1
# Copyright (C) 2005 Canonical Ltd
3
# This program is free software; you can redistribute it and/or modify
4
# it under the terms of the GNU General Public License as published by
5
# the Free Software Foundation; either version 2 of the License, or
6
# (at your option) any later version.
8
# This program is distributed in the hope that it will be useful,
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
# GNU General Public License for more details.
13
# You should have received a copy of the GNU General Public License
14
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17
"""Tree classes, representing directory at point in time.
21
from cStringIO import StringIO
29
from bzrlib.decorators import needs_read_lock
30
from bzrlib.errors import BzrError, BzrCheckError
31
from bzrlib import errors
32
from bzrlib.inventory import Inventory
33
from bzrlib.inter import InterObject
34
from bzrlib.osutils import fingerprint_file
35
import bzrlib.revision
36
from bzrlib.trace import mutter, note
40
"""Abstract file tree.
42
There are several subclasses:
44
* `WorkingTree` exists as files on disk editable by the user.
46
* `RevisionTree` is a tree as recorded at some point in the past.
48
Trees contain an `Inventory` object, and also know how to retrieve
49
file texts mentioned in the inventory, either from a working
50
directory or from a store.
52
It is possible for trees to contain files that are not described
53
in their inventory or vice versa; for this use `filenames()`.
55
Trees can be compared, etc, regardless of whether they are working
56
trees or versioned trees.
59
def changes_from(self, other, want_unchanged=False, specific_files=None,
60
extra_trees=None, require_versioned=False, include_root=False):
61
"""Return a TreeDelta of the changes from other to this tree.
63
:param other: A tree to compare with.
64
:param specific_files: An optional list of file paths to restrict the
65
comparison to. When mapping filenames to ids, all matches in all
66
trees (including optional extra_trees) are used, and all children of
67
matched directories are included.
68
:param want_unchanged: An optional boolean requesting the inclusion of
69
unchanged entries in the result.
70
:param extra_trees: An optional list of additional trees to use when
71
mapping the contents of specific_files (paths) to file_ids.
72
:param require_versioned: An optional boolean (defaults to False). When
73
supplied and True all the 'specific_files' must be versioned, or
74
a PathsNotVersionedError will be thrown.
76
The comparison will be performed by an InterTree object looked up on
79
# Martin observes that Tree.changes_from returns a TreeDelta and this
80
# may confuse people, because the class name of the returned object is
81
# a synonym of the object referenced in the method name.
82
return InterTree.get(other, self).compare(
83
want_unchanged=want_unchanged,
84
specific_files=specific_files,
85
extra_trees=extra_trees,
86
require_versioned=require_versioned,
87
include_root=include_root
90
def _iter_changes(self, from_tree, include_unchanged=False,
91
specific_files=None, pb=None, extra_trees=None,
92
require_versioned=True):
93
intertree = InterTree.get(from_tree, self)
94
return intertree._iter_changes(include_unchanged, specific_files, pb,
95
extra_trees, require_versioned)
98
"""Get a list of the conflicts in the tree.
100
Each conflict is an instance of bzrlib.conflicts.Conflict.
104
def get_parent_ids(self):
105
"""Get the parent ids for this tree.
107
:return: a list of parent ids. [] is returned to indicate
108
a tree with no parents.
109
:raises: BzrError if the parents are not known.
111
raise NotImplementedError(self.get_parent_ids)
113
def has_filename(self, filename):
114
"""True if the tree has given filename."""
115
raise NotImplementedError()
117
def has_id(self, file_id):
118
file_id = osutils.safe_file_id(file_id)
119
return self.inventory.has_id(file_id)
121
__contains__ = has_id
123
def has_or_had_id(self, file_id):
124
file_id = osutils.safe_file_id(file_id)
125
if file_id == self.inventory.root.file_id:
127
return self.inventory.has_id(file_id)
130
return iter(self.inventory)
132
def id2path(self, file_id):
133
file_id = osutils.safe_file_id(file_id)
134
return self.inventory.id2path(file_id)
136
def is_control_filename(self, filename):
137
"""True if filename is the name of a control file in this tree.
139
:param filename: A filename within the tree. This is a relative path
140
from the root of this tree.
