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, NoSuchId
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
import bzrlib.revision
40
39
from bzrlib.symbol_versioning import deprecated_function, deprecated_in
41
from bzrlib.trace import mutter, note
40
from bzrlib.trace import note
44
43
class Tree(object):
95
94
want_unversioned=want_unversioned,
98
@symbol_versioning.deprecated_method(symbol_versioning.one_three)
99
def _iter_changes(self, *args, **kwargs):
100
return self.iter_changes(*args, **kwargs)
102
97
def iter_changes(self, from_tree, include_unchanged=False,
103
98
specific_files=None, pb=None, extra_trees=None,
104
99
require_versioned=True, want_unversioned=False):
100
"""See InterTree.iter_changes"""
105
101
intertree = InterTree.get(from_tree, self)
106
102
return intertree.iter_changes(include_unchanged, specific_files, pb,
107
103
extra_trees, require_versioned, want_unversioned=want_unversioned)
200
194
The yield order (ignoring root) would be::
201
195
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.
203
201
return self.inventory.iter_entries_by_dir(
204
specific_file_ids=specific_file_ids)
202
specific_file_ids=specific_file_ids, yield_parents=yield_parents)
206
204
def iter_references(self):
207
for path, entry in self.iter_entries_by_dir():
208
if entry.kind == 'tree-reference':
209
yield path, entry.file_id
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
211
210
def kind(self, file_id):
212
211
raise NotImplementedError("Tree subclass %s must implement kind"
223
222
def path_content_summary(self, path):
224
223
"""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.)
226
228
:param path: A relative path within the tree.
227
229
:return: A tuple containing kind, size, exec, sha1-or-link.
228
230
Kind is always present (see tree.kind()).
229
size is present if kind is file, None otherwise.
231
size is present if kind is file and the size of the
232
canonical form can be cheaply determined, None otherwise.
230
233
exec is None unless kind is file and the platform supports the 'x'
232
235
sha1-or-link is the link target if kind is symlink, or the sha1 if
265
268
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)
267
284
def get_file_text(self, file_id, path=None):
268
285
"""Return the byte content of a file.
387
404
bit_iter = iter(path.split("/"))
388
405
for elt in bit_iter:
389
406
lelt = elt.lower()
390
408
for child in self.iter_children(cur_id):
410
# XXX: it seem like if the child is known to be in the
411
# tree, we shouldn't need to go from its id back to
412
# its path -- mbp 2010-02-11
414
# XXX: it seems like we could be more efficient
415
# by just directly looking up the original name and
416
# only then searching all children; also by not
417
# chopping paths so much. -- mbp 2010-02-11
392
418
child_base = os.path.basename(self.id2path(child))
393
if child_base.lower() == lelt:
419
if (child_base == elt):
420
# if we found an exact match, we can stop now; if
421
# we found an approximate match we need to keep
422
# searching because there might be an exact match
395
cur_path = osutils.pathjoin(cur_path, child_base)
425
new_path = osutils.pathjoin(cur_path, child_base)
427
elif child_base.lower() == lelt:
429
new_path = osutils.pathjoin(cur_path, child_base)
398
431
# before a change is committed we can see this error...
401
436
# got to the end of this directory and no entries matched.
402
437
# Return what matched so far, plus the rest as specified.
627
656
def supports_content_filtering(self):
659
def _content_filter_stack(self, path=None, file_id=None):
660
"""The stack of content filters for a path if filtering is supported.
662
Readers will be applied in first-to-last order.
663
Writers will be applied in last-to-first order.
664
Either the path or the file-id needs to be provided.
666
:param path: path relative to the root of the tree
668
:param file_id: file_id or None if unknown
669
:return: the list of filters - [] if there are none
671
filter_pref_names = filters._get_registered_names()
672
if len(filter_pref_names) == 0:
675
path = self.id2path(file_id)
676
prefs = self.iter_search_rules([path], filter_pref_names).next()
677
stk = filters._get_filter_stack_for(prefs)
678
if 'filters' in debug.debug_flags:
679
note("*** %s content-filter: %s => %r" % (path,prefs,stk))
682
def _content_filter_stack_provider(self):
683
"""A function that returns a stack of ContentFilters.
