17
17
"""Tree classes, representing directory at point in time.
21
from collections import deque
20
from __future__ import absolute_import
22
from .lazy_import import lazy_import
23
lazy_import(globals(), """
25
27
conflicts as _mod_conflicts,
30
31
revision as _mod_revision,
33
from bzrlib.decorators import needs_read_lock
34
from bzrlib.errors import BzrError, NoSuchId
35
from bzrlib import errors
36
from bzrlib.inventory import InventoryFile
37
from bzrlib.inter import InterObject
38
from bzrlib.osutils import fingerprint_file
39
from bzrlib.symbol_versioning import deprecated_function, deprecated_in
40
from bzrlib.trace import note
35
from breezy.i18n import gettext
43
from .inter import InterObject
50
class FileTimestampUnavailable(errors.BzrError):
52
_fmt = "The filestamp for %(path)s is not available."
56
def __init__(self, path):
60
class TreeEntry(object):
61
"""An entry that implements the minimum interface used by commands.
64
def __eq__(self, other):
65
# yes, this is ugly, TODO: best practice __eq__ style.
66
return (isinstance(other, TreeEntry)
67
and other.__class__ == self.__class__)
69
def kind_character(self):
73
class TreeDirectory(TreeEntry):
74
"""See TreeEntry. This is a directory in a working tree."""
76
def kind_character(self):
80
class TreeFile(TreeEntry):
81
"""See TreeEntry. This is a regular file in a working tree."""
83
def kind_character(self):
87
class TreeLink(TreeEntry):
88
"""See TreeEntry. This is a symlink in a working tree."""
90
def kind_character(self):
94
class TreeReference(TreeEntry):
95
"""See TreeEntry. This is a reference to a nested tree in a working tree."""
97
def kind_character(self):
43
101
class Tree(object):
50
108
* `RevisionTree` is a tree as recorded at some point in the past.
52
Trees contain an `Inventory` object, and also know how to retrieve
53
file texts mentioned in the inventory, either from a working
54
directory or from a store.
56
It is possible for trees to contain files that are not described
57
in their inventory or vice versa; for this use `filenames()`.
59
110
Trees can be compared, etc, regardless of whether they are working
60
111
trees or versioned trees.
114
def supports_rename_tracking(self):
115
"""Whether this tree supports rename tracking.
117
This defaults to True, but some implementations may want to override
122
def has_versioned_directories(self):
123
"""Whether this tree can contain explicitly versioned directories.
125
This defaults to True, but some implementations may want to override
63
130
def changes_from(self, other, want_unchanged=False, specific_files=None,
64
extra_trees=None, require_versioned=False, include_root=False,
65
want_unversioned=False):
131
extra_trees=None, require_versioned=False, include_root=False,
132
want_unversioned=False):
66
133
"""Return a TreeDelta of the changes from other to this tree.
68
135
:param other: A tree to compare with.
147
return iter(self.inventory)
149
204
def all_file_ids(self):
150
205
"""Iterate through all file ids, including ids for missing files."""
151
return set(self.inventory)
206
raise NotImplementedError(self.all_file_ids)
208
def all_versioned_paths(self):
209
"""Iterate through all paths, including paths for missing files."""
210
raise NotImplementedError(self.all_versioned_paths)
153
212
def id2path(self, file_id):
154
213
"""Return the path for a file id.
156
215
:raises NoSuchId:
158
return self.inventory.id2path(file_id)
160
def is_control_filename(self, filename):
161
"""True if filename is the name of a control file in this tree.
163
:param filename: A filename within the tree. This is a relative path
164
from the root of this tree.
166
This is true IF and ONLY IF the filename is part of the meta data
167
that bzr controls in this tree. I.E. a random .bzr directory placed
168
on disk will not be a control file for this tree.
170
return self.bzrdir.is_control_filename(filename)
173
def iter_entries_by_dir(self, specific_file_ids=None, yield_parents=False):
217
raise NotImplementedError(self.id2path)
219
def iter_entries_by_dir(self, specific_files=None):
174
220
"""Walk the tree in 'by_dir' order.
176
222
This will yield each entry in the tree as a (path, entry) tuple.
194
240
The yield order (ignoring root) would be::
195
242
a, f, a/b, a/d, a/b/c, a/d/e, f/g
197
:param yield_parents: If True, yield the parents from the root leading
198
down to specific_file_ids that have been requested. This has no
199
impact if specific_file_ids is None.
201
return self.inventory.iter_entries_by_dir(
202
specific_file_ids=specific_file_ids, yield_parents=yield_parents)
244
raise NotImplementedError(self.iter_entries_by_dir)
246
def iter_child_entries(self, path, file_id=None):
247
"""Iterate over the children of a directory or tree reference.
