84
97
specific_files=specific_files,
85
98
extra_trees=extra_trees,
86
99
require_versioned=require_versioned,
87
include_root=include_root
100
include_root=include_root,
101
want_unversioned=want_unversioned,
90
def _iter_changes(self, from_tree, include_unchanged=False,
91
specific_file_ids=None, pb=None):
104
def iter_changes(self, from_tree, include_unchanged=False,
105
specific_files=None, pb=None, extra_trees=None,
106
require_versioned=True, want_unversioned=False):
107
"""See InterTree.iter_changes"""
92
108
intertree = InterTree.get(from_tree, self)
93
return intertree._iter_changes(from_tree, self, include_unchanged,
94
specific_file_ids, pb)
109
return intertree.iter_changes(include_unchanged, specific_files, pb,
110
extra_trees, require_versioned, want_unversioned=want_unversioned)
96
112
def conflicts(self):
97
113
"""Get a list of the conflicts in the tree.
99
115
Each conflict is an instance of bzrlib.conflicts.Conflict.
117
return _mod_conflicts.ConflictList()
120
"""For trees that can have unversioned files, return all such paths."""
103
123
def get_parent_ids(self):
104
"""Get the parent ids for this tree.
124
"""Get the parent ids for this tree.
106
126
:return: a list of parent ids. [] is returned to indicate
107
127
a tree with no parents.
108
128
:raises: BzrError if the parents are not known.
110
130
raise NotImplementedError(self.get_parent_ids)
112
132
def has_filename(self, filename):
113
133
"""True if the tree has given filename."""
114
raise NotImplementedError()
134
raise NotImplementedError(self.has_filename)
116
136
def has_id(self, file_id):
117
return self.inventory.has_id(file_id)
137
raise NotImplementedError(self.has_id)
119
__contains__ = has_id
139
@deprecated_method(deprecated_in((2, 4, 0)))
140
def __contains__(self, file_id):
141
return self.has_id(file_id)
121
143
def has_or_had_id(self, file_id):
122
if file_id == self.inventory.root.file_id:
124
return self.inventory.has_id(file_id)
127
return iter(self.inventory)
144
raise NotImplementedError(self.has_or_had_id)
146
def is_ignored(self, filename):
147
"""Check whether the filename is ignored by this tree.
149
:param filename: The relative filename within the tree.
150
:return: True if the filename is ignored.
154
def all_file_ids(self):
155
"""Iterate through all file ids, including ids for missing files."""
156
raise NotImplementedError(self.all_file_ids)
129
158
def id2path(self, file_id):
130
return self.inventory.id2path(file_id)
132
def is_control_filename(self, filename):
133
"""True if filename is the name of a control file in this tree.
135
:param filename: A filename within the tree. This is a relative path
136
from the root of this tree.
138
This is true IF and ONLY IF the filename is part of the meta data
139
that bzr controls in this tree. I.E. a random .bzr directory placed
140
on disk will not be a control file for this tree.
159
"""Return the path for a file id.
142
return self.bzrdir.is_control_filename(filename)
163
raise NotImplementedError(self.id2path)
145
def iter_entries_by_dir(self, specific_file_ids=None):
165
def iter_entries_by_dir(self, specific_file_ids=None, yield_parents=False):
146
166
"""Walk the tree in 'by_dir' order.
148
This will yield each entry in the tree as a (path, entry) tuple. The
149
order that they are yielded is: the contents of a directory are
150
preceeded by the parent of a directory, and all the contents of a
151
directory are grouped together.
153
return self.inventory.iter_entries_by_dir(
154
specific_file_ids=specific_file_ids)
168
This will yield each entry in the tree as a (path, entry) tuple.
169
The order that they are yielded is:
171
Directories are walked in a depth-first lexicographical order,
172
however, whenever a directory is reached, all of its direct child
173
nodes are yielded in lexicographical order before yielding the
176
For example, in the tree::
186
The yield order (ignoring root) would be::
188
a, f, a/b, a/d, a/b/c, a/d/e, f/g
190
:param yield_parents: If True, yield the parents from the root leading
191
down to specific_file_ids that have been requested. This has no
192
impact if specific_file_ids is None.
194
raise NotImplementedError(self.iter_entries_by_dir)
196
def list_files(self, include_root=False, from_dir=None, recursive=True):
197
"""List all files in this tree.
199
:param include_root: Whether to include the entry for the tree root
200
:param from_dir: Directory under which to list files
201
:param recursive: Whether to list files recursively
202
:return: iterator over tuples of (path, versioned, kind, file_id,
205
raise NotImplementedError(self.list_files)
207
def iter_references(self):
208
if self.supports_tree_reference():
209
for path, entry in self.iter_entries_by_dir():
210
if entry.kind == 'tree-reference':
211
yield path, entry.file_id
156
213
def kind(self, file_id):
157
raise NotImplementedError("subclasses must implement kind")
214
raise NotImplementedError("Tree subclass %s must implement kind"
215
% self.__class__.__name__)
217
def stored_kind(self, file_id):
218
"""File kind stored for this file_id.
220
May not match kind on disk for working trees. Always available
221
for versioned files, even when the file itself is missing.
223
return self.kind(file_id)
225
def path_content_summary(self, path):
226
"""Get a summary of the information about path.
228
All the attributes returned are for the canonical form, not the
229
convenient form (if content filters are in use.)
231
:param path: A relative path within the tree.
232
:return: A tuple containing kind, size, exec, sha1-or-link.
233
Kind is always present (see tree.kind()).
234
size is present if kind is file and the size of the
235
canonical form can be cheaply determined, None otherwise.
236
exec is None unless kind is file and the platform supports the 'x'
238
sha1-or-link is the link target if kind is symlink, or the sha1 if
239
it can be obtained without reading the file.
241
raise NotImplementedError(self.path_content_summary)
243
def get_reference_revision(self, file_id, path=None):
244
raise NotImplementedError("Tree subclass %s must implement "
245
"get_reference_revision"
246
% self.__class__.__name__)
159
248
def _comparison_data(self, entry, path):
160
249
"""Return a tuple of kind, executable, stat_value for a file.
170
259
def _file_size(self, entry, stat_value):
171
260
raise NotImplementedError(self._file_size)
173
def _get_inventory(self):
174
return self._inventory
176
def get_file(self, file_id):
177
"""Return a file object for the file file_id in the tree."""
262
def get_file(self, file_id, path=None):
263
"""Return a file object for the file file_id in the tree.
265
If both file_id and path are defined, it is implementation defined as
266
to which one is used.
178
268
raise NotImplementedError(self.get_file)
180
def get_file_by_path(self, path):
181
return self.get_file(self._inventory.path2id(path))
183
def annotate_iter(self, file_id):
184
"""Return an iterator of revision_id, line tuples
270
def get_file_with_stat(self, file_id, path=None):
271
"""Get a file handle and stat object for file_id.
273
The default implementation returns (self.get_file, None) for backwards
276
:param file_id: The file id to read.
277
:param path: The path of the file, if it is known.
278
:return: A tuple (file_handle, stat_value_or_None). If the tree has
279
no stat facility, or need for a stat cache feedback during commit,
280
it may return None for the second element of the tuple.
