145
def all_file_ids(self):
146
"""Iterate through all file ids, including ids for missing files."""
147
raise NotImplementedError(self.all_file_ids)
147
return iter(self.inventory)
149
149
def id2path(self, file_id):
150
150
"""Return the path for a file id.
152
152
:raises NoSuchId:
154
raise NotImplementedError(self.id2path)
156
def iter_entries_by_dir(self, specific_file_ids=None, yield_parents=False):
154
file_id = osutils.safe_file_id(file_id)
155
return self.inventory.id2path(file_id)
157
def is_control_filename(self, filename):
158
"""True if filename is the name of a control file in this tree.
160
:param filename: A filename within the tree. This is a relative path
161
from the root of this tree.
163
This is true IF and ONLY IF the filename is part of the meta data
164
that bzr controls in this tree. I.E. a random .bzr directory placed
165
on disk will not be a control file for this tree.
167
return self.bzrdir.is_control_filename(filename)
170
def iter_entries_by_dir(self, specific_file_ids=None):
157
171
"""Walk the tree in 'by_dir' order.
159
This will yield each entry in the tree as a (path, entry) tuple.
160
The order that they are yielded is:
162
Directories are walked in a depth-first lexicographical order,
163
however, whenever a directory is reached, all of its direct child
164
nodes are yielded in lexicographical order before yielding the
167
For example, in the tree::
177
The yield order (ignoring root) would be::
179
a, f, a/b, a/d, a/b/c, a/d/e, f/g
181
:param yield_parents: If True, yield the parents from the root leading
182
down to specific_file_ids that have been requested. This has no
183
impact if specific_file_ids is None.
185
raise NotImplementedError(self.iter_entries_by_dir)
187
def list_files(self, include_root=False, from_dir=None, recursive=True):
188
"""List all files in this tree.
190
:param include_root: Whether to include the entry for the tree root
191
:param from_dir: Directory under which to list files
192
:param recursive: Whether to list files recursively
193
:return: iterator over tuples of (path, versioned, kind, file_id,
196
raise NotImplementedError(self.list_files)
173
This will yield each entry in the tree as a (path, entry) tuple. The
174
order that they are yielded is: the contents of a directory are
175
preceeded by the parent of a directory, and all the contents of a
176
directory are grouped together.
178
return self.inventory.iter_entries_by_dir(
179
specific_file_ids=specific_file_ids)
198
181
def iter_references(self):
199
if self.supports_tree_reference():
200
for path, entry in self.iter_entries_by_dir():
201
if entry.kind == 'tree-reference':
202
yield path, entry.file_id
182
for path, entry in self.iter_entries_by_dir():
183
if entry.kind == 'tree-reference':
184
yield path, entry.file_id
204
186
def kind(self, file_id):
205
187
raise NotImplementedError("Tree subclass %s must implement kind"
206
188
% self.__class__.__name__)
208
def stored_kind(self, file_id):
209
"""File kind stored for this file_id.
211
May not match kind on disk for working trees. Always available
212
for versioned files, even when the file itself is missing.
214
return self.kind(file_id)
216
def path_content_summary(self, path):
217
"""Get a summary of the information about path.
219
All the attributes returned are for the canonical form, not the
220
convenient form (if content filters are in use.)
222
:param path: A relative path within the tree.
223
:return: A tuple containing kind, size, exec, sha1-or-link.
224
Kind is always present (see tree.kind()).
225
size is present if kind is file and the size of the
226
canonical form can be cheaply determined, None otherwise.
227
exec is None unless kind is file and the platform supports the 'x'
229
sha1-or-link is the link target if kind is symlink, or the sha1 if
230
it can be obtained without reading the file.
232
raise NotImplementedError(self.path_content_summary)
234
190
def get_reference_revision(self, file_id, path=None):
235
191
raise NotImplementedError("Tree subclass %s must implement "
236
192
"get_reference_revision"
317
223
raise NotImplementedError(self.get_file_mtime)
319
def get_file_size(self, file_id):
320
"""Return the size of a file in bytes.
322
This applies only to regular files. If invoked on directories or
323
symlinks, it will return None.
324
:param file_id: The file-id of the file
326
raise NotImplementedError(self.get_file_size)
328
225
def get_file_by_path(self, path):
329
raise NotImplementedError(self.get_file_by_path)
331
def is_executable(self, file_id, path=None):
332
"""Check if a file is executable.
334
:param file_id: The handle for this file.
335
:param path: The path that this file can be found at.
336
These must point to the same object.
338
raise NotImplementedError(self.is_executable)
340
def iter_files_bytes(self, desired_files):
341
"""Iterate through file contents.
343
Files will not necessarily be returned in the order they occur in
344
desired_files. No specific order is guaranteed.
346
Yields pairs of identifier, bytes_iterator. identifier is an opaque
347
value supplied by the caller as part of desired_files. It should
348
uniquely identify the file version in the caller's context. (Examples:
349
an index number or a TreeTransform trans_id.)
351
bytes_iterator is an iterable of bytestrings for the file. The
352
kind of iterable and length of the bytestrings are unspecified, but for
353
this implementation, it is a tuple containing a single bytestring with
354
the complete text of the file.
356
:param desired_files: a list of (file_id, identifier) pairs
358
for file_id, identifier in desired_files:
359
# We wrap the string in a tuple so that we can return an iterable
360
# of bytestrings. (Technically, a bytestring is also an iterable
361
# of bytestrings, but iterating through each character is not
363
cur_file = (self.get_file_text(file_id),)
364
yield identifier, cur_file
366
def get_symlink_target(self, file_id, path=None):
226
return self.get_file(self._inventory.path2id(path))
228
def get_symlink_target(self, file_id):
367
229
"""Get the target for a given file_id.
