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
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):
45
44
"""Abstract file tree.
47
46
There are several subclasses:
49
48
* `WorkingTree` exists as files on disk editable by the user.
51
50
* `RevisionTree` is a tree as recorded at some point in the past.
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)
109
105
def conflicts(self):
110
106
"""Get a list of the conflicts in the tree.
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
256
259
def _get_inventory(self):
257
260
return self._inventory
259
262
def get_file(self, file_id, path=None):
260
263
"""Return a file object for the file file_id in the tree.
262
265
If both file_id and path are defined, it is implementation defined as
263
266
to which one is used.
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.
346
363
raise NotImplementedError(self.get_symlink_target)
365
def get_canonical_inventory_paths(self, paths):
366
"""Like get_canonical_inventory_path() but works on multiple items.
368
:param paths: A sequence of paths relative to the root of the tree.
369
:return: A list of paths, with each item the corresponding input path
370
adjusted to account for existing elements that match case
373
return list(self._yield_canonical_inventory_paths(paths))
375
def get_canonical_inventory_path(self, path):
376
"""Returns the first inventory item that case-insensitively matches path.
378
If a path matches exactly, it is returned. If no path matches exactly
379
but more than one path matches case-insensitively, it is implementation
380
defined which is returned.
382
If no path matches case-insensitively, the input path is returned, but
383
with as many path entries that do exist changed to their canonical
386
If you need to resolve many names from the same tree, you should
387
use get_canonical_inventory_paths() to avoid O(N) behaviour.
389
:param path: A paths relative to the root of the tree.
390
:return: The input path adjusted to account for existing elements
391
that match case insensitively.
393
return self._yield_canonical_inventory_paths([path]).next()
395
def _yield_canonical_inventory_paths(self, paths):
397
# First, if the path as specified exists exactly, just use it.
398
if self.path2id(path) is not None:
402
cur_id = self.get_root_id()
404
bit_iter = iter(path.split("/"))
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
418
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
425
new_path = osutils.pathjoin(cur_path, child_base)
427
elif child_base.lower() == lelt:
429
new_path = osutils.pathjoin(cur_path, child_base)
431
# before a change is committed we can see this error...
436
# got to the end of this directory and no entries matched.
437
# Return what matched so far, plus the rest as specified.
438
cur_path = osutils.pathjoin(cur_path, elt, *list(bit_iter))
348
443
def get_root_id(self):
349
444
"""Return the file_id for the root of this tree."""
350
445
raise NotImplementedError(self.get_root_id)
537
630
This yields all the data about the contents of a directory at a time.
538
631
After each directory has been yielded, if the caller has mutated the
539
632
list to exclude some directories, they are then not descended into.
541
634
The data yielded is of the form:
542
635
((directory-relpath, directory-path-from-root, directory-fileid),
543
[(relpath, basename, kind, lstat, path_from_tree_root, file_id,
636
[(relpath, basename, kind, lstat, path_from_tree_root, file_id,
544
637
versioned_kind), ...]),
545
638
- directory-relpath is the containing dirs relpath from prefix
546
639
- directory-path-from-root is the containing dirs path from /
565
658
raise NotImplementedError(self.walkdirs)
660
def supports_content_filtering(self):
663
def _content_filter_stack(self, path=None, file_id=None):
664
"""The stack of content filters for a path if filtering is supported.
666
Readers will be applied in first-to-last order.
667
Writers will be applied in last-to-first order.
668
Either the path or the file-id needs to be provided.
670
:param path: path relative to the root of the tree
672
:param file_id: file_id or None if unknown
673
:return: the list of filters - [] if there are none
675
filter_pref_names = filters._get_registered_names()
676
if len(filter_pref_names) == 0:
679
path = self.id2path(file_id)
680
prefs = self.iter_search_rules([path], filter_pref_names).next()
681
stk = filters._get_filter_stack_for(prefs)
682
if 'filters' in debug.debug_flags:
683
note("*** %s content-filter: %s => %r" % (path,prefs,stk))
686
def _content_filter_stack_provider(self):
687
"""A function that returns a stack of ContentFilters.
