771
784
# it works for all trees.
787
def _changes_from_entries(self, source_entry, target_entry, source_path,
789
"""Generate a iter_changes tuple between source_entry and target_entry.
791
:param source_entry: An inventory entry from self.source, or None.
792
:param target_entry: An inventory entry from self.target, or None.
793
:param source_path: The path of source_entry.
794
:param target_path: The path of target_entry.
795
:return: A tuple, item 0 of which is an iter_changes result tuple, and
796
item 1 is True if there are any changes in the result tuple.
798
if source_entry is None:
799
if target_entry is None:
801
file_id = target_entry.file_id
803
file_id = source_entry.file_id
804
if source_entry is not None:
805
source_versioned = True
806
source_name = source_entry.name
807
source_parent = source_entry.parent_id
808
source_kind, source_executable, source_stat = \
809
self.source._comparison_data(source_entry, source_path)
811
source_versioned = False
815
source_executable = None
816
if target_entry is not None:
817
target_versioned = True
818
target_name = target_entry.name
819
target_parent = target_entry.parent_id
820
target_kind, target_executable, target_stat = \
821
self.target._comparison_data(target_entry, target_path)
823
target_versioned = False
827
target_executable = None
828
versioned = (source_versioned, target_versioned)
829
kind = (source_kind, target_kind)
830
changed_content = False
831
if source_kind != target_kind:
832
changed_content = True
833
elif source_kind == 'file':
834
if not self.file_content_matches(
835
source_path, target_path,
836
source_stat, target_stat):
837
changed_content = True
838
elif source_kind == 'symlink':
839
if (self.source.get_symlink_target(source_path) !=
840
self.target.get_symlink_target(target_path)):
841
changed_content = True
842
elif source_kind == 'tree-reference':
843
if (self.source.get_reference_revision(source_path)
844
!= self.target.get_reference_revision(target_path)):
845
changed_content = True
846
parent = (source_parent, target_parent)
847
name = (source_name, target_name)
848
executable = (source_executable, target_executable)
849
if (changed_content is not False or versioned[0] != versioned[1] or
850
parent[0] != parent[1] or name[0] != name[1] or
851
executable[0] != executable[1]):
856
file_id, (source_path, target_path), changed_content,
857
versioned, parent, name, kind, executable), changes
774
859
def compare(self, want_unchanged=False, specific_files=None,
775
860
extra_trees=None, require_versioned=False, include_root=False,
776
861
want_unversioned=False):
836
920
output. An unversioned file is defined as one with (False, False)
837
921
for the versioned pair.
839
raise NotImplementedError(self.iter_changes)
926
extra_trees = list(extra_trees)
927
# The ids of items we need to examine to insure delta consistency.
928
precise_file_ids = set()
929
changed_file_ids = []
930
if specific_files == []:
931
target_specific_files = []
932
source_specific_files = []
934
target_specific_files = self.target.find_related_paths_across_trees(
935
specific_files, [self.source] + extra_trees,
936
require_versioned=require_versioned)
937
source_specific_files = self.source.find_related_paths_across_trees(
938
specific_files, [self.target] + extra_trees,
939
require_versioned=require_versioned)
940
if specific_files is not None:
941
# reparented or added entries must have their parents included
942
# so that valid deltas can be created. The seen_parents set
943
# tracks the parents that we need to have.
944
# The seen_dirs set tracks directory entries we've yielded.
945
# After outputting version object in to_entries we set difference
946
# the two seen sets and start checking parents.
950
all_unversioned = sorted([(p.split('/'), p) for p in
952
if specific_files is None or
953
osutils.is_inside_any(specific_files, p)])
954
all_unversioned = deque(all_unversioned)
956
all_unversioned = deque()
958
from_entries_by_dir = list(self.source.iter_entries_by_dir(
959
specific_files=source_specific_files))
960
from_data = dict(from_entries_by_dir)
961
to_entries_by_dir = list(self.target.iter_entries_by_dir(
962
specific_files=target_specific_files))
963
path_equivs = self.find_source_paths([p for p, e in to_entries_by_dir])
964
num_entries = len(from_entries_by_dir) + len(to_entries_by_dir)
966
# the unversioned path lookup only occurs on real trees - where there
967
# can be extras. So the fake_entry is solely used to look up
968
# executable it values when execute is not supported.
969
fake_entry = TreeFile()
970
for target_path, target_entry in to_entries_by_dir:
971
while (all_unversioned and
972
all_unversioned[0][0] < target_path.split('/')):
973
unversioned_path = all_unversioned.popleft()
974
target_kind, target_executable, target_stat = \
975
self.target._comparison_data(
976
fake_entry, unversioned_path[1])
978
None, (None, unversioned_path[1]), True, (False, False),
980
(None, unversioned_path[0][-1]),
982
(None, target_executable))
983
source_path = path_equivs[target_path]
984
if source_path is not None:
985
source_entry = from_data.get(source_path)
988
result, changes = self._changes_from_entries(
989
source_entry, target_entry, source_path=source_path, target_path=target_path)
990
to_paths[result.file_id] = result.path[1]
992
if result.versioned[0]:
995
pb.update('comparing files', entry_count, num_entries)
996
if changes or include_unchanged:
997
if specific_files is not None:
998
precise_file_ids.add(result.parent_id[1])
999
changed_file_ids.append(result.file_id)
