170
165
return self.bzrdir.is_control_filename(filename)
173
def iter_entries_by_dir(self, specific_file_ids=None, yield_parents=False):
168
def iter_entries_by_dir(self, specific_file_ids=None):
174
169
"""Walk the tree in 'by_dir' order.
176
This will yield each entry in the tree as a (path, entry) tuple.
177
The order that they are yielded is:
179
Directories are walked in a depth-first lexicographical order,
180
however, whenever a directory is reached, all of its direct child
181
nodes are yielded in lexicographical order before yielding the
184
For example, in the tree::
194
The yield order (ignoring root) would be::
195
a, f, a/b, a/d, a/b/c, a/d/e, f/g
197
:param yield_parents: If True, yield the parents from the root leading
198
down to specific_file_ids that have been requested. This has no
199
impact if specific_file_ids is None.
171
This will yield each entry in the tree as a (path, entry) tuple. The
172
order that they are yielded is: the contents of a directory are
173
preceeded by the parent of a directory, and all the contents of a
174
directory are grouped together.
201
176
return self.inventory.iter_entries_by_dir(
202
specific_file_ids=specific_file_ids, yield_parents=yield_parents)
177
specific_file_ids=specific_file_ids)
204
179
def iter_references(self):
205
if self.supports_tree_reference():
206
for path, entry in self.iter_entries_by_dir():
207
if entry.kind == 'tree-reference':
208
yield path, entry.file_id
180
for path, entry in self.iter_entries_by_dir():
181
if entry.kind == 'tree-reference':
182
yield path, entry.file_id
210
184
def kind(self, file_id):
211
185
raise NotImplementedError("Tree subclass %s must implement kind"
212
186
% self.__class__.__name__)
214
def stored_kind(self, file_id):
215
"""File kind stored for this file_id.
217
May not match kind on disk for working trees. Always available
218
for versioned files, even when the file itself is missing.
220
return self.kind(file_id)
222
188
def path_content_summary(self, path):
223
189
"""Get a summary of the information about path.
225
All the attributes returned are for the canonical form, not the
226
convenient form (if content filters are in use.)
228
191
:param path: A relative path within the tree.
229
192
:return: A tuple containing kind, size, exec, sha1-or-link.
230
193
Kind is always present (see tree.kind()).
231
size is present if kind is file and the size of the
232
canonical form can be cheaply determined, None otherwise.
194
size is present if kind is file, None otherwise.
233
195
exec is None unless kind is file and the platform supports the 'x'
235
197
sha1-or-link is the link target if kind is symlink, or the sha1 if
363
278
raise NotImplementedError(self.get_symlink_target)
365
def get_canonical_inventory_paths(self, paths):
366
"""Like get_canonical_inventory_path() but works on multiple items.
368
:param paths: A sequence of paths relative to the root of the tree.
369
:return: A list of paths, with each item the corresponding input path
370
adjusted to account for existing elements that match case
373
return list(self._yield_canonical_inventory_paths(paths))
375
def get_canonical_inventory_path(self, path):
376
"""Returns the first inventory item that case-insensitively matches path.
378
If a path matches exactly, it is returned. If no path matches exactly
379
but more than one path matches case-insensitively, it is implementation
380
defined which is returned.
382
If no path matches case-insensitively, the input path is returned, but
383
with as many path entries that do exist changed to their canonical
386
If you need to resolve many names from the same tree, you should
387
use get_canonical_inventory_paths() to avoid O(N) behaviour.
389
:param path: A paths relative to the root of the tree.
390
:return: The input path adjusted to account for existing elements
391
that match case insensitively.
393
return self._yield_canonical_inventory_paths([path]).next()
395
def _yield_canonical_inventory_paths(self, paths):
397
# First, if the path as specified exists exactly, just use it.
398
if self.path2id(path) is not None:
402
cur_id = self.get_root_id()
404
bit_iter = iter(path.split("/"))
408
for child in self.iter_children(cur_id):
410
# XXX: it seem like if the child is known to be in the
411
# tree, we shouldn't need to go from its id back to
412
# its path -- mbp 2010-02-11
414
# XXX: it seems like we could be more efficient
415
# by just directly looking up the original name and
416
# only then searching all children; also by not
417
# chopping paths so much. -- mbp 2010-02-11
418
child_base = os.path.basename(self.id2path(child))
419
if (child_base == elt):
420
# if we found an exact match, we can stop now; if
421
# we found an approximate match we need to keep
422
# searching because there might be an exact match
425
new_path = osutils.pathjoin(cur_path, child_base)
427
elif child_base.lower() == lelt:
429
new_path = osutils.pathjoin(cur_path, child_base)
