1
# Copyright (C) 2007 Canonical Ltd
3
# This program is free software; you can redistribute it and/or modify
4
# it under the terms of the GNU General Public License as published by
5
# the Free Software Foundation; either version 2 of the License, or
6
# (at your option) any later version.
8
# This program is distributed in the hope that it will be useful,
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
# GNU General Public License for more details.
13
# You should have received a copy of the GNU General Public License
14
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
28
STEP_UNIQUE_SEARCHER_EVERY = 5
30
# DIAGRAM of terminology
40
# In this diagram, relative to G and H:
41
# A, B, C, D, E are common ancestors.
42
# C, D and E are border ancestors, because each has a non-common descendant.
43
# D and E are least common ancestors because none of their descendants are
45
# C is not a least common ancestor because its descendant, E, is a common
48
# The find_unique_lca algorithm will pick A in two steps:
49
# 1. find_lca('G', 'H') => ['D', 'E']
50
# 2. Since len(['D', 'E']) > 1, find_lca('D', 'E') => ['A']
53
class DictParentsProvider(object):
54
"""A parents provider for Graph objects."""
56
def __init__(self, ancestry):
57
self.ancestry = ancestry
60
return 'DictParentsProvider(%r)' % self.ancestry
62
def get_parent_map(self, keys):
63
"""See _StackedParentsProvider.get_parent_map"""
64
ancestry = self.ancestry
65
return dict((k, ancestry[k]) for k in keys if k in ancestry)
68
class _StackedParentsProvider(object):
70
def __init__(self, parent_providers):
71
self._parent_providers = parent_providers
74
return "_StackedParentsProvider(%r)" % self._parent_providers
76
def get_parent_map(self, keys):
77
"""Get a mapping of keys => parents
79
A dictionary is returned with an entry for each key present in this
80
source. If this source doesn't have information about a key, it should
83
[NULL_REVISION] is used as the parent of the first user-committed
84
revision. Its parent list is empty.
86
:param keys: An iterable returning keys to check (eg revision_ids)
87
:return: A dictionary mapping each key to its parents
91
for parents_provider in self._parent_providers:
92
new_found = parents_provider.get_parent_map(remaining)
93
found.update(new_found)
94
remaining.difference_update(new_found)
100
class CachingParentsProvider(object):
101
"""A parents provider which will cache the revision => parents as a dict.
103
This is useful for providers which have an expensive look up.
105
Either a ParentsProvider or a get_parent_map-like callback may be
106
supplied. If it provides extra un-asked-for parents, they will be cached,
107
but filtered out of get_parent_map.
109
The cache is enabled by default, but may be disabled and re-enabled.
111
def __init__(self, parent_provider=None, get_parent_map=None):
114
:param parent_provider: The ParentProvider to use. It or
115
get_parent_map must be supplied.
116
:param get_parent_map: The get_parent_map callback to use. It or
117
parent_provider must be supplied.
119
self._real_provider = parent_provider
120
if get_parent_map is None:
121
self._get_parent_map = self._real_provider.get_parent_map
123
self._get_parent_map = get_parent_map
125
self.enable_cache(True)
128
return "%s(%r)" % (self.__class__.__name__, self._real_provider)
130
def enable_cache(self, cache_misses=True):
132
if self._cache is not None:
133
raise AssertionError('Cache enabled when already enabled.')
135
self._cache_misses = cache_misses
136
self.missing_keys = set()
138
def disable_cache(self):
139
"""Disable and clear the cache."""
141
self._cache_misses = None
142
self.missing_keys = set()
144
def get_cached_map(self):
145
"""Return any cached get_parent_map values."""
146
if self._cache is None:
148
return dict(self._cache)
150
def get_parent_map(self, keys):
151
"""See _StackedParentsProvider.get_parent_map."""
154
cache = self._get_parent_map(keys)
156
needed_revisions = set(key for key in keys if key not in cache)
157
# Do not ask for negatively cached keys
158
needed_revisions.difference_update(self.missing_keys)
160
parent_map = self._get_parent_map(needed_revisions)
161
cache.update(parent_map)
162
if self._cache_misses:
163
for key in needed_revisions:
164
if key not in parent_map:
165
self.note_missing_key(key)
168
value = cache.get(key)
169
if value is not None:
173
def note_missing_key(self, key):
174
"""Note that key is a missing key."""
175
if self._cache_misses:
176
self.missing_keys.add(key)
180
"""Provide incremental access to revision graphs.
182
This is the generic implementation; it is intended to be subclassed to
183
specialize it for other repository types.
186
def __init__(self, parents_provider):
187
"""Construct a Graph that uses several graphs as its input
189
This should not normally be invoked directly, because there may be
190
specialized implementations for particular repository types. See
191
Repository.get_graph().
193
:param parents_provider: An object providing a get_parent_map call
194
conforming to the behavior of
195
StackedParentsProvider.get_parent_map.
197
if getattr(parents_provider, 'get_parents', None) is not None:
198
self.get_parents = parents_provider.get_parents
199
if getattr(parents_provider, 'get_parent_map', None) is not None:
200
self.get_parent_map = parents_provider.get_parent_map
201
self._parents_provider = parents_provider
204
return 'Graph(%r)' % self._parents_provider
206
def find_lca(self, *revisions):
207
"""Determine the lowest common ancestors of the provided revisions
209
A lowest common ancestor is a common ancestor none of whose
210
descendants are common ancestors. In graphs, unlike trees, there may
211
be multiple lowest common ancestors.
213
This algorithm has two phases. Phase 1 identifies border ancestors,
214
and phase 2 filters border ancestors to determine lowest common
217
In phase 1, border ancestors are identified, using a breadth-first
218
search starting at the bottom of the graph. Searches are stopped
219
whenever a node or one of its descendants is determined to be common
221
In phase 2, the border ancestors are filtered to find the least
222
common ancestors. This is done by searching the ancestries of each
225
Phase 2 is perfomed on the principle that a border ancestor that is
226
not an ancestor of any other border ancestor is a least common
229
Searches are stopped when they find a node that is determined to be a
230
common ancestor of all border ancestors, because this shows that it
231
cannot be a descendant of any border ancestor.
233
The scaling of this operation should be proportional to
234
1. The number of uncommon ancestors
235
2. The number of border ancestors
236
3. The length of the shortest path between a border ancestor and an
237
ancestor of all border ancestors.
239
border_common, common, sides = self._find_border_ancestors(revisions)
240
# We may have common ancestors that can be reached from each other.
241
# - ask for the heads of them to filter it down to only ones that
242
# cannot be reached from each other - phase 2.
243
return self.heads(border_common)
245
def find_difference(self, left_revision, right_revision):
246
"""Determine the graph difference between two revisions"""
247
border, common, searchers = self._find_border_ancestors(
248
[left_revision, right_revision])
249
self._search_for_extra_common(common, searchers)
250
left = searchers[0].seen
251
right = searchers[1].seen
252
return (left.difference(right), right.difference(left))
254
def find_distance_to_null(self, target_revision_id, known_revision_ids):
255
"""Find the left-hand distance to the NULL_REVISION.