369
231
It is assumed that the caller already knows that file_id is referencing
371
233
:param file_id: Handle for the symlink entry.
372
:param path: The path of the file.
373
If both file_id and path are supplied, an implementation may use
375
234
:return: The path the symlink points to.
377
236
raise NotImplementedError(self.get_symlink_target)
379
def get_root_id(self):
380
"""Return the file_id for the root of this tree."""
381
raise NotImplementedError(self.get_root_id)
383
def annotate_iter(self, file_id,
384
default_revision=_mod_revision.CURRENT_REVISION):
385
"""Return an iterator of revision_id, line tuples.
238
def annotate_iter(self, file_id):
239
"""Return an iterator of revision_id, line tuples
387
241
For working trees (and mutable trees in general), the special
388
242
revision_id 'current:' will be used for lines that are new in this
389
243
tree, e.g. uncommitted changes.
390
244
:param file_id: The file to produce an annotated version from
391
:param default_revision: For lines that don't match a basis, mark them
392
with this revision id. Not all implementations will make use of
395
246
raise NotImplementedError(self.annotate_iter)
397
def _get_plan_merge_data(self, file_id, other, base):
398
from bzrlib import versionedfile
399
vf = versionedfile._PlanMergeVersionedFile(file_id)
400
last_revision_a = self._get_file_revision(file_id, vf, 'this:')
401
last_revision_b = other._get_file_revision(file_id, vf, 'other:')
403
last_revision_base = None
405
last_revision_base = base._get_file_revision(file_id, vf, 'base:')
406
return vf, last_revision_a, last_revision_b, last_revision_base
408
def plan_file_merge(self, file_id, other, base=None):
409
"""Generate a merge plan based on annotations.
411
If the file contains uncommitted changes in this tree, they will be
412
attributed to the 'current:' pseudo-revision. If the file contains
413
uncommitted changes in the other tree, they will be assigned to the
414
'other:' pseudo-revision.
416
data = self._get_plan_merge_data(file_id, other, base)
417
vf, last_revision_a, last_revision_b, last_revision_base = data
418
return vf.plan_merge(last_revision_a, last_revision_b,
421
def plan_file_lca_merge(self, file_id, other, base=None):
422
"""Generate a merge plan based lca-newness.
424
If the file contains uncommitted changes in this tree, they will be
425
attributed to the 'current:' pseudo-revision. If the file contains
426
uncommitted changes in the other tree, they will be assigned to the
427
'other:' pseudo-revision.
429
data = self._get_plan_merge_data(file_id, other, base)
430
vf, last_revision_a, last_revision_b, last_revision_base = data
431
return vf.plan_lca_merge(last_revision_a, last_revision_b,
434
def _iter_parent_trees(self):
435
"""Iterate through parent trees, defaulting to Tree.revision_tree."""
436
for revision_id in self.get_parent_ids():
438
yield self.revision_tree(revision_id)
439
except errors.NoSuchRevisionInTree:
440
yield self.repository.revision_tree(revision_id)
442
def _get_file_revision(self, file_id, vf, tree_revision):
443
"""Ensure that file_id, tree_revision is in vf to plan the merge."""
445
if getattr(self, '_repository', None) is None:
446
last_revision = tree_revision
447
parent_keys = [(file_id, t.get_file_revision(file_id)) for t in
448
self._iter_parent_trees()]
449
vf.add_lines((file_id, last_revision), parent_keys,
450
self.get_file_lines(file_id))
451
repo = self.branch.repository
454
last_revision = self.get_file_revision(file_id)
455
base_vf = self._repository.texts
456
if base_vf not in vf.fallback_versionedfiles:
457
vf.fallback_versionedfiles.append(base_vf)
248
def plan_file_merge(self, file_id, other):
249
"""Generate a merge plan based on annotations
251
If the file contains uncommitted changes in this tree, they will be
252
attributed to the 'current:' pseudo-revision. If the file contains
253
uncommitted changes in the other tree, they will be assigned to the
254
'other:' pseudo-revision.
256
from bzrlib import merge
257
annotated_a = list(self.annotate_iter(file_id,
258
_mod_revision.CURRENT_REVISION))
259
annotated_b = list(other.annotate_iter(file_id, 'other:'))
260
ancestors_a = self._get_ancestors(_mod_revision.CURRENT_REVISION)
261
ancestors_b = other._get_ancestors('other:')
262
return merge._plan_annotate_merge(annotated_a, annotated_b,
263
ancestors_a, ancestors_b)
265
inventory = property(_get_inventory,
266
doc="Inventory of this Tree")
460
268
def _check_retrieved(self, ie, f):
461
269
if not __debug__:
463
fp = osutils.fingerprint_file(f)
271
fp = fingerprint_file(f)
466
274
if ie.text_size is not None:
467
275
if ie.text_size != fp['size']:
468
raise errors.BzrError(
469
"mismatched size for file %r in %r" %
470
(ie.file_id, self._store),
276
raise BzrError("mismatched size for file %r in %r" % (ie.file_id, self._store),
471
277
["inventory expects %d bytes" % ie.text_size,
472
278
"file is actually %d bytes" % fp['size'],
473
279
"store is probably damaged/corrupt"])
475
281
if ie.text_sha1 != fp['sha1']:
476
raise errors.BzrError("wrong SHA-1 for file %r in %r" %
477
(ie.file_id, self._store),
282
raise BzrError("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
478
283
["inventory expects %s" % ie.text_sha1,
479
284
"file is actually %s" % fp['sha1'],
480
285
"store is probably damaged/corrupt"])
482
288
def path2id(self, path):
483
289
"""Return the id for path in this tree."""
