306
301
:param file_id: The file_id of the file.
307
302
:param path: The path of the file.
309
303
If both file_id and path are supplied, an implementation may use
312
306
return osutils.split_lines(self.get_file_text(file_id, path))
314
def get_file_verifier(self, file_id, path=None, stat_value=None):
315
"""Return a verifier for a file.
317
The default implementation returns a sha1.
319
:param file_id: The handle for this file.
320
:param path: The path that this file can be found at.
321
These must point to the same object.
322
:param stat_value: Optional stat value for the object
323
:return: Tuple with verifier name and verifier data
325
return ("SHA1", self.get_file_sha1(file_id, path=path,
326
stat_value=stat_value))
328
def get_file_sha1(self, file_id, path=None, stat_value=None):
329
"""Return the SHA1 file for a file.
331
:note: callers should use get_file_verifier instead
332
where possible, as the underlying repository implementation may
333
have quicker access to a non-sha1 verifier.
335
:param file_id: The handle for this file.
336
:param path: The path that this file can be found at.
337
These must point to the same object.
338
:param stat_value: Optional stat value for the object
340
raise NotImplementedError(self.get_file_sha1)
342
308
def get_file_mtime(self, file_id, path=None):
343
309
"""Return the modification time for a file.
392
352
cur_file = (self.get_file_text(file_id),)
393
353
yield identifier, cur_file
395
def get_symlink_target(self, file_id, path=None):
355
def get_symlink_target(self, file_id):
396
356
"""Get the target for a given file_id.
398
358
It is assumed that the caller already knows that file_id is referencing
400
360
:param file_id: Handle for the symlink entry.
401
:param path: The path of the file.
402
If both file_id and path are supplied, an implementation may use
404
361
:return: The path the symlink points to.
406
363
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))
408
443
def get_root_id(self):
409
444
"""Return the file_id for the root of this tree."""
410
445
raise NotImplementedError(self.get_root_id)
468
503
except errors.NoSuchRevisionInTree:
469
504
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()
471
515
def _get_file_revision(self, file_id, vf, tree_revision):
472
516
"""Ensure that file_id, tree_revision is in vf to plan the merge."""
474
518
if getattr(self, '_repository', None) is None:
475
519
last_revision = tree_revision
476
parent_keys = [(file_id, t.get_file_revision(file_id)) for t in
520
parent_keys = [(file_id, self._file_revision(t, file_id)) for t in
477
521
self._iter_parent_trees()]
478
522
vf.add_lines((file_id, last_revision), parent_keys,
479
523
self.get_file_lines(file_id))
480
524
repo = self.branch.repository
481
525
base_vf = repo.texts
483
last_revision = self.get_file_revision(file_id)
527
last_revision = self._file_revision(self, file_id)
484
528
base_vf = self._repository.texts
485
529
if base_vf not in vf.fallback_versionedfiles:
486
530
vf.fallback_versionedfiles.append(base_vf)
487
531
return last_revision
533
inventory = property(_get_inventory,
534
doc="Inventory of this Tree")
489
536
def _check_retrieved(self, ie, f):
490
537
if not __debug__:
492
fp = osutils.fingerprint_file(f)
539
fp = fingerprint_file(f)
495
542
if ie.text_size is not None:
496
543
if ie.text_size != fp['size']:
497
raise errors.BzrError(
498
"mismatched size for file %r in %r" %
499
(ie.file_id, self._store),
544
raise BzrError("mismatched size for file %r in %r" % (ie.file_id, self._store),
500
545
["inventory expects %d bytes" % ie.text_size,
501
546
"file is actually %d bytes" % fp['size'],
502
547
"store is probably damaged/corrupt"])
504
549
if ie.text_sha1 != fp['sha1']:
505
raise errors.BzrError("wrong SHA-1 for file %r in %r" %
506
(ie.file_id, self._store),
550
raise BzrError("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
507
551
["inventory expects %s" % ie.text_sha1,
508
552
"file is actually %s" % fp['sha1'],
509
553
"store is probably damaged/corrupt"])
511
556
def path2id(self, path):
512
557
"""Return the id for path in this tree."""