249
:param path: Path of the directory
250
:param file_id: Optional file id of the directory/tree-reference
251
:raise NoSuchId: When the file_id does not exist
252
:return: Iterator over entries in the directory
254
raise NotImplementedError(self.iter_child_entries)
256
def list_files(self, include_root=False, from_dir=None, recursive=True):
257
"""List all files in this tree.
259
:param include_root: Whether to include the entry for the tree root
260
:param from_dir: Directory under which to list files
261
:param recursive: Whether to list files recursively
262
:return: iterator over tuples of (path, versioned, kind, file_id,
265
raise NotImplementedError(self.list_files)
204
267
def iter_references(self):
205
268
if self.supports_tree_reference():
207
270
if entry.kind == 'tree-reference':
208
271
yield path, entry.file_id
210
def kind(self, file_id):
273
def kind(self, path, file_id=None):
211
274
raise NotImplementedError("Tree subclass %s must implement kind"
212
% self.__class__.__name__)
275
% self.__class__.__name__)
214
def stored_kind(self, file_id):
277
def stored_kind(self, path, file_id=None):
215
278
"""File kind stored for this file_id.
217
280
May not match kind on disk for working trees. Always available
218
281
for versioned files, even when the file itself is missing.
220
return self.kind(file_id)
283
return self.kind(path, file_id)
222
285
def path_content_summary(self, path):
223
286
"""Get a summary of the information about path.
268
325
raise NotImplementedError(self.get_file)
270
def get_file_with_stat(self, file_id, path=None):
327
def get_file_with_stat(self, path, file_id=None):
271
328
"""Get a file handle and stat object for file_id.
273
330
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.
333
:param path: The path of the file.
334
:param file_id: The file id to read, if it is known.
278
335
:return: A tuple (file_handle, stat_value_or_None). If the tree has
279
336
no stat facility, or need for a stat cache feedback during commit,
280
337
it may return None for the second element of the tuple.
282
return (self.get_file(file_id, path), None)
339
return (self.get_file(path, file_id), None)
284
def get_file_text(self, file_id, path=None):
341
def get_file_text(self, path, file_id=None):
285
342
"""Return the byte content of a file.
344
:param path: The path of the file.
287
345
:param file_id: The file_id of the file.
288
:param path: The path of the file.
289
347
If both file_id and path are supplied, an implementation may use
350
:returns: A single byte string for the whole file.
292
my_file = self.get_file(file_id, path)
352
with self.get_file(path, file_id) as my_file:
294
353
return my_file.read()
298
def get_file_lines(self, file_id, path=None):
355
def get_file_lines(self, path, file_id=None):
299
356
"""Return the content of a file, as lines.
358
:param path: The path of the file.
301
359
:param file_id: The file_id of the file.
302
:param path: The path of the file.
303
361
If both file_id and path are supplied, an implementation may use
306
return osutils.split_lines(self.get_file_text(file_id, path))
308
def get_file_mtime(self, file_id, path=None):
364
return osutils.split_lines(self.get_file_text(path, file_id))
366
def get_file_verifier(self, path, file_id=None, stat_value=None):
367
"""Return a verifier for a file.
369
The default implementation returns a sha1.
371
:param file_id: The handle for this file.
372
:param path: The path that this file can be found at.
373
These must point to the same object.
374
:param stat_value: Optional stat value for the object
375
:return: Tuple with verifier name and verifier data
377
return ("SHA1", self.get_file_sha1(path, file_id,
378
stat_value=stat_value))
380
def get_file_sha1(self, path, file_id=None, stat_value=None):
381
"""Return the SHA1 file for a file.
383
:note: callers should use get_file_verifier instead
384
where possible, as the underlying repository implementation may
385
have quicker access to a non-sha1 verifier.
387
:param path: The path that this file can be found at.
388
:param file_id: The handle for this file.
389
These must point to the same object.
390
:param stat_value: Optional stat value for the object
392
raise NotImplementedError(self.get_file_sha1)
394
def get_file_mtime(self, path, file_id=None):
309
395
"""Return the modification time for a file.
397
:param path: The path that this file can be found at.
311
398
:param file_id: The handle for this file.
312
:param path: The path that this file can be found at.
313
399
These must point to the same object.
315
401
raise NotImplementedError(self.get_file_mtime)
317
def get_file_size(self, file_id):
403
def get_file_size(self, path, file_id=None):
318
404
"""Return the size of a file in bytes.
320
406
This applies only to regular files. If invoked on directories or
342
434
this implementation, it is a tuple containing a single bytestring with
343
435
the complete text of the file.
