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 visit, a supported api."""
496
return self._walk_internal()
498
def iter_lines_added_or_present_in_versions(self, version_ids=None):
499
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
500
if version_ids is None:
501
version_ids = self.versions()
502
version_ids = set(version_ids)
503
for lineno, inserted, deletes, line in self._walk_internal(version_ids):
504
# if inserted not in version_ids then it was inserted before the
505
# versions we care about, but because weaves cannot represent ghosts
506
# properly, we do not filter down to that
507
# if inserted not in version_ids: continue
513
#@deprecated_method(zero_eight)
514
def walk(self, version_ids=None):
515
"""See VersionedFile.walk."""
516
return self._walk_internal(version_ids)
518
def _walk_internal(self, version_ids=None):
519
"""Helper method for weave actions."""
524
lineno = 0 # line of weave, 0-based
526
for l in self._weave:
527
if isinstance(l, tuple):
531
istack.append(self._idx_to_name(v))
535
assert self._idx_to_name(v) not in dset
536
dset.add(self._idx_to_name(v))
538
dset.remove(self._idx_to_name(v))
540
raise WeaveFormatError('unexpected instruction %r' % v)
542
assert isinstance(l, basestring)
544
yield lineno, istack[-1], dset.copy(), l
548
raise WeaveFormatError("unclosed insertion blocks "
549
"at end of weave: %s" % istack)
551
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
554
def _extract(self, versions):
555
"""Yield annotation of lines in included set.
557
Yields a sequence of tuples (origin, lineno, text), where
558
origin is the origin version, lineno the index in the weave,
559
and text the text of the line.
561
The set typically but not necessarily corresponds to a version.
564
if not isinstance(i, int):
567
included = self._inclusions(versions)
572
lineno = 0 # line of weave, 0-based
578
WFE = WeaveFormatError
580
for l in self._weave:
581
if isinstance(l, tuple):
585
assert v not in istack
599
assert isinstance(l, basestring)
601
isactive = (not dset) and istack and (istack[-1] in included)
603
result.append((istack[-1], lineno, l))
606
raise WeaveFormatError("unclosed insertion blocks "
607
"at end of weave: %s" % istack)
609
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
613
@deprecated_method(zero_eight)
614
def get_iter(self, name_or_index):
615
"""Deprecated, please do not use. Lookups are not not needed.
617
Please use get_lines now.
619
return self._get_iter(self._maybe_lookup(name_or_index))
621
@deprecated_method(zero_eight)
622
def maybe_lookup(self, name_or_index):
623
"""Deprecated, please do not use. Lookups are not not needed."""
624
return self._maybe_lookup(name_or_index)
626
def _maybe_lookup(self, name_or_index):
627
"""Convert possible symbolic name to index, or pass through indexes.
631
if isinstance(name_or_index, (int, long)):
634
return self._lookup(name_or_index)
636
def _get_iter(self, version_id):
637
"""Yield lines for the specified version."""
638
incls = [self._maybe_lookup(version_id)]
643
# We don't have sha1 sums for multiple entries
645
for origin, lineno, line in self._extract(incls):
650
expected_sha1 = self._sha1s[index]
651
measured_sha1 = cur_sha.hexdigest()
652
if measured_sha1 != expected_sha1:
653
raise errors.WeaveInvalidChecksum(
654
'file %s, revision %s, expected: %s, measured %s'
655
% (self._weave_name, self._names[index],
656
expected_sha1, measured_sha1))
658
@deprecated_method(zero_eight)
659
def get(self, version_id):
660
"""Please use either Weave.get_text or Weave.get_lines as desired."""
661
return self.get_lines(version_id)
663
def get_lines(self, version_id):
664
"""See VersionedFile.get_lines()."""
665
return list(self._get_iter(version_id))
667
def get_sha1(self, name):
668
"""Get the stored sha1 sum for the given revision.
670
:param name: The name of the version to lookup
672
return self._sha1s[self._lookup(name)]
674
@deprecated_method(zero_eight)
675
def numversions(self):
676
"""How many versions are in this weave?
678
Deprecated in favour of num_versions.
680
return self.num_versions()
682
def num_versions(self):
683
"""How many versions are in this weave?"""
