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
assert isinstance(version_id, basestring)
316
self._check_lines_not_unicode(lines)
317
self._check_lines_are_lines(lines)
319
sha1 = sha_strings(lines)
320
if sha1 == nostore_sha:
321
raise errors.ExistingContent
322
if version_id in self._name_map:
323
return self._check_repeated_add(version_id, parents, lines, sha1)
325
self._check_versions(parents)
326
## self._check_lines(lines)
327
new_version = len(self._parents)
329
# if we abort after here the (in-memory) weave will be corrupt because only
330
# some fields are updated
331
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
332
# - Robert Collins 20060226
333
self._parents.append(parents[:])
334
self._sha1s.append(sha1)
335
self._names.append(version_id)
336
self._name_map[version_id] = new_version
340
# special case; adding with no parents revision; can do
341
# this more quickly by just appending unconditionally.
342
# even more specially, if we're adding an empty text we
343
# need do nothing at all.
345
self._weave.append(('{', new_version))
346
self._weave.extend(lines)
347
self._weave.append(('}', None))
350
if len(parents) == 1:
351
pv = list(parents)[0]
352
if sha1 == self._sha1s[pv]:
353
# special case: same as the single parent
357
ancestors = self._inclusions(parents)
361
# basis a list of (origin, lineno, line)
364
for origin, lineno, line in self._extract(ancestors):
365
basis_lineno.append(lineno)
366
basis_lines.append(line)
368
# another small special case: a merge, producing the same text
370
if lines == basis_lines:
373
# add a sentinel, because we can also match against the final line
374
basis_lineno.append(len(self._weave))
376
# XXX: which line of the weave should we really consider
377
# matches the end of the file? the current code says it's the
378
# last line of the weave?
380
#print 'basis_lines:', basis_lines
381
#print 'new_lines: ', lines
383
s = self._matcher(None, basis_lines, lines)
385
# offset gives the number of lines that have been inserted
386
# into the weave up to the current point; if the original edit instruction
387
# says to change line A then we actually change (A+offset)
390
for tag, i1, i2, j1, j2 in s.get_opcodes():
391
# i1,i2 are given in offsets within basis_lines; we need to map them
392
# back to offsets within the entire weave
393
#print 'raw match', tag, i1, i2, j1, j2
397
i1 = basis_lineno[i1]
398
i2 = basis_lineno[i2]
400
assert 0 <= j1 <= j2 <= len(lines)
402
#print tag, i1, i2, j1, j2
404
# the deletion and insertion are handled separately.
405
# first delete the region.
407
self._weave.insert(i1+offset, ('[', new_version))
408
self._weave.insert(i2+offset+1, (']', new_version))
412
# there may have been a deletion spanning up to
413
# i2; we want to insert after this region to make sure
414
# we don't destroy ourselves
416
self._weave[i:i] = ([('{', new_version)]
419
offset += 2 + (j2 - j1)
422
def _clone_text(self, new_version_id, old_version_id, parents):
423
"""See VersionedFile.clone_text."""
424
old_lines = self.get_text(old_version_id)
425
self.add_lines(new_version_id, parents, old_lines)
427
def _inclusions(self, versions):
428
"""Return set of all ancestors of given version(s)."""
429
if not len(versions):
432
for v in xrange(max(versions), 0, -1):
434
# include all its parents
435
i.update(self._parents[v])
437
## except IndexError:
438
## raise ValueError("version %d not present in weave" % v)
440
def get_ancestry(self, version_ids, topo_sorted=True):
441
"""See VersionedFile.get_ancestry."""
442
if isinstance(version_ids, basestring):
443
version_ids = [version_ids]
444
i = self._inclusions([self._lookup(v) for v in version_ids])
445
return [self._idx_to_name(v) for v in i]
447
def _check_lines(self, text):
448
if not isinstance(text, list):
449
raise ValueError("text should be a list, not %s" % type(text))
452
if not isinstance(l, basestring):
453
raise ValueError("text line should be a string or unicode, not %s"
458
def _check_versions(self, indexes):
459
"""Check everything in the sequence of indexes is valid"""
464
raise IndexError("invalid version number %r" % i)
466
def _compatible_parents(self, my_parents, other_parents):
467
"""During join check that other_parents are joinable with my_parents.
