352
392
cur_file = (self.get_file_text(file_id),)
353
393
yield identifier, cur_file
355
def get_symlink_target(self, file_id):
395
def get_symlink_target(self, file_id, path=None):
356
396
"""Get the target for a given file_id.
358
398
It is assumed that the caller already knows that file_id is referencing
360
400
: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
361
404
:return: The path the symlink points to.
363
406
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
408
def get_root_id(self):
444
409
"""Return the file_id for the root of this tree."""
445
410
raise NotImplementedError(self.get_root_id)
503
468
except errors.NoSuchRevisionInTree:
504
469
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
471
def _get_file_revision(self, file_id, vf, tree_revision):
516
472
"""Ensure that file_id, tree_revision is in vf to plan the merge."""
518
474
if getattr(self, '_repository', None) is None:
519
475
last_revision = tree_revision
520
parent_keys = [(file_id, self._file_revision(t, file_id)) for t in
476
parent_keys = [(file_id, t.get_file_revision(file_id)) for t in
521
477
self._iter_parent_trees()]
522
478
vf.add_lines((file_id, last_revision), parent_keys,
523
self.get_file(file_id).readlines())
479
self.get_file_lines(file_id))
524
480
repo = self.branch.repository
525
481
base_vf = repo.texts
527
last_revision = self._file_revision(self, file_id)
483
last_revision = self.get_file_revision(file_id)
528
484
base_vf = self._repository.texts
529
485
if base_vf not in vf.fallback_versionedfiles:
530
486
vf.fallback_versionedfiles.append(base_vf)
531
487
return last_revision
533
inventory = property(_get_inventory,
534
doc="Inventory of this Tree")
536
489
def _check_retrieved(self, ie, f):
537
490
if not __debug__:
539
fp = fingerprint_file(f)
492
fp = osutils.fingerprint_file(f)
542
495
if ie.text_size is not None:
543
496
if ie.text_size != fp['size']:
544
raise BzrError("mismatched size for file %r in %r" % (ie.file_id, self._store),
497
raise errors.BzrError(
498
"mismatched size for file %r in %r" %
499
(ie.file_id, self._store),
545
500
["inventory expects %d bytes" % ie.text_size,
546
501
"file is actually %d bytes" % fp['size'],
547
502
"store is probably damaged/corrupt"])
549
504
if ie.text_sha1 != fp['sha1']:
550
raise BzrError("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
505
raise errors.BzrError("wrong SHA-1 for file %r in %r" %
506
(ie.file_id, self._store),
551
507
["inventory expects %s" % ie.text_sha1,
552
508
"file is actually %s" % fp['sha1'],
553
509
"store is probably damaged/corrupt"])
556
511
def path2id(self, path):
557
512
"""Return the id for path in this tree."""
558
return self._inventory.path2id(path)
513
raise NotImplementedError(self.path2id)
560
515
def paths2ids(self, paths, trees=[], require_versioned=True):
561
516
"""Return all the ids that can be reached by walking from paths.
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.
682
class InventoryTree(Tree):
683
"""A tree that relies on an inventory for its metadata.
685
Trees contain an `Inventory` object, and also know how to retrieve
686
file texts mentioned in the inventory, either from a working
687
directory or from a store.
689
It is possible for trees to contain files that are not described
690
in their inventory or vice versa; for this use `filenames()`.
692
Subclasses should set the _inventory attribute, which is considered
693
private to external API users.
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)
696
def get_canonical_inventory_paths(self, paths):
697
"""Like get_canonical_inventory_path() but works on multiple items.
699
:param paths: A sequence of paths relative to the root of the tree.
700
:return: A list of paths, with each item the corresponding input path
701
adjusted to account for existing elements that match case
704
return list(self._yield_canonical_inventory_paths(paths))
706
def get_canonical_inventory_path(self, path):
707
"""Returns the first inventory item that case-insensitively matches path.
709
If a path matches exactly, it is returned. If no path matches exactly
710
but more than one path matches case-insensitively, it is implementation
711
defined which is returned.
713
If no path matches case-insensitively, the input path is returned, but
714
with as many path entries that do exist changed to their canonical
717
If you need to resolve many names from the same tree, you should
718
use get_canonical_inventory_paths() to avoid O(N) behaviour.
720
:param path: A paths relative to the root of the tree.
721
:return: The input path adjusted to account for existing elements
722
that match case insensitively.
724
return self._yield_canonical_inventory_paths([path]).next()
726
def _yield_canonical_inventory_paths(self, paths):
728
# First, if the path as specified exists exactly, just use it.
729
if self.path2id(path) is not None:
733
cur_id = self.get_root_id()
735
bit_iter = iter(path.split("/"))
739
for child in self.iter_children(cur_id):
741
# XXX: it seem like if the child is known to be in the
742
# tree, we shouldn't need to go from its id back to
743
# its path -- mbp 2010-02-11
745
# XXX: it seems like we could be more efficient
746
# by just directly looking up the original name and
747
# only then searching all children; also by not
748
# chopping paths so much. -- mbp 2010-02-11
749
child_base = os.path.basename(self.id2path(child))
750
if (child_base == elt):
751
# if we found an exact match, we can stop now; if
752
# we found an approximate match we need to keep
753
# searching because there might be an exact match
756
new_path = osutils.pathjoin(cur_path, child_base)
758
elif child_base.lower() == lelt:
760
new_path = osutils.pathjoin(cur_path, child_base)
761
except errors.NoSuchId:
762
# before a change is committed we can see this error...
767
# got to the end of this directory and no entries matched.
768
# Return what matched so far, plus the rest as specified.
769
cur_path = osutils.pathjoin(cur_path, elt, *list(bit_iter))
774
@deprecated_method(deprecated_in((2, 5, 0)))
775
def _get_inventory(self):
776
return self._inventory
778
inventory = property(_get_inventory,
779
doc="Inventory of this Tree")
781
def _get_root_inventory(self):
782
return self._inventory
784
root_inventory = property(_get_root_inventory,
785
doc="Root inventory of this tree")
787
def _unpack_file_id(self, file_id):
788
"""Find the inventory and inventory file id for a tree file id.
