13
13
# You should have received a copy of the GNU General Public License
14
14
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
15
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17
17
"""Tree classes, representing directory at point in time.
20
from __future__ import absolute_import
24
from .lazy_import import lazy_import
25
lazy_import(globals(), """
29
conflicts as _mod_conflicts,
21
from collections import deque
22
from cStringIO import StringIO
33
28
revision as _mod_revision,
37
from breezy.i18n import gettext
45
from .inter import InterObject
52
class FileTimestampUnavailable(errors.BzrError):
54
_fmt = "The filestamp for %(path)s is not available."
58
def __init__(self, path):
62
class TreeEntry(object):
63
"""An entry that implements the minimum interface used by commands.
66
def __eq__(self, other):
67
# yes, this is ugly, TODO: best practice __eq__ style.
68
return (isinstance(other, TreeEntry)
69
and other.__class__ == self.__class__)
71
def kind_character(self):
75
class TreeDirectory(TreeEntry):
76
"""See TreeEntry. This is a directory in a working tree."""
78
def kind_character(self):
82
class TreeFile(TreeEntry):
83
"""See TreeEntry. This is a regular file in a working tree."""
85
def kind_character(self):
89
class TreeLink(TreeEntry):
90
"""See TreeEntry. This is a symlink in a working tree."""
92
def kind_character(self):
96
class TreeReference(TreeEntry):
97
"""See TreeEntry. This is a reference to a nested tree in a working tree."""
99
def kind_character(self):
29
conflicts as _mod_conflicts,
32
from bzrlib.decorators import needs_read_lock
33
from bzrlib.errors import BzrError, BzrCheckError
34
from bzrlib import errors
35
from bzrlib.inventory import Inventory, InventoryFile
36
from bzrlib.inter import InterObject
37
from bzrlib.osutils import fingerprint_file
38
import bzrlib.revision
39
from bzrlib.trace import mutter, note
103
42
class Tree(object):
104
43
"""Abstract file tree.
106
45
There are several subclasses:
108
47
* `WorkingTree` exists as files on disk editable by the user.
110
49
* `RevisionTree` is a tree as recorded at some point in the past.
51
Trees contain an `Inventory` object, and also know how to retrieve
52
file texts mentioned in the inventory, either from a working
53
directory or from a store.
55
It is possible for trees to contain files that are not described
56
in their inventory or vice versa; for this use `filenames()`.
112
58
Trees can be compared, etc, regardless of whether they are working
113
59
trees or versioned trees.
116
def supports_rename_tracking(self):
117
"""Whether this tree supports rename tracking.
119
This defaults to True, but some implementations may want to override
124
def has_versioned_directories(self):
125
"""Whether this tree can contain explicitly versioned directories.
127
This defaults to True, but some implementations may want to override
132
62
def changes_from(self, other, want_unchanged=False, specific_files=None,
133
63
extra_trees=None, require_versioned=False, include_root=False,
134
64
want_unversioned=False):
206
def all_file_ids(self):
207
"""Iterate through all file ids, including ids for missing files."""
208
raise NotImplementedError(self.all_file_ids)
210
def all_versioned_paths(self):
211
"""Iterate through all paths, including paths for missing files."""
212
raise NotImplementedError(self.all_versioned_paths)
146
return iter(self.inventory)
214
148
def id2path(self, file_id):
215
149
"""Return the path for a file id.
217
151
:raises NoSuchId:
219
raise NotImplementedError(self.id2path)
221
def iter_entries_by_dir(self, specific_files=None):
153
return self.inventory.id2path(file_id)
155
def is_control_filename(self, filename):
156
"""True if filename is the name of a control file in this tree.
158
:param filename: A filename within the tree. This is a relative path
159
from the root of this tree.
161
This is true IF and ONLY IF the filename is part of the meta data
162
that bzr controls in this tree. I.E. a random .bzr directory placed
163
on disk will not be a control file for this tree.
165
return self.bzrdir.is_control_filename(filename)
168
def iter_entries_by_dir(self, specific_file_ids=None):
222
169
"""Walk the tree in 'by_dir' order.
224
This will yield each entry in the tree as a (path, entry) tuple.
225
The order that they are yielded is:
227
Directories are walked in a depth-first lexicographical order,
228
however, whenever a directory is reached, all of its direct child
229
nodes are yielded in lexicographical order before yielding the
232
For example, in the tree::
242
The yield order (ignoring root) would be::
244
a, f, a/b, a/d, a/b/c, a/d/e, f/g
246
raise NotImplementedError(self.iter_entries_by_dir)
248
def iter_child_entries(self, path, file_id=None):
249
"""Iterate over the children of a directory or tree reference.
251
:param path: Path of the directory
252
:param file_id: Optional file id of the directory/tree-reference
253
:raise NoSuchId: When the file_id does not exist
254
:return: Iterator over entries in the directory
256
raise NotImplementedError(self.iter_child_entries)
258
def list_files(self, include_root=False, from_dir=None, recursive=True):
259
"""List all files in this tree.
261
:param include_root: Whether to include the entry for the tree root
262
:param from_dir: Directory under which to list files
263
:param recursive: Whether to list files recursively
264
:return: iterator over tuples of (path, versioned, kind, file_id,
267
raise NotImplementedError(self.list_files)
171
This will yield each entry in the tree as a (path, entry) tuple. The
172
order that they are yielded is: the contents of a directory are
173
preceeded by the parent of a directory, and all the contents of a
174
directory are grouped together.
176
return self.inventory.iter_entries_by_dir(
177
specific_file_ids=specific_file_ids)
269
179
def iter_references(self):
270
if self.supports_tree_reference():
271
for path, entry in self.iter_entries_by_dir():
272
if entry.kind == 'tree-reference':
273
yield path, entry.file_id
180
for path, entry in self.iter_entries_by_dir():
181
if entry.kind == 'tree-reference':
182
yield path, entry.file_id
275
def kind(self, path, file_id=None):
184
def kind(self, file_id):
276
185
raise NotImplementedError("Tree subclass %s must implement kind"
277
186
% self.__class__.__name__)
279
def stored_kind(self, path, file_id=None):
280
"""File kind stored for this file_id.