282
return (self.get_file(file_id, path), None)
284
def get_file_text(self, file_id, path=None):
285
"""Return the byte content of a file.
287
:param file_id: The file_id of the file.
288
:param path: The path of the file.
290
If both file_id and path are supplied, an implementation may use
293
:returns: A single byte string for the whole file.
295
my_file = self.get_file(file_id, path)
297
return my_file.read()
301
def get_file_lines(self, file_id, path=None):
302
"""Return the content of a file, as lines.
304
:param file_id: The file_id of the file.
305
:param path: The path of the file.
307
If both file_id and path are supplied, an implementation may use
310
return osutils.split_lines(self.get_file_text(file_id, path))
312
def get_file_verifier(self, file_id, path=None, stat_value=None):
313
"""Return a verifier for a file.
315
The default implementation returns a sha1.
317
:param file_id: The handle for this file.
318
:param path: The path that this file can be found at.
319
These must point to the same object.
320
:param stat_value: Optional stat value for the object
321
:return: Tuple with verifier name and verifier data
323
return ("SHA1", self.get_file_sha1(file_id, path=path,
324
stat_value=stat_value))
326
def get_file_sha1(self, file_id, path=None, stat_value=None):
327
"""Return the SHA1 file for a file.
329
:note: callers should use get_file_verifier instead
330
where possible, as the underlying repository implementation may
331
have quicker access to a non-sha1 verifier.
333
:param file_id: The handle for this file.
334
:param path: The path that this file can be found at.
335
These must point to the same object.
336
:param stat_value: Optional stat value for the object
338
raise NotImplementedError(self.get_file_sha1)
340
def get_file_mtime(self, file_id, path=None):
341
"""Return the modification time for a file.
343
:param file_id: The handle for this file.
344
:param path: The path that this file can be found at.
345
These must point to the same object.
347
raise NotImplementedError(self.get_file_mtime)
349
def get_file_size(self, file_id):
350
"""Return the size of a file in bytes.
352
This applies only to regular files. If invoked on directories or
353
symlinks, it will return None.
354
:param file_id: The file-id of the file
356
raise NotImplementedError(self.get_file_size)
358
def is_executable(self, file_id, path=None):
359
"""Check if a file is executable.
361
:param file_id: The handle for this file.
362
:param path: The path that this file can be found at.
363
These must point to the same object.
365
raise NotImplementedError(self.is_executable)
367
def iter_files_bytes(self, desired_files):
368
"""Iterate through file contents.
370
Files will not necessarily be returned in the order they occur in
371
desired_files. No specific order is guaranteed.
373
Yields pairs of identifier, bytes_iterator. identifier is an opaque
374
value supplied by the caller as part of desired_files. It should
375
uniquely identify the file version in the caller's context. (Examples:
376
an index number or a TreeTransform trans_id.)
378
bytes_iterator is an iterable of bytestrings for the file. The
379
kind of iterable and length of the bytestrings are unspecified, but for
380
this implementation, it is a tuple containing a single bytestring with
381
the complete text of the file.
383
:param desired_files: a list of (file_id, identifier) pairs
385
for file_id, identifier in desired_files:
386
# We wrap the string in a tuple so that we can return an iterable
387
# of bytestrings. (Technically, a bytestring is also an iterable
388
# of bytestrings, but iterating through each character is not
390
cur_file = (self.get_file_text(file_id),)
391
yield identifier, cur_file
393
def get_symlink_target(self, file_id, path=None):
394
"""Get the target for a given file_id.
396
It is assumed that the caller already knows that file_id is referencing
398
:param file_id: Handle for the symlink entry.
399
:param path: The path of the file.
400
If both file_id and path are supplied, an implementation may use
402
:return: The path the symlink points to.
404
raise NotImplementedError(self.get_symlink_target)
406
def get_root_id(self):
407
"""Return the file_id for the root of this tree."""
408
raise NotImplementedError(self.get_root_id)
410
def annotate_iter(self, file_id,
411
default_revision=_mod_revision.CURRENT_REVISION):
412
"""Return an iterator of revision_id, line tuples.
186
414
For working trees (and mutable trees in general), the special
187
415
revision_id 'current:' will be used for lines that are new in this
188
416
tree, e.g. uncommitted changes.
189
417
:param file_id: The file to produce an annotated version from
418
:param default_revision: For lines that don't match a basis, mark them
419
with this revision id. Not all implementations will make use of
191
422
raise NotImplementedError(self.annotate_iter)
193
inventory = property(_get_inventory,
194
doc="Inventory of this Tree")
424
def _get_plan_merge_data(self, file_id, other, base):
425
from bzrlib import versionedfile
426
vf = versionedfile._PlanMergeVersionedFile(file_id)
427
last_revision_a = self._get_file_revision(file_id, vf, 'this:')
428
last_revision_b = other._get_file_revision(file_id, vf, 'other:')
430
last_revision_base = None
432
last_revision_base = base._get_file_revision(file_id, vf, 'base:')
433
return vf, last_revision_a, last_revision_b, last_revision_base
435
def plan_file_merge(self, file_id, other, base=None):
436
"""Generate a merge plan based on annotations.
438
If the file contains uncommitted changes in this tree, they will be
439
attributed to the 'current:' pseudo-revision. If the file contains
440
uncommitted changes in the other tree, they will be assigned to the
441
'other:' pseudo-revision.
443
data = self._get_plan_merge_data(file_id, other, base)
444
vf, last_revision_a, last_revision_b, last_revision_base = data
445
return vf.plan_merge(last_revision_a, last_revision_b,
448
def plan_file_lca_merge(self, file_id, other, base=None):
449
"""Generate a merge plan based lca-newness.
451
If the file contains uncommitted changes in this tree, they will be
452
attributed to the 'current:' pseudo-revision. If the file contains
453
uncommitted changes in the other tree, they will be assigned to the
454
'other:' pseudo-revision.
456
data = self._get_plan_merge_data(file_id, other, base)
457
vf, last_revision_a, last_revision_b, last_revision_base = data
458
return vf.plan_lca_merge(last_revision_a, last_revision_b,
461
def _iter_parent_trees(self):
462
"""Iterate through parent trees, defaulting to Tree.revision_tree."""
463
for revision_id in self.get_parent_ids():
465
yield self.revision_tree(revision_id)
466
except errors.NoSuchRevisionInTree:
467
yield self.repository.revision_tree(revision_id)
469
def _get_file_revision(self, file_id, vf, tree_revision):
470
"""Ensure that file_id, tree_revision is in vf to plan the merge."""
