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
self._parents.append(parents[:])
319
self._sha1s.append(sha1)
320
self._names.append(version_id)
321
self._name_map[version_id] = new_version
325
# special case; adding with no parents revision; can do
326
# this more quickly by just appending unconditionally.
327
# even more specially, if we're adding an empty text we
328
# need do nothing at all.
330
self._weave.append(('{', new_version))
331
self._weave.extend(lines)
332
self._weave.append(('}', None))
335
if len(parents) == 1:
336
pv = list(parents)[0]
337
if sha1 == self._sha1s[pv]:
338
# special case: same as the single parent
342
ancestors = self._inclusions(parents)
346
# basis a list of (origin, lineno, line)
349
for origin, lineno, line in self._extract(ancestors):
350
basis_lineno.append(lineno)
351
basis_lines.append(line)
353
# another small special case: a merge, producing the same text
355
if lines == basis_lines:
358
# add a sentinal, because we can also match against the final line
359
basis_lineno.append(len(self._weave))
361
# XXX: which line of the weave should we really consider
362
# matches the end of the file? the current code says it's the
363
# last line of the weave?
365
#print 'basis_lines:', basis_lines
366
#print 'new_lines: ', lines
368
s = SequenceMatcher(None, basis_lines, lines)
370
# offset gives the number of lines that have been inserted
371
# into the weave up to the current point; if the original edit instruction
372
# says to change line A then we actually change (A+offset)
375
for tag, i1, i2, j1, j2 in s.get_opcodes():
376
# i1,i2 are given in offsets within basis_lines; we need to map them
377
# back to offsets within the entire weave
378
#print 'raw match', tag, i1, i2, j1, j2
382
i1 = basis_lineno[i1]
383
i2 = basis_lineno[i2]
385
assert 0 <= j1 <= j2 <= len(lines)
387
#print tag, i1, i2, j1, j2
389
# the deletion and insertion are handled separately.
390
# first delete the region.
392
self._weave.insert(i1+offset, ('[', new_version))
393
self._weave.insert(i2+offset+1, (']', new_version))
397
# there may have been a deletion spanning up to
398
# i2; we want to insert after this region to make sure
399
# we don't destroy ourselves
401
self._weave[i:i] = ([('{', new_version)]
404
offset += 2 + (j2 - j1)
407
def clone_text(self, new_version_id, old_version_id, parents):
408
"""See VersionedFile.clone_text."""
409
old_lines = self.get_text(old_version_id)
410
self.add_lines(new_version_id, parents, old_lines)
412
def _inclusions(self, versions):
413
"""Return set of all ancestors of given version(s)."""
415
for v in xrange(max(versions), 0, -1):
417
# include all its parents
418
i.update(self._parents[v])
420
## except IndexError:
421
## raise ValueError("version %d not present in weave" % v)
423
@deprecated_method(zero_eight)
424
def inclusions(self, version_ids):
425
"""Deprecated - see VersionedFile.get_ancestry for the replacement."""
428
if isinstance(version_ids[0], int):
429
return [self._idx_to_name(v) for v in self._inclusions(version_ids)]
431
return self.get_ancestry(version_ids)
433
def get_ancestry(self, version_ids):
434
"""See VersionedFile.get_ancestry."""
435
if isinstance(version_ids, basestring):
436
version_ids = [version_ids]
437
i = self._inclusions([self._lookup(v) for v in version_ids])
438
return [self._idx_to_name(v) for v in i]
440
def _check_lines(self, text):
441
if not isinstance(text, list):
442
raise ValueError("text should be a list, not %s" % type(text))
445
if not isinstance(l, basestring):
446
raise ValueError("text line should be a string or unicode, not %s"
451
def _check_versions(self, indexes):
452
"""Check everything in the sequence of indexes is valid"""
457
raise IndexError("invalid version number %r" % i)
459
def annotate(self, version_id):
460
if isinstance(version_id, int):
461
warn('Weave.annotate(int) is deprecated. Please use version names'
462
' in all circumstances as of 0.8',
467
for origin, lineno, text in self._extract([version_id]):
468
result.append((origin, text))
471
return super(Weave, self).annotate(version_id)
473
def annotate_iter(self, version_id):
474
"""Yield list of (version-id, line) pairs for the specified version.
476
The index indicates when the line originated in the weave."""
477
incls = [self._lookup(version_id)]
478
for origin, lineno, text in self._extract(incls):
479
yield self._idx_to_name(origin), text
481
@deprecated_method(zero_eight)
483
"""_walk has become walk, a supported api."""
