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 .inter import InterObject
54
class FileTimestampUnavailable(errors.BzrError):
56
_fmt = "The filestamp for %(path)s is not available."
60
def __init__(self, path):
64
class TreeEntry(object):
65
"""An entry that implements the minimum interface used by commands.
68
def __eq__(self, other):
69
# yes, this is ugly, TODO: best practice __eq__ style.
70
return (isinstance(other, TreeEntry)
71
and other.__class__ == self.__class__)
73
def kind_character(self):
77
class TreeDirectory(TreeEntry):
78
"""See TreeEntry. This is a directory in a working tree."""
80
def kind_character(self):
84
class TreeFile(TreeEntry):
85
"""See TreeEntry. This is a regular file in a working tree."""
87
def kind_character(self):
91
class TreeLink(TreeEntry):
92
"""See TreeEntry. This is a symlink in a working tree."""
94
def kind_character(self):
43
98
class Tree(object):
50
105
* `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
107
Trees can be compared, etc, regardless of whether they are working
60
108
trees or versioned trees.
111
def has_versioned_directories(self):
112
"""Whether this tree can contain explicitly versioned directories.
114
This defaults to True, but some implementations may want to override
63
119
def changes_from(self, other, want_unchanged=False, specific_files=None,
64
120
extra_trees=None, require_versioned=False, include_root=False,
65
121
want_unversioned=False):
147
return iter(self.inventory)
149
193
def all_file_ids(self):
150
194
"""Iterate through all file ids, including ids for missing files."""
151
return set(self.inventory)
195
raise NotImplementedError(self.all_file_ids)
197
def all_versioned_paths(self):
198
"""Iterate through all paths, including paths for missing files."""
199
raise NotImplementedError(self.all_versioned_paths)
153
201
def id2path(self, file_id):
154
202
"""Return the path for a file id.
156
204
: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)
206
raise NotImplementedError(self.id2path)
173
208
def iter_entries_by_dir(self, specific_file_ids=None, yield_parents=False):
174
209
"""Walk the tree in 'by_dir' order.
194
229
The yield order (ignoring root) would be::
195
231
a, f, a/b, a/d, a/b/c, a/d/e, f/g
197
233
:param yield_parents: If True, yield the parents from the root leading
198
234
down to specific_file_ids that have been requested. This has no
199
235
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)
237
raise NotImplementedError(self.iter_entries_by_dir)
239
def iter_child_entries(self, path, file_id=None):
240
"""Iterate over the children of a directory or tree reference.
242
:param path: Path of the directory
243
:param file_id: Optional file id of the directory/tree-reference
244
:raise NoSuchId: When the file_id does not exist
245
:return: Iterator over entries in the directory
247
raise NotImplementedError(self.iter_child_entries)
249
def list_files(self, include_root=False, from_dir=None, recursive=True):
250
"""List all files in this tree.
252
:param include_root: Whether to include the entry for the tree root
253
:param from_dir: Directory under which to list files
254
:param recursive: Whether to list files recursively
255
:return: iterator over tuples of (path, versioned, kind, file_id,
258
raise NotImplementedError(self.list_files)
204
260
def iter_references(self):
205
261
if self.supports_tree_reference():
207
263
if entry.kind == 'tree-reference':
208
264
yield path, entry.file_id
210
def kind(self, file_id):
266
def kind(self, path, file_id=None):
211
267
raise NotImplementedError("Tree subclass %s must implement kind"
212
268
% self.__class__.__name__)
214
def stored_kind(self, file_id):
270
def stored_kind(self, path, file_id=None):
215
271
"""File kind stored for this file_id.
217
273
May not match kind on disk for working trees. Always available
218
274
for versioned files, even when the file itself is missing.
220
return self.kind(file_id)
276
return self.kind(path, file_id)
222
278
def path_content_summary(self, path):
223
279
"""Get a summary of the information about path.
268
321
raise NotImplementedError(self.get_file)
270
def get_file_with_stat(self, file_id, path=None):
323
def get_file_with_stat(self, path, file_id=None):
271
324
"""Get a file handle and stat object for file_id.
273
326
The default implementation returns (self.get_file, None) for backwards
276
:param file_id: The file id to read.
277
:param path: The path of the file, if it is known.
329
:param path: The path of the file.
330
:param file_id: The file id to read, if it is known.
278
331
:return: A tuple (file_handle, stat_value_or_None). If the tree has
279
332
no stat facility, or need for a stat cache feedback during commit,
280
333
it may return None for the second element of the tuple.
