3
# Copyright (C) 2005 Canonical Ltd
5
# This program is free software; you can redistribute it and/or modify
6
# it under the terms of the GNU General Public License as published by
7
# the Free Software Foundation; either version 2 of the License, or
8
# (at your option) any later version.
10
# This program is distributed in the hope that it will be useful,
11
# but WITHOUT ANY WARRANTY; without even the implied warranty of
12
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13
# GNU General Public License for more details.
15
# You should have received a copy of the GNU General Public License
16
# along with this program; if not, write to the Free Software
17
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
19
# Author: Martin Pool <mbp@canonical.com>
22
"""Weave - storage of related text file versions"""
25
# XXX: If we do weaves this way, will a merge still behave the same
26
# way if it's done in a different order? That's a pretty desirable
29
# TODO: Nothing here so far assumes the lines are really \n newlines,
30
# rather than being split up in some other way. We could accommodate
31
# binaries, perhaps by naively splitting on \n or perhaps using
32
# something like a rolling checksum.
34
# TODO: End marker for each version so we can stop reading?
36
# TODO: Check that no insertion occurs inside a deletion that was
37
# active in the version of the insertion.
39
# TODO: In addition to the SHA-1 check, perhaps have some code that
40
# checks structural constraints of the weave: ie that insertions are
41
# properly nested, that there is no text outside of an insertion, that
42
# insertions or deletions are not repeated, etc.
44
# TODO: Parallel-extract that passes back each line along with a
45
# description of which revisions include it. Nice for checking all
46
# shas or calculating stats in parallel.
48
# TODO: Using a single _extract routine and then processing the output
49
# is probably inefficient. It's simple enough that we can afford to
50
# have slight specializations for different ways its used: annotate,
51
# basis for add, get, etc.
53
# TODO: Probably the API should work only in names to hide the integer
54
# indexes from the user.
56
# TODO: Is there any potential performance win by having an add()
57
# variant that is passed a pre-cooked version of the single basis
60
# TODO: Reweave can possibly be made faster by remembering diffs
61
# where the basis and destination are unchanged.
63
# FIXME: Sometimes we will be given a parents list for a revision
64
# that includes some redundant parents (i.e. already a parent of
65
# something in the list.) We should eliminate them. This can
66
# be done fairly efficiently because the sequence numbers constrain
67
# the possible relationships.
69
# FIXME: the conflict markers should be *7* characters
72
from cStringIO import StringIO
81
from bzrlib.trace import mutter
82
from bzrlib.errors import (WeaveError, WeaveFormatError, WeaveParentMismatch,
83
RevisionAlreadyPresent,
85
WeaveRevisionAlreadyPresent,
86
WeaveRevisionNotPresent,
88
import bzrlib.errors as errors
89
from bzrlib.osutils import sha_strings
90
import bzrlib.patiencediff
91
from bzrlib.tsort import topo_sort
92
from bzrlib.versionedfile import VersionedFile, InterVersionedFile
93
from bzrlib.weavefile import _read_weave_v5, write_weave_v5
96
class Weave(VersionedFile):
97
"""weave - versioned text file storage.
99
A Weave manages versions of line-based text files, keeping track
100
of the originating version for each line.
102
To clients the "lines" of the file are represented as a list of strings.
103
These strings will typically have terminal newline characters, but
104
this is not required. In particular files commonly do not have a newline
105
at the end of the file.
107
Texts can be identified in either of two ways:
109
* a nonnegative index number.
111
* a version-id string.
113
Typically the index number will be valid only inside this weave and
114
the version-id is used to reference it in the larger world.
116
The weave is represented as a list mixing edit instructions and
117
literal text. Each entry in _weave can be either a string (or
118
unicode), or a tuple. If a string, it means that the given line
119
should be output in the currently active revisions.
121
If a tuple, it gives a processing instruction saying in which
122
revisions the enclosed lines are active. The tuple has the form
123
(instruction, version).
125
The instruction can be '{' or '}' for an insertion block, and '['
126
and ']' for a deletion block respectively. The version is the
127
integer version index. There is no replace operator, only deletes
128
and inserts. For '}', the end of an insertion, there is no
129
version parameter because it always closes the most recently
134
* A later version can delete lines that were introduced by any
135
number of ancestor versions; this implies that deletion
136
instructions can span insertion blocks without regard to the
137
insertion block's nesting.