484
raise NotImplementedError(self.path2id)
290
return self._inventory.path2id(path)
486
292
def paths2ids(self, paths, trees=[], require_versioned=True):
487
293
"""Return all the ids that can be reached by walking from paths.
489
Each path is looked up in this tree and any extras provided in
295
Each path is looked up in each this tree and any extras provided in
490
296
trees, and this is repeated recursively: the children in an extra tree
491
297
of a directory that has been renamed under a provided path in this tree
492
are all returned, even if none exist under a provided path in this
298
are all returned, even if none exist until a provided path in this
493
299
tree, and vice versa.
495
301
:param paths: An iterable of paths to start converting to ids from.
580
387
raise NotImplementedError(self.walkdirs)
582
def supports_content_filtering(self):
390
class EmptyTree(Tree):
393
self._inventory = Inventory(root_id=None)
394
symbol_versioning.warn('EmptyTree is deprecated as of bzr 0.9 please'
395
' use repository.revision_tree instead.',
396
DeprecationWarning, stacklevel=2)
398
def get_parent_ids(self):
401
def get_symlink_target(self, file_id):
404
def has_filename(self, filename):
585
def _content_filter_stack(self, path=None, file_id=None):
586
"""The stack of content filters for a path if filtering is supported.
588
Readers will be applied in first-to-last order.
589
Writers will be applied in last-to-first order.
590
Either the path or the file-id needs to be provided.
592
:param path: path relative to the root of the tree
594
:param file_id: file_id or None if unknown
595
:return: the list of filters - [] if there are none
597
filter_pref_names = filters._get_registered_names()
598
if len(filter_pref_names) == 0:
601
path = self.id2path(file_id)
602
prefs = self.iter_search_rules([path], filter_pref_names).next()
603
stk = filters._get_filter_stack_for(prefs)
604
if 'filters' in debug.debug_flags:
605
trace.note("*** %s content-filter: %s => %r" % (path,prefs,stk))
608
def _content_filter_stack_provider(self):
609
"""A function that returns a stack of ContentFilters.
611
The function takes a path (relative to the top of the tree) and a
612
file-id as parameters.
614
:return: None if content filtering is not supported by this tree.
616
if self.supports_content_filtering():
617
return lambda path, file_id: \
618
self._content_filter_stack(path, file_id)
622
def iter_search_rules(self, path_names, pref_names=None,
623
_default_searcher=None):
624
"""Find the preferences for filenames in a tree.
626
:param path_names: an iterable of paths to find attributes for.
627
Paths are given relative to the root of the tree.
628
:param pref_names: the list of preferences to lookup - None for all
629
:param _default_searcher: private parameter to assist testing - don't use
630
:return: an iterator of tuple sequences, one per path-name.
631
See _RulesSearcher.get_items for details on the tuple sequence.
633
if _default_searcher is None:
634
_default_searcher = rules._per_user_searcher
635
searcher = self._get_rules_searcher(_default_searcher)
636
if searcher is not None:
637
if pref_names is not None:
638
for path in path_names:
639
yield searcher.get_selected_items(path, pref_names)
641
for path in path_names:
642
yield searcher.get_items(path)
644
def _get_rules_searcher(self, default_searcher):
645
"""Get the RulesSearcher for this tree given the default one."""
646
searcher = default_searcher
650
class InventoryTree(Tree):
651
"""A tree that relies on an inventory for its metadata.
653
Trees contain an `Inventory` object, and also know how to retrieve
654
file texts mentioned in the inventory, either from a working
655
directory or from a store.
657
It is possible for trees to contain files that are not described
658
in their inventory or vice versa; for this use `filenames()`.
660
Subclasses should set the _inventory attribute, which is considered
661
private to external API users.
407
def kind(self, file_id):
408
file_id = osutils.safe_file_id(file_id)
409
assert self._inventory[file_id].kind == "directory"
412
def list_files(self, include_root=False):
415
def __contains__(self, file_id):
416
file_id = osutils.safe_file_id(file_id)
417
return (file_id in self._inventory)
419
def get_file_sha1(self, file_id, path=None, stat_value=None):
423
######################################################################
426
# TODO: Merge these two functions into a single one that can operate
427
# on either a whole tree or a set of files.
429
# TODO: Return the diff in order by filename, not by category or in
430
# random order. Can probably be done by lock-stepping through the
431
# filenames from both trees.
434
def file_status(filename, old_tree, new_tree):
435
"""Return single-letter status, old and new names for a file.
437
The complexity here is in deciding how to represent renames;
438
many complex cases are possible.
664
def get_canonical_inventory_paths(self, paths):
665
"""Like get_canonical_inventory_path() but works on multiple items.
667
:param paths: A sequence of paths relative to the root of the tree.
668
:return: A list of paths, with each item the corresponding input path
669
adjusted to account for existing elements that match case
672
return list(self._yield_canonical_inventory_paths(paths))
674
def get_canonical_inventory_path(self, path):
675
"""Returns the first inventory item that case-insensitively matches path.
677
If a path matches exactly, it is returned. If no path matches exactly
678
but more than one path matches case-insensitively, it is implementation
679
defined which is returned.
681
If no path matches case-insensitively, the input path is returned, but
682
with as many path entries that do exist changed to their canonical
685
If you need to resolve many names from the same tree, you should
686
use get_canonical_inventory_paths() to avoid O(N) behaviour.
688
:param path: A paths relative to the root of the tree.
689
:return: The input path adjusted to account for existing elements
690
that match case insensitively.
