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.
70
from cStringIO import StringIO
71
from difflib import SequenceMatcher
76
from bzrlib.trace import mutter
77
from bzrlib.errors import (WeaveError, WeaveFormatError, WeaveParentMismatch,
78
RevisionAlreadyPresent,
80
WeaveRevisionAlreadyPresent,
81
WeaveRevisionNotPresent,
83
import bzrlib.errors as errors
84
from bzrlib.osutils import sha_strings
85
from bzrlib.symbol_versioning import *
86
from bzrlib.tsort import topo_sort
87
from bzrlib.versionedfile import VersionedFile, InterVersionedFile
88
from bzrlib.weavefile import _read_weave_v5, write_weave_v5
91
class Weave(VersionedFile):
92
"""weave - versioned text file storage.
94
A Weave manages versions of line-based text files, keeping track
95
of the originating version for each line.
97
To clients the "lines" of the file are represented as a list of strings.
98
These strings will typically have terminal newline characters, but
99
this is not required. In particular files commonly do not have a newline
100
at the end of the file.
102
Texts can be identified in either of two ways:
104
* a nonnegative index number.
106
* a version-id string.
108
Typically the index number will be valid only inside this weave and
109
the version-id is used to reference it in the larger world.
111
The weave is represented as a list mixing edit instructions and
112
literal text. Each entry in _weave can be either a string (or
113
unicode), or a tuple. If a string, it means that the given line
114
should be output in the currently active revisions.
116
If a tuple, it gives a processing instruction saying in which
117
revisions the enclosed lines are active. The tuple has the form
118
(instruction, version).
120
The instruction can be '{' or '}' for an insertion block, and '['
121
and ']' for a deletion block respectively. The version is the
122
integer version index. There is no replace operator, only deletes
123
and inserts. For '}', the end of an insertion, there is no
124
version parameter because it always closes the most recently
129
* A later version can delete lines that were introduced by any
130
number of ancestor versions; this implies that deletion
131
instructions can span insertion blocks without regard to the
132
insertion block's nesting.
134
* Similarly, deletions need not be properly nested with regard to
135
each other, because they might have been generated by
136
independent revisions.
138
* Insertions are always made by inserting a new bracketed block
139
into a single point in the previous weave. This implies they
140
can nest but not overlap, and the nesting must always have later
141
insertions on the inside.
143
* It doesn't seem very useful to have an active insertion
144
inside an inactive insertion, but it might happen.
146
* Therefore, all instructions are always"considered"; that
147
is passed onto and off the stack. An outer inactive block
148
doesn't disable an inner block.
150
* Lines are enabled if the most recent enclosing insertion is
151
active and none of the enclosing deletions are active.
153
* There is no point having a deletion directly inside its own
154
insertion; you might as well just not write it. And there
155
should be no way to get an earlier version deleting a later
159
Text of the weave; list of control instruction tuples and strings.
162
List of parents, indexed by version number.
163
It is only necessary to store the minimal set of parents for
164
each version; the parent's parents are implied.
167
List of hex SHA-1 of each version.
170
List of symbolic names for each version. Each should be unique.
173
For each name, the version number.
176
Descriptive name of this weave; typically the filename if known.
180
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
183
def __init__(self, weave_name=None):
189
self._weave_name = weave_name
192
return "Weave(%r)" % self._weave_name
195
"""Return a deep copy of self.
197
The copy can be modified without affecting the original weave."""
199
other._weave = self._weave[:]
200
other._parents = self._parents[:]
201
other._sha1s = self._sha1s[:]
202
other._names = self._names[:]
203
other._name_map = self._name_map.copy()
204
other._weave_name = self._weave_name
207
def __eq__(self, other):
208
if not isinstance(other, Weave):
210
return self._parents == other._parents \
211
and self._weave == other._weave \
212
and self._sha1s == other._sha1s
214
def __ne__(self, other):
215
return not self.__eq__(other)
217
@deprecated_method(zero_eight)
218
def idx_to_name(self, index):
219
"""Old public interface, the public interface is all names now."""
222
def _idx_to_name(self, version):
223
return self._names[version]
225
@deprecated_method(zero_eight)
226
def lookup(self, name):
227
"""Backwards compatability thunk:
229
Return name, as name is valid in the api now, and spew deprecation
234
def _lookup(self, name):
235
"""Convert symbolic version name to index."""
