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 accomodate
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.
71
from cStringIO import StringIO
72
from difflib import SequenceMatcher
77
from bzrlib.trace import mutter
78
from bzrlib.errors import (WeaveError, WeaveFormatError, WeaveParentMismatch,
79
RevisionAlreadyPresent,
81
WeaveRevisionAlreadyPresent,
82
WeaveRevisionNotPresent,
84
import bzrlib.errors as errors
85
from bzrlib.osutils import sha_strings
86
from bzrlib.symbol_versioning import *
87
from bzrlib.tsort import topo_sort
88
from bzrlib.versionedfile import VersionedFile, InterVersionedFile
89
from bzrlib.weavefile import _read_weave_v5, write_weave_v5
92
class Weave(VersionedFile):
93
"""weave - versioned text file storage.
95
A Weave manages versions of line-based text files, keeping track
96
of the originating version for each line.
98
To clients the "lines" of the file are represented as a list of strings.
99
These strings will typically have terminal newline characters, but
100
this is not required. In particular files commonly do not have a newline
101
at the end of the file.
103
Texts can be identified in either of two ways:
105
* a nonnegative index number.
107
* a version-id string.
109
Typically the index number will be valid only inside this weave and
110
the version-id is used to reference it in the larger world.
112
The weave is represented as a list mixing edit instructions and
113
literal text. Each entry in _weave can be either a string (or
114
unicode), or a tuple. If a string, it means that the given line
115
should be output in the currently active revisions.
117
If a tuple, it gives a processing instruction saying in which
118
revisions the enclosed lines are active. The tuple has the form
119
(instruction, version).
121
The instruction can be '{' or '}' for an insertion block, and '['
122
and ']' for a deletion block respectively. The version is the
123
integer version index. There is no replace operator, only deletes
124
and inserts. For '}', the end of an insertion, there is no
125
version parameter because it always closes the most recently
130
* A later version can delete lines that were introduced by any
131
number of ancestor versions; this implies that deletion
132
instructions can span insertion blocks without regard to the
133
insertion block's nesting.
135
* Similarly, deletions need not be properly nested with regard to
136
each other, because they might have been generated by
137
independent revisions.
139
* Insertions are always made by inserting a new bracketed block
140
into a single point in the previous weave. This implies they
141
can nest but not overlap, and the nesting must always have later
142
insertions on the inside.
144
* It doesn't seem very useful to have an active insertion
145
inside an inactive insertion, but it might happen.
147
* Therefore, all instructions are always"considered"; that
148
is passed onto and off the stack. An outer inactive block
149
doesn't disable an inner block.
151
* Lines are enabled if the most recent enclosing insertion is
152
active and none of the enclosing deletions are active.
154
* There is no point having a deletion directly inside its own
155
insertion; you might as well just not write it. And there
156
should be no way to get an earlier version deleting a later
160
Text of the weave; list of control instruction tuples and strings.
163
List of parents, indexed by version number.
164
It is only necessary to store the minimal set of parents for
165
each version; the parent's parents are implied.
168
List of hex SHA-1 of each version.
171
List of symbolic names for each version. Each should be unique.
174
For each name, the version number.
177
Descriptive name of this weave; typically the filename if known.
181
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
184
def __init__(self, weave_name=None, access_mode='w'):
185
super(Weave, self).__init__(access_mode)
191
self._weave_name = weave_name
194
return "Weave(%r)" % self._weave_name
197
"""Return a deep copy of self.
199
The copy can be modified without affecting the original weave."""
201
other._weave = self._weave[:]
202
other._parents = self._parents[:]
203
other._sha1s = self._sha1s[:]
204
other._names = self._names[:]
205
other._name_map = self._name_map.copy()
206
other._weave_name = self._weave_name
209
def __eq__(self, other):
210
if not isinstance(other, Weave):
212
return self._parents == other._parents \
213
and self._weave == other._weave \
214
and self._sha1s == other._sha1s
216
def __ne__(self, other):
217
return not self.__eq__(other)
219
@deprecated_method(zero_eight)
220
def idx_to_name(self, index):
221
"""Old public interface, the public interface is all names now."""