257
(This can also be considered the revno of a branch at
260
:param target_revision_id: A revision_id which we would like to know
262
:param known_revision_ids: [(revision_id, revno)] A list of known
263
revno, revision_id tuples. We'll use this to seed the search.
265
# Map from revision_ids to a known value for their revno
266
known_revnos = dict(known_revision_ids)
267
cur_tip = target_revision_id
269
NULL_REVISION = revision.NULL_REVISION
270
known_revnos[NULL_REVISION] = 0
272
searching_known_tips = list(known_revnos.keys())
274
unknown_searched = {}
276
while cur_tip not in known_revnos:
277
unknown_searched[cur_tip] = num_steps
279
to_search = set([cur_tip])
280
to_search.update(searching_known_tips)
281
parent_map = self.get_parent_map(to_search)
282
parents = parent_map.get(cur_tip, None)
283
if not parents: # An empty list or None is a ghost
284
raise errors.GhostRevisionsHaveNoRevno(target_revision_id,
288
for revision_id in searching_known_tips:
289
parents = parent_map.get(revision_id, None)
293
next_revno = known_revnos[revision_id] - 1
294
if next in unknown_searched:
295
# We have enough information to return a value right now
296
return next_revno + unknown_searched[next]
297
if next in known_revnos:
299
known_revnos[next] = next_revno
300
next_known_tips.append(next)
301
searching_known_tips = next_known_tips
303
# We reached a known revision, so just add in how many steps it took to
305
return known_revnos[cur_tip] + num_steps
307
def find_unique_ancestors(self, unique_revision, common_revisions):
308
"""Find the unique ancestors for a revision versus others.
310
This returns the ancestry of unique_revision, excluding all revisions
311
in the ancestry of common_revisions. If unique_revision is in the
312
ancestry, then the empty set will be returned.
314
:param unique_revision: The revision_id whose ancestry we are
316
XXX: Would this API be better if we allowed multiple revisions on
318
:param common_revisions: Revision_ids of ancestries to exclude.
319
:return: A set of revisions in the ancestry of unique_revision
321
if unique_revision in common_revisions:
324
# Algorithm description
325
# 1) Walk backwards from the unique node and all common nodes.
326
# 2) When a node is seen by both sides, stop searching it in the unique
327
# walker, include it in the common walker.
328
# 3) Stop searching when there are no nodes left for the unique walker.
329
# At this point, you have a maximal set of unique nodes. Some of
330
# them may actually be common, and you haven't reached them yet.
331
# 4) Start new searchers for the unique nodes, seeded with the
332
# information you have so far.
333
# 5) Continue searching, stopping the common searches when the search
334
# tip is an ancestor of all unique nodes.
335
# 6) Aggregate together unique searchers when they are searching the
336
# same tips. When all unique searchers are searching the same node,
337
# stop move it to a single 'all_unique_searcher'.
338
# 7) The 'all_unique_searcher' represents the very 'tip' of searching.
339
# Most of the time this produces very little important information.
340
# So don't step it as quickly as the other searchers.
341
# 8) Search is done when all common searchers have completed.
343
unique_searcher, common_searcher = self._find_initial_unique_nodes(
344
[unique_revision], common_revisions)
346
unique_nodes = unique_searcher.seen.difference(common_searcher.seen)
350
(all_unique_searcher,
351
unique_tip_searchers) = self._make_unique_searchers(unique_nodes,
352
unique_searcher, common_searcher)
354
self._refine_unique_nodes(unique_searcher, all_unique_searcher,
355
unique_tip_searchers, common_searcher)
356
true_unique_nodes = unique_nodes.difference(common_searcher.seen)
357
if 'graph' in debug.debug_flags:
358
trace.mutter('Found %d truly unique nodes out of %d',
359
len(true_unique_nodes), len(unique_nodes))
360
return true_unique_nodes
362
def _find_initial_unique_nodes(self, unique_revisions, common_revisions):
363
"""Steps 1-3 of find_unique_ancestors.
365
Find the maximal set of unique nodes. Some of these might actually
366
still be common, but we are sure that there are no other unique nodes.
368
:return: (unique_searcher, common_searcher)
371
unique_searcher = self._make_breadth_first_searcher(unique_revisions)
372
# we know that unique_revisions aren't in common_revisions, so skip
374
unique_searcher.next()
375
common_searcher = self._make_breadth_first_searcher(common_revisions)
377
# As long as we are still finding unique nodes, keep searching
378
while unique_searcher._next_query:
379
next_unique_nodes = set(unique_searcher.step())
380
next_common_nodes = set(common_searcher.step())
382
# Check if either searcher encounters new nodes seen by the other
384
unique_are_common_nodes = next_unique_nodes.intersection(
385
common_searcher.seen)
386
unique_are_common_nodes.update(
387
next_common_nodes.intersection(unique_searcher.seen))
388
if unique_are_common_nodes:
389
ancestors = unique_searcher.find_seen_ancestors(
390
unique_are_common_nodes)
391
# TODO: This is a bit overboard, we only really care about
392
# the ancestors of the tips because the rest we
393
# already know. This is *correct* but causes us to
394
# search too much ancestry.
395
ancestors.update(common_searcher.find_seen_ancestors(ancestors))
396
unique_searcher.stop_searching_any(ancestors)
397
common_searcher.start_searching(ancestors)
399
return unique_searcher, common_searcher
401
def _make_unique_searchers(self, unique_nodes, unique_searcher,
403
"""Create a searcher for all the unique search tips (step 4).
405
As a side effect, the common_searcher will stop searching any nodes
406
that are ancestors of the unique searcher tips.
