1
# (C) 2005, 2006 Canonical Limited.
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
17
"""Reconcilers are able to fix some potential data errors in a branch."""
20
__all__ = ['reconcile', 'Reconciler', 'RepoReconciler']
24
import bzrlib.errors as errors
25
import bzrlib.progress
26
from bzrlib.trace import mutter
27
from bzrlib.tsort import TopoSorter
28
import bzrlib.ui as ui
32
"""Reconcile the data in dir.
34
Currently this is limited to a inventory 'reweave'.
36
This is a convenience method, for using a Reconciler object.
38
Directly using Reconciler is recommended for library users that
39
desire fine grained control or analysis of the found issues.
41
reconciler = Reconciler(dir)
42
reconciler.reconcile()
45
class Reconciler(object):
46
"""Reconcilers are used to reconcile existing data."""
48
def __init__(self, dir):
52
"""Perform reconciliation.
54
After reconciliation the following attributes document found issues:
55
inconsistent_parents: The number of revisions in the repository whose
56
ancestry was being reported incorrectly.
57
garbage_inventories: The number of inventory objects without revisions
58
that were garbage collected.
60
self.pb = ui.ui_factory.nested_progress_bar()
67
"""Helper function for performing reconciliation."""
68
self.repo = self.bzrdir.find_repository()
69
self.pb.note('Reconciling repository %s',
70
self.repo.bzrdir.root_transport.base)
71
repo_reconciler = RepoReconciler(self.repo)
72
repo_reconciler.reconcile()
73
self.inconsistent_parents = repo_reconciler.inconsistent_parents
74
self.garbage_inventories = repo_reconciler.garbage_inventories
75
self.pb.note('Reconciliation complete.')
78
class RepoReconciler(object):
79
"""Reconciler that reconciles a repository.
81
Currently this consists of an inventory reweave with revision cross-checks.
84
def __init__(self, repo):
88
"""Perform reconciliation.
90
After reconciliation the following attributes document found issues:
91
inconsistent_parents: The number of revisions in the repository whose
92
ancestry was being reported incorrectly.
93
garbage_inventories: The number of inventory objects without revisions
94
that were garbage collected.
96
self.repo.lock_write()
98
self.pb = ui.ui_factory.nested_progress_bar()
100
self._reconcile_steps()
106
def _reconcile_steps(self):
107
"""Perform the steps to reconcile this repository."""
108
self._reweave_inventory()
110
def _reweave_inventory(self):
111
"""Regenerate the inventory weave for the repository from scratch."""
112
# local because its really a wart we want to hide
113
from bzrlib.weave import WeaveFile, Weave
114
transaction = self.repo.get_transaction()
115
self.pb.update('Reading inventory data.')
116
self.inventory = self.repo.get_inventory_weave()
117
# the total set of revisions to process
118
self.pending = set([rev_id for rev_id in self.repo._revision_store.all_revision_ids(transaction)])
120
# mapping from revision_id to parents
122
# errors that we detect
123
self.inconsistent_parents = 0
124
# we need the revision id of each revision and its available parents list
125
self._setup_steps(len(self.pending))
126
for rev_id in self.pending:
127
# put a revision into the graph.
128
self._graph_revision(rev_id)
129
self._check_garbage_inventories()
130
if not self.inconsistent_parents and not self.garbage_inventories:
131
self.pb.note('Inventory ok.')
133
self.pb.update('Backing up inventory...', 0, 0)
134
self.repo.control_weaves.copy(self.inventory, 'inventory.backup', self.repo.get_transaction())
135
self.pb.note('Backup Inventory created.')
