« back to all changes in this revision
Viewing changes to bzrlib/tree.py
- browse files at revision 4570.2.3
- compare with another revision
- compare with revision 489
- download diff
- download tarball
- view history from revision 4570.2.3
- Committer: Robert Collins
- Date: 2009-07-29 04:57:50 UTC
- mto: This revision was merged to the branch mainline in revision 4649.
- Revision ID: robertc@robertcollins.net-20090729045750-qp1iw48rp7jyrfh2
Change the way iter_changes treats specific files to prevent InconsistentDeltas.
This is the internal core code that handles specific file operations like ``bzr st
FILENAME`` or ``bzr commit FILENAME``, and has been changed to include the
parent directories if they have altered, and when a directory stops being
a directory its children are always included. This fixes a number of
causes for ``InconsistentDelta`` errors, and permits faster commit of
specific paths. (Robert Collins, #347649)
This is the internal core code that handles specific file operations like ``bzr st
FILENAME`` or ``bzr commit FILENAME``, and has been changed to include the
parent directories if they have altered, and when a directory stops being
a directory its children are always included. This fixes a number of
causes for ``InconsistentDelta`` errors, and permits faster commit of
specific paths. (Robert Collins, #347649)
added
removed
bzrlib/tree.py
170
170
return self.bzrdir.is_control_filename(filename)
171
171
172
172
@needs_read_lock
173
def iter_entries_by_dir(self, specific_file_ids=None):
173
def iter_entries_by_dir(self, specific_file_ids=None, yield_parents=False):
174
174
"""Walk the tree in 'by_dir' order.
175
175
176
176
This will yield each entry in the tree as a (path, entry) tuple.
193
193
194
194
The yield order (ignoring root) would be::
195
195
a, f, a/b, a/d, a/b/c, a/d/e, f/g
196
197
:param yield_parents: If True, yield the parents from the root leading
198
down to specific_file_ids that have been requested. This has no
199
impact if specific_file_ids is None.
196
200
"""
197
201
return self.inventory.iter_entries_by_dir(
198
specific_file_ids=specific_file_ids)
202
specific_file_ids=specific_file_ids, yield_parents=yield_parents)
199
203
200
204
def iter_references(self):
201
205
if self.supports_tree_reference():
846
850
847
851
_optimisers = []
848
852
853
def _changes_from_entries(self, source_entry, target_entry,
854
source_path=None, target_path=None):
855
"""Generate a iter_changes tuple between source_entry and target_entry.
856
857
:param source_entry: An inventory entry from self.source, or None.
858
:param target_entry: An inventory entry from self.target, or None.
859
:param source_path: The path of source_entry, if known. If not known
860
it will be looked up.
861
:param target_path: The path of target_entry, if known. If not known
862
it will be looked up.
863
:return: A tuple, item 0 of which is an iter_changes result tuple, and
864
item 1 is True if there are any changes in the result tuple.
865
"""
866
if source_entry is None:
867
if target_entry is None:
868
return None
869
file_id = target_entry.file_id
870
else:
871
file_id = source_entry.file_id
872
if source_entry is not None:
873
source_versioned = True
874
source_name = source_entry.name
875
source_parent = source_entry.parent_id
876
if source_path is None:
877
source_path = self.source.id2path(file_id)
878
source_kind, source_executable, source_stat = \
879
self.source._comparison_data(source_entry, source_path)
880
else:
881
source_versioned = False
882
source_name = None
883
source_parent = None
884
source_kind = None
885
source_executable = None
886
if target_entry is not None:
887
target_versioned = True
888
target_name = target_entry.name
889
target_parent = target_entry.parent_id
890
if target_path is None:
891
target_path = self.target.id2path(file_id)
892
target_kind, target_executable, target_stat = \
893
self.target._comparison_data(target_entry, target_path)
894
else:
895
target_versioned = False
896
target_name = None
897
target_parent = None
898
target_kind = None
899
target_executable = None
900
versioned = (source_versioned, target_versioned)
901
kind = (source_kind, target_kind)
902
changed_content = False
903
if source_kind != target_kind:
904
changed_content = True
905
elif source_kind == 'file':
906
if (self.source.get_file_sha1(file_id, source_path, source_stat) !=
907
self.target.get_file_sha1(file_id, target_path, target_stat)):
908
changed_content = True
909
elif source_kind == 'symlink':
910
if (self.source.get_symlink_target(file_id) !=
911
self.target.get_symlink_target(file_id)):
