352
394
cur_file = (self.get_file_text(file_id),)
353
395
yield identifier, cur_file
355
def get_symlink_target(self, file_id):
397
def get_symlink_target(self, file_id, path=None):
356
398
"""Get the target for a given file_id.
358
400
It is assumed that the caller already knows that file_id is referencing
360
402
:param file_id: Handle for the symlink entry.
403
:param path: The path of the file.
404
If both file_id and path are supplied, an implementation may use
361
406
:return: The path the symlink points to.
363
408
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
410
def get_root_id(self):
444
411
"""Return the file_id for the root of this tree."""
445
412
raise NotImplementedError(self.get_root_id)
503
470
except errors.NoSuchRevisionInTree:
504
471
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
473
def _get_file_revision(self, file_id, vf, tree_revision):
516
474
"""Ensure that file_id, tree_revision is in vf to plan the merge."""
518
476
if getattr(self, '_repository', None) is None:
519
477
last_revision = tree_revision
520
parent_keys = [(file_id, self._file_revision(t, file_id)) for t in
478
parent_keys = [(file_id, t.get_file_revision(file_id)) for t in
521
479
self._iter_parent_trees()]
522
480
vf.add_lines((file_id, last_revision), parent_keys,
523
self.get_file(file_id).readlines())
481
self.get_file_lines(file_id))
524
482
repo = self.branch.repository
525
483
base_vf = repo.texts
527
last_revision = self._file_revision(self, file_id)
485
last_revision = self.get_file_revision(file_id)
528
486
base_vf = self._repository.texts
529
487
if base_vf not in vf.fallback_versionedfiles:
530
488
vf.fallback_versionedfiles.append(base_vf)
531
489
return last_revision
533
inventory = property(_get_inventory,
534
doc="Inventory of this Tree")
536
491
def _check_retrieved(self, ie, f):
537
492
if not __debug__:
539
fp = fingerprint_file(f)
494
fp = osutils.fingerprint_file(f)
542
497
if ie.text_size is not None:
543
498
if ie.text_size != fp['size']:
544
raise BzrError("mismatched size for file %r in %r" % (ie.file_id, self._store),
499
raise errors.BzrError(
500
"mismatched size for file %r in %r" %
501
(ie.file_id, self._store),
545
502
["inventory expects %d bytes" % ie.text_size,
546
503
"file is actually %d bytes" % fp['size'],
547
504
"store is probably damaged/corrupt"])
549
506
if ie.text_sha1 != fp['sha1']:
550
raise BzrError("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
507
raise errors.BzrError("wrong SHA-1 for file %r in %r" %
508
(ie.file_id, self._store),
551
509
["inventory expects %s" % ie.text_sha1,
552
510
"file is actually %s" % fp['sha1'],
553
511
"store is probably damaged/corrupt"])
556
513
def path2id(self, path):
557
514
"""Return the id for path in this tree."""
558
return self._inventory.path2id(path)
515
raise NotImplementedError(self.path2id)
560
517
def paths2ids(self, paths, trees=[], require_versioned=True):
561
518
"""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.
684
class InventoryTree(Tree):
685
"""A tree that relies on an inventory for its metadata.
687
Trees contain an `Inventory` object, and also know how to retrieve
688
file texts mentioned in the inventory, either from a working
689
directory or from a store.
691
It is possible for trees to contain files that are not described
692
in their inventory or vice versa; for this use `filenames()`.
694
Subclasses should set the _inventory attribute, which is considered
695
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)
698
def get_canonical_inventory_paths(self, paths):
699
"""Like get_canonical_inventory_path() but works on multiple items.
701
:param paths: A sequence of paths relative to the root of the tree.
702
:return: A list of paths, with each item the corresponding input path
703
adjusted to account for existing elements that match case
706
return list(self._yield_canonical_inventory_paths(paths))
708
def get_canonical_inventory_path(self, path):
709
"""Returns the first inventory item that case-insensitively matches path.
711
If a path matches exactly, it is returned. If no path matches exactly
712
but more than one path matches case-insensitively, it is implementation
713
defined which is returned.
715
If no path matches case-insensitively, the input path is returned, but
716
with as many path entries that do exist changed to their canonical
719
If you need to resolve many names from the same tree, you should
720
use get_canonical_inventory_paths() to avoid O(N) behaviour.
722
:param path: A paths relative to the root of the tree.
723
:return: The input path adjusted to account for existing elements
724
that match case insensitively.
726
return next(self._yield_canonical_inventory_paths([path]))
728
def _yield_canonical_inventory_paths(self, paths):
730
# First, if the path as specified exists exactly, just use it.
731
if self.path2id(path) is not None:
735
cur_id = self.get_root_id()
737
bit_iter = iter(path.split("/"))
741
for child in self.iter_children(cur_id):
743
# XXX: it seem like if the child is known to be in the
744
# tree, we shouldn't need to go from its id back to
745
# its path -- mbp 2010-02-11
747
# XXX: it seems like we could be more efficient
748
# by just directly looking up the original name and
749
# only then searching all children; also by not
750
# chopping paths so much. -- mbp 2010-02-11
751
child_base = os.path.basename(self.id2path(child))
752
if (child_base == elt):
753
# if we found an exact match, we can stop now; if
754
# we found an approximate match we need to keep
755
# searching because there might be an exact match
758
new_path = osutils.pathjoin(cur_path, child_base)
760
elif child_base.lower() == lelt:
762
new_path = osutils.pathjoin(cur_path, child_base)
763
except errors.NoSuchId:
764
# before a change is committed we can see this error...
769
# got to the end of this directory and no entries matched.
770
# Return what matched so far, plus the rest as specified.
771
cur_path = osutils.pathjoin(cur_path, elt, *list(bit_iter))
776
def _get_root_inventory(self):
777
return self._inventory
779
root_inventory = property(_get_root_inventory,
780
doc="Root inventory of this tree")
782
def _unpack_file_id(self, file_id):
783
"""Find the inventory and inventory file id for a tree file id.