513
raise NotImplementedError(self.path2id)
558
return self._inventory.path2id(path)
515
560
def paths2ids(self, paths, trees=[], require_versioned=True):
516
561
"""Return all the ids that can be reached by walking from paths.
679
class InventoryTree(Tree):
680
"""A tree that relies on an inventory for its metadata.
682
Trees contain an `Inventory` object, and also know how to retrieve
683
file texts mentioned in the inventory, either from a working
684
directory or from a store.
686
It is possible for trees to contain files that are not described
687
in their inventory or vice versa; for this use `filenames()`.
689
Subclasses should set the _inventory attribute, which is considered
690
private to external API users.
728
######################################################################
731
# TODO: Merge these two functions into a single one that can operate
732
# on either a whole tree or a set of files.
734
# TODO: Return the diff in order by filename, not by category or in
735
# random order. Can probably be done by lock-stepping through the
736
# filenames from both trees.
739
def file_status(filename, old_tree, new_tree):
740
"""Return single-letter status, old and new names for a file.
742
The complexity here is in deciding how to represent renames;
743
many complex cases are possible.
693
def get_canonical_inventory_paths(self, paths):
694
"""Like get_canonical_inventory_path() but works on multiple items.
696
:param paths: A sequence of paths relative to the root of the tree.
697
:return: A list of paths, with each item the corresponding input path
698
adjusted to account for existing elements that match case
701
return list(self._yield_canonical_inventory_paths(paths))
703
def get_canonical_inventory_path(self, path):
704
"""Returns the first inventory item that case-insensitively matches path.
706
If a path matches exactly, it is returned. If no path matches exactly
707
but more than one path matches case-insensitively, it is implementation
708
defined which is returned.
710
If no path matches case-insensitively, the input path is returned, but
711
with as many path entries that do exist changed to their canonical
714
If you need to resolve many names from the same tree, you should
715
use get_canonical_inventory_paths() to avoid O(N) behaviour.
717
:param path: A paths relative to the root of the tree.
718
:return: The input path adjusted to account for existing elements
719
that match case insensitively.
721
return self._yield_canonical_inventory_paths([path]).next()
723
def _yield_canonical_inventory_paths(self, paths):
725
# First, if the path as specified exists exactly, just use it.
726
if self.path2id(path) is not None:
730
cur_id = self.get_root_id()
732
bit_iter = iter(path.split("/"))
736
for child in self.iter_children(cur_id):
738
# XXX: it seem like if the child is known to be in the
739
# tree, we shouldn't need to go from its id back to
740
# its path -- mbp 2010-02-11
742
# XXX: it seems like we could be more efficient
743
# by just directly looking up the original name and
744
# only then searching all children; also by not
745
# chopping paths so much. -- mbp 2010-02-11
746
child_base = os.path.basename(self.id2path(child))
747
if (child_base == elt):
748
# if we found an exact match, we can stop now; if
749
# we found an approximate match we need to keep
750
# searching because there might be an exact match
753
new_path = osutils.pathjoin(cur_path, child_base)
755
elif child_base.lower() == lelt:
757
new_path = osutils.pathjoin(cur_path, child_base)
758
except errors.NoSuchId:
759
# before a change is committed we can see this error...
764
# got to the end of this directory and no entries matched.
765
# Return what matched so far, plus the rest as specified.
766
cur_path = osutils.pathjoin(cur_path, elt, *list(bit_iter))
771
def _get_inventory(self):
772
return self._inventory
774
inventory = property(_get_inventory,
775
doc="Inventory of this Tree")
778
def path2id(self, path):
779
"""Return the id for path in this tree."""
780
return self._inventory.path2id(path)
782
def id2path(self, file_id):
783
"""Return the path for a file id.
787
return self.inventory.id2path(file_id)
789
def has_id(self, file_id):
790
return self.inventory.has_id(file_id)
792
def has_or_had_id(self, file_id):
793
return self.inventory.has_id(file_id)
795
def all_file_ids(self):
796
return set(self.inventory)
798
@deprecated_method(deprecated_in((2, 4, 0)))
800
return iter(self.inventory)
802
def filter_unversioned_files(self, paths):
803
"""Filter out paths that are versioned.
805
:return: set of paths.
