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., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
15
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
17
17
"""Tree classes, representing directory at point in time.
21
21
from collections import deque
22
from cStringIO import StringIO
25
24
from bzrlib import (
26
25
conflicts as _mod_conflicts,
29
30
revision as _mod_revision,
33
33
from bzrlib.decorators import needs_read_lock
34
from bzrlib.errors import BzrError, BzrCheckError
34
from bzrlib.errors import BzrError, NoSuchId
35
35
from bzrlib import errors
36
from bzrlib.inventory import Inventory, InventoryFile
36
from bzrlib.inventory import InventoryFile
37
37
from bzrlib.inter import InterObject
38
38
from bzrlib.osutils import fingerprint_file
39
39
import bzrlib.revision
40
from bzrlib.trace import mutter, note
40
from bzrlib.symbol_versioning import deprecated_function, deprecated_in
41
from bzrlib.trace import note
43
44
class Tree(object):
44
45
"""Abstract file tree.
46
47
There are several subclasses:
48
49
* `WorkingTree` exists as files on disk editable by the user.
50
51
* `RevisionTree` is a tree as recorded at some point in the past.
94
95
want_unversioned=want_unversioned,
97
@symbol_versioning.deprecated_method(symbol_versioning.one_three)
98
def _iter_changes(self, *args, **kwargs):
99
return self.iter_changes(*args, **kwargs)
101
98
def iter_changes(self, from_tree, include_unchanged=False,
102
99
specific_files=None, pb=None, extra_trees=None,
103
100
require_versioned=True, want_unversioned=False):
101
"""See InterTree.iter_changes"""
104
102
intertree = InterTree.get(from_tree, self)
105
103
return intertree.iter_changes(include_unchanged, specific_files, pb,
106
104
extra_trees, require_versioned, want_unversioned=want_unversioned)
108
106
def conflicts(self):
109
107
"""Get a list of the conflicts in the tree.
198
195
The yield order (ignoring root) would be::
199
196
a, f, a/b, a/d, a/b/c, a/d/e, f/g
198
:param yield_parents: If True, yield the parents from the root leading
199
down to specific_file_ids that have been requested. This has no
200
impact if specific_file_ids is None.
201
202
return self.inventory.iter_entries_by_dir(
202
specific_file_ids=specific_file_ids)
203
specific_file_ids=specific_file_ids, yield_parents=yield_parents)
204
205
def iter_references(self):
205
for path, entry in self.iter_entries_by_dir():
206
if entry.kind == 'tree-reference':
207
yield path, entry.file_id
206
if self.supports_tree_reference():
207
for path, entry in self.iter_entries_by_dir():
208
if entry.kind == 'tree-reference':
209
yield path, entry.file_id
209
211
def kind(self, file_id):
210
212
raise NotImplementedError("Tree subclass %s must implement kind"
254
260
def _get_inventory(self):
255
261
return self._inventory
257
263
def get_file(self, file_id, path=None):
258
264
"""Return a file object for the file file_id in the tree.
260
266
If both file_id and path are defined, it is implementation defined as
261
267
to which one is used.
263
269
raise NotImplementedError(self.get_file)
271
def get_file_with_stat(self, file_id, path=None):
272
"""Get a file handle and stat object for file_id.
274
The default implementation returns (self.get_file, None) for backwards
277
:param file_id: The file id to read.
278
:param path: The path of the file, if it is known.
279
:return: A tuple (file_handle, stat_value_or_None). If the tree has
280
no stat facility, or need for a stat cache feedback during commit,
281
it may return None for the second element of the tuple.
283
return (self.get_file(file_id, path), None)
285
def get_file_text(self, file_id, path=None):
286
"""Return the byte content of a file.
288
:param file_id: The file_id of the file.
289
:param path: The path of the file.
290
If both file_id and path are supplied, an implementation may use
293
my_file = self.get_file(file_id, path)
295
return my_file.read()
299
def get_file_lines(self, file_id, path=None):
300
"""Return the content of a file, as lines.