692
return self._yield_canonical_inventory_paths([path]).next()
694
def _yield_canonical_inventory_paths(self, paths):
696
# First, if the path as specified exists exactly, just use it.
697
if self.path2id(path) is not None:
701
cur_id = self.get_root_id()
703
bit_iter = iter(path.split("/"))
707
for child in self.iter_children(cur_id):
709
# XXX: it seem like if the child is known to be in the
710
# tree, we shouldn't need to go from its id back to
711
# its path -- mbp 2010-02-11
713
# XXX: it seems like we could be more efficient
714
# by just directly looking up the original name and
715
# only then searching all children; also by not
716
# chopping paths so much. -- mbp 2010-02-11
717
child_base = os.path.basename(self.id2path(child))
718
if (child_base == elt):
719
# if we found an exact match, we can stop now; if
720
# we found an approximate match we need to keep
721
# searching because there might be an exact match
724
new_path = osutils.pathjoin(cur_path, child_base)
726
elif child_base.lower() == lelt:
728
new_path = osutils.pathjoin(cur_path, child_base)
729
except errors.NoSuchId:
730
# before a change is committed we can see this error...
735
# got to the end of this directory and no entries matched.
736
# Return what matched so far, plus the rest as specified.
737
cur_path = osutils.pathjoin(cur_path, elt, *list(bit_iter))
742
def _get_inventory(self):
743
return self._inventory
745
inventory = property(_get_inventory,
746
doc="Inventory of this Tree")
749
def path2id(self, path):
750
"""Return the id for path in this tree."""
751
return self._inventory.path2id(path)
753
def id2path(self, file_id):
754
"""Return the path for a file id.
758
return self.inventory.id2path(file_id)
760
def has_id(self, file_id):
761
return self.inventory.has_id(file_id)
763
def has_or_had_id(self, file_id):
764
return self.inventory.has_id(file_id)
766
def all_file_ids(self):
767
return set(self.inventory)
769
@deprecated_method(deprecated_in((2, 4, 0)))
771
return iter(self.inventory)
773
def filter_unversioned_files(self, paths):
774
"""Filter out paths that are versioned.
776
:return: set of paths.
778
# NB: we specifically *don't* call self.has_filename, because for
779
# WorkingTrees that can indicate files that exist on disk but that
781
pred = self.inventory.has_filename
782
return set((p for p in paths if not pred(p)))
785
def iter_entries_by_dir(self, specific_file_ids=None, yield_parents=False):
786
"""Walk the tree in 'by_dir' order.
788
This will yield each entry in the tree as a (path, entry) tuple.
789
The order that they are yielded is:
791
See Tree.iter_entries_by_dir for details.
793
:param yield_parents: If True, yield the parents from the root leading
794
down to specific_file_ids that have been requested. This has no
795
impact if specific_file_ids is None.
797
return self.inventory.iter_entries_by_dir(
798
specific_file_ids=specific_file_ids, yield_parents=yield_parents)
800
def get_file_by_path(self, path):
801
return self.get_file(self._inventory.path2id(path), path)
440
old_inv = old_tree.inventory
441
new_inv = new_tree.inventory
442
new_id = new_inv.path2id(filename)
443
old_id = old_inv.path2id(filename)
445
if not new_id and not old_id:
446
# easy: doesn't exist in either; not versioned at all
447
if new_tree.is_ignored(filename):
448
return 'I', None, None
450
return '?', None, None
452
# There is now a file of this name, great.
455
# There is no longer a file of this name, but we can describe
456
# what happened to the file that used to have
457
# this name. There are two possibilities: either it was
458
# deleted entirely, or renamed.
460
if new_inv.has_id(old_id):
461
return 'X', old_inv.id2path(old_id), new_inv.id2path(old_id)
463
return 'D', old_inv.id2path(old_id), None
465
# if the file_id is new in this revision, it is added
466
if new_id and not old_inv.has_id(new_id):
469
# if there used to be a file of this name, but that ID has now
470
# disappeared, it is deleted
471
if old_id and not new_inv.has_id(old_id):
478
def find_renames(old_inv, new_inv):
479
for file_id in old_inv:
480
if file_id not in new_inv:
482
old_name = old_inv.id2path(file_id)
483
new_name = new_inv.id2path(file_id)
484
if old_name != new_name:
485
yield (old_name, new_name)
804
488
def find_ids_across_trees(filenames, trees, require_versioned=True):
805
489
"""Find the ids corresponding to specified filenames.
807
491
All matches in all trees will be used, and all children of matched
808
492
directories will be used.
880
565
Its instances have methods like 'compare' and contain references to the
881
566
source and target trees these operations are to be carried out on.
883
Clients of bzrlib should not need to use InterTree directly, rather they
568
clients of bzrlib should not need to use InterTree directly, rather they
884
569
should use the convenience methods on Tree such as 'Tree.compare()' which
885
570
will pass through to InterTree as appropriate.
888
# Formats that will be used to test this InterTree. If both are
889
# None, this InterTree will not be tested (e.g. because a complex
891
_matching_from_tree_format = None
892
_matching_to_tree_format = None
897
def is_compatible(kls, source, target):
898
# The default implementation is naive and uses the public API, so
899
# it works for all trees.
902
def _changes_from_entries(self, source_entry, target_entry,
903
source_path=None, target_path=None):
904
"""Generate a iter_changes tuple between source_entry and target_entry.
906
:param source_entry: An inventory entry from self.source, or None.
907
:param target_entry: An inventory entry from self.target, or None.
908
:param source_path: The path of source_entry, if known. If not known
909
it will be looked up.
910
:param target_path: The path of target_entry, if known. If not known
911
it will be looked up.
