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.
21
21
from collections import deque
22
from cStringIO import StringIO
24
25
from bzrlib import (
25
26
conflicts as _mod_conflicts,
30
29
revision as _mod_revision,
33
33
from bzrlib.decorators import needs_read_lock
34
from bzrlib.errors import BzrError, NoSuchId
34
from bzrlib.errors import BzrError, BzrCheckError, NoSuchId
35
35
from bzrlib import errors
36
from bzrlib.inventory import InventoryFile
36
from bzrlib.inventory import Inventory, InventoryFile
37
37
from bzrlib.inter import InterObject
38
38
from bzrlib.osutils import fingerprint_file
39
39
import bzrlib.revision
40
40
from bzrlib.symbol_versioning import deprecated_function, deprecated_in
41
from bzrlib.trace import note
41
from bzrlib.trace import mutter, note
44
44
class Tree(object):
194
200
The yield order (ignoring root) would be::
195
201
a, f, a/b, a/d, a/b/c, a/d/e, f/g
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.
201
203
return self.inventory.iter_entries_by_dir(
202
specific_file_ids=specific_file_ids, yield_parents=yield_parents)
204
specific_file_ids=specific_file_ids)
204
206
def iter_references(self):
205
if self.supports_tree_reference():
206
for path, entry in self.iter_entries_by_dir():
207
if entry.kind == 'tree-reference':
208
yield path, entry.file_id
207
for path, entry in self.iter_entries_by_dir():
208
if entry.kind == 'tree-reference':
209
yield path, entry.file_id
210
211
def kind(self, file_id):
211
212
raise NotImplementedError("Tree subclass %s must implement kind"
222
223
def path_content_summary(self, path):
223
224
"""Get a summary of the information about path.
225
All the attributes returned are for the canonical form, not the
226
convenient form (if content filters are in use.)
228
226
:param path: A relative path within the tree.
229
227
:return: A tuple containing kind, size, exec, sha1-or-link.
230
228
Kind is always present (see tree.kind()).
231
size is present if kind is file and the size of the
232
canonical form can be cheaply determined, None otherwise.
229
size is present if kind is file, None otherwise.
233
230
exec is None unless kind is file and the platform supports the 'x'
235
232
sha1-or-link is the link target if kind is symlink, or the sha1 if
268
265
raise NotImplementedError(self.get_file)
270
def get_file_with_stat(self, file_id, path=None):
271
"""Get a file handle and stat object for file_id.
273
The default implementation returns (self.get_file, None) for backwards
276
:param file_id: The file id to read.
277
:param path: The path of the file, if it is known.
278
:return: A tuple (file_handle, stat_value_or_None). If the tree has
279
no stat facility, or need for a stat cache feedback during commit,
280
it may return None for the second element of the tuple.
282
return (self.get_file(file_id, path), None)
284
267
def get_file_text(self, file_id, path=None):
285
268
"""Return the byte content of a file.
638
627
def supports_content_filtering(self):
641
def _content_filter_stack(self, path=None, file_id=None):
642
"""The stack of content filters for a path if filtering is supported.
644
Readers will be applied in first-to-last order.
645
Writers will be applied in last-to-first order.
646
Either the path or the file-id needs to be provided.
648
:param path: path relative to the root of the tree
650
:param file_id: file_id or None if unknown
651
:return: the list of filters - [] if there are none
653
filter_pref_names = filters._get_registered_names()
654
if len(filter_pref_names) == 0:
657
path = self.id2path(file_id)
658
prefs = self.iter_search_rules([path], filter_pref_names).next()
659
stk = filters._get_filter_stack_for(prefs)
660
if 'filters' in debug.debug_flags:
661
note("*** %s content-filter: %s => %r" % (path,prefs,stk))
664
def _content_filter_stack_provider(self):
665
"""A function that returns a stack of ContentFilters.
667
The function takes a path (relative to the top of the tree) and a
668
file-id as parameters.
670
:return: None if content filtering is not supported by this tree.
672
if self.supports_content_filtering():
673
return lambda path, file_id: \
674
self._content_filter_stack(path, file_id)
678
630
def iter_search_rules(self, path_names, pref_names=None,
679
_default_searcher=None):
631
_default_searcher=rules._per_user_searcher):
680
632
"""Find the preferences for filenames in a tree.
682
634
:param path_names: an iterable of paths to find attributes for.
851
802
will pass through to InterTree as appropriate.