684
l = len(self._parents)
685
assert l == len(self._sha1s)
688
__len__ = num_versions
690
def check(self, progress_bar=None):
691
# TODO evaluate performance hit of using string sets in this routine.
692
# check no circular inclusions
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
for version_id in version_ids:
760
if not other.has_version(version_id) and not ignore_missing:
761
raise RevisionNotPresent(version_id, self._weave_name)
763
version_ids = other.versions()
765
# two loops so that we do not change ourselves before verifying it
767
# work through in index order to make sure we get all dependencies
770
# get the selected versions only that are in other.versions.
771
version_ids = set(other.versions()).intersection(set(version_ids))
772
# pull in the referenced graph.
773
version_ids = other.get_ancestry(version_ids)
774
pending_graph = [(version, other.get_parents(version)) for
775
version in version_ids]
776
for name in topo_sort(pending_graph):
777
other_idx = other._name_map[name]
778
self._check_version_consistent(other, other_idx, name)
779
sha1 = other._sha1s[other_idx]
783
if name in self._name_map:
784
idx = self._lookup(name)
785
n1 = set(map(other._idx_to_name, other._parents[other_idx]))
786
n2 = set(map(self._idx_to_name, self._parents[idx]))
787
if sha1 == self._sha1s[idx] and n1 == n2:
790
names_to_join.append((other_idx, name))
797
for other_idx, name in names_to_join:
798
# TODO: If all the parents of the other version are already
799
# present then we can avoid some work by just taking the delta
800
# and adjusting the offsets.
801
new_parents = self._imported_parents(other, other_idx)
802
sha1 = other._sha1s[other_idx]
807
pb.update(msg, merged, len(names_to_join))
809
lines = other.get_lines(other_idx)
810
self._add(name, lines, new_parents, sha1)
812
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
813
merged, processed, self._weave_name, time.time()-time0))
815
def _imported_parents(self, other, other_idx):
816
"""Return list of parents in self corresponding to indexes in other."""
818
for parent_idx in other._parents[other_idx]:
819
parent_name = other._names[parent_idx]
820
if parent_name not in self._names:
821
# should not be possible
822
raise WeaveError("missing parent {%s} of {%s} in %r"
823
% (parent_name, other._name_map[other_idx], self))
824
new_parents.append(self._name_map[parent_name])
827
def _check_version_consistent(self, other, other_idx, name):
828
"""Check if a version in consistent in this and other.
830
To be consistent it must have:
833
* the same direct parents (by name, not index, and disregarding
836
If present & correct return True;
837
if not present in self return False;
838
if inconsistent raise error."""
839
this_idx = self._name_map.get(name, -1)
841
if self._sha1s[this_idx] != other._sha1s[other_idx]:
842
raise errors.WeaveTextDiffers(name, self, other)
843
self_parents = self._parents[this_idx]
844
other_parents = other._parents[other_idx]
845
n1 = set([self._names[i] for i in self_parents])
846
n2 = set([other._names[i] for i in other_parents])
847
if not self._compatible_parents(n1, n2):
848
raise WeaveParentMismatch("inconsistent parents "
849
"for version {%s}: %s vs %s" % (name, n1, n2))
855
@deprecated_method(zero_eight)
856
def reweave(self, other, pb=None, msg=None):
857
"""reweave has been superceded by plain use of join."""
858
return self.join(other, pb, msg)
860
def _reweave(self, other, pb, msg):
861
"""Reweave self with other - internal helper for join().
863
:param other: The other weave to merge
864
:param pb: An optional progress bar, indicating how far done we are
865
:param msg: An optional message for the progress
867
new_weave = _reweave(self, other, pb=pb, msg=msg)
868
for attr in self.__slots__:
869
if attr != '_weave_name':
870
setattr(self, attr, getattr(new_weave, attr))
873
class WeaveFile(Weave):
874
"""A WeaveFile represents a Weave on disk and writes on change."""
876
WEAVE_SUFFIX = '.weave'
878
def __init__(self, name, transport, mode=None, create=False):
879
"""Create a WeaveFile.
881
:param create: If not True, only open an existing knit.
883
super(WeaveFile, self).__init__(name)
884
self._transport = transport
887
_read_weave_v5(self._transport.get(name + WeaveFile.WEAVE_SUFFIX), self)
888
except errors.NoSuchFile:
894
def add_lines(self, version_id, parents, lines):
895
"""Add a version and save the weave."""
