81
93
specific_files=specific_files,
82
94
extra_trees=extra_trees,
83
95
require_versioned=require_versioned,
96
include_root=include_root,
97
want_unversioned=want_unversioned,
100
def iter_changes(self, from_tree, include_unchanged=False,
101
specific_files=None, pb=None, extra_trees=None,
102
require_versioned=True, want_unversioned=False):
103
intertree = InterTree.get(from_tree, self)
104
return intertree.iter_changes(include_unchanged, specific_files, pb,
105
extra_trees, require_versioned, want_unversioned=want_unversioned)
86
107
def conflicts(self):
87
108
"""Get a list of the conflicts in the tree.
89
110
Each conflict is an instance of bzrlib.conflicts.Conflict.
112
return _mod_conflicts.ConflictList()
115
"""For trees that can have unversioned files, return all such paths."""
93
118
def get_parent_ids(self):
94
"""Get the parent ids for this tree.
119
"""Get the parent ids for this tree.
96
121
:return: a list of parent ids. [] is returned to indicate
97
122
a tree with no parents.
98
123
:raises: BzrError if the parents are not known.
100
125
raise NotImplementedError(self.get_parent_ids)
102
127
def has_filename(self, filename):
103
128
"""True if the tree has given filename."""
104
raise NotImplementedError()
129
raise NotImplementedError(self.has_filename)
106
131
def has_id(self, file_id):
107
132
return self.inventory.has_id(file_id)
109
__contains__ = has_id
134
def __contains__(self, file_id):
135
return self.has_id(file_id)
111
137
def has_or_had_id(self, file_id):
112
138
if file_id == self.inventory.root.file_id:
114
140
return self.inventory.has_id(file_id)
142
def is_ignored(self, filename):
143
"""Check whether the filename is ignored by this tree.
145
:param filename: The relative filename within the tree.
146
:return: True if the filename is ignored.
116
150
def __iter__(self):
117
151
return iter(self.inventory)
153
def all_file_ids(self):
154
"""Iterate through all file ids, including ids for missing files."""
155
return set(self.inventory)
119
157
def id2path(self, file_id):
158
"""Return the path for a file id.
120
162
return self.inventory.id2path(file_id)
122
def iter_entries_by_dir(self):
164
def is_control_filename(self, filename):
165
"""True if filename is the name of a control file in this tree.
167
:param filename: A filename within the tree. This is a relative path
168
from the root of this tree.
170
This is true IF and ONLY IF the filename is part of the meta data
171
that bzr controls in this tree. I.E. a random .bzr directory placed
172
on disk will not be a control file for this tree.
174
return self.bzrdir.is_control_filename(filename)
177
def iter_entries_by_dir(self, specific_file_ids=None):
123
178
"""Walk the tree in 'by_dir' order.
125
This will yield each entry in the tree as a (path, entry) tuple. The
126
order that they are yielded is: the contents of a directory are
127
preceeded by the parent of a directory, and all the contents of a
128
directory are grouped together.
180
This will yield each entry in the tree as a (path, entry) tuple.
181
The order that they are yielded is:
183
Directories are walked in a depth-first lexicographical order,
184
however, whenever a directory is reached, all of its direct child
185
nodes are yielded in lexicographical order before yielding the
188
For example, in the tree::
198
The yield order (ignoring root) would be::
199
a, f, a/b, a/d, a/b/c, a/d/e, f/g
130
return self.inventory.iter_entries_by_dir()
201
return self.inventory.iter_entries_by_dir(
202
specific_file_ids=specific_file_ids)
204
def iter_references(self):
205
for path, entry in self.iter_entries_by_dir():
206
if entry.kind == 'tree-reference':
207
yield path, entry.file_id
132
209
def kind(self, file_id):
133
raise NotImplementedError("subclasses must implement kind")
210
raise NotImplementedError("Tree subclass %s must implement kind"
211
% self.__class__.__name__)
213
def stored_kind(self, file_id):
214
"""File kind stored for this file_id.
216
May not match kind on disk for working trees. Always available
217
for versioned files, even when the file itself is missing.
219
return self.kind(file_id)
221
def path_content_summary(self, path):
222
"""Get a summary of the information about path.