237
return self._name_map[name]
239
raise RevisionNotPresent(name, self._weave_name)
241
@deprecated_method(zero_eight)
242
def iter_names(self):
243
"""Deprecated convenience function, please see VersionedFile.names()."""
244
return iter(self.names())
246
@deprecated_method(zero_eight)
248
"""See Weave.versions for the current api."""
249
return self.versions()
252
"""See VersionedFile.versions."""
253
return self._names[:]
255
def has_version(self, version_id):
256
"""See VersionedFile.has_version."""
257
return self._name_map.has_key(version_id)
259
__contains__ = has_version
261
def get_parents(self, version_id):
262
"""See VersionedFile.get_parent."""
263
return map(self._idx_to_name, self._parents[self._lookup(version_id)])
265
def _check_repeated_add(self, name, parents, text, sha1):
266
"""Check that a duplicated add is OK.
268
If it is, return the (old) index; otherwise raise an exception.
270
idx = self._lookup(name)
271
if sorted(self._parents[idx]) != sorted(parents) \
272
or sha1 != self._sha1s[idx]:
273
raise RevisionAlreadyPresent(name, self._weave_name)
276
@deprecated_method(zero_eight)
277
def add_identical(self, old_rev_id, new_rev_id, parents):
278
"""Please use Weave.clone_text now."""
279
return self.clone_text(new_rev_id, old_rev_id, parents)
281
def add_lines(self, version_id, parents, lines):
282
"""See VersionedFile.add_lines."""
283
return self._add(version_id, lines, map(self._lookup, parents))
285
@deprecated_method(zero_eight)
286
def add(self, name, parents, text, sha1=None):
287
"""See VersionedFile.add_lines for the non deprecated api."""
288
return self._add(name, text, map(self._maybe_lookup, parents), sha1)
290
def _add(self, version_id, lines, parents, sha1=None):
291
"""Add a single text on top of the weave.
293
Returns the index number of the newly added version.
296
Symbolic name for this version.
297
(Typically the revision-id of the revision that added it.)
300
List or set of direct parent version numbers.
303
Sequence of lines to be added in the new version.
306
assert isinstance(version_id, basestring)
308
sha1 = sha_strings(lines)
309
if version_id in self._name_map:
310
return self._check_repeated_add(version_id, parents, lines, sha1)
312
self._check_versions(parents)
313
## self._check_lines(lines)
314
new_version = len(self._parents)
316
# if we abort after here the (in-memory) weave will be corrupt because only
317
# some fields are updated
318
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
319
# - Robert Collins 20060226
320
self._parents.append(parents[:])
321
self._sha1s.append(sha1)
322
self._names.append(version_id)
323
self._name_map[version_id] = new_version
327
# special case; adding with no parents revision; can do
328
# this more quickly by just appending unconditionally.
329
# even more specially, if we're adding an empty text we
330
# need do nothing at all.
332
self._weave.append(('{', new_version))
333
self._weave.extend(lines)
334
self._weave.append(('}', None))
337
if len(parents) == 1:
338
pv = list(parents)[0]
339
if sha1 == self._sha1s[pv]:
340
# special case: same as the single parent
344
ancestors = self._inclusions(parents)
348
# basis a list of (origin, lineno, line)
351
for origin, lineno, line in self._extract(ancestors):
352
basis_lineno.append(lineno)
353
basis_lines.append(line)
355
# another small special case: a merge, producing the same text
357
if lines == basis_lines:
360
# add a sentinal, because we can also match against the final line
361
basis_lineno.append(len(self._weave))
363
# XXX: which line of the weave should we really consider
364
# matches the end of the file? the current code says it's the
365
# last line of the weave?
367
#print 'basis_lines:', basis_lines
368
#print 'new_lines: ', lines
370
s = SequenceMatcher(None, basis_lines, lines)
372
# offset gives the number of lines that have been inserted
373
# into the weave up to the current point; if the original edit instruction
374
# says to change line A then we actually change (A+offset)
377
for tag, i1, i2, j1, j2 in s.get_opcodes():
378
# i1,i2 are given in offsets within basis_lines; we need to map them
379
# back to offsets within the entire weave
380
#print 'raw match', tag, i1, i2, j1, j2
384
i1 = basis_lineno[i1]
385
i2 = basis_lineno[i2]
387
assert 0 <= j1 <= j2 <= len(lines)
