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_parents(self, version_id):
252
"""See VersionedFile.get_parent."""
253
return map(self._idx_to_name, self._parents[self._lookup(version_id)])
255
def _check_repeated_add(self, name, parents, text, sha1):
256
"""Check that a duplicated add is OK.
258
If it is, return the (old) index; otherwise raise an exception.
260
idx = self._lookup(name)
261
if sorted(self._parents[idx]) != sorted(parents) \
262
or sha1 != self._sha1s[idx]:
263
raise RevisionAlreadyPresent(name, self._weave_name)
266
def _add_lines(self, version_id, parents, lines, parent_texts,
267
left_matching_blocks, nostore_sha, random_id, check_content):
268
"""See VersionedFile.add_lines."""
269
idx = self._add(version_id, lines, map(self._lookup, parents),
270
nostore_sha=nostore_sha)
271
return sha_strings(lines), sum(map(len, lines)), idx
273
def _add(self, version_id, lines, parents, sha1=None, nostore_sha=None):
274
"""Add a single text on top of the weave.
276
Returns the index number of the newly added version.
279
Symbolic name for this version.
280
(Typically the revision-id of the revision that added it.)
283
List or set of direct parent version numbers.
286
Sequence of lines to be added in the new version.
288
:param nostore_sha: See VersionedFile.add_lines.
290
assert isinstance(version_id, basestring)
291
self._check_lines_not_unicode(lines)
292
self._check_lines_are_lines(lines)
294
sha1 = sha_strings(lines)
295
if sha1 == nostore_sha:
296
raise errors.ExistingContent
297
if version_id in self._name_map:
298
return self._check_repeated_add(version_id, parents, lines, sha1)
300
self._check_versions(parents)
301
## self._check_lines(lines)
302
new_version = len(self._parents)
304
# if we abort after here the (in-memory) weave will be corrupt because only
305
# some fields are updated
306
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
307
# - Robert Collins 20060226
308
self._parents.append(parents[:])
309
self._sha1s.append(sha1)
310
self._names.append(version_id)
311
self._name_map[version_id] = new_version
315
# special case; adding with no parents revision; can do
316
# this more quickly by just appending unconditionally.
317
# even more specially, if we're adding an empty text we
318
# need do nothing at all.
320
self._weave.append(('{', new_version))
321
self._weave.extend(lines)
322
self._weave.append(('}', None))
325
if len(parents) == 1:
326
pv = list(parents)[0]
327
if sha1 == self._sha1s[pv]:
328
# special case: same as the single parent
332
ancestors = self._inclusions(parents)
336
# basis a list of (origin, lineno, line)
339
for origin, lineno, line in self._extract(ancestors):
340
basis_lineno.append(lineno)
341
basis_lines.append(line)
343
# another small special case: a merge, producing the same text
345
if lines == basis_lines:
348
# add a sentinel, because we can also match against the final line
349
basis_lineno.append(len(self._weave))
351
# XXX: which line of the weave should we really consider
352
# matches the end of the file? the current code says it's the
353
# last line of the weave?
355
#print 'basis_lines:', basis_lines
356
#print 'new_lines: ', lines
358
s = self._matcher(None, basis_lines, lines)
360
# offset gives the number of lines that have been inserted
361
# into the weave up to the current point; if the original edit instruction
362
# says to change line A then we actually change (A+offset)
365
for tag, i1, i2, j1, j2 in s.get_opcodes():
366
# i1,i2 are given in offsets within basis_lines; we need to map them
367
# back to offsets within the entire weave
368
#print 'raw match', tag, i1, i2, j1, j2
372
i1 = basis_lineno[i1]
373
i2 = basis_lineno[i2]
375
assert 0 <= j1 <= j2 <= len(lines)
377
#print tag, i1, i2, j1, j2
379
# the deletion and insertion are handled separately.
380
# first delete the region.
382
self._weave.insert(i1+offset, ('[', new_version))
383
self._weave.insert(i2+offset+1, (']', new_version))
387
# there may have been a deletion spanning up to
388
# i2; we want to insert after this region to make sure
389
# we don't destroy ourselves
391
self._weave[i:i] = ([('{', new_version)]
394
offset += 2 + (j2 - j1)
397
def _clone_text(self, new_version_id, old_version_id, parents):
398
"""See VersionedFile.clone_text."""