224
:param path: A relative path within the tree.
225
:return: A tuple containing kind, size, exec, sha1-or-link.
226
Kind is always present (see tree.kind()).
227
size is present if kind is file, None otherwise.
228
exec is None unless kind is file and the platform supports the 'x'
230
sha1-or-link is the link target if kind is symlink, or the sha1 if
231
it can be obtained without reading the file.
233
raise NotImplementedError(self.path_content_summary)
235
def get_reference_revision(self, file_id, path=None):
236
raise NotImplementedError("Tree subclass %s must implement "
237
"get_reference_revision"
238
% self.__class__.__name__)
240
def _comparison_data(self, entry, path):
241
"""Return a tuple of kind, executable, stat_value for a file.
243
entry may be None if there is no inventory entry for the file, but
244
path must always be supplied.
246
kind is None if there is no file present (even if an inventory id is
247
present). executable is False for non-file entries.
249
raise NotImplementedError(self._comparison_data)
251
def _file_size(self, entry, stat_value):
252
raise NotImplementedError(self._file_size)
135
254
def _get_inventory(self):
136
255
return self._inventory
257
def get_file(self, file_id, path=None):
258
"""Return a file object for the file file_id in the tree.
260
If both file_id and path are defined, it is implementation defined as
261
to which one is used.
263
raise NotImplementedError(self.get_file)
265
def get_file_text(self, file_id, path=None):
266
"""Return the byte content of a file.
268
:param file_id: The file_id of the file.
269
:param path: The path of the file.
270
If both file_id and path are supplied, an implementation may use
273
my_file = self.get_file(file_id, path)
275
return my_file.read()
279
def get_file_lines(self, file_id, path=None):
280
"""Return the content of a file, as lines.
282
:param file_id: The file_id of the file.
283
:param path: The path of the file.
284
If both file_id and path are supplied, an implementation may use
287
return osutils.split_lines(self.get_file_text(file_id, path))
289
def get_file_mtime(self, file_id, path=None):
290
"""Return the modification time for a file.
292
:param file_id: The handle for this file.
293
:param path: The path that this file can be found at.
294
These must point to the same object.
296
raise NotImplementedError(self.get_file_mtime)
298
def get_file_size(self, file_id):
299
"""Return the size of a file in bytes.
301
This applies only to regular files. If invoked on directories or
302
symlinks, it will return None.
303
:param file_id: The file-id of the file
305
raise NotImplementedError(self.get_file_size)
138
307
def get_file_by_path(self, path):
139
return self.get_file(self._inventory.path2id(path))
308
return self.get_file(self._inventory.path2id(path), path)
310
def iter_files_bytes(self, desired_files):
311
"""Iterate through file contents.
313
Files will not necessarily be returned in the order they occur in
314
desired_files. No specific order is guaranteed.
316
Yields pairs of identifier, bytes_iterator. identifier is an opaque
317
value supplied by the caller as part of desired_files. It should
318
uniquely identify the file version in the caller's context. (Examples:
319
an index number or a TreeTransform trans_id.)
321
bytes_iterator is an iterable of bytestrings for the file. The
322
kind of iterable and length of the bytestrings are unspecified, but for
323
this implementation, it is a tuple containing a single bytestring with
324
the complete text of the file.
326
:param desired_files: a list of (file_id, identifier) pairs
328
for file_id, identifier in desired_files:
329
# We wrap the string in a tuple so that we can return an iterable
330
# of bytestrings. (Technically, a bytestring is also an iterable
331
# of bytestrings, but iterating through each character is not
333
cur_file = (self.get_file_text(file_id),)
334
yield identifier, cur_file
336
def get_symlink_target(self, file_id):
337
"""Get the target for a given file_id.
339
It is assumed that the caller already knows that file_id is referencing
341
:param file_id: Handle for the symlink entry.
342
:return: The path the symlink points to.
344
raise NotImplementedError(self.get_symlink_target)
346
def get_canonical_inventory_paths(self, paths):
347
"""Like get_canonical_inventory_path() but works on multiple items.
349
:param paths: A sequence of paths relative to the root of the tree.
