140
145
def __iter__(self):
141
"""Yield all file ids in this tree."""
142
raise NotImplementedError(self.__iter__)
144
def all_file_ids(self):
145
"""Iterate through all file ids, including ids for missing files."""
146
return set(self.inventory)
146
return iter(self.inventory)
148
148
def id2path(self, file_id):
149
149
"""Return the path for a file id.
151
151
:raises NoSuchId:
153
raise NotImplementedError(self.id2path)
155
def iter_entries_by_dir(self, specific_file_ids=None, yield_parents=False):
153
file_id = osutils.safe_file_id(file_id)
154
return self.inventory.id2path(file_id)
156
def is_control_filename(self, filename):
157
"""True if filename is the name of a control file in this tree.
159
:param filename: A filename within the tree. This is a relative path
160
from the root of this tree.
162
This is true IF and ONLY IF the filename is part of the meta data
163
that bzr controls in this tree. I.E. a random .bzr directory placed
164
on disk will not be a control file for this tree.
166
return self.bzrdir.is_control_filename(filename)
169
def iter_entries_by_dir(self, specific_file_ids=None):
156
170
"""Walk the tree in 'by_dir' order.
158
This will yield each entry in the tree as a (path, entry) tuple.
159
The order that they are yielded is:
161
Directories are walked in a depth-first lexicographical order,
162
however, whenever a directory is reached, all of its direct child
163
nodes are yielded in lexicographical order before yielding the
166
For example, in the tree::
176
The yield order (ignoring root) would be::
177
a, f, a/b, a/d, a/b/c, a/d/e, f/g
179
:param yield_parents: If True, yield the parents from the root leading
180
down to specific_file_ids that have been requested. This has no
181
impact if specific_file_ids is None.
183
raise NotImplementedError(self.iter_entries_by_dir)
185
def list_files(self, include_root=False, from_dir=None, recursive=True):
186
"""List all files in this tree.
188
:param include_root: Whether to include the entry for the tree root
189
:param from_dir: Directory under which to list files
190
:param recursive: Whether to list files recursively
191
:return: iterator over tuples of (path, versioned, kind, file_id,
194
raise NotImplementedError(self.list_files)
172
This will yield each entry in the tree as a (path, entry) tuple. The
173
order that they are yielded is: the contents of a directory are
174
preceeded by the parent of a directory, and all the contents of a
175
directory are grouped together.
177
return self.inventory.iter_entries_by_dir(
178
specific_file_ids=specific_file_ids)
196
180
def iter_references(self):
197
if self.supports_tree_reference():
198
for path, entry in self.iter_entries_by_dir():
199
if entry.kind == 'tree-reference':
200
yield path, entry.file_id
181
for path, entry in self.iter_entries_by_dir():
182
if entry.kind == 'tree-reference':
183
yield path, entry.file_id
202
185
def kind(self, file_id):
203
186
raise NotImplementedError("Tree subclass %s must implement kind"
204
187
% self.__class__.__name__)
206
def stored_kind(self, file_id):
207
"""File kind stored for this file_id.
209
May not match kind on disk for working trees. Always available
210
for versioned files, even when the file itself is missing.
212
return self.kind(file_id)
214
def path_content_summary(self, path):
215
"""Get a summary of the information about path.
217
All the attributes returned are for the canonical form, not the
218
convenient form (if content filters are in use.)
220
:param path: A relative path within the tree.
221
:return: A tuple containing kind, size, exec, sha1-or-link.
222
Kind is always present (see tree.kind()).
223
size is present if kind is file and the size of the
224
canonical form can be cheaply determined, None otherwise.
225
exec is None unless kind is file and the platform supports the 'x'
227
sha1-or-link is the link target if kind is symlink, or the sha1 if
228
it can be obtained without reading the file.
230
raise NotImplementedError(self.path_content_summary)
232
189
def get_reference_revision(self, file_id, path=None):
233
190
raise NotImplementedError("Tree subclass %s must implement "
234
191
"get_reference_revision"
370
235
raise NotImplementedError(self.get_symlink_target)
373
def get_root_id(self):
374
"""Return the file_id for the root of this tree."""
375
raise NotImplementedError(self.get_root_id)
377
def annotate_iter(self, file_id,
378
default_revision=_mod_revision.CURRENT_REVISION):
379
"""Return an iterator of revision_id, line tuples.
237
def annotate_iter(self, file_id):
238
"""Return an iterator of revision_id, line tuples
381
240
For working trees (and mutable trees in general), the special
382
241
revision_id 'current:' will be used for lines that are new in this
383
242
tree, e.g. uncommitted changes.
384
243
:param file_id: The file to produce an annotated version from
385
:param default_revision: For lines that don't match a basis, mark them
386
with this revision id. Not all implementations will make use of
389
245
raise NotImplementedError(self.annotate_iter)
391
def _get_plan_merge_data(self, file_id, other, base):
392
from bzrlib import versionedfile
393
vf = versionedfile._PlanMergeVersionedFile(file_id)
394
last_revision_a = self._get_file_revision(file_id, vf, 'this:')
395
last_revision_b = other._get_file_revision(file_id, vf, 'other:')
397
last_revision_base = None
399
last_revision_base = base._get_file_revision(file_id, vf, 'base:')
400
return vf, last_revision_a, last_revision_b, last_revision_base
402
def plan_file_merge(self, file_id, other, base=None):
403
"""Generate a merge plan based on annotations.
