17
17
"""Tree classes, representing directory at point in time.
20
from __future__ import absolute_import
21
from collections import deque
24
from .lazy_import import lazy_import
25
lazy_import(globals(), """
25
29
conflicts as _mod_conflicts,
30
33
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
37
from breezy.bzr import (
40
from breezy.i18n import gettext
48
from .decorators import needs_read_lock
49
from .inter import InterObject
55
class FileTimestampUnavailable(errors.BzrError):
57
_fmt = "The filestamp for %(path)s is not available."
61
def __init__(self, path):
43
65
class Tree(object):
50
72
* `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
74
Trees can be compared, etc, regardless of whether they are working
60
75
trees or versioned trees.
78
def has_versioned_directories(self):
79
"""Whether this tree can contain explicitly versioned directories.
81
This defaults to True, but some implementations may want to override
63
86
def changes_from(self, other, want_unchanged=False, specific_files=None,
64
87
extra_trees=None, require_versioned=False, include_root=False,
65
88
want_unversioned=False):
127
150
raise NotImplementedError(self.has_filename)
129
152
def has_id(self, file_id):
130
return self.inventory.has_id(file_id)
132
def __contains__(self, file_id):
133
return self.has_id(file_id)
153
raise NotImplementedError(self.has_id)
135
155
def has_or_had_id(self, file_id):
136
return self.inventory.has_id(file_id)
156
raise NotImplementedError(self.has_or_had_id)
138
158
def is_ignored(self, filename):
139
159
"""Check whether the filename is ignored by this tree.
147
return iter(self.inventory)
149
166
def all_file_ids(self):
150
167
"""Iterate through all file ids, including ids for missing files."""
151
return set(self.inventory)
168
raise NotImplementedError(self.all_file_ids)
153
170
def id2path(self, file_id):
154
171
"""Return the path for a file id.
156
173
: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)
175
raise NotImplementedError(self.id2path)
173
177
def iter_entries_by_dir(self, specific_file_ids=None, yield_parents=False):
174
178
"""Walk the tree in 'by_dir' order.
194
198
The yield order (ignoring root) would be::
195
200
a, f, a/b, a/d, a/b/c, a/d/e, f/g
197
202
:param yield_parents: If True, yield the parents from the root leading
198
203
down to specific_file_ids that have been requested. This has no
199
204
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)
206
raise NotImplementedError(self.iter_entries_by_dir)
208
def iter_child_entries(self, file_id, path=None):
209
"""Iterate over the children of a directory or tree reference.
211
:param file_id: File id of the directory/tree-reference
212
:param path: Optional path of the directory
213
:raise NoSuchId: When the file_id does not exist
214
:return: Iterator over entries in the directory
216
raise NotImplementedError(self.iter_child_entries)
218
def list_files(self, include_root=False, from_dir=None, recursive=True):
219
"""List all files in this tree.
221
:param include_root: Whether to include the entry for the tree root
222
:param from_dir: Directory under which to list files
223
:param recursive: Whether to list files recursively
224
:return: iterator over tuples of (path, versioned, kind, file_id,
227
raise NotImplementedError(self.list_files)
204
229
def iter_references(self):
205
230
if self.supports_tree_reference():
301
326
:param file_id: The file_id of the file.
302
327
:param path: The path of the file.
303
329
If both file_id and path are supplied, an implementation may use
306
332
return osutils.split_lines(self.get_file_text(file_id, path))
334
def get_file_verifier(self, file_id, path=None, stat_value=None):
335
"""Return a verifier for a file.
337
The default implementation returns a sha1.
339
:param file_id: The handle for this file.
340
:param path: The path that this file can be found at.
341
These must point to the same object.
342
:param stat_value: Optional stat value for the object
343
:return: Tuple with verifier name and verifier data
345
return ("SHA1", self.get_file_sha1(file_id, path=path,
346
stat_value=stat_value))
348
def get_file_sha1(self, file_id, path=None, stat_value=None):
349
"""Return the SHA1 file for a file.
351
:note: callers should use get_file_verifier instead
352
where possible, as the underlying repository implementation may
353
have quicker access to a non-sha1 verifier.
355
:param file_id: The handle for this file.
356
:param path: The path that this file can be found at.
357
These must point to the same object.