350
:return: A list of paths, with each item the corresponding input path
351
adjusted to account for existing elements that match case
354
return list(self._yield_canonical_inventory_paths(paths))
356
def get_canonical_inventory_path(self, path):
357
"""Returns the first inventory item that case-insensitively matches path.
359
If a path matches exactly, it is returned. If no path matches exactly
360
but more than one path matches case-insensitively, it is implementation
361
defined which is returned.
363
If no path matches case-insensitively, the input path is returned, but
364
with as many path entries that do exist changed to their canonical
367
If you need to resolve many names from the same tree, you should
368
use get_canonical_inventory_paths() to avoid O(N) behaviour.
370
:param path: A paths relative to the root of the tree.
371
:return: The input path adjusted to account for existing elements
372
that match case insensitively.
374
return self._yield_canonical_inventory_paths([path]).next()
376
def _yield_canonical_inventory_paths(self, paths):
378
# First, if the path as specified exists exactly, just use it.
379
if self.path2id(path) is not None:
383
cur_id = self.get_root_id()
385
bit_iter = iter(path.split("/"))
388
for child in self.iter_children(cur_id):
390
child_base = os.path.basename(self.id2path(child))
391
if child_base.lower() == lelt:
393
cur_path = osutils.pathjoin(cur_path, child_base)
396
# before a change is committed we can see this error...
399
# got to the end of this directory and no entries matched.
400
# Return what matched so far, plus the rest as specified.
401
cur_path = osutils.pathjoin(cur_path, elt, *list(bit_iter))
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.
414
For working trees (and mutable trees in general), the special
415
revision_id 'current:' will be used for lines that are new in this
416
tree, e.g. uncommitted changes.
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
422
raise NotImplementedError(self.annotate_iter)
424
def _get_plan_merge_data(self, file_id, other, base):
425
from bzrlib import merge, 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)
470
def _file_revision(revision_tree, file_id):
471
"""Determine the revision associated with a file in a given tree."""
472
revision_tree.lock_read()
474
return revision_tree.inventory[file_id].revision
476
revision_tree.unlock()
478
def _get_file_revision(self, file_id, vf, tree_revision):
479
"""Ensure that file_id, tree_revision is in vf to plan the merge."""
481
if getattr(self, '_repository', None) is None:
482
last_revision = tree_revision
483
parent_keys = [(file_id, self._file_revision(t, file_id)) for t in
484
self._iter_parent_trees()]
485
vf.add_lines((file_id, last_revision), parent_keys,
486
self.get_file(file_id).readlines())
487
repo = self.branch.repository
490
last_revision = self._file_revision(self, file_id)
491
base_vf = self._repository.texts
492
if base_vf not in vf.fallback_versionedfiles:
493
vf.fallback_versionedfiles.append(base_vf)
141
496
inventory = property(_get_inventory,
142
497
doc="Inventory of this Tree")
144
499
def _check_retrieved(self, ie, f):
145
500
if not __debug__:
147
502
fp = fingerprint_file(f)
150
505
if ie.text_size is not None:
151
506
if ie.text_size != fp['size']:
152
507
raise BzrError("mismatched size for file %r in %r" % (ie.file_id, self._store),
182
575
def filter_unversioned_files(self, paths):
183
"""Filter out paths that are not versioned.
576
"""Filter out paths that are versioned.
185
578
:return: set of paths.
187
580
# NB: we specifically *don't* call self.has_filename, because for
188
# WorkingTrees that can indicate files that exist on disk but that
581
# WorkingTrees that can indicate files that exist on disk but that
189
582
# are not versioned.
190
583
pred = self.inventory.has_filename
191
584
return set((p for p in paths if not pred(p)))
195
from bzrlib.revisiontree import RevisionTree
198
class EmptyTree(Tree):
201
self._inventory = Inventory()
202
warn('EmptyTree is deprecated as of bzr 0.9 please use '
203
'repository.revision_tree instead.',
204
DeprecationWarning, stacklevel=2)
206
def get_parent_ids(self):
209
def get_symlink_target(self, file_id):
212
def has_filename(self, filename):
586
def walkdirs(self, prefix=""):
587
"""Walk the contents of this tree from path down.