139
* Similarly, deletions need not be properly nested with regard to
140
each other, because they might have been generated by
141
independent revisions.
143
* Insertions are always made by inserting a new bracketed block
144
into a single point in the previous weave. This implies they
145
can nest but not overlap, and the nesting must always have later
146
insertions on the inside.
148
* It doesn't seem very useful to have an active insertion
149
inside an inactive insertion, but it might happen.
151
* Therefore, all instructions are always"considered"; that
152
is passed onto and off the stack. An outer inactive block
153
doesn't disable an inner block.
155
* Lines are enabled if the most recent enclosing insertion is
156
active and none of the enclosing deletions are active.
158
* There is no point having a deletion directly inside its own
159
insertion; you might as well just not write it. And there
160
should be no way to get an earlier version deleting a later
164
Text of the weave; list of control instruction tuples and strings.
167
List of parents, indexed by version number.
168
It is only necessary to store the minimal set of parents for
169
each version; the parent's parents are implied.
172
List of hex SHA-1 of each version.
175
List of symbolic names for each version. Each should be unique.
178
For each name, the version number.
181
Descriptive name of this weave; typically the filename if known.
185
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
186
'_weave_name', '_matcher']
188
def __init__(self, weave_name=None, access_mode='w', matcher=None, get_scope=None):
191
:param get_scope: A callable that returns an opaque object to be used
192
for detecting when this weave goes out of scope (should stop
193
answering requests or allowing mutation).
195
super(Weave, self).__init__(access_mode)
201
self._weave_name = weave_name
203
self._matcher = bzrlib.patiencediff.PatienceSequenceMatcher
205
self._matcher = matcher
206
if get_scope is None:
207
get_scope = lambda:None
208
self._get_scope = get_scope
209
self._scope = get_scope()
210
self._access_mode = access_mode
213
return "Weave(%r)" % self._weave_name
215
def _check_write_ok(self):
216
"""Is the versioned file marked as 'finished' ? Raise if it is."""
217
if self._get_scope() != self._scope:
218
raise errors.OutSideTransaction()
219
if self._access_mode != 'w':
220
raise errors.ReadOnlyObjectDirtiedError(self)
223
"""Return a deep copy of self.
225
The copy can be modified without affecting the original weave."""
227
other._weave = self._weave[:]
228
other._parents = self._parents[:]
229
other._sha1s = self._sha1s[:]
230
other._names = self._names[:]
231
other._name_map = self._name_map.copy()
232
other._weave_name = self._weave_name
235
def __eq__(self, other):
236
if not isinstance(other, Weave):
238
return self._parents == other._parents \
239
and self._weave == other._weave \
240
and self._sha1s == other._sha1s
242
def __ne__(self, other):
243
return not self.__eq__(other)
245
def _idx_to_name(self, version):
246
return self._names[version]
248
def _lookup(self, name):
249
"""Convert symbolic version name to index."""
250
self.check_not_reserved_id(name)
252
return self._name_map[name]
254
raise RevisionNotPresent(name, self._weave_name)
257
"""See VersionedFile.versions."""
258
return self._names[:]
260
def has_version(self, version_id):
261
"""See VersionedFile.has_version."""
262
return (version_id in self._name_map)
264
__contains__ = has_version
266
def get_parent_map(self, version_ids):
267
"""See VersionedFile.get_parent_map."""
269
for version_id in version_ids:
271
result[version_id] = tuple(
272
map(self._idx_to_name, self._parents[self._lookup(version_id)]))
273
except RevisionNotPresent:
277
def get_parents_with_ghosts(self, version_id):
278
raise NotImplementedError(self.get_parents_with_ghosts)
280
def _check_repeated_add(self, name, parents, text, sha1):
281
"""Check that a duplicated add is OK.
283
If it is, return the (old) index; otherwise raise an exception.
285
idx = self._lookup(name)
286
if sorted(self._parents[idx]) != sorted(parents) \
287
or sha1 != self._sha1s[idx]:
288
raise RevisionAlreadyPresent(name, self._weave_name)
291
def _add_lines(self, version_id, parents, lines, parent_texts,
292
left_matching_blocks, nostore_sha, random_id, check_content):
293
"""See VersionedFile.add_lines."""