358
:param stat_value: Optional stat value for the object
360
raise NotImplementedError(self.get_file_sha1)
308
362
def get_file_mtime(self, file_id, path=None):
309
363
"""Return the modification time for a file.
324
378
raise NotImplementedError(self.get_file_size)
326
def get_file_by_path(self, path):
327
return self.get_file(self._inventory.path2id(path), path)
380
def is_executable(self, file_id, path=None):
381
"""Check if a file is executable.
383
:param file_id: The handle for this file.
384
:param path: The path that this file can be found at.
385
These must point to the same object.
387
raise NotImplementedError(self.is_executable)
329
389
def iter_files_bytes(self, desired_files):
330
390
"""Iterate through file contents.
352
412
cur_file = (self.get_file_text(file_id),)
353
413
yield identifier, cur_file
355
def get_symlink_target(self, file_id):
415
def get_symlink_target(self, file_id, path=None):
356
416
"""Get the target for a given file_id.
358
418
It is assumed that the caller already knows that file_id is referencing
360
420
:param file_id: Handle for the symlink entry.
421
:param path: The path of the file.
422
If both file_id and path are supplied, an implementation may use
361
424
:return: The path the symlink points to.
363
426
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
428
def get_root_id(self):
444
429
"""Return the file_id for the root of this tree."""
445
430
raise NotImplementedError(self.get_root_id)
503
488
except errors.NoSuchRevisionInTree:
504
489
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
491
def _get_file_revision(self, file_id, vf, tree_revision):
516
492
"""Ensure that file_id, tree_revision is in vf to plan the merge."""
518
494
if getattr(self, '_repository', None) is None:
519
495
last_revision = tree_revision
520
parent_keys = [(file_id, self._file_revision(t, file_id)) for t in
496
parent_keys = [(file_id, t.get_file_revision(file_id)) for t in
521
497
self._iter_parent_trees()]
522
498
vf.add_lines((file_id, last_revision), parent_keys,
523
self.get_file(file_id).readlines())
499
self.get_file_lines(file_id))
524
500
repo = self.branch.repository
525
501
base_vf = repo.texts
527
last_revision = self._file_revision(self, file_id)
503
last_revision = self.get_file_revision(file_id)
528
504
base_vf = self._repository.texts
529
505
if base_vf not in vf.fallback_versionedfiles:
530
506
vf.fallback_versionedfiles.append(base_vf)
531
507
return last_revision
533
inventory = property(_get_inventory,
534
doc="Inventory of this Tree")
536
509
def _check_retrieved(self, ie, f):
537
510
if not __debug__:
539
fp = fingerprint_file(f)
512
fp = osutils.fingerprint_file(f)
542
515
if ie.text_size is not None:
543
516
if ie.text_size != fp['size']:
544
raise BzrError("mismatched size for file %r in %r" % (ie.file_id, self._store),
517
raise errors.BzrError(
518
"mismatched size for file %r in %r" %
519
(ie.file_id, self._store),
545
520
["inventory expects %d bytes" % ie.text_size,
546
521
"file is actually %d bytes" % fp['size'],
547
522
"store is probably damaged/corrupt"])
549
524
if ie.text_sha1 != fp['sha1']:
550
raise BzrError("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
525
raise errors.BzrError("wrong SHA-1 for file %r in %r" %
526
(ie.file_id, self._store),
551
527
["inventory expects %s" % ie.text_sha1,
552
528
"file is actually %s" % fp['sha1'],
553
529
"store is probably damaged/corrupt"])
556
531
def path2id(self, path):
557
532
"""Return the id for path in this tree."""
558
return self._inventory.path2id(path)
533
raise NotImplementedError(self.path2id)
560
535
def paths2ids(self, paths, trees=[], require_versioned=True):
561
536
"""Return all the ids that can be reached by walking from paths.
577
552
return find_ids_across_trees(paths, [self] + list(trees), require_versioned)
579
554
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():
555
"""Iterate over the file ids of the children of an entry.
557
:param file_id: File id of the entry
558
:return: Iterator over child file ids.
560
raise NotImplementedError(self.iter_children)
584
562
def lock_read(self):
563
"""Lock this tree for multiple read only operations.
565
:return: A breezy.lock.LogicalLockResult.
567
return lock.LogicalLockResult(self.unlock)
587
569
def revision_tree(self, revision_id):
588
570
"""Obtain a revision tree for the revision revision_id.
