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', 'KnitReconciler']
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
31
def reconcile(dir, other=None):
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
:param other: another bzrdir to reconcile against.
43
reconciler = Reconciler(dir, other=other)
44
reconciler.reconcile()
47
class Reconciler(object):
48
"""Reconcilers are used to reconcile existing data."""
50
def __init__(self, dir, other=None):
51
"""Create a Reconciler."""
55
"""Perform reconciliation.
57
After reconciliation the following attributes document found issues:
58
inconsistent_parents: The number of revisions in the repository whose
59
ancestry was being reported incorrectly.
60
garbage_inventories: The number of inventory objects without revisions
61
that were garbage collected.
63
self.pb = ui.ui_factory.nested_progress_bar()
70
"""Helper function for performing reconciliation."""
71
self.repo = self.bzrdir.find_repository()
72
self.pb.note('Reconciling repository %s',
73
self.repo.bzrdir.root_transport.base)
74
repo_reconciler = self.repo.reconcile(thorough=True)
75
self.inconsistent_parents = repo_reconciler.inconsistent_parents
76
self.garbage_inventories = repo_reconciler.garbage_inventories
77
self.pb.note('Reconciliation complete.')
80
class RepoReconciler(object):
81
"""Reconciler that reconciles a repository.
83
Currently this consists of an inventory reweave with revision cross-checks.
86
def __init__(self, repo, other=None, thorough=False):
87
"""Construct a RepoReconciler.
89
:param thorough: perform a thorough check which may take longer but
90
will correct non-data loss issues such as incorrect
93
self.garbage_inventories = 0
94
self.inconsistent_parents = 0
96
self.thorough = thorough
99
"""Perform reconciliation.
101
After reconciliation the following attributes document found issues:
102
inconsistent_parents: The number of revisions in the repository whose
103
ancestry was being reported incorrectly.
104
garbage_inventories: The number of inventory objects without revisions
105
that were garbage collected.
107
self.repo.lock_write()
109
self.pb = ui.ui_factory.nested_progress_bar()
111
self._reconcile_steps()
117
def _reconcile_steps(self):
118
"""Perform the steps to reconcile this repository."""
119
self._reweave_inventory()
121
def _reweave_inventory(self):
122
"""Regenerate the inventory weave for the repository from scratch.
124
This is a smart function: it will only do the reweave if doing it
125
will correct data issues. The self.thorough flag controls whether
126
only data-loss causing issues (!self.thorough) or all issues
127
(self.thorough) are treated as requiring the reweave.
129
# local because needing to know about WeaveFile is a wart we want to hide
130
from bzrlib.weave import WeaveFile, Weave
131
transaction = self.repo.get_transaction()
132
self.pb.update('Reading inventory data.')
133
self.inventory = self.repo.get_inventory_weave()
134
# the total set of revisions to process
135
self.pending = set([rev_id for rev_id in self.repo._revision_store.all_revision_ids(transaction)])
137
# mapping from revision_id to parents
139
# errors that we detect
140
self.inconsistent_parents = 0
141
# we need the revision id of each revision and its available parents list
142
self._setup_steps(len(self.pending))
143
for rev_id in self.pending:
144
# put a revision into the graph.
145
self._graph_revision(rev_id)
146
self._check_garbage_inventories()
147
# if there are no inconsistent_parents and
148
# (no garbage inventories or we are not doing a thorough check)
149
if (not self.inconsistent_parents and
150
(not self.garbage_inventories or not self.thorough)):
151
self.pb.note('Inventory ok.')
153
self.pb.update('Backing up inventory...', 0, 0)
154
self.repo.control_weaves.copy(self.inventory, 'inventory.backup', self.repo.get_transaction())
155
self.pb.note('Backup Inventory created.')
156
# asking for '' should never return a non-empty weave
157
new_inventory_vf = self.repo.control_weaves.get_empty('inventory.new',
158
self.repo.get_transaction())
160
# we have topological order of revisions and non ghost parents ready.
161
self._setup_steps(len(self._rev_graph))
162
for rev_id in TopoSorter(self._rev_graph.items()).iter_topo_order():
163
parents = self._rev_graph[rev_id]
164
# double check this really is in topological order.
165
unavailable = [p for p in parents if p not in new_inventory_vf]
166
assert len(unavailable) == 0
167
# this entry has all the non ghost parents in the inventory
169
self._reweave_step('adding inventories')
170
if isinstance(new_inventory_vf, WeaveFile):