282
return (self.get_file(file_id, path), None)
335
return (self.get_file(path, file_id), None)
284
def get_file_text(self, file_id, path=None):
337
def get_file_text(self, path, file_id=None):
285
338
"""Return the byte content of a file.
340
:param path: The path of the file.
287
341
:param file_id: The file_id of the file.
288
:param path: The path of the file.
289
343
If both file_id and path are supplied, an implementation may use
346
:returns: A single byte string for the whole file.
292
my_file = self.get_file(file_id, path)
348
my_file = self.get_file(path, file_id)
294
350
return my_file.read()
298
def get_file_lines(self, file_id, path=None):
354
def get_file_lines(self, path, file_id=None):
299
355
"""Return the content of a file, as lines.
357
:param path: The path of the file.
301
358
:param file_id: The file_id of the file.
302
:param path: The path of the file.
303
360
If both file_id and path are supplied, an implementation may use
306
return osutils.split_lines(self.get_file_text(file_id, path))
308
def get_file_mtime(self, file_id, path=None):
363
return osutils.split_lines(self.get_file_text(path, file_id))
365
def get_file_verifier(self, path, file_id=None, stat_value=None):
366
"""Return a verifier for a file.
368
The default implementation returns a sha1.
370
:param file_id: The handle for this file.
371
:param path: The path that this file can be found at.
372
These must point to the same object.
373
:param stat_value: Optional stat value for the object
374
:return: Tuple with verifier name and verifier data
376
return ("SHA1", self.get_file_sha1(path, file_id,
377
stat_value=stat_value))
379
def get_file_sha1(self, path, file_id=None, stat_value=None):
380
"""Return the SHA1 file for a file.
382
:note: callers should use get_file_verifier instead
383
where possible, as the underlying repository implementation may
384
have quicker access to a non-sha1 verifier.
386
:param path: The path that this file can be found at.
387
:param file_id: The handle for this file.
388
These must point to the same object.
389
:param stat_value: Optional stat value for the object
391
raise NotImplementedError(self.get_file_sha1)
393
def get_file_mtime(self, path, file_id=None):
309
394
"""Return the modification time for a file.
396
:param path: The path that this file can be found at.
311
397
:param file_id: The handle for this file.
312
:param path: The path that this file can be found at.
313
398
These must point to the same object.
315
400
raise NotImplementedError(self.get_file_mtime)
317
def get_file_size(self, file_id):
402
def get_file_size(self, path, file_id=None):
318
403
"""Return the size of a file in bytes.
320
405
This applies only to regular files. If invoked on directories or
349
440
# of bytestrings. (Technically, a bytestring is also an iterable
350
441
# of bytestrings, but iterating through each character is not
352
cur_file = (self.get_file_text(file_id),)
443
# TODO(jelmer): Pass paths into iter_files_bytes
444
path = self.id2path(file_id)
445
cur_file = (self.get_file_text(path, file_id),)
353
446
yield identifier, cur_file
355
def get_symlink_target(self, file_id):
448
def get_symlink_target(self, path, file_id=None):
356
449
"""Get the target for a given file_id.
358
451
It is assumed that the caller already knows that file_id is referencing
360
453
:param file_id: Handle for the symlink entry.
454
:param path: The path of the file.
455
If both file_id and path are supplied, an implementation may use
361
457
:return: The path the symlink points to.
363
459
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
461
def get_root_id(self):
444
462
"""Return the file_id for the root of this tree."""
445
463
raise NotImplementedError(self.get_root_id)
447
def annotate_iter(self, file_id,
465
def annotate_iter(self, path, file_id=None,
448
466
default_revision=_mod_revision.CURRENT_REVISION):
449
467
"""Return an iterator of revision_id, line tuples.
