3
# Copyright (C) 2005 Canonical Ltd
5
# This program is free software; you can redistribute it and/or modify
6
# it under the terms of the GNU General Public License as published by
7
# the Free Software Foundation; either version 2 of the License, or
8
# (at your option) any later version.
10
# This program is distributed in the hope that it will be useful,
11
# but WITHOUT ANY WARRANTY; without even the implied warranty of
12
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13
# GNU General Public License for more details.
15
# You should have received a copy of the GNU General Public License
16
# along with this program; if not, write to the Free Software
17
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
19
# Author: Martin Pool <mbp@canonical.com>
22
"""Weave - storage of related text file versions"""
25
# XXX: If we do weaves this way, will a merge still behave the same
26
# way if it's done in a different order? That's a pretty desirable
29
# TODO: Nothing here so far assumes the lines are really \n newlines,
30
# rather than being split up in some other way. We could accommodate
31
# binaries, perhaps by naively splitting on \n or perhaps using
32
# something like a rolling checksum.
34
# TODO: End marker for each version so we can stop reading?
36
# TODO: Check that no insertion occurs inside a deletion that was
37
# active in the version of the insertion.
39
# TODO: In addition to the SHA-1 check, perhaps have some code that
40
# checks structural constraints of the weave: ie that insertions are
41
# properly nested, that there is no text outside of an insertion, that
42
# insertions or deletions are not repeated, etc.
44
# TODO: Parallel-extract that passes back each line along with a
45
# description of which revisions include it. Nice for checking all
46
# shas or calculating stats in parallel.
48
# TODO: Using a single _extract routine and then processing the output
49
# is probably inefficient. It's simple enough that we can afford to
50
# have slight specializations for different ways its used: annotate,
51
# basis for add, get, etc.
53
# TODO: Probably the API should work only in names to hide the integer
54
# indexes from the user.
56
# TODO: Is there any potential performance win by having an add()
57
# variant that is passed a pre-cooked version of the single basis
60
# TODO: Reweave can possibly be made faster by remembering diffs
61
# where the basis and destination are unchanged.
63
# FIXME: Sometimes we will be given a parents list for a revision
64
# that includes some redundant parents (i.e. already a parent of
65
# something in the list.) We should eliminate them. This can
66
# be done fairly efficiently because the sequence numbers constrain
67
# the possible relationships.
69
# FIXME: the conflict markers should be *7* characters
72
from cStringIO import StringIO
80
from bzrlib.trace import mutter
81
from bzrlib.errors import (WeaveError, WeaveFormatError, WeaveParentMismatch,
82
RevisionAlreadyPresent,
84
WeaveRevisionAlreadyPresent,
85
WeaveRevisionNotPresent,
87
import bzrlib.errors as errors
88
from bzrlib.osutils import sha, sha_strings
89
import bzrlib.patiencediff
90
from bzrlib.tsort import topo_sort
91
from bzrlib.versionedfile import VersionedFile, InterVersionedFile
92
from bzrlib.weavefile import _read_weave_v5, write_weave_v5
95
class Weave(VersionedFile):
96
"""weave - versioned text file storage.
98
A Weave manages versions of line-based text files, keeping track
99
of the originating version for each line.
101
To clients the "lines" of the file are represented as a list of strings.
102
These strings will typically have terminal newline characters, but
103
this is not required. In particular files commonly do not have a newline
104
at the end of the file.
106
Texts can be identified in either of two ways:
108
* a nonnegative index number.
110
* a version-id string.
112
Typically the index number will be valid only inside this weave and
113
the version-id is used to reference it in the larger world.
115
The weave is represented as a list mixing edit instructions and
116
literal text. Each entry in _weave can be either a string (or
117
unicode), or a tuple. If a string, it means that the given line
118
should be output in the currently active revisions.
120
If a tuple, it gives a processing instruction saying in which
121
revisions the enclosed lines are active. The tuple has the form
122
(instruction, version).