854
# Formats that will be used to test this InterTree. If both are
855
# None, this InterTree will not be tested (e.g. because a complex
857
_matching_from_tree_format = None
858
_matching_to_tree_format = None
862
def _changes_from_entries(self, source_entry, target_entry,
863
source_path=None, target_path=None):
864
"""Generate a iter_changes tuple between source_entry and target_entry.
866
:param source_entry: An inventory entry from self.source, or None.
867
:param target_entry: An inventory entry from self.target, or None.
868
:param source_path: The path of source_entry, if known. If not known
869
it will be looked up.
870
:param target_path: The path of target_entry, if known. If not known
871
it will be looked up.
872
:return: A tuple, item 0 of which is an iter_changes result tuple, and
873
item 1 is True if there are any changes in the result tuple.
875
if source_entry is None:
876
if target_entry is None:
878
file_id = target_entry.file_id
880
file_id = source_entry.file_id
881
if source_entry is not None:
882
source_versioned = True
883
source_name = source_entry.name
884
source_parent = source_entry.parent_id
885
if source_path is None:
886
source_path = self.source.id2path(file_id)
887
source_kind, source_executable, source_stat = \
888
self.source._comparison_data(source_entry, source_path)
890
source_versioned = False
894
source_executable = None
895
if target_entry is not None:
896
target_versioned = True
897
target_name = target_entry.name
898
target_parent = target_entry.parent_id
899
if target_path is None:
900
target_path = self.target.id2path(file_id)
901
target_kind, target_executable, target_stat = \
902
self.target._comparison_data(target_entry, target_path)
904
target_versioned = False
908
target_executable = None
909
versioned = (source_versioned, target_versioned)
910
kind = (source_kind, target_kind)
911
changed_content = False
912
if source_kind != target_kind:
913
changed_content = True
914
elif source_kind == 'file':
915
if (self.source.get_file_sha1(file_id, source_path, source_stat) !=
916
self.target.get_file_sha1(file_id, target_path, target_stat)):
917
changed_content = True
918
elif source_kind == 'symlink':
919
if (self.source.get_symlink_target(file_id) !=
920
self.target.get_symlink_target(file_id)):
921
changed_content = True
922
# XXX: Yes, the indentation below is wrong. But fixing it broke
923
# test_merge.TestMergerEntriesLCAOnDisk.
924
# test_nested_tree_subtree_renamed_and_modified. We'll wait for
925
# the fix from bzr.dev -- vila 2009026
926
elif source_kind == 'tree-reference':
927
if (self.source.get_reference_revision(file_id, source_path)
928
!= self.target.get_reference_revision(file_id, target_path)):
929
changed_content = True
930
parent = (source_parent, target_parent)
931
name = (source_name, target_name)
932
executable = (source_executable, target_executable)
933
if (changed_content is not False or versioned[0] != versioned[1]
934
or parent[0] != parent[1] or name[0] != name[1] or
935
executable[0] != executable[1]):
939
return (file_id, (source_path, target_path), changed_content,
940
versioned, parent, name, kind, executable), changes
943
808
def compare(self, want_unchanged=False, specific_files=None,
944
809
extra_trees=None, require_versioned=False, include_root=False,
1005
872
:param require_versioned: Raise errors.PathsNotVersionedError if a
1006
873
path in the specific_files list is not versioned in one of
1007
874
source, target or extra_trees.
1008
:param specific_files: An optional list of file paths to restrict the
1009
comparison to. When mapping filenames to ids, all matches in all
1010
trees (including optional extra_trees) are used, and all children
1011
of matched directories are included. The parents in the target tree
1012
of the specific files up to and including the root of the tree are
1013
always evaluated for changes too.
1014
875
:param want_unversioned: Should unversioned files be returned in the
1015
876
output. An unversioned file is defined as one with (False, False)
1016
877
for the versioned pair.
1018
880
lookup_trees = [self.source]
1020
882
lookup_trees.extend(extra_trees)
1021
# The ids of items we need to examine to insure delta consistency.
1022
precise_file_ids = set()
1023
changed_file_ids = []
1024
883
if specific_files == []:
1025
884
specific_file_ids = []
1027
886
specific_file_ids = self.target.paths2ids(specific_files,
1028
887
lookup_trees, require_versioned=require_versioned)
1029
if specific_files is not None:
1030
# reparented or added entries must have their parents included
1031
# so that valid deltas can be created. The seen_parents set
1032
# tracks the parents that we need to have.
1033
# The seen_dirs set tracks directory entries we've yielded.
1034
# After outputting version object in to_entries we set difference
1035
# the two seen sets and start checking parents.
1036
seen_parents = set()
1038
888
if want_unversioned:
1039
889
all_unversioned = sorted([(p.split('/'), p) for p in
1040
890
self.target.extras()
1055
905
# can be extras. So the fake_entry is solely used to look up
1056
906
# executable it values when execute is not supported.
1057
907
fake_entry = InventoryFile('unused', 'unused', 'unused')
1058
for target_path, target_entry in to_entries_by_dir:
1059
while (all_unversioned and
1060
all_unversioned[0][0] < target_path.split('/')):
908
for to_path, to_entry in to_entries_by_dir:
909
while all_unversioned and all_unversioned[0][0] < to_path.split('/'):
1061
910
unversioned_path = all_unversioned.popleft()
1062
target_kind, target_executable, target_stat = \
911
to_kind, to_executable, to_stat = \
1063
912
self.target._comparison_data(fake_entry, unversioned_path[1])
1064
913
yield (None, (None, unversioned_path[1]), True, (False, False),
1066
915
(None, unversioned_path[0][-1]),
1067
(None, target_kind),
1068
(None, target_executable))
1069
source_path, source_entry = from_data.get(target_entry.file_id,
1071
result, changes = self._changes_from_entries(source_entry,
1072
target_entry, source_path=source_path, target_path=target_path)
1073
to_paths[result[0]] = result[1][1]
917
(None, to_executable))
918
file_id = to_entry.file_id
919
to_paths[file_id] = to_path
1074
920
entry_count += 1
921
changed_content = False
922
from_path, from_entry = from_data.get(file_id, (None, None))
923
from_versioned = (from_entry is not None)
924
if from_entry is not None:
925
from_versioned = True
926
from_name = from_entry.name
927
from_parent = from_entry.parent_id
928
from_kind, from_executable, from_stat = \
929
self.source._comparison_data(from_entry, from_path)
1076
930
entry_count += 1
932
from_versioned = False
936
from_executable = None
937
versioned = (from_versioned, True)
938
to_kind, to_executable, to_stat = \
939
self.target._comparison_data(to_entry, to_path)
940
kind = (from_kind, to_kind)
941
if kind[0] != kind[1]:
942
changed_content = True
943
elif from_kind == 'file':
944
from_size = self.source._file_size(from_entry, from_stat)
945
to_size = self.target._file_size(to_entry, to_stat)
946
if from_size != to_size:
947
changed_content = True
948
elif (self.source.get_file_sha1(file_id, from_path, from_stat) !=
949
self.target.get_file_sha1(file_id, to_path, to_stat)):
950
changed_content = True
951
elif from_kind == 'symlink':
952
if (self.source.get_symlink_target(file_id) !=
953
self.target.get_symlink_target(file_id)):
954
changed_content = True
955
elif from_kind == 'tree-reference':
956
if (self.source.get_reference_revision(file_id, from_path)
957
!= self.target.get_reference_revision(file_id, to_path)):
958
changed_content = True
959
parent = (from_parent, to_entry.parent_id)
960
name = (from_name, to_entry.name)
961
executable = (from_executable, to_executable)
1077
962
if pb is not None:
1078
963
pb.update('comparing files', entry_count, num_entries)
1079
if changes or include_unchanged:
1080
if specific_file_ids is not None:
1081
new_parent_id = result[4][1]
1082
precise_file_ids.add(new_parent_id)
1083
changed_file_ids.append(result[0])
1085
# Ensure correct behaviour for reparented/added specific files.
1086
if specific_files is not None:
1087
# Record output dirs
1088
if result[6][1] == 'directory':
1089
seen_dirs.add(result[0])