408
:return: (all_unique_searcher, unique_tip_searchers)
410
unique_tips = self._remove_simple_descendants(unique_nodes,
411
self.get_parent_map(unique_nodes))
413
if len(unique_tips) == 1:
414
unique_tip_searchers = []
415
ancestor_all_unique = unique_searcher.find_seen_ancestors(unique_tips)
417
unique_tip_searchers = []
418
for tip in unique_tips:
419
revs_to_search = unique_searcher.find_seen_ancestors([tip])
420
revs_to_search.update(
421
common_searcher.find_seen_ancestors(revs_to_search))
422
searcher = self._make_breadth_first_searcher(revs_to_search)
423
# We don't care about the starting nodes.
424
searcher._label = tip
426
unique_tip_searchers.append(searcher)
428
ancestor_all_unique = None
429
for searcher in unique_tip_searchers:
430
if ancestor_all_unique is None:
431
ancestor_all_unique = set(searcher.seen)
433
ancestor_all_unique = ancestor_all_unique.intersection(
435
# Collapse all the common nodes into a single searcher
436
all_unique_searcher = self._make_breadth_first_searcher(
438
if ancestor_all_unique:
439
# We've seen these nodes in all the searchers, so we'll just go to
441
all_unique_searcher.step()
443
# Stop any search tips that are already known as ancestors of the
445
stopped_common = common_searcher.stop_searching_any(
446
common_searcher.find_seen_ancestors(ancestor_all_unique))
449
for searcher in unique_tip_searchers:
450
total_stopped += len(searcher.stop_searching_any(
451
searcher.find_seen_ancestors(ancestor_all_unique)))
452
if 'graph' in debug.debug_flags:
453
trace.mutter('For %d unique nodes, created %d + 1 unique searchers'
454
' (%d stopped search tips, %d common ancestors'
455
' (%d stopped common)',
456
len(unique_nodes), len(unique_tip_searchers),
457
total_stopped, len(ancestor_all_unique),
459
return all_unique_searcher, unique_tip_searchers
461
def _step_unique_and_common_searchers(self, common_searcher,
462
unique_tip_searchers,
464
"""Step all the searchers"""
465
newly_seen_common = set(common_searcher.step())
466
newly_seen_unique = set()
467
for searcher in unique_tip_searchers:
468
next = set(searcher.step())
469
next.update(unique_searcher.find_seen_ancestors(next))
470
next.update(common_searcher.find_seen_ancestors(next))
471
for alt_searcher in unique_tip_searchers:
472
if alt_searcher is searcher:
474
next.update(alt_searcher.find_seen_ancestors(next))
475
searcher.start_searching(next)
476
newly_seen_unique.update(next)
477
return newly_seen_common, newly_seen_unique
479
def _find_nodes_common_to_all_unique(self, unique_tip_searchers,
481
newly_seen_unique, step_all_unique):
482
"""Find nodes that are common to all unique_tip_searchers.
484
If it is time, step the all_unique_searcher, and add its nodes to the
487
common_to_all_unique_nodes = newly_seen_unique.copy()
488
for searcher in unique_tip_searchers:
489
common_to_all_unique_nodes.intersection_update(searcher.seen)
490
common_to_all_unique_nodes.intersection_update(
491
all_unique_searcher.seen)
492
# Step all-unique less frequently than the other searchers.
493
# In the common case, we don't need to spider out far here, so
494
# avoid doing extra work.
496
tstart = time.clock()
497
nodes = all_unique_searcher.step()
498
common_to_all_unique_nodes.update(nodes)
499
if 'graph' in debug.debug_flags:
500
tdelta = time.clock() - tstart
501
trace.mutter('all_unique_searcher step() took %.3fs'
502
'for %d nodes (%d total), iteration: %s',
503
tdelta, len(nodes), len(all_unique_searcher.seen),
504
all_unique_searcher._iterations)
505
return common_to_all_unique_nodes
507
def _collapse_unique_searchers(self, unique_tip_searchers,
508
common_to_all_unique_nodes):
509
"""Combine searchers that are searching the same tips.
511
When two searchers are searching the same tips, we can stop one of the
512
searchers. We also know that the maximal set of common ancestors is the
513
intersection of the two original searchers.
515
:return: A list of searchers that are searching unique nodes.
517
# Filter out searchers that don't actually search different
518
# nodes. We already have the ancestry intersection for them
519
unique_search_tips = {}
520
for searcher in unique_tip_searchers:
521
stopped = searcher.stop_searching_any(common_to_all_unique_nodes)
522
will_search_set = frozenset(searcher._next_query)
523
if not will_search_set:
524
if 'graph' in debug.debug_flags:
525
trace.mutter('Unique searcher %s was stopped.'
526
' (%s iterations) %d nodes stopped',
528
searcher._iterations,
530
elif will_search_set not in unique_search_tips:
531
# This searcher is searching a unique set of nodes, let it
532
unique_search_tips[will_search_set] = [searcher]
534
unique_search_tips[will_search_set].append(searcher)
535
# TODO: it might be possible to collapse searchers faster when they
536
# only have *some* search tips in common.
537
next_unique_searchers = []
538
for searchers in unique_search_tips.itervalues():
539
if len(searchers) == 1:
540
# Searching unique tips, go for it
541
next_unique_searchers.append(searchers[0])
543
# These searchers have started searching the same tips, we
544
# don't need them to cover the same ground. The
545
# intersection of their ancestry won't change, so create a
546
# new searcher, combining their histories.
547
next_searcher = searchers[0]
548
for searcher in searchers[1:]:
549
next_searcher.seen.intersection_update(searcher.seen)
550
if 'graph' in debug.debug_flags:
551
trace.mutter('Combining %d searchers into a single'
552
' searcher searching %d nodes with'
555
len(next_searcher._next_query),
556
len(next_searcher.seen))
557
next_unique_searchers.append(next_searcher)
558
return next_unique_searchers
560
def _refine_unique_nodes(self, unique_searcher, all_unique_searcher,
561
unique_tip_searchers, common_searcher):
562
"""Steps 5-8 of find_unique_ancestors.
564
This function returns when common_searcher has stopped searching for
567
# We step the ancestor_all_unique searcher only every
568
# STEP_UNIQUE_SEARCHER_EVERY steps.
569
step_all_unique_counter = 0
570
# While we still have common nodes to search
571
while common_searcher._next_query:
573
newly_seen_unique) = self._step_unique_and_common_searchers(
574
common_searcher, unique_tip_searchers, unique_searcher)
575
# These nodes are common ancestors of all unique nodes
576
common_to_all_unique_nodes = self._find_nodes_common_to_all_unique(
577
unique_tip_searchers, all_unique_searcher, newly_seen_unique,
578
step_all_unique_counter==0)
579
step_all_unique_counter = ((step_all_unique_counter + 1)
580
% STEP_UNIQUE_SEARCHER_EVERY)
582
if newly_seen_common:
583
# If a 'common' node is an ancestor of all unique searchers, we
584
# can stop searching it.
585
common_searcher.stop_searching_any(
586
all_unique_searcher.seen.intersection(newly_seen_common))
587
if common_to_all_unique_nodes:
588
common_to_all_unique_nodes.update(
589
common_searcher.find_seen_ancestors(
590
common_to_all_unique_nodes))
591
# The all_unique searcher can start searching the common nodes
592
# but everyone else can stop.
593
# This is the sort of thing where we would like to not have it
594
# start_searching all of the nodes, but only mark all of them
595
# as seen, and have it search only the actual tips. Otherwise
596
# it is another get_parent_map() traversal for it to figure out
597
# what we already should know.
598
all_unique_searcher.start_searching(common_to_all_unique_nodes)
599
common_searcher.stop_searching_any(common_to_all_unique_nodes)
601
next_unique_searchers = self._collapse_unique_searchers(
602
unique_tip_searchers, common_to_all_unique_nodes)
603
if len(unique_tip_searchers) != len(next_unique_searchers):
604
if 'graph' in debug.debug_flags:
605
trace.mutter('Collapsed %d unique searchers => %d'
607
len(unique_tip_searchers),
608
len(next_unique_searchers),
609
all_unique_searcher._iterations)
610
unique_tip_searchers = next_unique_searchers
612
def get_parent_map(self, revisions):
613
"""Get a map of key:parent_list for revisions.