472
if getattr(self, '_repository', None) is None:
473
last_revision = tree_revision
474
parent_keys = [(file_id, t.get_file_revision(file_id)) for t in
475
self._iter_parent_trees()]
476
vf.add_lines((file_id, last_revision), parent_keys,
477
self.get_file_lines(file_id))
478
repo = self.branch.repository
481
last_revision = self.get_file_revision(file_id)
482
base_vf = self._repository.texts
483
if base_vf not in vf.fallback_versionedfiles:
484
vf.fallback_versionedfiles.append(base_vf)
196
487
def _check_retrieved(self, ie, f):
197
488
if not __debug__:
199
fp = fingerprint_file(f)
490
fp = osutils.fingerprint_file(f)
202
493
if ie.text_size is not None:
203
494
if ie.text_size != fp['size']:
204
raise BzrError("mismatched size for file %r in %r" % (ie.file_id, self._store),
495
raise errors.BzrError(
496
"mismatched size for file %r in %r" %
497
(ie.file_id, self._store),
205
498
["inventory expects %d bytes" % ie.text_size,
206
499
"file is actually %d bytes" % fp['size'],
207
500
"store is probably damaged/corrupt"])
209
502
if ie.text_sha1 != fp['sha1']:
210
raise BzrError("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
503
raise errors.BzrError("wrong SHA-1 for file %r in %r" %
504
(ie.file_id, self._store),
211
505
["inventory expects %s" % ie.text_sha1,
212
506
"file is actually %s" % fp['sha1'],
213
507
"store is probably damaged/corrupt"])
215
509
def path2id(self, path):
216
510
"""Return the id for path in this tree."""
217
return self._inventory.path2id(path)
219
def print_file(self, file_id):
220
"""Print file with id `file_id` to stdout."""
222
sys.stdout.write(self.get_file_text(file_id))
511
raise NotImplementedError(self.path2id)
513
def paths2ids(self, paths, trees=[], require_versioned=True):
514
"""Return all the ids that can be reached by walking from paths.
516
Each path is looked up in this tree and any extras provided in
517
trees, and this is repeated recursively: the children in an extra tree
518
of a directory that has been renamed under a provided path in this tree
519
are all returned, even if none exist under a provided path in this
520
tree, and vice versa.
522
:param paths: An iterable of paths to start converting to ids from.
523
Alternatively, if paths is None, no ids should be calculated and None
524
will be returned. This is offered to make calling the api unconditional
525
for code that *might* take a list of files.
526
:param trees: Additional trees to consider.
527
:param require_versioned: If False, do not raise NotVersionedError if
528
an element of paths is not versioned in this tree and all of trees.
530
return find_ids_across_trees(paths, [self] + list(trees), require_versioned)
532
def iter_children(self, file_id):
533
entry = self.iter_entries_by_dir([file_id]).next()[1]
534
for child in getattr(entry, 'children', {}).itervalues():
224
537
def lock_read(self):
538
"""Lock this tree for multiple read only operations.
540
:return: A bzrlib.lock.LogicalLockResult.
227
544
def revision_tree(self, revision_id):
294
607
raise NotImplementedError(self.walkdirs)
297
class EmptyTree(Tree):
300
self._inventory = Inventory(root_id=None)
301
symbol_versioning.warn('EmptyTree is deprecated as of bzr 0.9 please'
302
' use repository.revision_tree instead.',
303
DeprecationWarning, stacklevel=2)
305
def get_parent_ids(self):
308
def get_symlink_target(self, file_id):
311
def has_filename(self, filename):
609
def supports_content_filtering(self):
314
def kind(self, file_id):
315
assert self._inventory[file_id].kind == "directory"
318
def list_files(self, include_root=False):
321
def __contains__(self, file_id):
322
return (file_id in self._inventory)
324
def get_file_sha1(self, file_id, path=None, stat_value=None):
328
######################################################################
331
# TODO: Merge these two functions into a single one that can operate
332
# on either a whole tree or a set of files.
334
# TODO: Return the diff in order by filename, not by category or in
335
# random order. Can probably be done by lock-stepping through the
336
# filenames from both trees.
339
def file_status(filename, old_tree, new_tree):
340
"""Return single-letter status, old and new names for a file.
342
The complexity here is in deciding how to represent renames;
343
many complex cases are possible.
612
def _content_filter_stack(self, path=None, file_id=None):
613
"""The stack of content filters for a path if filtering is supported.
615
Readers will be applied in first-to-last order.
616
Writers will be applied in last-to-first order.
617
Either the path or the file-id needs to be provided.
619
:param path: path relative to the root of the tree
621
:param file_id: file_id or None if unknown
622
:return: the list of filters - [] if there are none
624
filter_pref_names = filters._get_registered_names()
625
if len(filter_pref_names) == 0:
628
path = self.id2path(file_id)
629
prefs = self.iter_search_rules([path], filter_pref_names).next()
630
stk = filters._get_filter_stack_for(prefs)
631
if 'filters' in debug.debug_flags:
632
trace.note(gettext("*** {0} content-filter: {1} => {2!r}").format(path,prefs,stk))
635
def _content_filter_stack_provider(self):
636
"""A function that returns a stack of ContentFilters.
638
The function takes a path (relative to the top of the tree) and a
639
file-id as parameters.
641
:return: None if content filtering is not supported by this tree.
643
if self.supports_content_filtering():
644
return lambda path, file_id: \
645
self._content_filter_stack(path, file_id)
649
def iter_search_rules(self, path_names, pref_names=None,
650
_default_searcher=None):
651
"""Find the preferences for filenames in a tree.
653
:param path_names: an iterable of paths to find attributes for.
654
Paths are given relative to the root of the tree.
655
:param pref_names: the list of preferences to lookup - None for all
656
:param _default_searcher: private parameter to assist testing - don't use
657
:return: an iterator of tuple sequences, one per path-name.
658
See _RulesSearcher.get_items for details on the tuple sequence.
660
if _default_searcher is None:
661
_default_searcher = rules._per_user_searcher
662
searcher = self._get_rules_searcher(_default_searcher)
663
if searcher is not None:
664
if pref_names is not None:
665
for path in path_names:
666
yield searcher.get_selected_items(path, pref_names)
668
for path in path_names:
669
yield searcher.get_items(path)
671
def _get_rules_searcher(self, default_searcher):
672
"""Get the RulesSearcher for this tree given the default one."""
673
searcher = default_searcher
677
class InventoryTree(Tree):
678
"""A tree that relies on an inventory for its metadata.
680
Trees contain an `Inventory` object, and also know how to retrieve
681
file texts mentioned in the inventory, either from a working
682
directory or from a store.
684
It is possible for trees to contain files that are not described
685
in their inventory or vice versa; for this use `filenames()`.
687
Subclasses should set the _inventory attribute, which is considered
688
private to external API users.
345
old_inv = old_tree.inventory
346
new_inv = new_tree.inventory
347
new_id = new_inv.path2id(filename)
348
old_id = old_inv.path2id(filename)
350
if not new_id and not old_id:
351
# easy: doesn't exist in either; not versioned at all
352
if new_tree.is_ignored(filename):
353
return 'I', None, None
355
return '?', None, None
357
# There is now a file of this name, great.
360
# There is no longer a file of this name, but we can describe
361
# what happened to the file that used to have
362
# this name. There are two possibilities: either it was
363
# deleted entirely, or renamed.
365
if new_inv.has_id(old_id):
366
return 'X', old_inv.id2path(old_id), new_inv.id2path(old_id)
368
return 'D', old_inv.id2path(old_id), None
370
# if the file_id is new in this revision, it is added
371
if new_id and not old_inv.has_id(new_id):
374
# if there used to be a file of this name, but that ID has now
375
# disappeared, it is deleted
376
if old_id and not new_inv.has_id(old_id):
383
def find_renames(old_inv, new_inv):
384
for file_id in old_inv:
385
if file_id not in new_inv:
387
old_name = old_inv.id2path(file_id)
388
new_name = new_inv.id2path(file_id)
389
if old_name != new_name:
390
yield (old_name, new_name)
691
def get_canonical_inventory_paths(self, paths):
692
"""Like get_canonical_inventory_path() but works on multiple items.