405
If the file contains uncommitted changes in this tree, they will be
406
attributed to the 'current:' pseudo-revision. If the file contains
407
uncommitted changes in the other tree, they will be assigned to the
408
'other:' pseudo-revision.
410
data = self._get_plan_merge_data(file_id, other, base)
411
vf, last_revision_a, last_revision_b, last_revision_base = data
412
return vf.plan_merge(last_revision_a, last_revision_b,
415
def plan_file_lca_merge(self, file_id, other, base=None):
416
"""Generate a merge plan based lca-newness.
418
If the file contains uncommitted changes in this tree, they will be
419
attributed to the 'current:' pseudo-revision. If the file contains
420
uncommitted changes in the other tree, they will be assigned to the
421
'other:' pseudo-revision.
423
data = self._get_plan_merge_data(file_id, other, base)
424
vf, last_revision_a, last_revision_b, last_revision_base = data
425
return vf.plan_lca_merge(last_revision_a, last_revision_b,
428
def _iter_parent_trees(self):
429
"""Iterate through parent trees, defaulting to Tree.revision_tree."""
430
for revision_id in self.get_parent_ids():
432
yield self.revision_tree(revision_id)
433
except errors.NoSuchRevisionInTree:
434
yield self.repository.revision_tree(revision_id)
436
def _get_file_revision(self, file_id, vf, tree_revision):
437
"""Ensure that file_id, tree_revision is in vf to plan the merge."""
439
if getattr(self, '_repository', None) is None:
440
last_revision = tree_revision
441
parent_keys = [(file_id, t.get_file_revision(file_id)) for t in
442
self._iter_parent_trees()]
443
vf.add_lines((file_id, last_revision), parent_keys,
444
self.get_file_lines(file_id))
445
repo = self.branch.repository
448
last_revision = self.get_file_revision(file_id)
449
base_vf = self._repository.texts
450
if base_vf not in vf.fallback_versionedfiles:
451
vf.fallback_versionedfiles.append(base_vf)
247
inventory = property(_get_inventory,
248
doc="Inventory of this Tree")
454
250
def _check_retrieved(self, ie, f):
455
251
if not __debug__:
457
fp = osutils.fingerprint_file(f)
253
fp = fingerprint_file(f)
460
256
if ie.text_size is not None:
461
257
if ie.text_size != fp['size']:
462
raise errors.BzrError(
463
"mismatched size for file %r in %r" %
464
(ie.file_id, self._store),
258
raise BzrError("mismatched size for file %r in %r" % (ie.file_id, self._store),
465
259
["inventory expects %d bytes" % ie.text_size,
466
260
"file is actually %d bytes" % fp['size'],
467
261
"store is probably damaged/corrupt"])
469
263
if ie.text_sha1 != fp['sha1']:
470
raise errors.BzrError("wrong SHA-1 for file %r in %r" %
471
(ie.file_id, self._store),
264
raise BzrError("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
472
265
["inventory expects %s" % ie.text_sha1,
473
266
"file is actually %s" % fp['sha1'],
474
267
"store is probably damaged/corrupt"])
476
270
def path2id(self, path):
477
271
"""Return the id for path in this tree."""
478
raise NotImplementedError(self.path2id)
272
return self._inventory.path2id(path)
480
274
def paths2ids(self, paths, trees=[], require_versioned=True):
481
275
"""Return all the ids that can be reached by walking from paths.
483
Each path is looked up in this tree and any extras provided in
277
Each path is looked up in each this tree and any extras provided in
484
278
trees, and this is repeated recursively: the children in an extra tree
485
279
of a directory that has been renamed under a provided path in this tree
486
are all returned, even if none exist under a provided path in this
280
are all returned, even if none exist until a provided path in this
487
281
tree, and vice versa.
489
283
:param paths: An iterable of paths to start converting to ids from.
574
369
raise NotImplementedError(self.walkdirs)
576
def supports_content_filtering(self):
372
class EmptyTree(Tree):
375
self._inventory = Inventory(root_id=None)
376
symbol_versioning.warn('EmptyTree is deprecated as of bzr 0.9 please'
377
' use repository.revision_tree instead.',
378
DeprecationWarning, stacklevel=2)
380
def get_parent_ids(self):
383
def get_symlink_target(self, file_id):
386
def has_filename(self, filename):
579
def _content_filter_stack(self, path=None, file_id=None):
580
"""The stack of content filters for a path if filtering is supported.
582
Readers will be applied in first-to-last order.
583
Writers will be applied in last-to-first order.
584
Either the path or the file-id needs to be provided.