431
# before a change is committed we can see this error...
436
# got to the end of this directory and no entries matched.
437
# Return what matched so far, plus the rest as specified.
438
cur_path = osutils.pathjoin(cur_path, elt, *list(bit_iter))
443
280
def get_root_id(self):
444
281
"""Return the file_id for the root of this tree."""
445
282
raise NotImplementedError(self.get_root_id)
447
def annotate_iter(self, file_id,
448
default_revision=_mod_revision.CURRENT_REVISION):
284
def annotate_iter(self, file_id):
449
285
"""Return an iterator of revision_id, line tuples.
451
287
For working trees (and mutable trees in general), the special
452
288
revision_id 'current:' will be used for lines that are new in this
453
289
tree, e.g. uncommitted changes.
454
290
:param file_id: The file to produce an annotated version from
455
:param default_revision: For lines that don't match a basis, mark them
456
with this revision id. Not all implementations will make use of
459
292
raise NotImplementedError(self.annotate_iter)
461
def _get_plan_merge_data(self, file_id, other, base):
462
from bzrlib import versionedfile
294
def plan_file_merge(self, file_id, other):
295
"""Generate a merge plan based on annotations.
297
If the file contains uncommitted changes in this tree, they will be
298
attributed to the 'current:' pseudo-revision. If the file contains
299
uncommitted changes in the other tree, they will be assigned to the
300
'other:' pseudo-revision.
302
from bzrlib import merge, versionedfile
463
303
vf = versionedfile._PlanMergeVersionedFile(file_id)
464
304
last_revision_a = self._get_file_revision(file_id, vf, 'this:')
465
305
last_revision_b = other._get_file_revision(file_id, vf, 'other:')
467
last_revision_base = None
469
last_revision_base = base._get_file_revision(file_id, vf, 'base:')
470
return vf, last_revision_a, last_revision_b, last_revision_base
472
def plan_file_merge(self, file_id, other, base=None):
473
"""Generate a merge plan based on annotations.
475
If the file contains uncommitted changes in this tree, they will be
476
attributed to the 'current:' pseudo-revision. If the file contains
477
uncommitted changes in the other tree, they will be assigned to the
478
'other:' pseudo-revision.
480
data = self._get_plan_merge_data(file_id, other, base)
481
vf, last_revision_a, last_revision_b, last_revision_base = data
482
return vf.plan_merge(last_revision_a, last_revision_b,
485
def plan_file_lca_merge(self, file_id, other, base=None):
486
"""Generate a merge plan based lca-newness.
488
If the file contains uncommitted changes in this tree, they will be
489
attributed to the 'current:' pseudo-revision. If the file contains
490
uncommitted changes in the other tree, they will be assigned to the
491
'other:' pseudo-revision.
493
data = self._get_plan_merge_data(file_id, other, base)
494
vf, last_revision_a, last_revision_b, last_revision_base = data
495
return vf.plan_lca_merge(last_revision_a, last_revision_b,
498
def _iter_parent_trees(self):
499
"""Iterate through parent trees, defaulting to Tree.revision_tree."""
500
for revision_id in self.get_parent_ids():
502
yield self.revision_tree(revision_id)
503
except errors.NoSuchRevisionInTree:
504
yield self.repository.revision_tree(revision_id)
507
def _file_revision(revision_tree, file_id):
508
"""Determine the revision associated with a file in a given tree."""
509
revision_tree.lock_read()
511
return revision_tree.inventory[file_id].revision
513
revision_tree.unlock()
306
return vf.plan_merge(last_revision_a, last_revision_b)
515
308
def _get_file_revision(self, file_id, vf, tree_revision):
516
"""Ensure that file_id, tree_revision is in vf to plan the merge."""