142
This is true IF and ONLY IF the filename is part of the meta data
143
that bzr controls in this tree. I.E. a random .bzr directory placed
144
on disk will not be a control file for this tree.
146
return self.bzrdir.is_control_filename(filename)
149
def iter_entries_by_dir(self, specific_file_ids=None):
150
"""Walk the tree in 'by_dir' order.
152
This will yield each entry in the tree as a (path, entry) tuple. The
153
order that they are yielded is: the contents of a directory are
154
preceeded by the parent of a directory, and all the contents of a
155
directory are grouped together.
157
return self.inventory.iter_entries_by_dir(
158
specific_file_ids=specific_file_ids)
160
def iter_reference_entries(self):
161
for path, entry in self.iter_entries_by_dir():
162
if entry.kind == 'tree-reference':
165
def kind(self, file_id):
166
raise NotImplementedError("subclasses must implement kind")
168
def get_reference_revision(self, entry, path=None):
169
raise NotImplementedError("subclasses must implement "
170
"get_reference_revision")
172
def _comparison_data(self, entry, path):
173
"""Return a tuple of kind, executable, stat_value for a file.
175
entry may be None if there is no inventory entry for the file, but
176
path must always be supplied.
178
kind is None if there is no file present (even if an inventory id is
179
present). executable is False for non-file entries.
181
raise NotImplementedError(self._comparison_data)
183
def _file_size(self, entry, stat_value):
184
raise NotImplementedError(self._file_size)
186
def _get_inventory(self):
187
return self._inventory
189
def get_file(self, file_id):
190
"""Return a file object for the file file_id in the tree."""
191
raise NotImplementedError(self.get_file)
193
def get_file_by_path(self, path):
194
return self.get_file(self._inventory.path2id(path))
196
def get_symlink_target(self, file_id):
197
"""Get the target for a given file_id.
199
It is assumed that the caller already knows that file_id is referencing
201
:param file_id: Handle for the symlink entry.
202
:return: The path the symlink points to.
204
raise NotImplementedError(self.get_symlink_target)
206
def annotate_iter(self, file_id):
207
"""Return an iterator of revision_id, line tuples
209
For working trees (and mutable trees in general), the special
210
revision_id 'current:' will be used for lines that are new in this
211
tree, e.g. uncommitted changes.
212
:param file_id: The file to produce an annotated version from
214
raise NotImplementedError(self.annotate_iter)
216
inventory = property(_get_inventory,
217
doc="Inventory of this Tree")
219
def _check_retrieved(self, ie, f):
222
fp = fingerprint_file(f)
225
if ie.text_size is not None:
226
if ie.text_size != fp['size']:
227
raise BzrError("mismatched size for file %r in %r" % (ie.file_id, self._store),
228
["inventory expects %d bytes" % ie.text_size,
229
"file is actually %d bytes" % fp['size'],
230
"store is probably damaged/corrupt"])
232
if ie.text_sha1 != fp['sha1']:
233
raise BzrError("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
234
["inventory expects %s" % ie.text_sha1,
235
"file is actually %s" % fp['sha1'],
236
"store is probably damaged/corrupt"])
238
def path2id(self, path):
239
"""Return the id for path in this tree."""
240
return self._inventory.path2id(path)
242
def paths2ids(self, paths, trees=[], require_versioned=True):
243
"""Return all the ids that can be reached by walking from paths.
245
Each path is looked up in each this tree and any extras provided in
246
trees, and this is repeated recursively: the children in an extra tree
247
of a directory that has been renamed under a provided path in this tree
248
are all returned, even if none exist until a provided path in this
249
tree, and vice versa.
251
:param paths: An iterable of paths to start converting to ids from.
252
Alternatively, if paths is None, no ids should be calculated and None
253
will be returned. This is offered to make calling the api unconditional
254
for code that *might* take a list of files.
255
:param trees: Additional trees to consider.
256
:param require_versioned: If False, do not raise NotVersionedError if
257
an element of paths is not versioned in this tree and all of trees.
259
return find_ids_across_trees(paths, [self] + list(trees), require_versioned)
261
def print_file(self, file_id):
262
"""Print file with id `file_id` to stdout."""