294
idx = self._add(version_id, lines, map(self._lookup, parents),
295
nostore_sha=nostore_sha)
296
return sha_strings(lines), sum(map(len, lines)), idx
298
def _add(self, version_id, lines, parents, sha1=None, nostore_sha=None):
299
"""Add a single text on top of the weave.
301
Returns the index number of the newly added version.
304
Symbolic name for this version.
305
(Typically the revision-id of the revision that added it.)
308
List or set of direct parent version numbers.
311
Sequence of lines to be added in the new version.
313
:param nostore_sha: See VersionedFile.add_lines.
315
self._check_lines_not_unicode(lines)
316
self._check_lines_are_lines(lines)
318
sha1 = sha_strings(lines)
319
if sha1 == nostore_sha:
320
raise errors.ExistingContent
321
if version_id in self._name_map:
322
return self._check_repeated_add(version_id, parents, lines, sha1)
324
self._check_versions(parents)
325
## self._check_lines(lines)
326
new_version = len(self._parents)
328
# if we abort after here the (in-memory) weave will be corrupt because only
329
# some fields are updated
330
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
331
# - Robert Collins 20060226
332
self._parents.append(parents[:])
333
self._sha1s.append(sha1)
334
self._names.append(version_id)
335
self._name_map[version_id] = new_version
339
# special case; adding with no parents revision; can do
340
# this more quickly by just appending unconditionally.
341
# even more specially, if we're adding an empty text we
342
# need do nothing at all.
344
self._weave.append(('{', new_version))
345
self._weave.extend(lines)
346
self._weave.append(('}', None))
349
if len(parents) == 1:
350
pv = list(parents)[0]
351
if sha1 == self._sha1s[pv]:
352
# special case: same as the single parent
356
ancestors = self._inclusions(parents)
360
# basis a list of (origin, lineno, line)
363
for origin, lineno, line in self._extract(ancestors):
364
basis_lineno.append(lineno)
365
basis_lines.append(line)
367
# another small special case: a merge, producing the same text
369
if lines == basis_lines:
372
# add a sentinel, because we can also match against the final line
373
basis_lineno.append(len(self._weave))
375
# XXX: which line of the weave should we really consider
376
# matches the end of the file? the current code says it's the
377
# last line of the weave?
379
#print 'basis_lines:', basis_lines
380
#print 'new_lines: ', lines
382
s = self._matcher(None, basis_lines, lines)
384
# offset gives the number of lines that have been inserted
385
# into the weave up to the current point; if the original edit instruction
386
# says to change line A then we actually change (A+offset)
389
for tag, i1, i2, j1, j2 in s.get_opcodes():
390
# i1,i2 are given in offsets within basis_lines; we need to map them
391
# back to offsets within the entire weave
392
#print 'raw match', tag, i1, i2, j1, j2
395
i1 = basis_lineno[i1]
396
i2 = basis_lineno[i2]
397
# the deletion and insertion are handled separately.
398
# first delete the region.
400
self._weave.insert(i1+offset, ('[', new_version))
401
self._weave.insert(i2+offset+1, (']', new_version))
405
# there may have been a deletion spanning up to
406
# i2; we want to insert after this region to make sure
407
# we don't destroy ourselves
409
self._weave[i:i] = ([('{', new_version)]
412
offset += 2 + (j2 - j1)
415
def _inclusions(self, versions):
416
"""Return set of all ancestors of given version(s)."""
417
if not len(versions):
420
for v in xrange(max(versions), 0, -1):
422
# include all its parents
423
i.update(self._parents[v])
425
## except IndexError:
426
## raise ValueError("version %d not present in weave" % v)
428
def get_ancestry(self, version_ids, topo_sorted=True):
429
"""See VersionedFile.get_ancestry."""
430
if isinstance(version_ids, basestring):
431
version_ids = [version_ids]
432
i = self._inclusions([self._lookup(v) for v in version_ids])
433
return [self._idx_to_name(v) for v in i]
435
def _check_lines(self, text):
436
if not isinstance(text, list):
437
raise ValueError("text should be a list, not %s" % type(text))
440
if not isinstance(l, basestring):
441
raise ValueError("text line should be a string or unicode, not %s"
446
def _check_versions(self, indexes):
447
"""Check everything in the sequence of indexes is valid"""
452
raise IndexError("invalid version number %r" % i)
454
def _compatible_parents(self, my_parents, other_parents):
455
"""During join check that other_parents are joinable with my_parents.