586
:param path: path relative to the root of the tree
588
:param file_id: file_id or None if unknown
589
:return: the list of filters - [] if there are none
591
filter_pref_names = filters._get_registered_names()
592
if len(filter_pref_names) == 0:
595
path = self.id2path(file_id)
596
prefs = self.iter_search_rules([path], filter_pref_names).next()
597
stk = filters._get_filter_stack_for(prefs)
598
if 'filters' in debug.debug_flags:
599
trace.note("*** %s content-filter: %s => %r" % (path,prefs,stk))
602
def _content_filter_stack_provider(self):
603
"""A function that returns a stack of ContentFilters.
605
The function takes a path (relative to the top of the tree) and a
606
file-id as parameters.
608
:return: None if content filtering is not supported by this tree.
610
if self.supports_content_filtering():
611
return lambda path, file_id: \
612
self._content_filter_stack(path, file_id)
616
def iter_search_rules(self, path_names, pref_names=None,
617
_default_searcher=None):
618
"""Find the preferences for filenames in a tree.
620
:param path_names: an iterable of paths to find attributes for.
621
Paths are given relative to the root of the tree.
622
:param pref_names: the list of preferences to lookup - None for all
623
:param _default_searcher: private parameter to assist testing - don't use
624
:return: an iterator of tuple sequences, one per path-name.
625
See _RulesSearcher.get_items for details on the tuple sequence.
627
if _default_searcher is None:
628
_default_searcher = rules._per_user_searcher
629
searcher = self._get_rules_searcher(_default_searcher)
630
if searcher is not None:
631
if pref_names is not None:
632
for path in path_names:
633
yield searcher.get_selected_items(path, pref_names)
635
for path in path_names:
636
yield searcher.get_items(path)
638
def _get_rules_searcher(self, default_searcher):
639
"""Get the RulesSearcher for this tree given the default one."""
640
searcher = default_searcher
644
class InventoryTree(Tree):
645
"""A tree that relies on an inventory for its metadata.
647
Trees contain an `Inventory` object, and also know how to retrieve
648
file texts mentioned in the inventory, either from a working
649
directory or from a store.
651
It is possible for trees to contain files that are not described
652
in their inventory or vice versa; for this use `filenames()`.
654
Subclasses should set the _inventory attribute, which is considered
655
private to external API users.
658
def get_canonical_inventory_paths(self, paths):
659
"""Like get_canonical_inventory_path() but works on multiple items.
661
:param paths: A sequence of paths relative to the root of the tree.
662
:return: A list of paths, with each item the corresponding input path
663
adjusted to account for existing elements that match case
666
return list(self._yield_canonical_inventory_paths(paths))
668
def get_canonical_inventory_path(self, path):
669
"""Returns the first inventory item that case-insensitively matches path.
671
If a path matches exactly, it is returned. If no path matches exactly
672
but more than one path matches case-insensitively, it is implementation
673
defined which is returned.
675
If no path matches case-insensitively, the input path is returned, but
676
with as many path entries that do exist changed to their canonical
679
If you need to resolve many names from the same tree, you should
680
use get_canonical_inventory_paths() to avoid O(N) behaviour.
682
:param path: A paths relative to the root of the tree.
683
:return: The input path adjusted to account for existing elements
684
that match case insensitively.
686
return self._yield_canonical_inventory_paths([path]).next()
688
def _yield_canonical_inventory_paths(self, paths):
690
# First, if the path as specified exists exactly, just use it.
691
if self.path2id(path) is not None:
695
cur_id = self.get_root_id()
697
bit_iter = iter(path.split("/"))
701
for child in self.iter_children(cur_id):
703
# XXX: it seem like if the child is known to be in the
704
# tree, we shouldn't need to go from its id back to
705
# its path -- mbp 2010-02-11
707
# XXX: it seems like we could be more efficient
708
# by just directly looking up the original name and
709
# only then searching all children; also by not
710
# chopping paths so much. -- mbp 2010-02-11
711
child_base = os.path.basename(self.id2path(child))
712
if (child_base == elt):
713
# if we found an exact match, we can stop now; if
714
# we found an approximate match we need to keep
715
# searching because there might be an exact match
718
new_path = osutils.pathjoin(cur_path, child_base)
720
elif child_base.lower() == lelt:
722
new_path = osutils.pathjoin(cur_path, child_base)
723
except errors.NoSuchId:
724
# before a change is committed we can see this error...
729
# got to the end of this directory and no entries matched.
730
# Return what matched so far, plus the rest as specified.
731
cur_path = osutils.pathjoin(cur_path, elt, *list(bit_iter))
736
def _get_inventory(self):
737
return self._inventory
739
inventory = property(_get_inventory,
740
doc="Inventory of this Tree")
743
def path2id(self, path):
744
"""Return the id for path in this tree."""
745
return self._inventory.path2id(path)
747
def id2path(self, file_id):
748
"""Return the path for a file id.
752
return self.inventory.id2path(file_id)
754
def has_id(self, file_id):
755
return self.inventory.has_id(file_id)
757
def has_or_had_id(self, file_id):
758
return self.inventory.has_id(file_id)
761
return iter(self.inventory)
763
def filter_unversioned_files(self, paths):
764
"""Filter out paths that are versioned.