807
# NB: we specifically *don't* call self.has_filename, because for
808
# WorkingTrees that can indicate files that exist on disk but that
810
pred = self.inventory.has_filename
811
return set((p for p in paths if not pred(p)))
814
def iter_entries_by_dir(self, specific_file_ids=None, yield_parents=False):
815
"""Walk the tree in 'by_dir' order.
817
This will yield each entry in the tree as a (path, entry) tuple.
818
The order that they are yielded is:
820
See Tree.iter_entries_by_dir for details.
822
:param yield_parents: If True, yield the parents from the root leading
823
down to specific_file_ids that have been requested. This has no
824
impact if specific_file_ids is None.
826
return self.inventory.iter_entries_by_dir(
827
specific_file_ids=specific_file_ids, yield_parents=yield_parents)
829
@deprecated_method(deprecated_in((2, 5, 0)))
830
def get_file_by_path(self, path):
831
return self.get_file(self.path2id(path), path)
745
old_inv = old_tree.inventory
746
new_inv = new_tree.inventory
747
new_id = new_inv.path2id(filename)
748
old_id = old_inv.path2id(filename)
750
if not new_id and not old_id:
751
# easy: doesn't exist in either; not versioned at all
752
if new_tree.is_ignored(filename):
753
return 'I', None, None
755
return '?', None, None
757
# There is now a file of this name, great.
760
# There is no longer a file of this name, but we can describe
761
# what happened to the file that used to have
762
# this name. There are two possibilities: either it was
763
# deleted entirely, or renamed.
764
if new_inv.has_id(old_id):
765
return 'X', old_inv.id2path(old_id), new_inv.id2path(old_id)
767
return 'D', old_inv.id2path(old_id), None
769
# if the file_id is new in this revision, it is added
770
if new_id and not old_inv.has_id(new_id):
773
# if there used to be a file of this name, but that ID has now
774
# disappeared, it is deleted
775
if old_id and not new_inv.has_id(old_id):
781
@deprecated_function(deprecated_in((1, 9, 0)))
782
def find_renames(old_inv, new_inv):
783
for file_id in old_inv:
784
if file_id not in new_inv:
786
old_name = old_inv.id2path(file_id)
787
new_name = new_inv.id2path(file_id)
788
if old_name != new_name:
789
yield (old_name, new_name)
834
792
def find_ids_across_trees(filenames, trees, require_versioned=True):
1302
1254
changed_file_ids.add(result[0])
1306
def file_content_matches(self, source_file_id, target_file_id,
1307
source_path=None, target_path=None, source_stat=None, target_stat=None):
1308
"""Check if two files are the same in the source and target trees.
1310
This only checks that the contents of the files are the same,
1311
it does not touch anything else.
1313
:param source_file_id: File id of the file in the source tree
1314
:param target_file_id: File id of the file in the target tree
1315
:param source_path: Path of the file in the source tree
1316
:param target_path: Path of the file in the target tree
1317
:param source_stat: Optional stat value of the file in the source tree
1318
:param target_stat: Optional stat value of the file in the target tree
1319
:return: Boolean indicating whether the files have the same contents
1321
source_verifier_kind, source_verifier_data = self.source.get_file_verifier(
1322
source_file_id, source_path, source_stat)
1323
target_verifier_kind, target_verifier_data = self.target.get_file_verifier(
1324
target_file_id, target_path, target_stat)
1325
if source_verifier_kind == target_verifier_kind:
1326
return (source_verifier_data == target_verifier_data)
1327
# Fall back to SHA1 for now
1328
if source_verifier_kind != "SHA1":
1329
source_sha1 = self.source.get_file_sha1(source_file_id,
1330
source_path, source_stat)
1332
source_sha1 = source_verifier_data
1333
if target_verifier_kind != "SHA1":
1334
target_sha1 = self.target.get_file_sha1(target_file_id,
1335
target_path, target_stat)
1337
target_sha1 = target_verifier_data
1338
return (source_sha1 == target_sha1)
1340
InterTree.register_optimiser(InterTree)
1343
1258
class MultiWalker(object):
1344
1259
"""Walk multiple trees simultaneously, getting combined results."""
added
removed
bzrlib/tree.py