589
This yields all the data about the contents of a directory at a time.
590
After each directory has been yielded, if the caller has mutated the
591
list to exclude some directories, they are then not descended into.
593
The data yielded is of the form:
594
((directory-relpath, directory-path-from-root, directory-fileid),
595
[(relpath, basename, kind, lstat, path_from_tree_root, file_id,
596
versioned_kind), ...]),
597
- directory-relpath is the containing dirs relpath from prefix
598
- directory-path-from-root is the containing dirs path from /
599
- directory-fileid is the id of the directory if it is versioned.
600
- relpath is the relative path within the subtree being walked.
601
- basename is the basename
602
- kind is the kind of the file now. If unknonwn then the file is not
603
present within the tree - but it may be recorded as versioned. See
605
- lstat is the stat data *if* the file was statted.
606
- path_from_tree_root is the path from the root of the tree.
607
- file_id is the file_id if the entry is versioned.
608
- versioned_kind is the kind of the file as last recorded in the
609
versioning system. If 'unknown' the file is not versioned.
610
One of 'kind' and 'versioned_kind' must not be 'unknown'.
612
:param prefix: Start walking from prefix within the tree rather than
613
at the root. This allows one to walk a subtree but get paths that are
614
relative to a tree rooted higher up.
615
:return: an iterator over the directory data.
617
raise NotImplementedError(self.walkdirs)
619
def supports_content_filtering(self):
215
def kind(self, file_id):
216
assert self._inventory[file_id].kind == "directory"
219
def list_files(self):
222
def __contains__(self, file_id):
223
return (file_id in self._inventory)
225
def get_file_sha1(self, file_id, path=None):
622
def _content_filter_stack(self, path=None, file_id=None):
623
"""The stack of content filters for a path if filtering is supported.
625
Readers will be applied in first-to-last order.
626
Writers will be applied in last-to-first order.
627
Either the path or the file-id needs to be provided.
629
:param path: path relative to the root of the tree
631
:param file_id: file_id or None if unknown
632
:return: the list of filters - [] if there are none
634
filter_pref_names = filters._get_registered_names()
635
if len(filter_pref_names) == 0:
638
path = self.id2path(file_id)
639
prefs = self.iter_search_rules([path], filter_pref_names).next()
640
stk = filters._get_filter_stack_for(prefs)
641
if 'filters' in debug.debug_flags:
642
note("*** %s content-filter: %s => %r" % (path,prefs,stk))
645
def _content_filter_stack_provider(self):
646
"""A function that returns a stack of ContentFilters.
648
The function takes a path (relative to the top of the tree) and a
649
file-id as parameters.
651
:return: None if content filtering is not supported by this tree.
653
if self.supports_content_filtering():
654
return lambda path, file_id: \
655
self._content_filter_stack(path, file_id)
659
def iter_search_rules(self, path_names, pref_names=None,
660
_default_searcher=rules._per_user_searcher):
661
"""Find the preferences for filenames in a tree.
663
:param path_names: an iterable of paths to find attributes for.
664
Paths are given relative to the root of the tree.
665
:param pref_names: the list of preferences to lookup - None for all
666
:param _default_searcher: private parameter to assist testing - don't use
667
:return: an iterator of tuple sequences, one per path-name.
668
See _RulesSearcher.get_items for details on the tuple sequence.
670
searcher = self._get_rules_searcher(_default_searcher)
671
if searcher is not None:
672
if pref_names is not None:
673
for path in path_names:
674
yield searcher.get_selected_items(path, pref_names)
676
for path in path_names:
677
yield searcher.get_items(path)
680
def _get_rules_searcher(self, default_searcher):
681
"""Get the RulesSearcher for this tree given the default one."""
682
searcher = default_searcher
229
686
######################################################################
394
851
:param require_versioned: An optional boolean (defaults to False). When
395
852
supplied and True all the 'specific_files' must be versioned, or
396
853
a PathsNotVersionedError will be thrown.
854
:param want_unversioned: Scan for unversioned paths.