766
:return: set of paths.
768
# NB: we specifically *don't* call self.has_filename, because for
769
# WorkingTrees that can indicate files that exist on disk but that
771
pred = self.inventory.has_filename
772
return set((p for p in paths if not pred(p)))
775
def iter_entries_by_dir(self, specific_file_ids=None, yield_parents=False):
776
"""Walk the tree in 'by_dir' order.
778
This will yield each entry in the tree as a (path, entry) tuple.
779
The order that they are yielded is:
781
See Tree.iter_entries_by_dir for details.
783
:param yield_parents: If True, yield the parents from the root leading
784
down to specific_file_ids that have been requested. This has no
785
impact if specific_file_ids is None.
787
return self.inventory.iter_entries_by_dir(
788
specific_file_ids=specific_file_ids, yield_parents=yield_parents)
790
def get_file_by_path(self, path):
791
return self.get_file(self._inventory.path2id(path), path)
389
def kind(self, file_id):
390
file_id = osutils.safe_file_id(file_id)
391
assert self._inventory[file_id].kind == "directory"
394
def list_files(self, include_root=False):
397
def __contains__(self, file_id):
398
file_id = osutils.safe_file_id(file_id)
399
return (file_id in self._inventory)
401
def get_file_sha1(self, file_id, path=None, stat_value=None):
794
405
######################################################################
923
547
Its instances have methods like 'compare' and contain references to the
924
548
source and target trees these operations are to be carried out on.
926
Clients of bzrlib should not need to use InterTree directly, rather they
550
clients of bzrlib should not need to use InterTree directly, rather they
927
551
should use the convenience methods on Tree such as 'Tree.compare()' which
928
552
will pass through to InterTree as appropriate.
931
# Formats that will be used to test this InterTree. If both are
932
# None, this InterTree will not be tested (e.g. because a complex
934
_matching_from_tree_format = None
935
_matching_to_tree_format = None
939
def _changes_from_entries(self, source_entry, target_entry,
940
source_path=None, target_path=None):
941
"""Generate a iter_changes tuple between source_entry and target_entry.
943
:param source_entry: An inventory entry from self.source, or None.
944
:param target_entry: An inventory entry from self.target, or None.
945
:param source_path: The path of source_entry, if known. If not known
946
it will be looked up.
947
:param target_path: The path of target_entry, if known. If not known
948
it will be looked up.
949
:return: A tuple, item 0 of which is an iter_changes result tuple, and
950
item 1 is True if there are any changes in the result tuple.
952
if source_entry is None:
953
if target_entry is None:
955
file_id = target_entry.file_id
957
file_id = source_entry.file_id
958
if source_entry is not None:
959
source_versioned = True
960
source_name = source_entry.name
961
source_parent = source_entry.parent_id
962
if source_path is None:
963
source_path = self.source.id2path(file_id)
964
source_kind, source_executable, source_stat = \
965
self.source._comparison_data(source_entry, source_path)
967
source_versioned = False
971
source_executable = None
972
if target_entry is not None:
973
target_versioned = True
974
target_name = target_entry.name
975
target_parent = target_entry.parent_id
976
if target_path is None:
977
target_path = self.target.id2path(file_id)
978
target_kind, target_executable, target_stat = \
979
self.target._comparison_data(target_entry, target_path)
981
target_versioned = False
985
target_executable = None
986
versioned = (source_versioned, target_versioned)
987
kind = (source_kind, target_kind)
988
changed_content = False
989
if source_kind != target_kind:
990
changed_content = True
991
elif source_kind == 'file':
992
if (self.source.get_file_sha1(file_id, source_path, source_stat) !=
993
self.target.get_file_sha1(file_id, target_path, target_stat)):
994
changed_content = True
995
elif source_kind == 'symlink':
996
if (self.source.get_symlink_target(file_id) !=
997
self.target.get_symlink_target(file_id)):
998
changed_content = True
999
# XXX: Yes, the indentation below is wrong. But fixing it broke
1000
# test_merge.TestMergerEntriesLCAOnDisk.
1001
# test_nested_tree_subtree_renamed_and_modified. We'll wait for
1002
# the fix from bzr.dev -- vila 2009026
1003
elif source_kind == 'tree-reference':
1004
if (self.source.get_reference_revision(file_id, source_path)
1005
!= self.target.get_reference_revision(file_id, target_path)):
1006
changed_content = True
1007
parent = (source_parent, target_parent)
1008
name = (source_name, target_name)
1009
executable = (source_executable, target_executable)
1010
if (changed_content is not False or versioned[0] != versioned[1]
1011
or parent[0] != parent[1] or name[0] != name[1] or
1012
executable[0] != executable[1]):
1016
return (file_id, (source_path, target_path), changed_content,
1017
versioned, parent, name, kind, executable), changes
1019
557
@needs_read_lock
1020
558
def compare(self, want_unchanged=False, specific_files=None,
1021
559
extra_trees=None, require_versioned=False, include_root=False,
1082
621
:param require_versioned: Raise errors.PathsNotVersionedError if a
1083
622
path in the specific_files list is not versioned in one of
1084
623
source, target or extra_trees.