518
if getattr(self, '_repository', None) is None:
309
def file_revision(revision_tree):
310
revision_tree.lock_read()
312
return revision_tree.inventory[file_id].revision
314
revision_tree.unlock()
316
def iter_parent_trees():
317
for revision_id in self.get_parent_ids():
319
yield self.revision_tree(revision_id)
321
yield self.repository.revision_tree(revision_id)
323
if getattr(self, '_get_weave', None) is None:
519
324
last_revision = tree_revision
520
parent_keys = [(file_id, self._file_revision(t, file_id)) for t in
521
self._iter_parent_trees()]
522
vf.add_lines((file_id, last_revision), parent_keys,
325
parent_revisions = [file_revision(t) for t in iter_parent_trees()]
326
vf.add_lines(last_revision, parent_revisions,
523
327
self.get_file(file_id).readlines())
524
328
repo = self.branch.repository
329
transaction = repo.get_transaction()
330
base_vf = repo.weave_store.get_weave(file_id, transaction)
527
last_revision = self._file_revision(self, file_id)
528
base_vf = self._repository.texts
529
if base_vf not in vf.fallback_versionedfiles:
530
vf.fallback_versionedfiles.append(base_vf)
332
last_revision = file_revision(self)
333
base_vf = self._get_weave(file_id)
334
vf.fallback_versionedfiles.append(base_vf)
531
335
return last_revision
533
337
inventory = property(_get_inventory,
654
458
raise NotImplementedError(self.walkdirs)
656
def supports_content_filtering(self):
461
class EmptyTree(Tree):
464
self._inventory = Inventory(root_id=None)
465
symbol_versioning.warn('EmptyTree is deprecated as of bzr 0.9 please'
466
' use repository.revision_tree instead.',
467
DeprecationWarning, stacklevel=2)
469
def get_parent_ids(self):
472
def get_symlink_target(self, file_id):
475
def has_filename(self, filename):
659
def _content_filter_stack(self, path=None, file_id=None):
660
"""The stack of content filters for a path if filtering is supported.
662
Readers will be applied in first-to-last order.
663
Writers will be applied in last-to-first order.
664
Either the path or the file-id needs to be provided.
666
:param path: path relative to the root of the tree
668
:param file_id: file_id or None if unknown
669
:return: the list of filters - [] if there are none
671
filter_pref_names = filters._get_registered_names()
672
if len(filter_pref_names) == 0:
675
path = self.id2path(file_id)
676
prefs = self.iter_search_rules([path], filter_pref_names).next()
677
stk = filters._get_filter_stack_for(prefs)
678
if 'filters' in debug.debug_flags:
679
note("*** %s content-filter: %s => %r" % (path,prefs,stk))
682
def _content_filter_stack_provider(self):
683
"""A function that returns a stack of ContentFilters.
685
The function takes a path (relative to the top of the tree) and a
686
file-id as parameters.
688
:return: None if content filtering is not supported by this tree.
690
if self.supports_content_filtering():
691
return lambda path, file_id: \
692
self._content_filter_stack(path, file_id)
696
def iter_search_rules(self, path_names, pref_names=None,
697
_default_searcher=None):
698
"""Find the preferences for filenames in a tree.
700
:param path_names: an iterable of paths to find attributes for.
701
Paths are given relative to the root of the tree.
702
:param pref_names: the list of preferences to lookup - None for all
703
:param _default_searcher: private parameter to assist testing - don't use
704
:return: an iterator of tuple sequences, one per path-name.
705
See _RulesSearcher.get_items for details on the tuple sequence.
707
if _default_searcher is None:
708
_default_searcher = rules._per_user_searcher
709
searcher = self._get_rules_searcher(_default_searcher)
710
if searcher is not None:
711
if pref_names is not None:
712
for path in path_names:
713
yield searcher.get_selected_items(path, pref_names)
715
for path in path_names:
716
yield searcher.get_items(path)
718
def _get_rules_searcher(self, default_searcher):
719
"""Get the RulesSearcher for this tree given the default one."""
720
searcher = default_searcher
478
def kind(self, file_id):
479
assert self._inventory[file_id].kind == "directory"
482
def list_files(self, include_root=False):
485
def __contains__(self, file_id):
486
return (file_id in self._inventory)
488
def get_file_sha1(self, file_id, path=None, stat_value=None):
724
492
######################################################################
869
639
will pass through to InterTree as appropriate.