263
file_id = osutils.safe_file_id(file_id)
265
sys.stdout.write(self.get_file_text(file_id))
270
def revision_tree(self, revision_id):
271
"""Obtain a revision tree for the revision revision_id.
273
The intention of this method is to allow access to possibly cached
274
tree data. Implementors of this method should raise NoSuchRevision if
275
the tree is not locally available, even if they could obtain the
276
tree via a repository or some other means. Callers are responsible
277
for finding the ultimate source for a revision tree.
279
:param revision_id: The revision_id of the requested tree.
281
:raises: NoSuchRevision if the tree cannot be obtained.
283
raise errors.NoSuchRevisionInTree(self, revision_id)
286
"""What files are present in this tree and unknown.
288
:return: an iterator over the unknown files.
295
def filter_unversioned_files(self, paths):
296
"""Filter out paths that are not versioned.
298
:return: set of paths.
300
# NB: we specifically *don't* call self.has_filename, because for
301
# WorkingTrees that can indicate files that exist on disk but that
303
pred = self.inventory.has_filename
304
return set((p for p in paths if not pred(p)))
306
def walkdirs(self, prefix=""):
307
"""Walk the contents of this tree from path down.
309
This yields all the data about the contents of a directory at a time.
310
After each directory has been yielded, if the caller has mutated the
311
list to exclude some directories, they are then not descended into.
313
The data yielded is of the form:
314
((directory-relpath, directory-path-from-root, directory-fileid),
315
[(relpath, basename, kind, lstat, path_from_tree_root, file_id,
316
versioned_kind), ...]),
317
- directory-relpath is the containing dirs relpath from prefix
318
- directory-path-from-root is the containing dirs path from /
319
- directory-fileid is the id of the directory if it is versioned.
320
- relpath is the relative path within the subtree being walked.
321
- basename is the basename
322
- kind is the kind of the file now. If unknonwn then the file is not
323
present within the tree - but it may be recorded as versioned. See
325
- lstat is the stat data *if* the file was statted.
326
- path_from_tree_root is the path from the root of the tree.
327
- file_id is the file_id is the entry is versioned.
328
- versioned_kind is the kind of the file as last recorded in the
329
versioning system. If 'unknown' the file is not versioned.
330
One of 'kind' and 'versioned_kind' must not be 'unknown'.
332
:param prefix: Start walking from prefix within the tree rather than
333
at the root. This allows one to walk a subtree but get paths that are
334
relative to a tree rooted higher up.
335
:return: an iterator over the directory data.
337
raise NotImplementedError(self.walkdirs)
340
class EmptyTree(Tree):
343
self._inventory = Inventory(root_id=None)
344
symbol_versioning.warn('EmptyTree is deprecated as of bzr 0.9 please'
345
' use repository.revision_tree instead.',
346
DeprecationWarning, stacklevel=2)
348
def get_parent_ids(self):
351
def get_symlink_target(self, file_id):
354
def has_filename(self, filename):
357
def kind(self, file_id):
358
file_id = osutils.safe_file_id(file_id)
359
assert self._inventory[file_id].kind == "directory"
362
def list_files(self, include_root=False):
365
def __contains__(self, file_id):
366
file_id = osutils.safe_file_id(file_id)
367
return (file_id in self._inventory)
369
def get_file_sha1(self, file_id, path=None, stat_value=None):
373
######################################################################
376
# TODO: Merge these two functions into a single one that can operate
377
# on either a whole tree or a set of files.
379
# TODO: Return the diff in order by filename, not by category or in
380
# random order. Can probably be done by lock-stepping through the
381
# filenames from both trees.
384
def file_status(filename, old_tree, new_tree):
385
"""Return single-letter status, old and new names for a file.
387
The complexity here is in deciding how to represent renames;
388
many complex cases are possible.