224
def _idx_to_name(self, version):
225
return self._names[version]
227
@deprecated_method(zero_eight)
228
def lookup(self, name):
229
"""Backwards compatability thunk:
231
Return name, as name is valid in the api now, and spew deprecation
236
def _lookup(self, name):
237
"""Convert symbolic version name to index."""
239
return self._name_map[name]
241
raise RevisionNotPresent(name, self._weave_name)
243
@deprecated_method(zero_eight)
244
def iter_names(self):
245
"""Deprecated convenience function, please see VersionedFile.names()."""
246
return iter(self.names())
248
@deprecated_method(zero_eight)
250
"""See Weave.versions for the current api."""
251
return self.versions()
254
"""See VersionedFile.versions."""
255
return self._names[:]
257
def has_version(self, version_id):
258
"""See VersionedFile.has_version."""
259
return self._name_map.has_key(version_id)
261
__contains__ = has_version
263
def get_parents(self, version_id):
264
"""See VersionedFile.get_parent."""
265
return map(self._idx_to_name, self._parents[self._lookup(version_id)])
267
def _check_repeated_add(self, name, parents, text, sha1):
268
"""Check that a duplicated add is OK.
270
If it is, return the (old) index; otherwise raise an exception.
272
idx = self._lookup(name)
273
if sorted(self._parents[idx]) != sorted(parents) \
274
or sha1 != self._sha1s[idx]:
275
raise RevisionAlreadyPresent(name, self._weave_name)
278
@deprecated_method(zero_eight)
279
def add_identical(self, old_rev_id, new_rev_id, parents):
280
"""Please use Weave.clone_text now."""
281
return self.clone_text(new_rev_id, old_rev_id, parents)
283
def _add_lines(self, version_id, parents, lines, parent_texts):
284
"""See VersionedFile.add_lines."""
285
return self._add(version_id, lines, map(self._lookup, parents))
287
@deprecated_method(zero_eight)
288
def add(self, name, parents, text, sha1=None):
289
"""See VersionedFile.add_lines for the non deprecated api."""
290
return self._add(name, text, map(self._maybe_lookup, parents), sha1)
292
def _add(self, version_id, lines, parents, sha1=None):
293
"""Add a single text on top of the weave.
295
Returns the index number of the newly added version.
298
Symbolic name for this version.
299
(Typically the revision-id of the revision that added it.)
302
List or set of direct parent version numbers.
305
Sequence of lines to be added in the new version.
308
assert isinstance(version_id, basestring)
310
sha1 = sha_strings(lines)
311
if version_id in self._name_map:
312
return self._check_repeated_add(version_id, parents, lines, sha1)
314
self._check_versions(parents)
315
## self._check_lines(lines)
316
new_version = len(self._parents)
318
# if we abort after here the (in-memory) weave will be corrupt because only
319
# some fields are updated
320
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
321
# - Robert Collins 20060226
322
self._parents.append(parents[:])
323
self._sha1s.append(sha1)
324
self._names.append(version_id)
325
self._name_map[version_id] = new_version
329
# special case; adding with no parents revision; can do
330
# this more quickly by just appending unconditionally.
331
# even more specially, if we're adding an empty text we
332
# need do nothing at all.
334
self._weave.append(('{', new_version))
335
self._weave.extend(lines)
336
self._weave.append(('}', None))
339
if len(parents) == 1:
340
pv = list(parents)[0]
341
if sha1 == self._sha1s[pv]:
342
# special case: same as the single parent
346
ancestors = self._inclusions(parents)
350
# basis a list of (origin, lineno, line)
353
for origin, lineno, line in self._extract(ancestors):
354
basis_lineno.append(lineno)
355
basis_lines.append(line)
357
# another small special case: a merge, producing the same text
359
if lines == basis_lines:
362
# add a sentinal, because we can also match against the final line
363
basis_lineno.append(len(self._weave))