912
changed_content = True
913
# XXX: Yes, the indentation below is wrong. But fixing it broke
914
# test_merge.TestMergerEntriesLCAOnDisk.
915
# test_nested_tree_subtree_renamed_and_modified. We'll wait for
916
# the fix from bzr.dev -- vila 2009026
917
elif source_kind == 'tree-reference':
918
if (self.source.get_reference_revision(file_id, source_path)
919
!= self.target.get_reference_revision(file_id, target_path)):
920
changed_content = True
921
parent = (source_parent, target_parent)
922
name = (source_name, target_name)
923
executable = (source_executable, target_executable)
924
if (changed_content is not False or versioned[0] != versioned[1]
925
or parent[0] != parent[1] or name[0] != name[1] or
926
executable[0] != executable[1]):
927
changes = True
928
else:
929
changes = False
930
return (file_id, (source_path, target_path), changed_content,
931
versioned, parent, name, kind, executable), changes
932
849
933
@needs_read_lock
850
934
def compare(self, want_unchanged=False, specific_files=None,
851
935
extra_trees=None, require_versioned=False, include_root=False,
914
998
:param require_versioned: Raise errors.PathsNotVersionedError if a
915
999
path in the specific_files list is not versioned in one of
916
1000
source, target or extra_trees.
1001
:param specific_files: An optional list of file paths to restrict the
1002
comparison to. When mapping filenames to ids, all matches in all
1003
trees (including optional extra_trees) are used, and all children
1004
of matched directories are included. The parents in the target tree
1005
of the specific files up to and including the root of the tree are
1006
always evaluated for changes too.
917
1007
:param want_unversioned: Should unversioned files be returned in the
918
1008
output. An unversioned file is defined as one with (False, False)
919
1009
for the versioned pair.
921
1011
lookup_trees = [self.source]
922
1012
if extra_trees:
923
1013
lookup_trees.extend(extra_trees)
1014
# The ids of items we need to examine to insure delta consistency.
1015
precise_file_ids = set()
1016
changed_file_ids = []
924
1017
if specific_files == []:
925
1018
specific_file_ids = []
926
1019
else:
927
1020
specific_file_ids = self.target.paths2ids(specific_files,
928
1021
lookup_trees, require_versioned=require_versioned)
1022
if specific_files is not None:
1023
# reparented or added entries must have their parents included
1024
# so that valid deltas can be created. The seen_parents set
1025
# tracks the parents that we need to have.
1026
# The seen_dirs set tracks directory entries we've yielded.
1027
# After outputting version object in to_entries we set difference
1028
# the two seen sets and start checking parents.
1029
seen_parents = set()
1030
seen_dirs = set()
929
1031
if want_unversioned:
930
1032
all_unversioned = sorted([(p.split('/'), p) for p in
931
1033
self.target.extras()
