980
876
target_kind, target_executable, target_stat = \
981
877
self.target._comparison_data(
982
878
fake_entry, unversioned_path[1])
984
None, (None, unversioned_path[1]), True, (False, False),
986
(None, unversioned_path[0][-1]),
988
(None, target_executable))
989
source_path = path_equivs[target_path]
990
if source_path is not None:
991
source_entry = from_data.get(source_path)
994
result, changes = self._changes_from_entries(
995
source_entry, target_entry, source_path=source_path, target_path=target_path)
996
to_paths[result.file_id] = result.path[1]
879
yield (None, (None, unversioned_path[1]), True, (False, False),
881
(None, unversioned_path[0][-1]),
883
(None, target_executable))
884
source_path, source_entry = from_data.get(target_entry.file_id,
886
result, changes = self._changes_from_entries(source_entry,
887
target_entry, source_path=source_path, target_path=target_path)
888
to_paths[result[0]] = result[1][1]
998
if result.versioned[0]:
1000
892
if pb is not None:
1001
893
pb.update('comparing files', entry_count, num_entries)
1002
894
if changes or include_unchanged:
1003
895
if specific_files is not None:
1004
precise_file_ids.add(result.parent_id[1])
1005
changed_file_ids.append(result.file_id)
896
new_parent_id = result[4][1]
897
precise_file_ids.add(new_parent_id)
898
changed_file_ids.append(result[0])
1007
900
# Ensure correct behaviour for reparented/added specific files.
1008
901
if specific_files is not None:
1009
902
# Record output dirs
1010
if result.kind[1] == 'directory':
1011
seen_dirs.add(result.file_id)
903
if result[6][1] == 'directory':
904
seen_dirs.add(result[0])
1012
905
# Record parents of reparented/added entries.
1013
if not result.versioned[0] or result.is_reparented():
1014
seen_parents.add(result.parent_id[1])
906
versioned = result[3]
908
if not versioned[0] or parents[0] != parents[1]:
909
seen_parents.add(parents[1])
1015
910
while all_unversioned:
1016
911
# yield any trailing unversioned paths
1017
912
unversioned_path = all_unversioned.popleft()
1018
913
to_kind, to_executable, to_stat = \
1019
914
self.target._comparison_data(fake_entry, unversioned_path[1])
1021
None, (None, unversioned_path[1]), True, (False, False),
1023
(None, unversioned_path[0][-1]),
1025
(None, to_executable))
915
yield (None, (None, unversioned_path[1]), True, (False, False),
917
(None, unversioned_path[0][-1]),
919
(None, to_executable))
1026
920
# Yield all remaining source paths
1027
921
for path, from_entry in from_entries_by_dir:
1028
922
file_id = from_entry.file_id
1029
923
if file_id in to_paths:
1030
924
# already returned
1032
to_path = self.find_target_path(path)
926
to_path = find_previous_path(self.source, self.target, path)
1033
927
entry_count += 1
1034
928
if pb is not None:
1035
929
pb.update('comparing files', entry_count, num_entries)
1183
1077
# Fall back to SHA1 for now
1184
1078
if source_verifier_kind != "SHA1":
1185
1079
source_sha1 = self.source.get_file_sha1(
1186
source_path, source_stat)
1080
source_path, source_file_id, source_stat)
1188
1082
source_sha1 = source_verifier_data
1189
1083
if target_verifier_kind != "SHA1":
1190
1084
target_sha1 = self.target.get_file_sha1(
1191
target_path, target_stat)
1085
target_path, target_file_id, target_stat)
1193
1087
target_sha1 = target_verifier_data
1194
1088
return (source_sha1 == target_sha1)
1196
def find_target_path(self, path, recurse='none'):
1197
"""Find target tree path.
1199
:param path: Path to search for (exists in source)
1200
:return: path in target, or None if there is no equivalent path.
