_b:
541
yield 'killed-both', line
543
yield 'killed-a', line
545
yield 'killed-b', line
547
yield 'unchanged', line
548
elif insert in inc_a:
549
if deleteset & inc_a:
550
yield 'ghost-a', line
554
elif insert in inc_b:
555
if deleteset & inc_b:
556
yield 'ghost-b', line
560
# not in either revision
561
yield 'irrelevant', line
563
yield 'unchanged', '' # terminator
565
def _extract(self, versions):
566
"""Yield annotation of lines in included set.
568
Yields a sequence of tuples (origin, lineno, text), where
569
origin is the origin version, lineno the index in the weave,
570
and text the text of the line.
572
The set typically but not necessarily corresponds to a version.
575
if not isinstance(i, int):
578
included = self._inclusions(versions)
584
lineno = 0 # line of weave, 0-based
590
WFE = WeaveFormatError
593
# 449 0 4474.6820 2356.5590 bzrlib.weave:556(_extract)
594
# +285282 0 1676.8040 1676.8040 +<isinstance>
595
# 1.6 seconds in 'isinstance'.
596
# changing the first isinstance:
597
# 449 0 2814.2660 1577.1760 bzrlib.weave:556(_extract)
598
# +140414 0 762.8050 762.8050 +<isinstance>
599
# note that the inline time actually dropped (less function calls)
600
# and total processing time was halved.
601
# we're still spending ~1/4 of the method in isinstance though.
602
# so lets hard code the acceptable string classes we expect:
603
# 449 0 1202.9420 786.2930 bzrlib.weave:556(_extract)
604
# +71352 0 377.5560 377.5560 +<method 'append' of 'list'
606
# yay, down to ~1/4 the initial extract time, and our inline time
607
# has shrunk again, with isinstance no longer dominating.
608
# tweaking the stack inclusion test to use a set gives:
609
# 449 0 1122.8030 713.0080 bzrlib.weave:556(_extract)
610
# +71352 0 354.9980 354.9980 +<method 'append' of 'list'
612
# - a 5% win, or possibly just noise. However with large istacks that
613
# 'in' test could dominate, so I'm leaving this change in place -
614
# when its fast enough to consider profiling big datasets we can review.
619
for l in self._weave:
620
if l.__class__ == tuple:
628
iset.remove(istack.pop())
639
assert l.__class__ in (str, unicode)
641
isactive = (not dset) and istack and (istack[-1] in included)
643
result.append((istack[-1], lineno, l))
646
raise WeaveFormatError("unclosed insertion blocks "
647
"at end of weave: %s" % istack)
649
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
653
def _maybe_lookup(self, name_or_index):
654
"""Convert possible symbolic name to index, or pass through indexes.
658
if isinstance(name_or_index, (int, long)):
661
return self._lookup(name_or_index)
663
def get_lines(self, version_id):
664
"""See VersionedFile.get_lines()."""
665
int_index = self._maybe_lookup(version_id)
666
result = [line for (origin, lineno, line) in self._extract([int_index])]
667
expected_sha1 = self._sha1s[int_index]
668
measured_sha1 = sha_strings(result)
669
if measured_sha1 != expected_sha1:
670
raise errors.WeaveInvalidChecksum(
671
'file %s, revision %s, expected: %s, measured %s'
672
% (self._weave_name, version_id,
673
expected_sha1, measured_sha1))
676
def get_sha1(self, version_id):
677
"""See VersionedFile.get_sha1()."""
678
return self._sha1s[self._lookup(version_id)]
680
def get_sha1s(self, version_ids):
681
"""See VersionedFile.get_sha1s()."""
682
return [self._sha1s[self._lookup(v)] for v in version_ids]
684
def num_versions(self):
685
"""How many versions are in this weave?"""
686
l = len(self._parents)
687
assert l == len(self._sha1s)
690
__len__ = num_versions
692
def check(self, progress_bar=None):