389
#print tag, i1, i2, j1, j2
391
# the deletion and insertion are handled separately.
392
# first delete the region.
394
self._weave.insert(i1+offset, ('[', new_version))
395
self._weave.insert(i2+offset+1, (']', new_version))
399
# there may have been a deletion spanning up to
400
# i2; we want to insert after this region to make sure
401
# we don't destroy ourselves
403
self._weave[i:i] = ([('{', new_version)]
406
offset += 2 + (j2 - j1)
409
def clone_text(self, new_version_id, old_version_id, parents):
410
"""See VersionedFile.clone_text."""
411
old_lines = self.get_text(old_version_id)
412
self.add_lines(new_version_id, parents, old_lines)
414
def _inclusions(self, versions):
415
"""Return set of all ancestors of given version(s)."""
416
if not len(versions):
419
for v in xrange(max(versions), 0, -1):
421
# include all its parents
422
i.update(self._parents[v])
424
## except IndexError:
425
## raise ValueError("version %d not present in weave" % v)
427
@deprecated_method(zero_eight)
428
def inclusions(self, version_ids):
429
"""Deprecated - see VersionedFile.get_ancestry for the replacement."""
432
if isinstance(version_ids[0], int):
433
return [self._idx_to_name(v) for v in self._inclusions(version_ids)]
435
return self.get_ancestry(version_ids)
437
def get_ancestry(self, version_ids):
438
"""See VersionedFile.get_ancestry."""
439
if isinstance(version_ids, basestring):
440
version_ids = [version_ids]
441
i = self._inclusions([self._lookup(v) for v in version_ids])
442
return [self._idx_to_name(v) for v in i]
444
def _check_lines(self, text):
445
if not isinstance(text, list):
446
raise ValueError("text should be a list, not %s" % type(text))
449
if not isinstance(l, basestring):
450
raise ValueError("text line should be a string or unicode, not %s"
455
def _check_versions(self, indexes):
456
"""Check everything in the sequence of indexes is valid"""
461
raise IndexError("invalid version number %r" % i)
463
def _compatible_parents(self, my_parents, other_parents):
464
"""During join check that other_parents are joinable with my_parents.
466
Joinable is defined as 'is a subset of' - supersets may require
467
regeneration of diffs, but subsets do not.
469
return len(other_parents.difference(my_parents)) == 0
471
def annotate(self, version_id):
472
if isinstance(version_id, int):
473
warn('Weave.annotate(int) is deprecated. Please use version names'
474
' in all circumstances as of 0.8',
479
for origin, lineno, text in self._extract([version_id]):
480
result.append((origin, text))
483
return super(Weave, self).annotate(version_id)
485
def annotate_iter(self, version_id):
486
"""Yield list of (version-id, line) pairs for the specified version.
488
The index indicates when the line originated in the weave."""
489
incls = [self._lookup(version_id)]
490
for origin, lineno, text in self._extract(incls):
491
yield self._idx_to_name(origin), text
493
@deprecated_method(zero_eight)
495
"""_walk has become walk, a supported api."""
498
def walk(self, version_ids=None):
499
"""See VersionedFile.walk."""
504
lineno = 0 # line of weave, 0-based
506
for l in self._weave:
507
if isinstance(l, tuple):
511
istack.append(self._idx_to_name(v))
515
assert self._idx_to_name(v) not in dset
516
dset.add(self._idx_to_name(v))
518
dset.remove(self._idx_to_name(v))
520
raise WeaveFormatError('unexpected instruction %r' % v)
522
assert isinstance(l, basestring)
524
yield lineno, istack[-1], dset.copy(), 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 _extract(self, versions):
535
"""Yield annotation of lines in included set.
537
Yields a sequence of tuples (origin, lineno, text), where
538
origin is the origin version, lineno the index in the weave,
539
and text the text of the line.
541
The set typically but not necessarily corresponds to a version.
544
if not isinstance(i, int):
547
included = self._inclusions(versions)
552
lineno = 0 # line of weave, 0-based
558
WFE = WeaveFormatError
560
for l in self._weave:
561
if isinstance(l, tuple):
565
assert v not in istack
579
assert isinstance(l, basestring)
581
isactive = (not dset) and istack and (istack[-1] in included)
583
result.append((istack[-1], lineno, l))
586
raise WeaveFormatError("unclosed insertion blocks "
587
"at end of weave: %s" % istack)
589
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
593
@deprecated_method(zero_eight)
594
def get_iter(self, name_or_index):
595
"""Deprecated, please do not use. Lookups are not not needed.