872
# Formats that will be used to test this InterTree. If both are
873
# None, this InterTree will not be tested (e.g. because a complex
875
_matching_from_tree_format = None
876
_matching_to_tree_format = None
880
def _changes_from_entries(self, source_entry, target_entry,
881
source_path=None, target_path=None):
882
"""Generate a iter_changes tuple between source_entry and target_entry.
884
:param source_entry: An inventory entry from self.source, or None.
885
:param target_entry: An inventory entry from self.target, or None.
886
:param source_path: The path of source_entry, if known. If not known
887
it will be looked up.
888
:param target_path: The path of target_entry, if known. If not known
889
it will be looked up.
890
:return: A tuple, item 0 of which is an iter_changes result tuple, and
891
item 1 is True if there are any changes in the result tuple.
893
if source_entry is None:
894
if target_entry is None:
896
file_id = target_entry.file_id
898
file_id = source_entry.file_id
899
if source_entry is not None:
900
source_versioned = True
901
source_name = source_entry.name
902
source_parent = source_entry.parent_id
903
if source_path is None:
904
source_path = self.source.id2path(file_id)
905
source_kind, source_executable, source_stat = \
906
self.source._comparison_data(source_entry, source_path)
908
source_versioned = False
912
source_executable = None
913
if target_entry is not None:
914
target_versioned = True
915
target_name = target_entry.name
916
target_parent = target_entry.parent_id
917
if target_path is None:
918
target_path = self.target.id2path(file_id)
919
target_kind, target_executable, target_stat = \
920
self.target._comparison_data(target_entry, target_path)
922
target_versioned = False
926
target_executable = None
927
versioned = (source_versioned, target_versioned)
928
kind = (source_kind, target_kind)
929
changed_content = False
930
if source_kind != target_kind:
931
changed_content = True
932
elif source_kind == 'file':
933
if (self.source.get_file_sha1(file_id, source_path, source_stat) !=
934
self.target.get_file_sha1(file_id, target_path, target_stat)):
935
changed_content = True
936
elif source_kind == 'symlink':
937
if (self.source.get_symlink_target(file_id) !=
938
self.target.get_symlink_target(file_id)):
939
changed_content = True
940
# XXX: Yes, the indentation below is wrong. But fixing it broke
941
# test_merge.TestMergerEntriesLCAOnDisk.
942
# test_nested_tree_subtree_renamed_and_modified. We'll wait for
943
# the fix from bzr.dev -- vila 2009026
944
elif source_kind == 'tree-reference':
945
if (self.source.get_reference_revision(file_id, source_path)
946
!= self.target.get_reference_revision(file_id, target_path)):
947
changed_content = True
948
parent = (source_parent, target_parent)
949
name = (source_name, target_name)
950
executable = (source_executable, target_executable)
951
if (changed_content is not False or versioned[0] != versioned[1]
952
or parent[0] != parent[1] or name[0] != name[1] or
953
executable[0] != executable[1]):
957
return (file_id, (source_path, target_path), changed_content,
958
versioned, parent, name, kind, executable), changes
961
645
def compare(self, want_unchanged=False, specific_files=None,
962
646
extra_trees=None, require_versioned=False, include_root=False,
1063
732
all_unversioned = deque()
1065
from_entries_by_dir = list(self.source.iter_entries_by_dir(
734
from_entries_by_dir = list(self.source.inventory.iter_entries_by_dir(
1066
735
specific_file_ids=specific_file_ids))
1067
736
from_data = dict((e.file_id, (p, e)) for p, e in from_entries_by_dir)
1068
to_entries_by_dir = list(self.target.iter_entries_by_dir(
737
to_entries_by_dir = list(self.target.inventory.iter_entries_by_dir(
1069
738
specific_file_ids=specific_file_ids))
1070
739
num_entries = len(from_entries_by_dir) + len(to_entries_by_dir)