689
The function takes a path (relative to the top of the tree) and a
690
file-id as parameters.
692
:return: None if content filtering is not supported by this tree.
694
if self.supports_content_filtering():
695
return lambda path, file_id: \
696
self._content_filter_stack(path, file_id)
567
700
def iter_search_rules(self, path_names, pref_names=None,
568
_default_searcher=rules._per_user_searcher):
701
_default_searcher=None):
569
702
"""Find the preferences for filenames in a tree.
571
704
:param path_names: an iterable of paths to find attributes for.
584
719
for path in path_names:
585
720
yield searcher.get_items(path)
588
722
def _get_rules_searcher(self, default_searcher):
589
723
"""Get the RulesSearcher for this tree given the default one."""
590
724
searcher = default_searcher
594
class EmptyTree(Tree):
597
self._inventory = Inventory(root_id=None)
598
symbol_versioning.warn('EmptyTree is deprecated as of bzr 0.9 please'
599
' use repository.revision_tree instead.',
600
DeprecationWarning, stacklevel=2)
602
def get_parent_ids(self):
605
def get_symlink_target(self, file_id):
608
def has_filename(self, filename):
611
def kind(self, file_id):
614
def list_files(self, include_root=False):
617
def __contains__(self, file_id):
618
return (file_id in self._inventory)
620
def get_file_sha1(self, file_id, path=None, stat_value=None):
624
728
######################################################################
769
873
will pass through to InterTree as appropriate.
876
# Formats that will be used to test this InterTree. If both are
877
# None, this InterTree will not be tested (e.g. because a complex
879
_matching_from_tree_format = None
880
_matching_to_tree_format = None
884
def _changes_from_entries(self, source_entry, target_entry,
885
source_path=None, target_path=None):
886
"""Generate a iter_changes tuple between source_entry and target_entry.
888
:param source_entry: An inventory entry from self.source, or None.
889
:param target_entry: An inventory entry from self.target, or None.
890
:param source_path: The path of source_entry, if known. If not known
891
it will be looked up.
892
:param target_path: The path of target_entry, if known. If not known
893
it will be looked up.
894
:return: A tuple, item 0 of which is an iter_changes result tuple, and
895
item 1 is True if there are any changes in the result tuple.
897
if source_entry is None:
898
if target_entry is None:
900
file_id = target_entry.file_id
902
file_id = source_entry.file_id
903
if source_entry is not None:
904
source_versioned = True
905
source_name = source_entry.name
906
source_parent = source_entry.parent_id
907
if source_path is None:
908
source_path = self.source.id2path(file_id)
909
source_kind, source_executable, source_stat = \
910
self.source._comparison_data(source_entry, source_path)
912
source_versioned = False
916
source_executable = None
917
if target_entry is not None:
918
target_versioned = True
919
target_name = target_entry.name
920
target_parent = target_entry.parent_id
921
if target_path is None:
922
target_path = self.target.id2path(file_id)
923
target_kind, target_executable, target_stat = \
924
self.target._comparison_data(target_entry, target_path)
926
target_versioned = False
930
target_executable = None
931
versioned = (source_versioned, target_versioned)
932
kind = (source_kind, target_kind)
933
changed_content = False
934
if source_kind != target_kind:
935
changed_content = True
936
elif source_kind == 'file':
937
if (self.source.get_file_sha1(file_id, source_path, source_stat) !=
938
self.target.get_file_sha1(file_id, target_path, target_stat)):
939
changed_content = True
940
elif source_kind == 'symlink':
941
if (self.source.get_symlink_target(file_id) !=
942
self.target.get_symlink_target(file_id)):