282
May not match kind on disk for working trees. Always available
283
for versioned files, even when the file itself is missing.
285
return self.kind(path, file_id)
287
188
def path_content_summary(self, path):
288
189
"""Get a summary of the information about path.
290
All the attributes returned are for the canonical form, not the
291
convenient form (if content filters are in use.)
293
191
:param path: A relative path within the tree.
294
192
:return: A tuple containing kind, size, exec, sha1-or-link.
295
193
Kind is always present (see tree.kind()).
296
size is present if kind is file and the size of the
297
canonical form can be cheaply determined, None otherwise.
194
size is present if kind is file, None otherwise.
298
195
exec is None unless kind is file and the platform supports the 'x'
300
197
sha1-or-link is the link target if kind is symlink, or the sha1 if
319
216
raise NotImplementedError(self._comparison_data)
321
def get_file(self, path, file_id=None):
218
def _file_size(self, entry, stat_value):
219
raise NotImplementedError(self._file_size)
221
def _get_inventory(self):
222
return self._inventory
224
def get_file(self, file_id, path=None):
322
225
"""Return a file object for the file file_id in the tree.
324
227
If both file_id and path are defined, it is implementation defined as
325
228
to which one is used.
327
230
raise NotImplementedError(self.get_file)
329
def get_file_with_stat(self, path, file_id=None):
330
"""Get a file handle and stat object for file_id.
332
The default implementation returns (self.get_file, None) for backwards
335
:param path: The path of the file.
336
:param file_id: The file id to read, if it is known.
337
:return: A tuple (file_handle, stat_value_or_None). If the tree has
338
no stat facility, or need for a stat cache feedback during commit,
339
it may return None for the second element of the tuple.
341
return (self.get_file(path, file_id), None)
343
def get_file_text(self, path, file_id=None):
344
"""Return the byte content of a file.
346
:param path: The path of the file.
347
:param file_id: The file_id of the file.
349
If both file_id and path are supplied, an implementation may use
352
:returns: A single byte string for the whole file.
354
with self.get_file(path, file_id) as my_file:
355
return my_file.read()
357
def get_file_lines(self, path, file_id=None):
358
"""Return the content of a file, as lines.
360
:param path: The path of the file.
361
:param file_id: The file_id of the file.
363
If both file_id and path are supplied, an implementation may use
366
return osutils.split_lines(self.get_file_text(path, file_id))
368
def get_file_verifier(self, path, file_id=None, stat_value=None):
369
"""Return a verifier for a file.
371
The default implementation returns a sha1.
373
:param file_id: The handle for this file.
374
:param path: The path that this file can be found at.
375
These must point to the same object.
376
:param stat_value: Optional stat value for the object
377
:return: Tuple with verifier name and verifier data
379
return ("SHA1", self.get_file_sha1(path, file_id,
380
stat_value=stat_value))
382
def get_file_sha1(self, path, file_id=None, stat_value=None):
383
"""Return the SHA1 file for a file.
385
:note: callers should use get_file_verifier instead
386
where possible, as the underlying repository implementation may
387
have quicker access to a non-sha1 verifier.
389
:param path: The path that this file can be found at.
390
:param file_id: The handle for this file.
391
These must point to the same object.
392
:param stat_value: Optional stat value for the object
394
raise NotImplementedError(self.get_file_sha1)
396
def get_file_mtime(self, path, file_id=None):
232
def get_file_mtime(self, file_id, path=None):
397
233
"""Return the modification time for a file.
235
:param file_id: The handle for this file.
399
236
:param path: The path that this file can be found at.
400
:param file_id: The handle for this file.
401
237
These must point to the same object.
403
239
raise NotImplementedError(self.get_file_mtime)
405
def get_file_size(self, path, file_id=None):
406
"""Return the size of a file in bytes.
408
This applies only to regular files. If invoked on directories or
409
symlinks, it will return None.
410
:param file_id: The file-id of the file
412
raise NotImplementedError(self.get_file_size)
414
def is_executable(self, path, file_id=None):
415
"""Check if a file is executable.
417
:param path: The path that this file can be found at.
418
:param file_id: The handle for this file.
419
These must point to the same object.
421
raise NotImplementedError(self.is_executable)
241
def get_file_by_path(self, path):
242
return self.get_file(self._inventory.path2id(path), path)
423
244
def iter_files_bytes(self, desired_files):
424
245
"""Iterate through file contents.
463
281
"""Return the file_id for the root of this tree."""
464
282
raise NotImplementedError(self.get_root_id)
466
def annotate_iter(self, path, file_id=None,
467
default_revision=_mod_revision.CURRENT_REVISION):
284
def annotate_iter(self, file_id):
468
285
"""Return an iterator of revision_id, line tuples.
470
287
For working trees (and mutable trees in general), the special
471
288
revision_id 'current:' will be used for lines that are new in this
472
289
tree, e.g. uncommitted changes.
473
290
:param file_id: The file to produce an annotated version from
474
:param default_revision: For lines that don't match a basis, mark them
475
with this revision id. Not all implementations will make use of
478
292
raise NotImplementedError(self.annotate_iter)
480
def _iter_parent_trees(self):
481
"""Iterate through parent trees, defaulting to Tree.revision_tree."""
482
for revision_id in self.get_parent_ids():
484
yield self.revision_tree(revision_id)
485
except errors.NoSuchRevisionInTree:
486
yield self.repository.revision_tree(revision_id)
294
def plan_file_merge(self, file_id, other):
295
"""Generate a merge plan based on annotations.
297
If the file contains uncommitted changes in this tree, they will be
298
attributed to the 'current:' pseudo-revision. If the file contains
299
uncommitted changes in the other tree, they will be assigned to the
300
'other:' pseudo-revision.