693
# TODO evaluate performance hit of using string sets in this routine.
694
# TODO: check no circular inclusions
695
# TODO: create a nested progress bar
696
for version in range(self.num_versions()):
697
inclusions = list(self._parents[version])
700
if inclusions[-1] >= version:
701
raise WeaveFormatError("invalid included version %d for index %d"
702
% (inclusions[-1], version))
704
# try extracting all versions; parallel extraction is used
705
nv = self.num_versions()
710
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
711
# The problem is that set membership is much more expensive
712
name = self._idx_to_name(i)
713
sha1s[name] = sha.new()
715
new_inc = set([name])
716
for p in self._parents[i]:
717
new_inc.update(inclusions[self._idx_to_name(p)])
719
assert set(new_inc) == set(self.get_ancestry(name)), \
720
'failed %s != %s' % (set(new_inc), set(self.get_ancestry(name)))
721
inclusions[name] = new_inc
723
nlines = len(self._weave)
725
update_text = 'checking weave'
727
short_name = os.path.basename(self._weave_name)
728
update_text = 'checking %s' % (short_name,)
729
update_text = update_text[:25]
731
for lineno, insert, deleteset, line in self._walk_internal():
733
progress_bar.update(update_text, lineno, nlines)
735
for name, name_inclusions in inclusions.items():
736
# The active inclusion must be an ancestor,
737
# and no ancestors must have deleted this line,
738
# because we don't support resurrection.
739
if (insert in name_inclusions) and not (deleteset & name_inclusions):
740
sha1s[name].update(line)
743
version = self._idx_to_name(i)
744
hd = sha1s[version].hexdigest()
745
expected = self._sha1s[i]
747
raise errors.WeaveInvalidChecksum(
748
"mismatched sha1 for version %s: "
749
"got %s, expected %s"
750
% (version, hd, expected))
752
# TODO: check insertions are properly nested, that there are
753
# no lines outside of insertion blocks, that deletions are
754
# properly paired, etc.
756
def _join(self, other, pb, msg, version_ids, ignore_missing):
757
"""Worker routine for join()."""
758
if not other.versions():
759
return # nothing to update, easy
762
# versions is never none, InterWeave checks this.
765
# two loops so that we do not change ourselves before verifying it
767
# work through in index order to make sure we get all dependencies
770
# get the selected versions only that are in other.versions.
771
version_ids = set(other.versions()).intersection(set(version_ids))
772
# pull in the referenced graph.
773
version_ids = other.get_ancestry(version_ids)
774
pending_parents = other.get_parent_map(version_ids)
775
pending_graph = pending_parents.items()
776
if len(pending_graph) != len(version_ids):
777
raise RevisionNotPresent(
778
set(version_ids) - set(pending_parents.keys()), self)
779
for name in tsort.topo_sort(pending_graph):
780
other_idx = other._name_map[name]
781
# returns True if we have it, False if we need it.
782
if not self._check_version_consistent(other, other_idx, name):
783
names_to_join.append((other_idx, name))
791
for other_idx, name in names_to_join:
792
# TODO: If all the parents of the other version are already
793
# present then we can avoid some work by just taking the delta
794
# and adjusting the offsets.
795
new_parents = self._imported_parents(other, other_idx)
796
sha1 = other._sha1s[other_idx]
801
pb.update(msg, merged, len(names_to_join))
803
lines = other.get_lines(other_idx)
804
self._add(name, lines, new_parents, sha1)
806
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
807
merged, processed, self._weave_name, time.time()-time0))
809
def _imported_parents(self, other, other_idx):
810
"""Return list of parents in self corresponding to indexes in other."""
812
for parent_idx in other._parents[other_idx]:
813
parent_name = other._names[parent_idx]
814
if parent_name not in self._name_map:
815
# should not be possible
816
raise WeaveError("missing parent {%s} of {%s} in %r"
817
% (parent_name, other._name_map[other_idx], self))
818
new_parents.append(self._name_map[parent_name])
821
def _check_version_consistent(self, other, other_idx, name):
822
"""Check if a version in consistent in this and other.
824
To be consistent it must have:
827
* the same direct parents (by name, not index, and disregarding
830
If present & correct return True;
831
if not present in self return False;
832
if inconsistent raise error."""
833
this_idx = self._name_map.get(name, -1)
835
if self._sha1s[this_idx] != other._sha1s[other_idx]:
836
raise errors.WeaveTextDiffers(name, self, other)
837
self_parents = self._parents[this_idx]
838
other_parents = other._parents[other_idx]
839
n1 = set([self._names[i] for i in self_parents])
840
n2 = set([other._names[i] for i in other_parents])
841
if not self._compatible_parents(n1, n2):
842
raise WeaveParentMismatch("inconsistent parents "
843
"for version {%s}: %s vs %s" % (name, n1, n2))
849
def _reweave(self, other, pb, msg):
850
"""Reweave self with other - internal helper for join().
852
:param other: The other weave to merge
853
:param pb: An optional progress bar, indicating how far done we are
854
:param msg: An optional message for the progress
856
new_weave = _reweave(self, other, pb=pb, msg=msg)
857
self._copy_weave_content(new_weave)
859
def _copy_weave_content(self, otherweave):
860
"""adsorb the content from otherweave."""