943
changed_content = True
944
# XXX: Yes, the indentation below is wrong. But fixing it broke
945
# test_merge.TestMergerEntriesLCAOnDisk.
946
# test_nested_tree_subtree_renamed_and_modified. We'll wait for
947
# the fix from bzr.dev -- vila 2009026
948
elif source_kind == 'tree-reference':
949
if (self.source.get_reference_revision(file_id, source_path)
950
!= self.target.get_reference_revision(file_id, target_path)):
951
changed_content = True
952
parent = (source_parent, target_parent)
953
name = (source_name, target_name)
954
executable = (source_executable, target_executable)
955
if (changed_content is not False or versioned[0] != versioned[1]
956
or parent[0] != parent[1] or name[0] != name[1] or
957
executable[0] != executable[1]):
961
return (file_id, (source_path, target_path), changed_content,
962
versioned, parent, name, kind, executable), changes
775
965
def compare(self, want_unchanged=False, specific_files=None,
776
966
extra_trees=None, require_versioned=False, include_root=False,
839
1027
:param require_versioned: Raise errors.PathsNotVersionedError if a
840
1028
path in the specific_files list is not versioned in one of
841
1029
source, target or extra_trees.
1030
:param specific_files: An optional list of file paths to restrict the
1031
comparison to. When mapping filenames to ids, all matches in all
1032
trees (including optional extra_trees) are used, and all children
1033
of matched directories are included. The parents in the target tree
1034
of the specific files up to and including the root of the tree are
1035
always evaluated for changes too.
842
1036
:param want_unversioned: Should unversioned files be returned in the
843
1037
output. An unversioned file is defined as one with (False, False)
844
1038
for the versioned pair.
847
1040
lookup_trees = [self.source]
849
1042
lookup_trees.extend(extra_trees)
1043
# The ids of items we need to examine to insure delta consistency.
1044
precise_file_ids = set()
1045
changed_file_ids = []
850
1046
if specific_files == []:
851
1047
specific_file_ids = []
853
1049
specific_file_ids = self.target.paths2ids(specific_files,
854
1050
lookup_trees, require_versioned=require_versioned)
1051
if specific_files is not None:
1052
# reparented or added entries must have their parents included
1053
# so that valid deltas can be created. The seen_parents set
1054
# tracks the parents that we need to have.
1055
# The seen_dirs set tracks directory entries we've yielded.
1056
# After outputting version object in to_entries we set difference
1057
# the two seen sets and start checking parents.
1058
seen_parents = set()
855
1060
if want_unversioned:
856
1061
all_unversioned = sorted([(p.split('/'), p) for p in
857
1062
self.target.extras()
868
1073
specific_file_ids=specific_file_ids))
869
1074
num_entries = len(from_entries_by_dir) + len(to_entries_by_dir)