615
This implementation delegates to get_parents, for old parent_providers
616
that do not supply get_parent_map.
619
for rev, parents in self.get_parents(revisions):
620
if parents is not None:
621
result[rev] = parents
624
def _make_breadth_first_searcher(self, revisions):
625
return _BreadthFirstSearcher(revisions, self)
627
def _find_border_ancestors(self, revisions):
628
"""Find common ancestors with at least one uncommon descendant.
630
Border ancestors are identified using a breadth-first
631
search starting at the bottom of the graph. Searches are stopped
632
whenever a node or one of its descendants is determined to be common.
634
This will scale with the number of uncommon ancestors.
636
As well as the border ancestors, a set of seen common ancestors and a
637
list of sets of seen ancestors for each input revision is returned.
638
This allows calculation of graph difference from the results of this
641
if None in revisions:
642
raise errors.InvalidRevisionId(None, self)
643
common_ancestors = set()
644
searchers = [self._make_breadth_first_searcher([r])
646
active_searchers = searchers[:]
647
border_ancestors = set()
651
for searcher in searchers:
652
new_ancestors = searcher.step()
654
newly_seen.update(new_ancestors)
656
for revision in newly_seen:
657
if revision in common_ancestors:
658
# Not a border ancestor because it was seen as common
660
new_common.add(revision)
662
for searcher in searchers:
663
if revision not in searcher.seen:
666
# This is a border because it is a first common that we see
667
# after walking for a while.
668
border_ancestors.add(revision)
669
new_common.add(revision)
671
for searcher in searchers:
672
new_common.update(searcher.find_seen_ancestors(new_common))
673
for searcher in searchers:
674
searcher.start_searching(new_common)
675
common_ancestors.update(new_common)
677
# Figure out what the searchers will be searching next, and if
678
# there is only 1 set being searched, then we are done searching,
679
# since all searchers would have to be searching the same data,
680
# thus it *must* be in common.
681
unique_search_sets = set()
682
for searcher in searchers:
683
will_search_set = frozenset(searcher._next_query)
684
if will_search_set not in unique_search_sets:
685
# This searcher is searching a unique set of nodes, let it
686
unique_search_sets.add(will_search_set)
688
if len(unique_search_sets) == 1:
689
nodes = unique_search_sets.pop()
690
uncommon_nodes = nodes.difference(common_ancestors)
692
raise AssertionError("Somehow we ended up converging"
693
" without actually marking them as"
696
"\nuncommon_nodes: %s"
697
% (revisions, uncommon_nodes))
699
return border_ancestors, common_ancestors, searchers
701
def heads(self, keys):
702
"""Return the heads from amongst keys.
704
This is done by searching the ancestries of each key. Any key that is
705
reachable from another key is not returned; all the others are.
707
This operation scales with the relative depth between any two keys. If
708
any two keys are completely disconnected all ancestry of both sides
711
:param keys: An iterable of keys.
712
:return: A set of the heads. Note that as a set there is no ordering
713
information. Callers will need to filter their input to create
714
order if they need it.
716
candidate_heads = set(keys)
717
if revision.NULL_REVISION in candidate_heads:
718
# NULL_REVISION is only a head if it is the only entry
719
candidate_heads.remove(revision.NULL_REVISION)
720
if not candidate_heads:
721
return set([revision.NULL_REVISION])
722
if len(candidate_heads) < 2:
723
return candidate_heads
724
searchers = dict((c, self._make_breadth_first_searcher([c]))
725
for c in candidate_heads)
726
active_searchers = dict(searchers)
727
# skip over the actual candidate for each searcher
728
for searcher in active_searchers.itervalues():
730
# The common walker finds nodes that are common to two or more of the
731
# input keys, so that we don't access all history when a currently
732
# uncommon search point actually meets up with something behind a
733
# common search point. Common search points do not keep searches
734
# active; they just allow us to make searches inactive without
735
# accessing all history.
736
common_walker = self._make_breadth_first_searcher([])
737
while len(active_searchers) > 0:
742
except StopIteration:
743
# No common points being searched at this time.
745
for candidate in active_searchers.keys():
747
searcher = active_searchers[candidate]
749
# rare case: we deleted candidate in a previous iteration
750
# through this for loop, because it was determined to be
751
# a descendant of another candidate.
754
ancestors.update(searcher.next())
755
except StopIteration:
756
del active_searchers[candidate]
758
# process found nodes
760
for ancestor in ancestors:
761
if ancestor in candidate_heads:
762
candidate_heads.remove(ancestor)
763
del searchers[ancestor]
764
if ancestor in active_searchers:
765
del active_searchers[ancestor]
766
# it may meet up with a known common node
767
if ancestor in common_walker.seen:
768
# some searcher has encountered our known common nodes:
770
ancestor_set = set([ancestor])
771
for searcher in searchers.itervalues():
772
searcher.stop_searching_any(ancestor_set)
774
# or it may have been just reached by all the searchers:
775
for searcher in searchers.itervalues():
776
if ancestor not in searcher.seen:
779
# The final active searcher has just reached this node,
780
# making it be known as a descendant of all candidates,
781
# so we can stop searching it, and any seen ancestors
782
new_common.add(ancestor)
783
for searcher in searchers.itervalues():
785
searcher.find_seen_ancestors([ancestor])
786
searcher.stop_searching_any(seen_ancestors)
787
common_walker.start_searching(new_common)
788
return candidate_heads
790
def find_merge_order(self, tip_revision_id, lca_revision_ids):
791
"""Find the order that each revision was merged into tip.
793
This basically just walks backwards with a stack, and walks left-first
794
until it finds a node to stop.