457
Joinable is defined as 'is a subset of' - supersets may require
458
regeneration of diffs, but subsets do not.
460
return len(other_parents.difference(my_parents)) == 0
462
def annotate(self, version_id):
463
"""Return a list of (version-id, line) tuples for version_id.
465
The index indicates when the line originated in the weave."""
466
incls = [self._lookup(version_id)]
467
return [(self._idx_to_name(origin), text) for origin, lineno, text in
468
self._extract(incls)]
470
def iter_lines_added_or_present_in_versions(self, version_ids=None,
472
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
473
if version_ids is None:
474
version_ids = self.versions()
475
version_ids = set(version_ids)
476
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
477
# if inserted not in version_ids then it was inserted before the
478
# versions we care about, but because weaves cannot represent ghosts
479
# properly, we do not filter down to that
480
# if inserted not in version_ids: continue
482
yield line + '\n', inserted
486
def _walk_internal(self, version_ids=None):
487
"""Helper method for weave actions."""
492
lineno = 0 # line of weave, 0-based
494
for l in self._weave:
495
if l.__class__ == tuple:
499
istack.append(self._names[v])
503
dset.add(self._names[v])
505
dset.remove(self._names[v])
507
raise WeaveFormatError('unexpected instruction %r' % v)
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:
627
iset.remove(istack.pop())
635
raise AssertionError()
638
isactive = (not dset) and istack and (istack[-1] in included)
640
result.append((istack[-1], lineno, l))
643
raise WeaveFormatError("unclosed insertion blocks "
644
"at end of weave: %s" % istack)
646
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
650
def _maybe_lookup(self, name_or_index):
651
"""Convert possible symbolic name to index, or pass through indexes.
655
if isinstance(name_or_index, (int, long)):
658
return self._lookup(name_or_index)
660
def get_lines(self, version_id):
661
"""See VersionedFile.get_lines()."""
662
int_index = self._maybe_lookup(version_id)
663
result = [line for (origin, lineno, line) in self._extract([int_index])]
664
expected_sha1 = self._sha1s[int_index]
665
measured_sha1 = sha_strings(result)
666
if measured_sha1 != expected_sha1:
667
raise errors.WeaveInvalidChecksum(
668
'file %s, revision %s, expected: %s, measured %s'
669
% (self._weave_name, version_id,
670
expected_sha1, measured_sha1))
673
def get_sha1s(self, version_ids):
674
"""See VersionedFile.get_sha1s()."""
675
return [self._sha1s[self._lookup(v)] for v in version_ids]
677
def num_versions(self):
678
"""How many versions are in this weave?"""
679
l = len(self._parents)
682
__len__ = num_versions
684
def check(self, progress_bar=None):
685
# TODO evaluate performance hit of using string sets in this routine.
686
# TODO: check no circular inclusions
687
# TODO: create a nested progress bar
688
for version in range(self.num_versions()):
689
inclusions = list(self._parents[version])
692
if inclusions[-1] >= version:
693
raise WeaveFormatError("invalid included version %d for index %d"
694
% (inclusions[-1], version))
696
# try extracting all versions; parallel extraction is used
697
nv = self.num_versions()
702
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
703
# The problem is that set membership is much more expensive
704
name = self._idx_to_name(i)
705
sha1s[name] = sha.new()
707
new_inc = set([name])
708
for p in self._parents[i]:
709
new_inc.update(inclusions[self._idx_to_name(p)])
711
if set(new_inc) != set(self.get_ancestry(name)):
712
raise AssertionError(
714
% (set(new_inc), set(self.get_ancestry(name))))
715
inclusions[name] = new_inc
717
nlines = len(self._weave)
719
update_text = 'checking weave'
721
short_name = os.path.basename(self._weave_name)
722
update_text = 'checking %s' % (short_name,)
723
update_text = update_text[:25]
725
for lineno, insert, deleteset, line in self._walk_internal():
727
progress_bar.update(update_text, lineno, nlines)
729
for name, name_inclusions in inclusions.items():