685
The function takes a path (relative to the top of the tree) and a
686
file-id as parameters.
688
:return: None if content filtering is not supported by this tree.
690
if self.supports_content_filtering():
691
return lambda path, file_id: \
692
self._content_filter_stack(path, file_id)
630
696
def iter_search_rules(self, path_names, pref_names=None,
631
_default_searcher=rules._per_user_searcher):
697
_default_searcher=None):
632
698
"""Find the preferences for filenames in a tree.
634
700
:param path_names: an iterable of paths to find attributes for.
802
869
will pass through to InterTree as appropriate.
872
# Formats that will be used to test this InterTree. If both are
873
# None, this InterTree will not be tested (e.g. because a complex
875
_matching_from_tree_format = None
876
_matching_to_tree_format = None
880
def _changes_from_entries(self, source_entry, target_entry,
881
source_path=None, target_path=None):
882
"""Generate a iter_changes tuple between source_entry and target_entry.
884
:param source_entry: An inventory entry from self.source, or None.
885
:param target_entry: An inventory entry from self.target, or None.
886
:param source_path: The path of source_entry, if known. If not known
887
it will be looked up.
888
:param target_path: The path of target_entry, if known. If not known
889
it will be looked up.
890
:return: A tuple, item 0 of which is an iter_changes result tuple, and
891
item 1 is True if there are any changes in the result tuple.
893
if source_entry is None:
894
if target_entry is None:
896
file_id = target_entry.file_id
898
file_id = source_entry.file_id
899
if source_entry is not None:
900
source_versioned = True
901
source_name = source_entry.name
902
source_parent = source_entry.parent_id
903
if source_path is None:
904
source_path = self.source.id2path(file_id)
905
source_kind, source_executable, source_stat = \
906
self.source._comparison_data(source_entry, source_path)
908
source_versioned = False
912
source_executable = None
913
if target_entry is not None:
914
target_versioned = True
915
target_name = target_entry.name
916
target_parent = target_entry.parent_id
917
if target_path is None:
918
target_path = self.target.id2path(file_id)
919
target_kind, target_executable, target_stat = \
920
self.target._comparison_data(target_entry, target_path)
922
target_versioned = False
926
target_executable = None
927
versioned = (source_versioned, target_versioned)
928
kind = (source_kind, target_kind)
929
changed_content = False
930
if source_kind != target_kind:
931
changed_content = True
932
elif source_kind == 'file':
933
if (self.source.get_file_sha1(file_id, source_path, source_stat) !=
934
self.target.get_file_sha1(file_id, target_path, target_stat)):
935
changed_content = True
936
elif source_kind == 'symlink':
937
if (self.source.get_symlink_target(file_id) !=
938
self.target.get_symlink_target(file_id)):
939
changed_content = True
940
# XXX: Yes, the indentation below is wrong. But fixing it broke
941
# test_merge.TestMergerEntriesLCAOnDisk.
942
# test_nested_tree_subtree_renamed_and_modified. We'll wait for
943
# the fix from bzr.dev -- vila 2009026
944
elif source_kind == 'tree-reference':
945
if (self.source.get_reference_revision(file_id, source_path)
946
!= self.target.get_reference_revision(file_id, target_path)):
947
changed_content = True
948
parent = (source_parent, target_parent)
949
name = (source_name, target_name)
950
executable = (source_executable, target_executable)
951
if (changed_content is not False or versioned[0] != versioned[1]
952
or parent[0] != parent[1] or name[0] != name[1] or
953
executable[0] != executable[1]):
957
return (file_id, (source_path, target_path), changed_content,
958
versioned, parent, name, kind, executable), changes
808
961
def compare(self, want_unchanged=False, specific_files=None,
809
962
extra_trees=None, require_versioned=False, include_root=False,
872
1023
:param require_versioned: Raise errors.PathsNotVersionedError if a
873
1024
path in the specific_files list is not versioned in one of
874
1025
source, target or extra_trees.
