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"""
74
from bzrlib.trace import mutter
75
from bzrlib.errors import WeaveError, WeaveFormatError, WeaveParentMismatch, \
76
WeaveRevisionNotPresent, WeaveRevisionAlreadyPresent
77
from bzrlib.tsort import topo_sort
101
80
class Weave(object):
102
81
"""weave - versioned text file storage.
183
162
For each name, the version number.
165
Descriptive name of this weave; typically the filename if known.
186
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map']
169
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
170
'_weave_name', '_matcher']
172
def __init__(self, weave_name=None, matcher=None):
190
174
self._parents = []
193
177
self._name_map = {}
178
self._weave_name = weave_name
180
self._matcher = difflib.SequenceMatcher
182
self._matcher = matcher
186
"""Return a deep copy of self.
188
The copy can be modified without affecting the original weave."""
190
other._weave = self._weave[:]
191
other._parents = self._parents[:]
192
other._sha1s = self._sha1s[:]
193
other._names = self._names[:]
194
other._name_map = self._name_map.copy()
195
other._weave_name = self._weave_name
196
198
def __eq__(self, other):
197
199
if not isinstance(other, Weave):
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.
225
258
List or set of direct parent version numbers.
228
Sequence of lines to be added in the new version."""
261
Sequence of lines to be added in the new version.
263
sha -- SHA-1 of the file, if known. This is trusted to be
266
from bzrlib.osutils import sha_strings
230
268
assert isinstance(name, basestring)
270
sha1 = sha_strings(text)
231
271
if name in self._name_map:
232
raise WeaveError("name %r already present in weave" % name)
272
return self._check_repeated_add(name, parents, text, sha1)
274
parents = map(self.maybe_lookup, parents)
234
275
self._check_versions(parents)
235
276
## self._check_lines(text)
236
277
new_version = len(self._parents)
243
280
# if we abort after here the (in-memory) weave will be corrupt because only
244
281
# some fields are updated
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):
393
439
raise IndexError("invalid version number %r" % i)
396
def annotate(self, index):
397
return list(self.annotate_iter(index))
400
def annotate_iter(self, version):
442
def annotate(self, name_or_index):
443
return list(self.annotate_iter(name_or_index))
446
def annotate_iter(self, name_or_index):
401
447
"""Yield list of (index-id, line) pairs for the specified version.
403
449
The index indicates when the line originated in the weave."""
404
for origin, lineno, text in self._extract([version]):
450
incls = [self.maybe_lookup(name_or_index)]
451
for origin, lineno, text in self._extract(incls):
405
452
yield origin, text
409
455
"""Walk the weave.
490
539
result.append((istack[-1], lineno, l))
494
raise WFE("unclosed insertion blocks at end of weave",
542
raise WeaveFormatError("unclosed insertion blocks "
543
"at end of weave: %s" % istack)
497
raise WFE("unclosed deletion blocks at end of weave",
545
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
504
def get_iter(self, version):
550
def get_iter(self, name_or_index):
505
551
"""Yield lines for the specified version."""
506
for origin, lineno, line in self._extract([version]):
552
incls = [self.maybe_lookup(name_or_index)]
553
for origin, lineno, line in self._extract(incls):
510
def get(self, index):
511
return list(self.get_iter(index))
557
def get_text(self, name_or_index):
558
return ''.join(self.get_iter(name_or_index))
559
assert isinstance(version, int)
562
def get_lines(self, name_or_index):
563
return list(self.get_iter(name_or_index))
514
569
def mash_iter(self, included):
515
570
"""Return composed version of multiple included versions."""
571
included = map(self.maybe_lookup, included)
516
572
for origin, lineno, text in self._extract(included):
762
def join(self, other):
763
"""Integrate versions from other into this weave.
765
The resulting weave contains all the history of both weaves;
766
any version you could retrieve from either self or other can be
767
retrieved from self after this call.
769
It is illegal for the two weaves to contain different values
770
or different parents for any version. See also reweave().
772
if other.numversions() == 0:
773
return # nothing to update, easy
774
# two loops so that we do not change ourselves before verifying it
776
# work through in index order to make sure we get all dependencies
777
for other_idx, name in enumerate(other._names):
778
if self._check_version_consistent(other, other_idx, name):
780
for other_idx, name in enumerate(other._names):
781
# TODO: If all the parents of the other version are already
782
# present then we can avoid some work by just taking the delta
783
# and adjusting the offsets.
784
new_parents = self._imported_parents(other, other_idx)
785
lines = other.get_lines(other_idx)
786
sha1 = other._sha1s[other_idx]
787
self.add(name, new_parents, lines, sha1)
790
def _imported_parents(self, other, other_idx):
791
"""Return list of parents in self corresponding to indexes in other."""
793
for parent_idx in other._parents[other_idx]:
794
parent_name = other._names[parent_idx]
795
if parent_name not in self._names:
796
# should not be possible
797
raise WeaveError("missing parent {%s} of {%s} in %r"
798
% (parent_name, other._name_map[other_idx], self))
799
new_parents.append(self._name_map[parent_name])
802
def _check_version_consistent(self, other, other_idx, name):
803
"""Check if a version in consistent in this and other.
805
To be consistent it must have:
808
* the same direct parents (by name, not index, and disregarding
811
If present & correct return True;
812
if not present in self return False;
813
if inconsistent raise error."""
814
this_idx = self._name_map.get(name, -1)
816
if self._sha1s[this_idx] != other._sha1s[other_idx]:
817
raise WeaveError("inconsistent texts for version {%s} "
818
"when joining weaves"
820
self_parents = self._parents[this_idx]
821
other_parents = other._parents[other_idx]
822
n1 = [self._names[i] for i in self_parents]
823
n2 = [other._names[i] for i in other_parents]
827
raise WeaveParentMismatch("inconsistent parents "
828
"for version {%s}: %s vs %s" % (name, n1, n2))
834
def reweave(self, other):
835
"""Reweave self with other."""
836
new_weave = reweave(self, other)
837
for attr in self.__slots__:
838
setattr(self, attr, getattr(new_weave, attr))
842
"""Combine two weaves and return the result.
844
This works even if a revision R has different parents in
845
wa and wb. In the resulting weave all the parents are given.
847
This is done by just building up a new weave, maintaining ordering
848
of the versions in the two inputs. More efficient approaches
849
might be possible but it should only be necessary to do
850
this operation rarely, when a new previously ghost version is
855
queue_a = range(wa.numversions())
856
queue_b = range(wb.numversions())
857
# first determine combined parents of all versions
858
# map from version name -> all parent names
859
combined_parents = _reweave_parent_graphs(wa, wb)
860
mutter("combined parents: %r", combined_parents)
861
order = topo_sort(combined_parents.iteritems())
862
mutter("order to reweave: %r", order)
864
if name in wa._name_map:
865
lines = wa.get_lines(name)
866
if name in wb._name_map:
867
assert lines == wb.get_lines(name)
869
lines = wb.get_lines(name)
870
wr.add(name, combined_parents[name], lines)
874
def _reweave_parent_graphs(wa, wb):
875
"""Return combined parent ancestry for two weaves.
877
Returned as a list of (version_name, set(parent_names))"""
879
for weave in [wa, wb]:
880
for idx, name in enumerate(weave._names):
881
p = combined.setdefault(name, set())
882
p.update(map(weave.idx_to_name, weave._parents[idx]))
712
886
def weave_toc(w):
added
removed
bzrlib/weave.py