124
The instruction can be '{' or '}' for an insertion block, and '['
125
and ']' for a deletion block respectively. The version is the
126
integer version index. There is no replace operator, only deletes
127
and inserts. For '}', the end of an insertion, there is no
128
version parameter because it always closes the most recently
133
* A later version can delete lines that were introduced by any
134
number of ancestor versions; this implies that deletion
135
instructions can span insertion blocks without regard to the
136
insertion block's nesting.
138
* Similarly, deletions need not be properly nested with regard to
139
each other, because they might have been generated by
140
independent revisions.
142
* Insertions are always made by inserting a new bracketed block
143
into a single point in the previous weave. This implies they
144
can nest but not overlap, and the nesting must always have later
145
insertions on the inside.
147
* It doesn't seem very useful to have an active insertion
148
inside an inactive insertion, but it might happen.
150
* Therefore, all instructions are always"considered"; that
151
is passed onto and off the stack. An outer inactive block
152
doesn't disable an inner block.
154
* Lines are enabled if the most recent enclosing insertion is
155
active and none of the enclosing deletions are active.
157
* There is no point having a deletion directly inside its own
158
insertion; you might as well just not write it. And there
159
should be no way to get an earlier version deleting a later
163
Text of the weave; list of control instruction tuples and strings.
166
List of parents, indexed by version number.
167
It is only necessary to store the minimal set of parents for
168
each version; the parent's parents are implied.
171
List of hex SHA-1 of each version.
174
List of symbolic names for each version. Each should be unique.
177
For each name, the version number.
180
Descriptive name of this weave; typically the filename if known.
184
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
185
'_weave_name', '_matcher']
187
def __init__(self, weave_name=None, access_mode='w', matcher=None):
188
super(Weave, self).__init__(access_mode)
194
self._weave_name = weave_name
196
self._matcher = bzrlib.patiencediff.PatienceSequenceMatcher
198
self._matcher = matcher
201
return "Weave(%r)" % self._weave_name
204
"""Return a deep copy of self.
206
The copy can be modified without affecting the original weave."""
208
other._weave = self._weave[:]
209
other._parents = self._parents[:]
210
other._sha1s = self._sha1s[:]
211
other._names = self._names[:]
212
other._name_map = self._name_map.copy()
213
other._weave_name = self._weave_name
216
def __eq__(self, other):
217
if not isinstance(other, Weave):
219
return self._parents == other._parents \
220
and self._weave == other._weave \
221
and self._sha1s == other._sha1s
223
def __ne__(self, other):
224
return not self.__eq__(other)
226
def _idx_to_name(self, version):
227
return self._names[version]
229
def _lookup(self, name):
230
"""Convert symbolic version name to index."""
231
self.check_not_reserved_id(name)
233
return self._name_map[name]
235
raise RevisionNotPresent(name, self._weave_name)
238
"""See VersionedFile.versions."""
239
return self._names[:]
241
def has_version(self, version_id):
242
"""See VersionedFile.has_version."""
243
return (version_id in self._name_map)
245
__contains__ = has_version
247
def get_parents(self, version_id):
248
"""See VersionedFile.get_parent."""
249
return map(self._idx_to_name, self._parents[self._lookup(version_id)])
251
def _check_repeated_add(self, name, parents, text, sha1):
252
"""Check that a duplicated add is OK.
254
If it is, return the (old) index; otherwise raise an exception.
256
idx = self._lookup(name)
257
if sorted(self._parents[idx]) != sorted(parents) \
258
or sha1 != self._sha1s[idx]:
259
raise RevisionAlreadyPresent(name, self._weave_name)
262
def _add_lines(self, version_id, parents, lines, parent_texts,
263
left_matching_blocks, nostore_sha, random_id, check_content):
264
"""See VersionedFile.add_lines."""
265
idx = self._add(version_id, lines, map(self._lookup, parents),
266
nostore_sha=nostore_sha)
267
return sha_strings(lines), sum(map(len, lines)), idx
269
def _add(self, version_id, lines, parents, sha1=None, nostore_sha=None):
270
"""Add a single text on top of the weave.
272
Returns the index number of the newly added version.
275
Symbolic name for this version.
276
(Typically the revision-id of the revision that added it.)
279
List or set of direct parent version numbers.
282
Sequence of lines to be added in the new version.