459
477
raise NotImplementedError(self.annotate_iter)
461
479
def _get_plan_merge_data(self, file_id, other, base):
462
from bzrlib import versionedfile
480
from .bzr import versionedfile
463
481
vf = versionedfile._PlanMergeVersionedFile(file_id)
464
last_revision_a = self._get_file_revision(file_id, vf, 'this:')
465
last_revision_b = other._get_file_revision(file_id, vf, 'other:')
482
last_revision_a = self._get_file_revision(
483
self.id2path(file_id), file_id, vf, 'this:')
484
last_revision_b = other._get_file_revision(
485
other.id2path(file_id), file_id, vf, 'other:')
467
487
last_revision_base = None
469
last_revision_base = base._get_file_revision(file_id, vf, 'base:')
489
last_revision_base = base._get_file_revision(
490
base.id2path(file_id), file_id, vf, 'base:')
470
491
return vf, last_revision_a, last_revision_b, last_revision_base
472
493
def plan_file_merge(self, file_id, other, base=None):
503
524
except errors.NoSuchRevisionInTree:
504
525
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
def _get_file_revision(self, file_id, vf, tree_revision):
527
def _get_file_revision(self, path, file_id, vf, tree_revision):
516
528
"""Ensure that file_id, tree_revision is in vf to plan the merge."""
518
529
if getattr(self, '_repository', None) is None:
519
530
last_revision = tree_revision
520
parent_keys = [(file_id, self._file_revision(t, file_id)) for t in
531
parent_keys = [(file_id, t.get_file_revision(path, file_id)) for t in
521
532
self._iter_parent_trees()]
522
533
vf.add_lines((file_id, last_revision), parent_keys,
523
self.get_file(file_id).readlines())
534
self.get_file_lines(path, file_id))
524
535
repo = self.branch.repository
525
536
base_vf = repo.texts
527
last_revision = self._file_revision(self, file_id)
538
last_revision = self.get_file_revision(path, file_id)
528
539
base_vf = self._repository.texts
529
540
if base_vf not in vf.fallback_versionedfiles:
530
541
vf.fallback_versionedfiles.append(base_vf)
531
542
return last_revision
533
inventory = property(_get_inventory,
534
doc="Inventory of this Tree")
536
544
def _check_retrieved(self, ie, f):
537
545
if not __debug__:
539
fp = fingerprint_file(f)
547
fp = osutils.fingerprint_file(f)
542
550
if ie.text_size is not None:
543
551
if ie.text_size != fp['size']:
544
raise BzrError("mismatched size for file %r in %r" % (ie.file_id, self._store),
552
raise errors.BzrError(
553
"mismatched size for file %r in %r" %
554
(ie.file_id, self._store),
545
555
["inventory expects %d bytes" % ie.text_size,
546
556
"file is actually %d bytes" % fp['size'],
547
557
"store is probably damaged/corrupt"])
549
559
if ie.text_sha1 != fp['sha1']:
550
raise BzrError("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
560
raise errors.BzrError("wrong SHA-1 for file %r in %r" %
561
(ie.file_id, self._store),
551
562
["inventory expects %s" % ie.text_sha1,
552
563
"file is actually %s" % fp['sha1'],
553
564
"store is probably damaged/corrupt"])
556
566
def path2id(self, path):
557
567
"""Return the id for path in this tree."""
558
return self._inventory.path2id(path)
568
raise NotImplementedError(self.path2id)
570
def is_versioned(self, path):
571
"""Check whether path is versioned.
573
:param path: Path to check
576
return self.path2id(path) is not None
560
578
def paths2ids(self, paths, trees=[], require_versioned=True):
561
579
"""Return all the ids that can be reached by walking from paths.
577
595
return find_ids_across_trees(paths, [self] + list(trees), require_versioned)
579
597
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():
598
"""Iterate over the file ids of the children of an entry.
600
:param file_id: File id of the entry
601
:return: Iterator over child file ids.
603
raise NotImplementedError(self.iter_children)
584
605
def lock_read(self):
606
"""Lock this tree for multiple read only operations.
608
:return: A breezy.lock.LogicalLockResult.
610
return lock.LogicalLockResult(self.unlock)
587
612
def revision_tree(self, revision_id):
588
613
"""Obtain a revision tree for the revision revision_id.
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
745
def find_ids_across_trees(filenames, trees, require_versioned=True):
789
746
"""Find the ids corresponding to specified filenames.