399
old_lines = self.get_text(old_version_id)
400
self.add_lines(new_version_id, parents, old_lines)
402
def _inclusions(self, versions):
403
"""Return set of all ancestors of given version(s)."""
404
if not len(versions):
407
for v in xrange(max(versions), 0, -1):
409
# include all its parents
410
i.update(self._parents[v])
412
## except IndexError:
413
## raise ValueError("version %d not present in weave" % v)
415
def get_ancestry(self, version_ids, topo_sorted=True):
416
"""See VersionedFile.get_ancestry."""
417
if isinstance(version_ids, basestring):
418
version_ids = [version_ids]
419
i = self._inclusions([self._lookup(v) for v in version_ids])
420
return [self._idx_to_name(v) for v in i]
422
def _check_lines(self, text):
423
if not isinstance(text, list):
424
raise ValueError("text should be a list, not %s" % type(text))
427
if not isinstance(l, basestring):
428
raise ValueError("text line should be a string or unicode, not %s"
433
def _check_versions(self, indexes):
434
"""Check everything in the sequence of indexes is valid"""
439
raise IndexError("invalid version number %r" % i)
441
def _compatible_parents(self, my_parents, other_parents):
442
"""During join check that other_parents are joinable with my_parents.
444
Joinable is defined as 'is a subset of' - supersets may require
445
regeneration of diffs, but subsets do not.
447
return len(other_parents.difference(my_parents)) == 0
449
def annotate_iter(self, version_id):
450
"""Yield list of (version-id, line) pairs for the specified version.
452
The index indicates when the line originated in the weave."""
453
incls = [self._lookup(version_id)]
454
for origin, lineno, text in self._extract(incls):
455
yield self._idx_to_name(origin), text
457
def iter_lines_added_or_present_in_versions(self, version_ids=None,
459
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
460
if version_ids is None:
461
version_ids = self.versions()
462
version_ids = set(version_ids)
463
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
464
# if inserted not in version_ids then it was inserted before the
465
# versions we care about, but because weaves cannot represent ghosts
466
# properly, we do not filter down to that
467
# if inserted not in version_ids: continue
469
yield line + '\n', inserted
473
def _walk_internal(self, version_ids=None):
474
"""Helper method for weave actions."""
479
lineno = 0 # line of weave, 0-based
481
for l in self._weave:
482
if l.__class__ == tuple:
486
istack.append(self._names[v])
490
assert self._names[v] not in dset
491
dset.add(self._names[v])
493
dset.remove(self._names[v])
495
raise WeaveFormatError('unexpected instruction %r' % v)
497
assert l.__class__ in (str, unicode)
499
yield lineno, istack[-1], frozenset(dset), l
503
raise WeaveFormatError("unclosed insertion blocks "
504
"at end of weave: %s" % istack)
506
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
509
def plan_merge(self, ver_a, ver_b):
510
"""Return pseudo-annotation indicating how the two versions merge.
512
This is computed between versions a and b and their common
515
Weave lines present in none of them are skipped entirely.
517
inc_a = set(self.get_ancestry([ver_a]))
518
inc_b = set(self.get_ancestry([ver_b]))
519
inc_c = inc_a & inc_b
521
for lineno, insert, deleteset, line in self._walk_internal([ver_a, ver_b]):
522
if deleteset & inc_c:
523
# killed in parent; can't be in either a or b
524
# not relevant to our work
525
yield 'killed-base', line
526
elif insert in inc_c:
527
# was inserted in base
528
killed_a = bool(deleteset & inc_a)
529
killed_b = bool(deleteset & inc_b)
530
if killed_a and killed_b:
531
yield 'killed-both', line
533
yield 'killed-a', line
535
yield 'killed-b', line
537
yield 'unchanged', line
538
elif insert in inc_a:
539
if deleteset & inc_a:
540
yield 'ghost-a', line
544
elif insert in inc_b:
545
if deleteset & inc_b:
546
yield 'ghost-b', line
550
# not in either revision
551
yield 'irrelevant', line
553
yield 'unchanged', '' # terminator
555
def _extract(self, versions):
556
"""Yield annotation of lines in included set.
558
Yields a sequence of tuples (origin, lineno, text), where
559
origin is the origin version, lineno the index in the weave,
560
and text the text of the line.
562
The set typically but not necessarily corresponds to a version.