365
# XXX: which line of the weave should we really consider
366
# matches the end of the file? the current code says it's the
367
# last line of the weave?
369
#print 'basis_lines:', basis_lines
370
#print 'new_lines: ', lines
372
s = SequenceMatcher(None, basis_lines, lines)
374
# offset gives the number of lines that have been inserted
375
# into the weave up to the current point; if the original edit instruction
376
# says to change line A then we actually change (A+offset)
379
for tag, i1, i2, j1, j2 in s.get_opcodes():
380
# i1,i2 are given in offsets within basis_lines; we need to map them
381
# back to offsets within the entire weave
382
#print 'raw match', tag, i1, i2, j1, j2
386
i1 = basis_lineno[i1]
387
i2 = basis_lineno[i2]
389
assert 0 <= j1 <= j2 <= len(lines)
391
#print tag, i1, i2, j1, j2
393
# the deletion and insertion are handled separately.
394
# first delete the region.
396
self._weave.insert(i1+offset, ('[', new_version))
397
self._weave.insert(i2+offset+1, (']', new_version))
401
# there may have been a deletion spanning up to
402
# i2; we want to insert after this region to make sure
403
# we don't destroy ourselves
405
self._weave[i:i] = ([('{', new_version)]
408
offset += 2 + (j2 - j1)
411
def _clone_text(self, new_version_id, old_version_id, parents):
412
"""See VersionedFile.clone_text."""
413
old_lines = self.get_text(old_version_id)
414
self.add_lines(new_version_id, parents, old_lines)
416
def _inclusions(self, versions):
417
"""Return set of all ancestors of given version(s)."""
418
if not len(versions):
421
for v in xrange(max(versions), 0, -1):
423
# include all its parents
424
i.update(self._parents[v])
426
## except IndexError:
427
## raise ValueError("version %d not present in weave" % v)
429
@deprecated_method(zero_eight)
430
def inclusions(self, version_ids):
431
"""Deprecated - see VersionedFile.get_ancestry for the replacement."""
434
if isinstance(version_ids[0], int):
435
return [self._idx_to_name(v) for v in self._inclusions(version_ids)]
437
return self.get_ancestry(version_ids)
439
def get_ancestry(self, version_ids):
440
"""See VersionedFile.get_ancestry."""
441
if isinstance(version_ids, basestring):
442
version_ids = [version_ids]
443
i = self._inclusions([self._lookup(v) for v in version_ids])
444
return [self._idx_to_name(v) for v in i]
446
def _check_lines(self, text):
447
if not isinstance(text, list):
448
raise ValueError("text should be a list, not %s" % type(text))
451
if not isinstance(l, basestring):
452
raise ValueError("text line should be a string or unicode, not %s"
457
def _check_versions(self, indexes):
458
"""Check everything in the sequence of indexes is valid"""
463
raise IndexError("invalid version number %r" % i)
465
def _compatible_parents(self, my_parents, other_parents):
466
"""During join check that other_parents are joinable with my_parents.
468
Joinable is defined as 'is a subset of' - supersets may require
469
regeneration of diffs, but subsets do not.
471
return len(other_parents.difference(my_parents)) == 0
473
def annotate(self, version_id):
474
if isinstance(version_id, int):
475
warn('Weave.annotate(int) is deprecated. Please use version names'
476
' in all circumstances as of 0.8',
481
for origin, lineno, text in self._extract([version_id]):
482
result.append((origin, text))
485
return super(Weave, self).annotate(version_id)
487
def annotate_iter(self, version_id):
488
"""Yield list of (version-id, line) pairs for the specified version.
490
The index indicates when the line originated in the weave."""
491
incls = [self._lookup(version_id)]
492
for origin, lineno, text in self._extract(incls):
493
yield self._idx_to_name(origin), text
495
@deprecated_method(zero_eight)
497
"""_walk has become visit, a supported api."""
498
return self._walk_internal()
500
def iter_lines_added_or_present_in_versions(self, version_ids=None):
501
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
502
if version_ids is None:
503
version_ids = self.versions()
504
version_ids = set(version_ids)
505
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
506
# if inserted not in version_ids then it was inserted before the
507
# versions we care about, but because weaves cannot represent ghosts
508
# properly, we do not filter down to that
509
# if inserted not in version_ids: continue
515
#@deprecated_method(zero_eight)
516
def walk(self, version_ids=None):
517
"""See VersionedFile.walk."""