398
# NB: show_status depends on being able to pass in non-versioned files and
399
# report them as unknown
400
trees = (self.source, self.target)
856
# NB: show_status depends on being able to pass in non-versioned files
857
# and report them as unknown
858
trees = (self.source,)
401
859
if extra_trees is not None:
402
860
trees = trees + tuple(extra_trees)
403
specific_file_ids = find_ids_across_trees(specific_files,
404
trees, require_versioned=require_versioned)
861
# target is usually the newer tree:
862
specific_file_ids = self.target.paths2ids(specific_files, trees,
863
require_versioned=require_versioned)
405
864
if specific_files and not specific_file_ids:
406
865
# All files are unversioned, so just return an empty delta
407
866
# _compare_trees would think we want a complete delta
408
return delta.TreeDelta()
867
result = delta.TreeDelta()
868
fake_entry = InventoryFile('unused', 'unused', 'unused')
869
result.unversioned = [(path, None,
870
self.target._comparison_data(fake_entry, path)[0]) for path in
409
873
return delta._compare_trees(self.source, self.target, want_unchanged,
874
specific_files, include_root, extra_trees=extra_trees,
875
require_versioned=require_versioned,
876
want_unversioned=want_unversioned)
878
def iter_changes(self, include_unchanged=False,
879
specific_files=None, pb=None, extra_trees=[],
880
require_versioned=True, want_unversioned=False):
881
"""Generate an iterator of changes between trees.
884
(file_id, (path_in_source, path_in_target),
885
changed_content, versioned, parent, name, kind,
888
Changed_content is True if the file's content has changed. This
889
includes changes to its kind, and to a symlink's target.
891
versioned, parent, name, kind, executable are tuples of (from, to).
892
If a file is missing in a tree, its kind is None.
894
Iteration is done in parent-to-child order, relative to the target
897
There is no guarantee that all paths are in sorted order: the
898
requirement to expand the search due to renames may result in children
899
that should be found early being found late in the search, after
900
lexically later results have been returned.
901
:param require_versioned: Raise errors.PathsNotVersionedError if a
902
path in the specific_files list is not versioned in one of
903
source, target or extra_trees.
904
:param want_unversioned: Should unversioned files be returned in the
905
output. An unversioned file is defined as one with (False, False)
906
for the versioned pair.
908
lookup_trees = [self.source]
910
lookup_trees.extend(extra_trees)
911
if specific_files == []:
912
specific_file_ids = []
914
specific_file_ids = self.target.paths2ids(specific_files,
915
lookup_trees, require_versioned=require_versioned)
917
all_unversioned = sorted([(p.split('/'), p) for p in
919
if specific_files is None or
920
osutils.is_inside_any(specific_files, p)])
921
all_unversioned = deque(all_unversioned)
923
all_unversioned = deque()
925
from_entries_by_dir = list(self.source.iter_entries_by_dir(
926
specific_file_ids=specific_file_ids))
927
from_data = dict((e.file_id, (p, e)) for p, e in from_entries_by_dir)
928
to_entries_by_dir = list(self.target.iter_entries_by_dir(
929
specific_file_ids=specific_file_ids))
930
num_entries = len(from_entries_by_dir) + len(to_entries_by_dir)