871
# the unversioned path lookup only occurs on real trees - where there
1076
# the unversioned path lookup only occurs on real trees - where there
872
1077
# can be extras. So the fake_entry is solely used to look up
873
1078
# executable it values when execute is not supported.
874
1079
fake_entry = InventoryFile('unused', 'unused', 'unused')
875
for to_path, to_entry in to_entries_by_dir:
876
while all_unversioned and all_unversioned[0][0] < to_path.split('/'):
1080
for target_path, target_entry in to_entries_by_dir:
1081
while (all_unversioned and
1082
all_unversioned[0][0] < target_path.split('/')):
877
1083
unversioned_path = all_unversioned.popleft()
878
to_kind, to_executable, to_stat = \
1084
target_kind, target_executable, target_stat = \
879
1085
self.target._comparison_data(fake_entry, unversioned_path[1])
880
1086
yield (None, (None, unversioned_path[1]), True, (False, False),
882
1088
(None, unversioned_path[0][-1]),
884
(None, to_executable))
885
file_id = to_entry.file_id
886
to_paths[file_id] = to_path
1089
(None, target_kind),
1090
(None, target_executable))
1091
source_path, source_entry = from_data.get(target_entry.file_id,
1093
result, changes = self._changes_from_entries(source_entry,
1094
target_entry, source_path=source_path, target_path=target_path)
1095
to_paths[result[0]] = result[1][1]
887
1096
entry_count += 1
888
changed_content = False
889
from_path, from_entry = from_data.get(file_id, (None, None))
890
from_versioned = (from_entry is not None)
891
if from_entry is not None:
892
from_versioned = True
893
from_name = from_entry.name
894
from_parent = from_entry.parent_id
895
from_kind, from_executable, from_stat = \
896
self.source._comparison_data(from_entry, from_path)
897
1098
entry_count += 1
899
from_versioned = False
903
from_executable = None
904
versioned = (from_versioned, True)
905
to_kind, to_executable, to_stat = \
906
self.target._comparison_data(to_entry, to_path)
907
kind = (from_kind, to_kind)
908
if kind[0] != kind[1]:
909
changed_content = True
910
elif from_kind == 'file':
911
from_size = self.source._file_size(from_entry, from_stat)
912
to_size = self.target._file_size(to_entry, to_stat)
913
if from_size != to_size:
914
changed_content = True
915
elif (self.source.get_file_sha1(file_id, from_path, from_stat) !=
916
self.target.get_file_sha1(file_id, to_path, to_stat)):
917
changed_content = True
918
elif from_kind == 'symlink':
919
if (self.source.get_symlink_target(file_id) !=
920
self.target.get_symlink_target(file_id)):
921
changed_content = True
922
elif from_kind == 'tree-reference':
923
if (self.source.get_reference_revision(file_id, from_path)
924
!= self.target.get_reference_revision(file_id, to_path)):
925
changed_content = True
926
parent = (from_parent, to_entry.parent_id)
927
name = (from_name, to_entry.name)
928
executable = (from_executable, to_executable)
929
1099
if pb is not None:
930
1100
pb.update('comparing files', entry_count, num_entries)
931
if (changed_content is not False or versioned[0] != versioned[1]
932
or parent[0] != parent[1] or name[0] != name[1] or
933
executable[0] != executable[1] or include_unchanged):
934
yield (file_id, (from_path, to_path), changed_content,
935
versioned, parent, name, kind, executable)
1101
if changes or include_unchanged:
1102
if specific_file_ids is not None:
1103
new_parent_id = result[4][1]
1104
precise_file_ids.add(new_parent_id)
1105
changed_file_ids.append(result[0])
1107
# Ensure correct behaviour for reparented/added specific files.
1108
if specific_files is not None:
1109
# Record output dirs
1110
if result[6][1] == 'directory':
1111
seen_dirs.add(result[0])
1112
# Record parents of reparented/added entries.
1113
versioned = result[3]
1115
if not versioned[0] or parents[0] != parents[1]:
1116
seen_parents.add(parents[1])
937
1117
while all_unversioned:
938
1118
# yield any trailing unversioned paths
939
1119
unversioned_path = all_unversioned.popleft()
944
1124
(None, unversioned_path[0][-1]),
945
1125
(None, to_kind),
946
1126
(None, to_executable))
948
def get_to_path(to_entry):
949
if to_entry.parent_id is None:
950
to_path = '' # the root
952
if to_entry.parent_id not in to_paths:
954
return get_to_path(self.target.inventory[to_entry.parent_id])
955
to_path = osutils.pathjoin(to_paths[to_entry.parent_id],
957
to_paths[to_entry.file_id] = to_path
1127
# Yield all remaining source paths
960
1128
for path, from_entry in from_entries_by_dir:
961
1129
file_id = from_entry.file_id
962
1130
if file_id in to_paths:
963
1131
# already returned
965
if not file_id in self.target.all_file_ids():
1133
if file_id not in self.target.all_file_ids():
966
1134
# common case - paths we have not emitted are not present in
970
to_path = get_to_path(self.target.inventory[file_id])
1138
to_path = self.target.id2path(file_id)
971
1139
entry_count += 1
972
1140
if pb is not None:
973
1141
pb.update('comparing files', entry_count, num_entries)
980
1148
executable = (from_executable, None)