284
:param nostore_sha: See VersionedFile.add_lines.
286
assert isinstance(version_id, basestring)
287
self._check_lines_not_unicode(lines)
288
self._check_lines_are_lines(lines)
290
sha1 = sha_strings(lines)
291
if sha1 == nostore_sha:
292
raise errors.ExistingContent
293
if version_id in self._name_map:
294
return self._check_repeated_add(version_id, parents, lines, sha1)
296
self._check_versions(parents)
297
## self._check_lines(lines)
298
new_version = len(self._parents)
300
# if we abort after here the (in-memory) weave will be corrupt because only
301
# some fields are updated
302
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
303
# - Robert Collins 20060226
304
self._parents.append(parents[:])
305
self._sha1s.append(sha1)
306
self._names.append(version_id)
307
self._name_map[version_id] = new_version
311
# special case; adding with no parents revision; can do
312
# this more quickly by just appending unconditionally.
313
# even more specially, if we're adding an empty text we
314
# need do nothing at all.
316
self._weave.append(('{', new_version))
317
self._weave.extend(lines)
318
self._weave.append(('}', None))
321
if len(parents) == 1:
322
pv = list(parents)[0]
323
if sha1 == self._sha1s[pv]:
324
# special case: same as the single parent
328
ancestors = self._inclusions(parents)
332
# basis a list of (origin, lineno, line)
335
for origin, lineno, line in self._extract(ancestors):
336
basis_lineno.append(lineno)
337
basis_lines.append(line)
339
# another small special case: a merge, producing the same text
341
if lines == basis_lines:
344
# add a sentinel, because we can also match against the final line
345
basis_lineno.append(len(self._weave))
347
# XXX: which line of the weave should we really consider
348
# matches the end of the file? the current code says it's the
349
# last line of the weave?
351
#print 'basis_lines:', basis_lines
352
#print 'new_lines: ', lines
354
s = self._matcher(None, basis_lines, lines)
356
# offset gives the number of lines that have been inserted
357
# into the weave up to the current point; if the original edit instruction
358
# says to change line A then we actually change (A+offset)
361
for tag, i1, i2, j1, j2 in s.get_opcodes():
362
# i1,i2 are given in offsets within basis_lines; we need to map them
363
# back to offsets within the entire weave
364
#print 'raw match', tag, i1, i2, j1, j2
368
i1 = basis_lineno[i1]
369
i2 = basis_lineno[i2]
371
assert 0 <= j1 <= j2 <= len(lines)
373
#print tag, i1, i2, j1, j2
375
# the deletion and insertion are handled separately.
376
# first delete the region.
378
self._weave.insert(i1+offset, ('[', new_version))
379
self._weave.insert(i2+offset+1, (']', new_version))
383
# there may have been a deletion spanning up to
384
# i2; we want to insert after this region to make sure
385
# we don't destroy ourselves
387
self._weave[i:i] = ([('{', new_version)]
390
offset += 2 + (j2 - j1)
393
def _clone_text(self, new_version_id, old_version_id, parents):
394
"""See VersionedFile.clone_text."""
395
old_lines = self.get_text(old_version_id)
396
self.add_lines(new_version_id, parents, old_lines)
398
def _inclusions(self, versions):
399
"""Return set of all ancestors of given version(s)."""
400
if not len(versions):
403
for v in xrange(max(versions), 0, -1):
405
# include all its parents
406
i.update(self._parents[v])
408
## except IndexError:
409
## raise ValueError("version %d not present in weave" % v)
411
def get_ancestry(self, version_ids, topo_sorted=True):
412
"""See VersionedFile.get_ancestry."""
413
if isinstance(version_ids, basestring):
414
version_ids = [version_ids]
415
i = self._inclusions([self._lookup(v) for v in version_ids])
416
return [self._idx_to_name(v) for v in i]
418
def _check_lines(self, text):
419
if not isinstance(text, list):
420
raise ValueError("text should be a list, not %s" % type(text))
423
if not isinstance(l, basestring):
424
raise ValueError("text line should be a string or unicode, not %s"
429
def _check_versions(self, indexes):
430
"""Check everything in the sequence of indexes is valid"""
435
raise IndexError("invalid version number %r" % i)
437
def _compatible_parents(self, my_parents, other_parents):
438
"""During join check that other_parents are joinable with my_parents.