565
if not isinstance(i, int):
568
included = self._inclusions(versions)
574
lineno = 0 # line of weave, 0-based
580
WFE = WeaveFormatError
583
# 449 0 4474.6820 2356.5590 bzrlib.weave:556(_extract)
584
# +285282 0 1676.8040 1676.8040 +<isinstance>
585
# 1.6 seconds in 'isinstance'.
586
# changing the first isinstance:
587
# 449 0 2814.2660 1577.1760 bzrlib.weave:556(_extract)
588
# +140414 0 762.8050 762.8050 +<isinstance>
589
# note that the inline time actually dropped (less function calls)
590
# and total processing time was halved.
591
# we're still spending ~1/4 of the method in isinstance though.
592
# so lets hard code the acceptable string classes we expect:
593
# 449 0 1202.9420 786.2930 bzrlib.weave:556(_extract)
594
# +71352 0 377.5560 377.5560 +<method 'append' of 'list'
596
# yay, down to ~1/4 the initial extract time, and our inline time
597
# has shrunk again, with isinstance no longer dominating.
598
# tweaking the stack inclusion test to use a set gives:
599
# 449 0 1122.8030 713.0080 bzrlib.weave:556(_extract)
600
# +71352 0 354.9980 354.9980 +<method 'append' of 'list'
602
# - a 5% win, or possibly just noise. However with large istacks that
603
# 'in' test could dominate, so I'm leaving this change in place -
604
# when its fast enough to consider profiling big datasets we can review.
609
for l in self._weave:
610
if l.__class__ == tuple:
618
iset.remove(istack.pop())
629
assert l.__class__ in (str, unicode)
631
isactive = (not dset) and istack and (istack[-1] in included)
633
result.append((istack[-1], lineno, l))
636
raise WeaveFormatError("unclosed insertion blocks "
637
"at end of weave: %s" % istack)
639
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
643
def _maybe_lookup(self, name_or_index):
644
"""Convert possible symbolic name to index, or pass through indexes.
648
if isinstance(name_or_index, (int, long)):
651
return self._lookup(name_or_index)
653
def get_lines(self, version_id):
654
"""See VersionedFile.get_lines()."""
655
int_index = self._maybe_lookup(version_id)
656
result = [line for (origin, lineno, line) in self._extract([int_index])]
657
expected_sha1 = self._sha1s[int_index]
658
measured_sha1 = sha_strings(result)
659
if measured_sha1 != expected_sha1:
660
raise errors.WeaveInvalidChecksum(
661
'file %s, revision %s, expected: %s, measured %s'
662
% (self._weave_name, version_id,
663
expected_sha1, measured_sha1))
666
def get_sha1(self, version_id):
667
"""See VersionedFile.get_sha1()."""
668
return self._sha1s[self._lookup(version_id)]
670
def get_sha1s(self, version_ids):
671
"""See VersionedFile.get_sha1s()."""
672
return [self._sha1s[self._lookup(v)] for v in version_ids]
674
def num_versions(self):
675
"""How many versions are in this weave?"""
676
l = len(self._parents)
677
assert l == len(self._sha1s)
680
__len__ = num_versions
682
def check(self, progress_bar=None):
683
# TODO evaluate performance hit of using string sets in this routine.
684
# TODO: check no circular inclusions
685
# TODO: create a nested progress bar
686
for version in range(self.num_versions()):
687
inclusions = list(self._parents[version])
690
if inclusions[-1] >= version:
691
raise WeaveFormatError("invalid included version %d for index %d"
692
% (inclusions[-1], version))
694
# try extracting all versions; parallel extraction is used
695
nv = self.num_versions()
700
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
701
# The problem is that set membership is much more expensive
702
name = self._idx_to_name(i)
703
sha1s[name] = sha.new()
705
new_inc = set([name])
706
for p in self._parents[i]:
707
new_inc.update(inclusions[self._idx_to_name(p)])
709
assert set(new_inc) == set(self.get_ancestry(name)), \
710
'failed %s != %s' % (set(new_inc), set(self.get_ancestry(name)))
711
inclusions[name] = new_inc
713
nlines = len(self._weave)
715
update_text = 'checking weave'
717
short_name = os.path.basename(self._weave_name)
718
update_text = 'checking %s' % (short_name,)
719
update_text = update_text[:25]
721
for lineno, insert, deleteset, line in self._walk_internal():
723
progress_bar.update(update_text, lineno, nlines)
725
for name, name_inclusions in inclusions.items():