981
1149
changed_content = from_kind is not None
982
1150
# the parent's path is necessarily known at this point.
1151
changed_file_ids.append(file_id)
983
1152
yield(file_id, (path, to_path), changed_content, versioned, parent,
984
1153
name, kind, executable)
1154
changed_file_ids = set(changed_file_ids)
1155
if specific_file_ids is not None:
1156
for result in self._handle_precise_ids(precise_file_ids,
1160
def _get_entry(self, tree, file_id):
1161
"""Get an inventory entry from a tree, with missing entries as None.
1163
If the tree raises NotImplementedError on accessing .inventory, then
1164
this is worked around using iter_entries_by_dir on just the file id
1167
:param tree: The tree to lookup the entry in.
1168
:param file_id: The file_id to lookup.
1171
inventory = tree.inventory
1172
except NotImplementedError:
1173
# No inventory available.
1175
iterator = tree.iter_entries_by_dir(specific_file_ids=[file_id])
1176
return iterator.next()[1]
1177
except StopIteration:
1181
return inventory[file_id]
1182
except errors.NoSuchId:
1185
def _handle_precise_ids(self, precise_file_ids, changed_file_ids,
1186
discarded_changes=None):
1187
"""Fill out a partial iter_changes to be consistent.
1189
:param precise_file_ids: The file ids of parents that were seen during
1191
:param changed_file_ids: The file ids of already emitted items.
1192
:param discarded_changes: An optional dict of precalculated
1193
iter_changes items which the partial iter_changes had not output
1195
:return: A generator of iter_changes items to output.
1197
# process parents of things that had changed under the users
1198
# requested paths to prevent incorrect paths or parent ids which
1199
# aren't in the tree.
1200
while precise_file_ids:
1201
precise_file_ids.discard(None)
1202
# Don't emit file_ids twice
1203
precise_file_ids.difference_update(changed_file_ids)
1204
if not precise_file_ids:
1206
# If the there was something at a given output path in source, we
1207
# have to include the entry from source in the delta, or we would
1208
# be putting this entry into a used path.
1210
for parent_id in precise_file_ids:
1212
paths.append(self.target.id2path(parent_id))
1213
except errors.NoSuchId:
1214
# This id has been dragged in from the source by delta
1215
# expansion and isn't present in target at all: we don't
1216
# need to check for path collisions on it.
1219
old_id = self.source.path2id(path)
1220
precise_file_ids.add(old_id)
1221
precise_file_ids.discard(None)
1222
current_ids = precise_file_ids
1223
precise_file_ids = set()
1224
# We have to emit all of precise_file_ids that have been altered.
1225
# We may have to output the children of some of those ids if any
1226
# directories have stopped being directories.
1227
for file_id in current_ids:
1229
if discarded_changes:
1230
result = discarded_changes.get(file_id)
1235
old_entry = self._get_entry(self.source, file_id)
1236
new_entry = self._get_entry(self.target, file_id)
1237
result, changes = self._changes_from_entries(
1238
old_entry, new_entry)
1241
# Get this parents parent to examine.
1242
new_parent_id = result[4][1]
1243
precise_file_ids.add(new_parent_id)
1245
if (result[6][0] == 'directory' and
1246
result[6][1] != 'directory'):
1247
# This stopped being a directory, the old children have
1249
if old_entry is None:
1250
# Reusing a discarded change.
1251
old_entry = self._get_entry(self.source, file_id)
1252
for child in old_entry.children.values():
1253
precise_file_ids.add(child.file_id)
1254
changed_file_ids.add(result[0])
987
1258
class MultiWalker(object):
added
removed
bzrlib/tree.py