932
# the unversioned path lookup only occurs on real trees - where there
933
# can be extras. So the fake_entry is solely used to look up
934
# executable it values when execute is not supported.
935
fake_entry = InventoryFile('unused', 'unused', 'unused')
936
for to_path, to_entry in to_entries_by_dir:
937
while all_unversioned and all_unversioned[0][0] < to_path.split('/'):
938
unversioned_path = all_unversioned.popleft()
939
to_kind, to_executable, to_stat = \
940
self.target._comparison_data(fake_entry, unversioned_path[1])
941
yield (None, (None, unversioned_path[1]), True, (False, False),
943
(None, unversioned_path[0][-1]),
945
(None, to_executable))
946
file_id = to_entry.file_id
947
to_paths[file_id] = to_path
949
changed_content = False
950
from_path, from_entry = from_data.get(file_id, (None, None))
951
from_versioned = (from_entry is not None)
952
if from_entry is not None:
953
from_versioned = True
954
from_name = from_entry.name
955
from_parent = from_entry.parent_id
956
from_kind, from_executable, from_stat = \
957
self.source._comparison_data(from_entry, from_path)
960
from_versioned = False
964
from_executable = None
965
versioned = (from_versioned, True)
966
to_kind, to_executable, to_stat = \
967
self.target._comparison_data(to_entry, to_path)
968
kind = (from_kind, to_kind)
969
if kind[0] != kind[1]:
970
changed_content = True
971
elif from_kind == 'file':
972
if (self.source.get_file_sha1(file_id, from_path, from_stat) !=
973
self.target.get_file_sha1(file_id, to_path, to_stat)):
974
changed_content = True
975
elif from_kind == 'symlink':
976
if (self.source.get_symlink_target(file_id) !=
977
self.target.get_symlink_target(file_id)):
978
changed_content = True
979
# XXX: Yes, the indentation below is wrong. But fixing it broke
980
# test_merge.TestMergerEntriesLCAOnDisk.
981
# test_nested_tree_subtree_renamed_and_modified. We'll wait for
982
# the fix from bzr.dev -- vila 2009026
983
elif from_kind == 'tree-reference':
984
if (self.source.get_reference_revision(file_id, from_path)
985
!= self.target.get_reference_revision(file_id, to_path)):
986
changed_content = True
987
parent = (from_parent, to_entry.parent_id)
988
name = (from_name, to_entry.name)
989
executable = (from_executable, to_executable)
991
pb.update('comparing files', entry_count, num_entries)
992
if (changed_content is not False or versioned[0] != versioned[1]
993
or parent[0] != parent[1] or name[0] != name[1] or
994
executable[0] != executable[1] or include_unchanged):
995
yield (file_id, (from_path, to_path), changed_content,
996
versioned, parent, name, kind, executable)
998
while all_unversioned:
999
# yield any trailing unversioned paths
1000
unversioned_path = all_unversioned.popleft()
1001
to_kind, to_executable, to_stat = \
1002
self.target._comparison_data(fake_entry, unversioned_path[1])
1003
yield (None, (None, unversioned_path[1]), True, (False, False),
1005
(None, unversioned_path[0][-1]),
1007
(None, to_executable))
1009
def get_to_path(to_entry):
1010
if to_entry.parent_id is None:
1011
to_path = '' # the root
1013
if to_entry.parent_id not in to_paths:
1015
return get_to_path(self.target.inventory[to_entry.parent_id])
1016
to_path = osutils.pathjoin(to_paths[to_entry.parent_id],
1018
to_paths[to_entry.file_id] = to_path
1021
for path, from_entry in from_entries_by_dir:
1022
file_id = from_entry.file_id
1023
if file_id in to_paths:
1026
if not file_id in self.target.all_file_ids():
1027
# common case - paths we have not emitted are not present in
1031
to_path = get_to_path(self.target.inventory[file_id])
1034
pb.update('comparing files', entry_count, num_entries)
1035
versioned = (True, False)
1036
parent = (from_entry.parent_id, None)
1037
name = (from_entry.name, None)
1038
from_kind, from_executable, stat_value = \
1039
self.source._comparison_data(from_entry, path)
1040
kind = (from_kind, None)
1041
executable = (from_executable, None)
1042
changed_content = from_kind is not None
1043
# the parent's path is necessarily known at this point.
1044
yield(file_id, (path, to_path), changed_content, versioned, parent,
1045
name, kind, executable)
1048
class MultiWalker(object):
1049
"""Walk multiple trees simultaneously, getting combined results."""
1051
# Note: This could be written to not assume you can do out-of-order
1052
# lookups. Instead any nodes that don't match in all trees could be
1053
# marked as 'deferred', and then returned in the final cleanup loop.
1054
# For now, I think it is "nicer" to return things as close to the
1055
# "master_tree" order as we can.
1057
def __init__(self, master_tree, other_trees):
1058
"""Create a new MultiWalker.
1060
All trees being walked must implement "iter_entries_by_dir()", such
1061
that they yield (path, object) tuples, where that object will have a
1062
'.file_id' member, that can be used to check equality.