796
if len(lca_revision_ids) == 1:
797
return list(lca_revision_ids)
798
looking_for = set(lca_revision_ids)
799
# TODO: Is there a way we could do this "faster" by batching up the
800
# get_parent_map requests?
801
# TODO: Should we also be culling the ancestry search right away? We
802
# could add looking_for to the "stop" list, and walk their
803
# ancestry in batched mode. The flip side is it might mean we walk a
804
# lot of "stop" nodes, rather than only the minimum.
805
# Then again, without it we may trace back into ancestry we could have
807
stack = [tip_revision_id]
810
while stack and looking_for:
813
if next in looking_for:
815
looking_for.remove(next)
816
if len(looking_for) == 1:
817
found.append(looking_for.pop())
820
parent_ids = self.get_parent_map([next]).get(next, None)
821
if not parent_ids: # Ghost, nothing to search here
823
for parent_id in reversed(parent_ids):
824
# TODO: (performance) We see the parent at this point, but we
825
# wait to mark it until later to make sure we get left
826
# parents before right parents. However, instead of
827
# waiting until we have traversed enough parents, we
828
# could instead note that we've found it, and once all
829
# parents are in the stack, just reverse iterate the
831
if parent_id not in stop:
832
# this will need to be searched
833
stack.append(parent_id)
837
def find_unique_lca(self, left_revision, right_revision,
839
"""Find a unique LCA.
841
Find lowest common ancestors. If there is no unique common
842
ancestor, find the lowest common ancestors of those ancestors.
844
Iteration stops when a unique lowest common ancestor is found.
845
The graph origin is necessarily a unique lowest common ancestor.
847
Note that None is not an acceptable substitute for NULL_REVISION.
848
in the input for this method.
850
:param count_steps: If True, the return value will be a tuple of
851
(unique_lca, steps) where steps is the number of times that
852
find_lca was run. If False, only unique_lca is returned.
854
revisions = [left_revision, right_revision]
858
lca = self.find_lca(*revisions)
866
raise errors.NoCommonAncestor(left_revision, right_revision)
869
def iter_ancestry(self, revision_ids):
870
"""Iterate the ancestry of this revision.
872
:param revision_ids: Nodes to start the search
873
:return: Yield tuples mapping a revision_id to its parents for the
874
ancestry of revision_id.
875
Ghosts will be returned with None as their parents, and nodes
876
with no parents will have NULL_REVISION as their only parent. (As
877
defined by get_parent_map.)
878
There will also be a node for (NULL_REVISION, ())
880
pending = set(revision_ids)
883
processed.update(pending)
884
next_map = self.get_parent_map(pending)
886
for item in next_map.iteritems():
888
next_pending.update(p for p in item[1] if p not in processed)
889
ghosts = pending.difference(next_map)
892
pending = next_pending
894
def iter_topo_order(self, revisions):
895
"""Iterate through the input revisions in topological order.
897
This sorting only ensures that parents come before their children.
898
An ancestor may sort after a descendant if the relationship is not
899
visible in the supplied list of revisions.
901
sorter = tsort.TopoSorter(self.get_parent_map(revisions))
902
return sorter.iter_topo_order()
904
def is_ancestor(self, candidate_ancestor, candidate_descendant):
905
"""Determine whether a revision is an ancestor of another.
907
We answer this using heads() as heads() has the logic to perform the
908
smallest number of parent lookups to determine the ancestral
909
relationship between N revisions.
911
return set([candidate_descendant]) == self.heads(
912
[candidate_ancestor, candidate_descendant])
914
def is_between(self, revid, lower_bound_revid, upper_bound_revid):
915
"""Determine whether a revision is between two others.
917
returns true if and only if:
918
lower_bound_revid <= revid <= upper_bound_revid
920
return ((upper_bound_revid is None or
921
self.is_ancestor(revid, upper_bound_revid)) and
922
(lower_bound_revid is None or
923
self.is_ancestor(lower_bound_revid, revid)))
925
def _search_for_extra_common(self, common, searchers):
926
"""Make sure that unique nodes are genuinely unique.
928
After _find_border_ancestors, all nodes marked "common" are indeed
929
common. Some of the nodes considered unique are not, due to history
930
shortcuts stopping the searches early.
932
We know that we have searched enough when all common search tips are
933
descended from all unique (uncommon) nodes because we know that a node
934
cannot be an ancestor of its own ancestor.
936
:param common: A set of common nodes
937
:param searchers: The searchers returned from _find_border_ancestors
941
# A) The passed in searchers should all be on the same tips, thus
942
# they should be considered the "common" searchers.
943
# B) We find the difference between the searchers, these are the
944
# "unique" nodes for each side.
945
# C) We do a quick culling so that we only start searching from the
946
# more interesting unique nodes. (A unique ancestor is more
947
# interesting than any of its children.)
948
# D) We start searching for ancestors common to all unique nodes.
949
# E) We have the common searchers stop searching any ancestors of
951
# F) When there are no more common search tips, we stop
953
# TODO: We need a way to remove unique_searchers when they overlap with
954
# other unique searchers.
955
if len(searchers) != 2:
956
raise NotImplementedError(
957
"Algorithm not yet implemented for > 2 searchers")
958
common_searchers = searchers
959
left_searcher = searchers[0]
960
right_searcher = searchers[1]
961
unique = left_searcher.seen.symmetric_difference(right_searcher.seen)
962
if not unique: # No unique nodes, nothing to do
964
total_unique = len(unique)
965
unique = self._remove_simple_descendants(unique,
966
self.get_parent_map(unique))
967
simple_unique = len(unique)
969
unique_searchers = []
970
for revision_id in unique:
971
if revision_id in left_searcher.seen:
972
parent_searcher = left_searcher
974
parent_searcher = right_searcher
975
revs_to_search = parent_searcher.find_seen_ancestors([revision_id])
976
if not revs_to_search: # XXX: This shouldn't be possible
977
revs_to_search = [revision_id]
978
searcher = self._make_breadth_first_searcher(revs_to_search)
979
# We don't care about the starting nodes.
981
unique_searchers.append(searcher)
983
# possible todo: aggregate the common searchers into a single common
984
# searcher, just make sure that we include the nodes into the .seen
985
# properties of the original searchers
987
ancestor_all_unique = None
988
for searcher in unique_searchers:
989
if ancestor_all_unique is None:
990
ancestor_all_unique = set(searcher.seen)
992
ancestor_all_unique = ancestor_all_unique.intersection(
995
trace.mutter('Started %s unique searchers for %s unique revisions',
996
simple_unique, total_unique)
998
while True: # If we have no more nodes we have nothing to do
999
newly_seen_common = set()
1000
for searcher in common_searchers:
1001
newly_seen_common.update(searcher.step())
1002
newly_seen_unique = set()
1003
for searcher in unique_searchers:
1004
newly_seen_unique.update(searcher.step())
1005
new_common_unique = set()
1006
for revision in newly_seen_unique:
1007
for searcher in unique_searchers:
1008
if revision not in searcher.seen:
1011
# This is a border because it is a first common that we see
1012
# after walking for a while.
1013
new_common_unique.add(revision)
1014
if newly_seen_common:
1015
# These are nodes descended from one of the 'common' searchers.
1016
# Make sure all searchers are on the same page
1017
for searcher in common_searchers:
1018
newly_seen_common.update(
1019
searcher.find_seen_ancestors(newly_seen_common))
1020
# We start searching the whole ancestry. It is a bit wasteful,
1021
# though. We really just want to mark all of these nodes as
1022
# 'seen' and then start just the tips. However, it requires a
1023
# get_parent_map() call to figure out the tips anyway, and all
1024
# redundant requests should be fairly fast.
1025
for searcher in common_searchers:
1026
searcher.start_searching(newly_seen_common)
1028
# If a 'common' node is an ancestor of all unique searchers, we
1029
# can stop searching it.
1030
stop_searching_common = ancestor_all_unique.intersection(
1032
if stop_searching_common:
1033
for searcher in common_searchers:
1034
searcher.stop_searching_any(stop_searching_common)
1035
if new_common_unique:
1036
# We found some ancestors that are common
1037
for searcher in unique_searchers:
1038
new_common_unique.update(
1039
searcher.find_seen_ancestors(new_common_unique))
1040
# Since these are common, we can grab another set of ancestors
1042
for searcher in common_searchers:
1043
new_common_unique.update(
1044
searcher.find_seen_ancestors(new_common_unique))
1046
# We can tell all of the unique searchers to start at these
1047
# nodes, and tell all of the common searchers to *stop*
1048
# searching these nodes
1049
for searcher in unique_searchers:
1050
searcher.start_searching(new_common_unique)
1051
for searcher in common_searchers:
1052
searcher.stop_searching_any(new_common_unique)
1053
ancestor_all_unique.update(new_common_unique)
1055
# Filter out searchers that don't actually search different
1056
# nodes. We already have the ancestry intersection for them
1057
next_unique_searchers = []
1058
unique_search_sets = set()
1059
for searcher in unique_searchers:
1060
will_search_set = frozenset(searcher._next_query)
1061
if will_search_set not in unique_search_sets:
1062
# This searcher is searching a unique set of nodes, let it
1063
unique_search_sets.add(will_search_set)
1064
next_unique_searchers.append(searcher)
1065
unique_searchers = next_unique_searchers
1066
for searcher in common_searchers:
1067
if searcher._next_query:
1070
# All common searcher have stopped searching
1073
def _remove_simple_descendants(self, revisions, parent_map):
1074
"""remove revisions which are children of other ones in the set
1076
This doesn't do any graph searching, it just checks the immediate
1077
parent_map to find if there are any children which can be removed.
1079
:param revisions: A set of revision_ids
1080
:return: A set of revision_ids with the children removed
1082
simple_ancestors = revisions.copy()
1083
# TODO: jam 20071214 we *could* restrict it to searching only the
1084
# parent_map of revisions already present in 'revisions', but
1085
# considering the general use case, I think this is actually
1088
# This is the same as the following loop. I don't know that it is any
1090
## simple_ancestors.difference_update(r for r, p_ids in parent_map.iteritems()
1091
## if p_ids is not None and revisions.intersection(p_ids))
1092
## return simple_ancestors
1094
# Yet Another Way, invert the parent map (which can be cached)
1096
## for revision_id, parent_ids in parent_map.iteritems():
1097
## for p_id in parent_ids:
1098
## descendants.setdefault(p_id, []).append(revision_id)
1099
## for revision in revisions.intersection(descendants):
1100
## simple_ancestors.difference_update(descendants[revision])
1101
## return simple_ancestors
1102
for revision, parent_ids in parent_map.iteritems():
1103
if parent_ids is None:
1105
for parent_id in parent_ids:
1106
if parent_id in revisions:
1107
# This node has a parent present in the set, so we can
1109
simple_ancestors.discard(revision)
1111
return simple_ancestors
1114
class HeadsCache(object):
1115
"""A cache of results for graph heads calls."""
1117
def __init__(self, graph):
1121
def heads(self, keys):
1122
"""Return the heads of keys.
1124
This matches the API of Graph.heads(), specifically the return value is
1125
a set which can be mutated, and ordering of the input is not preserved
1128
:see also: Graph.heads.
1129
:param keys: The keys to calculate heads for.
1130
:return: A set containing the heads, which may be mutated without
1131
affecting future lookups.
1133
keys = frozenset(keys)
1135
return set(self._heads[keys])
1137
heads = self.graph.heads(keys)
1138
self._heads[keys] = heads
1142
class FrozenHeadsCache(object):
1143
"""Cache heads() calls, assuming the caller won't modify them."""
1145
def __init__(self, graph):
1149
def heads(self, keys):
1150
"""Return the heads of keys.
1152
Similar to Graph.heads(). The main difference is that the return value
1153
is a frozen set which cannot be mutated.
1155
:see also: Graph.heads.
1156
:param keys: The keys to calculate heads for.
1157
:return: A frozenset containing the heads.
1159
keys = frozenset(keys)
1161
return self._heads[keys]
1163
heads = frozenset(self.graph.heads(keys))
1164
self._heads[keys] = heads
1167
def cache(self, keys, heads):
1168
"""Store a known value."""
1169
self._heads[frozenset(keys)] = frozenset(heads)
1172
class _BreadthFirstSearcher(object):
1173
"""Parallel search breadth-first the ancestry of revisions.
1175
This class implements the iterator protocol, but additionally
1176
1. provides a set of seen ancestors, and
1177
2. allows some ancestries to be unsearched, via stop_searching_any
1180
def __init__(self, revisions, parents_provider):
1181
self._iterations = 0
1182
self._next_query = set(revisions)
1184
self._started_keys = set(self._next_query)
1185
self._stopped_keys = set()
1186
self._parents_provider = parents_provider
1187
self._returning = 'next_with_ghosts'
1188
self._current_present = set()
1189
self._current_ghosts = set()
1190
self._current_parents = {}
1193
if self._iterations:
1194
prefix = "searching"
1197
search = '%s=%r' % (prefix, list(self._next_query))
1198
return ('_BreadthFirstSearcher(iterations=%d, %s,'
1199
' seen=%r)' % (self._iterations, search, list(self.seen)))
1201
def get_result(self):
1202
"""Get a SearchResult for the current state of this searcher.
1204
:return: A SearchResult for this search so far. The SearchResult is
1205
static - the search can be advanced and the search result will not
1206
be invalidated or altered.