912
:return: A tuple, item 0 of which is an iter_changes result tuple, and
913
item 1 is True if there are any changes in the result tuple.
915
if source_entry is None:
916
if target_entry is None:
918
file_id = target_entry.file_id
920
file_id = source_entry.file_id
921
if source_entry is not None:
922
source_versioned = True
923
source_name = source_entry.name
924
source_parent = source_entry.parent_id
925
if source_path is None:
926
source_path = self.source.id2path(file_id)
927
source_kind, source_executable, source_stat = \
928
self.source._comparison_data(source_entry, source_path)
930
source_versioned = False
934
source_executable = None
935
if target_entry is not None:
936
target_versioned = True
937
target_name = target_entry.name
938
target_parent = target_entry.parent_id
939
if target_path is None:
940
target_path = self.target.id2path(file_id)
941
target_kind, target_executable, target_stat = \
942
self.target._comparison_data(target_entry, target_path)
944
target_versioned = False
948
target_executable = None
949
versioned = (source_versioned, target_versioned)
950
kind = (source_kind, target_kind)
951
changed_content = False
952
if source_kind != target_kind:
953
changed_content = True
954
elif source_kind == 'file':
955
if (self.source.get_file_sha1(file_id, source_path, source_stat) !=
956
self.target.get_file_sha1(file_id, target_path, target_stat)):
957
changed_content = True
958
elif source_kind == 'symlink':
959
if (self.source.get_symlink_target(file_id) !=
960
self.target.get_symlink_target(file_id)):
961
changed_content = True
962
# XXX: Yes, the indentation below is wrong. But fixing it broke
963
# test_merge.TestMergerEntriesLCAOnDisk.
964
# test_nested_tree_subtree_renamed_and_modified. We'll wait for
965
# the fix from bzr.dev -- vila 2009026
966
elif source_kind == 'tree-reference':
967
if (self.source.get_reference_revision(file_id, source_path)
968
!= self.target.get_reference_revision(file_id, target_path)):
969
changed_content = True
970
parent = (source_parent, target_parent)
971
name = (source_name, target_name)
972
executable = (source_executable, target_executable)
973
if (changed_content is not False or versioned[0] != versioned[1]
974
or parent[0] != parent[1] or name[0] != name[1] or
975
executable[0] != executable[1]):
979
return (file_id, (source_path, target_path), changed_content,
980
versioned, parent, name, kind, executable), changes
983
576
def compare(self, want_unchanged=False, specific_files=None,
984
577
extra_trees=None, require_versioned=False, include_root=False,
1045
640
:param require_versioned: Raise errors.PathsNotVersionedError if a
1046
641
path in the specific_files list is not versioned in one of
1047
642
source, target or extra_trees.
1048
:param specific_files: An optional list of file paths to restrict the
1049
comparison to. When mapping filenames to ids, all matches in all
1050
trees (including optional extra_trees) are used, and all children
1051
of matched directories are included. The parents in the target tree
1052
of the specific files up to and including the root of the tree are
1053
always evaluated for changes too.
1054
643
:param want_unversioned: Should unversioned files be returned in the
1055
644
output. An unversioned file is defined as one with (False, False)
1056
645
for the versioned pair.
1058
648
lookup_trees = [self.source]
1060
650
lookup_trees.extend(extra_trees)
1061
# The ids of items we need to examine to insure delta consistency.
1062
precise_file_ids = set()
1063
changed_file_ids = []
1064
if specific_files == []:
1065
specific_file_ids = []
1067
specific_file_ids = self.target.paths2ids(specific_files,
1068
lookup_trees, require_versioned=require_versioned)
1069
if specific_files is not None:
1070
# reparented or added entries must have their parents included
1071
# so that valid deltas can be created. The seen_parents set
1072
# tracks the parents that we need to have.
1073
# The seen_dirs set tracks directory entries we've yielded.
1074
# After outputting version object in to_entries we set difference
1075
# the two seen sets and start checking parents.
1076
seen_parents = set()
651
specific_file_ids = self.target.paths2ids(specific_files,
652
lookup_trees, require_versioned=require_versioned)
1078
653
if want_unversioned:
1079
all_unversioned = sorted([(p.split('/'), p) for p in
1080
self.target.extras()
1081
if specific_files is None or
654
all_unversioned = sorted([(p.split('/'), p) for p in self.target.extras()
655
if not specific_files or
1082
656
osutils.is_inside_any(specific_files, p)])
1083
all_unversioned = collections.deque(all_unversioned)
657
all_unversioned = deque(all_unversioned)
1085
all_unversioned = collections.deque()
659
all_unversioned = deque()
1087
from_entries_by_dir = list(self.source.iter_entries_by_dir(
661
from_entries_by_dir = list(self.source.inventory.iter_entries_by_dir(
1088
662
specific_file_ids=specific_file_ids))
1089
663
from_data = dict((e.file_id, (p, e)) for p, e in from_entries_by_dir)
1090
to_entries_by_dir = list(self.target.iter_entries_by_dir(
664
to_entries_by_dir = list(self.target.inventory.iter_entries_by_dir(
1091
665
specific_file_ids=specific_file_ids))
1092
666
num_entries = len(from_entries_by_dir) + len(to_entries_by_dir)