675
653
path = self.id2path(file_id)
676
prefs = self.iter_search_rules([path], filter_pref_names).next()
654
prefs = next(self.iter_search_rules([path], filter_pref_names))
677
655
stk = filters._get_filter_stack_for(prefs)
678
656
if 'filters' in debug.debug_flags:
679
note("*** %s content-filter: %s => %r" % (path,prefs,stk))
657
trace.note(gettext("*** {0} content-filter: {1} => {2!r}").format(path,prefs,stk))
682
660
def _content_filter_stack_provider(self):
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
702
def find_ids_across_trees(filenames, trees, require_versioned=True):
789
703
"""Find the ids corresponding to specified filenames.
930
850
if source_kind != target_kind:
931
851
changed_content = True
932
852
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)):
853
if not self.file_content_matches(file_id, file_id, source_path,
854
target_path, source_stat, target_stat):
935
855
changed_content = True
936
856
elif source_kind == 'symlink':
937
857
if (self.source.get_symlink_target(file_id) !=
938
858
self.target.get_symlink_target(file_id)):
939
859
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)):
860
elif source_kind == 'tree-reference':
861
if (self.source.get_reference_revision(file_id, source_path)
862
!= self.target.get_reference_revision(file_id, target_path)):
947
863
changed_content = True
948
864
parent = (source_parent, target_parent)
949
865
name = (source_name, target_name)
987
903
# All files are unversioned, so just return an empty delta
988
904
# _compare_trees would think we want a complete delta
989
905
result = delta.TreeDelta()
990
fake_entry = InventoryFile('unused', 'unused', 'unused')
906
fake_entry = inventory.InventoryFile('unused', 'unused', 'unused')
991
907
result.unversioned = [(path, None,
992
908
self.target._comparison_data(fake_entry, path)[0]) for path in
1072
988
# the unversioned path lookup only occurs on real trees - where there
1073
989
# can be extras. So the fake_entry is solely used to look up
1074
990
# executable it values when execute is not supported.
1075
fake_entry = InventoryFile('unused', 'unused', 'unused')
991
fake_entry = inventory.InventoryFile('unused', 'unused', 'unused')
1076
992
for target_path, target_entry in to_entries_by_dir:
1077
993
while (all_unversioned and
1078
994
all_unversioned[0][0] < target_path.split('/')):
1245
1161
if old_entry is None:
1246
1162
# Reusing a discarded change.
1247
1163
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)
1164
precise_file_ids.update(
1165
self.source.iter_children(file_id))
1250
1166
changed_file_ids.add(result[0])
1170
def file_content_matches(self, source_file_id, target_file_id,
1171
source_path=None, target_path=None, source_stat=None, target_stat=None):
1172
"""Check if two files are the same in the source and target trees.
1174
This only checks that the contents of the files are the same,
1175
it does not touch anything else.
1177
:param source_file_id: File id of the file in the source tree
1178
:param target_file_id: File id of the file in the target tree
1179
:param source_path: Path of the file in the source tree
1180
:param target_path: Path of the file in the target tree
1181
:param source_stat: Optional stat value of the file in the source tree
1182
:param target_stat: Optional stat value of the file in the target tree
1183
:return: Boolean indicating whether the files have the same contents
1185
source_verifier_kind, source_verifier_data = self.source.get_file_verifier(
1186
source_file_id, source_path, source_stat)
1187
target_verifier_kind, target_verifier_data = self.target.get_file_verifier(
1188
target_file_id, target_path, target_stat)
1189
if source_verifier_kind == target_verifier_kind:
1190
return (source_verifier_data == target_verifier_data)
1191
# Fall back to SHA1 for now
1192
if source_verifier_kind != "SHA1":
1193
source_sha1 = self.source.get_file_sha1(source_file_id,
1194
source_path, source_stat)
1196
source_sha1 = source_verifier_data
1197
if target_verifier_kind != "SHA1":
1198
target_sha1 = self.target.get_file_sha1(target_file_id,
1199
target_path, target_stat)
1201
target_sha1 = target_verifier_data
1202
return (source_sha1 == target_sha1)
1204
InterTree.register_optimiser(InterTree)
1254
1207
class MultiWalker(object):
1255
1208
"""Walk multiple trees simultaneously, getting combined results."""
added
removed
breezy/tree.py