1201
:raise NoSuchFile: If the path doesn't exist in source
1203
file_id = self.source.path2id(path)
1205
raise errors.NoSuchFile(path)
1207
return self.target.id2path(file_id, recurse=recurse)
1208
except errors.NoSuchId:
1211
def find_source_path(self, path, recurse='none'):
1212
"""Find the source tree path.
1214
:param path: Path to search for (exists in target)
1215
:return: path in source, or None if there is no equivalent path.
1216
:raise NoSuchFile: if the path doesn't exist in target
1218
file_id = self.target.path2id(path)
1220
raise errors.NoSuchFile(path)
1222
return self.source.id2path(file_id, recurse=recurse)
1223
except errors.NoSuchId:
1226
def find_target_paths(self, paths, recurse='none'):
1227
"""Find target tree paths.
1229
:param paths: Iterable over paths in target to search for
1230
:return: Dictionary mapping from source paths to paths in target , or
1231
None if there is no equivalent path.
1235
ret[path] = self.find_target_path(path, recurse=recurse)
1238
def find_source_paths(self, paths, recurse='none'):
1239
"""Find source tree paths.
1241
:param paths: Iterable over paths in target to search for
1242
:return: Dictionary mapping from target paths to paths in source, or
1243
None if there is no equivalent path.
1247
ret[path] = self.find_source_path(path, recurse=recurse)
1251
1091
InterTree.register_optimiser(InterTree)
1254
def find_previous_paths(from_tree, to_tree, paths, recurse='none'):
1094
class MultiWalker(object):
1095
"""Walk multiple trees simultaneously, getting combined results."""
1097
# Note: This could be written to not assume you can do out-of-order
1098
# lookups. Instead any nodes that don't match in all trees could be
1099
# marked as 'deferred', and then returned in the final cleanup loop.
1100
# For now, I think it is "nicer" to return things as close to the
1101
# "master_tree" order as we can.
1103
def __init__(self, master_tree, other_trees):
1104
"""Create a new MultiWalker.
1106
All trees being walked must implement "iter_entries_by_dir()", such
1107
that they yield (path, object) tuples, where that object will have a
1108
'.file_id' member, that can be used to check equality.
1110
:param master_tree: All trees will be 'slaved' to the master_tree such
1111
that nodes in master_tree will be used as 'first-pass' sync points.
1112
Any nodes that aren't in master_tree will be merged in a second
1114
:param other_trees: A list of other trees to walk simultaneously.
1116
self._master_tree = master_tree
1117
self._other_trees = other_trees
1119
# Keep track of any nodes that were properly processed just out of
1120
# order, that way we don't return them at the end, we don't have to
1121
# track *all* processed file_ids, just the out-of-order ones
1122
self._out_of_order_processed = set()
1125
def _step_one(iterator):
1126
"""Step an iter_entries_by_dir iterator.
1128
:return: (has_more, path, ie)
1129
If has_more is False, path and ie will be None.
1132
path, ie = next(iterator)
1133
except StopIteration:
1134
return False, None, None
1136
return True, path, ie
1139
def _lt_path_by_dirblock(path1, path2):
1140
"""Compare two paths based on what directory they are in.
1142
This generates a sort order, such that all children of a directory are
1143
sorted together, and grandchildren are in the same order as the
1144
children appear. But all grandchildren come after all children.
1146
:param path1: first path
1147
:param path2: the second path
1148
:return: negative number if ``path1`` comes first,
1149
0 if paths are equal
1150
and a positive number if ``path2`` sorts first
1152
# Shortcut this special case
1155
# This is stolen from _dirstate_helpers_py.py, only switching it to
1156
# Unicode objects. Consider using encode_utf8() and then using the
1157
# optimized versions, or maybe writing optimized unicode versions.
1158
if not isinstance(path1, text_type):
1159
raise TypeError("'path1' must be a unicode string, not %s: %r"
1160
% (type(path1), path1))
1161
if not isinstance(path2, text_type):
1162
raise TypeError("'path2' must be a unicode string, not %s: %r"
1163
% (type(path2), path2))
1164
return (MultiWalker._path_to_key(path1) <
1165
MultiWalker._path_to_key(path2))
1168
def _path_to_key(path):
1169
dirname, basename = osutils.split(path)
1170
return (dirname.split(u'/'), basename)
1172
def _lookup_by_file_id(self, extra_entries, other_tree, file_id):
1173
"""Lookup an inventory entry by file_id.