1085
:param specific_files: An optional list of file paths to restrict the
1086
comparison to. When mapping filenames to ids, all matches in all
1087
trees (including optional extra_trees) are used, and all children
1088
of matched directories are included. The parents in the target tree
1089
of the specific files up to and including the root of the tree are
1090
always evaluated for changes too.
1091
624
:param want_unversioned: Should unversioned files be returned in the
1092
625
output. An unversioned file is defined as one with (False, False)
1093
626
for the versioned pair.
1095
629
lookup_trees = [self.source]
1097
631
lookup_trees.extend(extra_trees)
1098
# The ids of items we need to examine to insure delta consistency.
1099
precise_file_ids = set()
1100
changed_file_ids = []
1101
if specific_files == []:
1102
specific_file_ids = []
1104
specific_file_ids = self.target.paths2ids(specific_files,
1105
lookup_trees, require_versioned=require_versioned)
1106
if specific_files is not None:
1107
# reparented or added entries must have their parents included
1108
# so that valid deltas can be created. The seen_parents set
1109
# tracks the parents that we need to have.
1110
# The seen_dirs set tracks directory entries we've yielded.
1111
# After outputting version object in to_entries we set difference
1112
# the two seen sets and start checking parents.
1113
seen_parents = set()
632
specific_file_ids = self.target.paths2ids(specific_files,
633
lookup_trees, require_versioned=require_versioned)
1115
634
if want_unversioned:
1116
all_unversioned = sorted([(p.split('/'), p) for p in
1117
self.target.extras()
1118
if specific_files is None or
635
all_unversioned = sorted([(p.split('/'), p) for p in self.target.extras()
636
if not specific_files or
1119
637
osutils.is_inside_any(specific_files, p)])
1120
all_unversioned = collections.deque(all_unversioned)
638
all_unversioned = deque(all_unversioned)
1122
all_unversioned = collections.deque()
640
all_unversioned = deque()
1124
from_entries_by_dir = list(self.source.iter_entries_by_dir(
642
from_entries_by_dir = list(self.source.inventory.iter_entries_by_dir(
1125
643
specific_file_ids=specific_file_ids))
1126
644
from_data = dict((e.file_id, (p, e)) for p, e in from_entries_by_dir)
1127
to_entries_by_dir = list(self.target.iter_entries_by_dir(
645
to_entries_by_dir = list(self.target.inventory.iter_entries_by_dir(
1128
646
specific_file_ids=specific_file_ids))
1129
647
num_entries = len(from_entries_by_dir) + len(to_entries_by_dir)
1131
# the unversioned path lookup only occurs on real trees - where there
649
# the unversioned path lookup only occurs on real trees - where there
1132
650
# can be extras. So the fake_entry is solely used to look up
1133
651
# executable it values when execute is not supported.
1134
fake_entry = inventory.InventoryFile('unused', 'unused', 'unused')
1135
for target_path, target_entry in to_entries_by_dir:
1136
while (all_unversioned and
1137
all_unversioned[0][0] < target_path.split('/')):
652
fake_entry = InventoryFile('unused', 'unused', 'unused')
653
for to_path, to_entry in to_entries_by_dir:
654
while all_unversioned and all_unversioned[0][0] < to_path.split('/'):
1138
655
unversioned_path = all_unversioned.popleft()
1139
target_kind, target_executable, target_stat = \
656
to_kind, to_executable, to_stat = \
1140
657
self.target._comparison_data(fake_entry, unversioned_path[1])
1141
658
yield (None, (None, unversioned_path[1]), True, (False, False),
1143
660
(None, unversioned_path[0][-1]),
1144
(None, target_kind),
1145
(None, target_executable))
1146
source_path, source_entry = from_data.get(target_entry.file_id,
1148
result, changes = self._changes_from_entries(source_entry,
1149
target_entry, source_path=source_path, target_path=target_path)
1150
to_paths[result[0]] = result[1][1]
662
(None, to_executable))
663
file_id = to_entry.file_id
664
to_paths[file_id] = to_path
1151
665
entry_count += 1
666
changed_content = False
667
from_path, from_entry = from_data.get(file_id, (None, None))
668
from_versioned = (from_entry is not None)
669
if from_entry is not None:
670
from_versioned = True
671
from_name = from_entry.name
672
from_parent = from_entry.parent_id
673
from_kind, from_executable, from_stat = \
674
self.source._comparison_data(from_entry, from_path)
1153
675
entry_count += 1
677
from_versioned = False
681
from_executable = None
682
versioned = (from_versioned, True)
683
to_kind, to_executable, to_stat = \
684
self.target._comparison_data(to_entry, to_path)
685
kind = (from_kind, to_kind)
686
if kind[0] != kind[1]:
687
changed_content = True
688
elif from_kind == 'file':
689
from_size = self.source._file_size(from_entry, from_stat)
690
to_size = self.target._file_size(to_entry, to_stat)
691
if from_size != to_size:
692
changed_content = True
693
elif (self.source.get_file_sha1(file_id, from_path, from_stat) !=
694
self.target.get_file_sha1(file_id, to_path, to_stat)):
695
changed_content = True
696
elif from_kind == 'symlink':
697
if (self.source.get_symlink_target(file_id) !=
698
self.target.get_symlink_target(file_id)):
699
changed_content = True
700
elif from_kind == 'tree-reference':
701
if (self.source.get_reference_revision(file_id, from_path)
702
!= self.target.get_reference_revision(file_id, to_path)):
703
changed_content = True
704
parent = (from_parent, to_entry.parent_id)
705
name = (from_name, to_entry.name)
706
executable = (from_executable, to_executable)
1154
707
if pb is not None:
1155
708
pb.update('comparing files', entry_count, num_entries)
1156
if changes or include_unchanged:
1157
if specific_file_ids is not None:
1158
new_parent_id = result[4][1]
1159
precise_file_ids.add(new_parent_id)
1160
changed_file_ids.append(result[0])