345
:param desired_files: a list of (file_id, identifier) pairs
437
:param desired_files: a list of (path, identifier) pairs
347
for file_id, identifier in desired_files:
439
for path, identifier in desired_files:
348
440
# We wrap the string in a tuple so that we can return an iterable
349
441
# of bytestrings. (Technically, a bytestring is also an iterable
350
442
# of bytestrings, but iterating through each character is not
352
cur_file = (self.get_file_text(file_id),)
444
cur_file = (self.get_file_text(path),)
353
445
yield identifier, cur_file
355
def get_symlink_target(self, file_id):
447
def get_symlink_target(self, path, file_id=None):
356
448
"""Get the target for a given file_id.
358
450
It is assumed that the caller already knows that file_id is referencing
360
452
:param file_id: Handle for the symlink entry.
453
:param path: The path of the file.
454
If both file_id and path are supplied, an implementation may use
361
456
:return: The path the symlink points to.
363
458
raise NotImplementedError(self.get_symlink_target)
365
def get_canonical_inventory_paths(self, paths):
366
"""Like get_canonical_inventory_path() but works on multiple items.
368
:param paths: A sequence of paths relative to the root of the tree.
369
:return: A list of paths, with each item the corresponding input path
370
adjusted to account for existing elements that match case
373
return list(self._yield_canonical_inventory_paths(paths))
375
def get_canonical_inventory_path(self, path):
376
"""Returns the first inventory item that case-insensitively matches path.
378
If a path matches exactly, it is returned. If no path matches exactly
379
but more than one path matches case-insensitively, it is implementation
380
defined which is returned.
382
If no path matches case-insensitively, the input path is returned, but
383
with as many path entries that do exist changed to their canonical
386
If you need to resolve many names from the same tree, you should
387
use get_canonical_inventory_paths() to avoid O(N) behaviour.
389
:param path: A paths relative to the root of the tree.
390
:return: The input path adjusted to account for existing elements
391
that match case insensitively.
393
return self._yield_canonical_inventory_paths([path]).next()
395
def _yield_canonical_inventory_paths(self, paths):
397
# First, if the path as specified exists exactly, just use it.
398
if self.path2id(path) is not None:
402
cur_id = self.get_root_id()
404
bit_iter = iter(path.split("/"))
408
for child in self.iter_children(cur_id):
410
# XXX: it seem like if the child is known to be in the
411
# tree, we shouldn't need to go from its id back to
412
# its path -- mbp 2010-02-11
414
# XXX: it seems like we could be more efficient
415
# by just directly looking up the original name and
416
# only then searching all children; also by not
417
# chopping paths so much. -- mbp 2010-02-11
418
child_base = os.path.basename(self.id2path(child))
419
if (child_base == elt):
420
# if we found an exact match, we can stop now; if
421
# we found an approximate match we need to keep
422
# searching because there might be an exact match
425
new_path = osutils.pathjoin(cur_path, child_base)
427
elif child_base.lower() == lelt:
429
new_path = osutils.pathjoin(cur_path, child_base)
431
# before a change is committed we can see this error...
436
# got to the end of this directory and no entries matched.
437
# Return what matched so far, plus the rest as specified.
438
cur_path = osutils.pathjoin(cur_path, elt, *list(bit_iter))
443
460
def get_root_id(self):
444
461
"""Return the file_id for the root of this tree."""
445
462
raise NotImplementedError(self.get_root_id)
447
def annotate_iter(self, file_id,
464
def annotate_iter(self, path, file_id=None,
448
465
default_revision=_mod_revision.CURRENT_REVISION):
449
466
"""Return an iterator of revision_id, line tuples.
459
476
raise NotImplementedError(self.annotate_iter)
461
def _get_plan_merge_data(self, file_id, other, base):
462
from bzrlib import versionedfile
463
vf = versionedfile._PlanMergeVersionedFile(file_id)
464
last_revision_a = self._get_file_revision(file_id, vf, 'this:')
465
last_revision_b = other._get_file_revision(file_id, vf, 'other:')
467
last_revision_base = None
469
last_revision_base = base._get_file_revision(file_id, vf, 'base:')
470
return vf, last_revision_a, last_revision_b, last_revision_base
472
def plan_file_merge(self, file_id, other, base=None):
473
"""Generate a merge plan based on annotations.
475
If the file contains uncommitted changes in this tree, they will be
476
attributed to the 'current:' pseudo-revision. If the file contains
477
uncommitted changes in the other tree, they will be assigned to the
478
'other:' pseudo-revision.
480
data = self._get_plan_merge_data(file_id, other, base)
481
vf, last_revision_a, last_revision_b, last_revision_base = data
482
return vf.plan_merge(last_revision_a, last_revision_b,
485
def plan_file_lca_merge(self, file_id, other, base=None):
486
"""Generate a merge plan based lca-newness.
488
If the file contains uncommitted changes in this tree, they will be
489
attributed to the 'current:' pseudo-revision. If the file contains
490
uncommitted changes in the other tree, they will be assigned to the
491
'other:' pseudo-revision.