694
:param paths: A sequence of paths relative to the root of the tree.
695
:return: A list of paths, with each item the corresponding input path
696
adjusted to account for existing elements that match case
699
return list(self._yield_canonical_inventory_paths(paths))
701
def get_canonical_inventory_path(self, path):
702
"""Returns the first inventory item that case-insensitively matches path.
704
If a path matches exactly, it is returned. If no path matches exactly
705
but more than one path matches case-insensitively, it is implementation
706
defined which is returned.
708
If no path matches case-insensitively, the input path is returned, but
709
with as many path entries that do exist changed to their canonical
712
If you need to resolve many names from the same tree, you should
713
use get_canonical_inventory_paths() to avoid O(N) behaviour.
715
:param path: A paths relative to the root of the tree.
716
:return: The input path adjusted to account for existing elements
717
that match case insensitively.
719
return self._yield_canonical_inventory_paths([path]).next()
721
def _yield_canonical_inventory_paths(self, paths):
723
# First, if the path as specified exists exactly, just use it.
724
if self.path2id(path) is not None:
728
cur_id = self.get_root_id()
730
bit_iter = iter(path.split("/"))
734
for child in self.iter_children(cur_id):
736
# XXX: it seem like if the child is known to be in the
737
# tree, we shouldn't need to go from its id back to
738
# its path -- mbp 2010-02-11
740
# XXX: it seems like we could be more efficient
741
# by just directly looking up the original name and
742
# only then searching all children; also by not
743
# chopping paths so much. -- mbp 2010-02-11
744
child_base = os.path.basename(self.id2path(child))
745
if (child_base == elt):
746
# if we found an exact match, we can stop now; if
747
# we found an approximate match we need to keep
748
# searching because there might be an exact match
751
new_path = osutils.pathjoin(cur_path, child_base)
753
elif child_base.lower() == lelt:
755
new_path = osutils.pathjoin(cur_path, child_base)
756
except errors.NoSuchId:
757
# before a change is committed we can see this error...
762
# got to the end of this directory and no entries matched.
763
# Return what matched so far, plus the rest as specified.
764
cur_path = osutils.pathjoin(cur_path, elt, *list(bit_iter))
769
def _get_inventory(self):
770
return self._inventory
772
inventory = property(_get_inventory,
773
doc="Inventory of this Tree")
776
def path2id(self, path):
777
"""Return the id for path in this tree."""
778
return self._inventory.path2id(path)
780
def id2path(self, file_id):
781
"""Return the path for a file id.
785
return self.inventory.id2path(file_id)
787
def has_id(self, file_id):
788
return self.inventory.has_id(file_id)
790
def has_or_had_id(self, file_id):
791
return self.inventory.has_id(file_id)
793
def all_file_ids(self):
794
return set(self.inventory)
796
@deprecated_method(deprecated_in((2, 4, 0)))
798
return iter(self.inventory)
800
def filter_unversioned_files(self, paths):
801
"""Filter out paths that are versioned.
803
:return: set of paths.
805
# NB: we specifically *don't* call self.has_filename, because for
806
# WorkingTrees that can indicate files that exist on disk but that
808
pred = self.inventory.has_filename
809
return set((p for p in paths if not pred(p)))
812
def iter_entries_by_dir(self, specific_file_ids=None, yield_parents=False):
813
"""Walk the tree in 'by_dir' order.
815
This will yield each entry in the tree as a (path, entry) tuple.
816
The order that they are yielded is:
818
See Tree.iter_entries_by_dir for details.
820
:param yield_parents: If True, yield the parents from the root leading
821
down to specific_file_ids that have been requested. This has no
822
impact if specific_file_ids is None.
824
return self.inventory.iter_entries_by_dir(
825
specific_file_ids=specific_file_ids, yield_parents=yield_parents)
827
@deprecated_method(deprecated_in((2, 5, 0)))
828
def get_file_by_path(self, path):
829
return self.get_file(self.path2id(path), path)
393
832
def find_ids_across_trees(filenames, trees, require_versioned=True):
394
833
"""Find the ids corresponding to specified filenames.
396
835
All matches in all trees will be used, and all children of matched
397
836
directories will be used.
408
847
specified_path_ids = _find_ids_across_trees(filenames, trees,
409
848
require_versioned)
410
849
return _find_children_across_trees(specified_path_ids, trees)
411
# specified_ids = [id for path, id in _find_path_ids_across_trees(filenames, trees, require_versioned)]
412
# return _find_children_across_trees(specified_ids, trees)
414
def find_path_ids_across_trees(filenames, trees, require_versioned=True):
415
"""Find the paths and ids corresponding to specified filenames.
417
All matches in all trees will be used, and all children of matched
418
directories will be included
420
:param filenames: The filenames to find file_ids for
421
:param trees: The trees to find file_ids within
422
:param require_versioned: if true, all specified filenames must occur in
424
:return: a set of (path, file ids) for the specified filenames and their
425
children. The returned path is the path of the id in the first tree
426
that contains it. This matters when files have been moved
430
# This function needs to know the ids for filenames in all trees, then
431
# search for those same files and children in all the other trees.
432
# it is complicated by the same path in two trees being able to have
433
# different ids, which might both be present in both trees.
434
# consider two trees, which have had 'mv foo bar' and 'mv baz foo' done
435
# in this case, a diff of 'foo' should should changes to both the current
436
# 'bar' and the current 'foo' which was baz. Its arguable that if
437
# the situation is 'mv parent/foo bar' and 'mv baz parent/foo', that
438
# we should return the current bar and the current parent/foo' - at the
439
# moment we do, but we loop around all ids and all trees: I*T checks.
441
# Updating this algorithm to be fast in the common case:
442
# nothing has moved, all files have the same id in parent, child and there
443
# are only two trees (or one is working tree and the others are parents).
444
# walk the dirstate. as we find each path, gather the paths of that
445
# id in all trees. add a mapping from the id to the path in those trees.
446
# now lookup children by id, again in all trees; for these trees that
447
# nothing has moved in, the id->path mapping will allow us to find the
448
# parent trivially. To answer 'has anything been moved' in one of the
449
# dirstate parent trees though, we will need to stare harder at it.
451
# Now, given a path index, that is trivial for any one tree, and given
452
# that we can ask for additional data from a dirstate tree, its a single
453
# pass, though it will require scanning the entire tree to find paths
454
# that were at the current location.
455
# ideal results?: There are three things: tree, path, id. Pathologically
456
# we can have completely disjoint ids for each tree; but we cannot have
457
# disjoin paths for each tree, except if we scan each tree for the
458
# different ids from other trees.
460
specified_path_ids = _find_ids_across_trees(filenames, trees,
462
return _find_path_id_children_across_trees(specified_path_ids, trees)
465
852
def _find_ids_across_trees(filenames, trees, require_versioned):
466
853
"""Find the ids corresponding to specified filenames.
468
855
All matches in all trees will be used, but subdirectories are not scanned.
