22
22
"""Weave - storage of related text file versions"""
24
# before intset (r923) 2000 versions in 41.5s
25
# with intset (r926) 2000 versions in 93s !!!
26
# better to just use plain sets.
28
# making _extract build and return a list, rather than being a generator
31
# with python -O, r923 does 2000 versions in 36.87s
33
# with optimizations to avoid mutating lists - 35.75! I guess copying
34
# all the elements every time costs more than the small manipulations.
35
# a surprisingly small change.
37
# r931, which avoids using a generator for extract, does 36.98s
39
# with memoized inclusions, takes 41.49s; not very good
41
# with slots, takes 37.35s; without takes 39.16, a bit surprising
43
# with the delta calculation mixed in with the add method, rather than
44
# separated, takes 36.78s
46
# with delta folded in and mutation of the list, 36.13s
48
# with all this and simplification of add code, 33s
54
# TODO: Perhaps have copy method for Weave instances?
56
25
# XXX: If we do weaves this way, will a merge still behave the same
57
26
# way if it's done in a different order? That's a pretty desirable
75
44
# TODO: Parallel-extract that passes back each line along with a
76
45
# description of which revisions include it. Nice for checking all
46
# shas or calculating stats in parallel.
79
48
# TODO: Using a single _extract routine and then processing the output
80
49
# is probably inefficient. It's simple enough that we can afford to
81
50
# have slight specializations for different ways its used: annotate,
82
51
# basis for add, get, etc.
84
# TODO: Perhaps the API should work only in names to hide the integer
85
# indexes from the user?
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.
93
class WeaveError(Exception):
94
"""Exception in processing weave"""
97
class WeaveFormatError(WeaveError):
98
"""Weave invariant violated"""
72
from difflib import SequenceMatcher
74
from bzrlib.trace import mutter
75
from bzrlib.errors import (WeaveError, WeaveFormatError, WeaveParentMismatch,
76
WeaveRevisionNotPresent, WeaveRevisionAlreadyPresent)
77
import bzrlib.errors as errors
78
from bzrlib.tsort import topo_sort
101
81
class Weave(object):
102
82
"""weave - versioned text file storage.
204
206
def __ne__(self, other):
205
207
return not self.__eq__(other)
209
def __contains__(self, name):
210
return self._name_map.has_key(name)
212
def maybe_lookup(self, name_or_index):
213
"""Convert possible symbolic name to index, or pass through indexes."""
214
if isinstance(name_or_index, (int, long)):
217
return self.lookup(name_or_index)
208
220
def lookup(self, name):
221
"""Convert symbolic version name to index."""
210
223
return self._name_map[name]
212
raise WeaveError("name %s not present in weave" % name)
225
raise WeaveRevisionNotPresent(name, self)
228
return self._names[:]
230
def iter_names(self):
231
"""Yield a list of all names in this weave."""
232
return iter(self._names)
234
def idx_to_name(self, version):
235
return self._names[version]
237
def _check_repeated_add(self, name, parents, text, sha1):
238
"""Check that a duplicated add is OK.
240
If it is, return the (old) index; otherwise raise an exception.
242
idx = self.lookup(name)
243
if sorted(self._parents[idx]) != sorted(parents) \
244
or sha1 != self._sha1s[idx]:
245
raise WeaveRevisionAlreadyPresent(name, self)
215
def add(self, name, parents, text):
248
def add(self, name, parents, text, sha1=None):
216
249
"""Add a single text on top of the weave.
218
251
Returns the index number of the newly added version.
335
371
return new_version
373
def add_identical(self, old_rev_id, new_rev_id, parents):
374
"""Add an identical text to old_rev_id as new_rev_id."""
375
old_lines = self.get(self.lookup(old_rev_id))
376
self.add(new_rev_id, parents, old_lines)
338
378
def inclusions(self, versions):
339
379
"""Return set of all ancestors of given version(s)."""
340
380
i = set(versions)
345
# include all its parents
346
i.update(self._parents[v])
350
raise ValueError("version %d not present in weave" % v)
381
for v in xrange(max(versions), 0, -1):
383
# include all its parents
384
i.update(self._parents[v])
386
## except IndexError:
387
## raise ValueError("version %d not present in weave" % v)
390
def parents(self, version):
391
return self._parents[version]
394
def parent_names(self, version):
395
"""Return version names for parents of a version."""
