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):
189
super(Weave, self).__init__(access_mode)
195
self._weave_name = weave_name
197
self._matcher = bzrlib.patiencediff.PatienceSequenceMatcher
199
self._matcher = matcher
202
return "Weave(%r)" % self._weave_name
205
"""Return a deep copy of self.
207
The copy can be modified without affecting the original weave."""
209
other._weave = self._weave[:]
210
other._parents = self._parents[:]
211
other._sha1s = self._sha1s[:]
212
other._names = self._names[:]
213
other._name_map = self._name_map.copy()
214
other._weave_name = self._weave_name
217
def __eq__(self, other):
218
if not isinstance(other, Weave):
220
return self._parents == other._parents \
221
and self._weave == other._weave \
222
and self._sha1s == other._sha1s
224
def __ne__(self, other):
225
return not self.__eq__(other)
227
def _idx_to_name(self, version):
228
return self._names[version]
230
def _lookup(self, name):
231
"""Convert symbolic version name to index."""
232
self.check_not_reserved_id(name)
234
return self._name_map[name]
236
raise RevisionNotPresent(name, self._weave_name)
239
"""See VersionedFile.versions."""
240
return self._names[:]
242
def has_version(self, version_id):
243
"""See VersionedFile.has_version."""
244
return (version_id in self._name_map)
246
__contains__ = has_version
248
def get_parent_map(self, version_ids):
249
"""See VersionedFile.get_parent_map."""
251
for version_id in version_ids:
253
result[version_id] = tuple(
254
map(self._idx_to_name, self._parents[self._lookup(version_id)]))
255
except RevisionNotPresent:
259
def get_parents_with_ghosts(self, version_id):
260
raise NotImplementedError(self.get_parents_with_ghosts)
262
def _check_repeated_add(self, name, parents, text, sha1):
263
"""Check that a duplicated add is OK.
265
If it is, return the (old) index; otherwise raise an exception.
267
idx = self._lookup(name)
268
if sorted(self._parents[idx]) != sorted(parents) \
269
or sha1 != self._sha1s[idx]:
270
raise RevisionAlreadyPresent(name, self._weave_name)
273
def _add_lines(self, version_id, parents, lines, parent_texts,
274
left_matching_blocks, nostore_sha, random_id, check_content):
275
"""See VersionedFile.add_lines."""
276
idx = self._add(version_id, lines, map(self._lookup, parents),
277
nostore_sha=nostore_sha)
278
return sha_strings(lines), sum(map(len, lines)), idx
280
def _add(self, version_id, lines, parents, sha1=None, nostore_sha=None):
281
"""Add a single text on top of the weave.
283
Returns the index number of the newly added version.
286
Symbolic name for this version.
287
(Typically the revision-id of the revision that added it.)
290
List or set of direct parent version numbers.
293
Sequence of lines to be added in the new version.
295
:param nostore_sha: See VersionedFile.add_lines.
297
assert isinstance(version_id, basestring)
298
self._check_lines_not_unicode(lines)
299
self._check_lines_are_lines(lines)
301
sha1 = sha_strings(lines)
302
if sha1 == nostore_sha:
303
raise errors.ExistingContent
304
if version_id in self._name_map:
305
return self._check_repeated_add(version_id, parents, lines, sha1)
307
self._check_versions(parents)
308
## self._check_lines(lines)
309
new_version = len(self._parents)
311
# if we abort after here the (in-memory) weave will be corrupt because only
312
# some fields are updated
313
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
314
# - Robert Collins 20060226
315
self._parents.append(parents[:])
316
self._sha1s.append(sha1)
317
self._names.append(version_id)
318
self._name_map[version_id] = new_version
322
# special case; adding with no parents revision; can do
323
# this more quickly by just appending unconditionally.
324
# even more specially, if we're adding an empty text we
325
# need do nothing at all.
327
self._weave.append(('{', new_version))
328
self._weave.extend(lines)
329
self._weave.append(('}', None))
332
if len(parents) == 1:
333
pv = list(parents)[0]
334
if sha1 == self._sha1s[pv]:
335
# special case: same as the single parent
339
ancestors = self._inclusions(parents)
343
# basis a list of (origin, lineno, line)
346
for origin, lineno, line in self._extract(ancestors):
347
basis_lineno.append(lineno)
348
basis_lines.append(line)
350
# another small special case: a merge, producing the same text
352
if lines == basis_lines:
355
# add a sentinel, because we can also match against the final line
356
basis_lineno.append(len(self._weave))