518
return self._walk_internal(version_ids)
520
def _walk_internal(self, version_ids=None):
521
"""Helper method for weave actions."""
526
lineno = 0 # line of weave, 0-based
528
for l in self._weave:
529
if isinstance(l, tuple):
533
istack.append(self._idx_to_name(v))
537
assert self._idx_to_name(v) not in dset
538
dset.add(self._idx_to_name(v))
540
dset.remove(self._idx_to_name(v))
542
raise WeaveFormatError('unexpected instruction %r' % v)
544
assert isinstance(l, basestring)
546
yield lineno, istack[-1], dset.copy(), l
550
raise WeaveFormatError("unclosed insertion blocks "
551
"at end of weave: %s" % istack)
553
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
556
def _extract(self, versions):
557
"""Yield annotation of lines in included set.
559
Yields a sequence of tuples (origin, lineno, text), where
560
origin is the origin version, lineno the index in the weave,
561
and text the text of the line.
563
The set typically but not necessarily corresponds to a version.
566
if not isinstance(i, int):
569
included = self._inclusions(versions)
575
lineno = 0 # line of weave, 0-based
581
WFE = WeaveFormatError
584
# 449 0 4474.6820 2356.5590 bzrlib.weave:556(_extract)
585
# +285282 0 1676.8040 1676.8040 +<isinstance>
586
# 1.6 seconds in 'isinstance'.
587
# changing the first isinstance:
588
# 449 0 2814.2660 1577.1760 bzrlib.weave:556(_extract)
589
# +140414 0 762.8050 762.8050 +<isinstance>
590
# note that the inline time actually dropped (less function calls)
591
# and total processing time was halved.
592
# we're still spending ~1/4 of the method in isinstance though.
593
# so lets hard code the acceptable string classes we expect:
594
# 449 0 1202.9420 786.2930 bzrlib.weave:556(_extract)
595
# +71352 0 377.5560 377.5560 +<method 'append' of 'list'
597
# yay, down to ~1/4 the initial extract time, and our inline time
598
# has shrunk again, with isinstance no longer dominating.
599
# tweaking the stack inclusion test to use a set gives:
600
# 449 0 1122.8030 713.0080 bzrlib.weave:556(_extract)
601
# +71352 0 354.9980 354.9980 +<method 'append' of 'list'
603
# - a 5% win, or possibly just noise. However with large istacks that
604
# 'in' test could dominate, so I'm leaving this change in place -
605
# when its fast enough to consider profiling big datasets we can review.
610
for l in self._weave:
611
if l.__class__ == tuple:
619
iset.remove(istack.pop())
630
assert l.__class__ in (str, unicode)
632
isactive = (not dset) and istack and (istack[-1] in included)
634
result.append((istack[-1], lineno, l))
637
raise WeaveFormatError("unclosed insertion blocks "
638
"at end of weave: %s" % istack)
640
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
644
@deprecated_method(zero_eight)
645
def get_iter(self, name_or_index):
646
"""Deprecated, please do not use. Lookups are not not needed.
648
Please use get_lines now.
650
return iter(self.get_lines(self._maybe_lookup(name_or_index)))
652
@deprecated_method(zero_eight)
653
def maybe_lookup(self, name_or_index):
654
"""Deprecated, please do not use. Lookups are not not needed."""
655
return self._maybe_lookup(name_or_index)
657
def _maybe_lookup(self, name_or_index):
658
"""Convert possible symbolic name to index, or pass through indexes.
662
if isinstance(name_or_index, (int, long)):
665
return self._lookup(name_or_index)
667
@deprecated_method(zero_eight)
668
def get(self, version_id):
669
"""Please use either Weave.get_text or Weave.get_lines as desired."""
670
return self.get_lines(version_id)
672
def get_lines(self, version_id):
673
"""See VersionedFile.get_lines()."""