440
Joinable is defined as 'is a subset of' - supersets may require
441
regeneration of diffs, but subsets do not.
443
return len(other_parents.difference(my_parents)) == 0
445
def annotate_iter(self, version_id):
446
"""Yield list of (version-id, line) pairs for the specified version.
448
The index indicates when the line originated in the weave."""
449
incls = [self._lookup(version_id)]
450
for origin, lineno, text in self._extract(incls):
451
yield self._idx_to_name(origin), text
453
def iter_lines_added_or_present_in_versions(self, version_ids=None,
455
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
456
if version_ids is None:
457
version_ids = self.versions()
458
version_ids = set(version_ids)
459
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
460
# if inserted not in version_ids then it was inserted before the
461
# versions we care about, but because weaves cannot represent ghosts
462
# properly, we do not filter down to that
463
# if inserted not in version_ids: continue
469
def _walk_internal(self, version_ids=None):
470
"""Helper method for weave actions."""
475
lineno = 0 # line of weave, 0-based
477
for l in self._weave:
478
if l.__class__ == tuple:
482
istack.append(self._names[v])
486
assert self._names[v] not in dset
487
dset.add(self._names[v])
489
dset.remove(self._names[v])
491
raise WeaveFormatError('unexpected instruction %r' % v)
493
assert l.__class__ in (str, unicode)
495
yield lineno, istack[-1], frozenset(dset), l
499
raise WeaveFormatError("unclosed insertion blocks "
500
"at end of weave: %s" % istack)
502
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
505
def plan_merge(self, ver_a, ver_b):
506
"""Return pseudo-annotation indicating how the two versions merge.
508
This is computed between versions a and b and their common
511
Weave lines present in none of them are skipped entirely.
513
inc_a = set(self.get_ancestry([ver_a]))
514
inc_b = set(self.get_ancestry([ver_b]))
515
inc_c = inc_a & inc_b
517
for lineno, insert, deleteset, line in self._walk_internal([ver_a, ver_b]):
518
if deleteset & inc_c:
519
# killed in parent; can't be in either a or b
520
# not relevant to our work
521
yield 'killed-base', line
522
elif insert in inc_c:
523
# was inserted in base
524
killed_a = bool(deleteset & inc_a)
525
killed_b = bool(deleteset & inc_b)
526
if killed_a and killed_b:
527
yield 'killed-both', line
529
yield 'killed-a', line
531
yield 'killed-b', line
533
yield 'unchanged', line
534
elif insert in inc_a:
535
if deleteset & inc_a:
536
yield 'ghost-a', line
540
elif insert in inc_b:
541
if deleteset & inc_b:
542
yield 'ghost-b', line
546
# not in either revision
547
yield 'irrelevant', line
549
yield 'unchanged', '' # terminator
551
def _extract(self, versions):
552
"""Yield annotation of lines in included set.
554
Yields a sequence of tuples (origin, lineno, text), where
555
origin is the origin version, lineno the index in the weave,
556
and text the text of the line.
558
The set typically but not necessarily corresponds to a version.
561
if not isinstance(i, int):
564
included = self._inclusions(versions)