470
857
:param filenames: The filenames to find file_ids for
471
858
:param trees: The trees to find file_ids within
472
859
:param require_versioned: if true, all specified filenames must occur in
473
860
at least one tree.
474
:return: a set of (path, file ids) for the specified filenames
861
:return: a set of file ids for the specified filenames
476
863
not_versioned = []
477
864
interesting_ids = set()
522
908
Its instances have methods like 'compare' and contain references to the
523
909
source and target trees these operations are to be carried out on.
525
clients of bzrlib should not need to use InterTree directly, rather they
911
Clients of bzrlib should not need to use InterTree directly, rather they
526
912
should use the convenience methods on Tree such as 'Tree.compare()' which
527
913
will pass through to InterTree as appropriate.
916
# Formats that will be used to test this InterTree. If both are
917
# None, this InterTree will not be tested (e.g. because a complex
919
_matching_from_tree_format = None
920
_matching_to_tree_format = None
925
def is_compatible(kls, source, target):
926
# The default implementation is naive and uses the public API, so
927
# it works for all trees.
930
def _changes_from_entries(self, source_entry, target_entry,
931
source_path=None, target_path=None):
932
"""Generate a iter_changes tuple between source_entry and target_entry.
934
:param source_entry: An inventory entry from self.source, or None.
935
:param target_entry: An inventory entry from self.target, or None.
936
:param source_path: The path of source_entry, if known. If not known
937
it will be looked up.
938
:param target_path: The path of target_entry, if known. If not known
939
it will be looked up.
940
:return: A tuple, item 0 of which is an iter_changes result tuple, and
941
item 1 is True if there are any changes in the result tuple.
943
if source_entry is None:
944
if target_entry is None:
946
file_id = target_entry.file_id
948
file_id = source_entry.file_id
949
if source_entry is not None:
950
source_versioned = True
951
source_name = source_entry.name
952
source_parent = source_entry.parent_id
953
if source_path is None:
954
source_path = self.source.id2path(file_id)
955
source_kind, source_executable, source_stat = \
956
self.source._comparison_data(source_entry, source_path)
958
source_versioned = False
962
source_executable = None
963
if target_entry is not None:
964
target_versioned = True
965
target_name = target_entry.name
966
target_parent = target_entry.parent_id
967
if target_path is None:
968
target_path = self.target.id2path(file_id)
969
target_kind, target_executable, target_stat = \
970
self.target._comparison_data(target_entry, target_path)
972
target_versioned = False
976
target_executable = None
977
versioned = (source_versioned, target_versioned)
978
kind = (source_kind, target_kind)
979
changed_content = False
980
if source_kind != target_kind:
981
changed_content = True
982
elif source_kind == 'file':
983
if not self.file_content_matches(file_id, file_id, source_path,
984
target_path, source_stat, target_stat):
985
changed_content = True
986
elif source_kind == 'symlink':
987
if (self.source.get_symlink_target(file_id) !=
988
self.target.get_symlink_target(file_id)):
989
changed_content = True
990
# XXX: Yes, the indentation below is wrong. But fixing it broke
991
# test_merge.TestMergerEntriesLCAOnDisk.
992
# test_nested_tree_subtree_renamed_and_modified. We'll wait for
993
# the fix from bzr.dev -- vila 2009026
994
elif source_kind == 'tree-reference':
995
if (self.source.get_reference_revision(file_id, source_path)
996
!= self.target.get_reference_revision(file_id, target_path)):
997
changed_content = True
998
parent = (source_parent, target_parent)
999
name = (source_name, target_name)
1000
executable = (source_executable, target_executable)
1001
if (changed_content is not False or versioned[0] != versioned[1]
1002
or parent[0] != parent[1] or name[0] != name[1] or
1003
executable[0] != executable[1]):
1007
return (file_id, (source_path, target_path), changed_content,
1008
versioned, parent, name, kind, executable), changes
532
1010
@needs_read_lock
533
1011
def compare(self, want_unchanged=False, specific_files=None,
534
extra_trees=None, require_versioned=False, include_root=False):
1012
extra_trees=None, require_versioned=False, include_root=False,
1013
want_unversioned=False):
535
1014
"""Return the changes from source to target.
537
1016
:return: A TreeDelta.
546
1025
:param require_versioned: An optional boolean (defaults to False). When
547
1026
supplied and True all the 'specific_files' must be versioned, or
548
1027
a PathsNotVersionedError will be thrown.
1028
:param want_unversioned: Scan for unversioned paths.
550
# NB: show_status depends on being able to pass in non-versioned files
551
# and report them as unknown
552
trees = (self.source, self.target)
1030
trees = (self.source,)
553
1031
if extra_trees is not None:
554
1032
trees = trees + tuple(extra_trees)
555
specific_file_ids = find_ids_across_trees(specific_files,
556
trees, require_versioned=require_versioned)
1033
# target is usually the newer tree:
1034
specific_file_ids = self.target.paths2ids(specific_files, trees,
1035
require_versioned=require_versioned)
557
1036
if specific_files and not specific_file_ids:
558
1037
# All files are unversioned, so just return an empty delta
559
1038
# _compare_trees would think we want a complete delta
560
return delta.TreeDelta()
1039
result = delta.TreeDelta()
1040
fake_entry = inventory.InventoryFile('unused', 'unused', 'unused')
1041
result.unversioned = [(path, None,
1042
self.target._comparison_data(fake_entry, path)[0]) for path in
561
1045
return delta._compare_trees(self.source, self.target, want_unchanged,
562
specific_file_ids, include_root)
1046
specific_files, include_root, extra_trees=extra_trees,
1047
require_versioned=require_versioned,
1048
want_unversioned=want_unversioned)
564
def _iter_changes(self, from_tree, to_tree, include_unchanged,
565
specific_file_ids, pb):
1050
def iter_changes(self, include_unchanged=False,
1051
specific_files=None, pb=None, extra_trees=[],
1052
require_versioned=True, want_unversioned=False):
566
1053
"""Generate an iterator of changes between trees.
568
1055
A tuple is returned:
569
(file_id, path, changed_content, versioned, parent, name, kind,
1056
(file_id, (path_in_source, path_in_target),
1057
changed_content, versioned, parent, name, kind,
572
Path is relative to the to_tree. changed_content is True if the file's
573
content has changed. This includes changes to its kind, and to
1060
Changed_content is True if the file's content has changed. This
1061
includes changes to its kind, and to a symlink's target.
576
1063
versioned, parent, name, kind, executable are tuples of (from, to).
577
1064
If a file is missing in a tree, its kind is None.
579
Iteration is done in parent-to-child order, relative to the to_tree.
1066
Iteration is done in parent-to-child order, relative to the target
1069
There is no guarantee that all paths are in sorted order: the
1070
requirement to expand the search due to renames may result in children
1071
that should be found early being found late in the search, after
1072
lexically later results have been returned.
1073
:param require_versioned: Raise errors.PathsNotVersionedError if a
1074
path in the specific_files list is not versioned in one of
1075
source, target or extra_trees.
1076
:param specific_files: An optional list of file paths to restrict the
1077
comparison to. When mapping filenames to ids, all matches in all
1078
trees (including optional extra_trees) are used, and all children
1079
of matched directories are included. The parents in the target tree
1080
of the specific files up to and including the root of the tree are
1081
always evaluated for changes too.