674
int_index = self._maybe_lookup(version_id)
675
result = [line for (origin, lineno, line) in self._extract([int_index])]
676
expected_sha1 = self._sha1s[int_index]
677
measured_sha1 = sha_strings(result)
678
if measured_sha1 != expected_sha1:
679
raise errors.WeaveInvalidChecksum(
680
'file %s, revision %s, expected: %s, measured %s'
681
% (self._weave_name, version_id,
682
expected_sha1, measured_sha1))
685
def get_sha1(self, name):
686
"""Get the stored sha1 sum for the given revision.
688
:param name: The name of the version to lookup
690
return self._sha1s[self._lookup(name)]
692
@deprecated_method(zero_eight)
693
def numversions(self):
694
"""How many versions are in this weave?
696
Deprecated in favour of num_versions.
698
return self.num_versions()
700
def num_versions(self):
701
"""How many versions are in this weave?"""
702
l = len(self._parents)
703
assert l == len(self._sha1s)
706
__len__ = num_versions
708
def check(self, progress_bar=None):
709
# TODO evaluate performance hit of using string sets in this routine.
710
# check no circular inclusions
711
for version in range(self.num_versions()):
712
inclusions = list(self._parents[version])
715
if inclusions[-1] >= version:
716
raise WeaveFormatError("invalid included version %d for index %d"
717
% (inclusions[-1], version))
719
# try extracting all versions; parallel extraction is used
720
nv = self.num_versions()
725
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
726
# The problem is that set membership is much more expensive
727
name = self._idx_to_name(i)
728
sha1s[name] = sha.new()
730
new_inc = set([name])
731
for p in self._parents[i]:
732
new_inc.update(inclusions[self._idx_to_name(p)])
734
assert set(new_inc) == set(self.get_ancestry(name)), \
735
'failed %s != %s' % (set(new_inc), set(self.get_ancestry(name)))
736
inclusions[name] = new_inc
738
nlines = len(self._weave)
740
update_text = 'checking weave'
742
short_name = os.path.basename(self._weave_name)
743
update_text = 'checking %s' % (short_name,)
744
update_text = update_text[:25]
746
for lineno, insert, deleteset, line in self._walk_internal():
748
progress_bar.update(update_text, lineno, nlines)
750
for name, name_inclusions in inclusions.items():
751
# The active inclusion must be an ancestor,
752
# and no ancestors must have deleted this line,
753
# because we don't support resurrection.
754
if (insert in name_inclusions) and not (deleteset & name_inclusions):
755
sha1s[name].update(line)
758
version = self._idx_to_name(i)
759
hd = sha1s[version].hexdigest()
760
expected = self._sha1s[i]
762
raise errors.WeaveInvalidChecksum(
763
"mismatched sha1 for version %s: "
764
"got %s, expected %s"
765
% (version, hd, expected))
767
# TODO: check insertions are properly nested, that there are
768
# no lines outside of insertion blocks, that deletions are
769
# properly paired, etc.
771
def _join(self, other, pb, msg, version_ids, ignore_missing):
772
"""Worker routine for join()."""
773
if not other.versions():
774
return # nothing to update, easy
777
for version_id in version_ids:
778
if not other.has_version(version_id) and not ignore_missing:
779
raise RevisionNotPresent(version_id, self._weave_name)
781
version_ids = other.versions()
783
# two loops so that we do not change ourselves before verifying it
785
# work through in index order to make sure we get all dependencies
788
# get the selected versions only that are in other.versions.
789
version_ids = set(other.versions()).intersection(set(version_ids))
790
# pull in the referenced graph.
791
version_ids = other.get_ancestry(version_ids)
792
pending_graph = [(version, other.get_parents(version)) for
793
version in version_ids]
794
for name in topo_sort(pending_graph):
795
other_idx = other._name_map[name]
796
# returns True if we have it, False if we need it.
797
if not self._check_version_consistent(other, other_idx, name):
798
names_to_join.append((other_idx, name))
807
for other_idx, name in names_to_join:
808
# TODO: If all the parents of the other version are already
809
# present then we can avoid some work by just taking the delta
810
# and adjusting the offsets.
811
new_parents = self._imported_parents(other, other_idx)
812
sha1 = other._sha1s[other_idx]
817
pb.update(msg, merged, len(names_to_join))
819
lines = other.get_lines(other_idx)
820
self._add(name, lines, new_parents, sha1)
822
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
823
merged, processed, self._weave_name, time.time()-time0))
825
def _imported_parents(self, other, other_idx):
826
"""Return list of parents in self corresponding to indexes in other."""