1090
# Record parents of reparented/added entries.
1091
versioned = result[3]
1093
if not versioned[0] or parents[0] != parents[1]:
1094
seen_parents.add(parents[1])
964
if (changed_content is not False or versioned[0] != versioned[1]
965
or parent[0] != parent[1] or name[0] != name[1] or
966
executable[0] != executable[1] or include_unchanged):
967
yield (file_id, (from_path, to_path), changed_content,
968
versioned, parent, name, kind, executable)
1095
970
while all_unversioned:
1096
971
# yield any trailing unversioned paths
1097
972
unversioned_path = all_unversioned.popleft()
1102
977
(None, unversioned_path[0][-1]),
1103
978
(None, to_kind),
1104
979
(None, to_executable))
1105
# Yield all remaining source paths
981
def get_to_path(to_entry):
982
if to_entry.parent_id is None:
983
to_path = '' # the root
985
if to_entry.parent_id not in to_paths:
987
return get_to_path(self.target.inventory[to_entry.parent_id])
988
to_path = osutils.pathjoin(to_paths[to_entry.parent_id],
990
to_paths[to_entry.file_id] = to_path
1106
993
for path, from_entry in from_entries_by_dir:
1107
994
file_id = from_entry.file_id
1108
995
if file_id in to_paths:
1109
996
# already returned
1111
if file_id not in self.target.all_file_ids():
998
if not file_id in self.target.all_file_ids():
1112
999
# common case - paths we have not emitted are not present in
1116
to_path = self.target.id2path(file_id)
1003
to_path = get_to_path(self.target.inventory[file_id])
1117
1004
entry_count += 1
1118
1005
if pb is not None:
1119
1006
pb.update('comparing files', entry_count, num_entries)
1126
1013
executable = (from_executable, None)
1127
1014
changed_content = from_kind is not None
1128
1015
# the parent's path is necessarily known at this point.
1129
changed_file_ids.append(file_id)
1130
1016
yield(file_id, (path, to_path), changed_content, versioned, parent,
1131
1017
name, kind, executable)
1132
changed_file_ids = set(changed_file_ids)
1133
if specific_file_ids is not None:
1134
for result in self._handle_precise_ids(precise_file_ids,
1138
def _get_entry(self, tree, file_id):
1139
"""Get an inventory entry from a tree, with missing entries as None.
1141
If the tree raises NotImplementedError on accessing .inventory, then
1142
this is worked around using iter_entries_by_dir on just the file id
1145
:param tree: The tree to lookup the entry in.
1146
:param file_id: The file_id to lookup.
1149
inventory = tree.inventory
1150
except NotImplementedError:
1151
# No inventory available.
1153
iterator = tree.iter_entries_by_dir(specific_file_ids=[file_id])
1154
return iterator.next()[1]
1155
except StopIteration:
1159
return inventory[file_id]
1160
except errors.NoSuchId:
1163
def _handle_precise_ids(self, precise_file_ids, changed_file_ids,
1164
discarded_changes=None):
1165
"""Fill out a partial iter_changes to be consistent.
1167
:param precise_file_ids: The file ids of parents that were seen during
1169
:param changed_file_ids: The file ids of already emitted items.
1170
:param discarded_changes: An optional dict of precalculated
1171
iter_changes items which the partial iter_changes had not output
1173
:return: A generator of iter_changes items to output.
1175
# process parents of things that had changed under the users
1176
# requested paths to prevent incorrect paths or parent ids which
1177
# aren't in the tree.
1178
while precise_file_ids:
1179
precise_file_ids.discard(None)
1180
# Don't emit file_ids twice
1181
precise_file_ids.difference_update(changed_file_ids)
1182
if not precise_file_ids:
1184
# If the there was something at a given output path in source, we
1185
# have to include the entry from source in the delta, or we would
1186
# be putting this entry into a used path.
1188
for parent_id in precise_file_ids:
1190
paths.append(self.target.id2path(parent_id))
1191
except errors.NoSuchId:
1192
# This id has been dragged in from the source by delta
1193
# expansion and isn't present in target at all: we don't
1194
# need to check for path collisions on it.
1197
old_id = self.source.path2id(path)
1198
precise_file_ids.add(old_id)
1199
precise_file_ids.discard(None)
1200
current_ids = precise_file_ids
1201
precise_file_ids = set()
1202
# We have to emit all of precise_file_ids that have been altered.
1203
# We may have to output the children of some of those ids if any
1204
# directories have stopped being directories.
1205
for file_id in current_ids:
1207
if discarded_changes:
1208
result = discarded_changes.get(file_id)
1213
old_entry = self._get_entry(self.source, file_id)
1214
new_entry = self._get_entry(self.target, file_id)
1215
result, changes = self._changes_from_entries(
1216
old_entry, new_entry)
1219
# Get this parents parent to examine.
1220
new_parent_id = result[4][1]
1221
precise_file_ids.add(new_parent_id)
1223
if (result[6][0] == 'directory' and
1224
result[6][1] != 'directory'):
1225
# This stopped being a directory, the old children have
1227
if old_entry is None:
1228
# Reusing a discarded change.
1229
old_entry = self._get_entry(self.source, file_id)
1230
for child in old_entry.children.values():
1231
precise_file_ids.add(child.file_id)
1232
changed_file_ids.add(result[0])
1236
1020
class MultiWalker(object):
added
removed
bzrlib/tree.py