302
:param file_id: The file_id of the file.
303
:param path: The path of the file.
304
If both file_id and path are supplied, an implementation may use
307
return osutils.split_lines(self.get_file_text(file_id, path))
265
309
def get_file_mtime(self, file_id, path=None):
266
310
"""Return the modification time for a file.
320
364
raise NotImplementedError(self.get_symlink_target)
366
def get_canonical_inventory_paths(self, paths):
367
"""Like get_canonical_inventory_path() but works on multiple items.
369
:param paths: A sequence of paths relative to the root of the tree.
370
:return: A list of paths, with each item the corresponding input path
371
adjusted to account for existing elements that match case
374
return list(self._yield_canonical_inventory_paths(paths))
376
def get_canonical_inventory_path(self, path):
377
"""Returns the first inventory item that case-insensitively matches path.
379
If a path matches exactly, it is returned. If no path matches exactly
380
but more than one path matches case-insensitively, it is implementation
381
defined which is returned.
383
If no path matches case-insensitively, the input path is returned, but
384
with as many path entries that do exist changed to their canonical
387
If you need to resolve many names from the same tree, you should
388
use get_canonical_inventory_paths() to avoid O(N) behaviour.
390
:param path: A paths relative to the root of the tree.
391
:return: The input path adjusted to account for existing elements
392
that match case insensitively.
394
return self._yield_canonical_inventory_paths([path]).next()
396
def _yield_canonical_inventory_paths(self, paths):
398
# First, if the path as specified exists exactly, just use it.
399
if self.path2id(path) is not None:
403
cur_id = self.get_root_id()
405
bit_iter = iter(path.split("/"))
409
for child in self.iter_children(cur_id):
411
# XXX: it seem like if the child is known to be in the
412
# tree, we shouldn't need to go from its id back to
413
# its path -- mbp 2010-02-11
415
# XXX: it seems like we could be more efficient
416
# by just directly looking up the original name and
417
# only then searching all children; also by not
418
# chopping paths so much. -- mbp 2010-02-11
419
child_base = os.path.basename(self.id2path(child))
420
if (child_base == elt):
421
# if we found an exact match, we can stop now; if
422
# we found an approximate match we need to keep
423
# searching because there might be an exact match
426
new_path = osutils.pathjoin(cur_path, child_base)
428
elif child_base.lower() == lelt:
430
new_path = osutils.pathjoin(cur_path, child_base)
432
# before a change is committed we can see this error...
437
# got to the end of this directory and no entries matched.
438
# Return what matched so far, plus the rest as specified.
439
cur_path = osutils.pathjoin(cur_path, elt, *list(bit_iter))
322
444
def get_root_id(self):
323
445
"""Return the file_id for the root of this tree."""
324
446
raise NotImplementedError(self.get_root_id)
534
655
raise NotImplementedError(self.walkdirs)
657
def supports_content_filtering(self):
660
def _content_filter_stack(self, path=None, file_id=None):
661
"""The stack of content filters for a path if filtering is supported.
663
Readers will be applied in first-to-last order.
664
Writers will be applied in last-to-first order.
665
Either the path or the file-id needs to be provided.
667
:param path: path relative to the root of the tree
669
:param file_id: file_id or None if unknown
670
:return: the list of filters - [] if there are none
672
filter_pref_names = filters._get_registered_names()
673
if len(filter_pref_names) == 0:
676
path = self.id2path(file_id)
677
prefs = self.iter_search_rules([path], filter_pref_names).next()
678
stk = filters._get_filter_stack_for(prefs)
679
if 'filters' in debug.debug_flags:
680
note("*** %s content-filter: %s => %r" % (path,prefs,stk))
683
def _content_filter_stack_provider(self):
684
"""A function that returns a stack of ContentFilters.
686
The function takes a path (relative to the top of the tree) and a
687
file-id as parameters.
689
:return: None if content filtering is not supported by this tree.