1026
:param specific_files: An optional list of file paths to restrict the
1027
comparison to. When mapping filenames to ids, all matches in all
1028
trees (including optional extra_trees) are used, and all children
1029
of matched directories are included. The parents in the target tree
1030
of the specific files up to and including the root of the tree are
1031
always evaluated for changes too.
875
1032
:param want_unversioned: Should unversioned files be returned in the
876
1033
output. An unversioned file is defined as one with (False, False)
877
1034
for the versioned pair.
880
1036
lookup_trees = [self.source]
882
1038
lookup_trees.extend(extra_trees)
1039
# The ids of items we need to examine to insure delta consistency.
1040
precise_file_ids = set()
1041
changed_file_ids = []
883
1042
if specific_files == []:
884
1043
specific_file_ids = []
886
1045
specific_file_ids = self.target.paths2ids(specific_files,
887
1046
lookup_trees, require_versioned=require_versioned)
1047
if specific_files is not None:
1048
# reparented or added entries must have their parents included
1049
# so that valid deltas can be created. The seen_parents set
1050
# tracks the parents that we need to have.
1051
# The seen_dirs set tracks directory entries we've yielded.
1052
# After outputting version object in to_entries we set difference
1053
# the two seen sets and start checking parents.
1054
seen_parents = set()
888
1056
if want_unversioned:
889
1057
all_unversioned = sorted([(p.split('/'), p) for p in
890
1058
self.target.extras()
905
1073
# can be extras. So the fake_entry is solely used to look up
906
1074
# executable it values when execute is not supported.
907
1075
fake_entry = InventoryFile('unused', 'unused', 'unused')
908
for to_path, to_entry in to_entries_by_dir:
909
while all_unversioned and all_unversioned[0][0] < to_path.split('/'):
1076
for target_path, target_entry in to_entries_by_dir:
1077
while (all_unversioned and
1078
all_unversioned[0][0] < target_path.split('/')):
910
1079
unversioned_path = all_unversioned.popleft()
911
to_kind, to_executable, to_stat = \
1080
target_kind, target_executable, target_stat = \
912
1081
self.target._comparison_data(fake_entry, unversioned_path[1])
913
1082
yield (None, (None, unversioned_path[1]), True, (False, False),
915
1084
(None, unversioned_path[0][-1]),
917
(None, to_executable))
918
file_id = to_entry.file_id
919
to_paths[file_id] = to_path
1085
(None, target_kind),
1086
(None, target_executable))
1087
source_path, source_entry = from_data.get(target_entry.file_id,
1089
result, changes = self._changes_from_entries(source_entry,
1090
target_entry, source_path=source_path, target_path=target_path)
1091
to_paths[result[0]] = result[1][1]
920
1092
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)
930
1094
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)
962
1095
if pb is not None:
963
1096
pb.update('comparing files', entry_count, num_entries)
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)
1097
if changes or include_unchanged:
1098
if specific_file_ids is not None:
1099
new_parent_id = result[4][1]
1100
precise_file_ids.add(new_parent_id)
1101
changed_file_ids.append(result[0])
1103
# Ensure correct behaviour for reparented/added specific files.
1104
if specific_files is not None:
1105
# Record output dirs
1106
if result[6][1] == 'directory':
1107
seen_dirs.add(result[0])
1108
# Record parents of reparented/added entries.
1109
versioned = result[3]
1111
if not versioned[0] or parents[0] != parents[1]:
1112
seen_parents.add(parents[1])
970
1113
while all_unversioned:
971
1114
# yield any trailing unversioned paths
972
1115
unversioned_path = all_unversioned.popleft()
977
1120
(None, unversioned_path[0][-1]),
978
1121
(None, to_kind),
979
1122
(None, to_executable))
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
1123
# Yield all remaining source paths
993
1124
for path, from_entry in from_entries_by_dir:
994
1125
file_id = from_entry.file_id
995
1126
if file_id in to_paths:
996
1127
# already returned
998
if not file_id in self.target.all_file_ids():
1129
if file_id not in self.target.all_file_ids():
999
1130
# common case - paths we have not emitted are not present in
1003
to_path = get_to_path(self.target.inventory[file_id])
1134
to_path = self.target.id2path(file_id)
1004
1135
entry_count += 1
1005
1136
if pb is not None:
1006
1137
pb.update('comparing files', entry_count, num_entries)
1013
1144
executable = (from_executable, None)