136
# asking for '' should never return a non-empty weave
137
new_inventory_vf = self.repo.control_weaves.get_empty('inventory.new',
138
self.repo.get_transaction())
140
# we have topological order of revisions and non ghost parents ready.
141
self._setup_steps(len(self._rev_graph))
142
for rev_id in TopoSorter(self._rev_graph.items()).iter_topo_order():
143
parents = self._rev_graph[rev_id]
144
# double check this really is in topological order.
145
unavailable = [p for p in parents if p not in new_inventory_vf]
146
assert len(unavailable) == 0
147
# this entry has all the non ghost parents in the inventory
149
self._reweave_step('adding inventories')
150
if isinstance(new_inventory_vf, WeaveFile):
151
# It's really a WeaveFile, but we call straight into the
152
# Weave's add method to disable the auto-write-out behaviour.
153
new_inventory_vf._check_write_ok()
154
Weave._add_lines(new_inventory_vf, rev_id, parents, self.inventory.get_lines(rev_id),
157
new_inventory_vf.add_lines(rev_id, parents, self.inventory.get_lines(rev_id))
159
if isinstance(new_inventory_vf, WeaveFile):
160
new_inventory_vf._save()
161
# if this worked, the set of new_inventory_vf.names should equal
163
assert set(new_inventory_vf.versions()) == self.pending
164
self.pb.update('Writing weave')
165
self.repo.control_weaves.copy(new_inventory_vf, 'inventory', self.repo.get_transaction())
166
self.repo.control_weaves.delete('inventory.new', self.repo.get_transaction())
167
self.inventory = None
168
self.pb.note('Inventory regenerated.')
170
def _setup_steps(self, new_total):
171
"""Setup the markers we need to control the progress bar."""
172
self.total = new_total
175
def _graph_revision(self, rev_id):
176
"""Load a revision into the revision graph."""
177
# pick a random revision
178
# analyse revision id rev_id and put it in the stack.
179
self._reweave_step('loading revisions')
180
rev = self.repo.get_revision_reconcile(rev_id)
181
assert rev.revision_id == rev_id
183
for parent in rev.parent_ids:
184
if self._parent_is_available(parent):
185
parents.append(parent)
187
mutter('found ghost %s', parent)
188
self._rev_graph[rev_id] = parents
189
if set(self.inventory.get_parents(rev_id)) != set(parents):
190
self.inconsistent_parents += 1
191
mutter('Inconsistent inventory parents: id {%s} '
192
'inventory claims %r, '
193
'available parents are %r, '
194
'unavailable parents are %r',
196
set(self.inventory.get_parents(rev_id)),
198
set(rev.parent_ids).difference(set(parents)))
200
def _check_garbage_inventories(self):
201
"""Check for garbage inventories which we cannot trust
203
We cant trust them because their pre-requisite file data may not
204
be present - all we know is that their revision was not installed.
206
inventories = set(self.inventory.versions())
207
revisions = set(self._rev_graph.keys())
208
garbage = inventories.difference(revisions)
209
self.garbage_inventories = len(garbage)
210
for revision_id in garbage:
211
mutter('Garbage inventory {%s} found.', revision_id)
213
def _parent_is_available(self, parent):
214
"""True if parent is a fully available revision
216
A fully available revision has a inventory and a revision object in the
219
return (parent in self._rev_graph or
220
(parent in self.inventory and self.repo.has_revision(parent)))
222
def _reweave_step(self, message):
223
"""Mark a single step of regeneration complete."""
224
self.pb.update(message, self.count, self.total)
228
class KnitReconciler(RepoReconciler):
229
"""Reconciler that reconciles a knit format repository.
231
This will detect garbage inventories and remove them.
233
Inconsistent parentage is checked for in the revision weave.
236
def _reconcile_steps(self):
237
"""Perform the steps to reconcile this repository."""
239
# knits never suffer this
240
self.inconsistent_parents = 0
243
def _load_indexes(self):
244
"""Load indexes for the reconciliation."""
245
self.transaction = self.repo.get_transaction()
246
self.pb.update('Reading indexes.', 0, 2)
247
self.inventory = self.repo.get_inventory_weave()
248
self.pb.update('Reading indexes.', 1, 2)
249
self.revisions = self.repo._revision_store.get_revision_file(self.transaction)
250
self.pb.update('Reading indexes.', 2, 2)
252
def _gc_inventory(self):
253
"""Remove inventories that are not referenced from the revision store."""