597
Please use get_lines now.
599
return self._get_iter(self._maybe_lookup(name_or_index))
601
@deprecated_method(zero_eight)
602
def maybe_lookup(self, name_or_index):
603
"""Deprecated, please do not use. Lookups are not not needed."""
604
return self._maybe_lookup(name_or_index)
606
def _maybe_lookup(self, name_or_index):
607
"""Convert possible symbolic name to index, or pass through indexes.
611
if isinstance(name_or_index, (int, long)):
614
return self._lookup(name_or_index)
616
def _get_iter(self, version_id):
617
"""Yield lines for the specified version."""
618
incls = [self._maybe_lookup(version_id)]
623
# We don't have sha1 sums for multiple entries
625
for origin, lineno, line in self._extract(incls):
630
expected_sha1 = self._sha1s[index]
631
measured_sha1 = cur_sha.hexdigest()
632
if measured_sha1 != expected_sha1:
633
raise errors.WeaveInvalidChecksum(
634
'file %s, revision %s, expected: %s, measured %s'
635
% (self._weave_name, self._names[index],
636
expected_sha1, measured_sha1))
638
@deprecated_method(zero_eight)
639
def get(self, version_id):
640
"""Please use either Weave.get_text or Weave.get_lines as desired."""
641
return self.get_lines(version_id)
643
def get_lines(self, version_id):
644
"""See VersionedFile.get_lines()."""
645
return list(self._get_iter(version_id))
647
def get_sha1(self, name):
648
"""Get the stored sha1 sum for the given revision.
650
:param name: The name of the version to lookup
652
return self._sha1s[self._lookup(name)]
654
@deprecated_method(zero_eight)
655
def numversions(self):
656
"""How many versions are in this weave?
658
Deprecated in favour of num_versions.
660
return self.num_versions()
662
def num_versions(self):
663
"""How many versions are in this weave?"""
664
l = len(self._parents)
665
assert l == len(self._sha1s)
668
__len__ = num_versions
670
def check(self, progress_bar=None):
671
# TODO evaluate performance hit of using string sets in this routine.
672
# check no circular inclusions
673
for version in range(self.num_versions()):
674
inclusions = list(self._parents[version])
677
if inclusions[-1] >= version:
678
raise WeaveFormatError("invalid included version %d for index %d"
679
% (inclusions[-1], version))
681
# try extracting all versions; parallel extraction is used
682
nv = self.num_versions()
687
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
688
# The problem is that set membership is much more expensive
689
name = self._idx_to_name(i)
690
sha1s[name] = sha.new()
692
new_inc = set([name])
693
for p in self._parents[i]:
694
new_inc.update(inclusions[self._idx_to_name(p)])
696
assert set(new_inc) == set(self.get_ancestry(name)), \
697
'failed %s != %s' % (set(new_inc), set(self.get_ancestry(name)))
698
inclusions[name] = new_inc
700
nlines = len(self._weave)
702
update_text = 'checking weave'
704
short_name = os.path.basename(self._weave_name)
705
update_text = 'checking %s' % (short_name,)
706
update_text = update_text[:25]
708
for lineno, insert, deleteset, line in self.walk():
710
progress_bar.update(update_text, lineno, nlines)
712
for name, name_inclusions in inclusions.items():
713
# The active inclusion must be an ancestor,
714
# and no ancestors must have deleted this line,
715
# because we don't support resurrection.
716
if (insert in name_inclusions) and not (deleteset & name_inclusions):
717
sha1s[name].update(line)
720
version = self._idx_to_name(i)
721
hd = sha1s[version].hexdigest()
722
expected = self._sha1s[i]
724
raise errors.WeaveInvalidChecksum(
725
"mismatched sha1 for version %s: "
726
"got %s, expected %s"
727
% (version, hd, expected))
729
# TODO: check insertions are properly nested, that there are
730
# no lines outside of insertion blocks, that deletions are
731
# properly paired, etc.
733
def _join(self, other, pb, msg, version_ids, ignore_missing):
734
"""Worker routine for join()."""
