1
# Copyright (C) 2005, 2006, 2008, 2009 Canonical Ltd
3
# This program is free software; you can redistribute it and/or modify
4
# it under the terms of the GNU General Public License as published by
5
# the Free Software Foundation; either version 2 of the License, or
6
# (at your option) any later version.
8
# This program is distributed in the hope that it will be useful,
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
# GNU General Public License for more details.
13
# You should have received a copy of the GNU General Public License
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
18
"""Copying of history from one branch to another.
20
The basic plan is that every branch knows the history of everything
21
that has merged into it. As the first step of a merge, pull, or
22
branch operation we copy history from the source into the destination
28
from bzrlib.lazy_import import lazy_import
29
lazy_import(globals(), """
40
from bzrlib.revision import NULL_REVISION
41
from bzrlib.trace import mutter
45
class RepoFetcher(object):
46
"""Pull revisions and texts from one repository to another.
48
This should not be used directly, it's essential a object to encapsulate
49
the logic in InterRepository.fetch().
52
def __init__(self, to_repository, from_repository, last_revision=None,
53
pb=None, find_ghosts=True, fetch_spec=None):
54
"""Create a repo fetcher.
56
:param last_revision: If set, try to limit to the data this revision
58
:param find_ghosts: If True search the entire history for ghosts.
59
:param pb: ProgressBar object to use; deprecated and ignored.
60
This method will just create one on top of the stack.
63
symbol_versioning.warn(
64
symbol_versioning.deprecated_in((1, 14, 0))
65
% "pb parameter to RepoFetcher.__init__")
66
# and for simplicity it is in fact ignored
67
if to_repository.has_same_location(from_repository):
68
# repository.fetch should be taking care of this case.
69
raise errors.BzrError('RepoFetcher run '
70
'between two objects at the same location: '
71
'%r and %r' % (to_repository, from_repository))
72
self.to_repository = to_repository
73
self.from_repository = from_repository
74
self.sink = to_repository._get_sink()
75
# must not mutate self._last_revision as its potentially a shared instance
76
self._last_revision = last_revision
77
self._fetch_spec = fetch_spec
78
self.find_ghosts = find_ghosts
79
self.from_repository.lock_read()
80
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
81
self.from_repository, self.from_repository._format,
82
self.to_repository, self.to_repository._format)
86
self.from_repository.unlock()
89
"""Primary worker function.
91
This initialises all the needed variables, and then fetches the
92
requested revisions, finally clearing the progress bar.
94
# Roughly this is what we're aiming for fetch to become:
96
# missing = self.sink.insert_stream(self.source.get_stream(search))
98
# missing = self.sink.insert_stream(self.source.get_items(missing))
101
self.file_ids_names = {}
102
pb = bzrlib.ui.ui_factory.nested_progress_bar()
103
pb.show_pct = pb.show_count = False
105
pb.update("Finding revisions", 0, 2)
106
search = self._revids_to_fetch()
109
pb.update("Fetching revisions", 1, 2)
110
self._fetch_everything_for_search(search)
114
def _fetch_everything_for_search(self, search):
115
"""Fetch all data for the given set of revisions."""
116
# The first phase is "file". We pass the progress bar for it directly
117
# into item_keys_introduced_by, which has more information about how
118
# that phase is progressing than we do. Progress updates for the other
119
# phases are taken care of in this function.
120
# XXX: there should be a clear owner of the progress reporting. Perhaps
121
# item_keys_introduced_by should have a richer API than it does at the
122
# moment, so that it can feed the progress information back to this
124
if (self.from_repository._format.rich_root_data and
125
not self.to_repository._format.rich_root_data):
126
raise errors.IncompatibleRepositories(
127
self.from_repository, self.to_repository,
128
"different rich-root support")
129
pb = bzrlib.ui.ui_factory.nested_progress_bar()
131
pb.update("Get stream source")
132
source = self.from_repository._get_source(
133
self.to_repository._format)
134
stream = source.get_stream(search)
135
from_format = self.from_repository._format
136
pb.update("Inserting stream")
137
resume_tokens, missing_keys = self.sink.insert_stream(
138
stream, from_format, [])
139
if self.to_repository._fallback_repositories:
141
self._parent_inventories(search.get_keys()))
143
pb.update("Missing keys")
144
stream = source.get_stream_for_missing_keys(missing_keys)
145
pb.update("Inserting missing keys")
146
resume_tokens, missing_keys = self.sink.insert_stream(
147
stream, from_format, resume_tokens)
149
raise AssertionError(
150
"second push failed to complete a fetch %r." % (
153
raise AssertionError(
154
"second push failed to commit the fetch %r." % (
156
pb.update("Finishing stream")
161
def _revids_to_fetch(self):
162
"""Determines the exact revisions needed from self.from_repository to
163
install self._last_revision in self.to_repository.
165
If no revisions need to be fetched, then this just returns None.