1082
:param want_unversioned: Should unversioned files be returned in the
1083
output. An unversioned file is defined as one with (False, False)
1084
for the versioned pair.
1086
lookup_trees = [self.source]
1088
lookup_trees.extend(extra_trees)
1089
# The ids of items we need to examine to insure delta consistency.
1090
precise_file_ids = set()
1091
changed_file_ids = []
1092
if specific_files == []:
1093
specific_file_ids = []
1095
specific_file_ids = self.target.paths2ids(specific_files,
1096
lookup_trees, require_versioned=require_versioned)
1097
if specific_files is not None:
1098
# reparented or added entries must have their parents included
1099
# so that valid deltas can be created. The seen_parents set
1100
# tracks the parents that we need to have.
1101
# The seen_dirs set tracks directory entries we've yielded.
1102
# After outputting version object in to_entries we set difference
1103
# the two seen sets and start checking parents.
1104
seen_parents = set()
1106
if want_unversioned:
1107
all_unversioned = sorted([(p.split('/'), p) for p in
1108
self.target.extras()
1109
if specific_files is None or
1110
osutils.is_inside_any(specific_files, p)])
1111
all_unversioned = collections.deque(all_unversioned)
1113
all_unversioned = collections.deque()
582
from_entries_by_dir = list(from_tree.inventory.iter_entries_by_dir(
1115
from_entries_by_dir = list(self.source.iter_entries_by_dir(
583
1116
specific_file_ids=specific_file_ids))
584
1117
from_data = dict((e.file_id, (p, e)) for p, e in from_entries_by_dir)
585
to_entries_by_dir = list(to_tree.inventory.iter_entries_by_dir(
1118
to_entries_by_dir = list(self.target.iter_entries_by_dir(
586
1119
specific_file_ids=specific_file_ids))
587
1120
num_entries = len(from_entries_by_dir) + len(to_entries_by_dir)
589
for to_path, to_entry in to_entries_by_dir:
590
file_id = to_entry.file_id
591
to_paths[file_id] = to_path
1122
# the unversioned path lookup only occurs on real trees - where there
1123
# can be extras. So the fake_entry is solely used to look up
1124
# executable it values when execute is not supported.
1125
fake_entry = inventory.InventoryFile('unused', 'unused', 'unused')
1126
for target_path, target_entry in to_entries_by_dir:
1127
while (all_unversioned and
1128
all_unversioned[0][0] < target_path.split('/')):
1129
unversioned_path = all_unversioned.popleft()
1130
target_kind, target_executable, target_stat = \
1131
self.target._comparison_data(fake_entry, unversioned_path[1])
1132
yield (None, (None, unversioned_path[1]), True, (False, False),
1134
(None, unversioned_path[0][-1]),
1135
(None, target_kind),
1136
(None, target_executable))
1137
source_path, source_entry = from_data.get(target_entry.file_id,
1139
result, changes = self._changes_from_entries(source_entry,
1140
target_entry, source_path=source_path, target_path=target_path)
1141
to_paths[result[0]] = result[1][1]
592
1142
entry_count += 1
593
changed_content = False
594
from_path, from_entry = from_data.get(file_id, (None, None))
595
from_versioned = (from_entry is not None)
596
if from_entry is not None:
597
from_versioned = True
598
from_name = from_entry.name
599
from_parent = from_entry.parent_id
600
from_kind, from_executable, from_stat = \
601
from_tree._comparison_data(from_entry, from_path)
602
1144
entry_count += 1
604
from_versioned = False
608
from_executable = None
609
versioned = (from_versioned, True)
1146
pb.update('comparing files', entry_count, num_entries)
1147
if changes or include_unchanged:
1148
if specific_file_ids is not None:
1149
new_parent_id = result[4][1]
1150
precise_file_ids.add(new_parent_id)
1151
changed_file_ids.append(result[0])
1153
# Ensure correct behaviour for reparented/added specific files.
1154
if specific_files is not None:
1155
# Record output dirs
1156
if result[6][1] == 'directory':
1157
seen_dirs.add(result[0])
1158
# Record parents of reparented/added entries.
1159
versioned = result[3]
1161
if not versioned[0] or parents[0] != parents[1]:
1162
seen_parents.add(parents[1])
1163
while all_unversioned:
1164
# yield any trailing unversioned paths
1165
unversioned_path = all_unversioned.popleft()
610
1166
to_kind, to_executable, to_stat = \
611
to_tree._comparison_data(to_entry, to_path)
612
kind = (from_kind, to_kind)
613
if kind[0] != kind[1]:
614
changed_content = True
615
elif from_kind == 'file':
616
from_size = from_tree._file_size(from_entry, from_stat)
617
to_size = to_tree._file_size(to_entry, to_stat)
618
if from_size != to_size:
619
changed_content = True
620
elif (from_tree.get_file_sha1(file_id, from_path, from_stat) !=
621
to_tree.get_file_sha1(file_id, to_path, to_stat)):
622
changed_content = True
623
elif from_kind == 'symlink':
624
if (from_tree.get_symlink_target(file_id) !=
625
to_tree.get_symlink_target(file_id)):
626
changed_content = True
627
parent = (from_parent, to_entry.parent_id)
628
name = (from_name, to_entry.name)
629
executable = (from_executable, to_executable)
631
pb.update('comparing files', entry_count, num_entries)
632
if (changed_content is not False or versioned[0] != versioned[1]
633
or parent[0] != parent[1] or name[0] != name[1] or
634
executable[0] != executable[1] or include_unchanged):
635
yield (file_id, to_path, changed_content, versioned, parent,
636
name, kind, executable)
638
def get_to_path(from_entry):
639
if from_entry.parent_id is None:
642
if from_entry.parent_id not in to_paths:
643
get_to_path(from_tree.inventory[from_entry.parent_id])
644
to_path = osutils.pathjoin(to_paths[from_entry.parent_id],
646
to_paths[from_entry.file_id] = to_path
1167
self.target._comparison_data(fake_entry, unversioned_path[1])
1168
yield (None, (None, unversioned_path[1]), True, (False, False),
1170
(None, unversioned_path[0][-1]),
1172
(None, to_executable))
1173
# Yield all remaining source paths
649
1174
for path, from_entry in from_entries_by_dir:
650
1175
file_id = from_entry.file_id
651
1176
if file_id in to_paths:
653
to_path = get_to_path(from_entry)
1179
if not self.target.has_id(file_id):
1180
# common case - paths we have not emitted are not present in
1184
to_path = self.target.id2path(file_id)
654
1185
entry_count += 1
655
1186
if pb is not None:
656
1187
pb.update('comparing files', entry_count, num_entries)
658
1189
parent = (from_entry.parent_id, None)
659
1190
name = (from_entry.name, None)
660
1191
from_kind, from_executable, stat_value = \
661
from_tree._comparison_data(from_entry, path)
1192
self.source._comparison_data(from_entry, path)
662
1193
kind = (from_kind, None)
663
1194
executable = (from_executable, None)
664
changed_content = True
1195
changed_content = from_kind is not None
665
1196
# the parent's path is necessarily known at this point.
666
yield(file_id, to_path, changed_content, versioned, parent,
1197
changed_file_ids.append(file_id)
1198
yield(file_id, (path, to_path), changed_content, versioned, parent,
667
1199
name, kind, executable)
670
# This was deprecated before 0.12, but did not have an official warning
671
@symbol_versioning.deprecated_function(symbol_versioning.zero_twelve)
672
def RevisionTree(*args, **kwargs):
673
"""RevisionTree has moved to bzrlib.revisiontree.RevisionTree()
675
Accessing it as bzrlib.tree.RevisionTree has been deprecated as of
678
from bzrlib.revisiontree import RevisionTree as _RevisionTree
679
return _RevisionTree(*args, **kwargs)
1200
changed_file_ids = set(changed_file_ids)
1201
if specific_file_ids is not None:
1202
for result in self._handle_precise_ids(precise_file_ids,
1206
def _get_entry(self, tree, file_id):
1207
"""Get an inventory entry from a tree, with missing entries as None.