896
super(WeaveFile, self).add_lines(version_id, parents, lines)
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(name + WeaveFile.WEAVE_SUFFIX, sio, self._mode)
907
def create_empty(self, name, transport, mode=None):
908
return WeaveFile(name, transport, mode, create=True)
911
"""Save the weave."""
913
write_weave_v5(self, sio)
915
self._transport.put(self._weave_name + WeaveFile.WEAVE_SUFFIX,
921
"""See VersionedFile.get_suffixes()."""
922
return [WeaveFile.WEAVE_SUFFIX]
924
def join(self, other, pb=None, msg=None, version_ids=None,
925
ignore_missing=False):
926
"""Join other into self and save."""
927
super(WeaveFile, self).join(other, pb, msg, version_ids, ignore_missing)
931
@deprecated_function(zero_eight)
932
def reweave(wa, wb, pb=None, msg=None):
933
"""reweaving is deprecation, please just use weave.join()."""
934
_reweave(wa, wb, pb, msg)
936
def _reweave(wa, wb, pb=None, msg=None):
937
"""Combine two weaves and return the result.
939
This works even if a revision R has different parents in
940
wa and wb. In the resulting weave all the parents are given.
942
This is done by just building up a new weave, maintaining ordering
943
of the versions in the two inputs. More efficient approaches
944
might be possible but it should only be necessary to do
945
this operation rarely, when a new previously ghost version is
948
:param pb: An optional progress bar, indicating how far done we are
949
:param msg: An optional message for the progress
953
queue_a = range(wa.num_versions())
954
queue_b = range(wb.num_versions())
955
# first determine combined parents of all versions
956
# map from version name -> all parent names
957
combined_parents = _reweave_parent_graphs(wa, wb)
958
mutter("combined parents: %r", combined_parents)
959
order = topo_sort(combined_parents.iteritems())
960
mutter("order to reweave: %r", order)
965
for idx, name in enumerate(order):
967
pb.update(msg, idx, len(order))
968
if name in wa._name_map:
969
lines = wa.get_lines(name)
970
if name in wb._name_map:
971
lines_b = wb.get_lines(name)
973
mutter('Weaves differ on content. rev_id {%s}', name)
974
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
976
lines = list(difflib.unified_diff(lines, lines_b,
977
wa._weave_name, wb._weave_name))
978
mutter('lines:\n%s', ''.join(lines))
979
raise errors.WeaveTextDiffers(name, wa, wb)
981
lines = wb.get_lines(name)
982
wr._add(name, lines, [wr._lookup(i) for i in combined_parents[name]])
985
def _reweave_parent_graphs(wa, wb):
986
"""Return combined parent ancestry for two weaves.
988
Returned as a list of (version_name, set(parent_names))"""
990
for weave in [wa, wb]:
991
for idx, name in enumerate(weave._names):
992
p = combined.setdefault(name, set())
993
p.update(map(weave._idx_to_name, weave._parents[idx]))
998
"""Show the weave's table-of-contents"""
999
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
1000
for i in (6, 50, 10, 10):
1003
for i in range(w.num_versions()):
1006
parent_str = ' '.join(map(str, w._parents[i]))
1007
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
1011
def weave_stats(weave_file, pb):
1012
from bzrlib.weavefile import read_weave
1014
wf = file(weave_file, 'rb')
1015
w = read_weave(wf, WeaveVersionedFile)
1016
# FIXME: doesn't work on pipes
1017
weave_size = wf.tell()
1021
for i in range(vers):
1022
pb.update('checking sizes', i, vers)
1023
for origin, lineno, line in w._extract([i]):
1028
print 'versions %9d' % vers
1029
print 'weave file %9d bytes' % weave_size
1030
print 'total contents %9d bytes' % total
1031
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1034
print 'average size %9d bytes' % avg
1035
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1039
print """bzr weave tool
1041
Experimental tool for weave algorithm.
1044
weave init WEAVEFILE
1045
Create an empty weave file
1046
weave get WEAVEFILE VERSION
1047
Write out specified version.
1048
weave check WEAVEFILE
1049
Check consistency of all versions.
1051
Display table of contents.
1052
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1053
Add NEWTEXT, with specified parent versions.
1054
weave annotate WEAVEFILE VERSION
1055
Display origin of each line.
1056
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1057
Auto-merge two versions and display conflicts.
1058
weave diff WEAVEFILE VERSION1 VERSION2
1059
Show differences between two versions.