735
if not other.versions():
736
return # nothing to update, easy
739
for version_id in version_ids:
740
if not self.has_version(version_id) and not ignore_missing:
741
raise RevisionNotPresent(version_id, self._weave_name)
743
version_ids = other.versions()
745
# two loops so that we do not change ourselves before verifying it
747
# work through in index order to make sure we get all dependencies
750
version_ids = set(other.versions()).intersection(set(version_ids))
751
pending_graph = [(version, other.get_parents(version)) for
752
version in version_ids]
753
for name in topo_sort(pending_graph):
754
other_idx = other._name_map[name]
755
self._check_version_consistent(other, other_idx, name)
756
sha1 = other._sha1s[other_idx]
760
if name in self._name_map:
761
idx = self._lookup(name)
762
n1 = set(map(other._idx_to_name, other._parents[other_idx]))
763
n2 = set(map(self._idx_to_name, self._parents[idx]))
764
if sha1 == self._sha1s[idx] and n1 == n2:
767
names_to_join.append((other_idx, name))
774
for other_idx, name in names_to_join:
775
# TODO: If all the parents of the other version are already
776
# present then we can avoid some work by just taking the delta
777
# and adjusting the offsets.
778
new_parents = self._imported_parents(other, other_idx)
779
sha1 = other._sha1s[other_idx]
784
pb.update(msg, merged, len(names_to_join))
786
lines = other.get_lines(other_idx)
787
self._add(name, lines, new_parents, sha1)
789
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
790
merged, processed, self._weave_name, time.time()-time0))
792
def _imported_parents(self, other, other_idx):
793
"""Return list of parents in self corresponding to indexes in other."""
795
for parent_idx in other._parents[other_idx]:
796
parent_name = other._names[parent_idx]
797
if parent_name not in self._names:
798
# should not be possible
799
raise WeaveError("missing parent {%s} of {%s} in %r"
800
% (parent_name, other._name_map[other_idx], self))
801
new_parents.append(self._name_map[parent_name])
804
def _check_version_consistent(self, other, other_idx, name):
805
"""Check if a version in consistent in this and other.
807
To be consistent it must have:
810
* the same direct parents (by name, not index, and disregarding
813
If present & correct return True;
814
if not present in self return False;
815
if inconsistent raise error."""
816
this_idx = self._name_map.get(name, -1)
818
if self._sha1s[this_idx] != other._sha1s[other_idx]:
819
raise errors.WeaveTextDiffers(name, self, other)
820
self_parents = self._parents[this_idx]
821
other_parents = other._parents[other_idx]
822
n1 = set([self._names[i] for i in self_parents])
823
n2 = set([other._names[i] for i in other_parents])
824
if not self._compatible_parents(n1, n2):
825
raise WeaveParentMismatch("inconsistent parents "
826
"for version {%s}: %s vs %s" % (name, n1, n2))
832
@deprecated_method(zero_eight)
833
def reweave(self, other, pb=None, msg=None):
834
"""reweave has been superceded by plain use of join."""
835
return self.join(other, pb, msg)
837
def _reweave(self, other, pb, msg):
838
"""Reweave self with other - internal helper for join().
840
:param other: The other weave to merge
841
:param pb: An optional progress bar, indicating how far done we are
842
:param msg: An optional message for the progress
844
new_weave = _reweave(self, other, pb=pb, msg=msg)
845
for attr in self.__slots__:
846
if attr != '_weave_name':
847
setattr(self, attr, getattr(new_weave, attr))
850
class WeaveFile(Weave):
851
"""A WeaveFile represents a Weave on disk and writes on change."""
853
WEAVE_SUFFIX = '.weave'
855
def __init__(self, name, transport, mode=None, create=False):
856
"""Create a WeaveFile.
858
:param create: If not True, only open an existing knit.
860
super(WeaveFile, self).__init__(name)
861
self._transport = transport
864
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
865
except errors.NoSuchFile:
871
def add_lines(self, version_id, parents, lines):
872
"""Add a version and save the weave."""
873
super(WeaveFile, self).add_lines(version_id, parents, lines)
876
def copy_to(self, name, transport):
877
"""See VersionedFile.copy_to()."""