1209
If the tree raises NotImplementedError on accessing .inventory, then
1210
this is worked around using iter_entries_by_dir on just the file id
1213
:param tree: The tree to lookup the entry in.
1214
:param file_id: The file_id to lookup.
1217
inventory = tree.inventory
1218
except NotImplementedError:
1219
# No inventory available.
1221
iterator = tree.iter_entries_by_dir(specific_file_ids=[file_id])
1222
return iterator.next()[1]
1223
except StopIteration:
1227
return inventory[file_id]
1228
except errors.NoSuchId:
1231
def _handle_precise_ids(self, precise_file_ids, changed_file_ids,
1232
discarded_changes=None):
1233
"""Fill out a partial iter_changes to be consistent.
1235
:param precise_file_ids: The file ids of parents that were seen during
1237
:param changed_file_ids: The file ids of already emitted items.
1238
:param discarded_changes: An optional dict of precalculated
1239
iter_changes items which the partial iter_changes had not output
1241
:return: A generator of iter_changes items to output.
1243
# process parents of things that had changed under the users
1244
# requested paths to prevent incorrect paths or parent ids which
1245
# aren't in the tree.
1246
while precise_file_ids:
1247
precise_file_ids.discard(None)
1248
# Don't emit file_ids twice
1249
precise_file_ids.difference_update(changed_file_ids)
1250
if not precise_file_ids:
1252
# If the there was something at a given output path in source, we
1253
# have to include the entry from source in the delta, or we would
1254
# be putting this entry into a used path.
1256
for parent_id in precise_file_ids:
1258
paths.append(self.target.id2path(parent_id))
1259
except errors.NoSuchId:
1260
# This id has been dragged in from the source by delta
1261
# expansion and isn't present in target at all: we don't
1262
# need to check for path collisions on it.
1265
old_id = self.source.path2id(path)
1266
precise_file_ids.add(old_id)
1267
precise_file_ids.discard(None)
1268
current_ids = precise_file_ids
1269
precise_file_ids = set()
1270
# We have to emit all of precise_file_ids that have been altered.
1271
# We may have to output the children of some of those ids if any
1272
# directories have stopped being directories.
1273
for file_id in current_ids:
1275
if discarded_changes:
1276
result = discarded_changes.get(file_id)
1281
old_entry = self._get_entry(self.source, file_id)
1282
new_entry = self._get_entry(self.target, file_id)
1283
result, changes = self._changes_from_entries(
1284
old_entry, new_entry)
1287
# Get this parents parent to examine.
1288
new_parent_id = result[4][1]
1289
precise_file_ids.add(new_parent_id)
1291
if (result[6][0] == 'directory' and
1292
result[6][1] != 'directory'):
1293
# This stopped being a directory, the old children have
1295
if old_entry is None:
1296
# Reusing a discarded change.
1297
old_entry = self._get_entry(self.source, file_id)
1298
for child in old_entry.children.values():
1299
precise_file_ids.add(child.file_id)
1300
changed_file_ids.add(result[0])
1304
def file_content_matches(self, source_file_id, target_file_id,
1305
source_path=None, target_path=None, source_stat=None, target_stat=None):
1306
"""Check if two files are the same in the source and target trees.
1308
This only checks that the contents of the files are the same,
1309
it does not touch anything else.
1311
:param source_file_id: File id of the file in the source tree
1312
:param target_file_id: File id of the file in the target tree
1313
:param source_path: Path of the file in the source tree
1314
:param target_path: Path of the file in the target tree
1315
:param source_stat: Optional stat value of the file in the source tree
1316
:param target_stat: Optional stat value of the file in the target tree
1317
:return: Boolean indicating whether the files have the same contents
1319
source_verifier_kind, source_verifier_data = self.source.get_file_verifier(
1320
source_file_id, source_path, source_stat)
1321
target_verifier_kind, target_verifier_data = self.target.get_file_verifier(
1322
target_file_id, target_path, target_stat)
1323
if source_verifier_kind == target_verifier_kind:
1324
return (source_verifier_data == target_verifier_data)
1325
# Fall back to SHA1 for now
1326
if source_verifier_kind != "SHA1":
1327
source_sha1 = self.source.get_file_sha1(source_file_id,
1328
source_path, source_stat)
1330
source_sha1 = source_verifier_data
1331
if target_verifier_kind != "SHA1":
1332
target_sha1 = self.target.get_file_sha1(target_file_id,
1333
target_path, target_stat)
1335
target_sha1 = target_verifier_data
1336
return (source_sha1 == target_sha1)
1338
InterTree.register_optimiser(InterTree)
1341
class MultiWalker(object):
1342
"""Walk multiple trees simultaneously, getting combined results."""
1344
# Note: This could be written to not assume you can do out-of-order
1345
# lookups. Instead any nodes that don't match in all trees could be
1346
# marked as 'deferred', and then returned in the final cleanup loop.
1347
# For now, I think it is "nicer" to return things as close to the
1348
# "master_tree" order as we can.
1350
def __init__(self, master_tree, other_trees):
1351
"""Create a new MultiWalker.
1353
All trees being walked must implement "iter_entries_by_dir()", such
1354
that they yield (path, object) tuples, where that object will have a
1355
'.file_id' member, that can be used to check equality.
1357
:param master_tree: All trees will be 'slaved' to the master_tree such
1358
that nodes in master_tree will be used as 'first-pass' sync points.
1359
Any nodes that aren't in master_tree will be merged in a second
1361
:param other_trees: A list of other trees to walk simultaneously.
1363
self._master_tree = master_tree
1364
self._other_trees = other_trees
1366
# Keep track of any nodes that were properly processed just out of
1367
# order, that way we don't return them at the end, we don't have to
1368
# track *all* processed file_ids, just the out-of-order ones
1369
self._out_of_order_processed = set()
1372
def _step_one(iterator):
1373
"""Step an iter_entries_by_dir iterator.
1375
:return: (has_more, path, ie)
1376
If has_more is False, path and ie will be None.
1379
path, ie = iterator.next()
1380
except StopIteration:
1381
return False, None, None
1383
return True, path, ie
1386
def _cmp_path_by_dirblock(path1, path2):
1387
"""Compare two paths based on what directory they are in.
