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
33
from bzrlib.revision import NULL_REVISION
34
from bzrlib.tsort import topo_sort
35
from bzrlib.trace import mutter
37
from bzrlib.versionedfile import FulltextContentFactory
40
class RepoFetcher(object):
41
"""Pull revisions and texts from one repository to another.
43
This should not be used directly, it's essential a object to encapsulate
44
the logic in InterRepository.fetch().
47
def __init__(self, to_repository, from_repository, last_revision=None,
48
pb=None, find_ghosts=True, fetch_spec=None):
49
"""Create a repo fetcher.
51
:param last_revision: If set, try to limit to the data this revision
53
:param find_ghosts: If True search the entire history for ghosts.
54
:param pb: ProgressBar object to use; deprecated and ignored.
55
This method will just create one on top of the stack.
58
symbol_versioning.warn(
59
symbol_versioning.deprecated_in((1, 14, 0))
60
% "pb parameter to RepoFetcher.__init__")
61
# and for simplicity it is in fact ignored
62
# repository.fetch has the responsibility for short-circuiting
63
# attempts to copy between a repository and itself.
64
self.to_repository = to_repository
65
self.from_repository = from_repository
66
self.sink = to_repository._get_sink()
67
# must not mutate self._last_revision as its potentially a shared instance
68
self._last_revision = last_revision
69
self._fetch_spec = fetch_spec
70
self.find_ghosts = find_ghosts
71
self.from_repository.lock_read()
72
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
73
self.from_repository, self.from_repository._format,
74
self.to_repository, self.to_repository._format)
78
self.from_repository.unlock()
81
"""Primary worker function.
83
This initialises all the needed variables, and then fetches the
84
requested revisions, finally clearing the progress bar.
86
# Roughly this is what we're aiming for fetch to become:
88
# missing = self.sink.insert_stream(self.source.get_stream(search))
90
# missing = self.sink.insert_stream(self.source.get_items(missing))
93
self.file_ids_names = {}
94
pb = bzrlib.ui.ui_factory.nested_progress_bar()
95
pb.show_pct = pb.show_count = False
97
pb.update("Finding revisions", 0, 2)
98
search = self._revids_to_fetch()
101
pb.update("Fetching revisions", 1, 2)
102
self._fetch_everything_for_search(search)
106
def _fetch_everything_for_search(self, search):
107
"""Fetch all data for the given set of revisions."""
108
# The first phase is "file". We pass the progress bar for it directly
109
# into item_keys_introduced_by, which has more information about how
110
# that phase is progressing than we do. Progress updates for the other
111
# phases are taken care of in this function.
112
# XXX: there should be a clear owner of the progress reporting. Perhaps
113
# item_keys_introduced_by should have a richer API than it does at the
114
# moment, so that it can feed the progress information back to this
116
if (self.from_repository._format.rich_root_data and
117
not self.to_repository._format.rich_root_data):
118
raise errors.IncompatibleRepositories(
119
self.from_repository, self.to_repository,
120
"different rich-root support")
121
pb = bzrlib.ui.ui_factory.nested_progress_bar()
123
pb.update("Get stream source")
124
source = self.from_repository._get_source(
125
self.to_repository._format)
126
stream = source.get_stream(search)
127
from_format = self.from_repository._format
128
pb.update("Inserting stream")
129
resume_tokens, missing_keys = self.sink.insert_stream(
130
stream, from_format, [])
131
if self.to_repository._fallback_repositories:
133
self._parent_inventories(search.get_keys()))
135
pb.update("Missing keys")
136
stream = source.get_stream_for_missing_keys(missing_keys)
137
pb.update("Inserting missing keys")
138
resume_tokens, missing_keys = self.sink.insert_stream(
139
stream, from_format, resume_tokens)
141
raise AssertionError(
142
"second push failed to complete a fetch %r." % (
145
raise AssertionError(
146
"second push failed to commit the fetch %r." % (
148
pb.update("Finishing stream")
153
def _revids_to_fetch(self):
154
"""Determines the exact revisions needed from self.from_repository to
155
install self._last_revision in self.to_repository.