254
self.pb.update('Checking unused inventories.', 0, 1)
255
self._check_garbage_inventories()
256
self.pb.update('Checking unused inventories.', 1, 3)
257
if not self.garbage_inventories:
258
self.pb.note('Inventory ok.')
260
self.pb.update('Backing up inventory...', 0, 0)
261
self.repo.control_weaves.copy(self.inventory, 'inventory.backup', self.transaction)
262
self.pb.note('Backup Inventory created.')
263
# asking for '' should never return a non-empty weave
264
new_inventory_vf = self.repo.control_weaves.get_empty('inventory.new',
267
# we have topological order of revisions and non ghost parents ready.
268
self._setup_steps(len(self.revisions))
269
for rev_id in TopoSorter(self.revisions.get_graph().items()).iter_topo_order():
270
parents = self.revisions.get_parents(rev_id)
271
# double check this really is in topological order.
272
unavailable = [p for p in parents if p not in new_inventory_vf]
273
assert len(unavailable) == 0
274
# this entry has all the non ghost parents in the inventory
276
self._reweave_step('adding inventories')
277
# ugly but needed, weaves are just way tooooo slow else.
278
new_inventory_vf.add_lines(rev_id, parents, self.inventory.get_lines(rev_id))
280
# if this worked, the set of new_inventory_vf.names should equal
282
assert set(new_inventory_vf.versions()) == set(self.revisions.versions())
283
self.pb.update('Writing weave')
284
self.repo.control_weaves.copy(new_inventory_vf, 'inventory', self.transaction)
285
self.repo.control_weaves.delete('inventory.new', self.transaction)
286
self.inventory = None
287
self.pb.note('Inventory regenerated.')
289
def _reinsert_revisions(self):
290
"""Correct the revision history for revisions in the revision knit."""
291
# the total set of revisions to process
292
self.pending = set(self.revisions.versions())
294
# mapping from revision_id to parents
296
# errors that we detect
297
self.inconsistent_parents = 0
298
# we need the revision id of each revision and its available parents list
299
self._setup_steps(len(self.pending))
300
for rev_id in self.pending:
301
# put a revision into the graph.
302
self._graph_revision(rev_id)
304
if not self.inconsistent_parents:
305
self.pb.note('Revision history accurate.')
307
self._setup_steps(len(self._rev_graph))
308
for rev_id, parents in self._rev_graph.items():
309
if parents != self.revisions.get_parents(rev_id):
310
self.revisions.fix_parents(rev_id, parents)
311
self._reweave_step('Fixing parents')
312
self.pb.note('Ancestry corrected.')
314
def _graph_revision(self, rev_id):
315
"""Load a revision into the revision graph."""
316
# pick a random revision
317
# analyse revision id rev_id and put it in the stack.
318
self._reweave_step('loading revisions')
319
rev = self.repo._revision_store.get_revision(rev_id, self.transaction)
320
assert rev.revision_id == rev_id
322
for parent in rev.parent_ids:
323
if self.revisions.has_version(parent):
324
parents.append(parent)
326
mutter('found ghost %s', parent)
327
self._rev_graph[rev_id] = parents
328
if set(self.inventory.get_parents(rev_id)) != set(parents):
329
self.inconsistent_parents += 1
330
mutter('Inconsistent inventory parents: id {%s} '
331
'inventory claims %r, '
332
'available parents are %r, '
333
'unavailable parents are %r',
335
set(self.inventory.get_parents(rev_id)),
337
set(rev.parent_ids).difference(set(parents)))
339
def _check_garbage_inventories(self):
340
"""Check for garbage inventories which we cannot trust
342
We cant trust them because their pre-requisite file data may not
343
be present - all we know is that their revision was not installed.
345
inventories = set(self.inventory.versions())
346
revisions = set(self.revisions.versions())
347
garbage = inventories.difference(revisions)
348
self.garbage_inventories = len(garbage)
349
for revision_id in garbage:
350
mutter('Garbage inventory {%s} found.', revision_id)
added
removed
bzrlib/reconcile.py