1014
1145
changed_content = from_kind is not None
1015
1146
# the parent's path is necessarily known at this point.
1147
changed_file_ids.append(file_id)
1016
1148
yield(file_id, (path, to_path), changed_content, versioned, parent,
1017
1149
name, kind, executable)
1150
changed_file_ids = set(changed_file_ids)
1151
if specific_file_ids is not None:
1152
for result in self._handle_precise_ids(precise_file_ids,
1156
def _get_entry(self, tree, file_id):
1157
"""Get an inventory entry from a tree, with missing entries as None.
1159
If the tree raises NotImplementedError on accessing .inventory, then
1160
this is worked around using iter_entries_by_dir on just the file id
1163
:param tree: The tree to lookup the entry in.
1164
:param file_id: The file_id to lookup.
1167
inventory = tree.inventory
1168
except NotImplementedError:
1169
# No inventory available.
1171
iterator = tree.iter_entries_by_dir(specific_file_ids=[file_id])
1172
return iterator.next()[1]
1173
except StopIteration:
1177
return inventory[file_id]
1178
except errors.NoSuchId:
1181
def _handle_precise_ids(self, precise_file_ids, changed_file_ids,
1182
discarded_changes=None):
1183
"""Fill out a partial iter_changes to be consistent.
1185
:param precise_file_ids: The file ids of parents that were seen during
1187
:param changed_file_ids: The file ids of already emitted items.
1188
:param discarded_changes: An optional dict of precalculated
1189
iter_changes items which the partial iter_changes had not output
1191
:return: A generator of iter_changes items to output.
1193
# process parents of things that had changed under the users
1194
# requested paths to prevent incorrect paths or parent ids which
1195
# aren't in the tree.
1196
while precise_file_ids:
1197
precise_file_ids.discard(None)
1198
# Don't emit file_ids twice
1199
precise_file_ids.difference_update(changed_file_ids)
1200
if not precise_file_ids:
1202
# If the there was something at a given output path in source, we
1203
# have to include the entry from source in the delta, or we would
1204
# be putting this entry into a used path.
1206
for parent_id in precise_file_ids:
1208
paths.append(self.target.id2path(parent_id))
1209
except errors.NoSuchId:
1210
# This id has been dragged in from the source by delta
1211
# expansion and isn't present in target at all: we don't
1212
# need to check for path collisions on it.
1215
old_id = self.source.path2id(path)
1216
precise_file_ids.add(old_id)
1217
precise_file_ids.discard(None)
1218
current_ids = precise_file_ids
1219
precise_file_ids = set()
1220
# We have to emit all of precise_file_ids that have been altered.
1221
# We may have to output the children of some of those ids if any
1222
# directories have stopped being directories.
1223
for file_id in current_ids:
1225
if discarded_changes:
1226
result = discarded_changes.get(file_id)
1231
old_entry = self._get_entry(self.source, file_id)
1232
new_entry = self._get_entry(self.target, file_id)
1233
result, changes = self._changes_from_entries(
1234
old_entry, new_entry)
1237
# Get this parents parent to examine.
1238
new_parent_id = result[4][1]
1239
precise_file_ids.add(new_parent_id)
1241
if (result[6][0] == 'directory' and
1242
result[6][1] != 'directory'):
1243
# This stopped being a directory, the old children have
1245
if old_entry is None:
1246
# Reusing a discarded change.
1247
old_entry = self._get_entry(self.source, file_id)
1248
for child in old_entry.children.values():
1249
precise_file_ids.add(child.file_id)
1250
changed_file_ids.add(result[0])
1020
1254
class MultiWalker(object):
added
removed
bzrlib/tree.py