1064
:param master_tree: All trees will be 'slaved' to the master_tree such
1065
that nodes in master_tree will be used as 'first-pass' sync points.
1066
Any nodes that aren't in master_tree will be merged in a second
1068
:param other_trees: A list of other trees to walk simultaneously.
1070
self._master_tree = master_tree
1071
self._other_trees = other_trees
1073
# Keep track of any nodes that were properly processed just out of
1074
# order, that way we don't return them at the end, we don't have to
1075
# track *all* processed file_ids, just the out-of-order ones
1076
self._out_of_order_processed = set()
1079
def _step_one(iterator):
1080
"""Step an iter_entries_by_dir iterator.
1082
:return: (has_more, path, ie)
1083
If has_more is False, path and ie will be None.
1086
path, ie = iterator.next()
1087
except StopIteration:
1088
return False, None, None
1090
return True, path, ie
1093
def _cmp_path_by_dirblock(path1, path2):
1094
"""Compare two paths based on what directory they are in.
1096
This generates a sort order, such that all children of a directory are
1097
sorted together, and grandchildren are in the same order as the
1098
children appear. But all grandchildren come after all children.
1100
:param path1: first path
1101
:param path2: the second path
1102
:return: negative number if ``path1`` comes first,
1103
0 if paths are equal
1104
and a positive number if ``path2`` sorts first
1106
# Shortcut this special case
1109
# This is stolen from _dirstate_helpers_py.py, only switching it to
1110
# Unicode objects. Consider using encode_utf8() and then using the
1111
# optimized versions, or maybe writing optimized unicode versions.
1112
if not isinstance(path1, unicode):
1113
raise TypeError("'path1' must be a unicode string, not %s: %r"
1114
% (type(path1), path1))
1115
if not isinstance(path2, unicode):
1116
raise TypeError("'path2' must be a unicode string, not %s: %r"
1117
% (type(path2), path2))
1118
return cmp(MultiWalker._path_to_key(path1),
1119
MultiWalker._path_to_key(path2))
1122
def _path_to_key(path):
1123
dirname, basename = osutils.split(path)
1124
return (dirname.split(u'/'), basename)
1126
def _lookup_by_file_id(self, extra_entries, other_tree, file_id):
1127
"""Lookup an inventory entry by file_id.
1129
This is called when an entry is missing in the normal order.
1130
Generally this is because a file was either renamed, or it was
1131
deleted/added. If the entry was found in the inventory and not in
1132
extra_entries, it will be added to self._out_of_order_processed
1134
:param extra_entries: A dictionary of {file_id: (path, ie)}. This
1135
should be filled with entries that were found before they were
1136
used. If file_id is present, it will be removed from the
1138
:param other_tree: The Tree to search, in case we didn't find the entry
1140
:param file_id: The file_id to look for
1141
:return: (path, ie) if found or (None, None) if not present.
1143
if file_id in extra_entries:
1144
return extra_entries.pop(file_id)
1145
# TODO: Is id2path better as the first call, or is
1146
# inventory[file_id] better as a first check?
1148
cur_path = other_tree.id2path(file_id)
1149
except errors.NoSuchId:
1151
if cur_path is None:
1154
self._out_of_order_processed.add(file_id)
1155
cur_ie = other_tree.inventory[file_id]
1156
return (cur_path, cur_ie)
1159
"""Match up the values in the different trees."""
1160
for result in self._walk_master_tree():
1162
self._finish_others()
1163
for result in self._walk_others():
1166
def _walk_master_tree(self):
1167
"""First pass, walk all trees in lock-step.
1169
When we are done, all nodes in the master_tree will have been
1170
processed. _other_walkers, _other_entries, and _others_extra will be
1171
set on 'self' for future processing.
