301
295
:param file_id: The file_id of the file.
302
296
:param path: The path of the file.
303
298
If both file_id and path are supplied, an implementation may use
306
301
return osutils.split_lines(self.get_file_text(file_id, path))
303
def get_file_verifier(self, file_id, path=None, stat_value=None):
304
"""Return a verifier for a file.
306
The default implementation returns a sha1.
308
:param file_id: The handle for this file.
309
:param path: The path that this file can be found at.
310
These must point to the same object.
311
:param stat_value: Optional stat value for the object
312
:return: Tuple with verifier name and verifier data
314
return ("SHA1", self.get_file_sha1(file_id, path=path,
315
stat_value=stat_value))
317
def get_file_sha1(self, file_id, path=None, stat_value=None):
318
"""Return the SHA1 file for a file.
320
:note: callers should use get_file_verifier instead
321
where possible, as the underlying repository implementation may
322
have quicker access to a non-sha1 verifier.
324
:param file_id: The handle for this file.
325
:param path: The path that this file can be found at.
326
These must point to the same object.
327
:param stat_value: Optional stat value for the object
329
raise NotImplementedError(self.get_file_sha1)
308
331
def get_file_mtime(self, file_id, path=None):
309
332
"""Return the modification time for a file.
352
384
cur_file = (self.get_file_text(file_id),)
353
385
yield identifier, cur_file
355
def get_symlink_target(self, file_id):
387
def get_symlink_target(self, file_id, path=None):
356
388
"""Get the target for a given file_id.
358
390
It is assumed that the caller already knows that file_id is referencing
360
392
:param file_id: Handle for the symlink entry.
393
:param path: The path of the file.
394
If both file_id and path are supplied, an implementation may use
361
396
:return: The path the symlink points to.
363
398
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
400
def get_root_id(self):
444
401
"""Return the file_id for the root of this tree."""
445
402
raise NotImplementedError(self.get_root_id)
503
460
except errors.NoSuchRevisionInTree:
504
461
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
463
def _get_file_revision(self, file_id, vf, tree_revision):
516
464
"""Ensure that file_id, tree_revision is in vf to plan the merge."""
518
466
if getattr(self, '_repository', None) is None:
519
467
last_revision = tree_revision
520
parent_keys = [(file_id, self._file_revision(t, file_id)) for t in
468
parent_keys = [(file_id, t.get_file_revision(file_id)) for t in
521
469
self._iter_parent_trees()]
522
470
vf.add_lines((file_id, last_revision), parent_keys,
523
self.get_file(file_id).readlines())
471
self.get_file_lines(file_id))
524
472
repo = self.branch.repository
525
473
base_vf = repo.texts
527
last_revision = self._file_revision(self, file_id)
475
last_revision = self.get_file_revision(file_id)
528
476
base_vf = self._repository.texts
529
477
if base_vf not in vf.fallback_versionedfiles:
530
478
vf.fallback_versionedfiles.append(base_vf)
531
479
return last_revision
533
inventory = property(_get_inventory,
534
doc="Inventory of this Tree")
536
481
def _check_retrieved(self, ie, f):
537
482
if not __debug__:
539
fp = fingerprint_file(f)
484
fp = osutils.fingerprint_file(f)
542
487
if ie.text_size is not None:
543
488
if ie.text_size != fp['size']:
544
raise BzrError("mismatched size for file %r in %r" % (ie.file_id, self._store),
489
raise errors.BzrError(
490
"mismatched size for file %r in %r" %
491
(ie.file_id, self._store),
545
492
["inventory expects %d bytes" % ie.text_size,
546
493
"file is actually %d bytes" % fp['size'],
547
494
"store is probably damaged/corrupt"])
549
496
if ie.text_sha1 != fp['sha1']:
550
raise BzrError("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
497
raise errors.BzrError("wrong SHA-1 for file %r in %r" %
498
(ie.file_id, self._store),
551
499
["inventory expects %s" % ie.text_sha1,
552
500
"file is actually %s" % fp['sha1'],
553
501
"store is probably damaged/corrupt"])
556
503
def path2id(self, path):
557
504
"""Return the id for path in this tree."""
558
return self._inventory.path2id(path)
505
raise NotImplementedError(self.path2id)
560
507
def paths2ids(self, paths, trees=[], require_versioned=True):
561
508
"""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.
671
class InventoryTree(Tree):
672
"""A tree that relies on an inventory for its metadata.
674
Trees contain an `Inventory` object, and also know how to retrieve
675
file texts mentioned in the inventory, either from a working
676
directory or from a store.
678
It is possible for trees to contain files that are not described
679
in their inventory or vice versa; for this use `filenames()`.
681
Subclasses should set the _inventory attribute, which is considered
682
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)
685
def get_canonical_inventory_paths(self, paths):
686
"""Like get_canonical_inventory_path() but works on multiple items.
688
:param paths: A sequence of paths relative to the root of the tree.
689
:return: A list of paths, with each item the corresponding input path
690
adjusted to account for existing elements that match case
693
return list(self._yield_canonical_inventory_paths(paths))
695
def get_canonical_inventory_path(self, path):
696
"""Returns the first inventory item that case-insensitively matches path.
698
If a path matches exactly, it is returned. If no path matches exactly
699
but more than one path matches case-insensitively, it is implementation
700
defined which is returned.