493
data = self._get_plan_merge_data(file_id, other, base)
494
vf, last_revision_a, last_revision_b, last_revision_base = data
495
return vf.plan_lca_merge(last_revision_a, last_revision_b,
498
478
def _iter_parent_trees(self):
499
479
"""Iterate through parent trees, defaulting to Tree.revision_tree."""
500
480
for revision_id in self.get_parent_ids():
503
483
except errors.NoSuchRevisionInTree:
504
484
yield self.repository.revision_tree(revision_id)
507
def _file_revision(revision_tree, file_id):
508
"""Determine the revision associated with a file in a given tree."""
509
revision_tree.lock_read()
511
return revision_tree.inventory[file_id].revision
513
revision_tree.unlock()
515
def _get_file_revision(self, file_id, vf, tree_revision):
516
"""Ensure that file_id, tree_revision is in vf to plan the merge."""
518
if getattr(self, '_repository', None) is None:
519
last_revision = tree_revision
520
parent_keys = [(file_id, self._file_revision(t, file_id)) for t in
521
self._iter_parent_trees()]
522
vf.add_lines((file_id, last_revision), parent_keys,
523
self.get_file(file_id).readlines())
524
repo = self.branch.repository
527
last_revision = self._file_revision(self, file_id)
528
base_vf = self._repository.texts
529
if base_vf not in vf.fallback_versionedfiles:
530
vf.fallback_versionedfiles.append(base_vf)
533
inventory = property(_get_inventory,
534
doc="Inventory of this Tree")
536
def _check_retrieved(self, ie, f):
539
fp = fingerprint_file(f)
542
if ie.text_size is not None:
543
if ie.text_size != fp['size']:
544
raise BzrError("mismatched size for file %r in %r" % (ie.file_id, self._store),
545
["inventory expects %d bytes" % ie.text_size,
546
"file is actually %d bytes" % fp['size'],
547
"store is probably damaged/corrupt"])
549
if ie.text_sha1 != fp['sha1']:
550
raise BzrError("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
551
["inventory expects %s" % ie.text_sha1,
552
"file is actually %s" % fp['sha1'],
553
"store is probably damaged/corrupt"])
556
486
def path2id(self, path):
557
487
"""Return the id for path in this tree."""
558
return self._inventory.path2id(path)
560
def paths2ids(self, paths, trees=[], require_versioned=True):
561
"""Return all the ids that can be reached by walking from paths.
563
Each path is looked up in this tree and any extras provided in
564
trees, and this is repeated recursively: the children in an extra tree
565
of a directory that has been renamed under a provided path in this tree
566
are all returned, even if none exist under a provided path in this
567
tree, and vice versa.
569
:param paths: An iterable of paths to start converting to ids from.
570
Alternatively, if paths is None, no ids should be calculated and None
571
will be returned. This is offered to make calling the api unconditional
572
for code that *might* take a list of files.
573
:param trees: Additional trees to consider.
574
:param require_versioned: If False, do not raise NotVersionedError if
575
an element of paths is not versioned in this tree and all of trees.
577
return find_ids_across_trees(paths, [self] + list(trees), require_versioned)
579
def iter_children(self, file_id):
580
entry = self.iter_entries_by_dir([file_id]).next()[1]
581
for child in getattr(entry, 'children', {}).itervalues():
488
raise NotImplementedError(self.path2id)
490
def is_versioned(self, path):
491
"""Check whether path is versioned.
493
:param path: Path to check
496
return self.path2id(path) is not None
498
def find_related_paths_across_trees(self, paths, trees=[],
499
require_versioned=True):
500
"""Find related paths in tree corresponding to specified filenames in any
503
All matches in all trees will be used, and all children of matched
504
directories will be used.
506
:param paths: The filenames to find related paths for (if None, returns
508
:param trees: The trees to find file_ids within
509
:param require_versioned: if true, all specified filenames must occur in
511
:return: a set of paths for the specified filenames and their children
514
raise NotImplementedError(self.find_related_paths_across_trees)
584
516
def lock_read(self):
517
"""Lock this tree for multiple read only operations.
519
:return: A breezy.lock.LogicalLockResult.
521
return lock.LogicalLockResult(self.unlock)
587
523
def revision_tree(self, revision_id):
588
524
"""Obtain a revision tree for the revision revision_id.
720
653
searcher = default_searcher
724
######################################################################
727
# TODO: Merge these two functions into a single one that can operate
728
# on either a whole tree or a set of files.
730
# TODO: Return the diff in order by filename, not by category or in
731
# random order. Can probably be done by lock-stepping through the
732
# filenames from both trees.
735
def file_status(filename, old_tree, new_tree):
736
"""Return single-letter status, old and new names for a file.