390
old_inv = old_tree.inventory
391
new_inv = new_tree.inventory
392
new_id = new_inv.path2id(filename)
393
old_id = old_inv.path2id(filename)
395
if not new_id and not old_id:
396
# easy: doesn't exist in either; not versioned at all
397
if new_tree.is_ignored(filename):
398
return 'I', None, None
400
return '?', None, None
402
# There is now a file of this name, great.
405
# There is no longer a file of this name, but we can describe
406
# what happened to the file that used to have
407
# this name. There are two possibilities: either it was
408
# deleted entirely, or renamed.
410
if new_inv.has_id(old_id):
411
return 'X', old_inv.id2path(old_id), new_inv.id2path(old_id)
413
return 'D', old_inv.id2path(old_id), None
415
# if the file_id is new in this revision, it is added
416
if new_id and not old_inv.has_id(new_id):
419
# if there used to be a file of this name, but that ID has now
420
# disappeared, it is deleted
421
if old_id and not new_inv.has_id(old_id):
428
def find_renames(old_inv, new_inv):
429
for file_id in old_inv:
430
if file_id not in new_inv:
432
old_name = old_inv.id2path(file_id)
433
new_name = new_inv.id2path(file_id)
434
if old_name != new_name:
435
yield (old_name, new_name)
438
def find_ids_across_trees(filenames, trees, require_versioned=True):
439
"""Find the ids corresponding to specified filenames.
441
All matches in all trees will be used, and all children of matched
442
directories will be used.
444
:param filenames: The filenames to find file_ids for (if None, returns
446
:param trees: The trees to find file_ids within
447
:param require_versioned: if true, all specified filenames must occur in
449
:return: a set of file ids for the specified filenames and their children.
453
specified_path_ids = _find_ids_across_trees(filenames, trees,
455
return _find_children_across_trees(specified_path_ids, trees)
456
# specified_ids = [id for path, id in _find_path_ids_across_trees(filenames, trees, require_versioned)]
457
# return _find_children_across_trees(specified_ids, trees)
459
def find_path_ids_across_trees(filenames, trees, require_versioned=True):
460
"""Find the paths and ids corresponding to specified filenames.
462
All matches in all trees will be used, and all children of matched
463
directories will be included
465
:param filenames: The filenames to find file_ids for
466
:param trees: The trees to find file_ids within
467
:param require_versioned: if true, all specified filenames must occur in
469
:return: a set of (path, file ids) for the specified filenames and their
470
children. The returned path is the path of the id in the first tree
471
that contains it. This matters when files have been moved
475
# This function needs to know the ids for filenames in all trees, then
476
# search for those same files and children in all the other trees.
477
# it is complicated by the same path in two trees being able to have
478
# different ids, which might both be present in both trees.
479
# consider two trees, which have had 'mv foo bar' and 'mv baz foo' done
480
# in this case, a diff of 'foo' should should changes to both the current
481
# 'bar' and the current 'foo' which was baz. Its arguable that if
482
# the situation is 'mv parent/foo bar' and 'mv baz parent/foo', that
483
# we should return the current bar and the current parent/foo' - at the
484
# moment we do, but we loop around all ids and all trees: I*T checks.
486
# Updating this algorithm to be fast in the common case:
487
# nothing has moved, all files have the same id in parent, child and there
488
# are only two trees (or one is working tree and the others are parents).
489
# walk the dirstate. as we find each path, gather the paths of that
490
# id in all trees. add a mapping from the id to the path in those trees.
491
# now lookup children by id, again in all trees; for these trees that
492
# nothing has moved in, the id->path mapping will allow us to find the
493
# parent trivially. To answer 'has anything been moved' in one of the
494
# dirstate parent trees though, we will need to stare harder at it.
496
# Now, given a path index, that is trivial for any one tree, and given
497
# that we can ask for additional data from a dirstate tree, its a single
498
# pass, though it will require scanning the entire tree to find paths
499
# that were at the current location.
500
# ideal results?: There are three things: tree, path, id. Pathologically
501
# we can have completely disjoint ids for each tree; but we cannot have
502
# disjoin paths for each tree, except if we scan each tree for the
503
# different ids from other trees.
505
specified_path_ids = _find_ids_across_trees(filenames, trees,
507
return _find_path_id_children_across_trees(specified_path_ids, trees)
510
def _find_ids_across_trees(filenames, trees, require_versioned):
511
"""Find the ids corresponding to specified filenames.
513
All matches in all trees will be used, but subdirectories are not scanned.