486
def walk(self, version_ids=None):
487
"""See VersionedFile.walk."""
492
lineno = 0 # line of weave, 0-based
494
for l in self._weave:
495
if isinstance(l, tuple):
499
istack.append(self._idx_to_name(v))
503
assert self._idx_to_name(v) not in dset
504
dset.add(self._idx_to_name(v))
506
dset.remove(self._idx_to_name(v))
508
raise WeaveFormatError('unexpected instruction %r' % v)
510
assert isinstance(l, basestring)
512
yield lineno, istack[-1], dset.copy(), l
516
raise WeaveFormatError("unclosed insertion blocks "
517
"at end of weave: %s" % istack)
519
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
522
def _extract(self, versions):
523
"""Yield annotation of lines in included set.
525
Yields a sequence of tuples (origin, lineno, text), where
526
origin is the origin version, lineno the index in the weave,
527
and text the text of the line.
529
The set typically but not necessarily corresponds to a version.
532
if not isinstance(i, int):
535
included = self._inclusions(versions)
540
lineno = 0 # line of weave, 0-based
546
WFE = WeaveFormatError
548
for l in self._weave:
549
if isinstance(l, tuple):
553
assert v not in istack
567
assert isinstance(l, basestring)
569
isactive = (not dset) and istack and (istack[-1] in included)
571
result.append((istack[-1], lineno, l))
574
raise WeaveFormatError("unclosed insertion blocks "
575
"at end of weave: %s" % istack)
577
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
581
@deprecated_method(zero_eight)
582
def get_iter(self, name_or_index):
583
"""Deprecated, please do not use. Lookups are not not needed.
585
Please use get_lines now.
587
return self._get_iter(self._maybe_lookup(name_or_index))
589
@deprecated_method(zero_eight)
590
def maybe_lookup(self, name_or_index):
591
"""Deprecated, please do not use. Lookups are not not needed."""
592
return self._maybe_lookup(name_or_index)
594
def _maybe_lookup(self, name_or_index):
595
"""Convert possible symbolic name to index, or pass through indexes.
599
if isinstance(name_or_index, (int, long)):
602
return self._lookup(name_or_index)
604
def _get_iter(self, version_id):
605
"""Yield lines for the specified version."""
606
incls = [self._maybe_lookup(version_id)]
611
# We don't have sha1 sums for multiple entries
613
for origin, lineno, line in self._extract(incls):
618
expected_sha1 = self._sha1s[index]
619
measured_sha1 = cur_sha.hexdigest()
620
if measured_sha1 != expected_sha1:
621
raise errors.WeaveInvalidChecksum(
622
'file %s, revision %s, expected: %s, measured %s'
623
% (self._weave_name, self._names[index],
624
expected_sha1, measured_sha1))
626
@deprecated_method(zero_eight)
627
def get(self, version_id):
628
"""Please use either Weave.get_text or Weave.get_lines as desired."""
629
return self.get_lines(version_id)
631
def get_lines(self, version_id):
632
"""See VersionedFile.get_lines()."""
633
return list(self._get_iter(version_id))
635
def get_sha1(self, name):
636
"""Get the stored sha1 sum for the given revision.
638
:param name: The name of the version to lookup
640
return self._sha1s[self._lookup(name)]
642
@deprecated_method(zero_eight)
643
def numversions(self):
644
"""How many versions are in this weave?
646
Deprecated in favour of num_versions.
648
return self.num_versions()
650
def num_versions(self):
651
"""How many versions are in this weave?"""
652
l = len(self._parents)
653
assert l == len(self._sha1s)
656
__len__ = num_versions
658
def check(self, progress_bar=None):
659
# TODO evaluate performance hit of using string sets in this routine.
660
# check no circular inclusions
661
for version in range(self.num_versions()):
662
inclusions = list(self._parents[version])
665
if inclusions[-1] >= version:
666
raise WeaveFormatError("invalid included version %d for index %d"
667
% (inclusions[-1], version))
669
# try extracting all versions; parallel extraction is used
670
nv = self.num_versions()
675
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
676
# The problem is that set membership is much more expensive
677
name = self._idx_to_name(i)
678
sha1s[name] = sha.new()
680
new_inc = set([name])
681
for p in self._parents[i]:
682
new_inc.update(inclusions[self._idx_to_name(p)])
684
assert set(new_inc) == set(self.get_ancestry(name)), \
685
'failed %s != %s' % (set(new_inc), set(self.get_ancestry(name)))
686
inclusions[name] = new_inc
688
nlines = len(self._weave)
690
update_text = 'checking weave'
692
short_name = os.path.basename(self._weave_name)
693
update_text = 'checking %s' % (short_name,)
694
update_text = update_text[:25]
696
for lineno, insert, deleteset, line in self.walk():
698
progress_bar.update(update_text, lineno, nlines)
700
for name, name_inclusions in inclusions.items():