570
lineno = 0 # line of weave, 0-based
576
WFE = WeaveFormatError
579
# 449 0 4474.6820 2356.5590 bzrlib.weave:556(_extract)
580
# +285282 0 1676.8040 1676.8040 +<isinstance>
581
# 1.6 seconds in 'isinstance'.
582
# changing the first isinstance:
583
# 449 0 2814.2660 1577.1760 bzrlib.weave:556(_extract)
584
# +140414 0 762.8050 762.8050 +<isinstance>
585
# note that the inline time actually dropped (less function calls)
586
# and total processing time was halved.
587
# we're still spending ~1/4 of the method in isinstance though.
588
# so lets hard code the acceptable string classes we expect:
589
# 449 0 1202.9420 786.2930 bzrlib.weave:556(_extract)
590
# +71352 0 377.5560 377.5560 +<method 'append' of 'list'
592
# yay, down to ~1/4 the initial extract time, and our inline time
593
# has shrunk again, with isinstance no longer dominating.
594
# tweaking the stack inclusion test to use a set gives:
595
# 449 0 1122.8030 713.0080 bzrlib.weave:556(_extract)
596
# +71352 0 354.9980 354.9980 +<method 'append' of 'list'
598
# - a 5% win, or possibly just noise. However with large istacks that
599
# 'in' test could dominate, so I'm leaving this change in place -
600
# when its fast enough to consider profiling big datasets we can review.
605
for l in self._weave:
606
if l.__class__ == tuple:
614
iset.remove(istack.pop())
625
assert l.__class__ in (str, unicode)
627
isactive = (not dset) and istack and (istack[-1] in included)
629
result.append((istack[-1], lineno, l))
632
raise WeaveFormatError("unclosed insertion blocks "
633
"at end of weave: %s" % istack)
635
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
639
def _maybe_lookup(self, name_or_index):
640
"""Convert possible symbolic name to index, or pass through indexes.
644
if isinstance(name_or_index, (int, long)):
647
return self._lookup(name_or_index)
649
def get_lines(self, version_id):
650
"""See VersionedFile.get_lines()."""
651
int_index = self._maybe_lookup(version_id)
652
result = [line for (origin, lineno, line) in self._extract([int_index])]
653
expected_sha1 = self._sha1s[int_index]
654
measured_sha1 = sha_strings(result)
655
if measured_sha1 != expected_sha1:
656
raise errors.WeaveInvalidChecksum(
657
'file %s, revision %s, expected: %s, measured %s'
658
% (self._weave_name, version_id,
659
expected_sha1, measured_sha1))
662
def get_sha1(self, version_id):
663
"""See VersionedFile.get_sha1()."""
664
return self._sha1s[self._lookup(version_id)]
666
def get_sha1s(self, version_ids):
667
"""See VersionedFile.get_sha1s()."""
668
return [self._sha1s[self._lookup(v)] for v in version_ids]
670
def num_versions(self):
671
"""How many versions are in this weave?"""
672
l = len(self._parents)
673
assert l == len(self._sha1s)
676
__len__ = num_versions
678
def check(self, progress_bar=None):
679
# TODO evaluate performance hit of using string sets in this routine.
680
# TODO: check no circular inclusions
681
# TODO: create a nested progress bar
682
for version in range(self.num_versions()):
683
inclusions = list(self._parents[version])
686
if inclusions[-1] >= version:
687
raise WeaveFormatError("invalid included version %d for index %d"
688
% (inclusions[-1], version))
690
# try extracting all versions; parallel extraction is used
691
nv = self.num_versions()
696
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
697
# The problem is that set membership is much more expensive
698
name = self._idx_to_name(i)
701
new_inc = set([name])
702
for p in self._parents[i]:
703
new_inc.update(inclusions[self._idx_to_name(p)])
705
assert set(new_inc) == set(self.get_ancestry(name)), \
706
'failed %s != %s' % (set(new_inc), set(self.get_ancestry(name)))
707
inclusions[name] = new_inc
709
nlines = len(self._weave)
711
update_text = 'checking weave'
713
short_name = os.path.basename(self._weave_name)
714
update_text = 'checking %s' % (short_name,)
715
update_text = update_text[:25]
717
for lineno, insert, deleteset, line in self._walk_internal():
719
progress_bar.update(update_text, lineno, nlines)
721
for name, name_inclusions in inclusions.items():
722
# The active inclusion must be an ancestor,
723
# and no ancestors must have deleted this line,
724
# because we don't support resurrection.
725
if (insert in name_inclusions) and not (deleteset & name_inclusions):
726
sha1s[name].update(line)
729
version = self._idx_to_name(i)
730
hd = sha1s[version].hexdigest()
731
expected = self._sha1s[i]
733
raise errors.WeaveInvalidChecksum(
734
"mismatched sha1 for version %s: "
735
"got %s, expected %s"
736
% (version, hd, expected))
738
# TODO: check insertions are properly nested, that there are
739
# no lines outside of insertion blocks, that deletions are
740
# properly paired, etc.
742
def _join(self, other, pb, msg, version_ids, ignore_missing):
743
"""Worker routine for join()."""