1162
# Ensure correct behaviour for reparented/added specific files.
1163
if specific_files is not None:
1164
# Record output dirs
1165
if result[6][1] == 'directory':
1166
seen_dirs.add(result[0])
1167
# Record parents of reparented/added entries.
1168
versioned = result[3]
1170
if not versioned[0] or parents[0] != parents[1]:
1171
seen_parents.add(parents[1])
709
if (changed_content is not False or versioned[0] != versioned[1]
710
or parent[0] != parent[1] or name[0] != name[1] or
711
executable[0] != executable[1] or include_unchanged):
712
yield (file_id, (from_path, to_path), changed_content,
713
versioned, parent, name, kind, executable)
1172
715
while all_unversioned:
1173
716
# yield any trailing unversioned paths
1174
717
unversioned_path = all_unversioned.popleft()
1201
756
self.source._comparison_data(from_entry, path)
1202
757
kind = (from_kind, None)
1203
758
executable = (from_executable, None)
1204
changed_content = from_kind is not None
759
changed_content = True
1205
760
# the parent's path is necessarily known at this point.
1206
changed_file_ids.append(file_id)
1207
761
yield(file_id, (path, to_path), changed_content, versioned, parent,
1208
762
name, kind, executable)
1209
changed_file_ids = set(changed_file_ids)
1210
if specific_file_ids is not None:
1211
for result in self._handle_precise_ids(precise_file_ids,
1215
def _get_entry(self, tree, file_id):
1216
"""Get an inventory entry from a tree, with missing entries as None.
1218
If the tree raises NotImplementedError on accessing .inventory, then
1219
this is worked around using iter_entries_by_dir on just the file id
1222
:param tree: The tree to lookup the entry in.
1223
:param file_id: The file_id to lookup.
1226
inventory = tree.inventory
1227
except NotImplementedError:
1228
# No inventory available.
1230
iterator = tree.iter_entries_by_dir(specific_file_ids=[file_id])
1231
return iterator.next()[1]
1232
except StopIteration:
1236
return inventory[file_id]
1237
except errors.NoSuchId:
1240
def _handle_precise_ids(self, precise_file_ids, changed_file_ids,
1241
discarded_changes=None):
1242
"""Fill out a partial iter_changes to be consistent.
1244
:param precise_file_ids: The file ids of parents that were seen during
1246
:param changed_file_ids: The file ids of already emitted items.
1247
:param discarded_changes: An optional dict of precalculated
1248
iter_changes items which the partial iter_changes had not output
1250
:return: A generator of iter_changes items to output.
1252
# process parents of things that had changed under the users
1253
# requested paths to prevent incorrect paths or parent ids which
1254
# aren't in the tree.
1255
while precise_file_ids:
1256
precise_file_ids.discard(None)
1257
# Don't emit file_ids twice
1258
precise_file_ids.difference_update(changed_file_ids)
1259
if not precise_file_ids:
1261
# If the there was something at a given output path in source, we
1262
# have to include the entry from source in the delta, or we would
1263
# be putting this entry into a used path.
1265
for parent_id in precise_file_ids:
1267
paths.append(self.target.id2path(parent_id))
1268
except errors.NoSuchId:
1269
# This id has been dragged in from the source by delta
1270
# expansion and isn't present in target at all: we don't
1271
# need to check for path collisions on it.
1274
old_id = self.source.path2id(path)
1275
precise_file_ids.add(old_id)
1276
precise_file_ids.discard(None)
1277
current_ids = precise_file_ids
1278
precise_file_ids = set()
1279
# We have to emit all of precise_file_ids that have been altered.
1280
# We may have to output the children of some of those ids if any
1281
# directories have stopped being directories.
1282
for file_id in current_ids:
1284
if discarded_changes:
1285
result = discarded_changes.get(file_id)
1290
old_entry = self._get_entry(self.source, file_id)
1291
new_entry = self._get_entry(self.target, file_id)
1292
result, changes = self._changes_from_entries(
1293
old_entry, new_entry)
1296
# Get this parents parent to examine.
1297
new_parent_id = result[4][1]
1298
precise_file_ids.add(new_parent_id)
1300
if (result[6][0] == 'directory' and
1301
result[6][1] != 'directory'):
1302
# This stopped being a directory, the old children have
1304
if old_entry is None:
1305
# Reusing a discarded change.
1306
old_entry = self._get_entry(self.source, file_id)
1307
for child in old_entry.children.values():
1308
precise_file_ids.add(child.file_id)
1309
changed_file_ids.add(result[0])
1313
class MultiWalker(object):
1314
"""Walk multiple trees simultaneously, getting combined results."""