701
# The active inclusion must be an ancestor,
702
# and no ancestors must have deleted this line,
703
# because we don't support resurrection.
704
if (insert in name_inclusions) and not (deleteset & name_inclusions):
705
sha1s[name].update(line)
708
version = self._idx_to_name(i)
709
hd = sha1s[version].hexdigest()
710
expected = self._sha1s[i]
712
raise errors.WeaveInvalidChecksum(
713
"mismatched sha1 for version %s: "
714
"got %s, expected %s"
715
% (version, hd, expected))
717
# TODO: check insertions are properly nested, that there are
718
# no lines outside of insertion blocks, that deletions are
719
# properly paired, etc.
721
def _join(self, other, pb, msg, version_ids):
722
"""Worker routine for join()."""
723
if not other.versions():
724
return # nothing to update, easy
727
for version_id in version_ids:
728
if not self.has_version(version_id):
729
raise RevisionNotPresent(version_id, self._weave_name)
730
assert version_ids == None
732
# two loops so that we do not change ourselves before verifying it
734
# work through in index order to make sure we get all dependencies
737
for other_idx, name in enumerate(other._names):
738
self._check_version_consistent(other, other_idx, name)
739
sha1 = other._sha1s[other_idx]
743
if name in self._name_map:
744
idx = self._lookup(name)
745
n1 = set(map(other._idx_to_name, other._parents[other_idx]))
746
n2 = set(map(self._idx_to_name, self._parents[idx]))
747
if sha1 == self._sha1s[idx] and n1 == n2:
750
names_to_join.append((other_idx, name))
757
for other_idx, name in names_to_join:
758
# TODO: If all the parents of the other version are already
759
# present then we can avoid some work by just taking the delta
760
# and adjusting the offsets.
761
new_parents = self._imported_parents(other, other_idx)
762
sha1 = other._sha1s[other_idx]
767
pb.update(msg, merged, len(names_to_join))
769
lines = other.get_lines(other_idx)
770
self._add(name, lines, new_parents, sha1)
772
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
773
merged, processed, self._weave_name, time.time()-time0))
775
def _imported_parents(self, other, other_idx):
776
"""Return list of parents in self corresponding to indexes in other."""
778
for parent_idx in other._parents[other_idx]:
779
parent_name = other._names[parent_idx]
780
if parent_name not in self._names:
781
# should not be possible
782
raise WeaveError("missing parent {%s} of {%s} in %r"
783
% (parent_name, other._name_map[other_idx], self))
784
new_parents.append(self._name_map[parent_name])
787
def _check_version_consistent(self, other, other_idx, name):
788
"""Check if a version in consistent in this and other.
790
To be consistent it must have:
793
* the same direct parents (by name, not index, and disregarding
796
If present & correct return True;
797
if not present in self return False;
798
if inconsistent raise error."""
799
this_idx = self._name_map.get(name, -1)
801
if self._sha1s[this_idx] != other._sha1s[other_idx]:
802
raise errors.WeaveTextDiffers(name, self, other)
803
self_parents = self._parents[this_idx]
804
other_parents = other._parents[other_idx]
805
n1 = set([self._names[i] for i in self_parents])
806
n2 = set([other._names[i] for i in other_parents])
808
raise WeaveParentMismatch("inconsistent parents "
809
"for version {%s}: %s vs %s" % (name, n1, n2))
815
@deprecated_method(zero_eight)
816
def reweave(self, other, pb=None, msg=None):
817
"""reweave has been superceded by plain use of join."""
818
return self.join(other, pb, msg)
820
def _reweave(self, other, pb, msg):
821
"""Reweave self with other - internal helper for join().
823
:param other: The other weave to merge
824
:param pb: An optional progress bar, indicating how far done we are
825
:param msg: An optional message for the progress
827
new_weave = _reweave(self, other, pb=pb, msg=msg)
828
for attr in self.__slots__:
829
if attr != '_weave_name':
830
setattr(self, attr, getattr(new_weave, attr))
833
class WeaveFile(Weave):
834
"""A WeaveFile represents a Weave on disk and writes on change."""