358
# XXX: which line of the weave should we really consider
359
# matches the end of the file? the current code says it's the
360
# last line of the weave?
362
#print 'basis_lines:', basis_lines
363
#print 'new_lines: ', lines
365
s = self._matcher(None, basis_lines, lines)
367
# offset gives the number of lines that have been inserted
368
# into the weave up to the current point; if the original edit instruction
369
# says to change line A then we actually change (A+offset)
372
for tag, i1, i2, j1, j2 in s.get_opcodes():
373
# i1,i2 are given in offsets within basis_lines; we need to map them
374
# back to offsets within the entire weave
375
#print 'raw match', tag, i1, i2, j1, j2
379
i1 = basis_lineno[i1]
380
i2 = basis_lineno[i2]
382
assert 0 <= j1 <= j2 <= len(lines)
384
#print tag, i1, i2, j1, j2
386
# the deletion and insertion are handled separately.
387
# first delete the region.
389
self._weave.insert(i1+offset, ('[', new_version))
390
self._weave.insert(i2+offset+1, (']', new_version))
394
# there may have been a deletion spanning up to
395
# i2; we want to insert after this region to make sure
396
# we don't destroy ourselves
398
self._weave[i:i] = ([('{', new_version)]
401
offset += 2 + (j2 - j1)
404
def _clone_text(self, new_version_id, old_version_id, parents):
405
"""See VersionedFile.clone_text."""
406
old_lines = self.get_text(old_version_id)
407
self.add_lines(new_version_id, parents, old_lines)
409
def _inclusions(self, versions):
410
"""Return set of all ancestors of given version(s)."""
411
if not len(versions):
414
for v in xrange(max(versions), 0, -1):
416
# include all its parents
417
i.update(self._parents[v])
419
## except IndexError:
420
## raise ValueError("version %d not present in weave" % v)
422
def get_ancestry(self, version_ids, topo_sorted=True):
423
"""See VersionedFile.get_ancestry."""
424
if isinstance(version_ids, basestring):
425
version_ids = [version_ids]
426
i = self._inclusions([self._lookup(v) for v in version_ids])
427
return [self._idx_to_name(v) for v in i]
429
def _check_lines(self, text):
430
if not isinstance(text, list):
431
raise ValueError("text should be a list, not %s" % type(text))
434
if not isinstance(l, basestring):
435
raise ValueError("text line should be a string or unicode, not %s"
440
def _check_versions(self, indexes):
441
"""Check everything in the sequence of indexes is valid"""
446
raise IndexError("invalid version number %r" % i)
448
def _compatible_parents(self, my_parents, other_parents):
449
"""During join check that other_parents are joinable with my_parents.
451
Joinable is defined as 'is a subset of' - supersets may require
452
regeneration of diffs, but subsets do not.
454
return len(other_parents.difference(my_parents)) == 0
456
def annotate_iter(self, version_id):
457
"""Yield list of (version-id, line) pairs for the specified version.
459
The index indicates when the line originated in the weave."""
460
incls = [self._lookup(version_id)]
461
for origin, lineno, text in self._extract(incls):
462
yield self._idx_to_name(origin), text
464
def iter_lines_added_or_present_in_versions(self, version_ids=None,
466
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
467
if version_ids is None:
468
version_ids = self.versions()
469
version_ids = set(version_ids)
470
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
471
# if inserted not in version_ids then it was inserted before the
472
# versions we care about, but because weaves cannot represent ghosts
473
# properly, we do not filter down to that
474
# if inserted not in version_ids: continue
476
yield line + '\n', inserted
480
def _walk_internal(self, version_ids=None):
481
"""Helper method for weave actions."""
486
lineno = 0 # line of weave, 0-based
488
for l in self._weave:
489
if l.__class__ == tuple:
493
istack.append(self._names[v])
497
assert self._names[v] not in dset
498
dset.add(self._names[v])
500
dset.remove(self._names[v])
502
raise WeaveFormatError('unexpected instruction %r' % v)
504
assert l.__class__ in (str, unicode)
506
yield lineno, istack[-1], frozenset(dset), l
510
raise WeaveFormatError("unclosed insertion blocks "
511
"at end of weave: %s" % istack)
513
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
516
def plan_merge(self, ver_a, ver_b):
517
"""Return pseudo-annotation indicating how the two versions merge.
519
This is computed between versions a and b and their common
522
Weave lines present in none of them are skipped entirely.