1072
# the unversioned path lookup only occurs on real trees - where there
741
# the unversioned path lookup only occurs on real trees - where there
1073
742
# can be extras. So the fake_entry is solely used to look up
1074
743
# executable it values when execute is not supported.
1075
744
fake_entry = InventoryFile('unused', 'unused', 'unused')
1076
for target_path, target_entry in to_entries_by_dir:
1077
while (all_unversioned and
1078
all_unversioned[0][0] < target_path.split('/')):
745
for to_path, to_entry in to_entries_by_dir:
746
while all_unversioned and all_unversioned[0][0] < to_path.split('/'):
1079
747
unversioned_path = all_unversioned.popleft()
1080
target_kind, target_executable, target_stat = \
748
to_kind, to_executable, to_stat = \
1081
749
self.target._comparison_data(fake_entry, unversioned_path[1])
1082
750
yield (None, (None, unversioned_path[1]), True, (False, False),
1084
752
(None, unversioned_path[0][-1]),
1085
(None, target_kind),
1086
(None, target_executable))
1087
source_path, source_entry = from_data.get(target_entry.file_id,
1089
result, changes = self._changes_from_entries(source_entry,
1090
target_entry, source_path=source_path, target_path=target_path)
1091
to_paths[result[0]] = result[1][1]
754
(None, to_executable))
755
file_id = to_entry.file_id
756
to_paths[file_id] = to_path
1092
757
entry_count += 1
758
changed_content = False
759
from_path, from_entry = from_data.get(file_id, (None, None))
760
from_versioned = (from_entry is not None)
761
if from_entry is not None:
762
from_versioned = True
763
from_name = from_entry.name
764
from_parent = from_entry.parent_id
765
from_kind, from_executable, from_stat = \
766
self.source._comparison_data(from_entry, from_path)
1094
767
entry_count += 1
769
from_versioned = False
773
from_executable = None
774
versioned = (from_versioned, True)
775
to_kind, to_executable, to_stat = \
776
self.target._comparison_data(to_entry, to_path)
777
kind = (from_kind, to_kind)
778
if kind[0] != kind[1]:
779
changed_content = True
780
elif from_kind == 'file':
781
from_size = self.source._file_size(from_entry, from_stat)
782
to_size = self.target._file_size(to_entry, to_stat)
783
if from_size != to_size:
784
changed_content = True
785
elif (self.source.get_file_sha1(file_id, from_path, from_stat) !=
786
self.target.get_file_sha1(file_id, to_path, to_stat)):
787
changed_content = True
788
elif from_kind == 'symlink':
789
if (self.source.get_symlink_target(file_id) !=
790
self.target.get_symlink_target(file_id)):
791
changed_content = True
792
elif from_kind == 'tree-reference':
793
if (self.source.get_reference_revision(file_id, from_path)
794
!= self.target.get_reference_revision(file_id, to_path)):
795
changed_content = True
796
parent = (from_parent, to_entry.parent_id)
797
name = (from_name, to_entry.name)
798
executable = (from_executable, to_executable)
1095
799
if pb is not None:
1096
800
pb.update('comparing files', entry_count, num_entries)
1097
if changes or include_unchanged:
1098
if specific_file_ids is not None:
1099
new_parent_id = result[4][1]
1100
precise_file_ids.add(new_parent_id)
1101
changed_file_ids.append(result[0])
1103
# Ensure correct behaviour for reparented/added specific files.
1104
if specific_files is not None:
1105
# Record output dirs
1106
if result[6][1] == 'directory':
1107
seen_dirs.add(result[0])
1108
# Record parents of reparented/added entries.
1109
versioned = result[3]
1111
if not versioned[0] or parents[0] != parents[1]:
1112
seen_parents.add(parents[1])
801
if (changed_content is not False or versioned[0] != versioned[1]
802
or parent[0] != parent[1] or name[0] != name[1] or
803
executable[0] != executable[1] or include_unchanged):
804
yield (file_id, (from_path, to_path), changed_content,
805
versioned, parent, name, kind, executable)
1113
807
while all_unversioned:
1114
808
# yield any trailing unversioned paths
1115
809
unversioned_path = all_unversioned.popleft()
1142
848
self.source._comparison_data(from_entry, path)
1143
849
kind = (from_kind, None)
1144
850
executable = (from_executable, None)
1145
changed_content = from_kind is not None
851
changed_content = True
1146
852
# the parent's path is necessarily known at this point.
1147
changed_file_ids.append(file_id)
1148
853
yield(file_id, (path, to_path), changed_content, versioned, parent,
1149
854
name, kind, executable)
1150
changed_file_ids = set(changed_file_ids)
1151
if specific_file_ids is not None:
1152
for result in self._handle_precise_ids(precise_file_ids,
1156
def _get_entry(self, tree, file_id):
1157
"""Get an inventory entry from a tree, with missing entries as None.