946
1048
# can be extras. So the fake_entry is solely used to look up
947
1049
# executable it values when execute is not supported.
948
1050
fake_entry = InventoryFile('unused', 'unused', 'unused')
949
for to_path, to_entry in to_entries_by_dir:
950
while all_unversioned and all_unversioned[0][0] < to_path.split('/'):
1051
for target_path, target_entry in to_entries_by_dir:
1052
while (all_unversioned and
1053
all_unversioned[0][0] < target_path.split('/')):
951
1054
unversioned_path = all_unversioned.popleft()
952
to_kind, to_executable, to_stat = \
1055
target_kind, target_executable, target_stat = \
953
1056
self.target._comparison_data(fake_entry, unversioned_path[1])
954
1057
yield (None, (None, unversioned_path[1]), True, (False, False),
955
1058
(None, None),
956
1059
(None, unversioned_path[0][-1]),
957
(None, to_kind),
958
(None, to_executable))
959
file_id = to_entry.file_id
960
to_paths[file_id] = to_path
1060
(None, target_kind),
1061
(None, target_executable))
1062
source_path, source_entry = from_data.get(target_entry.file_id,
1063
(None, None))
1064
result, changes = self._changes_from_entries(source_entry,
1065
target_entry, source_path=source_path, target_path=target_path)
1066
to_paths[result[0]] = result[1][1]
961
1067
entry_count += 1
962
changed_content = False
963
from_path, from_entry = from_data.get(file_id, (None, None))
964
from_versioned = (from_entry is not None)
965
if from_entry is not None:
966
from_versioned = True
967
from_name = from_entry.name
968
from_parent = from_entry.parent_id
969
from_kind, from_executable, from_stat = \
970
self.source._comparison_data(from_entry, from_path)
1068
if result[3][0]:
971
1069
entry_count += 1
972
else:
973
from_versioned = False
974
from_kind = None
975
from_parent = None
976
from_name = None
977
from_executable = None
978
versioned = (from_versioned, True)
979
to_kind, to_executable, to_stat = \
980
self.target._comparison_data(to_entry, to_path)
981
kind = (from_kind, to_kind)
982
if kind[0] != kind[1]:
983
changed_content = True
984
elif from_kind == 'file':
985
if (self.source.get_file_sha1(file_id, from_path, from_stat) !=
986
self.target.get_file_sha1(file_id, to_path, to_stat)):
987
changed_content = True
988
elif from_kind == 'symlink':
989
if (self.source.get_symlink_target(file_id) !=
990
self.target.get_symlink_target(file_id)):
991
changed_content = True
992
# XXX: Yes, the indentation below is wrong. But fixing it broke
993
# test_merge.TestMergerEntriesLCAOnDisk.
994
# test_nested_tree_subtree_renamed_and_modified. We'll wait for
995
# the fix from bzr.dev -- vila 2009026
996
elif from_kind == 'tree-reference':
997
if (self.source.get_reference_revision(file_id, from_path)
998
!= self.target.get_reference_revision(file_id, to_path)):
999
changed_content = True
1000
parent = (from_parent, to_entry.parent_id)
1001
name = (from_name, to_entry.name)
1002
executable = (from_executable, to_executable)
1003
1070
if pb is not None:
1004
1071
pb.update('comparing files', entry_count, num_entries)
1005
if (changed_content is not False or versioned[0] != versioned[1]
1006
or parent[0] != parent[1] or name[0] != name[1] or
1007
executable[0] != executable[1] or include_unchanged):
1008
yield (file_id, (from_path, to_path), changed_content,
1009
versioned, parent, name, kind, executable)
1010
1072
if changes or include_unchanged:
1073
if specific_file_ids is not None:
1074
new_parent_id = result[4][1]
1075
precise_file_ids.add(new_parent_id)
1076
changed_file_ids.append(result[0])
1077
yield result
1078
# Ensure correct behaviour for reparented/added specific files.
1079
if specific_files is not None:
1080
# Record output dirs
1081
if result[6][1] == 'directory':
1082
seen_dirs.add(result[0])
1083
# Record parents of reparented/added entries.
1084
versioned = result[3]
1085
parents = result[4]
1086
if not versioned[0] or parents[0] != parents[1]:
1087
seen_parents.add(parents[1])
1011
1088
while all_unversioned:
1012
1089
# yield any trailing unversioned paths
1013
1090
unversioned_path = all_unversioned.popleft()
1018
1095
(None, unversioned_path[0][-1]),
1019
1096
(None, to_kind),
1020
1097
(None, to_executable))
1021
1022
def get_to_path(to_entry):
1023
if to_entry.parent_id is None:
1024
to_path = '' # the root
1025
else:
1026
if to_entry.parent_id not in to_paths:
1027
# recurse up
1028
return get_to_path(self.target.inventory[to_entry.parent_id])
1029
to_path = osutils.pathjoin(to_paths[to_entry.parent_id],
1030
to_entry.name)
1031
to_paths[to_entry.file_id] = to_path
1032
return to_path
1033
1098
# Yield all remaining source paths
1034
1099
for path, from_entry in from_entries_by_dir:
1035
1100
file_id = from_entry.file_id
1036
1101
if file_id in to_paths:
1037
1102
# already returned
1038
1103
continue
1039
if not file_id in self.target.all_file_ids():
1104
if file_id not in self.target.all_file_ids():
1040
1105
# common case - paths we have not emitted are not present in
1041
1106
# target.
1042
1107
to_path = None
1043
1108
else:
1044
to_path = get_to_path(self.target.inventory[file_id])
1109
to_path = self.target.id2path(file_id)
1045
1110
entry_count += 1
1046
1111
if pb is not None:
1047
1112
pb.update('comparing files', entry_count, num_entries)
1054
1119
executable = (from_executable, None)