524
inc_a = set(self.get_ancestry([ver_a]))
525
inc_b = set(self.get_ancestry([ver_b]))
526
inc_c = inc_a & inc_b
528
for lineno, insert, deleteset, line in self._walk_internal([ver_a, ver_b]):
529
if deleteset & inc_c:
530
# killed in parent; can't be in either a or b
531
# not relevant to our work
532
yield 'killed-base', line
533
elif insert in inc_c:
534
# was inserted in base
535
killed_a = bool(deleteset & inc_a)
536
killed_b = bool(deleteset & inc_b)
537
if killed_a and killed_b:
538
yield 'killed-both', line
540
yield 'killed-a', line
542
yield 'killed-b', line
544
yield 'unchanged', line
545
elif insert in inc_a:
546
if deleteset & inc_a:
547
yield 'ghost-a', line
551
elif insert in inc_b:
552
if deleteset & inc_b:
553
yield 'ghost-b', line
557
# not in either revision
558
yield 'irrelevant', line
560
yield 'unchanged', '' # terminator
562
def _extract(self, versions):
563
"""Yield annotation of lines in included set.
565
Yields a sequence of tuples (origin, lineno, text), where
566
origin is the origin version, lineno the index in the weave,
567
and text the text of the line.
569
The set typically but not necessarily corresponds to a version.
572
if not isinstance(i, int):
575
included = self._inclusions(versions)
581
lineno = 0 # line of weave, 0-based
587
WFE = WeaveFormatError
590
# 449 0 4474.6820 2356.5590 bzrlib.weave:556(_extract)
591
# +285282 0 1676.8040 1676.8040 +<isinstance>
592
# 1.6 seconds in 'isinstance'.
593
# changing the first isinstance:
594
# 449 0 2814.2660 1577.1760 bzrlib.weave:556(_extract)
595
# +140414 0 762.8050 762.8050 +<isinstance>
596
# note that the inline time actually dropped (less function calls)
597
# and total processing time was halved.
598
# we're still spending ~1/4 of the method in isinstance though.
599
# so lets hard code the acceptable string classes we expect:
600
# 449 0 1202.9420 786.2930 bzrlib.weave:556(_extract)
601
# +71352 0 377.5560 377.5560 +<method 'append' of 'list'
603
# yay, down to ~1/4 the initial extract time, and our inline time
604
# has shrunk again, with isinstance no longer dominating.
605
# tweaking the stack inclusion test to use a set gives:
606
# 449 0 1122.8030 713.0080 bzrlib.weave:556(_extract)
607
# +71352 0 354.9980 354.9980 +<method 'append' of 'list'
609
# - a 5% win, or possibly just noise. However with large istacks that
610
# 'in' test could dominate, so I'm leaving this change in place -
611
# when its fast enough to consider profiling big datasets we can review.
616
for l in self._weave:
617
if l.__class__ == tuple:
625
iset.remove(istack.pop())
636
assert l.__class__ in (str, unicode)
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_sha1(self, version_id):
674
"""See VersionedFile.get_sha1()."""
675
return self._sha1s[self._lookup(version_id)]
677
def get_sha1s(self, version_ids):
678
"""See VersionedFile.get_sha1s()."""
679
return [self._sha1s[self._lookup(v)] for v in version_ids]
681
def num_versions(self):
682
"""How many versions are in this weave?"""
683
l = len(self._parents)
684
assert l == len(self._sha1s)
687
__len__ = num_versions
689
def check(self, progress_bar=None):
690
# TODO evaluate performance hit of using string sets in this routine.
691
# TODO: check no circular inclusions
692
# TODO: create a nested progress bar
693
for version in range(self.num_versions()):
694
inclusions = list(self._parents[version])
697
if inclusions[-1] >= version:
698
raise WeaveFormatError("invalid included version %d for index %d"
699
% (inclusions[-1], version))
701
# try extracting all versions; parallel extraction is used
702
nv = self.num_versions()
707
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
708
# The problem is that set membership is much more expensive
709
name = self._idx_to_name(i)
710
sha1s[name] = sha.new()
712
new_inc = set([name])
713
for p in self._parents[i]:
714
new_inc.update(inclusions[self._idx_to_name(p)])
716
assert set(new_inc) == set(self.get_ancestry(name)), \
717
'failed %s != %s' % (set(new_inc), set(self.get_ancestry(name)))
718
inclusions[name] = new_inc
720
nlines = len(self._weave)
722
update_text = 'checking weave'
724
short_name = os.path.basename(self._weave_name)
725
update_text = 'checking %s' % (short_name,)
726
update_text = update_text[:25]
728
for lineno, insert, deleteset, line in self._walk_internal():
730
progress_bar.update(update_text, lineno, nlines)
732
for name, name_inclusions in inclusions.items():