1208
if self._returning == 'next':
1209
# We have to know the current nodes children to be able to list the
1210
# exclude keys for them. However, while we could have a second
1211
# look-ahead result buffer and shuffle things around, this method
1212
# is typically only called once per search - when memoising the
1213
# results of the search.
1214
found, ghosts, next, parents = self._do_query(self._next_query)
1215
# pretend we didn't query: perhaps we should tweak _do_query to be
1216
# entirely stateless?
1217
self.seen.difference_update(next)
1218
next_query = next.union(ghosts)
1220
next_query = self._next_query
1221
excludes = self._stopped_keys.union(next_query)
1222
included_keys = self.seen.difference(excludes)
1223
return SearchResult(self._started_keys, excludes, len(included_keys),
1229
except StopIteration:
1233
"""Return the next ancestors of this revision.
1235
Ancestors are returned in the order they are seen in a breadth-first
1236
traversal. No ancestor will be returned more than once. Ancestors are
1237
returned before their parentage is queried, so ghosts and missing
1238
revisions (including the start revisions) are included in the result.
1239
This can save a round trip in LCA style calculation by allowing
1240
convergence to be detected without reading the data for the revision
1241
the convergence occurs on.
1243
:return: A set of revision_ids.
1245
if self._returning != 'next':
1246
# switch to returning the query, not the results.
1247
self._returning = 'next'
1248
self._iterations += 1
1251
if len(self._next_query) == 0:
1252
raise StopIteration()
1253
# We have seen what we're querying at this point as we are returning
1254
# the query, not the results.
1255
self.seen.update(self._next_query)
1256
return self._next_query
1258
def next_with_ghosts(self):
1259
"""Return the next found ancestors, with ghosts split out.
1261
Ancestors are returned in the order they are seen in a breadth-first
1262
traversal. No ancestor will be returned more than once. Ancestors are
1263
returned only after asking for their parents, which allows us to detect
1264
which revisions are ghosts and which are not.
1266
:return: A tuple with (present ancestors, ghost ancestors) sets.
1268
if self._returning != 'next_with_ghosts':
1269
# switch to returning the results, not the current query.
1270
self._returning = 'next_with_ghosts'
1272
if len(self._next_query) == 0:
1273
raise StopIteration()
1275
return self._current_present, self._current_ghosts
1278
"""Advance the search.
1280
Updates self.seen, self._next_query, self._current_present,
1281
self._current_ghosts, self._current_parents and self._iterations.
1283
self._iterations += 1
1284
found, ghosts, next, parents = self._do_query(self._next_query)
1285
self._current_present = found
1286
self._current_ghosts = ghosts
1287
self._next_query = next
1288
self._current_parents = parents
1289
# ghosts are implicit stop points, otherwise the search cannot be
1290
# repeated when ghosts are filled.
1291
self._stopped_keys.update(ghosts)
1293
def _do_query(self, revisions):
1294
"""Query for revisions.
1296
Adds revisions to the seen set.
1298
:param revisions: Revisions to query.
1299
:return: A tuple: (set(found_revisions), set(ghost_revisions),
1300
set(parents_of_found_revisions), dict(found_revisions:parents)).
1302
found_revisions = set()
1303
parents_of_found = set()
1304
# revisions may contain nodes that point to other nodes in revisions:
1305
# we want to filter them out.
1306
self.seen.update(revisions)
1307
parent_map = self._parents_provider.get_parent_map(revisions)
1308
found_revisions.update(parent_map)
1309
for rev_id, parents in parent_map.iteritems():
1312
new_found_parents = [p for p in parents if p not in self.seen]
1313
if new_found_parents:
1314
# Calling set.update() with an empty generator is actually
1316
parents_of_found.update(new_found_parents)
1317
ghost_revisions = revisions - found_revisions
1318
return found_revisions, ghost_revisions, parents_of_found, parent_map
1323
def find_seen_ancestors(self, revisions):
1324
"""Find ancestors of these revisions that have already been seen.
1326
This function generally makes the assumption that querying for the
1327
parents of a node that has already been queried is reasonably cheap.
1328
(eg, not a round trip to a remote host).
1330
# TODO: Often we might ask one searcher for its seen ancestors, and
1331
# then ask another searcher the same question. This can result in
1332
# searching the same revisions repeatedly if the two searchers
1333
# have a lot of overlap.
1334
all_seen = self.seen
1335
pending = set(revisions).intersection(all_seen)
1336
seen_ancestors = set(pending)
1338
if self._returning == 'next':
1339
# self.seen contains what nodes have been returned, not what nodes
1340
# have been queried. We don't want to probe for nodes that haven't
1341
# been searched yet.
1342
not_searched_yet = self._next_query
1344
not_searched_yet = ()
1345
pending.difference_update(not_searched_yet)
1346
get_parent_map = self._parents_provider.get_parent_map
1348
parent_map = get_parent_map(pending)
1350
# We don't care if it is a ghost, since it can't be seen if it is
1352
for parent_ids in parent_map.itervalues():
1353
all_parents.extend(parent_ids)
1354
next_pending = all_seen.intersection(all_parents).difference(seen_ancestors)
1355
seen_ancestors.update(next_pending)
1356
next_pending.difference_update(not_searched_yet)
1357
pending = next_pending
1359
return seen_ancestors
1361
def stop_searching_any(self, revisions):
1363
Remove any of the specified revisions from the search list.
1365
None of the specified revisions are required to be present in the
1368
It is okay to call stop_searching_any() for revisions which were seen
1369
in previous iterations. It is the callers responsibility to call
1370
find_seen_ancestors() to make sure that current search tips that are
1371
ancestors of those revisions are also stopped. All explicitly stopped
1372
revisions will be excluded from the search result's get_keys(), though.
1374
# TODO: does this help performance?
1377
revisions = frozenset(revisions)
1378
if self._returning == 'next':
1379
stopped = self._next_query.intersection(revisions)
1380
self._next_query = self._next_query.difference(revisions)
1382
stopped_present = self._current_present.intersection(revisions)
1383
stopped = stopped_present.union(
1384
self._current_ghosts.intersection(revisions))
1385
self._current_present.difference_update(stopped)
1386
self._current_ghosts.difference_update(stopped)
1387
# stopping 'x' should stop returning parents of 'x', but
1388
# not if 'y' always references those same parents
1389
stop_rev_references = {}
1390
for rev in stopped_present:
1391
for parent_id in self._current_parents[rev]:
1392
if parent_id not in stop_rev_references:
1393
stop_rev_references[parent_id] = 0
1394
stop_rev_references[parent_id] += 1
1395
# if only the stopped revisions reference it, the ref count will be
1397
for parents in self._current_parents.itervalues():
1398
for parent_id in parents:
1400
stop_rev_references[parent_id] -= 1
1403
stop_parents = set()
1404
for rev_id, refs in stop_rev_references.iteritems():
1406
stop_parents.add(rev_id)
1407
self._next_query.difference_update(stop_parents)
1408
self._stopped_keys.update(stopped)
1409
self._stopped_keys.update(revisions)
1412
def start_searching(self, revisions):
1413
"""Add revisions to the search.