1316
# Note: This could be written to not assume you can do out-of-order
1317
# lookups. Instead any nodes that don't match in all trees could be
1318
# marked as 'deferred', and then returned in the final cleanup loop.
1319
# For now, I think it is "nicer" to return things as close to the
1320
# "master_tree" order as we can.
1322
def __init__(self, master_tree, other_trees):
1323
"""Create a new MultiWalker.
1325
All trees being walked must implement "iter_entries_by_dir()", such
1326
that they yield (path, object) tuples, where that object will have a
1327
'.file_id' member, that can be used to check equality.
1329
:param master_tree: All trees will be 'slaved' to the master_tree such
1330
that nodes in master_tree will be used as 'first-pass' sync points.
1331
Any nodes that aren't in master_tree will be merged in a second
1333
:param other_trees: A list of other trees to walk simultaneously.
1335
self._master_tree = master_tree
1336
self._other_trees = other_trees
1338
# Keep track of any nodes that were properly processed just out of
1339
# order, that way we don't return them at the end, we don't have to
1340
# track *all* processed file_ids, just the out-of-order ones
1341
self._out_of_order_processed = set()
1344
def _step_one(iterator):
1345
"""Step an iter_entries_by_dir iterator.
1347
:return: (has_more, path, ie)
1348
If has_more is False, path and ie will be None.
1351
path, ie = iterator.next()
1352
except StopIteration:
1353
return False, None, None
1355
return True, path, ie
1358
def _cmp_path_by_dirblock(path1, path2):
1359
"""Compare two paths based on what directory they are in.
1361
This generates a sort order, such that all children of a directory are
1362
sorted together, and grandchildren are in the same order as the
1363
children appear. But all grandchildren come after all children.
1365
:param path1: first path
1366
:param path2: the second path
1367
:return: negative number if ``path1`` comes first,
1368
0 if paths are equal
1369
and a positive number if ``path2`` sorts first
1371
# Shortcut this special case
1374
# This is stolen from _dirstate_helpers_py.py, only switching it to
1375
# Unicode objects. Consider using encode_utf8() and then using the
1376
# optimized versions, or maybe writing optimized unicode versions.
1377
if not isinstance(path1, unicode):
1378
raise TypeError("'path1' must be a unicode string, not %s: %r"
1379
% (type(path1), path1))
1380
if not isinstance(path2, unicode):
1381
raise TypeError("'path2' must be a unicode string, not %s: %r"
1382
% (type(path2), path2))
1383
return cmp(MultiWalker._path_to_key(path1),
1384
MultiWalker._path_to_key(path2))
1387
def _path_to_key(path):
1388
dirname, basename = osutils.split(path)
1389
return (dirname.split(u'/'), basename)
1391
def _lookup_by_file_id(self, extra_entries, other_tree, file_id):
1392
"""Lookup an inventory entry by file_id.
1394
This is called when an entry is missing in the normal order.
1395
Generally this is because a file was either renamed, or it was
1396
deleted/added. If the entry was found in the inventory and not in
1397
extra_entries, it will be added to self._out_of_order_processed
1399
:param extra_entries: A dictionary of {file_id: (path, ie)}. This
1400
should be filled with entries that were found before they were
1401
used. If file_id is present, it will be removed from the
1403
:param other_tree: The Tree to search, in case we didn't find the entry
1405
:param file_id: The file_id to look for
1406
:return: (path, ie) if found or (None, None) if not present.
1408
if file_id in extra_entries:
1409
return extra_entries.pop(file_id)
1410
# TODO: Is id2path better as the first call, or is
1411
# inventory[file_id] better as a first check?
1413
cur_path = other_tree.id2path(file_id)
1414
except errors.NoSuchId:
1416
if cur_path is None:
1419
self._out_of_order_processed.add(file_id)
1420
cur_ie = other_tree.inventory[file_id]
1421
return (cur_path, cur_ie)
1424
"""Match up the values in the different trees."""
1425
for result in self._walk_master_tree():
1427
self._finish_others()
1428
for result in self._walk_others():
1431
def _walk_master_tree(self):
1432
"""First pass, walk all trees in lock-step.
1434
When we are done, all nodes in the master_tree will have been
1435
processed. _other_walkers, _other_entries, and _others_extra will be
1436
set on 'self' for future processing.