861
for attr in self.__slots__:
862
if attr != '_weave_name':
863
setattr(self, attr, copy(getattr(otherweave, attr)))
866
class WeaveFile(Weave):
867
"""A WeaveFile represents a Weave on disk and writes on change."""
869
WEAVE_SUFFIX = '.weave'
871
def __init__(self, name, transport, filemode=None, create=False, access_mode='w'):
872
"""Create a WeaveFile.
874
:param create: If not True, only open an existing knit.
876
super(WeaveFile, self).__init__(name, access_mode)
877
self._transport = transport
878
self._filemode = filemode
880
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
881
except errors.NoSuchFile:
887
def _add_lines(self, version_id, parents, lines, parent_texts,
888
left_matching_blocks, nostore_sha, random_id, check_content):
889
"""Add a version and save the weave."""
890
self.check_not_reserved_id(version_id)
891
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
892
parent_texts, left_matching_blocks, nostore_sha, random_id,
897
def _clone_text(self, new_version_id, old_version_id, parents):
898
"""See VersionedFile.clone_text."""
899
super(WeaveFile, self)._clone_text(new_version_id, old_version_id, parents)
902
def copy_to(self, name, transport):
903
"""See VersionedFile.copy_to()."""
904
# as we are all in memory always, just serialise to the new place.
906
write_weave_v5(self, sio)
908
transport.put_file(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
910
def create_empty(self, name, transport, filemode=None):
911
return WeaveFile(name, transport, filemode, create=True)
914
"""Save the weave."""
915
self._check_write_ok()
917
write_weave_v5(self, sio)
919
self._transport.put_file(self._weave_name + WeaveFile.WEAVE_SUFFIX,
925
"""See VersionedFile.get_suffixes()."""
926
return [WeaveFile.WEAVE_SUFFIX]
928
def join(self, other, pb=None, msg=None, version_ids=None,
929
ignore_missing=False):
930
"""Join other into self and save."""
931
super(WeaveFile, self).join(other, pb, msg, version_ids, ignore_missing)
935
def _reweave(wa, wb, pb=None, msg=None):
936
"""Combine two weaves and return the result.
938
This works even if a revision R has different parents in
939
wa and wb. In the resulting weave all the parents are given.
941
This is done by just building up a new weave, maintaining ordering
942
of the versions in the two inputs. More efficient approaches
943
might be possible but it should only be necessary to do
944
this operation rarely, when a new previously ghost version is
947
:param pb: An optional progress bar, indicating how far done we are
948
:param msg: An optional message for the progress
952
queue_a = range(wa.num_versions())
953
queue_b = range(wb.num_versions())
954
# first determine combined parents of all versions
955
# map from version name -> all parent names
956
combined_parents = _reweave_parent_graphs(wa, wb)
957
mutter("combined parents: %r", combined_parents)
958
order = tsort.topo_sort(combined_parents.iteritems())
959
mutter("order to reweave: %r", order)
964
for idx, name in enumerate(order):
966
pb.update(msg, idx, len(order))
967
if name in wa._name_map:
968
lines = wa.get_lines(name)
969
if name in wb._name_map:
970
lines_b = wb.get_lines(name)
972
mutter('Weaves differ on content. rev_id {%s}', name)
973
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
975
lines = list(difflib.unified_diff(lines, lines_b,
976
wa._weave_name, wb._weave_name))
977
mutter('lines:\n%s', ''.join(lines))
978
raise errors.WeaveTextDiffers(name, wa, wb)
980
lines = wb.get_lines(name)
981
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
984
def _reweave_parent_graphs(wa, wb):
985
"""Return combined parent ancestry for two weaves.
987
Returned as a list of (version_name, set(parent_names))"""
989
for weave in [wa, wb]:
990
for idx, name in enumerate(weave._names):
991
p = combined.setdefault(name, set())
992
p.update(map(weave._idx_to_name, weave._parents[idx]))
997
"""Show the weave's table-of-contents"""
998
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
999
for i in (6, 50, 10, 10):
1002
for i in range(w.num_versions()):
1005
parent_str = ' '.join(map(str, w._parents[i]))
1006
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
1010
def weave_stats(weave_file, pb):
1011
from bzrlib.weavefile import read_weave
1013
wf = file(weave_file, 'rb')
1015
# FIXME: doesn't work on pipes
1016
weave_size = wf.tell()
1020
for i in range(vers):
1021
pb.update('checking sizes', i, vers)
1022
for origin, lineno, line in w._extract([i]):
1027
print 'versions %9d' % vers
1028
print 'weave file %9d bytes' % weave_size
1029
print 'total contents %9d bytes' % total
1030
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1033
print 'average size %9d bytes' % avg
1034
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1038
print """bzr weave tool
1040
Experimental tool for weave algorithm.