1175
This is called when an entry is missing in the normal order.
1176
Generally this is because a file was either renamed, or it was
1177
deleted/added. If the entry was found in the inventory and not in
1178
extra_entries, it will be added to self._out_of_order_processed
1180
:param extra_entries: A dictionary of {file_id: (path, ie)}. This
1181
should be filled with entries that were found before they were
1182
used. If file_id is present, it will be removed from the
1184
:param other_tree: The Tree to search, in case we didn't find the entry
1186
:param file_id: The file_id to look for
1187
:return: (path, ie) if found or (None, None) if not present.
1189
if file_id in extra_entries:
1190
return extra_entries.pop(file_id)
1191
# TODO: Is id2path better as the first call, or is
1192
# inventory[file_id] better as a first check?
1194
cur_path = other_tree.id2path(file_id)
1195
except errors.NoSuchId:
1197
if cur_path is None:
1200
self._out_of_order_processed.add(file_id)
1201
cur_ie = other_tree.root_inventory.get_entry(file_id)
1202
return (cur_path, cur_ie)
1205
"""Match up the values in the different trees."""
1206
for result in self._walk_master_tree():
1208
self._finish_others()
1209
for result in self._walk_others():
1212
def _walk_master_tree(self):
1213
"""First pass, walk all trees in lock-step.
1215
When we are done, all nodes in the master_tree will have been
1216
processed. _other_walkers, _other_entries, and _others_extra will be
1217
set on 'self' for future processing.
1219
# This iterator has the most "inlining" done, because it tends to touch
1220
# every file in the tree, while the others only hit nodes that don't
1222
master_iterator = self._master_tree.iter_entries_by_dir()
1224
other_walkers = [other.iter_entries_by_dir()
1225
for other in self._other_trees]
1226
other_entries = [self._step_one(walker) for walker in other_walkers]
1227
# Track extra nodes in the other trees
1228
others_extra = [{} for _ in range(len(self._other_trees))]
1230
master_has_more = True
1231
step_one = self._step_one
1232
lookup_by_file_id = self._lookup_by_file_id
1233
out_of_order_processed = self._out_of_order_processed
1235
while master_has_more:
1236
(master_has_more, path, master_ie) = step_one(master_iterator)
1237
if not master_has_more:
1240
file_id = master_ie.file_id
1242
other_values_append = other_values.append
1243
next_other_entries = []
1244
next_other_entries_append = next_other_entries.append
1245
for idx, (other_has_more, other_path, other_ie) in enumerate(other_entries):
1246
if not other_has_more:
1247
other_values_append(lookup_by_file_id(
1248
others_extra[idx], self._other_trees[idx], file_id))
1249
next_other_entries_append((False, None, None))
1250
elif file_id == other_ie.file_id:
1251
# This is the critical code path, as most of the entries
1252
# should match between most trees.
1253
other_values_append((other_path, other_ie))
1254
next_other_entries_append(step_one(other_walkers[idx]))