157
If no revisions need to be fetched, then this just returns None.
159
if self._fetch_spec is not None:
160
return self._fetch_spec
161
mutter('fetch up to rev {%s}', self._last_revision)
162
if self._last_revision is NULL_REVISION:
163
# explicit limit of no revisions needed
165
return self.to_repository.search_missing_revision_ids(
166
self.from_repository, self._last_revision,
167
find_ghosts=self.find_ghosts)
169
def _parent_inventories(self, revision_ids):
170
# Find all the parent revisions referenced by the stream, but
171
# not present in the stream, and make sure we send their
173
parent_maps = self.to_repository.get_parent_map(revision_ids)
175
map(parents.update, parent_maps.itervalues())
176
parents.discard(NULL_REVISION)
177
parents.difference_update(revision_ids)
178
missing_keys = set(('inventories', rev_id) for rev_id in parents)
182
class Inter1and2Helper(object):
183
"""Helper for operations that convert data from model 1 and 2
185
This is for use by fetchers and converters.
188
def __init__(self, source):
191
:param source: The repository data comes from
195
def iter_rev_trees(self, revs):
196
"""Iterate through RevisionTrees efficiently.
198
Additionally, the inventory's revision_id is set if unset.
200
Trees are retrieved in batches of 100, and then yielded in the order
203
:param revs: A list of revision ids
205
# In case that revs is not a list.
208
for tree in self.source.revision_trees(revs[:100]):
209
if tree.inventory.revision_id is None:
210
tree.inventory.revision_id = tree.get_revision_id()
214
def _find_root_ids(self, revs, parent_map, graph):
216
planned_versions = {}
217
for tree in self.iter_rev_trees(revs):
218
revision_id = tree.inventory.root.revision
219
root_id = tree.get_root_id()
220
planned_versions.setdefault(root_id, []).append(revision_id)
221
revision_root[revision_id] = root_id
222
# Find out which parents we don't already know root ids for
224
for revision_parents in parent_map.itervalues():
225
parents.update(revision_parents)
226
parents.difference_update(revision_root.keys() + [NULL_REVISION])
227
# Limit to revisions present in the versionedfile
228
parents = graph.get_parent_map(parents).keys()
229
for tree in self.iter_rev_trees(parents):
230
root_id = tree.get_root_id()
231
revision_root[tree.get_revision_id()] = root_id
232
return revision_root, planned_versions
234
def generate_root_texts(self, revs):
235
"""Generate VersionedFiles for all root ids.
237
:param revs: the revisions to include
239
graph = self.source.get_graph()
240
parent_map = graph.get_parent_map(revs)
241
rev_order = topo_sort(parent_map)
242
rev_id_to_root_id, root_id_to_rev_ids = self._find_root_ids(
243
revs, parent_map, graph)
244
root_id_order = [(rev_id_to_root_id[rev_id], rev_id) for rev_id in
246
# Guaranteed stable, this groups all the file id operations together
247
# retaining topological order within the revisions of a file id.
248
# File id splits and joins would invalidate this, but they don't exist
249
# yet, and are unlikely to in non-rich-root environments anyway.
250
root_id_order.sort(key=operator.itemgetter(0))
251
# Create a record stream containing the roots to create.
253
for key in root_id_order:
254
root_id, rev_id = key
255
rev_parents = parent_map[rev_id]
256
# We drop revision parents with different file-ids, because
257
# that represents a rename of the root to a different location
258
# - its not actually a parent for us. (We could look for that
259
# file id in the revision tree at considerably more expense,
260
# but for now this is sufficient (and reconcile will catch and
261
# correct this anyway).
262
# When a parent revision is a ghost, we guess that its root id
263
# was unchanged (rather than trimming it from the parent list).
264
parent_keys = tuple((root_id, parent) for parent in rev_parents
265
if parent != NULL_REVISION and
266
rev_id_to_root_id.get(parent, root_id) == root_id)
267
yield FulltextContentFactory(key, parent_keys, None, '')
268
return [('texts', yield_roots())]
added
removed
bzrlib/fetch.py