396
return map(self.idx_to_name, self._parents[self.lookup(version)])
353
399
def minimal_parents(self, version):
490
544
result.append((istack[-1], lineno, l))
494
raise WFE("unclosed insertion blocks at end of weave",
547
raise WeaveFormatError("unclosed insertion blocks "
548
"at end of weave: %s" % istack)
497
raise WFE("unclosed deletion blocks at end of weave",
550
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
504
def get_iter(self, version):
555
def get_iter(self, name_or_index):
505
556
"""Yield lines for the specified version."""
506
for origin, lineno, line in self._extract([version]):
557
incls = [self.maybe_lookup(name_or_index)]
562
# We don't have sha1 sums for multiple entries
564
for origin, lineno, line in self._extract(incls):
510
def get(self, index):
511
return list(self.get_iter(index))
569
expected_sha1 = self._sha1s[index]
570
measured_sha1 = cur_sha.hexdigest()
571
if measured_sha1 != expected_sha1:
572
raise errors.WeaveInvalidChecksum(
573
'file %s, revision %s, expected: %s, measured %s'
574
% (self._weave_name, self._names[index],
575
expected_sha1, measured_sha1))
578
def get_text(self, name_or_index):
579
return ''.join(self.get_iter(name_or_index))
580
assert isinstance(version, int)
583
def get_lines(self, name_or_index):
584
return list(self.get_iter(name_or_index))
590
def get_sha1(self, name):
591
"""Get the stored sha1 sum for the given revision.
593
:param name: The name of the version to lookup
595
return self._sha1s[self.lookup(name)]
514
597
def mash_iter(self, included):
515
598
"""Return composed version of multiple included versions."""
599
included = map(self.maybe_lookup, included)
516
600
for origin, lineno, text in self._extract(included):
546
629
raise WeaveFormatError("invalid included version %d for index %d"
547
630
% (inclusions[-1], version))
549
# try extracting all versions; this is a bit slow and parallel
550
# extraction could be used
632
# try extracting all versions; parallel extraction is used
551
633
nv = self.numversions()
634
sha1s = [sha.new() for i in range(nv)]
635
texts = [[] for i in range(nv)]
638
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
639
# The problem is that set membership is much more expensive
641
for p in self._parents[i]:
642
new_inc.update(inclusions[p])
644
#assert set(new_inc) == self.inclusions([i]), 'failed %s != %s' % (new_inc, self.inclusions([i]))
645
inclusions.append(new_inc)
647
nlines = len(self._weave)
649
update_text = 'checking weave'
651
short_name = os.path.basename(self._weave_name)
652
update_text = 'checking %s' % (short_name,)
653
update_text = update_text[:25]
655
for lineno, insert, deleteset, line in self._walk():
657
progress_bar.update(update_text, lineno, nlines)
659
for j, j_inc in enumerate(inclusions):
660
# The active inclusion must be an ancestor,
661
# and no ancestors must have deleted this line,
662
# because we don't support resurrection.
663
if (insert in j_inc) and not (deleteset & j_inc):
664
sha1s[j].update(line)
552
666
for version in range(nv):
554
progress_bar.update('checking text', version, nv)
556
for l in self.get_iter(version):
667
hd = sha1s[version].hexdigest()
559
668
expected = self._sha1s[version]
560
669
if hd != expected:
561
raise WeaveError("mismatched sha1 for version %d; "
562
"got %s, expected %s"
563
% (version, hd, expected))
670
raise errors.WeaveInvalidChecksum(
671
"mismatched sha1 for version %s: "
672
"got %s, expected %s"
673
% (self._names[version], hd, expected))
565
675
# TODO: check insertions are properly nested, that there are
566
676
# no lines outside of insertion blocks, that deletions are
567
677
# properly paired, etc.
571
def merge(self, merge_versions):
572
"""Automerge and mark conflicts between versions.
574
This returns a sequence, each entry describing alternatives
575
for a chunk of the file. Each of the alternatives is given as
578
If there is a chunk of the file where there's no diagreement,
579
only one alternative is given.
582
# approach: find the included versions common to all the
584
raise NotImplementedError()
588
679
def _delta(self, included, lines):
589
680
"""Return changes from basis to new revision.
703
798
'killed-both'), \
802
def join(self, other, pb=None, msg=None):
804
"""Integrate versions from other into this weave.
806
The resulting weave contains all the history of both weaves;
807
any version you could retrieve from either self or other can be
808
retrieved from self after this call.
810
It is illegal for the two weaves to contain different values
811
or different parents for any version. See also reweave().