1159
If the tree raises NotImplementedError on accessing .inventory, then
1160
this is worked around using iter_entries_by_dir on just the file id
1163
:param tree: The tree to lookup the entry in.
1164
:param file_id: The file_id to lookup.
1167
inventory = tree.inventory
1168
except NotImplementedError:
1169
# No inventory available.
1171
iterator = tree.iter_entries_by_dir(specific_file_ids=[file_id])
1172
return iterator.next()[1]
1173
except StopIteration:
1177
return inventory[file_id]
1178
except errors.NoSuchId:
1181
def _handle_precise_ids(self, precise_file_ids, changed_file_ids,
1182
discarded_changes=None):
1183
"""Fill out a partial iter_changes to be consistent.
1185
:param precise_file_ids: The file ids of parents that were seen during
1187
:param changed_file_ids: The file ids of already emitted items.
1188
:param discarded_changes: An optional dict of precalculated
1189
iter_changes items which the partial iter_changes had not output
1191
:return: A generator of iter_changes items to output.
1193
# process parents of things that had changed under the users
1194
# requested paths to prevent incorrect paths or parent ids which
1195
# aren't in the tree.
1196
while precise_file_ids:
1197
precise_file_ids.discard(None)
1198
# Don't emit file_ids twice
1199
precise_file_ids.difference_update(changed_file_ids)
1200
if not precise_file_ids:
1202
# If the there was something at a given output path in source, we
1203
# have to include the entry from source in the delta, or we would
1204
# be putting this entry into a used path.
1206
for parent_id in precise_file_ids:
1208
paths.append(self.target.id2path(parent_id))
1209
except errors.NoSuchId:
1210
# This id has been dragged in from the source by delta
1211
# expansion and isn't present in target at all: we don't
1212
# need to check for path collisions on it.
1215
old_id = self.source.path2id(path)
1216
precise_file_ids.add(old_id)
1217
precise_file_ids.discard(None)
1218
current_ids = precise_file_ids
1219
precise_file_ids = set()
1220
# We have to emit all of precise_file_ids that have been altered.
1221
# We may have to output the children of some of those ids if any
1222
# directories have stopped being directories.
1223
for file_id in current_ids:
1225
if discarded_changes:
1226
result = discarded_changes.get(file_id)
1231
old_entry = self._get_entry(self.source, file_id)
1232
new_entry = self._get_entry(self.target, file_id)
1233
result, changes = self._changes_from_entries(
1234
old_entry, new_entry)
1237
# Get this parents parent to examine.
1238
new_parent_id = result[4][1]
1239
precise_file_ids.add(new_parent_id)
1241
if (result[6][0] == 'directory' and
1242
result[6][1] != 'directory'):
1243
# This stopped being a directory, the old children have
1245
if old_entry is None:
1246
# Reusing a discarded change.
1247
old_entry = self._get_entry(self.source, file_id)
1248
for child in old_entry.children.values():
1249
precise_file_ids.add(child.file_id)
1250
changed_file_ids.add(result[0])
1254
class MultiWalker(object):
1255
"""Walk multiple trees simultaneously, getting combined results."""
1257
# Note: This could be written to not assume you can do out-of-order
1258
# lookups. Instead any nodes that don't match in all trees could be
1259
# marked as 'deferred', and then returned in the final cleanup loop.
1260
# For now, I think it is "nicer" to return things as close to the
1261
# "master_tree" order as we can.
1263
def __init__(self, master_tree, other_trees):
1264
"""Create a new MultiWalker.
1266
All trees being walked must implement "iter_entries_by_dir()", such
1267
that they yield (path, object) tuples, where that object will have a
1268
'.file_id' member, that can be used to check equality.
1270
:param master_tree: All trees will be 'slaved' to the master_tree such
1271
that nodes in master_tree will be used as 'first-pass' sync points.
1272
Any nodes that aren't in master_tree will be merged in a second
1274
:param other_trees: A list of other trees to walk simultaneously.