691
if self.supports_content_filtering():
692
return lambda path, file_id: \
693
self._content_filter_stack(path, file_id)
536
697
def iter_search_rules(self, path_names, pref_names=None,
537
_default_searcher=rules._per_user_searcher):
698
_default_searcher=None):
538
699
"""Find the preferences for filenames in a tree.
540
701
:param path_names: an iterable of paths to find attributes for.
553
716
for path in path_names:
554
717
yield searcher.get_items(path)
557
719
def _get_rules_searcher(self, default_searcher):
558
720
"""Get the RulesSearcher for this tree given the default one."""
559
721
searcher = default_searcher
563
class EmptyTree(Tree):
566
self._inventory = Inventory(root_id=None)
567
symbol_versioning.warn('EmptyTree is deprecated as of bzr 0.9 please'
568
' use repository.revision_tree instead.',
569
DeprecationWarning, stacklevel=2)
571
def get_parent_ids(self):
574
def get_symlink_target(self, file_id):
577
def has_filename(self, filename):
580
def kind(self, file_id):
583
def list_files(self, include_root=False):
586
def __contains__(self, file_id):
587
return (file_id in self._inventory)
589
def get_file_sha1(self, file_id, path=None, stat_value=None):
593
725
######################################################################
739
870
will pass through to InterTree as appropriate.
873
# Formats that will be used to test this InterTree. If both are
874
# None, this InterTree will not be tested (e.g. because a complex
876
_matching_from_tree_format = None
877
_matching_to_tree_format = None
881
def _changes_from_entries(self, source_entry, target_entry,
882
source_path=None, target_path=None):
883
"""Generate a iter_changes tuple between source_entry and target_entry.
885
:param source_entry: An inventory entry from self.source, or None.
886
:param target_entry: An inventory entry from self.target, or None.
887
:param source_path: The path of source_entry, if known. If not known
888
it will be looked up.
889
:param target_path: The path of target_entry, if known. If not known
890
it will be looked up.
891
:return: A tuple, item 0 of which is an iter_changes result tuple, and
892
item 1 is True if there are any changes in the result tuple.
894
if source_entry is None:
895
if target_entry is None:
897
file_id = target_entry.file_id
899
file_id = source_entry.file_id
900
if source_entry is not None:
901
source_versioned = True
902
source_name = source_entry.name
903
source_parent = source_entry.parent_id
904
if source_path is None:
905
source_path = self.source.id2path(file_id)
906
source_kind, source_executable, source_stat = \
907
self.source._comparison_data(source_entry, source_path)
909
source_versioned = False
913
source_executable = None
914
if target_entry is not None:
915
target_versioned = True
916
target_name = target_entry.name
917
target_parent = target_entry.parent_id
918
if target_path is None:
919
target_path = self.target.id2path(file_id)
920
target_kind, target_executable, target_stat = \
921
self.target._comparison_data(target_entry, target_path)
923
target_versioned = False
927
target_executable = None
928
versioned = (source_versioned, target_versioned)
929
kind = (source_kind, target_kind)
930
changed_content = False
931
if source_kind != target_kind:
932
changed_content = True
933
elif source_kind == 'file':
934
if (self.source.get_file_sha1(file_id, source_path, source_stat) !=
935
self.target.get_file_sha1(file_id, target_path, target_stat)):
936
changed_content = True
937
elif source_kind == 'symlink':
938
if (self.source.get_symlink_target(file_id) !=
939
self.target.get_symlink_target(file_id)):