744
if not other.versions():
745
return # nothing to update, easy
748
# versions is never none, InterWeave checks this.
751
# two loops so that we do not change ourselves before verifying it
753
# work through in index order to make sure we get all dependencies
756
# get the selected versions only that are in other.versions.
757
version_ids = set(other.versions()).intersection(set(version_ids))
758
# pull in the referenced graph.
759
version_ids = other.get_ancestry(version_ids)
760
pending_graph = [(version, other.get_parents(version)) for
761
version in version_ids]
762
for name in topo_sort(pending_graph):
763
other_idx = other._name_map[name]
764
# returns True if we have it, False if we need it.
765
if not self._check_version_consistent(other, other_idx, name):
766
names_to_join.append((other_idx, name))
775
for other_idx, name in names_to_join:
776
# TODO: If all the parents of the other version are already
777
# present then we can avoid some work by just taking the delta
778
# and adjusting the offsets.
779
new_parents = self._imported_parents(other, other_idx)
780
sha1 = other._sha1s[other_idx]
785
pb.update(msg, merged, len(names_to_join))
787
lines = other.get_lines(other_idx)
788
self._add(name, lines, new_parents, sha1)
790
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
791
merged, processed, self._weave_name, time.time()-time0))
793
def _imported_parents(self, other, other_idx):
794
"""Return list of parents in self corresponding to indexes in other."""
796
for parent_idx in other._parents[other_idx]:
797
parent_name = other._names[parent_idx]
798
if parent_name not in self._name_map:
799
# should not be possible
800
raise WeaveError("missing parent {%s} of {%s} in %r"
801
% (parent_name, other._name_map[other_idx], self))
802
new_parents.append(self._name_map[parent_name])
805
def _check_version_consistent(self, other, other_idx, name):
806
"""Check if a version in consistent in this and other.
808
To be consistent it must have:
811
* the same direct parents (by name, not index, and disregarding
814
If present & correct return True;
815
if not present in self return False;
816
if inconsistent raise error."""
817
this_idx = self._name_map.get(name, -1)
819
if self._sha1s[this_idx] != other._sha1s[other_idx]:
820
raise errors.WeaveTextDiffers(name, self, other)
821
self_parents = self._parents[this_idx]
822
other_parents = other._parents[other_idx]
823
n1 = set([self._names[i] for i in self_parents])
824
n2 = set([other._names[i] for i in other_parents])
825
if not self._compatible_parents(n1, n2):
826
raise WeaveParentMismatch("inconsistent parents "
827
"for version {%s}: %s vs %s" % (name, n1, n2))
833
def _reweave(self, other, pb, msg):
834
"""Reweave self with other - internal helper for join().
836
:param other: The other weave to merge
837
:param pb: An optional progress bar, indicating how far done we are
838
:param msg: An optional message for the progress
840
new_weave = _reweave(self, other, pb=pb, msg=msg)
841
self._copy_weave_content(new_weave)
843
def _copy_weave_content(self, otherweave):
844
"""adsorb the content from otherweave."""
845
for attr in self.__slots__:
846
if attr != '_weave_name':
847
setattr(self, attr, copy(getattr(otherweave, attr)))
850
class WeaveFile(Weave):
851
"""A WeaveFile represents a Weave on disk and writes on change."""
853
WEAVE_SUFFIX = '.weave'
855
def __init__(self, name, transport, filemode=None, create=False, access_mode='w'):
856
"""Create a WeaveFile.
858
:param create: If not True, only open an existing knit.
860
super(WeaveFile, self).__init__(name, access_mode)
861
self._transport = transport
862
self._filemode = filemode
864
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
865
except errors.NoSuchFile:
871
def _add_lines(self, version_id, parents, lines, parent_texts,
872
left_matching_blocks, nostore_sha, random_id, check_content):
873
"""Add a version and save the weave."""