828
for parent_idx in other._parents[other_idx]:
829
parent_name = other._names[parent_idx]
830
if parent_name not in self._name_map:
831
# should not be possible
832
raise WeaveError("missing parent {%s} of {%s} in %r"
833
% (parent_name, other._name_map[other_idx], self))
834
new_parents.append(self._name_map[parent_name])
837
def _check_version_consistent(self, other, other_idx, name):
838
"""Check if a version in consistent in this and other.
840
To be consistent it must have:
843
* the same direct parents (by name, not index, and disregarding
846
If present & correct return True;
847
if not present in self return False;
848
if inconsistent raise error."""
849
this_idx = self._name_map.get(name, -1)
851
if self._sha1s[this_idx] != other._sha1s[other_idx]:
852
raise errors.WeaveTextDiffers(name, self, other)
853
self_parents = self._parents[this_idx]
854
other_parents = other._parents[other_idx]
855
n1 = set([self._names[i] for i in self_parents])
856
n2 = set([other._names[i] for i in other_parents])
857
if not self._compatible_parents(n1, n2):
858
raise WeaveParentMismatch("inconsistent parents "
859
"for version {%s}: %s vs %s" % (name, n1, n2))
865
@deprecated_method(zero_eight)
866
def reweave(self, other, pb=None, msg=None):
867
"""reweave has been superceded by plain use of join."""
868
return self.join(other, pb, msg)
870
def _reweave(self, other, pb, msg):
871
"""Reweave self with other - internal helper for join().
873
:param other: The other weave to merge
874
:param pb: An optional progress bar, indicating how far done we are
875
:param msg: An optional message for the progress
877
new_weave = _reweave(self, other, pb=pb, msg=msg)
878
self._copy_weave_content(new_weave)
880
def _copy_weave_content(self, otherweave):
881
"""adsorb the content from otherweave."""
882
for attr in self.__slots__:
883
if attr != '_weave_name':
884
setattr(self, attr, copy(getattr(otherweave, attr)))
887
class WeaveFile(Weave):
888
"""A WeaveFile represents a Weave on disk and writes on change."""
890
WEAVE_SUFFIX = '.weave'
892
def __init__(self, name, transport, filemode=None, create=False, access_mode='w'):
893
"""Create a WeaveFile.
895
:param create: If not True, only open an existing knit.
897
super(WeaveFile, self).__init__(name, access_mode)
898
self._transport = transport
899
self._filemode = filemode
901
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
902
except errors.NoSuchFile:
908
def _add_lines(self, version_id, parents, lines, parent_texts):
909
"""Add a version and save the weave."""
910
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
915
def _clone_text(self, new_version_id, old_version_id, parents):
916
"""See VersionedFile.clone_text."""
917
super(WeaveFile, self)._clone_text(new_version_id, old_version_id, parents)
920
def copy_to(self, name, transport):
921
"""See VersionedFile.copy_to()."""
922
# as we are all in memory always, just serialise to the new place.
924
write_weave_v5(self, sio)
926
transport.put(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
928
def create_empty(self, name, transport, filemode=None):
929
return WeaveFile(name, transport, filemode, create=True)
932
"""Save the weave."""
933
self._check_write_ok()
935
write_weave_v5(self, sio)
937
self._transport.put(self._weave_name + WeaveFile.WEAVE_SUFFIX,
943
"""See VersionedFile.get_suffixes()."""
944
return [WeaveFile.WEAVE_SUFFIX]
946
def join(self, other, pb=None, msg=None, version_ids=None,
947
ignore_missing=False):
948
"""Join other into self and save."""