940
changed_content = True
941
# XXX: Yes, the indentation below is wrong. But fixing it broke
942
# test_merge.TestMergerEntriesLCAOnDisk.
943
# test_nested_tree_subtree_renamed_and_modified. We'll wait for
944
# the fix from bzr.dev -- vila 2009026
945
elif source_kind == 'tree-reference':
946
if (self.source.get_reference_revision(file_id, source_path)
947
!= self.target.get_reference_revision(file_id, target_path)):
948
changed_content = True
949
parent = (source_parent, target_parent)
950
name = (source_name, target_name)
951
executable = (source_executable, target_executable)
952
if (changed_content is not False or versioned[0] != versioned[1]
953
or parent[0] != parent[1] or name[0] != name[1] or
954
executable[0] != executable[1]):
958
return (file_id, (source_path, target_path), changed_content,
959
versioned, parent, name, kind, executable), changes
745
962
def compare(self, want_unchanged=False, specific_files=None,
746
963
extra_trees=None, require_versioned=False, include_root=False,
809
1024
:param require_versioned: Raise errors.PathsNotVersionedError if a
810
1025
path in the specific_files list is not versioned in one of
811
1026
source, target or extra_trees.
1027
:param specific_files: An optional list of file paths to restrict the
1028
comparison to. When mapping filenames to ids, all matches in all
1029
trees (including optional extra_trees) are used, and all children
1030
of matched directories are included. The parents in the target tree
1031
of the specific files up to and including the root of the tree are
1032
always evaluated for changes too.
812
1033
:param want_unversioned: Should unversioned files be returned in the
813
1034
output. An unversioned file is defined as one with (False, False)
814
1035
for the versioned pair.
817
1037
lookup_trees = [self.source]
819
1039
lookup_trees.extend(extra_trees)
1040
# The ids of items we need to examine to insure delta consistency.
1041
precise_file_ids = set()
1042
changed_file_ids = []
820
1043
if specific_files == []:
821
1044
specific_file_ids = []
823
1046
specific_file_ids = self.target.paths2ids(specific_files,
824
1047
lookup_trees, require_versioned=require_versioned)
1048
if specific_files is not None:
1049
# reparented or added entries must have their parents included
1050
# so that valid deltas can be created. The seen_parents set
1051
# tracks the parents that we need to have.
1052
# The seen_dirs set tracks directory entries we've yielded.
1053
# After outputting version object in to_entries we set difference
1054
# the two seen sets and start checking parents.
1055
seen_parents = set()
825
1057
if want_unversioned:
826
1058
all_unversioned = sorted([(p.split('/'), p) for p in
827
1059
self.target.extras()
832
1064
all_unversioned = deque()
834
from_entries_by_dir = list(self.source.inventory.iter_entries_by_dir(
1066
from_entries_by_dir = list(self.source.iter_entries_by_dir(
835
1067
specific_file_ids=specific_file_ids))
836
1068
from_data = dict((e.file_id, (p, e)) for p, e in from_entries_by_dir)
837
to_entries_by_dir = list(self.target.inventory.iter_entries_by_dir(
1069
to_entries_by_dir = list(self.target.iter_entries_by_dir(
838
1070
specific_file_ids=specific_file_ids))
839
1071
num_entries = len(from_entries_by_dir) + len(to_entries_by_dir)