930
893
if source_kind != target_kind:
931
894
changed_content = True
932
895
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)):
896
if not self.file_content_matches(file_id, file_id, source_path,
897
target_path, source_stat, target_stat):
935
898
changed_content = True
936
899
elif source_kind == 'symlink':
937
if (self.source.get_symlink_target(file_id) !=
938
self.target.get_symlink_target(file_id)):
900
if (self.source.get_symlink_target(source_path, file_id) !=
901
self.target.get_symlink_target(target_path, file_id)):
939
902
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)):
903
elif source_kind == 'tree-reference':
904
if (self.source.get_reference_revision(source_path, file_id)
905
!= self.target.get_reference_revision(target_path, file_id)):
947
906
changed_content = True
948
907
parent = (source_parent, target_parent)
949
908
name = (source_name, target_name)
980
938
trees = (self.source,)
981
939
if extra_trees is not None:
982
940
trees = trees + tuple(extra_trees)
983
# target is usually the newer tree:
984
specific_file_ids = self.target.paths2ids(specific_files, trees,
985
require_versioned=require_versioned)
986
if specific_files and not specific_file_ids:
987
# All files are unversioned, so just return an empty delta
988
# _compare_trees would think we want a complete delta
989
result = delta.TreeDelta()
990
fake_entry = InventoryFile('unused', 'unused', 'unused')
991
result.unversioned = [(path, None,
992
self.target._comparison_data(fake_entry, path)[0]) for path in
995
return delta._compare_trees(self.source, self.target, want_unchanged,
996
specific_files, include_root, extra_trees=extra_trees,
997
require_versioned=require_versioned,
998
want_unversioned=want_unversioned)
941
with self.lock_read():
942
# target is usually the newer tree:
943
specific_file_ids = self.target.paths2ids(specific_files, trees,
944
require_versioned=require_versioned)
945
if specific_files and not specific_file_ids:
946
# All files are unversioned, so just return an empty delta
947
# _compare_trees would think we want a complete delta
948
result = delta.TreeDelta()
949
fake_entry = inventory.InventoryFile('unused', 'unused', 'unused')
950
result.unversioned = [(path, None,
951
self.target._comparison_data(fake_entry, path)[0]) for path in
954
return delta._compare_trees(self.source, self.target, want_unchanged,
955
specific_files, include_root, extra_trees=extra_trees,
956
require_versioned=require_versioned,
957
want_unversioned=want_unversioned)
1000
959
def iter_changes(self, include_unchanged=False,
1001
960
specific_files=None, pb=None, extra_trees=[],
1239
1198
precise_file_ids.add(new_parent_id)
1241
1200
if (result[6][0] == 'directory' and
1242
result[6][1] != 'directory'):
1201
result[6][1] != 'directory'):
1243
1202
# This stopped being a directory, the old children have
1244
1203
# to be included.
1245
1204
if old_entry is None:
1246
1205
# Reusing a discarded change.
1247
1206
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)
1207
precise_file_ids.update(
1208
self.source.iter_children(file_id))
1250
1209
changed_file_ids.add(result[0])
1212
def file_content_matches(
1213
self, source_file_id, target_file_id, source_path=None,
1214
target_path=None, source_stat=None, target_stat=None):
1215
"""Check if two files are the same in the source and target trees.
1217
This only checks that the contents of the files are the same,
1218
it does not touch anything else.
1220
:param source_file_id: File id of the file in the source tree
1221
:param target_file_id: File id of the file in the target tree
1222
:param source_path: Path of the file in the source tree
1223
:param target_path: Path of the file in the target tree
1224
:param source_stat: Optional stat value of the file in the source tree
1225
:param target_stat: Optional stat value of the file in the target tree
1226
:return: Boolean indicating whether the files have the same contents
1228
with self.lock_read():
1229
if source_path is None:
1230
source_path = self.source.id2path(source_file_id)
1231
if target_path is None:
1232
target_path = self.target.id2path(target_file_id)
1233
source_verifier_kind, source_verifier_data = (
1234
self.source.get_file_verifier(
1235
source_path, source_file_id, source_stat))
1236
target_verifier_kind, target_verifier_data = (
1237
self.target.get_file_verifier(
1238
target_path, target_file_id, target_stat))
1239
if source_verifier_kind == target_verifier_kind:
1240
return (source_verifier_data == target_verifier_data)
1241
# Fall back to SHA1 for now
1242
if source_verifier_kind != "SHA1":
1243
source_sha1 = self.source.get_file_sha1(
1244
source_path, source_file_id, source_stat)
1246
source_sha1 = source_verifier_data
1247
if target_verifier_kind != "SHA1":
1248
target_sha1 = self.target.get_file_sha1(
1249
target_path, target_file_id, target_stat)
1251
target_sha1 = target_verifier_data
1252
return (source_sha1 == target_sha1)
1254
InterTree.register_optimiser(InterTree)
1254
1257
class MultiWalker(object):
1255
1258
"""Walk multiple trees simultaneously, getting combined results."""
added
removed
breezy/tree.py