1438
# This iterator has the most "inlining" done, because it tends to touch
1439
# every file in the tree, while the others only hit nodes that don't
1441
master_iterator = self._master_tree.iter_entries_by_dir()
1443
other_walkers = [other.iter_entries_by_dir()
1444
for other in self._other_trees]
1445
other_entries = [self._step_one(walker) for walker in other_walkers]
1446
# Track extra nodes in the other trees
1447
others_extra = [{} for i in xrange(len(self._other_trees))]
1449
master_has_more = True
1450
step_one = self._step_one
1451
lookup_by_file_id = self._lookup_by_file_id
1452
out_of_order_processed = self._out_of_order_processed
1454
while master_has_more:
1455
(master_has_more, path, master_ie) = step_one(master_iterator)
1456
if not master_has_more:
1459
file_id = master_ie.file_id
1461
other_values_append = other_values.append
1462
next_other_entries = []
1463
next_other_entries_append = next_other_entries.append
1464
for idx, (other_has_more, other_path, other_ie) in enumerate(other_entries):
1465
if not other_has_more:
1466
other_values_append(lookup_by_file_id(
1467
others_extra[idx], self._other_trees[idx], file_id))
1468
next_other_entries_append((False, None, None))
1469
elif file_id == other_ie.file_id:
1470
# This is the critical code path, as most of the entries
1471
# should match between most trees.
1472
other_values_append((other_path, other_ie))
1473
next_other_entries_append(step_one(other_walkers[idx]))
1475
# This walker did not match, step it until it either
1476
# matches, or we know we are past the current walker.
1477
other_walker = other_walkers[idx]
1478
other_extra = others_extra[idx]
1479
while (other_has_more and
1480
self._cmp_path_by_dirblock(other_path, path) < 0):
1481
other_file_id = other_ie.file_id
1482
if other_file_id not in out_of_order_processed:
1483
other_extra[other_file_id] = (other_path, other_ie)
1484
other_has_more, other_path, other_ie = \
1485
step_one(other_walker)
1486
if other_has_more and other_ie.file_id == file_id:
1487
# We ended up walking to this point, match and step
1489
other_values_append((other_path, other_ie))
1490
other_has_more, other_path, other_ie = \
1491
step_one(other_walker)
1493
# This record isn't in the normal order, see if it
1495
other_values_append(lookup_by_file_id(
1496
other_extra, self._other_trees[idx], file_id))
1497
next_other_entries_append((other_has_more, other_path,
1499
other_entries = next_other_entries
1501
# We've matched all the walkers, yield this datapoint
1502
yield path, file_id, master_ie, other_values
1503
self._other_walkers = other_walkers
1504
self._other_entries = other_entries
1505
self._others_extra = others_extra
1507
def _finish_others(self):
1508
"""Finish walking the other iterators, so we get all entries."""
1509
for idx, info in enumerate(self._other_entries):
1510
other_extra = self._others_extra[idx]
1511
(other_has_more, other_path, other_ie) = info
1512
while other_has_more:
1513
other_file_id = other_ie.file_id
1514
if other_file_id not in self._out_of_order_processed:
1515
other_extra[other_file_id] = (other_path, other_ie)
1516
other_has_more, other_path, other_ie = \
1517
self._step_one(self._other_walkers[idx])
1518
del self._other_entries
1520
def _walk_others(self):
1521
"""Finish up by walking all the 'deferred' nodes."""
1522
# TODO: One alternative would be to grab all possible unprocessed
1523
# file_ids, and then sort by path, and then yield them. That
1524
# might ensure better ordering, in case a caller strictly
1525
# requires parents before children.
1526
for idx, other_extra in enumerate(self._others_extra):
1527
others = sorted(other_extra.itervalues(),
1528
key=lambda x: self._path_to_key(x[0]))
1529
for other_path, other_ie in others:
1530
file_id = other_ie.file_id
1531
# We don't need to check out_of_order_processed here, because
1532
# the lookup_by_file_id will be removing anything processed
1533
# from the extras cache
1534
other_extra.pop(file_id)
1535
other_values = [(None, None) for i in xrange(idx)]
1536
other_values.append((other_path, other_ie))
1537
for alt_idx, alt_extra in enumerate(self._others_extra[idx+1:]):
1538
alt_idx = alt_idx + idx + 1
1539
alt_extra = self._others_extra[alt_idx]
1540
alt_tree = self._other_trees[alt_idx]
1541
other_values.append(self._lookup_by_file_id(
1542
alt_extra, alt_tree, file_id))
1543
yield other_path, file_id, None, other_values
765
# This was deprecated before 0.12, but did not have an official warning
766
@symbol_versioning.deprecated_function(symbol_versioning.zero_twelve)
767
def RevisionTree(*args, **kwargs):
768
"""RevisionTree has moved to bzrlib.revisiontree.RevisionTree()
770
Accessing it as bzrlib.tree.RevisionTree has been deprecated as of
773
from bzrlib.revisiontree import RevisionTree as _RevisionTree
774
return _RevisionTree(*args, **kwargs)
added
removed
bzrlib/tree.py