733
# The active inclusion must be an ancestor,
734
# and no ancestors must have deleted this line,
735
# because we don't support resurrection.
736
if (insert in name_inclusions) and not (deleteset & name_inclusions):
737
sha1s[name].update(line)
740
version = self._idx_to_name(i)
741
hd = sha1s[version].hexdigest()
742
expected = self._sha1s[i]
744
raise errors.WeaveInvalidChecksum(
745
"mismatched sha1 for version %s: "
746
"got %s, expected %s"
747
% (version, hd, expected))
749
# TODO: check insertions are properly nested, that there are
750
# no lines outside of insertion blocks, that deletions are
751
# properly paired, etc.
753
def _join(self, other, pb, msg, version_ids, ignore_missing):
754
"""Worker routine for join()."""
755
if not other.versions():
756
return # nothing to update, easy
759
# versions is never none, InterWeave checks this.
762
# two loops so that we do not change ourselves before verifying it
764
# work through in index order to make sure we get all dependencies
767
# get the selected versions only that are in other.versions.
768
version_ids = set(other.versions()).intersection(set(version_ids))
769
# pull in the referenced graph.
770
version_ids = other.get_ancestry(version_ids)
771
pending_parents = other.get_parent_map(version_ids)
772
pending_graph = pending_parents.items()
773
if len(pending_graph) != len(version_ids):
774
raise RevisionNotPresent(
775
set(version_ids) - set(pending_parents.keys()), self)
776
for name in topo_sort(pending_graph):
777
other_idx = other._name_map[name]
778
# returns True if we have it, False if we need it.
779
if not self._check_version_consistent(other, other_idx, name):
780
names_to_join.append((other_idx, name))
788
for other_idx, name in names_to_join:
789
# TODO: If all the parents of the other version are already
790
# present then we can avoid some work by just taking the delta
791
# and adjusting the offsets.
792
new_parents = self._imported_parents(other, other_idx)
793
sha1 = other._sha1s[other_idx]
798
pb.update(msg, merged, len(names_to_join))
800
lines = other.get_lines(other_idx)
801
self._add(name, lines, new_parents, sha1)
803
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
804
merged, processed, self._weave_name, time.time()-time0))
806
def _imported_parents(self, other, other_idx):
807
"""Return list of parents in self corresponding to indexes in other."""
809
for parent_idx in other._parents[other_idx]:
810
parent_name = other._names[parent_idx]
811
if parent_name not in self._name_map:
812
# should not be possible
813
raise WeaveError("missing parent {%s} of {%s} in %r"
814
% (parent_name, other._name_map[other_idx], self))
815
new_parents.append(self._name_map[parent_name])
818
def _check_version_consistent(self, other, other_idx, name):
819
"""Check if a version in consistent in this and other.
821
To be consistent it must have:
824
* the same direct parents (by name, not index, and disregarding
827
If present & correct return True;
828
if not present in self return False;
829
if inconsistent raise error."""
830
this_idx = self._name_map.get(name, -1)
832
if self._sha1s[this_idx] != other._sha1s[other_idx]:
833
raise errors.WeaveTextDiffers(name, self, other)
834
self_parents = self._parents[this_idx]
835
other_parents = other._parents[other_idx]
836
n1 = set([self._names[i] for i in self_parents])
837
n2 = set([other._names[i] for i in other_parents])
838
if not self._compatible_parents(n1, n2):
839
raise WeaveParentMismatch("inconsistent parents "
840
"for version {%s}: %s vs %s" % (name, n1, n2))
846
def _reweave(self, other, pb, msg):
847
"""Reweave self with other - internal helper for join().
849
:param other: The other weave to merge
850
:param pb: An optional progress bar, indicating how far done we are
851
:param msg: An optional message for the progress
853
new_weave = _reweave(self, other, pb=pb, msg=msg)
854
self._copy_weave_content(new_weave)
856
def _copy_weave_content(self, otherweave):
857
"""adsorb the content from otherweave."""
858
for attr in self.__slots__:
859
if attr != '_weave_name':
860
setattr(self, attr, copy(getattr(otherweave, attr)))
863
class WeaveFile(Weave):
864
"""A WeaveFile represents a Weave on disk and writes on change."""
