1
# Copyright (C) 2005 by 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., 59 Temple Place, Suite 330, Boston, MA 02111-1307 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
25
The copying is done in a slightly complicated order. We don't want to
26
add a revision to the store until everything it refers to is also
27
stored, so that if a revision is present we can totally recreate it.
28
However, we can't know what files are included in a revision until we
29
read its inventory. So we query the inventory store of the source for
30
the ids we need, and then pull those ids and finally actually join
35
import bzrlib.errors as errors
36
from bzrlib.errors import (InstallFailed, NoSuchRevision,
38
from bzrlib.trace import mutter
39
from bzrlib.reconcile import RepoReconciler
40
from bzrlib.revision import NULL_REVISION
41
from bzrlib.symbol_versioning import *
44
# TODO: Avoid repeatedly opening weaves so many times.
46
# XXX: This doesn't handle ghost (not present in branch) revisions at
47
# all yet. I'm not sure they really should be supported.
49
# NOTE: This doesn't copy revisions which may be present but not
50
# merged into the last revision. I'm not sure we want to do that.
52
# - get a list of revisions that need to be pulled in
53
# - for each one, pull in that revision file
54
# and get the inventory, and store the inventory with right
56
# - and get the ancestry, and store that with right parents too
57
# - and keep a note of all file ids and version seen
58
# - then go through all files; for each one get the weave,
59
# and add in all file versions
62
@deprecated_function(zero_eight)
63
def greedy_fetch(to_branch, from_branch, revision=None, pb=None):
64
"""Legacy API, please see branch.fetch(from_branch, last_revision, pb)."""
65
f = Fetcher(to_branch, from_branch, revision, pb)
66
return f.count_copied, f.failed_revisions
71
class RepoFetcher(object):
72
"""Pull revisions and texts from one repository to another.
75
if set, try to limit to the data this revision references.
78
count_copied -- number of revisions copied
80
This should not be used directory, its essential a object to encapsulate
81
the logic in InterRepository.fetch().
83
def __init__(self, to_repository, from_repository, last_revision=None, pb=None):
85
self.failed_revisions = []
87
if to_repository.control_files._transport.base == from_repository.control_files._transport.base:
88
# check that last_revision is in 'from' and then return a no-operation.
89
if last_revision not in (None, NULL_REVISION):
90
from_repository.get_revision(last_revision)
92
self.to_repository = to_repository
93
self.from_repository = from_repository
94
# must not mutate self._last_revision as its potentially a shared instance
95
self._last_revision = last_revision
97
self.pb = bzrlib.ui.ui_factory.nested_progress_bar()
98
self.nested_pb = self.pb
101
self.nested_pb = None
102
self.from_repository.lock_read()
104
self.to_repository.lock_write()
108
if self.nested_pb is not None:
109
self.nested_pb.finished()
110
self.to_repository.unlock()
112
self.from_repository.unlock()
115
"""Primary worker function.
117
This initialises all the needed variables, and then fetches the
118
requested revisions, finally clearing the progress bar.
120
self.to_weaves = self.to_repository.weave_store
121
self.to_control = self.to_repository.control_weaves
122
self.from_weaves = self.from_repository.weave_store
123
self.from_control = self.from_repository.control_weaves
125
self.file_ids_names = {}
128
revs = self._revids_to_fetch()
131
self.pb.update('Fetching text', 1, self.total_steps)
132
self._fetch_weave_texts(revs)
133
self.pb.update('Fetching inventories', 2, self.total_steps)
134
self._fetch_inventory_weave(revs)
135
self.pb.update('Fetching revisions', 3, self.total_steps)
136
self._fetch_revision_texts(revs)
137
self.pb.update('Fetching revisions', 4, self.total_steps)
138
self.count_copied += len(revs)
142
def _revids_to_fetch(self):
143
self.pb.update('Calculating needed data', 0, self.total_steps)
144
mutter('fetch up to rev {%s}', self._last_revision)
145
if self._last_revision is NULL_REVISION:
146
# explicit limit of no revisions needed
148
if (self._last_revision != None and
149
self.to_repository.has_revision(self._last_revision)):
153
return self.to_repository.missing_revision_ids(self.from_repository,
155
except errors.NoSuchRevision:
156
raise InstallFailed([self._last_revision])
158
def _fetch_weave_texts(self, revs):
159
texts_pb = bzrlib.ui.ui_factory.nested_progress_bar()
161
file_ids = self.from_repository.fileid_involved_by_set(revs)
163
num_file_ids = len(file_ids)
164
for file_id in file_ids:
165
texts_pb.update("fetch texts", count, num_file_ids)
168
to_weave = self.to_weaves.get_weave(file_id,
169
self.to_repository.get_transaction())
170
except errors.NoSuchFile:
171
# destination is empty, just copy it.
172
# this copies all the texts, which is useful and
173
# on per-file basis quite cheap.
174
self.to_weaves.copy_multi(
178
self.from_repository.get_transaction(),
179
self.to_repository.get_transaction())
181
# destination has contents, must merge
182
from_weave = self.from_weaves.get_weave(file_id,
183
self.from_repository.get_transaction())
184
# we fetch all the texts, because texts do
185
# not reference anything, and its cheap enough
186
to_weave.join(from_weave)
190
def _fetch_inventory_weave(self, revs):
191
inv_pb = bzrlib.ui.ui_factory.nested_progress_bar()
193
inv_pb.update("fetch inventory", 0, 2)
194
to_weave = self.to_control.get_weave('inventory',
195
self.to_repository.get_transaction())
197
# just merge, this is optimisable and its means we dont
198
# copy unreferenced data such as not-needed inventories.
199
self.pb.update("fetch inventory", 1, 2)
200
from_weave = self.from_repository.get_inventory_weave()
201
self.pb.update("fetch inventory", 2, 2)
202
# we fetch only the referenced inventories because we do not
203
# know for unselected inventories whether all their required
204
# texts are present in the other repository - it could be
206
to_weave.join(from_weave, msg='fetch inventory', version_ids=revs)
211
class GenericRepoFetcher(RepoFetcher):
212
"""This is a generic repo to repo fetcher.
214
This makes minimal assumptions about repo layout and contents.
215
It triggers a reconciliation after fetching to ensure integrity.
218
def _fetch_revision_texts(self, revs):
219
rev_pb = bzrlib.ui.ui_factory.nested_progress_bar()
221
self.to_transaction = self.to_repository.get_transaction()
225
rev_pb.update('fetch revisions', count, total)
227
sig_text = self.from_repository.get_signature_text(rev)
228
self.to_repository._revision_store.add_revision_signature_text(
229
rev, sig_text, self.to_transaction)
230
except errors.NoSuchRevision:
233
self.to_repository._revision_store.add_revision(
234
self.from_repository.get_revision(rev),
237
rev_pb.update('copying revisions', count, total)
238
# fixup inventory if needed:
239
# this is expensive because we have no inverse index to current ghosts.
240
# but on local disk its a few seconds and sftp push is already insane.
242
# FIXME: repository should inform if this is needed.
243
self.to_repository.reconcile()
248
class KnitRepoFetcher(RepoFetcher):
249
"""This is a knit format repository specific fetcher.
251
This differs from the GenericRepoFetcher by not doing a
252
reconciliation after copying, and using knit joining to
256
def _fetch_revision_texts(self, revs):
257
# may need to be a InterRevisionStore call here.
258
from_transaction = self.from_repository.get_transaction()
259
to_transaction = self.to_repository.get_transaction()
260
to_sf = self.to_repository._revision_store.get_signature_file(
262
from_sf = self.from_repository._revision_store.get_signature_file(
264
to_sf.join(from_sf, version_ids=revs, ignore_missing=True)
265
to_rf = self.to_repository._revision_store.get_revision_file(
267
from_rf = self.from_repository._revision_store.get_revision_file(
269
to_rf.join(from_rf, version_ids=revs)
272
class Fetcher(object):
273
"""Backwards compatability glue for branch.fetch()."""
275
@deprecated_method(zero_eight)
276
def __init__(self, to_branch, from_branch, last_revision=None, pb=None):
277
"""Please see branch.fetch()."""
278
to_branch.fetch(from_branch, last_revision, pb)
added
removed
bzrlib/fetch.py