785
:param file_id: The tree file id, as bytestring or tuple
786
:return: Inventory and inventory file id
788
if isinstance(file_id, tuple):
789
if len(file_id) != 1:
790
raise ValueError("nested trees not yet supported: %r" % file_id)
792
return self.root_inventory, file_id
795
def path2id(self, path):
796
"""Return the id for path in this tree."""
797
return self._path2inv_file_id(path)[1]
799
def _path2inv_file_id(self, path):
800
"""Lookup a inventory and inventory file id by path.
802
:param path: Path to look up
803
:return: tuple with inventory and inventory file id
805
# FIXME: Support nested trees
806
return self.root_inventory, self.root_inventory.path2id(path)
808
def id2path(self, file_id):
809
"""Return the path for a file id.
813
inventory, file_id = self._unpack_file_id(file_id)
814
return inventory.id2path(file_id)
816
def has_id(self, file_id):
817
inventory, file_id = self._unpack_file_id(file_id)
818
return inventory.has_id(file_id)
820
def has_or_had_id(self, file_id):
821
inventory, file_id = self._unpack_file_id(file_id)
822
return inventory.has_id(file_id)
824
def all_file_ids(self):
825
return {entry.file_id for path, entry in self.iter_entries_by_dir()}
827
def filter_unversioned_files(self, paths):
828
"""Filter out paths that are versioned.
830
:return: set of paths.
832
# NB: we specifically *don't* call self.has_filename, because for
833
# WorkingTrees that can indicate files that exist on disk but that
835
return set((p for p in paths if self.path2id(p) is None))
838
def iter_entries_by_dir(self, specific_file_ids=None, yield_parents=False):
839
"""Walk the tree in 'by_dir' order.
841
This will yield each entry in the tree as a (path, entry) tuple.
842
The order that they are yielded is:
844
See Tree.iter_entries_by_dir for details.
846
:param yield_parents: If True, yield the parents from the root leading
847
down to specific_file_ids that have been requested. This has no
848
impact if specific_file_ids is None.
850
if specific_file_ids is None:
851
inventory_file_ids = None
853
inventory_file_ids = []
854
for tree_file_id in specific_file_ids:
855
inventory, inv_file_id = self._unpack_file_id(tree_file_id)
856
if not inventory is self.root_inventory: # for now
857
raise AssertionError("%r != %r" % (
858
inventory, self.root_inventory))
859
inventory_file_ids.append(inv_file_id)
860
# FIXME: Handle nested trees
861
return self.root_inventory.iter_entries_by_dir(
862
specific_file_ids=inventory_file_ids, yield_parents=yield_parents)
865
def iter_child_entries(self, file_id, path=None):
866
inv, inv_file_id = self._unpack_file_id(file_id)
867
return inv[inv_file_id].children.itervalues()
869
def iter_children(self, file_id, path=None):
870
"""See Tree.iter_children."""
871
entry = self.iter_entries_by_dir([file_id]).next()[1]
872
for child in getattr(entry, 'children', {}).itervalues():
788
876
def find_ids_across_trees(filenames, trees, require_versioned=True):
1245
1335
if old_entry is None:
1246
1336
# Reusing a discarded change.
1247
1337
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)
1338
precise_file_ids.update(
1339
self.source.iter_children(file_id))
1250
1340
changed_file_ids.add(result[0])
1344
def file_content_matches(self, source_file_id, target_file_id,
1345
source_path=None, target_path=None, source_stat=None, target_stat=None):
1346
"""Check if two files are the same in the source and target trees.
1348
This only checks that the contents of the files are the same,
1349
it does not touch anything else.
1351
:param source_file_id: File id of the file in the source tree
1352
:param target_file_id: File id of the file in the target tree
1353
:param source_path: Path of the file in the source tree
1354
:param target_path: Path of the file in the target tree
1355
:param source_stat: Optional stat value of the file in the source tree
1356
:param target_stat: Optional stat value of the file in the target tree
1357
:return: Boolean indicating whether the files have the same contents
1359
source_verifier_kind, source_verifier_data = self.source.get_file_verifier(
1360
source_file_id, source_path, source_stat)
1361
target_verifier_kind, target_verifier_data = self.target.get_file_verifier(
1362
target_file_id, target_path, target_stat)
1363
if source_verifier_kind == target_verifier_kind:
1364
return (source_verifier_data == target_verifier_data)
1365
# Fall back to SHA1 for now
1366
if source_verifier_kind != "SHA1":
1367
source_sha1 = self.source.get_file_sha1(source_file_id,
1368
source_path, source_stat)
1370
source_sha1 = source_verifier_data
1371
if target_verifier_kind != "SHA1":
1372
target_sha1 = self.target.get_file_sha1(target_file_id,
1373
target_path, target_stat)
1375
target_sha1 = target_verifier_data
1376
return (source_sha1 == target_sha1)
1378
InterTree.register_optimiser(InterTree)
1254
1381
class MultiWalker(object):
1255
1382
"""Walk multiple trees simultaneously, getting combined results."""
added
removed
breezy/tree.py