469
Joinable is defined as 'is a subset of' - supersets may require
470
regeneration of diffs, but subsets do not.
472
return len(other_parents.difference(my_parents)) == 0
474
def annotate_iter(self, version_id):
475
"""Yield list of (version-id, line) pairs for the specified version.
477
The index indicates when the line originated in the weave."""
478
incls = [self._lookup(version_id)]
479
for origin, lineno, text in self._extract(incls):
480
yield self._idx_to_name(origin), text
482
def iter_lines_added_or_present_in_versions(self, version_ids=None,
484
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
485
if version_ids is None:
486
version_ids = self.versions()
487
version_ids = set(version_ids)
488
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
489
# if inserted not in version_ids then it was inserted before the
490
# versions we care about, but because weaves cannot represent ghosts
491
# properly, we do not filter down to that
492
# if inserted not in version_ids: continue
494
yield line + '\n', inserted
498
def _walk_internal(self, version_ids=None):
499
"""Helper method for weave actions."""
504
lineno = 0 # line of weave, 0-based
506
for l in self._weave:
507
if l.__class__ == tuple:
511
istack.append(self._names[v])
515
assert self._names[v] not in dset
516
dset.add(self._names[v])
518
dset.remove(self._names[v])
520
raise WeaveFormatError('unexpected instruction %r' % v)
522
assert l.__class__ in (str, unicode)
524
yield lineno, istack[-1], frozenset(dset), l
528
raise WeaveFormatError("unclosed insertion blocks "
529
"at end of weave: %s" % istack)
531
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
534
def plan_merge(self, ver_a, ver_b):
535
"""Return pseudo-annotation indicating how the two versions merge.
537
This is computed between versions a and b and their common
540
Weave lines present in none of them are skipped entirely.
542
inc_a = set(self.get_ancestry([ver_a]))
543
inc_b = set(self.get_ancestry([ver_b]))
544
inc_c = inc_a & inc_b
546
for lineno, insert, deleteset, line in self._walk_internal([ver_a, ver_b]):
547
if deleteset & inc_c:
548
# killed in parent; can't be in either a or b
549
# not relevant to our work
550
yield 'killed-base', line
551
elif insert in inc_c:
552
# was inserted in base
553
killed_a = bool(deleteset & inc_a)
554
killed_b = bool(deleteset & inc_b)
555
if killed_a and killed_b:
556
yield 'killed-both', line
558
yield 'killed-a', line
560
yield 'killed-b', line
562
yield 'unchanged', line
563
elif insert in inc_a:
564
if deleteset & inc_a:
565
yield 'ghost-a', line
569
elif insert in inc_b:
570
if deleteset & inc_b:
571
yield 'ghost-b', line
575
# not in either revision
576
yield 'irrelevant', line
578
yield 'unchanged', '' # terminator
580
def _extract(self, versions):
581
"""Yield annotation of lines in included set.
583
Yields a sequence of tuples (origin, lineno, text), where
584
origin is the origin version, lineno the index in the weave,
585
and text the text of the line.
587
The set typically but not necessarily corresponds to a version.
590
if not isinstance(i, int):
593
included = self._inclusions(versions)