1256
# This walker did not match, step it until it either
1257
# matches, or we know we are past the current walker.
1258
other_walker = other_walkers[idx]
1259
other_extra = others_extra[idx]
1260
while (other_has_more and
1261
self._lt_path_by_dirblock(other_path, path)):
1262
other_file_id = other_ie.file_id
1263
if other_file_id not in out_of_order_processed:
1264
other_extra[other_file_id] = (other_path, other_ie)
1265
other_has_more, other_path, other_ie = \
1266
step_one(other_walker)
1267
if other_has_more and other_ie.file_id == file_id:
1268
# We ended up walking to this point, match and step
1270
other_values_append((other_path, other_ie))
1271
other_has_more, other_path, other_ie = \
1272
step_one(other_walker)
1274
# This record isn't in the normal order, see if it
1276
other_values_append(lookup_by_file_id(
1277
other_extra, self._other_trees[idx], file_id))
1278
next_other_entries_append((other_has_more, other_path,
1280
other_entries = next_other_entries
1282
# We've matched all the walkers, yield this datapoint
1283
yield path, file_id, master_ie, other_values
1284
self._other_walkers = other_walkers
1285
self._other_entries = other_entries
1286
self._others_extra = others_extra
1288
def _finish_others(self):
1289
"""Finish walking the other iterators, so we get all entries."""
1290
for idx, info in enumerate(self._other_entries):
1291
other_extra = self._others_extra[idx]
1292
(other_has_more, other_path, other_ie) = info
1293
while other_has_more:
1294
other_file_id = other_ie.file_id
1295
if other_file_id not in self._out_of_order_processed:
1296
other_extra[other_file_id] = (other_path, other_ie)
1297
other_has_more, other_path, other_ie = \
1298
self._step_one(self._other_walkers[idx])
1299
del self._other_entries
1301
def _walk_others(self):
1302
"""Finish up by walking all the 'deferred' nodes."""
1303
# TODO: One alternative would be to grab all possible unprocessed
1304
# file_ids, and then sort by path, and then yield them. That
1305
# might ensure better ordering, in case a caller strictly
1306
# requires parents before children.
1307
for idx, other_extra in enumerate(self._others_extra):
1308
others = sorted(viewvalues(other_extra),
1309
key=lambda x: self._path_to_key(x[0]))
1310
for other_path, other_ie in others:
1311
file_id = other_ie.file_id
1312
# We don't need to check out_of_order_processed here, because
1313
# the lookup_by_file_id will be removing anything processed
1314
# from the extras cache
1315
other_extra.pop(file_id)
1316
other_values = [(None, None)] * idx
1317
other_values.append((other_path, other_ie))
1318
for alt_idx, alt_extra in enumerate(self._others_extra[idx + 1:]):
1319
alt_idx = alt_idx + idx + 1
1320
alt_extra = self._others_extra[alt_idx]
1321
alt_tree = self._other_trees[alt_idx]
1322
other_values.append(self._lookup_by_file_id(
1323
alt_extra, alt_tree, file_id))
1324
yield other_path, file_id, None, other_values
1327
def find_previous_paths(from_tree, to_tree, paths):
1255
1328
"""Find previous tree paths.
1257
1330
:param from_tree: From tree
1258
1331
:param to_tree: To tree
1259
:param paths: Iterable over paths in from_tree to search for
1332
:param paths: Iterable over paths to search for
1260
1333
:return: Dictionary mapping from from_tree paths to paths in to_tree, or
1261
1334
None if there is no equivalent path.
1263
return InterTree.get(to_tree, from_tree).find_source_paths(paths, recurse=recurse)
1266
def find_previous_path(from_tree, to_tree, path, recurse='none'):
1338
ret[path] = find_previous_path(from_tree, to_tree, path)
1342
def find_previous_path(from_tree, to_tree, path, file_id=None):
1267
1343
"""Find previous tree path.
1269
1345
:param from_tree: From tree
1270
1346
:param to_tree: To tree
1271
:param path: Path to search for (exists in from_tree)
1347
:param path: Path to search for
1272
1348
:return: path in to_tree, or None if there is no equivalent path.
1273
:raise NoSuchFile: If the path doesn't exist in from_tree
1275
return InterTree.get(to_tree, from_tree).find_source_path(
1276
path, recurse=recurse)
1351
file_id = from_tree.path2id(path)
1353
raise errors.NoSuchFile(path)
1355
return to_tree.id2path(file_id)
1356
except errors.NoSuchId:
1279
1360
def get_canonical_path(tree, path, normalize):
added
removed
breezy/tree.py