1094
# the unversioned path lookup only occurs on real trees - where there
668
# the unversioned path lookup only occurs on real trees - where there
1095
669
# can be extras. So the fake_entry is solely used to look up
1096
670
# executable it values when execute is not supported.
1097
fake_entry = inventory.InventoryFile('unused', 'unused', 'unused')
1098
for target_path, target_entry in to_entries_by_dir:
1099
while (all_unversioned and
1100
all_unversioned[0][0] < target_path.split('/')):
671
fake_entry = InventoryFile('unused', 'unused', 'unused')
672
for to_path, to_entry in to_entries_by_dir:
673
while all_unversioned and all_unversioned[0][0] < to_path.split('/'):
1101
674
unversioned_path = all_unversioned.popleft()
1102
target_kind, target_executable, target_stat = \
675
to_kind, to_executable, to_stat = \
1103
676
self.target._comparison_data(fake_entry, unversioned_path[1])
1104
677
yield (None, (None, unversioned_path[1]), True, (False, False),
1106
679
(None, unversioned_path[0][-1]),
1107
(None, target_kind),
1108
(None, target_executable))
1109
source_path, source_entry = from_data.get(target_entry.file_id,
1111
result, changes = self._changes_from_entries(source_entry,
1112
target_entry, source_path=source_path, target_path=target_path)
1113
to_paths[result[0]] = result[1][1]
681
(None, to_executable))
682
file_id = to_entry.file_id
683
to_paths[file_id] = to_path
1114
684
entry_count += 1
685
changed_content = False
686
from_path, from_entry = from_data.get(file_id, (None, None))
687
from_versioned = (from_entry is not None)
688
if from_entry is not None:
689
from_versioned = True
690
from_name = from_entry.name
691
from_parent = from_entry.parent_id
692
from_kind, from_executable, from_stat = \
693
self.source._comparison_data(from_entry, from_path)
1116
694
entry_count += 1
696
from_versioned = False
700
from_executable = None
701
versioned = (from_versioned, True)
702
to_kind, to_executable, to_stat = \
703
self.target._comparison_data(to_entry, to_path)
704
kind = (from_kind, to_kind)
705
if kind[0] != kind[1]:
706
changed_content = True
707
elif from_kind == 'file':
708
from_size = self.source._file_size(from_entry, from_stat)
709
to_size = self.target._file_size(to_entry, to_stat)
710
if from_size != to_size:
711
changed_content = True
712
elif (self.source.get_file_sha1(file_id, from_path, from_stat) !=
713
self.target.get_file_sha1(file_id, to_path, to_stat)):
714
changed_content = True
715
elif from_kind == 'symlink':
716
if (self.source.get_symlink_target(file_id) !=
717
self.target.get_symlink_target(file_id)):
718
changed_content = True
719
elif from_kind == 'tree-reference':
720
if (self.source.get_reference_revision(file_id, from_path)
721
!= self.target.get_reference_revision(file_id, to_path)):
722
changed_content = True
723
parent = (from_parent, to_entry.parent_id)
724
name = (from_name, to_entry.name)
725
executable = (from_executable, to_executable)
1117
726
if pb is not None:
1118
727
pb.update('comparing files', entry_count, num_entries)
1119
if changes or include_unchanged:
1120
if specific_file_ids is not None:
1121
new_parent_id = result[4][1]
1122
precise_file_ids.add(new_parent_id)
1123
changed_file_ids.append(result[0])
1125
# Ensure correct behaviour for reparented/added specific files.
1126
if specific_files is not None:
1127
# Record output dirs
1128
if result[6][1] == 'directory':
1129
seen_dirs.add(result[0])
1130
# Record parents of reparented/added entries.
1131
versioned = result[3]
1133
if not versioned[0] or parents[0] != parents[1]:
1134
seen_parents.add(parents[1])
728
if (changed_content is not False or versioned[0] != versioned[1]
729
or parent[0] != parent[1] or name[0] != name[1] or
730
executable[0] != executable[1] or include_unchanged):
731
yield (file_id, (from_path, to_path), changed_content,
732
versioned, parent, name, kind, executable)
1135
734
while all_unversioned:
1136
735
# yield any trailing unversioned paths
1137
736
unversioned_path = all_unversioned.popleft()
1164
775
self.source._comparison_data(from_entry, path)
1165
776
kind = (from_kind, None)
1166
777
executable = (from_executable, None)
1167
changed_content = from_kind is not None
778
changed_content = True
1168
779
# the parent's path is necessarily known at this point.
1169
changed_file_ids.append(file_id)
1170
780
yield(file_id, (path, to_path), changed_content, versioned, parent,
1171
781
name, kind, executable)
1172
changed_file_ids = set(changed_file_ids)
1173
if specific_file_ids is not None:
1174
for result in self._handle_precise_ids(precise_file_ids,
1178
def _get_entry(self, tree, file_id):
1179
"""Get an inventory entry from a tree, with missing entries as None.
1181
If the tree raises NotImplementedError on accessing .inventory, then
1182
this is worked around using iter_entries_by_dir on just the file id
1185
:param tree: The tree to lookup the entry in.
1186
:param file_id: The file_id to lookup.
1189
inventory = tree.inventory
1190
except NotImplementedError:
1191
# No inventory available.
1193
iterator = tree.iter_entries_by_dir(specific_file_ids=[file_id])
1194
return iterator.next()[1]
1195
except StopIteration:
1199
return inventory[file_id]
1200
except errors.NoSuchId:
1203
def _handle_precise_ids(self, precise_file_ids, changed_file_ids,
1204
discarded_changes=None):
1205
"""Fill out a partial iter_changes to be consistent.
1207
:param precise_file_ids: The file ids of parents that were seen during
1209
:param changed_file_ids: The file ids of already emitted items.
1210
:param discarded_changes: An optional dict of precalculated
1211
iter_changes items which the partial iter_changes had not output
1213
:return: A generator of iter_changes items to output.