841
# the unversioned path lookup only occurs on real trees - where there
1073
# the unversioned path lookup only occurs on real trees - where there
842
1074
# can be extras. So the fake_entry is solely used to look up
843
1075
# executable it values when execute is not supported.
844
1076
fake_entry = InventoryFile('unused', 'unused', 'unused')
845
for to_path, to_entry in to_entries_by_dir:
846
while all_unversioned and all_unversioned[0][0] < to_path.split('/'):
1077
for target_path, target_entry in to_entries_by_dir:
1078
while (all_unversioned and
1079
all_unversioned[0][0] < target_path.split('/')):
847
1080
unversioned_path = all_unversioned.popleft()
848
to_kind, to_executable, to_stat = \
1081
target_kind, target_executable, target_stat = \
849
1082
self.target._comparison_data(fake_entry, unversioned_path[1])
850
1083
yield (None, (None, unversioned_path[1]), True, (False, False),
852
1085
(None, unversioned_path[0][-1]),
854
(None, to_executable))
855
file_id = to_entry.file_id
856
to_paths[file_id] = to_path
1086
(None, target_kind),
1087
(None, target_executable))
1088
source_path, source_entry = from_data.get(target_entry.file_id,
1090
result, changes = self._changes_from_entries(source_entry,
1091
target_entry, source_path=source_path, target_path=target_path)
1092
to_paths[result[0]] = result[1][1]
857
1093
entry_count += 1
858
changed_content = False
859
from_path, from_entry = from_data.get(file_id, (None, None))
860
from_versioned = (from_entry is not None)
861
if from_entry is not None:
862
from_versioned = True
863
from_name = from_entry.name
864
from_parent = from_entry.parent_id
865
from_kind, from_executable, from_stat = \
866
self.source._comparison_data(from_entry, from_path)
867
1095
entry_count += 1
869
from_versioned = False
873
from_executable = None
874
versioned = (from_versioned, True)
875
to_kind, to_executable, to_stat = \
876
self.target._comparison_data(to_entry, to_path)
877
kind = (from_kind, to_kind)
878
if kind[0] != kind[1]:
879
changed_content = True
880
elif from_kind == 'file':
881
from_size = self.source._file_size(from_entry, from_stat)
882
to_size = self.target._file_size(to_entry, to_stat)
883
if from_size != to_size:
884
changed_content = True
885
elif (self.source.get_file_sha1(file_id, from_path, from_stat) !=
886
self.target.get_file_sha1(file_id, to_path, to_stat)):
887
changed_content = True
888
elif from_kind == 'symlink':
889
if (self.source.get_symlink_target(file_id) !=
890
self.target.get_symlink_target(file_id)):
891
changed_content = True
892
elif from_kind == 'tree-reference':
893
if (self.source.get_reference_revision(file_id, from_path)
894
!= self.target.get_reference_revision(file_id, to_path)):
895
changed_content = True
896
parent = (from_parent, to_entry.parent_id)
897
name = (from_name, to_entry.name)
898
executable = (from_executable, to_executable)
899
1096
if pb is not None:
900
1097
pb.update('comparing files', entry_count, num_entries)
901
if (changed_content is not False or versioned[0] != versioned[1]
902
or parent[0] != parent[1] or name[0] != name[1] or
903
executable[0] != executable[1] or include_unchanged):
904
yield (file_id, (from_path, to_path), changed_content,
905
versioned, parent, name, kind, executable)
1098
if changes or include_unchanged:
1099
if specific_file_ids is not None:
1100
new_parent_id = result[4][1]
1101
precise_file_ids.add(new_parent_id)
1102
changed_file_ids.append(result[0])
1104
# Ensure correct behaviour for reparented/added specific files.
1105
if specific_files is not None:
1106
# Record output dirs
1107
if result[6][1] == 'directory':
1108
seen_dirs.add(result[0])
1109
# Record parents of reparented/added entries.
1110
versioned = result[3]
1112
if not versioned[0] or parents[0] != parents[1]:
1113
seen_parents.add(parents[1])
907
1114
while all_unversioned:
908
1115
# yield any trailing unversioned paths
909
1116
unversioned_path = all_unversioned.popleft()
914
1121
(None, unversioned_path[0][-1]),
915
1122
(None, to_kind),
916
1123
(None, to_executable))
918
def get_to_path(to_entry):
919
if to_entry.parent_id is None:
920
to_path = '' # the root
922
if to_entry.parent_id not in to_paths:
924
return get_to_path(self.target.inventory[to_entry.parent_id])
925
to_path = osutils.pathjoin(to_paths[to_entry.parent_id],
927
to_paths[to_entry.file_id] = to_path
1124
# Yield all remaining source paths
930
1125
for path, from_entry in from_entries_by_dir:
931
1126
file_id = from_entry.file_id
932
1127
if file_id in to_paths:
933
1128
# already returned
935
if not file_id in self.target.inventory:
1130
if file_id not in self.target.all_file_ids():
936
1131
# common case - paths we have not emitted are not present in
940
to_path = get_to_path(self.target.inventory[file_id])
1135
to_path = self.target.id2path(file_id)
941
1136
entry_count += 1
942
1137
if pb is not None:
943
1138
pb.update('comparing files', entry_count, num_entries)
950
1145
executable = (from_executable, None)