866
WEAVE_SUFFIX = '.weave'
868
def __init__(self, name, transport, filemode=None, create=False, access_mode='w'):
869
"""Create a WeaveFile.
871
:param create: If not True, only open an existing knit.
873
super(WeaveFile, self).__init__(name, access_mode)
874
self._transport = transport
875
self._filemode = filemode
877
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
878
except errors.NoSuchFile:
884
def _add_lines(self, version_id, parents, lines, parent_texts,
885
left_matching_blocks, nostore_sha, random_id, check_content):
886
"""Add a version and save the weave."""
887
self.check_not_reserved_id(version_id)
888
result = super(WeaveFile, self)._add_lines(version_id, parents, lines,
889
parent_texts, left_matching_blocks, nostore_sha, random_id,
894
def _clone_text(self, new_version_id, old_version_id, parents):
895
"""See VersionedFile.clone_text."""
896
super(WeaveFile, self)._clone_text(new_version_id, old_version_id, parents)
899
def copy_to(self, name, transport):
900
"""See VersionedFile.copy_to()."""
901
# as we are all in memory always, just serialise to the new place.
903
write_weave_v5(self, sio)
905
transport.put_file(name + WeaveFile.WEAVE_SUFFIX, sio, self._filemode)
907
def create_empty(self, name, transport, filemode=None):
908
return WeaveFile(name, transport, filemode, create=True)
911
"""Save the weave."""
912
self._check_write_ok()
914
write_weave_v5(self, sio)
916
self._transport.put_file(self._weave_name + WeaveFile.WEAVE_SUFFIX,
922
"""See VersionedFile.get_suffixes()."""
923
return [WeaveFile.WEAVE_SUFFIX]
925
def join(self, other, pb=None, msg=None, version_ids=None,
926
ignore_missing=False):
927
"""Join other into self and save."""
928
super(WeaveFile, self).join(other, pb, msg, version_ids, ignore_missing)
932
def _reweave(wa, wb, pb=None, msg=None):
933
"""Combine two weaves and return the result.
935
This works even if a revision R has different parents in
936
wa and wb. In the resulting weave all the parents are given.
938
This is done by just building up a new weave, maintaining ordering
939
of the versions in the two inputs. More efficient approaches
940
might be possible but it should only be necessary to do
941
this operation rarely, when a new previously ghost version is
944
:param pb: An optional progress bar, indicating how far done we are
945
:param msg: An optional message for the progress
949
queue_a = range(wa.num_versions())
950
queue_b = range(wb.num_versions())
951
# first determine combined parents of all versions
952
# map from version name -> all parent names
953
combined_parents = _reweave_parent_graphs(wa, wb)
954
mutter("combined parents: %r", combined_parents)
955
order = topo_sort(combined_parents.iteritems())
956
mutter("order to reweave: %r", order)
961
for idx, name in enumerate(order):
963
pb.update(msg, idx, len(order))
964
if name in wa._name_map:
965
lines = wa.get_lines(name)
966
if name in wb._name_map:
967
lines_b = wb.get_lines(name)
969
mutter('Weaves differ on content. rev_id {%s}', name)
970
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
972
lines = list(difflib.unified_diff(lines, lines_b,
973
wa._weave_name, wb._weave_name))
974
mutter('lines:\n%s', ''.join(lines))
975
raise errors.WeaveTextDiffers(name, wa, wb)
977
lines = wb.get_lines(name)
978
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
981
def _reweave_parent_graphs(wa, wb):
982
"""Return combined parent ancestry for two weaves.
984
Returned as a list of (version_name, set(parent_names))"""
986
for weave in [wa, wb]:
987
for idx, name in enumerate(weave._names):
988
p = combined.setdefault(name, set())
989
p.update(map(weave._idx_to_name, weave._parents[idx]))
994
"""Show the weave's table-of-contents"""
995
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
996
for i in (6, 50, 10, 10):
999
for i in range(w.num_versions()):
1002
parent_str = ' '.join(map(str, w._parents[i]))
1003
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
1007
def weave_stats(weave_file, pb):
1008
from bzrlib.weavefile import read_weave
1010
wf = file(weave_file, 'rb')
1012
# FIXME: doesn't work on pipes
1013
weave_size = wf.tell()
1017
for i in range(vers):
1018
pb.update('checking sizes', i, vers)
1019
for origin, lineno, line in w._extract([i]):
1024
print 'versions %9d' % vers
1025
print 'weave file %9d bytes' % weave_size
1026
print 'total contents %9d bytes' % total
1027
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1030
print 'average size %9d bytes' % avg
1031
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1035
print """bzr weave tool
1037
Experimental tool for weave algorithm.