738
The complexity here is in deciding how to represent renames;
739
many complex cases are possible.
741
old_inv = old_tree.inventory
742
new_inv = new_tree.inventory
743
new_id = new_inv.path2id(filename)
744
old_id = old_inv.path2id(filename)
746
if not new_id and not old_id:
747
# easy: doesn't exist in either; not versioned at all
748
if new_tree.is_ignored(filename):
749
return 'I', None, None
751
return '?', None, None
753
# There is now a file of this name, great.
756
# There is no longer a file of this name, but we can describe
757
# what happened to the file that used to have
758
# this name. There are two possibilities: either it was
759
# deleted entirely, or renamed.
760
if new_inv.has_id(old_id):
761
return 'X', old_inv.id2path(old_id), new_inv.id2path(old_id)
763
return 'D', old_inv.id2path(old_id), None
765
# if the file_id is new in this revision, it is added
766
if new_id and not old_inv.has_id(new_id):
769
# if there used to be a file of this name, but that ID has now
770
# disappeared, it is deleted
771
if old_id and not new_inv.has_id(old_id):
777
@deprecated_function(deprecated_in((1, 9, 0)))
778
def find_renames(old_inv, new_inv):
779
for file_id in old_inv:
780
if file_id not in new_inv:
782
old_name = old_inv.id2path(file_id)
783
new_name = new_inv.id2path(file_id)
784
if old_name != new_name:
785
yield (old_name, new_name)
788
def find_ids_across_trees(filenames, trees, require_versioned=True):
789
"""Find the ids corresponding to specified filenames.
791
All matches in all trees will be used, and all children of matched
792
directories will be used.
794
:param filenames: The filenames to find file_ids for (if None, returns
796
:param trees: The trees to find file_ids within
797
:param require_versioned: if true, all specified filenames must occur in
799
:return: a set of file ids for the specified filenames and their children.
803
specified_path_ids = _find_ids_across_trees(filenames, trees,
805
return _find_children_across_trees(specified_path_ids, trees)
808
def _find_ids_across_trees(filenames, trees, require_versioned):
809
"""Find the ids corresponding to specified filenames.
811
All matches in all trees will be used, but subdirectories are not scanned.
813
:param filenames: The filenames to find file_ids for
814
:param trees: The trees to find file_ids within
815
:param require_versioned: if true, all specified filenames must occur in
817
:return: a set of file ids for the specified filenames
820
interesting_ids = set()
821
for tree_path in filenames:
824
file_id = tree.path2id(tree_path)
825
if file_id is not None:
826
interesting_ids.add(file_id)
829
not_versioned.append(tree_path)
830
if len(not_versioned) > 0 and require_versioned:
831
raise errors.PathsNotVersionedError(not_versioned)
832
return interesting_ids
835
def _find_children_across_trees(specified_ids, trees):
836
"""Return a set including specified ids and their children.
838
All matches in all trees will be used.
840
:param trees: The trees to find file_ids within
841
:return: a set containing all specified ids and their children
843
interesting_ids = set(specified_ids)
844
pending = interesting_ids
845
# now handle children of interesting ids
846
# we loop so that we handle all children of each id in both trees
847
while len(pending) > 0:
849
for file_id in pending:
851
if not tree.has_or_had_id(file_id):
853
for child_id in tree.iter_children(file_id):
854
if child_id not in interesting_ids:
855
new_pending.add(child_id)
856
interesting_ids.update(new_pending)
857
pending = new_pending
858
return interesting_ids
656
def archive(self, format, name, root='', subdir=None,
658
"""Create an archive of this tree.
660
:param format: Format name (e.g. 'tar')
661
:param name: target file name
662
:param root: Root directory name (or None)
663
:param subdir: Subdirectory to export (or None)
664
:return: Iterator over archive chunks
666
from .archive import create_archive
667
with self.lock_read():
668
return create_archive(format, self, name, root,
669
subdir, force_mtime=force_mtime)
672
def versionable_kind(cls, kind):
673
"""Check if this tree support versioning a specific file kind."""
674
return (kind in ('file', 'directory', 'symlink', 'tree-reference'))
861
677
class InterTree(InterObject):
880
def _changes_from_entries(self, source_entry, target_entry,
881
source_path=None, target_path=None):
697
def is_compatible(kls, source, target):
698
# The default implementation is naive and uses the public API, so
699
# it works for all trees.
702
def _changes_from_entries(self, source_entry, target_entry, source_path,
882
704
"""Generate a iter_changes tuple between source_entry and target_entry.
884
706
:param source_entry: An inventory entry from self.source, or None.
885
707
:param target_entry: An inventory entry from self.target, or None.