836
WEAVE_SUFFIX = '.weave'
838
def __init__(self, name, transport, mode=None):
839
super(WeaveFile, self).__init__(name)
840
self._transport = transport
843
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
844
except errors.NoSuchFile:
848
def add_lines(self, version_id, parents, lines):
849
"""Add a version and save the weave."""
850
super(WeaveFile, self).add_lines(version_id, parents, lines)
853
def copy_to(self, name, transport):
854
"""See VersionedFile.copy_to()."""
855
# as we are all in memory always, just serialise to the new place.
857
write_weave_v5(self, sio)
859
transport.put(name + WeaveFile.WEAVE_SUFFIX, sio, self._mode)
861
def create_empty(self, name, transport, mode=None):
862
return WeaveFile(name, transport, mode)
865
"""Save the weave."""
867
write_weave_v5(self, sio)
869
self._transport.put(self._weave_name + WeaveFile.WEAVE_SUFFIX,
875
"""See VersionedFile.get_suffixes()."""
876
return [WeaveFile.WEAVE_SUFFIX]
878
def join(self, other, pb=None, msg=None, version_ids=None):
879
"""Join other into self and save."""
880
super(WeaveFile, self).join(other, pb, msg, version_ids)
884
@deprecated_function(zero_eight)
885
def reweave(wa, wb, pb=None, msg=None):
886
"""reweaving is deprecation, please just use weave.join()."""
887
_reweave(wa, wb, pb, msg)
889
def _reweave(wa, wb, pb=None, msg=None):
890
"""Combine two weaves and return the result.
892
This works even if a revision R has different parents in
893
wa and wb. In the resulting weave all the parents are given.
895
This is done by just building up a new weave, maintaining ordering
896
of the versions in the two inputs. More efficient approaches
897
might be possible but it should only be necessary to do
898
this operation rarely, when a new previously ghost version is
901
:param pb: An optional progress bar, indicating how far done we are
902
:param msg: An optional message for the progress
906
queue_a = range(wa.num_versions())
907
queue_b = range(wb.num_versions())
908
# first determine combined parents of all versions
909
# map from version name -> all parent names
910
combined_parents = _reweave_parent_graphs(wa, wb)
911
mutter("combined parents: %r", combined_parents)
912
order = topo_sort(combined_parents.iteritems())
913
mutter("order to reweave: %r", order)
918
for idx, name in enumerate(order):
920
pb.update(msg, idx, len(order))
921
if name in wa._name_map:
922
lines = wa.get_lines(name)
923
if name in wb._name_map:
924
lines_b = wb.get_lines(name)
926
mutter('Weaves differ on content. rev_id {%s}', name)
927
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
929
lines = list(difflib.unified_diff(lines, lines_b,
930
wa._weave_name, wb._weave_name))
931
mutter('lines:\n%s', ''.join(lines))
932
raise errors.WeaveTextDiffers(name, wa, wb)
934
lines = wb.get_lines(name)
935
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
938
def _reweave_parent_graphs(wa, wb):
939
"""Return combined parent ancestry for two weaves.
941
Returned as a list of (version_name, set(parent_names))"""
943
for weave in [wa, wb]:
944
for idx, name in enumerate(weave._names):
945
p = combined.setdefault(name, set())
946
p.update(map(weave._idx_to_name, weave._parents[idx]))
951
"""Show the weave's table-of-contents"""
952
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
953
for i in (6, 50, 10, 10):
956
for i in range(w.num_versions()):
959
parent_str = ' '.join(map(str, w._parents[i]))
960
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
964
def weave_stats(weave_file, pb):
965
from bzrlib.weavefile import read_weave
967
wf = file(weave_file, 'rb')
968
w = read_weave(wf, WeaveVersionedFile)
969
# FIXME: doesn't work on pipes
970
weave_size = wf.tell()
974
for i in range(vers):
975
pb.update('checking sizes', i, vers)
976
for origin, lineno, line in w._extract([i]):
981
print 'versions %9d' % vers
982
print 'weave file %9d bytes' % weave_size
983
print 'total contents %9d bytes' % total
984
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
987
print 'average size %9d bytes' % avg
988
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
992
print """bzr weave tool
994
Experimental tool for weave algorithm.
998
Create an empty weave file
999
weave get WEAVEFILE VERSION
1000
Write out specified version.
1001
weave check WEAVEFILE
1002
Check consistency of all versions.
1004
Display table of contents.
1005
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1006
Add NEWTEXT, with specified parent versions.
1007
weave annotate WEAVEFILE VERSION
1008
Display origin of each line.