813
:param other: The other weave to pull into this one
814
:param pb: An optional progress bar
815
:param msg: An optional message to display for progress
817
if other.numversions() == 0:
818
return # nothing to update, easy
819
# two loops so that we do not change ourselves before verifying it
821
# work through in index order to make sure we get all dependencies
824
for other_idx, name in enumerate(other._names):
825
self._check_version_consistent(other, other_idx, name)
826
sha1 = other._sha1s[other_idx]
830
if name in self._name_map:
831
idx = self.lookup(name)
832
n1 = set(map(other.idx_to_name, other._parents[other_idx]))
833
n2 = set(map(self.idx_to_name, self._parents[idx]))
834
if sha1 == self._sha1s[idx] and n1 == n2:
837
names_to_join.append((other_idx, name))
844
for other_idx, name in names_to_join:
845
# TODO: If all the parents of the other version are already
846
# present then we can avoid some work by just taking the delta
847
# and adjusting the offsets.
848
new_parents = self._imported_parents(other, other_idx)
849
sha1 = other._sha1s[other_idx]
854
pb.update(msg, merged, len(names_to_join))
856
lines = other.get_lines(other_idx)
857
self.add(name, new_parents, lines, sha1)
859
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
860
merged, processed, self._weave_name, time.time( )-time0))
862
def _imported_parents(self, other, other_idx):
863
"""Return list of parents in self corresponding to indexes in other."""
865
for parent_idx in other._parents[other_idx]:
866
parent_name = other._names[parent_idx]
867
if parent_name not in self._names:
868
# should not be possible
869
raise WeaveError("missing parent {%s} of {%s} in %r"
870
% (parent_name, other._name_map[other_idx], self))
871
new_parents.append(self._name_map[parent_name])
874
def _check_version_consistent(self, other, other_idx, name):
875
"""Check if a version in consistent in this and other.
877
To be consistent it must have:
880
* the same direct parents (by name, not index, and disregarding
883
If present & correct return True;
884
if not present in self return False;
885
if inconsistent raise error."""
886
this_idx = self._name_map.get(name, -1)
888
if self._sha1s[this_idx] != other._sha1s[other_idx]:
889
raise WeaveError("inconsistent texts for version {%s} "
890
"when joining weaves"
892
self_parents = self._parents[this_idx]
893
other_parents = other._parents[other_idx]
894
n1 = set([self._names[i] for i in self_parents])
895
n2 = set([other._names[i] for i in other_parents])
897
raise WeaveParentMismatch("inconsistent parents "
898
"for version {%s}: %s vs %s" % (name, n1, n2))
904
def reweave(self, other, pb=None, msg=None):
905
"""Reweave self with other.
907
:param other: The other weave to merge
908
:param pb: An optional progress bar, indicating how far done we are
909
:param msg: An optional message for the progress
911
new_weave = reweave(self, other, pb=pb, msg=msg)
912
for attr in self.__slots__:
913
setattr(self, attr, getattr(new_weave, attr))
916
def reweave(wa, wb, pb=None, msg=None):
917
"""Combine two weaves and return the result.
919
This works even if a revision R has different parents in
920
wa and wb. In the resulting weave all the parents are given.
922
This is done by just building up a new weave, maintaining ordering
923
of the versions in the two inputs. More efficient approaches
924
might be possible but it should only be necessary to do
925
this operation rarely, when a new previously ghost version is
928
:param pb: An optional progress bar, indicating how far done we are
929
:param msg: An optional message for the progress
933
queue_a = range(wa.numversions())
934
queue_b = range(wb.numversions())
935
# first determine combined parents of all versions
936
# map from version name -> all parent names
937
combined_parents = _reweave_parent_graphs(wa, wb)
938
mutter("combined parents: %r", combined_parents)
939
order = topo_sort(combined_parents.iteritems())
940
mutter("order to reweave: %r", order)
945
for idx, name in enumerate(order):
947
pb.update(msg, idx, len(order))
948
if name in wa._name_map:
949
lines = wa.get_lines(name)
950
if name in wb._name_map:
951
lines_b = wb.get_lines(name)
953
mutter('Weaves differ on content. rev_id {%s}', name)
954
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
956
lines = list(difflib.unified_diff(lines, lines_b,
957
wa._weave_name, wb._weave_name))
958
mutter('lines:\n%s', ''.join(lines))
959
raise errors.WeaveTextDiffers(name, wa, wb)
961
lines = wb.get_lines(name)
962
wr.add(name, combined_parents[name], lines)
966
def _reweave_parent_graphs(wa, wb):
967
"""Return combined parent ancestry for two weaves.
969
Returned as a list of (version_name, set(parent_names))"""
971
for weave in [wa, wb]:
972
for idx, name in enumerate(weave._names):
973
p = combined.setdefault(name, set())
974
p.update(map(weave.idx_to_name, weave._parents[idx]))
712
978
def weave_toc(w):
added
removed
bzrlib/weave.py