949
super(WeaveFile, self).join(other, pb, msg, version_ids, ignore_missing)
953
@deprecated_function(zero_eight)
954
def reweave(wa, wb, pb=None, msg=None):
955
"""reweaving is deprecation, please just use weave.join()."""
956
_reweave(wa, wb, pb, msg)
958
def _reweave(wa, wb, pb=None, msg=None):
959
"""Combine two weaves and return the result.
961
This works even if a revision R has different parents in
962
wa and wb. In the resulting weave all the parents are given.
964
This is done by just building up a new weave, maintaining ordering
965
of the versions in the two inputs. More efficient approaches
966
might be possible but it should only be necessary to do
967
this operation rarely, when a new previously ghost version is
970
:param pb: An optional progress bar, indicating how far done we are
971
:param msg: An optional message for the progress
975
queue_a = range(wa.num_versions())
976
queue_b = range(wb.num_versions())
977
# first determine combined parents of all versions
978
# map from version name -> all parent names
979
combined_parents = _reweave_parent_graphs(wa, wb)
980
mutter("combined parents: %r", combined_parents)
981
order = topo_sort(combined_parents.iteritems())
982
mutter("order to reweave: %r", order)
987
for idx, name in enumerate(order):
989
pb.update(msg, idx, len(order))
990
if name in wa._name_map:
991
lines = wa.get_lines(name)
992
if name in wb._name_map:
993
lines_b = wb.get_lines(name)
995
mutter('Weaves differ on content. rev_id {%s}', name)
996
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
998
lines = list(difflib.unified_diff(lines, lines_b,
999
wa._weave_name, wb._weave_name))
1000
mutter('lines:\n%s', ''.join(lines))
1001
raise errors.WeaveTextDiffers(name, wa, wb)
1003
lines = wb.get_lines(name)
1004
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
1007
def _reweave_parent_graphs(wa, wb):
1008
"""Return combined parent ancestry for two weaves.
1010
Returned as a list of (version_name, set(parent_names))"""
1012
for weave in [wa, wb]:
1013
for idx, name in enumerate(weave._names):
1014
p = combined.setdefault(name, set())
1015
p.update(map(weave._idx_to_name, weave._parents[idx]))
1020
"""Show the weave's table-of-contents"""
1021
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
1022
for i in (6, 50, 10, 10):
1025
for i in range(w.num_versions()):
1028
parent_str = ' '.join(map(str, w._parents[i]))
1029
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
1033
def weave_stats(weave_file, pb):
1034
from bzrlib.weavefile import read_weave
1036
wf = file(weave_file, 'rb')
1037
w = read_weave(wf, WeaveVersionedFile)
1038
# FIXME: doesn't work on pipes
1039
weave_size = wf.tell()
1043
for i in range(vers):
1044
pb.update('checking sizes', i, vers)
1045
for origin, lineno, line in w._extract([i]):
1050
print 'versions %9d' % vers
1051
print 'weave file %9d bytes' % weave_size
1052
print 'total contents %9d bytes' % total
1053
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1056
print 'average size %9d bytes' % avg
1057
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1061
print """bzr weave tool
1063
Experimental tool for weave algorithm.
1066
weave init WEAVEFILE
1067
Create an empty weave file
1068
weave get WEAVEFILE VERSION
1069
Write out specified version.
1070
weave check WEAVEFILE
1071
Check consistency of all versions.
1073
Display table of contents.
1074
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1075
Add NEWTEXT, with specified parent versions.
1076
weave annotate WEAVEFILE VERSION
1077
Display origin of each line.
1078
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1079
Auto-merge two versions and display conflicts.
1080
weave diff WEAVEFILE VERSION1 VERSION2
1081
Show differences between two versions.
1085
% weave init foo.weave
1087
% weave add foo.weave ver0 < foo.txt
1090
(create updated version)
1092
% weave get foo.weave 0 | diff -u - foo.txt
1093
% weave add foo.weave ver1 0 < foo.txt
1096
% weave get foo.weave 0 > foo.txt (create forked version)
1098
% weave add foo.weave ver2 0 < foo.txt
1101
% weave merge foo.weave 1 2 > foo.txt (merge them)
1102
% vi foo.txt (resolve conflicts)
1103
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1115
# in case we're run directly from the subdirectory
1116
sys.path.append('..')