1055
1120
changed_content = from_kind is not None
1056
1121
# the parent's path is necessarily known at this point.
1122
changed_file_ids.append(file_id)
1057
1123
yield(file_id, (path, to_path), changed_content, versioned, parent,
1058
1124
name, kind, executable)
1125
changed_file_ids = set(changed_file_ids)
1126
if specific_file_ids is not None:
1127
for result in self._handle_precise_ids(precise_file_ids,
1128
changed_file_ids):
1129
yield result
1130
1131
def _get_entry(self, tree, file_id):
1132
"""Get an inventory entry from a tree, with missing entries as None.
1133
1134
If the tree raises NotImplementedError on accessing .inventory, then
1135
this is worked around using iter_entries_by_dir on just the file id
1136
desired.
1137
1138
:param tree: The tree to lookup the entry in.
1139
:param file_id: The file_id to lookup.
1140
"""
1141
try:
1142
inventory = tree.inventory
1143
except NotImplementedError:
1144
# No inventory available.
1145
try:
1146
iterator = tree.iter_entries_by_dir(specific_file_ids=[file_id])
1147
return iterator.next()[1]
1148
except StopIteration:
1149
return None
1150
else:
1151
try:
1152
return inventory[file_id]
1153
except errors.NoSuchId:
1154
return None
1155
1156
def _handle_precise_ids(self, precise_file_ids, changed_file_ids,
1157
discarded_changes=None):
1158
"""Fill out a partial iter_changes to be consistent.
1159
1160
:param precise_file_ids: The file ids of parents that were seen during
1161
the iter_changes.
1162
:param changed_file_ids: The file ids of already emitted items.
1163
:param discarded_changes: An optional dict of precalculated
1164
iter_changes items which the partial iter_changes had not output
1165
but had calculated.
1166
:return: A generator of iter_changes items to output.
1167
"""
1168
# process parents of things that had changed under the users
1169
# requested paths to prevent incorrect paths or parent ids which
1170
# aren't in the tree.
1171
while precise_file_ids:
1172
precise_file_ids.discard(None)
1173
# Don't emit file_ids twice
1174
precise_file_ids.difference_update(changed_file_ids)
1175
if not precise_file_ids:
1176
break
1177
# If the there was something at a given output path in source, we
1178
# have to include the entry from source in the delta, or we would
1179
# be putting this entry into a used path.
1180
paths = []
1181
for parent_id in precise_file_ids:
1182
try:
1183
paths.append(self.target.id2path(parent_id))
1184
except errors.NoSuchId:
1185
# This id has been dragged in from the source by delta
1186
# expansion and isn't present in target at all: we don't
1187
# need to check for path collisions on it.
1188
pass
1189
for path in paths:
1190
old_id = self.source.path2id(path)
1191
precise_file_ids.add(old_id)
1192
precise_file_ids.discard(None)
1193
current_ids = precise_file_ids
1194
precise_file_ids = set()
1195
# We have to emit all of precise_file_ids that have been altered.
1196
# We may have to output the children of some of those ids if any
1197
# directories have stopped being directories.
1198
for file_id in current_ids:
1199
# Examine file_id
1200
if discarded_changes:
1201
result = discarded_changes.get(file_id)
1202
old_entry = None
1203
else:
1204
result = None
1205
if result is None:
1206
old_entry = self._get_entry(self.source, file_id)
1207
new_entry = self._get_entry(self.target, file_id)
1208
result, changes = self._changes_from_entries(
1209
old_entry, new_entry)
1210
else:
1211
changes = True
1212
if changes:
1213
# Get this parents parent.
1214
new_parent_id = result[4][1]
1215
precise_file_ids.add(new_parent_id)
1216
if (result[6][0] == 'directory' and
1217
result[6][1] != 'directory'):
1218
# This stopped being a directory, the old children have
1219
# to be included.
1220
if old_entry is None:
1221
# Reusing a discarded change.
1222
old_entry = self._get_entry(self.source, file_id)
1223
for child in old_entry.children.values():
1224
precise_file_ids.add(child.file_id)
1225
changed_file_ids.add(result[0])
1226
yield result
1059
1227
1060
1228
1061
1229
class MultiWalker(object):
Loggerhead is a web-based interface for Breezy
Version: 2.0.1