1009
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1010
Auto-merge two versions and display conflicts.
1011
weave diff WEAVEFILE VERSION1 VERSION2
1012
Show differences between two versions.
1016
% weave init foo.weave
1018
% weave add foo.weave ver0 < foo.txt
1021
(create updated version)
1023
% weave get foo.weave 0 | diff -u - foo.txt
1024
% weave add foo.weave ver1 0 < foo.txt
1027
% weave get foo.weave 0 > foo.txt (create forked version)
1029
% weave add foo.weave ver2 0 < foo.txt
1032
% weave merge foo.weave 1 2 > foo.txt (merge them)
1033
% vi foo.txt (resolve conflicts)
1034
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1046
# in case we're run directly from the subdirectory
1047
sys.path.append('..')
1049
from bzrlib.weavefile import write_weave, read_weave
1050
from bzrlib.progress import ProgressBar
1065
return read_weave(file(argv[2], 'rb'))
1071
# at the moment, based on everything in the file
1073
parents = map(int, argv[4:])
1074
lines = sys.stdin.readlines()
1075
ver = w.add(name, parents, lines)
1076
write_weave(w, file(argv[2], 'wb'))
1077
print 'added version %r %d' % (name, ver)
1080
if os.path.exists(fn):
1081
raise IOError("file exists")
1083
write_weave(w, file(fn, 'wb'))
1084
elif cmd == 'get': # get one version
1086
sys.stdout.writelines(w.get_iter(int(argv[3])))
1089
from difflib import unified_diff
1092
v1, v2 = map(int, argv[3:5])
1095
diff_gen = unified_diff(lines1, lines2,
1096
'%s version %d' % (fn, v1),
1097
'%s version %d' % (fn, v2))
1098
sys.stdout.writelines(diff_gen)
1100
elif cmd == 'annotate':
1102
# newline is added to all lines regardless; too hard to get
1103
# reasonable formatting otherwise
1105
for origin, text in w.annotate(int(argv[3])):
1106
text = text.rstrip('\r\n')
1108
print ' | %s' % (text)
1110
print '%5d | %s' % (origin, text)
1116
elif cmd == 'stats':
1117
weave_stats(argv[2], ProgressBar())
1119
elif cmd == 'check':
1124
print '%d versions ok' % w.num_versions()
1126
elif cmd == 'inclusions':
1128
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1130
elif cmd == 'parents':
1132
print ' '.join(map(str, w._parents[int(argv[3])]))
1134
elif cmd == 'plan-merge':
1136
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1138
print '%14s | %s' % (state, line),
1140
elif cmd == 'merge':
1142
p = w.plan_merge(int(argv[3]), int(argv[4]))
1143
sys.stdout.writelines(w.weave_merge(p))
1146
raise ValueError('unknown command %r' % cmd)
1150
def profile_main(argv):
1151
import tempfile, hotshot, hotshot.stats
1153
prof_f = tempfile.NamedTemporaryFile()
1155
prof = hotshot.Profile(prof_f.name)
1157
ret = prof.runcall(main, argv)
1160
stats = hotshot.stats.load(prof_f.name)
1162
stats.sort_stats('cumulative')
1163
## XXX: Might like to write to stderr or the trace file instead but
1164
## print_stats seems hardcoded to stdout
1165
stats.print_stats(20)
1170
def lsprofile_main(argv):
1171
from bzrlib.lsprof import profile
1172
ret,stats = profile(main, argv)
1178
if __name__ == '__main__':
1180
if '--profile' in sys.argv:
1182
args.remove('--profile')
1183
sys.exit(profile_main(args))
1184
elif '--lsprof' in sys.argv:
1186
args.remove('--lsprof')
1187
sys.exit(lsprofile_main(args))
1189
sys.exit(main(sys.argv))
1192
class InterWeave(InterVersionedFile):
1193
"""Optimised code paths for weave to weave operations."""
1195
_matching_file_factory = staticmethod(WeaveFile)
1198
def is_compatible(source, target):
1199
"""Be compatible with weaves."""
1201
return (isinstance(source, Weave) and
1202
isinstance(target, Weave))
1203
except AttributeError:
1206
def join(self, pb=None, msg=None, version_ids=None):
1207
"""See InterVersionedFile.join."""
1209
self.target._join(self.source, pb, msg, version_ids)
1210
except errors.WeaveParentMismatch:
1211
self.target._reweave(self.source, pb, msg)
1214
InterVersionedFile.register_optimiser(InterWeave)
added
removed
bzrlib/weave.py