599
lineno = 0 # line of weave, 0-based
605
WFE = WeaveFormatError
608
# 449 0 4474.6820 2356.5590 bzrlib.weave:556(_extract)
609
# +285282 0 1676.8040 1676.8040 +<isinstance>
610
# 1.6 seconds in 'isinstance'.
611
# changing the first isinstance:
612
# 449 0 2814.2660 1577.1760 bzrlib.weave:556(_extract)
613
# +140414 0 762.8050 762.8050 +<isinstance>
614
# note that the inline time actually dropped (less function calls)
615
# and total processing time was halved.
616
# we're still spending ~1/4 of the method in isinstance though.
617
# so lets hard code the acceptable string classes we expect:
618
# 449 0 1202.9420 786.2930 bzrlib.weave:556(_extract)
619
# +71352 0 377.5560 377.5560 +<method 'append' of 'list'
621
# yay, down to ~1/4 the initial extract time, and our inline time
622
# has shrunk again, with isinstance no longer dominating.
623
# tweaking the stack inclusion test to use a set gives:
624
# 449 0 1122.8030 713.0080 bzrlib.weave:556(_extract)
625
# +71352 0 354.9980 354.9980 +<method 'append' of 'list'
627
# - a 5% win, or possibly just noise. However with large istacks that
628
# 'in' test could dominate, so I'm leaving this change in place -
629
# when its fast enough to consider profiling big datasets we can review.
634
for l in self._weave:
635
if l.__class__ == tuple:
643
iset.remove(istack.pop())
654
assert l.__class__ in (str, unicode)
656
isactive = (not dset) and istack and (istack[-1] in included)
658
result.append((istack[-1], lineno, l))
661
raise WeaveFormatError("unclosed insertion blocks "
662
"at end of weave: %s" % istack)
664
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
668
def _maybe_lookup(self, name_or_index):
669
"""Convert possible symbolic name to index, or pass through indexes.
673
if isinstance(name_or_index, (int, long)):
676
return self._lookup(name_or_index)
678
def get_lines(self, version_id):
679
"""See VersionedFile.get_lines()."""
680
int_index = self._maybe_lookup(version_id)
681
result = [line for (origin, lineno, line) in self._extract([int_index])]
682
expected_sha1 = self._sha1s[int_index]
683
measured_sha1 = sha_strings(result)
684
if measured_sha1 != expected_sha1:
685
raise errors.WeaveInvalidChecksum(
686
'file %s, revision %s, expected: %s, measured %s'
687
% (self._weave_name, version_id,
688
expected_sha1, measured_sha1))
691
def get_sha1s(self, version_ids):
692
"""See VersionedFile.get_sha1s()."""
693
return [self._sha1s[self._lookup(v)] for v in version_ids]
695
def num_versions(self):
696
"""How many versions are in this weave?"""
697
l = len(self._parents)
698
assert l == len(self._sha1s)
701
__len__ = num_versions
703
def check(self, progress_bar=None):
704
# TODO evaluate performance hit of using string sets in this routine.
705
# TODO: check no circular inclusions
706
# TODO: create a nested progress bar
707
for version in range(self.num_versions()):
708
inclusions = list(self._parents[version])
711
if inclusions[-1] >= version:
712
raise WeaveFormatError("invalid included version %d for index %d"
713
% (inclusions[-1], version))
715
# try extracting all versions; parallel extraction is used
716
nv = self.num_versions()
721
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
722
# The problem is that set membership is much more expensive
723
name = self._idx_to_name(i)
724
sha1s[name] = sha.new()
726
new_inc = set([name])
727
for p in self._parents[i]:
728
new_inc.update(inclusions[self._idx_to_name(p)])
730
assert set(new_inc) == set(self.get_ancestry(name)), \
731
'failed %s != %s' % (set(new_inc), set(self.get_ancestry(name)))
732
inclusions[name] = new_inc
734
nlines = len(self._weave)
736
update_text = 'checking weave'
738
short_name = os.path.basename(self._weave_name)
739
update_text = 'checking %s' % (short_name,)
740
update_text = update_text[:25]
742
for lineno, insert, deleteset, line in self._walk_internal():
744
progress_bar.update(update_text, lineno, nlines)
746
for name, name_inclusions in inclusions.items():
747
# The active inclusion must be an ancestor,
748
# and no ancestors must have deleted this line,
749
# because we don't support resurrection.
750
if (insert in name_inclusions) and not (deleteset & name_inclusions):
751
sha1s[name].update(line)
754
version = self._idx_to_name(i)
755
hd = sha1s[version].hexdigest()
756
expected = self._sha1s[i]
758
raise errors.WeaveInvalidChecksum(
759
"mismatched sha1 for version %s: "
760
"got %s, expected %s"
761
% (version, hd, expected))
763
# TODO: check insertions are properly nested, that there are
764
# no lines outside of insertion blocks, that deletions are
765
# properly paired, etc.
767
def _join(self, other, pb, msg, version_ids, ignore_missing):
768
"""Worker routine for join()."""