1389
This generates a sort order, such that all children of a directory are
1390
sorted together, and grandchildren are in the same order as the
1391
children appear. But all grandchildren come after all children.
1393
:param path1: first path
1394
:param path2: the second path
1395
:return: negative number if ``path1`` comes first,
1396
0 if paths are equal
1397
and a positive number if ``path2`` sorts first
1399
# Shortcut this special case
1402
# This is stolen from _dirstate_helpers_py.py, only switching it to
1403
# Unicode objects. Consider using encode_utf8() and then using the
1404
# optimized versions, or maybe writing optimized unicode versions.
1405
if not isinstance(path1, unicode):
1406
raise TypeError("'path1' must be a unicode string, not %s: %r"
1407
% (type(path1), path1))
1408
if not isinstance(path2, unicode):
1409
raise TypeError("'path2' must be a unicode string, not %s: %r"
1410
% (type(path2), path2))
1411
return cmp(MultiWalker._path_to_key(path1),
1412
MultiWalker._path_to_key(path2))
1415
def _path_to_key(path):
1416
dirname, basename = osutils.split(path)
1417
return (dirname.split(u'/'), basename)
1419
def _lookup_by_file_id(self, extra_entries, other_tree, file_id):
1420
"""Lookup an inventory entry by file_id.
1422
This is called when an entry is missing in the normal order.
1423
Generally this is because a file was either renamed, or it was
1424
deleted/added. If the entry was found in the inventory and not in
1425
extra_entries, it will be added to self._out_of_order_processed
1427
:param extra_entries: A dictionary of {file_id: (path, ie)}. This
1428
should be filled with entries that were found before they were
1429
used. If file_id is present, it will be removed from the
1431
:param other_tree: The Tree to search, in case we didn't find the entry
1433
:param file_id: The file_id to look for
1434
:return: (path, ie) if found or (None, None) if not present.
1436
if file_id in extra_entries:
1437
return extra_entries.pop(file_id)
1438
# TODO: Is id2path better as the first call, or is
1439
# inventory[file_id] better as a first check?
1441
cur_path = other_tree.id2path(file_id)
1442
except errors.NoSuchId:
1444
if cur_path is None:
1447
self._out_of_order_processed.add(file_id)
1448
cur_ie = other_tree.inventory[file_id]
1449
return (cur_path, cur_ie)
1452
"""Match up the values in the different trees."""
1453
for result in self._walk_master_tree():
1455
self._finish_others()
1456
for result in self._walk_others():
1459
def _walk_master_tree(self):
1460
"""First pass, walk all trees in lock-step.
1462
When we are done, all nodes in the master_tree will have been
1463
processed. _other_walkers, _other_entries, and _others_extra will be
1464
set on 'self' for future processing.
1466
# This iterator has the most "inlining" done, because it tends to touch
1467
# every file in the tree, while the others only hit nodes that don't
1469
master_iterator = self._master_tree.iter_entries_by_dir()
1471
other_walkers = [other.iter_entries_by_dir()
1472
for other in self._other_trees]
1473
other_entries = [self._step_one(walker) for walker in other_walkers]
1474
# Track extra nodes in the other trees
1475
others_extra = [{} for i in xrange(len(self._other_trees))]
1477
master_has_more = True
1478
step_one = self._step_one
1479
lookup_by_file_id = self._lookup_by_file_id
1480
out_of_order_processed = self._out_of_order_processed
1482
while master_has_more:
1483
(master_has_more, path, master_ie) = step_one(master_iterator)
1484
if not master_has_more:
1487
file_id = master_ie.file_id
1489
other_values_append = other_values.append
1490
next_other_entries = []
1491
next_other_entries_append = next_other_entries.append
1492
for idx, (other_has_more, other_path, other_ie) in enumerate(other_entries):
1493
if not other_has_more:
1494
other_values_append(lookup_by_file_id(
1495
others_extra[idx], self._other_trees[idx], file_id))
1496
next_other_entries_append((False, None, None))
1497
elif file_id == other_ie.file_id:
1498
# This is the critical code path, as most of the entries
1499
# should match between most trees.
1500
other_values_append((other_path, other_ie))
1501
next_other_entries_append(step_one(other_walkers[idx]))
1503
# This walker did not match, step it until it either
1504
# matches, or we know we are past the current walker.
1505
other_walker = other_walkers[idx]
1506
other_extra = others_extra[idx]
1507
while (other_has_more and
1508
self._cmp_path_by_dirblock(other_path, path) < 0):
1509
other_file_id = other_ie.file_id
1510
if other_file_id not in out_of_order_processed:
1511
other_extra[other_file_id] = (other_path, other_ie)
1512
other_has_more, other_path, other_ie = \
1513
step_one(other_walker)
1514
if other_has_more and other_ie.file_id == file_id:
1515
# We ended up walking to this point, match and step
1517
other_values_append((other_path, other_ie))
1518
other_has_more, other_path, other_ie = \
1519
step_one(other_walker)
1521
# This record isn't in the normal order, see if it
1523
other_values_append(lookup_by_file_id(
1524
other_extra, self._other_trees[idx], file_id))
1525
next_other_entries_append((other_has_more, other_path,
1527
other_entries = next_other_entries
1529
# We've matched all the walkers, yield this datapoint
1530
yield path, file_id, master_ie, other_values
1531
self._other_walkers = other_walkers
1532
self._other_entries = other_entries
1533
self._others_extra = others_extra
1535
def _finish_others(self):
1536
"""Finish walking the other iterators, so we get all entries."""
1537
for idx, info in enumerate(self._other_entries):
1538
other_extra = self._others_extra[idx]
1539
(other_has_more, other_path, other_ie) = info
1540
while other_has_more:
1541
other_file_id = other_ie.file_id
1542
if other_file_id not in self._out_of_order_processed:
1543
other_extra[other_file_id] = (other_path, other_ie)
1544
other_has_more, other_path, other_ie = \
1545
self._step_one(self._other_walkers[idx])
1546
del self._other_entries
1548
def _walk_others(self):
1549
"""Finish up by walking all the 'deferred' nodes."""
1550
# TODO: One alternative would be to grab all possible unprocessed
1551
# file_ids, and then sort by path, and then yield them. That
1552
# might ensure better ordering, in case a caller strictly
1553
# requires parents before children.
1554
for idx, other_extra in enumerate(self._others_extra):
1555
others = sorted(other_extra.itervalues(),
1556
key=lambda x: self._path_to_key(x[0]))
1557
for other_path, other_ie in others:
1558
file_id = other_ie.file_id
1559
# We don't need to check out_of_order_processed here, because
1560
# the lookup_by_file_id will be removing anything processed
1561
# from the extras cache
1562
other_extra.pop(file_id)
1563
other_values = [(None, None) for i in xrange(idx)]
1564
other_values.append((other_path, other_ie))
1565
for alt_idx, alt_extra in enumerate(self._others_extra[idx+1:]):
1566
alt_idx = alt_idx + idx + 1
1567
alt_extra = self._others_extra[alt_idx]
1568
alt_tree = self._other_trees[alt_idx]
1569
other_values.append(self._lookup_by_file_id(
1570
alt_extra, alt_tree, file_id))
1571
yield other_path, file_id, None, other_values
added
removed
bzrlib/tree.py