1001
# Ensure correct behaviour for reparented/added specific files.
1002
if specific_files is not None:
1003
# Record output dirs
1004
if result.kind[1] == 'directory':
1005
seen_dirs.add(result.file_id)
1006
# Record parents of reparented/added entries.
1007
if not result.versioned[0] or result.is_reparented():
1008
seen_parents.add(result.parent_id[1])
1009
while all_unversioned:
1010
# yield any trailing unversioned paths
1011
unversioned_path = all_unversioned.popleft()
1012
to_kind, to_executable, to_stat = \
1013
self.target._comparison_data(fake_entry, unversioned_path[1])
1015
None, (None, unversioned_path[1]), True, (False, False),
1017
(None, unversioned_path[0][-1]),
1019
(None, to_executable))
1020
# Yield all remaining source paths
1021
for path, from_entry in from_entries_by_dir:
1022
file_id = from_entry.file_id
1023
if file_id in to_paths:
1026
to_path = self.find_target_path(path)
1029
pb.update('comparing files', entry_count, num_entries)
1030
versioned = (True, False)
1031
parent = (from_entry.parent_id, None)
1032
name = (from_entry.name, None)
1033
from_kind, from_executable, stat_value = \
1034
self.source._comparison_data(from_entry, path)
1035
kind = (from_kind, None)
1036
executable = (from_executable, None)
1037
changed_content = from_kind is not None
1038
# the parent's path is necessarily known at this point.
1039
changed_file_ids.append(file_id)
1041
file_id, (path, to_path), changed_content, versioned, parent,
1042
name, kind, executable)
1043
changed_file_ids = set(changed_file_ids)
1044
if specific_files is not None:
1045
for result in self._handle_precise_ids(precise_file_ids,
1050
def _get_entry(tree, path):
1051
"""Get an inventory entry from a tree, with missing entries as None.
1053
If the tree raises NotImplementedError on accessing .inventory, then
1054
this is worked around using iter_entries_by_dir on just the file id
1057
:param tree: The tree to lookup the entry in.
1058
:param path: The path to look up
1060
# No inventory available.
1062
iterator = tree.iter_entries_by_dir(specific_files=[path])
1063
return next(iterator)[1]
1064
except StopIteration:
1067
def _handle_precise_ids(self, precise_file_ids, changed_file_ids,
1068
discarded_changes=None):
1069
"""Fill out a partial iter_changes to be consistent.
1071
:param precise_file_ids: The file ids of parents that were seen during
1073
:param changed_file_ids: The file ids of already emitted items.
1074
:param discarded_changes: An optional dict of precalculated
1075
iter_changes items which the partial iter_changes had not output
1077
:return: A generator of iter_changes items to output.
1079
# process parents of things that had changed under the users
1080
# requested paths to prevent incorrect paths or parent ids which
1081
# aren't in the tree.
1082
while precise_file_ids:
1083
precise_file_ids.discard(None)
1084
# Don't emit file_ids twice
1085
precise_file_ids.difference_update(changed_file_ids)
1086
if not precise_file_ids:
1088
# If the there was something at a given output path in source, we
1089
# have to include the entry from source in the delta, or we would
1090
# be putting this entry into a used path.
1092
for parent_id in precise_file_ids:
1094
paths.append(self.target.id2path(parent_id))
1095
except errors.NoSuchId:
1096
# This id has been dragged in from the source by delta
1097
# expansion and isn't present in target at all: we don't
1098
# need to check for path collisions on it.
1101
old_id = self.source.path2id(path)
1102
precise_file_ids.add(old_id)
1103
precise_file_ids.discard(None)
1104
current_ids = precise_file_ids
1105
precise_file_ids = set()
1106
# We have to emit all of precise_file_ids that have been altered.
1107
# We may have to output the children of some of those ids if any
1108
# directories have stopped being directories.
1109
for file_id in current_ids:
1111
if discarded_changes:
1112
result = discarded_changes.get(file_id)
1118
source_path = self.source.id2path(file_id)
1119
except errors.NoSuchId:
1123
source_entry = self._get_entry(
1124
self.source, source_path)
1126
target_path = self.target.id2path(file_id)
1127
except errors.NoSuchId:
1131
target_entry = self._get_entry(
1132
self.target, target_path)
1133
result, changes = self._changes_from_entries(
1134
source_entry, target_entry, source_path, target_path)
1137
# Get this parents parent to examine.
1138
new_parent_id = result.parent_id[1]
1139
precise_file_ids.add(new_parent_id)
1141
if (result.kind[0] == 'directory' and
1142
result.kind[1] != 'directory'):
1143
# This stopped being a directory, the old children have
1145
if source_entry is None:
1146
# Reusing a discarded change.
1147
source_entry = self._get_entry(
1148
self.source, result.path[0])
1149
precise_file_ids.update(
1151
for child in self.source.iter_child_entries(result.path[0]))
1152
changed_file_ids.add(result.file_id)
841
1155
def file_content_matches(
842
1156
self, source_path, target_path,
added
removed
breezy/tree.py