769
if not other.versions():
770
return # nothing to update, easy
773
# versions is never none, InterWeave checks this.
776
# two loops so that we do not change ourselves before verifying it
778
# work through in index order to make sure we get all dependencies
781
# get the selected versions only that are in other.versions.
782
version_ids = set(other.versions()).intersection(set(version_ids))
783
# pull in the referenced graph.
784
version_ids = other.get_ancestry(version_ids)
785
pending_parents = other.get_parent_map(version_ids)
786
pending_graph = pending_parents.items()
787
if len(pending_graph) != len(version_ids):
788
raise RevisionNotPresent(
789
set(version_ids) - set(pending_parents.keys()), self)
790
for name in topo_sort(pending_graph):
791
other_idx = other._name_map[name]
792
# returns True if we have it, False if we need it.
793
if not self._check_version_consistent(other, other_idx, name):
794
names_to_join.append((other_idx, name))
802
for other_idx, name in names_to_join:
803
# TODO: If all the parents of the other version are already
804
# present then we can avoid some work by just taking the delta
805
# and adjusting the offsets.
806
new_parents = self._imported_parents(other, other_idx)
807
sha1 = other._sha1s[other_idx]
812
pb.update(msg, merged, len(names_to_join))
814
lines = other.get_lines(other_idx)
815
self._add(name, lines, new_parents, sha1)
817
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
818
merged, processed, self._weave_name, time.time()-time0))
820
def _imported_parents(self, other, other_idx):
821
"""Return list of parents in self corresponding to indexes in other."""
823
for parent_idx in other._parents[other_idx]:
824
parent_name = other._names[parent_idx]
825
if parent_name not in self._name_map:
826
# should not be possible
827
raise WeaveError("missing parent {%s} of {%s} in %r"
828
% (parent_name, other._name_map[other_idx], self))
829
new_parents.append(self._name_map[parent_name])
832
def _check_version_consistent(self, other, other_idx, name):
833
"""Check if a version in consistent in this and other.
835
To be consistent it must have:
838
* the same direct parents (by name, not index, and disregarding
841
If present & correct return True;
842
if not present in self return False;
843
if inconsistent raise error."""
844
this_idx = self._name_map.get(name, -1)
846
if self._sha1s[this_idx] != other._sha1s[other_idx]:
847
raise errors.WeaveTextDiffers(name, self, other)
848
self_parents = self._parents[this_idx]
849
other_parents = other._parents[other_idx]
850
n1 = set([self._names[i] for i in self_parents])
851
n2 = set([other._names[i] for i in other_parents])
852
if not self._compatible_parents(n1, n2):
853
raise WeaveParentMismatch("inconsistent parents "
854
"for version {%s}: %s vs %s" % (name, n1, n2))
860
def _reweave(self, other, pb, msg):
861
"""Reweave self with other - internal helper for join().
863
:param other: The other weave to merge
864
:param pb: An optional progress bar, indicating how far done we are
865
:param msg: An optional message for the progress
867
new_weave = _reweave(self, other, pb=pb, msg=msg)
868
self._copy_weave_content(new_weave)
870
def _copy_weave_content(self, otherweave):
871
"""adsorb the content from otherweave."""
872
for attr in self.__slots__:
873
if attr != '_weave_name':
874
setattr(self, attr, copy(getattr(otherweave, attr)))
877
class WeaveFile(Weave):
878
"""A WeaveFile represents a Weave on disk and writes on change."""
880
WEAVE_SUFFIX = '.weave'
882
def __init__(self, name, transport, filemode=None, create=False, access_mode='w', get_scope=None):
883
"""Create a WeaveFile.
885
:param create: If not True, only open an existing knit.
887
super(WeaveFile, self).__init__(name, access_mode, get_scope=get_scope)
888
self._transport = transport
889
self._filemode = filemode
891
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
892
except errors.NoSuchFile:
898
def _add_lines(self, version_id, parents, lines, parent_texts,
899
left_matching_blocks, nostore_sha, random_id, check_content):
900
"""Add a version and save the weave."""