302
from bzrlib import merge
303
annotated_a = list(self.annotate_iter(file_id,
304
_mod_revision.CURRENT_REVISION))
305
annotated_b = list(other.annotate_iter(file_id, 'other:'))
306
ancestors_a = self._get_ancestors(_mod_revision.CURRENT_REVISION)
307
ancestors_b = other._get_ancestors('other:')
308
return merge._plan_annotate_merge(annotated_a, annotated_b,
309
ancestors_a, ancestors_b)
311
inventory = property(_get_inventory,
312
doc="Inventory of this Tree")
314
def _check_retrieved(self, ie, f):
317
fp = fingerprint_file(f)
320
if ie.text_size is not None:
321
if ie.text_size != fp['size']:
322
raise BzrError("mismatched size for file %r in %r" % (ie.file_id, self._store),
323
["inventory expects %d bytes" % ie.text_size,
324
"file is actually %d bytes" % fp['size'],
325
"store is probably damaged/corrupt"])
327
if ie.text_sha1 != fp['sha1']:
328
raise BzrError("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
329
["inventory expects %s" % ie.text_sha1,
330
"file is actually %s" % fp['sha1'],
331
"store is probably damaged/corrupt"])
488
334
def path2id(self, path):
489
335
"""Return the id for path in this tree."""
490
raise NotImplementedError(self.path2id)
492
def is_versioned(self, path):
493
"""Check whether path is versioned.
495
:param path: Path to check
498
return self.path2id(path) is not None
500
def find_related_paths_across_trees(self, paths, trees=[],
501
require_versioned=True):
502
"""Find related paths in tree corresponding to specified filenames in any
505
All matches in all trees will be used, and all children of matched
506
directories will be used.
508
:param paths: The filenames to find related paths for (if None, returns
510
:param trees: The trees to find file_ids within
511
:param require_versioned: if true, all specified filenames must occur in
513
:return: a set of paths for the specified filenames and their children
516
raise NotImplementedError(self.find_related_paths_across_trees)
336
return self._inventory.path2id(path)
338
def paths2ids(self, paths, trees=[], require_versioned=True):
339
"""Return all the ids that can be reached by walking from paths.
341
Each path is looked up in this tree and any extras provided in
342
trees, and this is repeated recursively: the children in an extra tree
343
of a directory that has been renamed under a provided path in this tree
344
are all returned, even if none exist under a provided path in this
345
tree, and vice versa.
347
:param paths: An iterable of paths to start converting to ids from.
348
Alternatively, if paths is None, no ids should be calculated and None
349
will be returned. This is offered to make calling the api unconditional
350
for code that *might* take a list of files.
351
:param trees: Additional trees to consider.
352
:param require_versioned: If False, do not raise NotVersionedError if
353
an element of paths is not versioned in this tree and all of trees.
355
return find_ids_across_trees(paths, [self] + list(trees), require_versioned)
357
def print_file(self, file_id):
358
"""Print file with id `file_id` to stdout."""
360
sys.stdout.write(self.get_file_text(file_id))
518
362
def lock_read(self):
519
"""Lock this tree for multiple read only operations.
521
:return: A breezy.lock.LogicalLockResult.
523
return lock.LogicalLockResult(self.unlock)
525
365
def revision_tree(self, revision_id):
526
366
"""Obtain a revision tree for the revision revision_id.
528
368
The intention of this method is to allow access to possibly cached
529
369
tree data. Implementors of this method should raise NoSuchRevision if
530
the tree is not locally available, even if they could obtain the
531
tree via a repository or some other means. Callers are responsible
370
the tree is not locally available, even if they could obtain the
371
tree via a repository or some other means. Callers are responsible
532
372
for finding the ultimate source for a revision tree.
534
374
:param revision_id: The revision_id of the requested tree.
591
432
raise NotImplementedError(self.walkdirs)
593
def supports_content_filtering(self):
435
class EmptyTree(Tree):
438
self._inventory = Inventory(root_id=None)
439
symbol_versioning.warn('EmptyTree is deprecated as of bzr 0.9 please'
440
' use repository.revision_tree instead.',
441
DeprecationWarning, stacklevel=2)
443
def get_parent_ids(self):
446
def get_symlink_target(self, file_id):
449
def has_filename(self, filename):
596
def _content_filter_stack(self, path=None):
597
"""The stack of content filters for a path if filtering is supported.
599
Readers will be applied in first-to-last order.
600
Writers will be applied in last-to-first order.
601
Either the path or the file-id needs to be provided.
603
:param path: path relative to the root of the tree
605
:return: the list of filters - [] if there are none
607
filter_pref_names = filters._get_registered_names()
608
if len(filter_pref_names) == 0:
610
prefs = next(self.iter_search_rules([path], filter_pref_names))
611
stk = filters._get_filter_stack_for(prefs)
612
if 'filters' in debug.debug_flags:
613
trace.note(gettext("*** {0} content-filter: {1} => {2!r}").format(path, prefs, stk))
616
def _content_filter_stack_provider(self):
617
"""A function that returns a stack of ContentFilters.
619
The function takes a path (relative to the top of the tree) and a
620
file-id as parameters.
622
:return: None if content filtering is not supported by this tree.
624
if self.supports_content_filtering():
625
return lambda path, file_id: \
626
self._content_filter_stack(path)
630
def iter_search_rules(self, path_names, pref_names=None,
631
_default_searcher=None):
632
"""Find the preferences for filenames in a tree.
634
:param path_names: an iterable of paths to find attributes for.
635
Paths are given relative to the root of the tree.
636
:param pref_names: the list of preferences to lookup - None for all
637
:param _default_searcher: private parameter to assist testing - don't use
638
:return: an iterator of tuple sequences, one per path-name.
639
See _RulesSearcher.get_items for details on the tuple sequence.
641
if _default_searcher is None:
642
_default_searcher = rules._per_user_searcher
643
searcher = self._get_rules_searcher(_default_searcher)
644
if searcher is not None:
645
if pref_names is not None:
646
for path in path_names:
647
yield searcher.get_selected_items(path, pref_names)
649
for path in path_names:
650
yield searcher.get_items(path)
652
def _get_rules_searcher(self, default_searcher):
653
"""Get the RulesSearcher for this tree given the default one."""