515
:param filenames: The filenames to find file_ids for
516
:param trees: The trees to find file_ids within
517
:param require_versioned: if true, all specified filenames must occur in
519
:return: a set of (path, file ids) for the specified filenames
522
interesting_ids = set()
523
for tree_path in filenames:
526
file_id = tree.path2id(tree_path)
527
if file_id is not None:
528
interesting_ids.add(file_id)
531
not_versioned.append(tree_path)
532
if len(not_versioned) > 0 and require_versioned:
533
raise errors.PathsNotVersionedError(not_versioned)
534
return interesting_ids
537
def _find_children_across_trees(specified_ids, trees):
538
"""Return a set including specified ids and their children
540
All matches in all trees will be used.
542
:param trees: The trees to find file_ids within
543
:return: a set containing all specified ids and their children
545
interesting_ids = set(specified_ids)
546
pending = interesting_ids
547
# now handle children of interesting ids
548
# we loop so that we handle all children of each id in both trees
549
while len(pending) > 0:
551
for file_id in pending:
553
if not tree.has_id(file_id):
555
entry = tree.inventory[file_id]
556
for child in getattr(entry, 'children', {}).itervalues():
557
if child.file_id not in interesting_ids:
558
new_pending.add(child.file_id)
559
interesting_ids.update(new_pending)
560
pending = new_pending
561
return interesting_ids
564
class InterTree(InterObject):
565
"""This class represents operations taking place between two Trees.
567
Its instances have methods like 'compare' and contain references to the
568
source and target trees these operations are to be carried out on.
570
clients of bzrlib should not need to use InterTree directly, rather they
571
should use the convenience methods on Tree such as 'Tree.compare()' which
572
will pass through to InterTree as appropriate.
578
def compare(self, want_unchanged=False, specific_files=None,
579
extra_trees=None, require_versioned=False, include_root=False):
580
"""Return the changes from source to target.
582
:return: A TreeDelta.
583
:param specific_files: An optional list of file paths to restrict the
584
comparison to. When mapping filenames to ids, all matches in all
585
trees (including optional extra_trees) are used, and all children of
586
matched directories are included.
587
:param want_unchanged: An optional boolean requesting the inclusion of
588
unchanged entries in the result.
589
:param extra_trees: An optional list of additional trees to use when
590
mapping the contents of specific_files (paths) to file_ids.
591
:param require_versioned: An optional boolean (defaults to False). When
592
supplied and True all the 'specific_files' must be versioned, or
593
a PathsNotVersionedError will be thrown.
595
# NB: show_status depends on being able to pass in non-versioned files
596
# and report them as unknown
597
trees = (self.source,)
598
if extra_trees is not None:
599
trees = trees + tuple(extra_trees)
600
# target is usually the newer tree:
601
specific_file_ids = self.target.paths2ids(specific_files, trees,
602
require_versioned=require_versioned)
603
if specific_files and not specific_file_ids:
604
# All files are unversioned, so just return an empty delta
605
# _compare_trees would think we want a complete delta
606
return delta.TreeDelta()
607
return delta._compare_trees(self.source, self.target, want_unchanged,
608
specific_files, include_root, extra_trees=extra_trees)
610
def _iter_changes(self, include_unchanged=False,
611
specific_files=None, pb=None, extra_trees=[],
612
require_versioned=True):
613
"""Generate an iterator of changes between trees.
616
(file_id, path, changed_content, versioned, parent, name, kind,
619
Path is relative to the target tree. changed_content is True if the
620
file's content has changed. This includes changes to its kind, and to
623
versioned, parent, name, kind, executable are tuples of (from, to).
624
If a file is missing in a tree, its kind is None.
626
Iteration is done in parent-to-child order, relative to the target
629
There is no guarantee that all paths are in sorted order: the
630
requirement to expand the search due to renames may result in children
631
that should be found early being found late in the search, after
632
lexically later results have been returned.
633
:param require_versioned: Raise errors.PathsNotVersionedError if a
634
path in the specific_files list is not versioned in one of
635
source, target or extra_trees.