901
self.check_not_reserved_id(version_id)
902
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
903
parent_texts, left_matching_blocks, nostore_sha, random_id,
908
def _clone_text(self, new_version_id, old_version_id, parents):
909
"""See VersionedFile.clone_text."""
910
super(WeaveFile, self)._clone_text(new_version_id, old_version_id, parents)
913
def copy_to(self, name, transport):
914
"""See VersionedFile.copy_to()."""
915
# as we are all in memory always, just serialise to the new place.
917
write_weave_v5(self, sio)
919
transport.put_file(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
922
"""Save the weave."""
923
self._check_write_ok()
925
write_weave_v5(self, sio)
927
self._transport.put_file(self._weave_name + WeaveFile.WEAVE_SUFFIX,
933
"""See VersionedFile.get_suffixes()."""
934
return [WeaveFile.WEAVE_SUFFIX]
936
def join(self, other, pb=None, msg=None, version_ids=None,
937
ignore_missing=False):
938
"""Join other into self and save."""
939
super(WeaveFile, self).join(other, pb, msg, version_ids, ignore_missing)
943
def _reweave(wa, wb, pb=None, msg=None):
944
"""Combine two weaves and return the result.
946
This works even if a revision R has different parents in
947
wa and wb. In the resulting weave all the parents are given.
949
This is done by just building up a new weave, maintaining ordering
950
of the versions in the two inputs. More efficient approaches
951
might be possible but it should only be necessary to do
952
this operation rarely, when a new previously ghost version is
955
:param pb: An optional progress bar, indicating how far done we are
956
:param msg: An optional message for the progress
960
queue_a = range(wa.num_versions())
961
queue_b = range(wb.num_versions())
962
# first determine combined parents of all versions
963
# map from version name -> all parent names
964
combined_parents = _reweave_parent_graphs(wa, wb)
965
mutter("combined parents: %r", combined_parents)
966
order = topo_sort(combined_parents.iteritems())
967
mutter("order to reweave: %r", order)
972
for idx, name in enumerate(order):
974
pb.update(msg, idx, len(order))
975
if name in wa._name_map:
976
lines = wa.get_lines(name)
977
if name in wb._name_map:
978
lines_b = wb.get_lines(name)
980
mutter('Weaves differ on content. rev_id {%s}', name)
981
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
983
lines = list(difflib.unified_diff(lines, lines_b,
984
wa._weave_name, wb._weave_name))
985
mutter('lines:\n%s', ''.join(lines))
986
raise errors.WeaveTextDiffers(name, wa, wb)
988
lines = wb.get_lines(name)
989
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
992
def _reweave_parent_graphs(wa, wb):
993
"""Return combined parent ancestry for two weaves.
995
Returned as a list of (version_name, set(parent_names))"""
997
for weave in [wa, wb]:
998
for idx, name in enumerate(weave._names):
999
p = combined.setdefault(name, set())
1000
p.update(map(weave._idx_to_name, weave._parents[idx]))
1005
"""Show the weave's table-of-contents"""
1006
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
1007
for i in (6, 50, 10, 10):
1010
for i in range(w.num_versions()):
1013
parent_str = ' '.join(map(str, w._parents[i]))
1014
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
1018
def weave_stats(weave_file, pb):
1019
from bzrlib.weavefile import read_weave
1021
wf = file(weave_file, 'rb')
1023
# FIXME: doesn't work on pipes
1024
weave_size = wf.tell()
1028
for i in range(vers):
1029
pb.update('checking sizes', i, vers)
1030
for origin, lineno, line in w._extract([i]):
1035
print 'versions %9d' % vers
1036
print 'weave file %9d bytes' % weave_size
1037
print 'total contents %9d bytes' % total
1038
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1041
print 'average size %9d bytes' % avg
1042
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1046
print """bzr weave tool
1048
Experimental tool for weave algorithm.
1051
weave init WEAVEFILE
1052
Create an empty weave file
1053
weave get WEAVEFILE VERSION
1054
Write out specified version.
1055
weave check WEAVEFILE
1056
Check consistency of all versions.