1173
# This iterator has the most "inlining" done, because it tends to touch
1174
# every file in the tree, while the others only hit nodes that don't
1176
master_iterator = self._master_tree.iter_entries_by_dir()
1178
other_walkers = [other.iter_entries_by_dir()
1179
for other in self._other_trees]
1180
other_entries = [self._step_one(walker) for walker in other_walkers]
1181
# Track extra nodes in the other trees
1182
others_extra = [{} for i in xrange(len(self._other_trees))]
1184
master_has_more = True
1185
step_one = self._step_one
1186
lookup_by_file_id = self._lookup_by_file_id
1187
out_of_order_processed = self._out_of_order_processed
1189
while master_has_more:
1190
(master_has_more, path, master_ie) = step_one(master_iterator)
1191
if not master_has_more:
1194
file_id = master_ie.file_id
1196
other_values_append = other_values.append
1197
next_other_entries = []
1198
next_other_entries_append = next_other_entries.append
1199
for idx, (other_has_more, other_path, other_ie) in enumerate(other_entries):
1200
if not other_has_more:
1201
other_values_append(lookup_by_file_id(
1202
others_extra[idx], self._other_trees[idx], file_id))
1203
next_other_entries_append((False, None, None))
1204
elif file_id == other_ie.file_id:
1205
# This is the critical code path, as most of the entries
1206
# should match between most trees.
1207
other_values_append((other_path, other_ie))
1208
next_other_entries_append(step_one(other_walkers[idx]))
1210
# This walker did not match, step it until it either
1211
# matches, or we know we are past the current walker.
1212
other_walker = other_walkers[idx]
1213
other_extra = others_extra[idx]
1214
while (other_has_more and
1215
self._cmp_path_by_dirblock(other_path, path) < 0):
1216
other_file_id = other_ie.file_id
1217
if other_file_id not in out_of_order_processed:
1218
other_extra[other_file_id] = (other_path, other_ie)
1219
other_has_more, other_path, other_ie = \
1220
step_one(other_walker)
1221
if other_has_more and other_ie.file_id == file_id:
1222
# We ended up walking to this point, match and step
1224
other_values_append((other_path, other_ie))
1225
other_has_more, other_path, other_ie = \
1226
step_one(other_walker)
1228
# This record isn't in the normal order, see if it
1230
other_values_append(lookup_by_file_id(
1231
other_extra, self._other_trees[idx], file_id))
1232
next_other_entries_append((other_has_more, other_path,
1234
other_entries = next_other_entries
1236
# We've matched all the walkers, yield this datapoint
1237
yield path, file_id, master_ie, other_values
1238
self._other_walkers = other_walkers
1239
self._other_entries = other_entries
1240
self._others_extra = others_extra
1242
def _finish_others(self):
1243
"""Finish walking the other iterators, so we get all entries."""
1244
for idx, info in enumerate(self._other_entries):
1245
other_extra = self._others_extra[idx]
1246
(other_has_more, other_path, other_ie) = info
1247
while other_has_more:
1248
other_file_id = other_ie.file_id
1249
if other_file_id not in self._out_of_order_processed:
1250
other_extra[other_file_id] = (other_path, other_ie)
1251
other_has_more, other_path, other_ie = \
1252
self._step_one(self._other_walkers[idx])
1253
del self._other_entries
1255
def _walk_others(self):
1256
"""Finish up by walking all the 'deferred' nodes."""
1257
# TODO: One alternative would be to grab all possible unprocessed
1258
# file_ids, and then sort by path, and then yield them. That
1259
# might ensure better ordering, in case a caller strictly
1260
# requires parents before children.
1261
for idx, other_extra in enumerate(self._others_extra):
1262
others = sorted(other_extra.itervalues(),
1263
key=lambda x: self._path_to_key(x[0]))
1264
for other_path, other_ie in others:
1265
file_id = other_ie.file_id
1266
# We don't need to check out_of_order_processed here, because
1267
# the lookup_by_file_id will be removing anything processed
1268
# from the extras cache
1269
other_extra.pop(file_id)
1270
other_values = [(None, None) for i in xrange(idx)]
1271
other_values.append((other_path, other_ie))
1272
for alt_idx, alt_extra in enumerate(self._others_extra[idx+1:]):
1273
alt_idx = alt_idx + idx + 1
1274
alt_extra = self._others_extra[alt_idx]
1275
alt_tree = self._other_trees[alt_idx]
1276
other_values.append(self._lookup_by_file_id(
1277
alt_extra, alt_tree, file_id))
1278
yield other_path, file_id, None, other_values
added
removed
bzrlib/tree.py