702
If no path matches case-insensitively, the input path is returned, but
703
with as many path entries that do exist changed to their canonical
706
If you need to resolve many names from the same tree, you should
707
use get_canonical_inventory_paths() to avoid O(N) behaviour.
709
:param path: A paths relative to the root of the tree.
710
:return: The input path adjusted to account for existing elements
711
that match case insensitively.
713
return self._yield_canonical_inventory_paths([path]).next()
715
def _yield_canonical_inventory_paths(self, paths):
717
# First, if the path as specified exists exactly, just use it.
718
if self.path2id(path) is not None:
722
cur_id = self.get_root_id()
724
bit_iter = iter(path.split("/"))
728
for child in self.iter_children(cur_id):
730
# XXX: it seem like if the child is known to be in the
731
# tree, we shouldn't need to go from its id back to
732
# its path -- mbp 2010-02-11
734
# XXX: it seems like we could be more efficient
735
# by just directly looking up the original name and
736
# only then searching all children; also by not
737
# chopping paths so much. -- mbp 2010-02-11
738
child_base = os.path.basename(self.id2path(child))
739
if (child_base == elt):
740
# if we found an exact match, we can stop now; if
741
# we found an approximate match we need to keep
742
# searching because there might be an exact match
745
new_path = osutils.pathjoin(cur_path, child_base)
747
elif child_base.lower() == lelt:
749
new_path = osutils.pathjoin(cur_path, child_base)
750
except errors.NoSuchId:
751
# before a change is committed we can see this error...
756
# got to the end of this directory and no entries matched.
757
# Return what matched so far, plus the rest as specified.
758
cur_path = osutils.pathjoin(cur_path, elt, *list(bit_iter))
763
def _get_inventory(self):
764
return self._inventory
766
inventory = property(_get_inventory,
767
doc="Inventory of this Tree")
770
def path2id(self, path):
771
"""Return the id for path in this tree."""
772
return self._inventory.path2id(path)
774
def id2path(self, file_id):
775
"""Return the path for a file id.
779
return self.inventory.id2path(file_id)
781
def has_id(self, file_id):
782
return self.inventory.has_id(file_id)
784
def has_or_had_id(self, file_id):
785
return self.inventory.has_id(file_id)
787
def all_file_ids(self):
788
return set(self.inventory)
790
@deprecated_method(deprecated_in((2, 4, 0)))
792
return iter(self.inventory)
794
def filter_unversioned_files(self, paths):
795
"""Filter out paths that are versioned.
797
:return: set of paths.
799
# NB: we specifically *don't* call self.has_filename, because for
800
# WorkingTrees that can indicate files that exist on disk but that
802
pred = self.inventory.has_filename
803
return set((p for p in paths if not pred(p)))
806
def iter_entries_by_dir(self, specific_file_ids=None, yield_parents=False):
807
"""Walk the tree in 'by_dir' order.
809
This will yield each entry in the tree as a (path, entry) tuple.
810
The order that they are yielded is:
812
See Tree.iter_entries_by_dir for details.
814
:param yield_parents: If True, yield the parents from the root leading
815
down to specific_file_ids that have been requested. This has no
816
impact if specific_file_ids is None.
818
return self.inventory.iter_entries_by_dir(
819
specific_file_ids=specific_file_ids, yield_parents=yield_parents)
821
def get_file_by_path(self, path):
822
return self.get_file(self._inventory.path2id(path), path)
788
825
def find_ids_across_trees(filenames, trees, require_versioned=True):
1250
1293
changed_file_ids.add(result[0])
1297
def file_content_matches(self, source_file_id, target_file_id,
1298
source_path=None, target_path=None, source_stat=None, target_stat=None):
1299
"""Check if two files are the same in the source and target trees.
1301
This only checks that the contents of the files are the same,
1302
it does not touch anything else.
1304
:param source_file_id: File id of the file in the source tree
1305
:param target_file_id: File id of the file in the target tree
1306
:param source_path: Path of the file in the source tree
1307
:param target_path: Path of the file in the target tree
1308
:param source_stat: Optional stat value of the file in the source tree
1309
:param target_stat: Optional stat value of the file in the target tree
1310
:return: Boolean indicating whether the files have the same contents
1312
source_verifier_kind, source_verifier_data = self.source.get_file_verifier(
1313
source_file_id, source_path, source_stat)
1314
target_verifier_kind, target_verifier_data = self.target.get_file_verifier(
1315
target_file_id, target_path, target_stat)
1316
if source_verifier_kind == target_verifier_kind:
1317
return (source_verifier_data == target_verifier_data)
1318
# Fall back to SHA1 for now
1319
if source_verifier_kind != "SHA1":
1320
source_sha1 = self.source.get_file_sha1(source_file_id,
1321
source_path, source_stat)
1323
source_sha1 = source_verifier_data
1324
if target_verifier_kind != "SHA1":
1325
target_sha1 = self.target.get_file_sha1(target_file_id,
1326
target_path, target_stat)
1328
target_sha1 = target_verifier_data
1329
return (source_sha1 == target_sha1)
1331
InterTree.register_optimiser(InterTree)
1254
1334
class MultiWalker(object):
1255
1335
"""Walk multiple trees simultaneously, getting combined results."""
added
removed
bzrlib/tree.py