1040
weave init WEAVEFILE
1041
Create an empty weave file
1042
weave get WEAVEFILE VERSION
1043
Write out specified version.
1044
weave check WEAVEFILE
1045
Check consistency of all versions.
1047
Display table of contents.
1048
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1049
Add NEWTEXT, with specified parent versions.
1050
weave annotate WEAVEFILE VERSION
1051
Display origin of each line.
1052
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1053
Auto-merge two versions and display conflicts.
1054
weave diff WEAVEFILE VERSION1 VERSION2
1055
Show differences between two versions.
1059
% weave init foo.weave
1061
% weave add foo.weave ver0 < foo.txt
1064
(create updated version)
1066
% weave get foo.weave 0 | diff -u - foo.txt
1067
% weave add foo.weave ver1 0 < foo.txt
1070
% weave get foo.weave 0 > foo.txt (create forked version)
1072
% weave add foo.weave ver2 0 < foo.txt
1075
% weave merge foo.weave 1 2 > foo.txt (merge them)
1076
% vi foo.txt (resolve conflicts)
1077
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1089
# in case we're run directly from the subdirectory
1090
sys.path.append('..')
1092
from bzrlib.weavefile import write_weave, read_weave
1093
from bzrlib.progress import ProgressBar
1108
return read_weave(file(argv[2], 'rb'))
1114
# at the moment, based on everything in the file
1116
parents = map(int, argv[4:])
1117
lines = sys.stdin.readlines()
1118
ver = w.add(name, parents, lines)
1119
write_weave(w, file(argv[2], 'wb'))
1120
print 'added version %r %d' % (name, ver)
1123
if os.path.exists(fn):
1124
raise IOError("file exists")
1126
write_weave(w, file(fn, 'wb'))
1127
elif cmd == 'get': # get one version
1129
sys.stdout.writelines(w.get_iter(int(argv[3])))
1134
v1, v2 = map(int, argv[3:5])
1137
diff_gen = bzrlib.patiencediff.unified_diff(lines1, lines2,
1138
'%s version %d' % (fn, v1),
1139
'%s version %d' % (fn, v2))
1140
sys.stdout.writelines(diff_gen)
1142
elif cmd == 'annotate':
1144
# newline is added to all lines regardless; too hard to get
1145
# reasonable formatting otherwise
1147
for origin, text in w.annotate(int(argv[3])):
1148
text = text.rstrip('\r\n')
1150
print ' | %s' % (text)
1152
print '%5d | %s' % (origin, text)
1158
elif cmd == 'stats':
1159
weave_stats(argv[2], ProgressBar())
1161
elif cmd == 'check':
1166
print '%d versions ok' % w.num_versions()
1168
elif cmd == 'inclusions':
1170
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1172
elif cmd == 'parents':
1174
print ' '.join(map(str, w._parents[int(argv[3])]))
1176
elif cmd == 'plan-merge':
1177
# replaced by 'bzr weave-plan-merge'
1179
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1181
print '%14s | %s' % (state, line),
1182
elif cmd == 'merge':
1183
# replaced by 'bzr weave-merge-text'
1185
p = w.plan_merge(int(argv[3]), int(argv[4]))
1186
sys.stdout.writelines(w.weave_merge(p))
1188
raise ValueError('unknown command %r' % cmd)
1191
if __name__ == '__main__':
1193
sys.exit(main(sys.argv))
1196
class InterWeave(InterVersionedFile):
1197
"""Optimised code paths for weave to weave operations."""
1199
_matching_file_from_factory = staticmethod(WeaveFile)
1200
_matching_file_to_factory = staticmethod(WeaveFile)
1203
def is_compatible(source, target):
1204
"""Be compatible with weaves."""
1206
return (isinstance(source, Weave) and
1207
isinstance(target, Weave))
1208
except AttributeError:
1211
def join(self, pb=None, msg=None, version_ids=None, ignore_missing=False):
1212
"""See InterVersionedFile.join."""
1213
version_ids = self._get_source_version_ids(version_ids, ignore_missing)
1214
if self.target.versions() == [] and version_ids is None:
1215
self.target._copy_weave_content(self.source)
1217
self.target._join(self.source, pb, msg, version_ids, ignore_missing)
1220
InterVersionedFile.register_optimiser(InterWeave)
added
removed
bzrlib/weave.py