1118
from bzrlib.weavefile import write_weave, read_weave
1119
from bzrlib.progress import ProgressBar
1134
return read_weave(file(argv[2], 'rb'))
1140
# at the moment, based on everything in the file
1142
parents = map(int, argv[4:])
1143
lines = sys.stdin.readlines()
1144
ver = w.add(name, parents, lines)
1145
write_weave(w, file(argv[2], 'wb'))
1146
print 'added version %r %d' % (name, ver)
1149
if os.path.exists(fn):
1150
raise IOError("file exists")
1152
write_weave(w, file(fn, 'wb'))
1153
elif cmd == 'get': # get one version
1155
sys.stdout.writelines(w.get_iter(int(argv[3])))
1158
from difflib import unified_diff
1161
v1, v2 = map(int, argv[3:5])
1164
diff_gen = unified_diff(lines1, lines2,
1165
'%s version %d' % (fn, v1),
1166
'%s version %d' % (fn, v2))
1167
sys.stdout.writelines(diff_gen)
1169
elif cmd == 'annotate':
1171
# newline is added to all lines regardless; too hard to get
1172
# reasonable formatting otherwise
1174
for origin, text in w.annotate(int(argv[3])):
1175
text = text.rstrip('\r\n')
1177
print ' | %s' % (text)
1179
print '%5d | %s' % (origin, text)
1185
elif cmd == 'stats':
1186
weave_stats(argv[2], ProgressBar())
1188
elif cmd == 'check':
1193
print '%d versions ok' % w.num_versions()
1195
elif cmd == 'inclusions':
1197
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1199
elif cmd == 'parents':
1201
print ' '.join(map(str, w._parents[int(argv[3])]))
1203
elif cmd == 'plan-merge':
1205
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1207
print '%14s | %s' % (state, line),
1209
elif cmd == 'merge':
1211
p = w.plan_merge(int(argv[3]), int(argv[4]))
1212
sys.stdout.writelines(w.weave_merge(p))
1215
raise ValueError('unknown command %r' % cmd)
1219
def profile_main(argv):
1220
import tempfile, hotshot, hotshot.stats
1222
prof_f = tempfile.NamedTemporaryFile()
1224
prof = hotshot.Profile(prof_f.name)
1226
ret = prof.runcall(main, argv)
1229
stats = hotshot.stats.load(prof_f.name)
1231
stats.sort_stats('cumulative')
1232
## XXX: Might like to write to stderr or the trace file instead but
1233
## print_stats seems hardcoded to stdout
1234
stats.print_stats(20)
1239
def lsprofile_main(argv):
1240
from bzrlib.lsprof import profile
1241
ret,stats = profile(main, argv)
1247
if __name__ == '__main__':
1249
if '--profile' in sys.argv:
1251
args.remove('--profile')
1252
sys.exit(profile_main(args))
1253
elif '--lsprof' in sys.argv:
1255
args.remove('--lsprof')
1256
sys.exit(lsprofile_main(args))
1258
sys.exit(main(sys.argv))
1261
class InterWeave(InterVersionedFile):
1262
"""Optimised code paths for weave to weave operations."""
1264
_matching_file_factory = staticmethod(WeaveFile)
1267
def is_compatible(source, target):
1268
"""Be compatible with weaves."""
1270
return (isinstance(source, Weave) and
1271
isinstance(target, Weave))
1272
except AttributeError:
1275
def join(self, pb=None, msg=None, version_ids=None, ignore_missing=False):
1276
"""See InterVersionedFile.join."""
1277
if self.target.versions() == []:
1279
self.target._copy_weave_content(self.source)
1282
self.target._join(self.source, pb, msg, version_ids, ignore_missing)
1283
except errors.WeaveParentMismatch:
1284
self.target._reweave(self.source, pb, msg)
1287
InterVersionedFile.register_optimiser(InterWeave)
added
removed
bzrlib/weave.py