726
# The active inclusion must be an ancestor,
727
# and no ancestors must have deleted this line,
728
# because we don't support resurrection.
729
if (insert in name_inclusions) and not (deleteset & name_inclusions):
730
sha1s[name].update(line)
733
version = self._idx_to_name(i)
734
hd = sha1s[version].hexdigest()
735
expected = self._sha1s[i]
737
raise errors.WeaveInvalidChecksum(
738
"mismatched sha1 for version %s: "
739
"got %s, expected %s"
740
% (version, hd, expected))
742
# TODO: check insertions are properly nested, that there are
743
# no lines outside of insertion blocks, that deletions are
744
# properly paired, etc.
746
def _join(self, other, pb, msg, version_ids, ignore_missing):
747
"""Worker routine for join()."""
748
if not other.versions():
749
return # nothing to update, easy
752
# versions is never none, InterWeave checks this.
755
# two loops so that we do not change ourselves before verifying it
757
# work through in index order to make sure we get all dependencies
760
# get the selected versions only that are in other.versions.
761
version_ids = set(other.versions()).intersection(set(version_ids))
762
# pull in the referenced graph.
763
version_ids = other.get_ancestry(version_ids)
764
pending_graph = [(version, other.get_parents(version)) for
765
version in version_ids]
766
for name in tsort.topo_sort(pending_graph):
767
other_idx = other._name_map[name]
768
# returns True if we have it, False if we need it.
769
if not self._check_version_consistent(other, other_idx, name):
770
names_to_join.append((other_idx, name))
779
for other_idx, name in names_to_join:
780
# TODO: If all the parents of the other version are already
781
# present then we can avoid some work by just taking the delta
782
# and adjusting the offsets.
783
new_parents = self._imported_parents(other, other_idx)
784
sha1 = other._sha1s[other_idx]
789
pb.update(msg, merged, len(names_to_join))
791
lines = other.get_lines(other_idx)
792
self._add(name, lines, new_parents, sha1)
794
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
795
merged, processed, self._weave_name, time.time()-time0))
797
def _imported_parents(self, other, other_idx):
798
"""Return list of parents in self corresponding to indexes in other."""
800
for parent_idx in other._parents[other_idx]:
801
parent_name = other._names[parent_idx]
802
if parent_name not in self._name_map:
803
# should not be possible
804
raise WeaveError("missing parent {%s} of {%s} in %r"
805
% (parent_name, other._name_map[other_idx], self))
806
new_parents.append(self._name_map[parent_name])
809
def _check_version_consistent(self, other, other_idx, name):
810
"""Check if a version in consistent in this and other.
812
To be consistent it must have:
815
* the same direct parents (by name, not index, and disregarding
818
If present & correct return True;
819
if not present in self return False;
820
if inconsistent raise error."""
821
this_idx = self._name_map.get(name, -1)
823
if self._sha1s[this_idx] != other._sha1s[other_idx]:
824
raise errors.WeaveTextDiffers(name, self, other)
825
self_parents = self._parents[this_idx]
826
other_parents = other._parents[other_idx]
827
n1 = set([self._names[i] for i in self_parents])
828
n2 = set([other._names[i] for i in other_parents])
829
if not self._compatible_parents(n1, n2):
830
raise WeaveParentMismatch("inconsistent parents "
831
"for version {%s}: %s vs %s" % (name, n1, n2))
837
def _reweave(self, other, pb, msg):
838
"""Reweave self with other - internal helper for join().
840
:param other: The other weave to merge
841
:param pb: An optional progress bar, indicating how far done we are
842
:param msg: An optional message for the progress
844
new_weave = _reweave(self, other, pb=pb, msg=msg)
845
self._copy_weave_content(new_weave)
847
def _copy_weave_content(self, otherweave):
848
"""adsorb the content from otherweave."""
849
for attr in self.__slots__:
850
if attr != '_weave_name':
851
setattr(self, attr, copy(getattr(otherweave, attr)))
854
class WeaveFile(Weave):
855
"""A WeaveFile represents a Weave on disk and writes on change."""
857
WEAVE_SUFFIX = '.weave'
859
def __init__(self, name, transport, filemode=None, create=False, access_mode='w'):
860
"""Create a WeaveFile.