886
:param source_path: The path of source_entry, if known. If not known
887
it will be looked up.
888
:param target_path: The path of target_entry, if known. If not known
889
it will be looked up.
708
:param source_path: The path of source_entry.
709
:param target_path: The path of target_entry.
890
710
:return: A tuple, item 0 of which is an iter_changes result tuple, and
891
711
item 1 is True if there are any changes in the result tuple.
930
746
if source_kind != target_kind:
931
747
changed_content = True
932
748
elif source_kind == 'file':
933
if (self.source.get_file_sha1(file_id, source_path, source_stat) !=
934
self.target.get_file_sha1(file_id, target_path, target_stat)):
749
if not self.file_content_matches(
750
source_path, target_path,
751
file_id, file_id, source_stat, target_stat):
935
752
changed_content = True
936
753
elif source_kind == 'symlink':
937
if (self.source.get_symlink_target(file_id) !=
938
self.target.get_symlink_target(file_id)):
939
changed_content = True
940
# XXX: Yes, the indentation below is wrong. But fixing it broke
941
# test_merge.TestMergerEntriesLCAOnDisk.
942
# test_nested_tree_subtree_renamed_and_modified. We'll wait for
943
# the fix from bzr.dev -- vila 2009026
944
elif source_kind == 'tree-reference':
945
if (self.source.get_reference_revision(file_id, source_path)
946
!= self.target.get_reference_revision(file_id, target_path)):
947
changed_content = True
754
if (self.source.get_symlink_target(source_path, file_id) !=
755
self.target.get_symlink_target(target_path, file_id)):
756
changed_content = True
757
elif source_kind == 'tree-reference':
758
if (self.source.get_reference_revision(source_path, file_id)
759
!= self.target.get_reference_revision(target_path, file_id)):
760
changed_content = True
948
761
parent = (source_parent, target_parent)
949
762
name = (source_name, target_name)
950
763
executable = (source_executable, target_executable)
951
if (changed_content is not False or versioned[0] != versioned[1]
952
or parent[0] != parent[1] or name[0] != name[1] or
953
executable[0] != executable[1]):
764
if (changed_content is not False or versioned[0] != versioned[1] or
765
parent[0] != parent[1] or name[0] != name[1] or
766
executable[0] != executable[1]):
957
770
return (file_id, (source_path, target_path), changed_content,
958
771
versioned, parent, name, kind, executable), changes
961
773
def compare(self, want_unchanged=False, specific_files=None,
962
extra_trees=None, require_versioned=False, include_root=False,
963
want_unversioned=False):
774
extra_trees=None, require_versioned=False, include_root=False,
775
want_unversioned=False):
964
776
"""Return the changes from source to target.
966
778
:return: A TreeDelta.
980
792
trees = (self.source,)
981
793
if extra_trees is not None:
982
794
trees = trees + tuple(extra_trees)
983
# target is usually the newer tree:
984
specific_file_ids = self.target.paths2ids(specific_files, trees,
985
require_versioned=require_versioned)
986
if specific_files and not specific_file_ids:
987
# All files are unversioned, so just return an empty delta
988
# _compare_trees would think we want a complete delta
989
result = delta.TreeDelta()
990
fake_entry = InventoryFile('unused', 'unused', 'unused')
991
result.unversioned = [(path, None,
992
self.target._comparison_data(fake_entry, path)[0]) for path in
995
return delta._compare_trees(self.source, self.target, want_unchanged,
996
specific_files, include_root, extra_trees=extra_trees,
997
require_versioned=require_versioned,
998
want_unversioned=want_unversioned)
795
with self.lock_read():
796
return delta._compare_trees(self.source, self.target, want_unchanged,
797
specific_files, include_root, extra_trees=extra_trees,
798
require_versioned=require_versioned,
799
want_unversioned=want_unversioned)
1000
801
def iter_changes(self, include_unchanged=False,
1001
specific_files=None, pb=None, extra_trees=[],
1002
require_versioned=True, want_unversioned=False):
802
specific_files=None, pb=None, extra_trees=[],
803
require_versioned=True, want_unversioned=False):
1003
804
"""Generate an iterator of changes between trees.
1005
806
A tuple is returned:
1033
834
output. An unversioned file is defined as one with (False, False)
1034
835
for the versioned pair.