730
# The active inclusion must be an ancestor,
731
# and no ancestors must have deleted this line,
732
# because we don't support resurrection.
733
if (insert in name_inclusions) and not (deleteset & name_inclusions):
734
sha1s[name].update(line)
737
version = self._idx_to_name(i)
738
hd = sha1s[version].hexdigest()
739
expected = self._sha1s[i]
741
raise errors.WeaveInvalidChecksum(
742
"mismatched sha1 for version %s: "
743
"got %s, expected %s"
744
% (version, hd, expected))
746
# TODO: check insertions are properly nested, that there are
747
# no lines outside of insertion blocks, that deletions are
748
# properly paired, etc.
750
def _join(self, other, pb, msg, version_ids, ignore_missing):
751
"""Worker routine for join()."""
752
if not other.versions():
753
return # nothing to update, easy
756
# versions is never none, InterWeave checks this.
759
# two loops so that we do not change ourselves before verifying it
761
# work through in index order to make sure we get all dependencies
764
# get the selected versions only that are in other.versions.
765
version_ids = set(other.versions()).intersection(set(version_ids))
766
# pull in the referenced graph.
767
version_ids = other.get_ancestry(version_ids)
768
pending_parents = other.get_parent_map(version_ids)
769
pending_graph = pending_parents.items()
770
if len(pending_graph) != len(version_ids):
771
raise RevisionNotPresent(
772
set(version_ids) - set(pending_parents.keys()), self)
773
for name in topo_sort(pending_graph):
774
other_idx = other._name_map[name]
775
# returns True if we have it, False if we need it.
776
if not self._check_version_consistent(other, other_idx, name):
777
names_to_join.append((other_idx, name))
785
for other_idx, name in names_to_join:
786
# TODO: If all the parents of the other version are already
787
# present then we can avoid some work by just taking the delta
788
# and adjusting the offsets.
789
new_parents = self._imported_parents(other, other_idx)
790
sha1 = other._sha1s[other_idx]
795
pb.update(msg, merged, len(names_to_join))
797
lines = other.get_lines(other_idx)
798
self._add(name, lines, new_parents, sha1)
800
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
801
merged, processed, self._weave_name, time.time()-time0))
803
def _imported_parents(self, other, other_idx):
804
"""Return list of parents in self corresponding to indexes in other."""
806
for parent_idx in other._parents[other_idx]:
807
parent_name = other._names[parent_idx]
808
if parent_name not in self._name_map:
809
# should not be possible
810
raise WeaveError("missing parent {%s} of {%s} in %r"
811
% (parent_name, other._name_map[other_idx], self))
812
new_parents.append(self._name_map[parent_name])
815
def _check_version_consistent(self, other, other_idx, name):
816
"""Check if a version in consistent in this and other.
818
To be consistent it must have:
821
* the same direct parents (by name, not index, and disregarding
824
If present & correct return True;
825
if not present in self return False;
826
if inconsistent raise error."""
827
this_idx = self._name_map.get(name, -1)
829
if self._sha1s[this_idx] != other._sha1s[other_idx]:
830
raise errors.WeaveTextDiffers(name, self, other)
831
self_parents = self._parents[this_idx]
832
other_parents = other._parents[other_idx]
833
n1 = set([self._names[i] for i in self_parents])
834
n2 = set([other._names[i] for i in other_parents])
835
if not self._compatible_parents(n1, n2):
836
raise WeaveParentMismatch("inconsistent parents "
837
"for version {%s}: %s vs %s" % (name, n1, n2))
843
def _reweave(self, other, pb, msg):
844
"""Reweave self with other - internal helper for join().
846
:param other: The other weave to merge
847
:param pb: An optional progress bar, indicating how far done we are
848
:param msg: An optional message for the progress
850
new_weave = _reweave(self, other, pb=pb, msg=msg)
851
self._copy_weave_content(new_weave)
853
def _copy_weave_content(self, otherweave):
854
"""adsorb the content from otherweave."""
855
for attr in self.__slots__:
856
if attr != '_weave_name':
857
setattr(self, attr, copy(getattr(otherweave, attr)))
860
class WeaveFile(Weave):
861
"""A WeaveFile represents a Weave on disk and writes on change."""