1063
% weave init foo.weave
1065
% weave add foo.weave ver0 < foo.txt
1068
(create updated version)
1070
% weave get foo.weave 0 | diff -u - foo.txt
1071
% weave add foo.weave ver1 0 < foo.txt
1074
% weave get foo.weave 0 > foo.txt (create forked version)
1076
% weave add foo.weave ver2 0 < foo.txt
1079
% weave merge foo.weave 1 2 > foo.txt (merge them)
1080
% vi foo.txt (resolve conflicts)
1081
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1093
# in case we're run directly from the subdirectory
1094
sys.path.append('..')
1096
from bzrlib.weavefile import write_weave, read_weave
1097
from bzrlib.progress import ProgressBar
1112
return read_weave(file(argv[2], 'rb'))
1118
# at the moment, based on everything in the file
1120
parents = map(int, argv[4:])
1121
lines = sys.stdin.readlines()
1122
ver = w.add(name, parents, lines)
1123
write_weave(w, file(argv[2], 'wb'))
1124
print 'added version %r %d' % (name, ver)
1127
if os.path.exists(fn):
1128
raise IOError("file exists")
1130
write_weave(w, file(fn, 'wb'))
1131
elif cmd == 'get': # get one version
1133
sys.stdout.writelines(w.get_iter(int(argv[3])))
1136
from difflib import unified_diff
1139
v1, v2 = map(int, argv[3:5])
1142
diff_gen = unified_diff(lines1, lines2,
1143
'%s version %d' % (fn, v1),
1144
'%s version %d' % (fn, v2))
1145
sys.stdout.writelines(diff_gen)
1147
elif cmd == 'annotate':
1149
# newline is added to all lines regardless; too hard to get
1150
# reasonable formatting otherwise
1152
for origin, text in w.annotate(int(argv[3])):
1153
text = text.rstrip('\r\n')
1155
print ' | %s' % (text)
1157
print '%5d | %s' % (origin, text)
1163
elif cmd == 'stats':
1164
weave_stats(argv[2], ProgressBar())
1166
elif cmd == 'check':
1171
print '%d versions ok' % w.num_versions()
1173
elif cmd == 'inclusions':
1175
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1177
elif cmd == 'parents':
1179
print ' '.join(map(str, w._parents[int(argv[3])]))
1181
elif cmd == 'plan-merge':
1183
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1185
print '%14s | %s' % (state, line),
1187
elif cmd == 'merge':
1189
p = w.plan_merge(int(argv[3]), int(argv[4]))
1190
sys.stdout.writelines(w.weave_merge(p))
1193
raise ValueError('unknown command %r' % cmd)
1197
def profile_main(argv):
1198
import tempfile, hotshot, hotshot.stats
1200
prof_f = tempfile.NamedTemporaryFile()
1202
prof = hotshot.Profile(prof_f.name)
1204
ret = prof.runcall(main, argv)
1207
stats = hotshot.stats.load(prof_f.name)
1209
stats.sort_stats('cumulative')
1210
## XXX: Might like to write to stderr or the trace file instead but
1211
## print_stats seems hardcoded to stdout
1212
stats.print_stats(20)
1217
def lsprofile_main(argv):
1218
from bzrlib.lsprof import profile
1219
ret,stats = profile(main, argv)
1225
if __name__ == '__main__':
1227
if '--profile' in sys.argv:
1229
args.remove('--profile')
1230
sys.exit(profile_main(args))
1231
elif '--lsprof' in sys.argv:
1233
args.remove('--lsprof')
1234
sys.exit(lsprofile_main(args))
1236
sys.exit(main(sys.argv))
1239
class InterWeave(InterVersionedFile):
1240
"""Optimised code paths for weave to weave operations."""
1242
_matching_file_factory = staticmethod(WeaveFile)
1245
def is_compatible(source, target):
1246
"""Be compatible with weaves."""
1248
return (isinstance(source, Weave) and
1249
isinstance(target, Weave))
1250
except AttributeError:
1253
def join(self, pb=None, msg=None, version_ids=None, ignore_missing=False):
1254
"""See InterVersionedFile.join."""
1256
self.target._join(self.source, pb, msg, version_ids, ignore_missing)
1257
except errors.WeaveParentMismatch:
1258
self.target._reweave(self.source, pb, msg)
1261
InterVersionedFile.register_optimiser(InterWeave)
added
removed
bzrlib/weave.py