1215
# process parents of things that had changed under the users
1216
# requested paths to prevent incorrect paths or parent ids which
1217
# aren't in the tree.
1218
while precise_file_ids:
1219
precise_file_ids.discard(None)
1220
# Don't emit file_ids twice
1221
precise_file_ids.difference_update(changed_file_ids)
1222
if not precise_file_ids:
1224
# If the there was something at a given output path in source, we
1225
# have to include the entry from source in the delta, or we would
1226
# be putting this entry into a used path.
1228
for parent_id in precise_file_ids:
1230
paths.append(self.target.id2path(parent_id))
1231
except errors.NoSuchId:
1232
# This id has been dragged in from the source by delta
1233
# expansion and isn't present in target at all: we don't
1234
# need to check for path collisions on it.
1237
old_id = self.source.path2id(path)
1238
precise_file_ids.add(old_id)
1239
precise_file_ids.discard(None)
1240
current_ids = precise_file_ids
1241
precise_file_ids = set()
1242
# We have to emit all of precise_file_ids that have been altered.
1243
# We may have to output the children of some of those ids if any
1244
# directories have stopped being directories.
1245
for file_id in current_ids:
1247
if discarded_changes:
1248
result = discarded_changes.get(file_id)
1253
old_entry = self._get_entry(self.source, file_id)
1254
new_entry = self._get_entry(self.target, file_id)
1255
result, changes = self._changes_from_entries(
1256
old_entry, new_entry)
1259
# Get this parents parent to examine.
1260
new_parent_id = result[4][1]
1261
precise_file_ids.add(new_parent_id)
1263
if (result[6][0] == 'directory' and
1264
result[6][1] != 'directory'):
1265
# This stopped being a directory, the old children have
1267
if old_entry is None:
1268
# Reusing a discarded change.
1269
old_entry = self._get_entry(self.source, file_id)
1270
for child in old_entry.children.values():
1271
precise_file_ids.add(child.file_id)
1272
changed_file_ids.add(result[0])
1276
InterTree.register_optimiser(InterTree)
1279
class MultiWalker(object):
1280
"""Walk multiple trees simultaneously, getting combined results."""
1282
# Note: This could be written to not assume you can do out-of-order
1283
# lookups. Instead any nodes that don't match in all trees could be
1284
# marked as 'deferred', and then returned in the final cleanup loop.
1285
# For now, I think it is "nicer" to return things as close to the
1286
# "master_tree" order as we can.
1288
def __init__(self, master_tree, other_trees):
1289
"""Create a new MultiWalker.
1291
All trees being walked must implement "iter_entries_by_dir()", such
1292
that they yield (path, object) tuples, where that object will have a
1293
'.file_id' member, that can be used to check equality.
1295
:param master_tree: All trees will be 'slaved' to the master_tree such
1296
that nodes in master_tree will be used as 'first-pass' sync points.
1297
Any nodes that aren't in master_tree will be merged in a second
1299
:param other_trees: A list of other trees to walk simultaneously.
1301
self._master_tree = master_tree
1302
self._other_trees = other_trees
1304
# Keep track of any nodes that were properly processed just out of
1305
# order, that way we don't return them at the end, we don't have to
1306
# track *all* processed file_ids, just the out-of-order ones
1307
self._out_of_order_processed = set()
1310
def _step_one(iterator):
1311
"""Step an iter_entries_by_dir iterator.
1313
:return: (has_more, path, ie)
1314
If has_more is False, path and ie will be None.
1317
path, ie = iterator.next()
1318
except StopIteration:
1319
return False, None, None
1321
return True, path, ie
1324
def _cmp_path_by_dirblock(path1, path2):
1325
"""Compare two paths based on what directory they are in.
1327
This generates a sort order, such that all children of a directory are
1328
sorted together, and grandchildren are in the same order as the
1329
children appear. But all grandchildren come after all children.
1331
:param path1: first path
1332
:param path2: the second path
1333
:return: negative number if ``path1`` comes first,
1334
0 if paths are equal
1335
and a positive number if ``path2`` sorts first
1337
# Shortcut this special case
1340
# This is stolen from _dirstate_helpers_py.py, only switching it to
1341
# Unicode objects. Consider using encode_utf8() and then using the
1342
# optimized versions, or maybe writing optimized unicode versions.
1343
if not isinstance(path1, unicode):
1344
raise TypeError("'path1' must be a unicode string, not %s: %r"
1345
% (type(path1), path1))
1346
if not isinstance(path2, unicode):
1347
raise TypeError("'path2' must be a unicode string, not %s: %r"
1348
% (type(path2), path2))
1349
return cmp(MultiWalker._path_to_key(path1),
1350
MultiWalker._path_to_key(path2))
1353
def _path_to_key(path):
1354
dirname, basename = osutils.split(path)
1355
return (dirname.split(u'/'), basename)
1357
def _lookup_by_file_id(self, extra_entries, other_tree, file_id):
1358
"""Lookup an inventory entry by file_id.
1360
This is called when an entry is missing in the normal order.
1361
Generally this is because a file was either renamed, or it was
1362
deleted/added. If the entry was found in the inventory and not in
1363
extra_entries, it will be added to self._out_of_order_processed
1365
:param extra_entries: A dictionary of {file_id: (path, ie)}. This
1366
should be filled with entries that were found before they were
1367
used. If file_id is present, it will be removed from the
1369
:param other_tree: The Tree to search, in case we didn't find the entry
1371
:param file_id: The file_id to look for
1372
:return: (path, ie) if found or (None, None) if not present.
1374
if file_id in extra_entries:
1375
return extra_entries.pop(file_id)
1376
# TODO: Is id2path better as the first call, or is
1377
# inventory[file_id] better as a first check?
1379
cur_path = other_tree.id2path(file_id)
1380
except errors.NoSuchId:
1382
if cur_path is None:
1385
self._out_of_order_processed.add(file_id)
1386
cur_ie = other_tree.inventory[file_id]
1387
return (cur_path, cur_ie)
1390
"""Match up the values in the different trees."""