1043
weave init WEAVEFILE
1044
Create an empty weave file
1045
weave get WEAVEFILE VERSION
1046
Write out specified version.
1047
weave check WEAVEFILE
1048
Check consistency of all versions.
1050
Display table of contents.
1051
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1052
Add NEWTEXT, with specified parent versions.
1053
weave annotate WEAVEFILE VERSION
1054
Display origin of each line.
1055
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1056
Auto-merge two versions and display conflicts.
1057
weave diff WEAVEFILE VERSION1 VERSION2
1058
Show differences between two versions.
1062
% weave init foo.weave
1064
% weave add foo.weave ver0 < foo.txt
1067
(create updated version)
1069
% weave get foo.weave 0 | diff -u - foo.txt
1070
% weave add foo.weave ver1 0 < foo.txt
1073
% weave get foo.weave 0 > foo.txt (create forked version)
1075
% weave add foo.weave ver2 0 < foo.txt
1078
% weave merge foo.weave 1 2 > foo.txt (merge them)
1079
% vi foo.txt (resolve conflicts)
1080
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1092
# in case we're run directly from the subdirectory
1093
sys.path.append('..')
1095
from bzrlib.weavefile import write_weave, read_weave
1096
from bzrlib.progress import ProgressBar
1111
return read_weave(file(argv[2], 'rb'))
1117
# at the moment, based on everything in the file
1119
parents = map(int, argv[4:])
1120
lines = sys.stdin.readlines()
1121
ver = w.add(name, parents, lines)
1122
write_weave(w, file(argv[2], 'wb'))
1123
print 'added version %r %d' % (name, ver)
1126
if os.path.exists(fn):
1127
raise IOError("file exists")
1129
write_weave(w, file(fn, 'wb'))
1130
elif cmd == 'get': # get one version
1132
sys.stdout.writelines(w.get_iter(int(argv[3])))
1137
v1, v2 = map(int, argv[3:5])
1140
diff_gen = bzrlib.patiencediff.unified_diff(lines1, lines2,
1141
'%s version %d' % (fn, v1),
1142
'%s version %d' % (fn, v2))
1143
sys.stdout.writelines(diff_gen)
1145
elif cmd == 'annotate':
1147
# newline is added to all lines regardless; too hard to get
1148
# reasonable formatting otherwise
1150
for origin, text in w.annotate(int(argv[3])):
1151
text = text.rstrip('\r\n')
1153
print ' | %s' % (text)
1155
print '%5d | %s' % (origin, text)
1161
elif cmd == 'stats':
1162
weave_stats(argv[2], ProgressBar())
1164
elif cmd == 'check':
1169
print '%d versions ok' % w.num_versions()
1171
elif cmd == 'inclusions':
1173
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1175
elif cmd == 'parents':
1177
print ' '.join(map(str, w._parents[int(argv[3])]))
1179
elif cmd == 'plan-merge':
1180
# replaced by 'bzr weave-plan-merge'
1182
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1184
print '%14s | %s' % (state, line),
1185
elif cmd == 'merge':
1186
# replaced by 'bzr weave-merge-text'
1188
p = w.plan_merge(int(argv[3]), int(argv[4]))
1189
sys.stdout.writelines(w.weave_merge(p))
1191
raise ValueError('unknown command %r' % cmd)
1194
if __name__ == '__main__':
1196
sys.exit(main(sys.argv))
1199
class InterWeave(InterVersionedFile):
1200
"""Optimised code paths for weave to weave operations."""
1202
_matching_file_from_factory = staticmethod(WeaveFile)
1203
_matching_file_to_factory = staticmethod(WeaveFile)
1206
def is_compatible(source, target):
1207
"""Be compatible with weaves."""
1209
return (isinstance(source, Weave) and
1210
isinstance(target, Weave))
1211
except AttributeError:
1214
def join(self, pb=None, msg=None, version_ids=None, ignore_missing=False):
1215
"""See InterVersionedFile.join."""
1216
version_ids = self._get_source_version_ids(version_ids, ignore_missing)
1217
if self.target.versions() == [] and version_ids is None:
1218
self.target._copy_weave_content(self.source)
1220
self.target._join(self.source, pb, msg, version_ids, ignore_missing)
1223
InterVersionedFile.register_optimiser(InterWeave)
added
removed
bzrlib/weave.py