863
WEAVE_SUFFIX = '.weave'
865
def __init__(self, name, transport, filemode=None, create=False, access_mode='w', get_scope=None):
866
"""Create a WeaveFile.
868
:param create: If not True, only open an existing knit.
870
super(WeaveFile, self).__init__(name, access_mode, get_scope=get_scope)
871
self._transport = transport
872
self._filemode = filemode
874
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
875
except errors.NoSuchFile:
881
def _add_lines(self, version_id, parents, lines, parent_texts,
882
left_matching_blocks, nostore_sha, random_id, check_content):
883
"""Add a version and save the weave."""
884
self.check_not_reserved_id(version_id)
885
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
886
parent_texts, left_matching_blocks, nostore_sha, random_id,
891
def copy_to(self, name, transport):
892
"""See VersionedFile.copy_to()."""
893
# as we are all in memory always, just serialise to the new place.
895
write_weave_v5(self, sio)
897
transport.put_file(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
900
"""Save the weave."""
901
self._check_write_ok()
903
write_weave_v5(self, sio)
905
self._transport.put_file(self._weave_name + WeaveFile.WEAVE_SUFFIX,
911
"""See VersionedFile.get_suffixes()."""
912
return [WeaveFile.WEAVE_SUFFIX]
914
def join(self, other, pb=None, msg=None, version_ids=None,
915
ignore_missing=False):
916
"""Join other into self and save."""
917
super(WeaveFile, self).join(other, pb, msg, version_ids, ignore_missing)
921
def _reweave(wa, wb, pb=None, msg=None):
922
"""Combine two weaves and return the result.
924
This works even if a revision R has different parents in
925
wa and wb. In the resulting weave all the parents are given.
927
This is done by just building up a new weave, maintaining ordering
928
of the versions in the two inputs. More efficient approaches
929
might be possible but it should only be necessary to do
930
this operation rarely, when a new previously ghost version is
933
:param pb: An optional progress bar, indicating how far done we are
934
:param msg: An optional message for the progress
938
queue_a = range(wa.num_versions())
939
queue_b = range(wb.num_versions())
940
# first determine combined parents of all versions
941
# map from version name -> all parent names
942
combined_parents = _reweave_parent_graphs(wa, wb)
943
mutter("combined parents: %r", combined_parents)
944
order = topo_sort(combined_parents.iteritems())
945
mutter("order to reweave: %r", order)
950
for idx, name in enumerate(order):
952
pb.update(msg, idx, len(order))
953
if name in wa._name_map:
954
lines = wa.get_lines(name)
955
if name in wb._name_map:
956
lines_b = wb.get_lines(name)
958
mutter('Weaves differ on content. rev_id {%s}', name)
959
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
961
lines = list(difflib.unified_diff(lines, lines_b,
962
wa._weave_name, wb._weave_name))
963
mutter('lines:\n%s', ''.join(lines))
964
raise errors.WeaveTextDiffers(name, wa, wb)
966
lines = wb.get_lines(name)
967
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
970
def _reweave_parent_graphs(wa, wb):
971
"""Return combined parent ancestry for two weaves.
973
Returned as a list of (version_name, set(parent_names))"""
975
for weave in [wa, wb]:
976
for idx, name in enumerate(weave._names):
977
p = combined.setdefault(name, set())
978
p.update(map(weave._idx_to_name, weave._parents[idx]))
983
"""Show the weave's table-of-contents"""
984
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
985
for i in (6, 50, 10, 10):
988
for i in range(w.num_versions()):
991
parent_str = ' '.join(map(str, w._parents[i]))
992
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
996
def weave_stats(weave_file, pb):
997
from bzrlib.weavefile import read_weave
999
wf = file(weave_file, 'rb')
1001
# FIXME: doesn't work on pipes
1002
weave_size = wf.tell()
1006
for i in range(vers):
1007
pb.update('checking sizes', i, vers)
1008
for origin, lineno, line in w._extract([i]):
1013
print 'versions %9d' % vers
1014
print 'weave file %9d bytes' % weave_size
1015
print 'total contents %9d bytes' % total
1016
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1019
print 'average size %9d bytes' % avg
1020
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1024
print """bzr weave tool
1026
Experimental tool for weave algorithm.