878
# as we are all in memory always, just serialise to the new place.
880
write_weave_v5(self, sio)
882
transport.put(name + WeaveFile.WEAVE_SUFFIX, sio, self._mode)
884
def create_empty(self, name, transport, mode=None):
885
return WeaveFile(name, transport, mode, create=True)
888
"""Save the weave."""
890
write_weave_v5(self, sio)
892
self._transport.put(self._weave_name + WeaveFile.WEAVE_SUFFIX,
898
"""See VersionedFile.get_suffixes()."""
899
return [WeaveFile.WEAVE_SUFFIX]
901
def join(self, other, pb=None, msg=None, version_ids=None,
902
ignore_missing=False):
903
"""Join other into self and save."""
904
super(WeaveFile, self).join(other, pb, msg, version_ids, ignore_missing)
908
@deprecated_function(zero_eight)
909
def reweave(wa, wb, pb=None, msg=None):
910
"""reweaving is deprecation, please just use weave.join()."""
911
_reweave(wa, wb, pb, msg)
913
def _reweave(wa, wb, pb=None, msg=None):
914
"""Combine two weaves and return the result.
916
This works even if a revision R has different parents in
917
wa and wb. In the resulting weave all the parents are given.
919
This is done by just building up a new weave, maintaining ordering
920
of the versions in the two inputs. More efficient approaches
921
might be possible but it should only be necessary to do
922
this operation rarely, when a new previously ghost version is
925
:param pb: An optional progress bar, indicating how far done we are
926
:param msg: An optional message for the progress
930
queue_a = range(wa.num_versions())
931
queue_b = range(wb.num_versions())
932
# first determine combined parents of all versions
933
# map from version name -> all parent names
934
combined_parents = _reweave_parent_graphs(wa, wb)
935
mutter("combined parents: %r", combined_parents)
936
order = topo_sort(combined_parents.iteritems())
937
mutter("order to reweave: %r", order)
942
for idx, name in enumerate(order):
944
pb.update(msg, idx, len(order))
945
if name in wa._name_map:
946
lines = wa.get_lines(name)
947
if name in wb._name_map:
948
lines_b = wb.get_lines(name)
950
mutter('Weaves differ on content. rev_id {%s}', name)
951
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
953
lines = list(difflib.unified_diff(lines, lines_b,
954
wa._weave_name, wb._weave_name))
955
mutter('lines:\n%s', ''.join(lines))
956
raise errors.WeaveTextDiffers(name, wa, wb)
958
lines = wb.get_lines(name)
959
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
962
def _reweave_parent_graphs(wa, wb):
963
"""Return combined parent ancestry for two weaves.
965
Returned as a list of (version_name, set(parent_names))"""
967
for weave in [wa, wb]:
968
for idx, name in enumerate(weave._names):
969
p = combined.setdefault(name, set())
970
p.update(map(weave._idx_to_name, weave._parents[idx]))
975
"""Show the weave's table-of-contents"""
976
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
977
for i in (6, 50, 10, 10):
980
for i in range(w.num_versions()):
983
parent_str = ' '.join(map(str, w._parents[i]))
984
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
988
def weave_stats(weave_file, pb):
989
from bzrlib.weavefile import read_weave
991
wf = file(weave_file, 'rb')
992
w = read_weave(wf, WeaveVersionedFile)
993
# FIXME: doesn't work on pipes
994
weave_size = wf.tell()
998
for i in range(vers):
999
pb.update('checking sizes', i, vers)
1000
for origin, lineno, line in w._extract([i]):
1005
print 'versions %9d' % vers
1006
print 'weave file %9d bytes' % weave_size
1007
print 'total contents %9d bytes' % total
1008
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1011
print 'average size %9d bytes' % avg
1012
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1016
print """bzr weave tool
1018
Experimental tool for weave algorithm.
1021
weave init WEAVEFILE
1022
Create an empty weave file
1023
weave get WEAVEFILE VERSION
1024
Write out specified version.
1025
weave check WEAVEFILE
1026
Check consistency of all versions.
1028
Display table of contents.
1029
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1030
Add NEWTEXT, with specified parent versions.
1031
weave annotate WEAVEFILE VERSION
1032
Display origin of each line.
1033
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1034
Auto-merge two versions and display conflicts.
1035
weave diff WEAVEFILE VERSION1 VERSION2
1036
Show differences between two versions.