167
if self._fetch_spec is not None:
168
return self._fetch_spec
169
mutter('fetch up to rev {%s}', self._last_revision)
170
if self._last_revision is NULL_REVISION:
171
# explicit limit of no revisions needed
173
return self.to_repository.search_missing_revision_ids(
174
self.from_repository, self._last_revision,
175
find_ghosts=self.find_ghosts)
177
def _parent_inventories(self, revision_ids):
178
# Find all the parent revisions referenced by the stream, but
179
# not present in the stream, and make sure we send their
181
parent_maps = self.to_repository.get_parent_map(revision_ids)
183
map(parents.update, parent_maps.itervalues())
184
parents.discard(NULL_REVISION)
185
parents.difference_update(revision_ids)
186
missing_keys = set(('inventories', rev_id) for rev_id in parents)
190
class Inter1and2Helper(object):
191
"""Helper for operations that convert data from model 1 and 2
193
This is for use by fetchers and converters.
196
def __init__(self, source):
199
:param source: The repository data comes from
203
def iter_rev_trees(self, revs):
204
"""Iterate through RevisionTrees efficiently.
206
Additionally, the inventory's revision_id is set if unset.
208
Trees are retrieved in batches of 100, and then yielded in the order
211
:param revs: A list of revision ids
213
# In case that revs is not a list.
216
for tree in self.source.revision_trees(revs[:100]):
217
if tree.inventory.revision_id is None:
218
tree.inventory.revision_id = tree.get_revision_id()
222
def _find_root_ids(self, revs, parent_map, graph):
224
for tree in self.iter_rev_trees(revs):
225
revision_id = tree.inventory.root.revision
226
root_id = tree.get_root_id()
227
revision_root[revision_id] = root_id
228
# Find out which parents we don't already know root ids for
230
for revision_parents in parent_map.itervalues():
231
parents.update(revision_parents)
232
parents.difference_update(revision_root.keys() + [NULL_REVISION])
233
# Limit to revisions present in the versionedfile
234
parents = graph.get_parent_map(parents).keys()
235
for tree in self.iter_rev_trees(parents):
236
root_id = tree.get_root_id()
237
revision_root[tree.get_revision_id()] = root_id
240
def generate_root_texts(self, revs):
241
"""Generate VersionedFiles for all root ids.
243
:param revs: the revisions to include
245
graph = self.source.get_graph()
246
parent_map = graph.get_parent_map(revs)
247
rev_order = tsort.topo_sort(parent_map)
248
rev_id_to_root_id = self._find_root_ids(revs, parent_map, graph)
249
root_id_order = [(rev_id_to_root_id[rev_id], rev_id) for rev_id in
251
# Guaranteed stable, this groups all the file id operations together
252
# retaining topological order within the revisions of a file id.
253
# File id splits and joins would invalidate this, but they don't exist
254
# yet, and are unlikely to in non-rich-root environments anyway.
255
root_id_order.sort(key=operator.itemgetter(0))
256
# Create a record stream containing the roots to create.
257
new_roots_stream = _new_root_data_stream(
258
root_id_order, rev_id_to_root_id, parent_map, self.source)
259
return [('texts', new_roots_stream)]
262
def _new_root_data_stream(
263
root_keys_to_create, rev_id_to_root_id_map, parent_map, repo):
264
for root_key in root_keys_to_create:
265
root_id, rev_id = root_key
266
parent_keys = _parent_keys_for_root_version(
267
root_id, rev_id, rev_id_to_root_id_map, parent_map, repo)
268
yield versionedfile.FulltextContentFactory(
269
root_key, parent_keys, None, '')
272
def _parent_keys_for_root_version(
273
root_id, rev_id, rev_id_to_root_id_map, parent_map, repo):
274
"""Get the parent keys for a given root id."""
275
# Include direct parents of the revision, but only if they used the same
276
# root_id and are heads.
277
rev_parents = parent_map[rev_id]
279
for parent_id in rev_parents:
280
if parent_id == NULL_REVISION:
282
if parent_id not in rev_id_to_root_id_map:
283
# We probably didn't read this revision, go spend the extra effort
286
tree = repo.revision_tree(parent_id)
287
except errors.NoSuchRevision:
288
# Ghost, fill out rev_id_to_root_id in case we encounter this
290
# But set parent_root_id to None since we don't really know
291
parent_root_id = None
293
parent_root_id = tree.get_root_id()
294
rev_id_to_root_id_map[parent_id] = None
296
# rev_id_to_root_id_map[parent_id] = parent_root_id
297
# memory consumption maybe?
299
parent_root_id = rev_id_to_root_id_map[parent_id]
300
if root_id == parent_root_id:
301
# With stacking we _might_ want to refer to a non-local revision,
302
# but this code path only applies when we have the full content
303
# available, so ghosts really are ghosts, not just the edge of
305
parent_ids.append(parent_id)
307
# root_id may be in the parent anyway.
309
tree = repo.revision_tree(parent_id)
310
except errors.NoSuchRevision:
311
# ghost, can't refer to it.
315
parent_ids.append(tree.inventory[root_id].revision)
316
except errors.NoSuchId:
319
# Drop non-head parents
320
heads = repo.get_graph().heads(parent_ids)
322
for parent_id in parent_ids:
323
if parent_id in heads and parent_id not in selected_ids:
324
selected_ids.append(parent_id)
325
parent_keys = [(root_id, parent_id) for parent_id in selected_ids]
added
removed
bzrlib/fetch.py