1276
self._master_tree = master_tree
1277
self._other_trees = other_trees
1279
# Keep track of any nodes that were properly processed just out of
1280
# order, that way we don't return them at the end, we don't have to
1281
# track *all* processed file_ids, just the out-of-order ones
1282
self._out_of_order_processed = set()
1285
def _step_one(iterator):
1286
"""Step an iter_entries_by_dir iterator.
1288
:return: (has_more, path, ie)
1289
If has_more is False, path and ie will be None.
1292
path, ie = iterator.next()
1293
except StopIteration:
1294
return False, None, None
1296
return True, path, ie
1299
def _cmp_path_by_dirblock(path1, path2):
1300
"""Compare two paths based on what directory they are in.
1302
This generates a sort order, such that all children of a directory are
1303
sorted together, and grandchildren are in the same order as the
1304
children appear. But all grandchildren come after all children.
1306
:param path1: first path
1307
:param path2: the second path
1308
:return: negative number if ``path1`` comes first,
1309
0 if paths are equal
1310
and a positive number if ``path2`` sorts first
1312
# Shortcut this special case
1315
# This is stolen from _dirstate_helpers_py.py, only switching it to
1316
# Unicode objects. Consider using encode_utf8() and then using the
1317
# optimized versions, or maybe writing optimized unicode versions.
1318
if not isinstance(path1, unicode):
1319
raise TypeError("'path1' must be a unicode string, not %s: %r"
1320
% (type(path1), path1))
1321
if not isinstance(path2, unicode):
1322
raise TypeError("'path2' must be a unicode string, not %s: %r"
1323
% (type(path2), path2))
1324
return cmp(MultiWalker._path_to_key(path1),
1325
MultiWalker._path_to_key(path2))
1328
def _path_to_key(path):
1329
dirname, basename = osutils.split(path)
1330
return (dirname.split(u'/'), basename)
1332
def _lookup_by_file_id(self, extra_entries, other_tree, file_id):
1333
"""Lookup an inventory entry by file_id.
1335
This is called when an entry is missing in the normal order.
1336
Generally this is because a file was either renamed, or it was
1337
deleted/added. If the entry was found in the inventory and not in
1338
extra_entries, it will be added to self._out_of_order_processed
1340
:param extra_entries: A dictionary of {file_id: (path, ie)}. This
1341
should be filled with entries that were found before they were
1342
used. If file_id is present, it will be removed from the
1344
:param other_tree: The Tree to search, in case we didn't find the entry
1346
:param file_id: The file_id to look for
1347
:return: (path, ie) if found or (None, None) if not present.
1349
if file_id in extra_entries:
1350
return extra_entries.pop(file_id)
1351
# TODO: Is id2path better as the first call, or is
1352
# inventory[file_id] better as a first check?
1354
cur_path = other_tree.id2path(file_id)
1355
except errors.NoSuchId:
1357
if cur_path is None:
1360
self._out_of_order_processed.add(file_id)
1361
cur_ie = other_tree.inventory[file_id]
1362
return (cur_path, cur_ie)
1365
"""Match up the values in the different trees."""
1366
for result in self._walk_master_tree():
1368
self._finish_others()
1369
for result in self._walk_others():
1372
def _walk_master_tree(self):
1373
"""First pass, walk all trees in lock-step.
1375
When we are done, all nodes in the master_tree will have been
1376
processed. _other_walkers, _other_entries, and _others_extra will be
1377
set on 'self' for future processing.
1379
# This iterator has the most "inlining" done, because it tends to touch
1380
# every file in the tree, while the others only hit nodes that don't
1382
master_iterator = self._master_tree.iter_entries_by_dir()
1384
other_walkers = [other.iter_entries_by_dir()
1385
for other in self._other_trees]
1386
other_entries = [self._step_one(walker) for walker in other_walkers]
1387
# Track extra nodes in the other trees
1388
others_extra = [{} for i in xrange(len(self._other_trees))]
1390
master_has_more = True
1391
step_one = self._step_one
1392
lookup_by_file_id = self._lookup_by_file_id
1393
out_of_order_processed = self._out_of_order_processed
1395
while master_has_more:
1396
(master_has_more, path, master_ie) = step_one(master_iterator)
1397
if not master_has_more:
1400
file_id = master_ie.file_id
1402
other_values_append = other_values.append
1403
next_other_entries = []
1404
next_other_entries_append = next_other_entries.append
1405
for idx, (other_has_more, other_path, other_ie) in enumerate(other_entries):
1406
if not other_has_more:
1407
other_values_append(lookup_by_file_id(
1408
others_extra[idx], self._other_trees[idx], file_id))
1409
next_other_entries_append((False, None, None))
1410
elif file_id == other_ie.file_id:
1411
# This is the critical code path, as most of the entries
1412
# should match between most trees.
1413
other_values_append((other_path, other_ie))
1414
next_other_entries_append(step_one(other_walkers[idx]))
1416
# This walker did not match, step it until it either
1417
# matches, or we know we are past the current walker.
1418
other_walker = other_walkers[idx]
1419
other_extra = others_extra[idx]
1420
while (other_has_more and
1421
self._cmp_path_by_dirblock(other_path, path) < 0):
1422
other_file_id = other_ie.file_id
1423
if other_file_id not in out_of_order_processed:
1424
other_extra[other_file_id] = (other_path, other_ie)
1425
other_has_more, other_path, other_ie = \
1426
step_one(other_walker)
1427
if other_has_more and other_ie.file_id == file_id:
1428
# We ended up walking to this point, match and step
1430
other_values_append((other_path, other_ie))
1431
other_has_more, other_path, other_ie = \
1432
step_one(other_walker)
1434
# This record isn't in the normal order, see if it
1436
other_values_append(lookup_by_file_id(
1437
other_extra, self._other_trees[idx], file_id))
1438
next_other_entries_append((other_has_more, other_path,
1440
other_entries = next_other_entries
1442
# We've matched all the walkers, yield this datapoint
1443
yield path, file_id, master_ie, other_values
1444
self._other_walkers = other_walkers
1445
self._other_entries = other_entries
1446
self._others_extra = others_extra
1448
def _finish_others(self):
1449
"""Finish walking the other iterators, so we get all entries."""
1450
for idx, info in enumerate(self._other_entries):
1451
other_extra = self._others_extra[idx]
1452
(other_has_more, other_path, other_ie) = info
1453
while other_has_more:
1454
other_file_id = other_ie.file_id
1455
if other_file_id not in self._out_of_order_processed:
1456
other_extra[other_file_id] = (other_path, other_ie)
1457
other_has_more, other_path, other_ie = \
1458
self._step_one(self._other_walkers[idx])
1459
del self._other_entries
1461
def _walk_others(self):
1462
"""Finish up by walking all the 'deferred' nodes."""
1463
# TODO: One alternative would be to grab all possible unprocessed
1464
# file_ids, and then sort by path, and then yield them. That
1465
# might ensure better ordering, in case a caller strictly
1466
# requires parents before children.
1467
for idx, other_extra in enumerate(self._others_extra):
1468
others = sorted(other_extra.itervalues(),
1469
key=lambda x: self._path_to_key(x[0]))
1470
for other_path, other_ie in others:
1471
file_id = other_ie.file_id
1472
# We don't need to check out_of_order_processed here, because
1473
# the lookup_by_file_id will be removing anything processed
1474
# from the extras cache
1475
other_extra.pop(file_id)
1476
other_values = [(None, None) for i in xrange(idx)]
1477
other_values.append((other_path, other_ie))
1478
for alt_idx, alt_extra in enumerate(self._others_extra[idx+1:]):
1479
alt_idx = alt_idx + idx + 1
1480
alt_extra = self._others_extra[alt_idx]
1481
alt_tree = self._other_trees[alt_idx]
1482
other_values.append(self._lookup_by_file_id(
1483
alt_extra, alt_tree, file_id))
1484
yield other_path, file_id, None, other_values
857
# This was deprecated before 0.12, but did not have an official warning
858
@symbol_versioning.deprecated_function(symbol_versioning.zero_twelve)
859
def RevisionTree(*args, **kwargs):
860
"""RevisionTree has moved to bzrlib.revisiontree.RevisionTree()
862
Accessing it as bzrlib.tree.RevisionTree has been deprecated as of
865
from bzrlib.revisiontree import RevisionTree as _RevisionTree
866
return _RevisionTree(*args, **kwargs)
added
removed
bzrlib/tree.py