1029
weave init WEAVEFILE
1030
Create an empty weave file
1031
weave get WEAVEFILE VERSION
1032
Write out specified version.
1033
weave check WEAVEFILE
1034
Check consistency of all versions.
1036
Display table of contents.
1037
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1038
Add NEWTEXT, with specified parent versions.
1039
weave annotate WEAVEFILE VERSION
1040
Display origin of each line.
1041
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1042
Auto-merge two versions and display conflicts.
1043
weave diff WEAVEFILE VERSION1 VERSION2
1044
Show differences between two versions.
1048
% weave init foo.weave
1050
% weave add foo.weave ver0 < foo.txt
1053
(create updated version)
1055
% weave get foo.weave 0 | diff -u - foo.txt
1056
% weave add foo.weave ver1 0 < foo.txt
1059
% weave get foo.weave 0 > foo.txt (create forked version)
1061
% weave add foo.weave ver2 0 < foo.txt
1064
% weave merge foo.weave 1 2 > foo.txt (merge them)
1065
% vi foo.txt (resolve conflicts)
1066
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1078
# in case we're run directly from the subdirectory
1079
sys.path.append('..')
1081
from bzrlib.weavefile import write_weave, read_weave
1082
from bzrlib.progress import ProgressBar
1097
return read_weave(file(argv[2], 'rb'))
1103
# at the moment, based on everything in the file
1105
parents = map(int, argv[4:])
1106
lines = sys.stdin.readlines()
1107
ver = w.add(name, parents, lines)
1108
write_weave(w, file(argv[2], 'wb'))
1109
print 'added version %r %d' % (name, ver)
1112
if os.path.exists(fn):
1113
raise IOError("file exists")
1115
write_weave(w, file(fn, 'wb'))
1116
elif cmd == 'get': # get one version
1118
sys.stdout.writelines(w.get_iter(int(argv[3])))
1123
v1, v2 = map(int, argv[3:5])
1126
diff_gen = bzrlib.patiencediff.unified_diff(lines1, lines2,
1127
'%s version %d' % (fn, v1),
1128
'%s version %d' % (fn, v2))
1129
sys.stdout.writelines(diff_gen)
1131
elif cmd == 'annotate':
1133
# newline is added to all lines regardless; too hard to get
1134
# reasonable formatting otherwise
1136
for origin, text in w.annotate(int(argv[3])):
1137
text = text.rstrip('\r\n')
1139
print ' | %s' % (text)
1141
print '%5d | %s' % (origin, text)
1147
elif cmd == 'stats':
1148
weave_stats(argv[2], ProgressBar())
1150
elif cmd == 'check':
1155
print '%d versions ok' % w.num_versions()
1157
elif cmd == 'inclusions':
1159
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1161
elif cmd == 'parents':
1163
print ' '.join(map(str, w._parents[int(argv[3])]))
1165
elif cmd == 'plan-merge':
1166
# replaced by 'bzr weave-plan-merge'
1168
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1170
print '%14s | %s' % (state, line),
1171
elif cmd == 'merge':
1172
# replaced by 'bzr weave-merge-text'
1174
p = w.plan_merge(int(argv[3]), int(argv[4]))
1175
sys.stdout.writelines(w.weave_merge(p))
1177
raise ValueError('unknown command %r' % cmd)
1180
if __name__ == '__main__':
1182
sys.exit(main(sys.argv))
1185
class InterWeave(InterVersionedFile):
1186
"""Optimised code paths for weave to weave operations."""
1188
_matching_file_from_factory = staticmethod(WeaveFile)
1189
_matching_file_to_factory = staticmethod(WeaveFile)
1192
def is_compatible(source, target):
1193
"""Be compatible with weaves."""
1195
return (isinstance(source, Weave) and
1196
isinstance(target, Weave))
1197
except AttributeError:
1200
def join(self, pb=None, msg=None, version_ids=None, ignore_missing=False):
1201
"""See InterVersionedFile.join."""
1202
version_ids = self._get_source_version_ids(version_ids, ignore_missing)
1203
if self.target.versions() == [] and version_ids is None:
1204
self.target._copy_weave_content(self.source)
1206
self.target._join(self.source, pb, msg, version_ids, ignore_missing)
1209
InterVersionedFile.register_optimiser(InterWeave)
added
removed
bzrlib/weave.py