654
searcher = default_searcher
657
def archive(self, format, name, root='', subdir=None,
659
"""Create an archive of this tree.
661
:param format: Format name (e.g. 'tar')
662
:param name: target file name
663
:param root: Root directory name (or None)
664
:param subdir: Subdirectory to export (or None)
665
:return: Iterator over archive chunks
667
from .archive import create_archive
668
with self.lock_read():
669
return create_archive(format, self, name, root,
670
subdir, force_mtime=force_mtime)
673
def versionable_kind(cls, kind):
674
"""Check if this tree support versioning a specific file kind."""
675
return (kind in ('file', 'directory', 'symlink', 'tree-reference'))
452
def kind(self, file_id):
453
assert self._inventory[file_id].kind == "directory"
456
def list_files(self, include_root=False):
459
def __contains__(self, file_id):
460
return (file_id in self._inventory)
462
def get_file_sha1(self, file_id, path=None, stat_value=None):
466
######################################################################
469
# TODO: Merge these two functions into a single one that can operate
470
# on either a whole tree or a set of files.
472
# TODO: Return the diff in order by filename, not by category or in
473
# random order. Can probably be done by lock-stepping through the
474
# filenames from both trees.
477
def file_status(filename, old_tree, new_tree):
478
"""Return single-letter status, old and new names for a file.
480
The complexity here is in deciding how to represent renames;
481
many complex cases are possible.
483
old_inv = old_tree.inventory
484
new_inv = new_tree.inventory
485
new_id = new_inv.path2id(filename)
486
old_id = old_inv.path2id(filename)
488
if not new_id and not old_id:
489
# easy: doesn't exist in either; not versioned at all
490
if new_tree.is_ignored(filename):
491
return 'I', None, None
493
return '?', None, None
495
# There is now a file of this name, great.
498
# There is no longer a file of this name, but we can describe
499
# what happened to the file that used to have
500
# this name. There are two possibilities: either it was
501
# deleted entirely, or renamed.
503
if new_inv.has_id(old_id):
504
return 'X', old_inv.id2path(old_id), new_inv.id2path(old_id)
506
return 'D', old_inv.id2path(old_id), None
508
# if the file_id is new in this revision, it is added
509
if new_id and not old_inv.has_id(new_id):
512
# if there used to be a file of this name, but that ID has now
513
# disappeared, it is deleted
514
if old_id and not new_inv.has_id(old_id):
521
def find_renames(old_inv, new_inv):
522
for file_id in old_inv:
523
if file_id not in new_inv:
525
old_name = old_inv.id2path(file_id)
526
new_name = new_inv.id2path(file_id)
527
if old_name != new_name:
528
yield (old_name, new_name)
531
def find_ids_across_trees(filenames, trees, require_versioned=True):
532
"""Find the ids corresponding to specified filenames.
534
All matches in all trees will be used, and all children of matched
535
directories will be used.
537
:param filenames: The filenames to find file_ids for (if None, returns
539
:param trees: The trees to find file_ids within
540
:param require_versioned: if true, all specified filenames must occur in
542
:return: a set of file ids for the specified filenames and their children.
546
specified_path_ids = _find_ids_across_trees(filenames, trees,
548
return _find_children_across_trees(specified_path_ids, trees)
551
def _find_ids_across_trees(filenames, trees, require_versioned):
552
"""Find the ids corresponding to specified filenames.
554
All matches in all trees will be used, but subdirectories are not scanned.
556
:param filenames: The filenames to find file_ids for
557
:param trees: The trees to find file_ids within
558
:param require_versioned: if true, all specified filenames must occur in
560
:return: a set of file ids for the specified filenames
563
interesting_ids = set()
564
for tree_path in filenames:
567
file_id = tree.path2id(tree_path)
568
if file_id is not None:
569
interesting_ids.add(file_id)
572
not_versioned.append(tree_path)
573
if len(not_versioned) > 0 and require_versioned:
574
raise errors.PathsNotVersionedError(not_versioned)
575
return interesting_ids
578
def _find_children_across_trees(specified_ids, trees):
579
"""Return a set including specified ids and their children.
581
All matches in all trees will be used.
583
:param trees: The trees to find file_ids within
584
:return: a set containing all specified ids and their children
586
interesting_ids = set(specified_ids)
587
pending = interesting_ids
588
# now handle children of interesting ids
589
# we loop so that we handle all children of each id in both trees
590
while len(pending) > 0:
592
for file_id in pending:
594
if not tree.has_id(file_id):
596
entry = tree.inventory[file_id]
597
for child in getattr(entry, 'children', {}).itervalues():
598
if child.file_id not in interesting_ids:
599
new_pending.add(child.file_id)
600
interesting_ids.update(new_pending)
601
pending = new_pending
602
return interesting_ids
678
605
class InterTree(InterObject):
681
608
Its instances have methods like 'compare' and contain references to the
682
609
source and target trees these operations are to be carried out on.
684
Clients of breezy should not need to use InterTree directly, rather they
611
Clients of bzrlib should not need to use InterTree directly, rather they
685
612
should use the convenience methods on Tree such as 'Tree.compare()' which
686
613
will pass through to InterTree as appropriate.
689
# Formats that will be used to test this InterTree. If both are
690
# None, this InterTree will not be tested (e.g. because a complex
692
_matching_from_tree_format = None
693
_matching_to_tree_format = None
698
def is_compatible(kls, source, target):
699
# The default implementation is naive and uses the public API, so
700
# it works for all trees.
703
def _changes_from_entries(self, source_entry, target_entry, source_path,
705
"""Generate a iter_changes tuple between source_entry and target_entry.
707
:param source_entry: An inventory entry from self.source, or None.
708
:param target_entry: An inventory entry from self.target, or None.
709
:param source_path: The path of source_entry.
710
:param target_path: The path of target_entry.
711
:return: A tuple, item 0 of which is an iter_changes result tuple, and
712
item 1 is True if there are any changes in the result tuple.