874
self.check_not_reserved_id(version_id)
875
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
876
parent_texts, left_matching_blocks, nostore_sha, random_id,
881
def _clone_text(self, new_version_id, old_version_id, parents):
882
"""See VersionedFile.clone_text."""
883
super(WeaveFile, self)._clone_text(new_version_id, old_version_id, parents)
886
def copy_to(self, name, transport):
887
"""See VersionedFile.copy_to()."""
888
# as we are all in memory always, just serialise to the new place.
890
write_weave_v5(self, sio)
892
transport.put_file(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
894
def create_empty(self, name, transport, filemode=None):
895
return WeaveFile(name, transport, filemode, create=True)
898
"""Save the weave."""
899
self._check_write_ok()
901
write_weave_v5(self, sio)
903
self._transport.put_file(self._weave_name + WeaveFile.WEAVE_SUFFIX,
909
"""See VersionedFile.get_suffixes()."""
910
return [WeaveFile.WEAVE_SUFFIX]
912
def join(self, other, pb=None, msg=None, version_ids=None,
913
ignore_missing=False):
914
"""Join other into self and save."""
915
super(WeaveFile, self).join(other, pb, msg, version_ids, ignore_missing)
919
def _reweave(wa, wb, pb=None, msg=None):
920
"""Combine two weaves and return the result.
922
This works even if a revision R has different parents in
923
wa and wb. In the resulting weave all the parents are given.
925
This is done by just building up a new weave, maintaining ordering
926
of the versions in the two inputs. More efficient approaches
927
might be possible but it should only be necessary to do
928
this operation rarely, when a new previously ghost version is
931
:param pb: An optional progress bar, indicating how far done we are
932
:param msg: An optional message for the progress
936
queue_a = range(wa.num_versions())
937
queue_b = range(wb.num_versions())
938
# first determine combined parents of all versions
939
# map from version name -> all parent names
940
combined_parents = _reweave_parent_graphs(wa, wb)
941
mutter("combined parents: %r", combined_parents)
942
order = topo_sort(combined_parents.iteritems())
943
mutter("order to reweave: %r", order)
948
for idx, name in enumerate(order):
950
pb.update(msg, idx, len(order))
951
if name in wa._name_map:
952
lines = wa.get_lines(name)
953
if name in wb._name_map:
954
lines_b = wb.get_lines(name)
956
mutter('Weaves differ on content. rev_id {%s}', name)
957
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
959
lines = list(difflib.unified_diff(lines, lines_b,
960
wa._weave_name, wb._weave_name))
961
mutter('lines:\n%s', ''.join(lines))
962
raise errors.WeaveTextDiffers(name, wa, wb)
964
lines = wb.get_lines(name)
965
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
968
def _reweave_parent_graphs(wa, wb):
969
"""Return combined parent ancestry for two weaves.
971
Returned as a list of (version_name, set(parent_names))"""
973
for weave in [wa, wb]:
974
for idx, name in enumerate(weave._names):
975
p = combined.setdefault(name, set())
976
p.update(map(weave._idx_to_name, weave._parents[idx]))
981
"""Show the weave's table-of-contents"""
982
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
983
for i in (6, 50, 10, 10):
986
for i in range(w.num_versions()):
989
parent_str = ' '.join(map(str, w._parents[i]))
990
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
994
def weave_stats(weave_file, pb):
995
from bzrlib.weavefile import read_weave
997
wf = file(weave_file, 'rb')
999
# FIXME: doesn't work on pipes
1000
weave_size = wf.tell()
1004
for i in range(vers):
1005
pb.update('checking sizes', i, vers)
1006
for origin, lineno, line in w._extract([i]):
1011
print 'versions %9d' % vers
1012
print 'weave file %9d bytes' % weave_size
1013
print 'total contents %9d bytes' % total
1014
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1017
print 'average size %9d bytes' % avg
1018
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1022
print """bzr weave tool
1024
Experimental tool for weave algorithm.
1027
weave init WEAVEFILE
1028
Create an empty weave file
1029
weave get WEAVEFILE VERSION
1030
Write out specified version.
1031
weave check WEAVEFILE
1032
Check consistency of all versions.