951
1146
changed_content = from_kind is not None
952
1147
# the parent's path is necessarily known at this point.
1148
changed_file_ids.append(file_id)
953
1149
yield(file_id, (path, to_path), changed_content, versioned, parent,
954
1150
name, kind, executable)
1151
changed_file_ids = set(changed_file_ids)
1152
if specific_file_ids is not None:
1153
for result in self._handle_precise_ids(precise_file_ids,
1157
def _get_entry(self, tree, file_id):
1158
"""Get an inventory entry from a tree, with missing entries as None.
1160
If the tree raises NotImplementedError on accessing .inventory, then
1161
this is worked around using iter_entries_by_dir on just the file id
1164
:param tree: The tree to lookup the entry in.
1165
:param file_id: The file_id to lookup.
1168
inventory = tree.inventory
1169
except NotImplementedError:
1170
# No inventory available.
1172
iterator = tree.iter_entries_by_dir(specific_file_ids=[file_id])
1173
return iterator.next()[1]
1174
except StopIteration:
1178
return inventory[file_id]
1179
except errors.NoSuchId:
1182
def _handle_precise_ids(self, precise_file_ids, changed_file_ids,
1183
discarded_changes=None):
1184
"""Fill out a partial iter_changes to be consistent.
1186
:param precise_file_ids: The file ids of parents that were seen during
1188
:param changed_file_ids: The file ids of already emitted items.
1189
:param discarded_changes: An optional dict of precalculated
1190
iter_changes items which the partial iter_changes had not output
1192
:return: A generator of iter_changes items to output.
1194
# process parents of things that had changed under the users
1195
# requested paths to prevent incorrect paths or parent ids which
1196
# aren't in the tree.
1197
while precise_file_ids:
1198
precise_file_ids.discard(None)
1199
# Don't emit file_ids twice
1200
precise_file_ids.difference_update(changed_file_ids)
1201
if not precise_file_ids:
1203
# If the there was something at a given output path in source, we
1204
# have to include the entry from source in the delta, or we would
1205
# be putting this entry into a used path.
1207
for parent_id in precise_file_ids:
1209
paths.append(self.target.id2path(parent_id))
1210
except errors.NoSuchId:
1211
# This id has been dragged in from the source by delta
1212
# expansion and isn't present in target at all: we don't
1213
# need to check for path collisions on it.
1216
old_id = self.source.path2id(path)
1217
precise_file_ids.add(old_id)
1218
precise_file_ids.discard(None)
1219
current_ids = precise_file_ids
1220
precise_file_ids = set()
1221
# We have to emit all of precise_file_ids that have been altered.
1222
# We may have to output the children of some of those ids if any
1223
# directories have stopped being directories.
1224
for file_id in current_ids:
1226
if discarded_changes:
1227
result = discarded_changes.get(file_id)
1232
old_entry = self._get_entry(self.source, file_id)
1233
new_entry = self._get_entry(self.target, file_id)
1234
result, changes = self._changes_from_entries(
1235
old_entry, new_entry)
1238
# Get this parents parent to examine.
1239
new_parent_id = result[4][1]
1240
precise_file_ids.add(new_parent_id)
1242
if (result[6][0] == 'directory' and
1243
result[6][1] != 'directory'):
1244
# This stopped being a directory, the old children have
1246
if old_entry is None:
1247
# Reusing a discarded change.
1248
old_entry = self._get_entry(self.source, file_id)
1249
for child in old_entry.children.values():
1250
precise_file_ids.add(child.file_id)
1251
changed_file_ids.add(result[0])
957
1255
class MultiWalker(object):
added
removed
bzrlib/tree.py