714
if source_entry is None:
715
if target_entry is None:
717
file_id = target_entry.file_id
719
file_id = source_entry.file_id
720
if source_entry is not None:
721
source_versioned = True
722
source_name = source_entry.name
723
source_parent = source_entry.parent_id
724
source_kind, source_executable, source_stat = \
725
self.source._comparison_data(source_entry, source_path)
727
source_versioned = False
731
source_executable = None
732
if target_entry is not None:
733
target_versioned = True
734
target_name = target_entry.name
735
target_parent = target_entry.parent_id
736
target_kind, target_executable, target_stat = \
737
self.target._comparison_data(target_entry, target_path)
739
target_versioned = False
743
target_executable = None
744
versioned = (source_versioned, target_versioned)
745
kind = (source_kind, target_kind)
746
changed_content = False
747
if source_kind != target_kind:
748
changed_content = True
749
elif source_kind == 'file':
750
if not self.file_content_matches(
751
source_path, target_path,
752
file_id, file_id, source_stat, target_stat):
753
changed_content = True
754
elif source_kind == 'symlink':
755
if (self.source.get_symlink_target(source_path, file_id) !=
756
self.target.get_symlink_target(target_path, file_id)):
757
changed_content = True
758
elif source_kind == 'tree-reference':
759
if (self.source.get_reference_revision(source_path, file_id)
760
!= self.target.get_reference_revision(target_path, file_id)):
761
changed_content = True
762
parent = (source_parent, target_parent)
763
name = (source_name, target_name)
764
executable = (source_executable, target_executable)
765
if (changed_content is not False or versioned[0] != versioned[1]
766
or parent[0] != parent[1] or name[0] != name[1] or
767
executable[0] != executable[1]):
771
return (file_id, (source_path, target_path), changed_content,
772
versioned, parent, name, kind, executable), changes
774
619
def compare(self, want_unchanged=False, specific_files=None,
775
620
extra_trees=None, require_versioned=False, include_root=False,
776
621
want_unversioned=False):
825
683
:param require_versioned: Raise errors.PathsNotVersionedError if a
826
684
path in the specific_files list is not versioned in one of
827
685
source, target or extra_trees.
828
:param specific_files: An optional list of file paths to restrict the
829
comparison to. When mapping filenames to ids, all matches in all
830
trees (including optional extra_trees) are used, and all children
831
of matched directories are included. The parents in the target tree
832
of the specific files up to and including the root of the tree are
833
always evaluated for changes too.
834
686
:param want_unversioned: Should unversioned files be returned in the
835
687
output. An unversioned file is defined as one with (False, False)
836
688
for the versioned pair.
841
extra_trees = list(extra_trees)
842
# The ids of items we need to examine to insure delta consistency.
843
precise_file_ids = set()
844
changed_file_ids = []
691
lookup_trees = [self.source]
693
lookup_trees.extend(extra_trees)
845
694
if specific_files == []:
846
target_specific_files = []
847
source_specific_files = []
695
specific_file_ids = []
849
target_specific_files = self.target.find_related_paths_across_trees(
850
specific_files, [self.source] + extra_trees,
851
require_versioned=require_versioned)
852
source_specific_files = self.source.find_related_paths_across_trees(
853
specific_files, [self.target] + extra_trees,
854
require_versioned=require_versioned)
855
if specific_files is not None:
856
# reparented or added entries must have their parents included
857
# so that valid deltas can be created. The seen_parents set
858
# tracks the parents that we need to have.
859
# The seen_dirs set tracks directory entries we've yielded.
860
# After outputting version object in to_entries we set difference
861
# the two seen sets and start checking parents.
697
specific_file_ids = self.target.paths2ids(specific_files,
698
lookup_trees, require_versioned=require_versioned)
864
699
if want_unversioned:
865
700
all_unversioned = sorted([(p.split('/'), p) for p in
866
701
self.target.extras()
867
702
if specific_files is None or
868
703
osutils.is_inside_any(specific_files, p)])
869
all_unversioned = collections.deque(all_unversioned)
704
all_unversioned = deque(all_unversioned)
871
all_unversioned = collections.deque()
706
all_unversioned = deque()
873
from_entries_by_dir = list(self.source.iter_entries_by_dir(
874
specific_files=source_specific_files))
708
from_entries_by_dir = list(self.source.inventory.iter_entries_by_dir(
709
specific_file_ids=specific_file_ids))
875
710
from_data = dict((e.file_id, (p, e)) for p, e in from_entries_by_dir)
876
to_entries_by_dir = list(self.target.iter_entries_by_dir(
877
specific_files=target_specific_files))
711
to_entries_by_dir = list(self.target.inventory.iter_entries_by_dir(
712
specific_file_ids=specific_file_ids))
878
713
num_entries = len(from_entries_by_dir) + len(to_entries_by_dir)