1034
Display table of contents.
1035
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1036
Add NEWTEXT, with specified parent versions.
1037
weave annotate WEAVEFILE VERSION
1038
Display origin of each line.
1039
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1040
Auto-merge two versions and display conflicts.
1041
weave diff WEAVEFILE VERSION1 VERSION2
1042
Show differences between two versions.
1046
% weave init foo.weave
1048
% weave add foo.weave ver0 < foo.txt
1051
(create updated version)
1053
% weave get foo.weave 0 | diff -u - foo.txt
1054
% weave add foo.weave ver1 0 < foo.txt
1057
% weave get foo.weave 0 > foo.txt (create forked version)
1059
% weave add foo.weave ver2 0 < foo.txt
1062
% weave merge foo.weave 1 2 > foo.txt (merge them)
1063
% vi foo.txt (resolve conflicts)
1064
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1076
# in case we're run directly from the subdirectory
1077
sys.path.append('..')
1079
from bzrlib.weavefile import write_weave, read_weave
1080
from bzrlib.progress import ProgressBar
1095
return read_weave(file(argv[2], 'rb'))
1101
# at the moment, based on everything in the file
1103
parents = map(int, argv[4:])
1104
lines = sys.stdin.readlines()
1105
ver = w.add(name, parents, lines)
1106
write_weave(w, file(argv[2], 'wb'))
1107
print 'added version %r %d' % (name, ver)
1110
if os.path.exists(fn):
1111
raise IOError("file exists")
1113
write_weave(w, file(fn, 'wb'))
1114
elif cmd == 'get': # get one version
1116
sys.stdout.writelines(w.get_iter(int(argv[3])))
1121
v1, v2 = map(int, argv[3:5])
1124
diff_gen = bzrlib.patiencediff.unified_diff(lines1, lines2,
1125
'%s version %d' % (fn, v1),
1126
'%s version %d' % (fn, v2))
1127
sys.stdout.writelines(diff_gen)
1129
elif cmd == 'annotate':
1131
# newline is added to all lines regardless; too hard to get
1132
# reasonable formatting otherwise
1134
for origin, text in w.annotate(int(argv[3])):
1135
text = text.rstrip('\r\n')
1137
print ' | %s' % (text)
1139
print '%5d | %s' % (origin, text)
1145
elif cmd == 'stats':
1146
weave_stats(argv[2], ProgressBar())
1148
elif cmd == 'check':
1153
print '%d versions ok' % w.num_versions()
1155
elif cmd == 'inclusions':
1157
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1159
elif cmd == 'parents':
1161
print ' '.join(map(str, w._parents[int(argv[3])]))
1163
elif cmd == 'plan-merge':
1164
# replaced by 'bzr weave-plan-merge'
1166
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1168
print '%14s | %s' % (state, line),
1169
elif cmd == 'merge':
1170
# replaced by 'bzr weave-merge-text'
1172
p = w.plan_merge(int(argv[3]), int(argv[4]))
1173
sys.stdout.writelines(w.weave_merge(p))
1175
raise ValueError('unknown command %r' % cmd)
1178
if __name__ == '__main__':
1180
sys.exit(main(sys.argv))
1183
class InterWeave(InterVersionedFile):
1184
"""Optimised code paths for weave to weave operations."""
1186
_matching_file_from_factory = staticmethod(WeaveFile)
1187
_matching_file_to_factory = staticmethod(WeaveFile)
1190
def is_compatible(source, target):
1191
"""Be compatible with weaves."""
1193
return (isinstance(source, Weave) and
1194
isinstance(target, Weave))
1195
except AttributeError:
1198
def join(self, pb=None, msg=None, version_ids=None, ignore_missing=False):
1199
"""See InterVersionedFile.join."""
1200
version_ids = self._get_source_version_ids(version_ids, ignore_missing)
1201
if self.target.versions() == [] and version_ids is None:
1202
self.target._copy_weave_content(self.source)
1204
self.target._join(self.source, pb, msg, version_ids, ignore_missing)
1207
InterVersionedFile.register_optimiser(InterWeave)
added
removed
bzrlib/weave.py