637
lookup_trees = [self.source]
639
lookup_trees.extend(extra_trees)
640
specific_file_ids = self.target.paths2ids(specific_files,
641
lookup_trees, require_versioned=require_versioned)
643
from_entries_by_dir = list(self.source.inventory.iter_entries_by_dir(
644
specific_file_ids=specific_file_ids))
645
from_data = dict((e.file_id, (p, e)) for p, e in from_entries_by_dir)
646
to_entries_by_dir = list(self.target.inventory.iter_entries_by_dir(
647
specific_file_ids=specific_file_ids))
648
num_entries = len(from_entries_by_dir) + len(to_entries_by_dir)
650
for to_path, to_entry in to_entries_by_dir:
651
file_id = to_entry.file_id
652
to_paths[file_id] = to_path
654
changed_content = False
655
from_path, from_entry = from_data.get(file_id, (None, None))
656
from_versioned = (from_entry is not None)
657
if from_entry is not None:
658
from_versioned = True
659
from_name = from_entry.name
660
from_parent = from_entry.parent_id
661
from_kind, from_executable, from_stat = \
662
self.source._comparison_data(from_entry, from_path)
665
from_versioned = False
669
from_executable = None
670
versioned = (from_versioned, True)
671
to_kind, to_executable, to_stat = \
672
self.target._comparison_data(to_entry, to_path)
673
kind = (from_kind, to_kind)
674
if kind[0] != kind[1]:
675
changed_content = True
676
elif from_kind == 'file':
677
from_size = self.source._file_size(from_entry, from_stat)
678
to_size = self.target._file_size(to_entry, to_stat)
679
if from_size != to_size:
680
changed_content = True
681
elif (self.source.get_file_sha1(file_id, from_path, from_stat) !=
682
self.target.get_file_sha1(file_id, to_path, to_stat)):
683
changed_content = True
684
elif from_kind == 'symlink':
685
if (self.source.get_symlink_target(file_id) !=
686
self.target.get_symlink_target(file_id)):
687
changed_content = True
688
elif from_kind == 'tree-reference':
689
if (self.source.get_reference_revision(from_entry, from_path)
690
!= self.target.get_reference_revision(to_entry, to_path)):
691
changed_content = True
692
parent = (from_parent, to_entry.parent_id)
693
name = (from_name, to_entry.name)
694
executable = (from_executable, to_executable)
696
pb.update('comparing files', entry_count, num_entries)
697
if (changed_content is not False or versioned[0] != versioned[1]
698
or parent[0] != parent[1] or name[0] != name[1] or
699
executable[0] != executable[1] or include_unchanged):
700
yield (file_id, to_path, changed_content, versioned, parent,
701
name, kind, executable)
703
def get_to_path(from_entry):
704
if from_entry.parent_id is None:
707
if from_entry.parent_id not in to_paths:
708
get_to_path(self.source.inventory[from_entry.parent_id])
709
to_path = osutils.pathjoin(to_paths[from_entry.parent_id],
711
to_paths[from_entry.file_id] = to_path
714
for path, from_entry in from_entries_by_dir:
715
file_id = from_entry.file_id
716
if file_id in to_paths:
718
to_path = get_to_path(from_entry)
721
pb.update('comparing files', entry_count, num_entries)
722
versioned = (True, False)
723
parent = (from_entry.parent_id, None)
724
name = (from_entry.name, None)
725
from_kind, from_executable, stat_value = \
726
self.source._comparison_data(from_entry, path)
727
kind = (from_kind, None)
728
executable = (from_executable, None)
729
changed_content = True
730
# the parent's path is necessarily known at this point.
731
yield(file_id, to_path, changed_content, versioned, parent,
732
name, kind, executable)
735
# This was deprecated before 0.12, but did not have an official warning
736
@symbol_versioning.deprecated_function(symbol_versioning.zero_twelve)
737
def RevisionTree(*args, **kwargs):
738
"""RevisionTree has moved to bzrlib.revisiontree.RevisionTree()
740
Accessing it as bzrlib.tree.RevisionTree has been deprecated as of
743
from bzrlib.revisiontree import RevisionTree as _RevisionTree
744
return _RevisionTree(*args, **kwargs)
added
removed
bzrlib/tree.py