880
# the unversioned path lookup only occurs on real trees - where there
715
# the unversioned path lookup only occurs on real trees - where there
881
716
# can be extras. So the fake_entry is solely used to look up
882
717
# executable it values when execute is not supported.
883
fake_entry = TreeFile()
884
for target_path, target_entry in to_entries_by_dir:
885
while (all_unversioned and
886
all_unversioned[0][0] < target_path.split('/')):
718
fake_entry = InventoryFile('unused', 'unused', 'unused')
719
for to_path, to_entry in to_entries_by_dir:
720
while all_unversioned and all_unversioned[0][0] < to_path.split('/'):
887
721
unversioned_path = all_unversioned.popleft()
888
target_kind, target_executable, target_stat = \
722
to_kind, to_executable, to_stat = \
889
723
self.target._comparison_data(fake_entry, unversioned_path[1])
890
724
yield (None, (None, unversioned_path[1]), True, (False, False),
892
726
(None, unversioned_path[0][-1]),
894
(None, target_executable))
895
source_path, source_entry = from_data.get(target_entry.file_id,
897
result, changes = self._changes_from_entries(source_entry,
898
target_entry, source_path=source_path, target_path=target_path)
899
to_paths[result[0]] = result[1][1]
728
(None, to_executable))
729
file_id = to_entry.file_id
730
to_paths[file_id] = to_path
732
changed_content = False
733
from_path, from_entry = from_data.get(file_id, (None, None))
734
from_versioned = (from_entry is not None)
735
if from_entry is not None:
736
from_versioned = True
737
from_name = from_entry.name
738
from_parent = from_entry.parent_id
739
from_kind, from_executable, from_stat = \
740
self.source._comparison_data(from_entry, from_path)
743
from_versioned = False
747
from_executable = None
748
versioned = (from_versioned, True)
749
to_kind, to_executable, to_stat = \
750
self.target._comparison_data(to_entry, to_path)
751
kind = (from_kind, to_kind)
752
if kind[0] != kind[1]:
753
changed_content = True
754
elif from_kind == 'file':
755
from_size = self.source._file_size(from_entry, from_stat)
756
to_size = self.target._file_size(to_entry, to_stat)
757
if from_size != to_size:
758
changed_content = True
759
elif (self.source.get_file_sha1(file_id, from_path, from_stat) !=
760
self.target.get_file_sha1(file_id, to_path, to_stat)):
761
changed_content = True
762
elif from_kind == 'symlink':
763
if (self.source.get_symlink_target(file_id) !=
764
self.target.get_symlink_target(file_id)):
765
changed_content = True
766
elif from_kind == 'tree-reference':
767
if (self.source.get_reference_revision(file_id, from_path)
768
!= self.target.get_reference_revision(file_id, to_path)):
769
changed_content = True
770
parent = (from_parent, to_entry.parent_id)
771
name = (from_name, to_entry.name)
772
executable = (from_executable, to_executable)
903
773
if pb is not None:
904
774
pb.update('comparing files', entry_count, num_entries)
905
if changes or include_unchanged:
906
if specific_files is not None:
907
new_parent_id = result[4][1]
908
precise_file_ids.add(new_parent_id)
909
changed_file_ids.append(result[0])
911
# Ensure correct behaviour for reparented/added specific files.
912
if specific_files is not None:
914
if result[6][1] == 'directory':
915
seen_dirs.add(result[0])
916
# Record parents of reparented/added entries.
917
versioned = result[3]
919
if not versioned[0] or parents[0] != parents[1]:
920
seen_parents.add(parents[1])
775
if (changed_content is not False or versioned[0] != versioned[1]
776
or parent[0] != parent[1] or name[0] != name[1] or
777
executable[0] != executable[1] or include_unchanged):
778
yield (file_id, (from_path, to_path), changed_content,
779
versioned, parent, name, kind, executable)
921
781
while all_unversioned:
922
782
# yield any trailing unversioned paths
923
783
unversioned_path = all_unversioned.popleft()
945
822
self.source._comparison_data(from_entry, path)
946
823
kind = (from_kind, None)
947
824
executable = (from_executable, None)
948
changed_content = from_kind is not None
825
changed_content = True
949
826
# the parent's path is necessarily known at this point.
950
changed_file_ids.append(file_id)
951
827
yield(file_id, (path, to_path), changed_content, versioned, parent,
952
828
name, kind, executable)
953
changed_file_ids = set(changed_file_ids)
954
if specific_files is not None:
955
for result in self._handle_precise_ids(precise_file_ids,
959
def _get_entry(self, tree, path):
960
"""Get an inventory entry from a tree, with missing entries as None.
962
If the tree raises NotImplementedError on accessing .inventory, then
963
this is worked around using iter_entries_by_dir on just the file id
966
:param tree: The tree to lookup the entry in.
967
:param file_id: The file_id to lookup.
969
# No inventory available.
971
iterator = tree.iter_entries_by_dir(specific_files=[path])
972
return next(iterator)[1]
973
except StopIteration:
976
def _handle_precise_ids(self, precise_file_ids, changed_file_ids,
977
discarded_changes=None):
978
"""Fill out a partial iter_changes to be consistent.
980
:param precise_file_ids: The file ids of parents that were seen during
982
:param changed_file_ids: The file ids of already emitted items.
983
:param discarded_changes: An optional dict of precalculated
984
iter_changes items which the partial iter_changes had not output
986
:return: A generator of iter_changes items to output.
988
# process parents of things that had changed under the users
989
# requested paths to prevent incorrect paths or parent ids which
990
# aren't in the tree.
991
while precise_file_ids:
992
precise_file_ids.discard(None)
993
# Don't emit file_ids twice
994
precise_file_ids.difference_update(changed_file_ids)
995
if not precise_file_ids:
997
# If the there was something at a given output path in source, we
998
# have to include the entry from source in the delta, or we would
999
# be putting this entry into a used path.
1001
for parent_id in precise_file_ids:
1003
paths.append(self.target.id2path(parent_id))
1004
except errors.NoSuchId:
1005
# This id has been dragged in from the source by delta
1006
# expansion and isn't present in target at all: we don't
1007
# need to check for path collisions on it.
1010
old_id = self.source.path2id(path)
1011
precise_file_ids.add(old_id)
1012
precise_file_ids.discard(None)
1013
current_ids = precise_file_ids
1014
precise_file_ids = set()
1015
# We have to emit all of precise_file_ids that have been altered.
1016
# We may have to output the children of some of those ids if any
1017
# directories have stopped being directories.
1018
for file_id in current_ids:
1020
if discarded_changes:
1021
result = discarded_changes.get(file_id)
1027
source_path = self.source.id2path(file_id)
1028
except errors.NoSuchId:
1032
source_entry = self._get_entry(self.source, source_path)
1034
target_path = self.target.id2path(file_id)
1035
except errors.NoSuchId:
1039
target_entry = self._get_entry(self.target, target_path)
1040
result, changes = self._changes_from_entries(
1041
source_entry, target_entry, source_path, target_path)
1044
# Get this parents parent to examine.
1045
new_parent_id = result[4][1]
1046
precise_file_ids.add(new_parent_id)
1048
if (result[6][0] == 'directory' and
1049
result[6][1] != 'directory'):
1050
# This stopped being a directory, the old children have
1052
if source_entry is None:
1053
# Reusing a discarded change.
1054
source_entry = self._get_entry(self.source, result[1][0])
1055
precise_file_ids.update(
1057
for child in self.source.iter_child_entries(result[1][0]))
1058
changed_file_ids.add(result[0])
1061
def file_content_matches(
1062
self, source_path, target_path,
1063
source_file_id=None, target_file_id=None,
1064
source_stat=None, target_stat=None):
1065
"""Check if two files are the same in the source and target trees.