862
:param create: If not True, only open an existing knit.
864
super(WeaveFile, self).__init__(name, access_mode)
865
self._transport = transport
866
self._filemode = filemode
868
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
869
except errors.NoSuchFile:
875
def _add_lines(self, version_id, parents, lines, parent_texts,
876
left_matching_blocks, nostore_sha, random_id, check_content):
877
"""Add a version and save the weave."""
878
self.check_not_reserved_id(version_id)
879
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
880
parent_texts, left_matching_blocks, nostore_sha, random_id,
885
def _clone_text(self, new_version_id, old_version_id, parents):
886
"""See VersionedFile.clone_text."""
887
super(WeaveFile, self)._clone_text(new_version_id, old_version_id, parents)
890
def copy_to(self, name, transport):
891
"""See VersionedFile.copy_to()."""
892
# as we are all in memory always, just serialise to the new place.
894
write_weave_v5(self, sio)
896
transport.put_file(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
898
def create_empty(self, name, transport, filemode=None):
899
return WeaveFile(name, transport, filemode, create=True)
902
"""Save the weave."""
903
self._check_write_ok()
905
write_weave_v5(self, sio)
907
self._transport.put_file(self._weave_name + WeaveFile.WEAVE_SUFFIX,
913
"""See VersionedFile.get_suffixes()."""
914
return [WeaveFile.WEAVE_SUFFIX]
916
def join(self, other, pb=None, msg=None, version_ids=None,
917
ignore_missing=False):
918
"""Join other into self and save."""
919
super(WeaveFile, self).join(other, pb, msg, version_ids, ignore_missing)
923
def _reweave(wa, wb, pb=None, msg=None):
924
"""Combine two weaves and return the result.
926
This works even if a revision R has different parents in
927
wa and wb. In the resulting weave all the parents are given.
929
This is done by just building up a new weave, maintaining ordering
930
of the versions in the two inputs. More efficient approaches
931
might be possible but it should only be necessary to do
932
this operation rarely, when a new previously ghost version is
935
:param pb: An optional progress bar, indicating how far done we are
936
:param msg: An optional message for the progress
940
queue_a = range(wa.num_versions())
941
queue_b = range(wb.num_versions())
942
# first determine combined parents of all versions
943
# map from version name -> all parent names
944
combined_parents = _reweave_parent_graphs(wa, wb)
945
mutter("combined parents: %r", combined_parents)
946
order = tsort.topo_sort(combined_parents.iteritems())
947
mutter("order to reweave: %r", order)
952
for idx, name in enumerate(order):
954
pb.update(msg, idx, len(order))
955
if name in wa._name_map:
956
lines = wa.get_lines(name)
957
if name in wb._name_map:
958
lines_b = wb.get_lines(name)
960
mutter('Weaves differ on content. rev_id {%s}', name)
961
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
963
lines = list(difflib.unified_diff(lines, lines_b,
964
wa._weave_name, wb._weave_name))
965
mutter('lines:\n%s', ''.join(lines))
966
raise errors.WeaveTextDiffers(name, wa, wb)
968
lines = wb.get_lines(name)
969
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
972
def _reweave_parent_graphs(wa, wb):
973
"""Return combined parent ancestry for two weaves.
975
Returned as a list of (version_name, set(parent_names))"""
977
for weave in [wa, wb]:
978
for idx, name in enumerate(weave._names):
979
p = combined.setdefault(name, set())
980
p.update(map(weave._idx_to_name, weave._parents[idx]))
985
"""Show the weave's table-of-contents"""
986
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
987
for i in (6, 50, 10, 10):
990
for i in range(w.num_versions()):
993
parent_str = ' '.join(map(str, w._parents[i]))
994
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
998
def weave_stats(weave_file, pb):
999
from bzrlib.weavefile import read_weave
1001
wf = file(weave_file, 'rb')
1003
# FIXME: doesn't work on pipes
1004
weave_size = wf.tell()
1008
for i in range(vers):
1009
pb.update('checking sizes', i, vers)
1010
for origin, lineno, line in w._extract([i]):
1015
print 'versions %9d' % vers
1016
print 'weave file %9d bytes' % weave_size
1017
print 'total contents %9d bytes' % total
1018
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1021
print 'average size %9d bytes' % avg
1022
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1026
print """bzr weave tool
1028
Experimental tool for weave algorithm.
