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
import bzrlib.revision
39
40
from bzrlib.symbol_versioning import deprecated_function, deprecated_in
40
from bzrlib.trace import note
41
from bzrlib.trace import mutter, 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,
98
@symbol_versioning.deprecated_method(symbol_versioning.one_three)
99
def _iter_changes(self, *args, **kwargs):
100
return self.iter_changes(*args, **kwargs)
97
102
def iter_changes(self, from_tree, include_unchanged=False,
98
103
specific_files=None, pb=None, extra_trees=None,
99
104
require_versioned=True, want_unversioned=False):
100
"""See InterTree.iter_changes"""
101
105
intertree = InterTree.get(from_tree, self)
102
106
return intertree.iter_changes(include_unchanged, specific_files, pb,
103
107
extra_trees, require_versioned, want_unversioned=want_unversioned)
105
109
def conflicts(self):
106
110
"""Get a list of the conflicts in the tree.
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
259
256
def _get_inventory(self):
260
257
return self._inventory
262
259
def get_file(self, file_id, path=None):
263
260
"""Return a file object for the file file_id in the tree.
265
262
If both file_id and path are defined, it is implementation defined as
266
263
to which one is used.
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.
404
387
bit_iter = iter(path.split("/"))
405
388
for elt in bit_iter:
406
389
lelt = elt.lower()
408
390
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
418
392
child_base = os.path.basename(self.id2path(child))
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
393
if child_base.lower() == lelt:
425
new_path = osutils.pathjoin(cur_path, child_base)
395
cur_path = osutils.pathjoin(cur_path, child_base)
427
elif child_base.lower() == lelt:
429
new_path = osutils.pathjoin(cur_path, child_base)
431
398
# before a change is committed we can see this error...
436
401
# got to the end of this directory and no entries matched.
437
402
# Return what matched so far, plus the rest as specified.
626
597
This yields all the data about the contents of a directory at a time.
627
598
After each directory has been yielded, if the caller has mutated the
628
599
list to exclude some directories, they are then not descended into.
630
601
The data yielded is of the form:
631
602
((directory-relpath, directory-path-from-root, directory-fileid),
632
[(relpath, basename, kind, lstat, path_from_tree_root, file_id,
603
[(relpath, basename, kind, lstat, path_from_tree_root, file_id,
633
604
versioned_kind), ...]),
634
605
- directory-relpath is the containing dirs relpath from prefix
635
606
- directory-path-from-root is the containing dirs path from /
656
627
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)
696
630
def iter_search_rules(self, path_names, pref_names=None,
697
_default_searcher=None):
631
_default_searcher=rules._per_user_searcher):
698
632
"""Find the preferences for filenames in a tree.
700
634
:param path_names: an iterable of paths to find attributes for.
715
647
for path in path_names:
716
648
yield searcher.get_items(path)
718
651
def _get_rules_searcher(self, default_searcher):
719
652
"""Get the RulesSearcher for this tree given the default one."""
720
653
searcher = default_searcher
657
class EmptyTree(Tree):
660
self._inventory = Inventory(root_id=None)
661
symbol_versioning.warn('EmptyTree is deprecated as of bzr 0.9 please'
662
' use repository.revision_tree instead.',
663
DeprecationWarning, stacklevel=2)
665
def get_parent_ids(self):
668
def get_symlink_target(self, file_id):
671
def has_filename(self, filename):
674
def kind(self, file_id):
677
def list_files(self, include_root=False):
680
def __contains__(self, file_id):
681
return (file_id in self._inventory)
683
def get_file_sha1(self, file_id, path=None, stat_value=None):
724
687
######################################################################
869
832
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
961
838
def compare(self, want_unchanged=False, specific_files=None,
962
839
extra_trees=None, require_versioned=False, include_root=False,
1023
902
:param require_versioned: Raise errors.PathsNotVersionedError if a
1024
903
path in the specific_files list is not versioned in one of
1025
904
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.
1032
905
:param want_unversioned: Should unversioned files be returned in the
1033
906
output. An unversioned file is defined as one with (False, False)
1034
907
for the versioned pair.