1067
This only checks that the contents of the files are the same,
1068
it does not touch anything else.
1070
:param source_path: Path of the file in the source tree
1071
:param target_path: Path of the file in the target tree
1072
:param source_file_id: Optional file id of the file in the source tree
1073
:param target_file_id: Optional file id of the file in the target tree
1074
:param source_stat: Optional stat value of the file in the source tree
1075
:param target_stat: Optional stat value of the file in the target tree
1076
:return: Boolean indicating whether the files have the same contents
1078
with self.lock_read():
1079
source_verifier_kind, source_verifier_data = (
1080
self.source.get_file_verifier(
1081
source_path, source_file_id, source_stat))
1082
target_verifier_kind, target_verifier_data = (
1083
self.target.get_file_verifier(
1084
target_path, target_file_id, target_stat))
1085
if source_verifier_kind == target_verifier_kind:
1086
return (source_verifier_data == target_verifier_data)
1087
# Fall back to SHA1 for now
1088
if source_verifier_kind != "SHA1":
1089
source_sha1 = self.source.get_file_sha1(
1090
source_path, source_file_id, source_stat)
1092
source_sha1 = source_verifier_data
1093
if target_verifier_kind != "SHA1":
1094
target_sha1 = self.target.get_file_sha1(
1095
target_path, target_file_id, target_stat)
1097
target_sha1 = target_verifier_data
1098
return (source_sha1 == target_sha1)
1100
InterTree.register_optimiser(InterTree)
1103
class MultiWalker(object):
1104
"""Walk multiple trees simultaneously, getting combined results."""
1106
# Note: This could be written to not assume you can do out-of-order
1107
# lookups. Instead any nodes that don't match in all trees could be
1108
# marked as 'deferred', and then returned in the final cleanup loop.
1109
# For now, I think it is "nicer" to return things as close to the
1110
# "master_tree" order as we can.
1112
def __init__(self, master_tree, other_trees):
1113
"""Create a new MultiWalker.
1115
All trees being walked must implement "iter_entries_by_dir()", such
1116
that they yield (path, object) tuples, where that object will have a
1117
'.file_id' member, that can be used to check equality.
1119
:param master_tree: All trees will be 'slaved' to the master_tree such
1120
that nodes in master_tree will be used as 'first-pass' sync points.
1121
Any nodes that aren't in master_tree will be merged in a second
1123
:param other_trees: A list of other trees to walk simultaneously.
1125
self._master_tree = master_tree
1126
self._other_trees = other_trees
1128
# Keep track of any nodes that were properly processed just out of
1129
# order, that way we don't return them at the end, we don't have to
1130
# track *all* processed file_ids, just the out-of-order ones
1131
self._out_of_order_processed = set()
1134
def _step_one(iterator):
1135
"""Step an iter_entries_by_dir iterator.
1137
:return: (has_more, path, ie)
1138
If has_more is False, path and ie will be None.
1141
path, ie = next(iterator)
1142
except StopIteration:
1143
return False, None, None
1145
return True, path, ie
1148
def _cmp_path_by_dirblock(path1, path2):
1149
"""Compare two paths based on what directory they are in.
1151
This generates a sort order, such that all children of a directory are
1152
sorted together, and grandchildren are in the same order as the
1153
children appear. But all grandchildren come after all children.
1155
:param path1: first path
1156
:param path2: the second path
1157
:return: negative number if ``path1`` comes first,
1158
0 if paths are equal
1159
and a positive number if ``path2`` sorts first
1161
# Shortcut this special case
1164
# This is stolen from _dirstate_helpers_py.py, only switching it to
1165
# Unicode objects. Consider using encode_utf8() and then using the
1166
# optimized versions, or maybe writing optimized unicode versions.
1167
if not isinstance(path1, text_type):
1168
raise TypeError("'path1' must be a unicode string, not %s: %r"
1169
% (type(path1), path1))
1170
if not isinstance(path2, text_type):
1171
raise TypeError("'path2' must be a unicode string, not %s: %r"
1172
% (type(path2), path2))
1173
return cmp(MultiWalker._path_to_key(path1),
1174
MultiWalker._path_to_key(path2))
1177
def _path_to_key(path):
1178
dirname, basename = osutils.split(path)
1179
return (dirname.split(u'/'), basename)
1181
def _lookup_by_file_id(self, extra_entries, other_tree, file_id):
1182
"""Lookup an inventory entry by file_id.
1184
This is called when an entry is missing in the normal order.
1185
Generally this is because a file was either renamed, or it was
1186
deleted/added. If the entry was found in the inventory and not in
1187
extra_entries, it will be added to self._out_of_order_processed
1189
:param extra_entries: A dictionary of {file_id: (path, ie)}. This
1190
should be filled with entries that were found before they were
1191
used. If file_id is present, it will be removed from the
1193
:param other_tree: The Tree to search, in case we didn't find the entry
1195
:param file_id: The file_id to look for
1196
:return: (path, ie) if found or (None, None) if not present.
1198
if file_id in extra_entries:
1199
return extra_entries.pop(file_id)
1200
# TODO: Is id2path better as the first call, or is
1201
# inventory[file_id] better as a first check?
1203
cur_path = other_tree.id2path(file_id)
1204
except errors.NoSuchId:
1206
if cur_path is None:
1209
self._out_of_order_processed.add(file_id)
1210
cur_ie = other_tree.root_inventory.get_entry(file_id)
1211
return (cur_path, cur_ie)
1214
"""Match up the values in the different trees."""
1215
for result in self._walk_master_tree():
1217
self._finish_others()
1218
for result in self._walk_others():
1221
def _walk_master_tree(self):
1222
"""First pass, walk all trees in lock-step.