1058
Display table of contents.
1059
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1060
Add NEWTEXT, with specified parent versions.
1061
weave annotate WEAVEFILE VERSION
1062
Display origin of each line.
1063
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1064
Auto-merge two versions and display conflicts.
1065
weave diff WEAVEFILE VERSION1 VERSION2
1066
Show differences between two versions.
1070
% weave init foo.weave
1072
% weave add foo.weave ver0 < foo.txt
1075
(create updated version)
1077
% weave get foo.weave 0 | diff -u - foo.txt
1078
% weave add foo.weave ver1 0 < foo.txt
1081
% weave get foo.weave 0 > foo.txt (create forked version)
1083
% weave add foo.weave ver2 0 < foo.txt
1086
% weave merge foo.weave 1 2 > foo.txt (merge them)
1087
% vi foo.txt (resolve conflicts)
1088
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1100
# in case we're run directly from the subdirectory
1101
sys.path.append('..')
1103
from bzrlib.weavefile import write_weave, read_weave
1104
from bzrlib.progress import ProgressBar
1119
return read_weave(file(argv[2], 'rb'))
1125
# at the moment, based on everything in the file
1127
parents = map(int, argv[4:])
1128
lines = sys.stdin.readlines()
1129
ver = w.add(name, parents, lines)
1130
write_weave(w, file(argv[2], 'wb'))
1131
print 'added version %r %d' % (name, ver)
1134
if os.path.exists(fn):
1135
raise IOError("file exists")
1137
write_weave(w, file(fn, 'wb'))
1138
elif cmd == 'get': # get one version
1140
sys.stdout.writelines(w.get_iter(int(argv[3])))
1145
v1, v2 = map(int, argv[3:5])
1148
diff_gen = bzrlib.patiencediff.unified_diff(lines1, lines2,
1149
'%s version %d' % (fn, v1),
1150
'%s version %d' % (fn, v2))
1151
sys.stdout.writelines(diff_gen)
1153
elif cmd == 'annotate':
1155
# newline is added to all lines regardless; too hard to get
1156
# reasonable formatting otherwise
1158
for origin, text in w.annotate(int(argv[3])):
1159
text = text.rstrip('\r\n')
1161
print ' | %s' % (text)
1163
print '%5d | %s' % (origin, text)
1169
elif cmd == 'stats':
1170
weave_stats(argv[2], ProgressBar())
1172
elif cmd == 'check':
1177
print '%d versions ok' % w.num_versions()
1179
elif cmd == 'inclusions':
1181
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1183
elif cmd == 'parents':
1185
print ' '.join(map(str, w._parents[int(argv[3])]))
1187
elif cmd == 'plan-merge':
1188
# replaced by 'bzr weave-plan-merge'
1190
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1192
print '%14s | %s' % (state, line),
1193
elif cmd == 'merge':
1194
# replaced by 'bzr weave-merge-text'
1196
p = w.plan_merge(int(argv[3]), int(argv[4]))
1197
sys.stdout.writelines(w.weave_merge(p))
1199
raise ValueError('unknown command %r' % cmd)
1202
if __name__ == '__main__':
1204
sys.exit(main(sys.argv))
1207
class InterWeave(InterVersionedFile):
1208
"""Optimised code paths for weave to weave operations."""
1210
_matching_file_from_factory = staticmethod(WeaveFile)
1211
_matching_file_to_factory = staticmethod(WeaveFile)
1214
def is_compatible(source, target):
1215
"""Be compatible with weaves."""
1217
return (isinstance(source, Weave) and
1218
isinstance(target, Weave))
1219
except AttributeError:
1222
def join(self, pb=None, msg=None, version_ids=None, ignore_missing=False):
1223
"""See InterVersionedFile.join."""
1224
version_ids = self._get_source_version_ids(version_ids, ignore_missing)
1225
if self.target.versions() == [] and version_ids is None:
1226
self.target._copy_weave_content(self.source)
1228
self.target._join(self.source, pb, msg, version_ids, ignore_missing)
1231
InterVersionedFile.register_optimiser(InterWeave)
added
removed
bzrlib/weave.py