1036
lookup_trees = [self.source]
1038
lookup_trees.extend(extra_trees)
840
extra_trees = list(extra_trees)
1039
841
# The ids of items we need to examine to insure delta consistency.
1040
842
precise_file_ids = set()
1041
843
changed_file_ids = []
1042
844
if specific_files == []:
1043
specific_file_ids = []
845
target_specific_files = []
846
source_specific_files = []
1045
specific_file_ids = self.target.paths2ids(specific_files,
1046
lookup_trees, require_versioned=require_versioned)
848
target_specific_files = self.target.find_related_paths_across_trees(
849
specific_files, [self.source] + extra_trees,
850
require_versioned=require_versioned)
851
source_specific_files = self.source.find_related_paths_across_trees(
852
specific_files, [self.target] + extra_trees,
853
require_versioned=require_versioned)
1047
854
if specific_files is not None:
1048
855
# reparented or added entries must have their parents included
1049
856
# so that valid deltas can be created. The seen_parents set
1055
862
seen_dirs = set()
1056
863
if want_unversioned:
1057
864
all_unversioned = sorted([(p.split('/'), p) for p in
1058
self.target.extras()
1059
if specific_files is None or
1060
osutils.is_inside_any(specific_files, p)])
1061
all_unversioned = deque(all_unversioned)
866
if specific_files is None or
867
osutils.is_inside_any(specific_files, p)])
868
all_unversioned = collections.deque(all_unversioned)
1063
all_unversioned = deque()
870
all_unversioned = collections.deque()
1065
872
from_entries_by_dir = list(self.source.iter_entries_by_dir(
1066
specific_file_ids=specific_file_ids))
873
specific_files=source_specific_files))
1067
874
from_data = dict((e.file_id, (p, e)) for p, e in from_entries_by_dir)
1068
875
to_entries_by_dir = list(self.target.iter_entries_by_dir(
1069
specific_file_ids=specific_file_ids))
876
specific_files=target_specific_files))
1070
877
num_entries = len(from_entries_by_dir) + len(to_entries_by_dir)
1072
879
# the unversioned path lookup only occurs on real trees - where there
1073
880
# can be extras. So the fake_entry is solely used to look up
1074
881
# executable it values when execute is not supported.
1075
fake_entry = InventoryFile('unused', 'unused', 'unused')
882
fake_entry = TreeFile()
1076
883
for target_path, target_entry in to_entries_by_dir:
1077
884
while (all_unversioned and
1078
all_unversioned[0][0] < target_path.split('/')):
885
all_unversioned[0][0] < target_path.split('/')):
1079
886
unversioned_path = all_unversioned.popleft()
1080
887
target_kind, target_executable, target_stat = \
1081
self.target._comparison_data(fake_entry, unversioned_path[1])
888
self.target._comparison_data(
889
fake_entry, unversioned_path[1])
1082
890
yield (None, (None, unversioned_path[1]), True, (False, False),
1084
(None, unversioned_path[0][-1]),
1085
(None, target_kind),
1086
(None, target_executable))
892
(None, unversioned_path[0][-1]),
894
(None, target_executable))
1087
895
source_path, source_entry = from_data.get(target_entry.file_id,
1089
897
result, changes = self._changes_from_entries(source_entry,
1090
target_entry, source_path=source_path, target_path=target_path)
898
target_entry, source_path=source_path, target_path=target_path)
1091
899
to_paths[result[0]] = result[1][1]
1092
900
entry_count += 1
1093
901
if result[3][0]:
1116
924
to_kind, to_executable, to_stat = \
1117
925
self.target._comparison_data(fake_entry, unversioned_path[1])
1118
926
yield (None, (None, unversioned_path[1]), True, (False, False),
1120
(None, unversioned_path[0][-1]),
1122
(None, to_executable))
928
(None, unversioned_path[0][-1]),
930
(None, to_executable))
1123
931
# Yield all remaining source paths
1124
932
for path, from_entry in from_entries_by_dir:
1125
933
file_id = from_entry.file_id
1126
934
if file_id in to_paths:
1127
935
# already returned
1129
if file_id not in self.target.all_file_ids():
1130
# common case - paths we have not emitted are not present in
1134
to_path = self.target.id2path(file_id)
937
to_path = find_previous_path(self.source, self.target, path)
1135
938
entry_count += 1
1136
939
if pb is not None:
1137
940
pb.update('comparing files', entry_count, num_entries)
1239
1048
precise_file_ids.add(new_parent_id)
1241
1050
if (result[6][0] == 'directory' and
1242
result[6][1] != 'directory'):
1051
result[6][1] != 'directory'):
1243
1052
# This stopped being a directory, the old children have
1244
1053
# to be included.
1245
if old_entry is None:
1054
if source_entry is None:
1246
1055
# Reusing a discarded change.
1247
old_entry = self._get_entry(self.source, file_id)
1248
for child in old_entry.children.values():
1249
precise_file_ids.add(child.file_id)
1056
source_entry = self._get_entry(
1057
self.source, result[1][0])
1058
precise_file_ids.update(
1060
for child in self.source.iter_child_entries(result[1][0]))
1250
1061
changed_file_ids.add(result[0])
1064
def file_content_matches(
1065
self, source_path, target_path,
1066
source_file_id=None, target_file_id=None,
1067
source_stat=None, target_stat=None):
1068
"""Check if two files are the same in the source and target trees.