1031
weave init WEAVEFILE
1032
Create an empty weave file
1033
weave get WEAVEFILE VERSION
1034
Write out specified version.
1035
weave check WEAVEFILE
1036
Check consistency of all versions.
1038
Display table of contents.
1039
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1040
Add NEWTEXT, with specified parent versions.
1041
weave annotate WEAVEFILE VERSION
1042
Display origin of each line.
1043
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1044
Auto-merge two versions and display conflicts.
1045
weave diff WEAVEFILE VERSION1 VERSION2
1046
Show differences between two versions.
1050
% weave init foo.weave
1052
% weave add foo.weave ver0 < foo.txt
1055
(create updated version)
1057
% weave get foo.weave 0 | diff -u - foo.txt
1058
% weave add foo.weave ver1 0 < foo.txt
1061
% weave get foo.weave 0 > foo.txt (create forked version)
1063
% weave add foo.weave ver2 0 < foo.txt
1066
% weave merge foo.weave 1 2 > foo.txt (merge them)
1067
% vi foo.txt (resolve conflicts)
1068
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1080
# in case we're run directly from the subdirectory
1081
sys.path.append('..')
1083
from bzrlib.weavefile import write_weave, read_weave
1084
from bzrlib.progress import ProgressBar
1099
return read_weave(file(argv[2], 'rb'))
1105
# at the moment, based on everything in the file
1107
parents = map(int, argv[4:])
1108
lines = sys.stdin.readlines()
1109
ver = w.add(name, parents, lines)
1110
write_weave(w, file(argv[2], 'wb'))
1111
print 'added version %r %d' % (name, ver)
1114
if os.path.exists(fn):
1115
raise IOError("file exists")
1117
write_weave(w, file(fn, 'wb'))
1118
elif cmd == 'get': # get one version
1120
sys.stdout.writelines(w.get_iter(int(argv[3])))
1125
v1, v2 = map(int, argv[3:5])
1128
diff_gen = bzrlib.patiencediff.unified_diff(lines1, lines2,
1129
'%s version %d' % (fn, v1),
1130
'%s version %d' % (fn, v2))
1131
sys.stdout.writelines(diff_gen)
1133
elif cmd == 'annotate':
1135
# newline is added to all lines regardless; too hard to get
1136
# reasonable formatting otherwise
1138
for origin, text in w.annotate(int(argv[3])):
1139
text = text.rstrip('\r\n')
1141
print ' | %s' % (text)
1143
print '%5d | %s' % (origin, text)
1149
elif cmd == 'stats':
1150
weave_stats(argv[2], ProgressBar())
1152
elif cmd == 'check':
1157
print '%d versions ok' % w.num_versions()
1159
elif cmd == 'inclusions':
1161
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1163
elif cmd == 'parents':
1165
print ' '.join(map(str, w._parents[int(argv[3])]))
1167
elif cmd == 'plan-merge':
1168
# replaced by 'bzr weave-plan-merge'
1170
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1172
print '%14s | %s' % (state, line),
1173
elif cmd == 'merge':
1174
# replaced by 'bzr weave-merge-text'
1176
p = w.plan_merge(int(argv[3]), int(argv[4]))
1177
sys.stdout.writelines(w.weave_merge(p))
1179
raise ValueError('unknown command %r' % cmd)
1182
if __name__ == '__main__':
1184
sys.exit(main(sys.argv))
1187
class InterWeave(InterVersionedFile):
1188
"""Optimised code paths for weave to weave operations."""
1190
_matching_file_from_factory = staticmethod(WeaveFile)
1191
_matching_file_to_factory = staticmethod(WeaveFile)
1194
def is_compatible(source, target):
1195
"""Be compatible with weaves."""
1197
return (isinstance(source, Weave) and
1198
isinstance(target, Weave))
1199
except AttributeError:
1202
def join(self, pb=None, msg=None, version_ids=None, ignore_missing=False):
1203
"""See InterVersionedFile.join."""
1204
version_ids = self._get_source_version_ids(version_ids, ignore_missing)
1205
if self.target.versions() == [] and version_ids is None:
1206
self.target._copy_weave_content(self.source)
1208
self.target._join(self.source, pb, msg, version_ids, ignore_missing)
1211
InterVersionedFile.register_optimiser(InterWeave)
added
removed
bzrlib/weave.py