1040
% weave init foo.weave
1042
% weave add foo.weave ver0 < foo.txt
1045
(create updated version)
1047
% weave get foo.weave 0 | diff -u - foo.txt
1048
% weave add foo.weave ver1 0 < foo.txt
1051
% weave get foo.weave 0 > foo.txt (create forked version)
1053
% weave add foo.weave ver2 0 < foo.txt
1056
% weave merge foo.weave 1 2 > foo.txt (merge them)
1057
% vi foo.txt (resolve conflicts)
1058
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1070
# in case we're run directly from the subdirectory
1071
sys.path.append('..')
1073
from bzrlib.weavefile import write_weave, read_weave
1074
from bzrlib.progress import ProgressBar
1089
return read_weave(file(argv[2], 'rb'))
1095
# at the moment, based on everything in the file
1097
parents = map(int, argv[4:])
1098
lines = sys.stdin.readlines()
1099
ver = w.add(name, parents, lines)
1100
write_weave(w, file(argv[2], 'wb'))
1101
print 'added version %r %d' % (name, ver)
1104
if os.path.exists(fn):
1105
raise IOError("file exists")
1107
write_weave(w, file(fn, 'wb'))
1108
elif cmd == 'get': # get one version
1110
sys.stdout.writelines(w.get_iter(int(argv[3])))
1113
from difflib import unified_diff
1116
v1, v2 = map(int, argv[3:5])
1119
diff_gen = unified_diff(lines1, lines2,
1120
'%s version %d' % (fn, v1),
1121
'%s version %d' % (fn, v2))
1122
sys.stdout.writelines(diff_gen)
1124
elif cmd == 'annotate':
1126
# newline is added to all lines regardless; too hard to get
1127
# reasonable formatting otherwise
1129
for origin, text in w.annotate(int(argv[3])):
1130
text = text.rstrip('\r\n')
1132
print ' | %s' % (text)
1134
print '%5d | %s' % (origin, text)
1140
elif cmd == 'stats':
1141
weave_stats(argv[2], ProgressBar())
1143
elif cmd == 'check':
1148
print '%d versions ok' % w.num_versions()
1150
elif cmd == 'inclusions':
1152
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1154
elif cmd == 'parents':
1156
print ' '.join(map(str, w._parents[int(argv[3])]))
1158
elif cmd == 'plan-merge':
1160
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1162
print '%14s | %s' % (state, line),
1164
elif cmd == 'merge':
1166
p = w.plan_merge(int(argv[3]), int(argv[4]))
1167
sys.stdout.writelines(w.weave_merge(p))
1170
raise ValueError('unknown command %r' % cmd)
1174
def profile_main(argv):
1175
import tempfile, hotshot, hotshot.stats
1177
prof_f = tempfile.NamedTemporaryFile()
1179
prof = hotshot.Profile(prof_f.name)
1181
ret = prof.runcall(main, argv)
1184
stats = hotshot.stats.load(prof_f.name)
1186
stats.sort_stats('cumulative')
1187
## XXX: Might like to write to stderr or the trace file instead but
1188
## print_stats seems hardcoded to stdout
1189
stats.print_stats(20)
1194
def lsprofile_main(argv):
1195
from bzrlib.lsprof import profile
1196
ret,stats = profile(main, argv)
1202
if __name__ == '__main__':
1204
if '--profile' in sys.argv:
1206
args.remove('--profile')
1207
sys.exit(profile_main(args))
1208
elif '--lsprof' in sys.argv:
1210
args.remove('--lsprof')
1211
sys.exit(lsprofile_main(args))
1213
sys.exit(main(sys.argv))
1216
class InterWeave(InterVersionedFile):
1217
"""Optimised code paths for weave to weave operations."""
1219
_matching_file_factory = staticmethod(WeaveFile)
1222
def is_compatible(source, target):
1223
"""Be compatible with weaves."""
1225
return (isinstance(source, Weave) and
1226
isinstance(target, Weave))
1227
except AttributeError:
1230
def join(self, pb=None, msg=None, version_ids=None, ignore_missing=False):
1231
"""See InterVersionedFile.join."""
1233
self.target._join(self.source, pb, msg, version_ids, ignore_missing)
1234
except errors.WeaveParentMismatch:
1235
self.target._reweave(self.source, pb, msg)
1238
InterVersionedFile.register_optimiser(InterWeave)
added
removed
bzrlib/weave.py