1070
This only checks that the contents of the files are the same,
1071
it does not touch anything else.
1073
:param source_path: Path of the file in the source tree
1074
:param target_path: Path of the file in the target tree
1075
:param source_file_id: Optional file id of the file in the source tree
1076
:param target_file_id: Optional file id of the file in the target tree
1077
:param source_stat: Optional stat value of the file in the source tree
1078
:param target_stat: Optional stat value of the file in the target tree
1079
:return: Boolean indicating whether the files have the same contents
1081
with self.lock_read():
1082
source_verifier_kind, source_verifier_data = (
1083
self.source.get_file_verifier(
1084
source_path, source_file_id, source_stat))
1085
target_verifier_kind, target_verifier_data = (
1086
self.target.get_file_verifier(
1087
target_path, target_file_id, target_stat))
1088
if source_verifier_kind == target_verifier_kind:
1089
return (source_verifier_data == target_verifier_data)
1090
# Fall back to SHA1 for now
1091
if source_verifier_kind != "SHA1":
1092
source_sha1 = self.source.get_file_sha1(
1093
source_path, source_file_id, source_stat)
1095
source_sha1 = source_verifier_data
1096
if target_verifier_kind != "SHA1":
1097
target_sha1 = self.target.get_file_sha1(
1098
target_path, target_file_id, target_stat)
1100
target_sha1 = target_verifier_data
1101
return (source_sha1 == target_sha1)
1104
InterTree.register_optimiser(InterTree)
1254
1107
class MultiWalker(object):
1255
1108
"""Walk multiple trees simultaneously, getting combined results."""
1473
1326
# the lookup_by_file_id will be removing anything processed
1474
1327
# from the extras cache
1475
1328
other_extra.pop(file_id)
1476
other_values = [(None, None) for i in xrange(idx)]
1329
other_values = [(None, None)] * idx
1477
1330
other_values.append((other_path, other_ie))
1478
for alt_idx, alt_extra in enumerate(self._others_extra[idx+1:]):
1331
for alt_idx, alt_extra in enumerate(self._others_extra[idx + 1:]):
1479
1332
alt_idx = alt_idx + idx + 1
1480
1333
alt_extra = self._others_extra[alt_idx]
1481
1334
alt_tree = self._other_trees[alt_idx]
1482
1335
other_values.append(self._lookup_by_file_id(
1483
alt_extra, alt_tree, file_id))
1336
alt_extra, alt_tree, file_id))
1484
1337
yield other_path, file_id, None, other_values
1340
def find_previous_paths(from_tree, to_tree, paths):
1341
"""Find previous tree paths.
1343
:param from_tree: From tree
1344
:param to_tree: To tree
1345
:param paths: Iterable over paths to search for
1346
:return: Dictionary mapping from from_tree paths to paths in to_tree, or
1347
None if there is no equivalent path.
1351
ret[path] = find_previous_path(from_tree, to_tree, path)
1355
def find_previous_path(from_tree, to_tree, path, file_id=None):
1356
"""Find previous tree path.
1358
:param from_tree: From tree
1359
:param to_tree: To tree
1360
:param path: Path to search for
1361
:return: path in to_tree, or None if there is no equivalent path.
1364
file_id = from_tree.path2id(path)
1366
raise errors.NoSuchFile(path)
1368
return to_tree.id2path(file_id)
1369
except errors.NoSuchId:
1373
def get_canonical_path(tree, path, normalize):
1374
"""Find the canonical path of an item, ignoring case.
1376
:param tree: Tree to traverse
1377
:param path: Case-insensitive path to look up
1378
:param normalize: Function to normalize a filename for comparison
1379
:return: The canonical path
1382
cur_id = tree.get_root_id()
1384
bit_iter = iter(path.split("/"))
1385
for elt in bit_iter:
1386
lelt = normalize(elt)
1389
for child in tree.iter_child_entries(cur_path, cur_id):
1391
if child.name == elt:
1392
# if we found an exact match, we can stop now; if
1393
# we found an approximate match we need to keep
1394
# searching because there might be an exact match
1396
cur_id = child.file_id
1397
new_path = osutils.pathjoin(cur_path, child.name)
1399
elif normalize(child.name) == lelt:
1400
cur_id = child.file_id
1401
new_path = osutils.pathjoin(cur_path, child.name)
1402
except errors.NoSuchId:
1403
# before a change is committed we can see this error...
1405
except errors.NotADirectory:
1410
# got to the end of this directory and no entries matched.
1411
# Return what matched so far, plus the rest as specified.
1412
cur_path = osutils.pathjoin(cur_path, elt, *list(bit_iter))