1415
The parents of revisions will be returned from the next call to next()
1416
or next_with_ghosts(). If next_with_ghosts was the most recently used
1417
next* call then the return value is the result of looking up the
1418
ghost/not ghost status of revisions. (A tuple (present, ghosted)).
1420
revisions = frozenset(revisions)
1421
self._started_keys.update(revisions)
1422
new_revisions = revisions.difference(self.seen)
1423
if self._returning == 'next':
1424
self._next_query.update(new_revisions)
1425
self.seen.update(new_revisions)
1427
# perform a query on revisions
1428
revs, ghosts, query, parents = self._do_query(revisions)
1429
self._stopped_keys.update(ghosts)
1430
self._current_present.update(revs)
1431
self._current_ghosts.update(ghosts)
1432
self._next_query.update(query)
1433
self._current_parents.update(parents)
1437
class SearchResult(object):
1438
"""The result of a breadth first search.
1440
A SearchResult provides the ability to reconstruct the search or access a
1441
set of the keys the search found.
1444
def __init__(self, start_keys, exclude_keys, key_count, keys):
1445
"""Create a SearchResult.
1447
:param start_keys: The keys the search started at.
1448
:param exclude_keys: The keys the search excludes.
1449
:param key_count: The total number of keys (from start to but not
1451
:param keys: The keys the search found. Note that in future we may get
1452
a SearchResult from a smart server, in which case the keys list is
1453
not necessarily immediately available.
1455
self._recipe = ('search', start_keys, exclude_keys, key_count)
1456
self._keys = frozenset(keys)
1458
def get_recipe(self):
1459
"""Return a recipe that can be used to replay this search.
1461
The recipe allows reconstruction of the same results at a later date
1462
without knowing all the found keys. The essential elements are a list
1463
of keys to start and to stop at. In order to give reproducible
1464
results when ghosts are encountered by a search they are automatically
1465
added to the exclude list (or else ghost filling may alter the
1468
:return: A tuple ('search', start_keys_set, exclude_keys_set,
1469
revision_count). To recreate the results of this search, create a
1470
breadth first searcher on the same graph starting at start_keys.
1471
Then call next() (or next_with_ghosts()) repeatedly, and on every
1472
result, call stop_searching_any on any keys from the exclude_keys
1473
set. The revision_count value acts as a trivial cross-check - the
1474
found revisions of the new search should have as many elements as
1475
revision_count. If it does not, then additional revisions have been
1476
ghosted since the search was executed the first time and the second
1482
"""Return the keys found in this search.
1484
:return: A set of keys.
1489
"""Return false if the search lists 1 or more revisions."""
1490
return self._recipe[3] == 0
1492
def refine(self, seen, referenced):
1493
"""Create a new search by refining this search.
1495
:param seen: Revisions that have been satisfied.
1496
:param referenced: Revision references observed while satisfying some
1499
start = self._recipe[1]
1500
exclude = self._recipe[2]
1501
count = self._recipe[3]
1502
keys = self.get_keys()
1503
# New heads = referenced + old heads - seen things - exclude
1504
pending_refs = set(referenced)
1505
pending_refs.update(start)
1506
pending_refs.difference_update(seen)
1507
pending_refs.difference_update(exclude)
1508
# New exclude = old exclude + satisfied heads
1509
seen_heads = start.intersection(seen)
1510
exclude.update(seen_heads)
1511
# keys gets seen removed
1513
# length is reduced by len(seen)
1515
return SearchResult(pending_refs, exclude, count, keys)
1518
class PendingAncestryResult(object):
1519
"""A search result that will reconstruct the ancestry for some graph heads.
1521
Unlike SearchResult, this doesn't hold the complete search result in
1522
memory, it just holds a description of how to generate it.
1525
def __init__(self, heads, repo):
1528
:param heads: an iterable of graph heads.
1529
:param repo: a repository to use to generate the ancestry for the given
1532
self.heads = frozenset(heads)
1535
def get_recipe(self):
1536
"""Return a recipe that can be used to replay this search.
1538
The recipe allows reconstruction of the same results at a later date.
1540
:seealso SearchResult.get_recipe:
1542
:return: A tuple ('proxy-search', start_keys_set, set(), -1)
1543
To recreate this result, create a PendingAncestryResult with the
1546
return ('proxy-search', self.heads, set(), -1)
1549
"""See SearchResult.get_keys.
1551
Returns all the keys for the ancestry of the heads, excluding
1554
return self._get_keys(self.repo.get_graph())
1556
def _get_keys(self, graph):
1557
NULL_REVISION = revision.NULL_REVISION
1558
keys = [key for (key, parents) in graph.iter_ancestry(self.heads)
1559
if key != NULL_REVISION]
1563
"""Return false if the search lists 1 or more revisions."""
1564
if revision.NULL_REVISION in self.heads:
1565
return len(self.heads) == 1
1567
return len(self.heads) == 0
1569
def refine(self, seen, referenced):
1570
"""Create a new search by refining this search.
1572
:param seen: Revisions that have been satisfied.
1573
:param referenced: Revision references observed while satisfying some
1576
referenced = self.heads.union(referenced)
1577
return PendingAncestryResult(referenced - seen, self.repo)
1580
def collapse_linear_regions(parent_map):
1581
"""Collapse regions of the graph that are 'linear'.
1587
can be collapsed by removing B and getting::
1591
:param parent_map: A dictionary mapping children to their parents
1592
:return: Another dictionary with 'linear' chains collapsed
1594
# Note: this isn't a strictly minimal collapse. For example:
1602
# Will not have 'D' removed, even though 'E' could fit. Also:
1608
# A and C are both kept because they are edges of the graph. We *could* get
1609
# rid of A if we wanted.
1617
# Will not have any nodes removed, even though you do have an
1618
# 'uninteresting' linear D->B and E->C
1620
for child, parents in parent_map.iteritems():
1621
children.setdefault(child, [])
1623
children.setdefault(p, []).append(child)
1625
orig_children = dict(children)
1627
result = dict(parent_map)
1628
for node in parent_map:
1629
parents = result[node]
1630
if len(parents) == 1:
1631
parent_children = children[parents[0]]
1632
if len(parent_children) != 1:
1633
# This is not the only child
1635
node_children = children[node]
1636
if len(node_children) != 1:
1638
child_parents = result.get(node_children[0], None)
1639
if len(child_parents) != 1:
1640
# This is not its only parent
1642
# The child of this node only points at it, and the parent only has
1643
# this as a child. remove this node, and join the others together
1644
result[node_children[0]] = parents
1645
children[parents[0]] = node_children
added
removed
bzrlib/graph.py