790
:param file_id: The tree file id, as bytestring or tuple
791
:return: Inventory and inventory file id
793
if isinstance(file_id, tuple):
794
if len(file_id) != 1:
795
raise ValueError("nested trees not yet supported: %r" % file_id)
797
return self.root_inventory, file_id
800
def path2id(self, path):
801
"""Return the id for path in this tree."""
802
return self._path2inv_file_id(path)[1]
804
def _path2inv_file_id(self, path):
805
"""Lookup a inventory and inventory file id by path.
807
:param path: Path to look up
808
:return: tuple with inventory and inventory file id
810
# FIXME: Support nested trees
811
return self.root_inventory, self.root_inventory.path2id(path)
813
def id2path(self, file_id):
814
"""Return the path for a file id.
818
inventory, file_id = self._unpack_file_id(file_id)
819
return inventory.id2path(file_id)
821
def has_id(self, file_id):
822
inventory, file_id = self._unpack_file_id(file_id)
823
return inventory.has_id(file_id)
825
def has_or_had_id(self, file_id):
826
inventory, file_id = self._unpack_file_id(file_id)
827
return inventory.has_id(file_id)
829
def all_file_ids(self):
831
[entry.file_id for path, entry in self.iter_entries_by_dir()])
833
@deprecated_method(deprecated_in((2, 4, 0)))
835
return iter(self.all_file_ids())
837
def filter_unversioned_files(self, paths):
838
"""Filter out paths that are versioned.
840
:return: set of paths.
842
# NB: we specifically *don't* call self.has_filename, because for
843
# WorkingTrees that can indicate files that exist on disk but that
845
return set((p for p in paths if self.path2id(p) is None))
848
def iter_entries_by_dir(self, specific_file_ids=None, yield_parents=False):
849
"""Walk the tree in 'by_dir' order.
851
This will yield each entry in the tree as a (path, entry) tuple.
852
The order that they are yielded is:
854
See Tree.iter_entries_by_dir for details.
856
:param yield_parents: If True, yield the parents from the root leading
857
down to specific_file_ids that have been requested. This has no
858
impact if specific_file_ids is None.
860
if specific_file_ids is None:
861
inventory_file_ids = None
863
inventory_file_ids = []
864
for tree_file_id in specific_file_ids:
865
inventory, inv_file_id = self._unpack_file_id(tree_file_id)
866
if not inventory is self.root_inventory: # for now
867
raise AssertionError("%r != %r" % (
868
inventory, self.root_inventory))
869
inventory_file_ids.append(inv_file_id)
870
# FIXME: Handle nested trees
871
return self.root_inventory.iter_entries_by_dir(
872
specific_file_ids=inventory_file_ids, yield_parents=yield_parents)
874
@deprecated_method(deprecated_in((2, 5, 0)))
875
def get_file_by_path(self, path):
876
return self.get_file(self.path2id(path), path)
878
def iter_children(self, file_id, path=None):
879
"""See Tree.iter_children."""
880
entry = self.iter_entries_by_dir([file_id]).next()[1]
881
for child in getattr(entry, 'children', {}).itervalues():
788
885
def find_ids_across_trees(filenames, trees, require_versioned=True):
1245
1344
if old_entry is None:
1246
1345
# Reusing a discarded change.
1247
1346
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)
1347
for child in self.source.iter_children(file_id):
1348
precise_file_ids.add(child)
1250
1349
changed_file_ids.add(result[0])
1353
def file_content_matches(self, source_file_id, target_file_id,
1354
source_path=None, target_path=None, source_stat=None, target_stat=None):
1355
"""Check if two files are the same in the source and target trees.
1357
This only checks that the contents of the files are the same,
1358
it does not touch anything else.
1360
:param source_file_id: File id of the file in the source tree
1361
:param target_file_id: File id of the file in the target tree
1362
:param source_path: Path of the file in the source tree
1363
:param target_path: Path of the file in the target tree
1364
:param source_stat: Optional stat value of the file in the source tree
1365
:param target_stat: Optional stat value of the file in the target tree
1366
:return: Boolean indicating whether the files have the same contents
1368
source_verifier_kind, source_verifier_data = self.source.get_file_verifier(
1369
source_file_id, source_path, source_stat)
1370
target_verifier_kind, target_verifier_data = self.target.get_file_verifier(
1371
target_file_id, target_path, target_stat)
1372
if source_verifier_kind == target_verifier_kind:
1373
return (source_verifier_data == target_verifier_data)
1374
# Fall back to SHA1 for now
1375
if source_verifier_kind != "SHA1":
1376
source_sha1 = self.source.get_file_sha1(source_file_id,
1377
source_path, source_stat)
1379
source_sha1 = source_verifier_data
1380
if target_verifier_kind != "SHA1":
1381
target_sha1 = self.target.get_file_sha1(target_file_id,
1382
target_path, target_stat)
1384
target_sha1 = target_verifier_data
1385
return (source_sha1 == target_sha1)
1387
InterTree.register_optimiser(InterTree)
1254
1390
class MultiWalker(object):
1255
1391
"""Walk multiple trees simultaneously, getting combined results."""
added
removed
bzrlib/tree.py