1391
for result in self._walk_master_tree():
1393
self._finish_others()
1394
for result in self._walk_others():
1397
def _walk_master_tree(self):
1398
"""First pass, walk all trees in lock-step.
1400
When we are done, all nodes in the master_tree will have been
1401
processed. _other_walkers, _other_entries, and _others_extra will be
1402
set on 'self' for future processing.
1404
# This iterator has the most "inlining" done, because it tends to touch
1405
# every file in the tree, while the others only hit nodes that don't
1407
master_iterator = self._master_tree.iter_entries_by_dir()
1409
other_walkers = [other.iter_entries_by_dir()
1410
for other in self._other_trees]
1411
other_entries = [self._step_one(walker) for walker in other_walkers]
1412
# Track extra nodes in the other trees
1413
others_extra = [{} for i in xrange(len(self._other_trees))]
1415
master_has_more = True
1416
step_one = self._step_one
1417
lookup_by_file_id = self._lookup_by_file_id
1418
out_of_order_processed = self._out_of_order_processed
1420
while master_has_more:
1421
(master_has_more, path, master_ie) = step_one(master_iterator)
1422
if not master_has_more:
1425
file_id = master_ie.file_id
1427
other_values_append = other_values.append
1428
next_other_entries = []
1429
next_other_entries_append = next_other_entries.append
1430
for idx, (other_has_more, other_path, other_ie) in enumerate(other_entries):
1431
if not other_has_more:
1432
other_values_append(lookup_by_file_id(
1433
others_extra[idx], self._other_trees[idx], file_id))
1434
next_other_entries_append((False, None, None))
1435
elif file_id == other_ie.file_id:
1436
# This is the critical code path, as most of the entries
1437
# should match between most trees.
1438
other_values_append((other_path, other_ie))
1439
next_other_entries_append(step_one(other_walkers[idx]))
1441
# This walker did not match, step it until it either
1442
# matches, or we know we are past the current walker.
1443
other_walker = other_walkers[idx]
1444
other_extra = others_extra[idx]
1445
while (other_has_more and
1446
self._cmp_path_by_dirblock(other_path, path) < 0):
1447
other_file_id = other_ie.file_id
1448
if other_file_id not in out_of_order_processed:
1449
other_extra[other_file_id] = (other_path, other_ie)
1450
other_has_more, other_path, other_ie = \
1451
step_one(other_walker)
1452
if other_has_more and other_ie.file_id == file_id:
1453
# We ended up walking to this point, match and step
1455
other_values_append((other_path, other_ie))
1456
other_has_more, other_path, other_ie = \
1457
step_one(other_walker)
1459
# This record isn't in the normal order, see if it
1461
other_values_append(lookup_by_file_id(
1462
other_extra, self._other_trees[idx], file_id))
1463
next_other_entries_append((other_has_more, other_path,
1465
other_entries = next_other_entries
1467
# We've matched all the walkers, yield this datapoint
1468
yield path, file_id, master_ie, other_values
1469
self._other_walkers = other_walkers
1470
self._other_entries = other_entries
1471
self._others_extra = others_extra
1473
def _finish_others(self):
1474
"""Finish walking the other iterators, so we get all entries."""
1475
for idx, info in enumerate(self._other_entries):
1476
other_extra = self._others_extra[idx]
1477
(other_has_more, other_path, other_ie) = info
1478
while other_has_more:
1479
other_file_id = other_ie.file_id
1480
if other_file_id not in self._out_of_order_processed:
1481
other_extra[other_file_id] = (other_path, other_ie)
1482
other_has_more, other_path, other_ie = \
1483
self._step_one(self._other_walkers[idx])
1484
del self._other_entries
1486
def _walk_others(self):
1487
"""Finish up by walking all the 'deferred' nodes."""
1488
# TODO: One alternative would be to grab all possible unprocessed
1489
# file_ids, and then sort by path, and then yield them. That
1490
# might ensure better ordering, in case a caller strictly
1491
# requires parents before children.
1492
for idx, other_extra in enumerate(self._others_extra):
1493
others = sorted(other_extra.itervalues(),
1494
key=lambda x: self._path_to_key(x[0]))
1495
for other_path, other_ie in others:
1496
file_id = other_ie.file_id
1497
# We don't need to check out_of_order_processed here, because
1498
# the lookup_by_file_id will be removing anything processed
1499
# from the extras cache
1500
other_extra.pop(file_id)
1501
other_values = [(None, None) for i in xrange(idx)]
1502
other_values.append((other_path, other_ie))
1503
for alt_idx, alt_extra in enumerate(self._others_extra[idx+1:]):
1504
alt_idx = alt_idx + idx + 1
1505
alt_extra = self._others_extra[alt_idx]
1506
alt_tree = self._other_trees[alt_idx]
1507
other_values.append(self._lookup_by_file_id(
1508
alt_extra, alt_tree, file_id))
1509
yield other_path, file_id, None, other_values
784
# This was deprecated before 0.12, but did not have an official warning
785
@symbol_versioning.deprecated_function(symbol_versioning.zero_twelve)
786
def RevisionTree(*args, **kwargs):
787
"""RevisionTree has moved to bzrlib.revisiontree.RevisionTree()
789
Accessing it as bzrlib.tree.RevisionTree has been deprecated as of
792
from bzrlib.revisiontree import RevisionTree as _RevisionTree
793
return _RevisionTree(*args, **kwargs)
added
removed
bzrlib/tree.py