1036
910
lookup_trees = [self.source]
1038
912
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 = []
1042
913
if specific_files == []:
1043
914
specific_file_ids = []
1045
916
specific_file_ids = self.target.paths2ids(specific_files,
1046
917
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()
1056
918
if want_unversioned:
1057
919
all_unversioned = sorted([(p.split('/'), p) for p in
1058
920
self.target.extras()
1069
931
specific_file_ids=specific_file_ids))
1070
932
num_entries = len(from_entries_by_dir) + len(to_entries_by_dir)
1072
# the unversioned path lookup only occurs on real trees - where there
934
# the unversioned path lookup only occurs on real trees - where there
1073
935
# can be extras. So the fake_entry is solely used to look up
1074
936
# executable it values when execute is not supported.
1075
937
fake_entry = InventoryFile('unused', 'unused', 'unused')
1076
for target_path, target_entry in to_entries_by_dir:
1077
while (all_unversioned and
1078
all_unversioned[0][0] < target_path.split('/')):
938
for to_path, to_entry in to_entries_by_dir:
939
while all_unversioned and all_unversioned[0][0] < to_path.split('/'):
1079
940
unversioned_path = all_unversioned.popleft()
1080
target_kind, target_executable, target_stat = \
941
to_kind, to_executable, to_stat = \
1081
942
self.target._comparison_data(fake_entry, unversioned_path[1])
1082
943
yield (None, (None, unversioned_path[1]), True, (False, False),
1084
945
(None, unversioned_path[0][-1]),
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]
947
(None, to_executable))
948
file_id = to_entry.file_id
949
to_paths[file_id] = to_path
1092
950
entry_count += 1
951
changed_content = False
952
from_path, from_entry = from_data.get(file_id, (None, None))
953
from_versioned = (from_entry is not None)
954
if from_entry is not None:
955
from_versioned = True
956
from_name = from_entry.name
957
from_parent = from_entry.parent_id
958
from_kind, from_executable, from_stat = \
959
self.source._comparison_data(from_entry, from_path)
1094
960
entry_count += 1
962
from_versioned = False
966
from_executable = None
967
versioned = (from_versioned, True)
968
to_kind, to_executable, to_stat = \
969
self.target._comparison_data(to_entry, to_path)
970
kind = (from_kind, to_kind)
971
if kind[0] != kind[1]:
972
changed_content = True
973
elif from_kind == 'file':
974
from_size = self.source._file_size(from_entry, from_stat)
975
to_size = self.target._file_size(to_entry, to_stat)
976
if from_size != to_size:
977
changed_content = True
978
elif (self.source.get_file_sha1(file_id, from_path, from_stat) !=
979
self.target.get_file_sha1(file_id, to_path, to_stat)):
980
changed_content = True
981
elif from_kind == 'symlink':
982
if (self.source.get_symlink_target(file_id) !=
983
self.target.get_symlink_target(file_id)):
984
changed_content = True
985
elif from_kind == 'tree-reference':
986
if (self.source.get_reference_revision(file_id, from_path)
987
!= self.target.get_reference_revision(file_id, to_path)):
988
changed_content = True
989
parent = (from_parent, to_entry.parent_id)
990
name = (from_name, to_entry.name)
991
executable = (from_executable, to_executable)
1095
992
if pb is not None:
1096
993
pb.update('comparing files', entry_count, num_entries)
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])
994
if (changed_content is not False or versioned[0] != versioned[1]
995
or parent[0] != parent[1] or name[0] != name[1] or
996
executable[0] != executable[1] or include_unchanged):
997
yield (file_id, (from_path, to_path), changed_content,
998
versioned, parent, name, kind, executable)
1113
1000
while all_unversioned:
1114
1001
# yield any trailing unversioned paths
1115
1002
unversioned_path = all_unversioned.popleft()
1120
1007
(None, unversioned_path[0][-1]),
1121
1008
(None, to_kind),
1122
1009
(None, to_executable))
1123
# Yield all remaining source paths
1011
def get_to_path(to_entry):
1012
if to_entry.parent_id is None:
1013
to_path = '' # the root
1015
if to_entry.parent_id not in to_paths:
1017
return get_to_path(self.target.inventory[to_entry.parent_id])
1018
to_path = osutils.pathjoin(to_paths[to_entry.parent_id],
1020
to_paths[to_entry.file_id] = to_path
1124
1023
for path, from_entry in from_entries_by_dir:
1125
1024
file_id = from_entry.file_id
1126
1025
if file_id in to_paths:
1127
1026
# already returned
1129
if file_id not in self.target.all_file_ids():
1028
if not file_id in self.target.all_file_ids():
1130
1029
# common case - paths we have not emitted are not present in
1134
to_path = self.target.id2path(file_id)
1033
to_path = get_to_path(self.target.inventory[file_id])
1135
1034
entry_count += 1
1136
1035
if pb is not None:
1137
1036
pb.update('comparing files', entry_count, num_entries)
1144
1043
executable = (from_executable, None)
1145
1044
changed_content = from_kind is not None
1146
1045
# the parent's path is necessarily known at this point.
1147
changed_file_ids.append(file_id)
1148
1046
yield(file_id, (path, to_path), changed_content, versioned, parent,
1149
1047
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])
1254
1050
class MultiWalker(object):
added
removed
bzrlib/tree.py