1224
When we are done, all nodes in the master_tree will have been
1225
processed. _other_walkers, _other_entries, and _others_extra will be
1226
set on 'self' for future processing.
1228
# This iterator has the most "inlining" done, because it tends to touch
1229
# every file in the tree, while the others only hit nodes that don't
1231
master_iterator = self._master_tree.iter_entries_by_dir()
1233
other_walkers = [other.iter_entries_by_dir()
1234
for other in self._other_trees]
1235
other_entries = [self._step_one(walker) for walker in other_walkers]
1236
# Track extra nodes in the other trees
1237
others_extra = [{} for _ in range(len(self._other_trees))]
1239
master_has_more = True
1240
step_one = self._step_one
1241
lookup_by_file_id = self._lookup_by_file_id
1242
out_of_order_processed = self._out_of_order_processed
1244
while master_has_more:
1245
(master_has_more, path, master_ie) = step_one(master_iterator)
1246
if not master_has_more:
1249
file_id = master_ie.file_id
1251
other_values_append = other_values.append
1252
next_other_entries = []
1253
next_other_entries_append = next_other_entries.append
1254
for idx, (other_has_more, other_path, other_ie) in enumerate(other_entries):
1255
if not other_has_more:
1256
other_values_append(lookup_by_file_id(
1257
others_extra[idx], self._other_trees[idx], file_id))
1258
next_other_entries_append((False, None, None))
1259
elif file_id == other_ie.file_id:
1260
# This is the critical code path, as most of the entries
1261
# should match between most trees.
1262
other_values_append((other_path, other_ie))
1263
next_other_entries_append(step_one(other_walkers[idx]))
1265
# This walker did not match, step it until it either
1266
# matches, or we know we are past the current walker.
1267
other_walker = other_walkers[idx]
1268
other_extra = others_extra[idx]
1269
while (other_has_more and
1270
self._cmp_path_by_dirblock(other_path, path) < 0):
1271
other_file_id = other_ie.file_id
1272
if other_file_id not in out_of_order_processed:
1273
other_extra[other_file_id] = (other_path, other_ie)
1274
other_has_more, other_path, other_ie = \
1275
step_one(other_walker)
1276
if other_has_more and other_ie.file_id == file_id:
1277
# We ended up walking to this point, match and step
1279
other_values_append((other_path, other_ie))
1280
other_has_more, other_path, other_ie = \
1281
step_one(other_walker)
1283
# This record isn't in the normal order, see if it
1285
other_values_append(lookup_by_file_id(
1286
other_extra, self._other_trees[idx], file_id))
1287
next_other_entries_append((other_has_more, other_path,
1289
other_entries = next_other_entries
1291
# We've matched all the walkers, yield this datapoint
1292
yield path, file_id, master_ie, other_values
1293
self._other_walkers = other_walkers
1294
self._other_entries = other_entries
1295
self._others_extra = others_extra
1297
def _finish_others(self):
1298
"""Finish walking the other iterators, so we get all entries."""
1299
for idx, info in enumerate(self._other_entries):
1300
other_extra = self._others_extra[idx]
1301
(other_has_more, other_path, other_ie) = info
1302
while other_has_more:
1303
other_file_id = other_ie.file_id
1304
if other_file_id not in self._out_of_order_processed:
1305
other_extra[other_file_id] = (other_path, other_ie)
1306
other_has_more, other_path, other_ie = \
1307
self._step_one(self._other_walkers[idx])
1308
del self._other_entries
1310
def _walk_others(self):
1311
"""Finish up by walking all the 'deferred' nodes."""
1312
# TODO: One alternative would be to grab all possible unprocessed
1313
# file_ids, and then sort by path, and then yield them. That
1314
# might ensure better ordering, in case a caller strictly
1315
# requires parents before children.
1316
for idx, other_extra in enumerate(self._others_extra):
1317
others = sorted(viewvalues(other_extra),
1318
key=lambda x: self._path_to_key(x[0]))
1319
for other_path, other_ie in others:
1320
file_id = other_ie.file_id
1321
# We don't need to check out_of_order_processed here, because
1322
# the lookup_by_file_id will be removing anything processed
1323
# from the extras cache
1324
other_extra.pop(file_id)
1325
other_values = [(None, None)] * idx
1326
other_values.append((other_path, other_ie))
1327
for alt_idx, alt_extra in enumerate(self._others_extra[idx+1:]):
1328
alt_idx = alt_idx + idx + 1
1329
alt_extra = self._others_extra[alt_idx]
1330
alt_tree = self._other_trees[alt_idx]
1331
other_values.append(self._lookup_by_file_id(
1332
alt_extra, alt_tree, file_id))
1333
yield other_path, file_id, None, other_values
1336
def find_previous_paths(from_tree, to_tree, paths):
1337
"""Find previous tree paths.
1339
:param from_tree: From tree
1340
:param to_tree: To tree
1341
:param paths: Iterable over paths to search for
1342
:return: Dictionary mapping from from_tree paths to paths in to_tree, or
1343
None if there is no equivalent path.
1347
ret[path] = find_previous_path(from_tree, to_tree, path)
1351
def find_previous_path(from_tree, to_tree, path, file_id=None):
1352
"""Find previous tree path.
1354
:param from_tree: From tree
1355
:param to_tree: To tree
1356
:param path: Path to search for
1357
:return: path in to_tree, or None if there is no equivalent path.
1360
file_id = from_tree.path2id(path)
1362
raise errors.NoSuchFile(path)
1364
return to_tree.id2path(file_id)
1365
except errors.NoSuchId:
831
# This was deprecated before 0.12, but did not have an official warning
832
@symbol_versioning.deprecated_function(symbol_versioning.zero_twelve)
833
def RevisionTree(*args, **kwargs):
834
"""RevisionTree has moved to bzrlib.revisiontree.RevisionTree()
836
Accessing it as bzrlib.tree.RevisionTree has been deprecated as of
839
from bzrlib.revisiontree import RevisionTree as _RevisionTree
840
return _RevisionTree(*args, **kwargs)
added
removed
bzrlib/tree.py