171
# It's really a WeaveFile, but we call straight into the
172
# Weave's add method to disable the auto-write-out behaviour.
173
# This is done to avoid a revision_count * time-to-write additional overhead on
175
new_inventory_vf._check_write_ok()
176
Weave._add_lines(new_inventory_vf, rev_id, parents, self.inventory.get_lines(rev_id),
179
new_inventory_vf.add_lines(rev_id, parents, self.inventory.get_lines(rev_id))
181
if isinstance(new_inventory_vf, WeaveFile):
182
new_inventory_vf._save()
183
# if this worked, the set of new_inventory_vf.names should equal
185
assert set(new_inventory_vf.versions()) == self.pending
186
self.pb.update('Writing weave')
187
self.repo.control_weaves.copy(new_inventory_vf, 'inventory', self.repo.get_transaction())
188
self.repo.control_weaves.delete('inventory.new', self.repo.get_transaction())
189
self.inventory = None
190
self.pb.note('Inventory regenerated.')
192
def _setup_steps(self, new_total):
193
"""Setup the markers we need to control the progress bar."""
194
self.total = new_total
197
def _graph_revision(self, rev_id):
198
"""Load a revision into the revision graph."""
199
# pick a random revision
200
# analyse revision id rev_id and put it in the stack.
201
self._reweave_step('loading revisions')
202
rev = self.repo.get_revision_reconcile(rev_id)
203
assert rev.revision_id == rev_id
205
for parent in rev.parent_ids:
206
if self._parent_is_available(parent):
207
parents.append(parent)
209
mutter('found ghost %s', parent)
210
self._rev_graph[rev_id] = parents
211
if self._parents_are_inconsistent(rev_id, parents):
212
self.inconsistent_parents += 1
213
mutter('Inconsistent inventory parents: id {%s} '
214
'inventory claims %r, '
215
'available parents are %r, '
216
'unavailable parents are %r',
218
set(self.inventory.get_parents(rev_id)),
220
set(rev.parent_ids).difference(set(parents)))
222
def _parents_are_inconsistent(self, rev_id, parents):
223
"""Return True if the parents list of rev_id does not match the weave.
225
This detect inconsistences based on the self.thorough value:
226
if thorough is on, the first parent value is checked as well as ghost
228
Otherwise only the ghost differences are evaluated.
230
weave_parents = self.inventory.get_parents(rev_id)
231
weave_missing_old_ghosts = set(weave_parents) != set(parents)
232
first_parent_is_wrong = (
233
len(weave_parents) and len(parents) and
234
parents[0] != weave_parents[0])
236
return weave_missing_old_ghosts or first_parent_is_wrong
238
return weave_missing_old_ghosts
240
def _check_garbage_inventories(self):
241
"""Check for garbage inventories which we cannot trust
243
We cant trust them because their pre-requisite file data may not
244
be present - all we know is that their revision was not installed.
246
if not self.thorough:
248
inventories = set(self.inventory.versions())
249
revisions = set(self._rev_graph.keys())
250
garbage = inventories.difference(revisions)
251
self.garbage_inventories = len(garbage)
252
for revision_id in garbage:
253
mutter('Garbage inventory {%s} found.', revision_id)
255
def _parent_is_available(self, parent):
256
"""True if parent is a fully available revision
258
A fully available revision has a inventory and a revision object in the
261
return (parent in self._rev_graph or
262
(parent in self.inventory and self.repo.has_revision(parent)))
264
def _reweave_step(self, message):
265
"""Mark a single step of regeneration complete."""
266
self.pb.update(message, self.count, self.total)
270
class KnitReconciler(RepoReconciler):
271
"""Reconciler that reconciles a knit format repository.
273
This will detect garbage inventories and remove them.
275
Inconsistent parentage is checked for in the revision weave.
278
def _reconcile_steps(self):
279
"""Perform the steps to reconcile this repository."""
282
# knits never suffer this
285
def _load_indexes(self):
286
"""Load indexes for the reconciliation."""
287
self.transaction = self.repo.get_transaction()
288
self.pb.update('Reading indexes.', 0, 2)
289
self.inventory = self.repo.get_inventory_weave()
290
self.pb.update('Reading indexes.', 1, 2)
291
self.revisions = self.repo._revision_store.get_revision_file(self.transaction)
292
self.pb.update('Reading indexes.', 2, 2)
294
def _gc_inventory(self):
295
"""Remove inventories that are not referenced from the revision store."""
296
self.pb.update('Checking unused inventories.', 0, 1)
297
self._check_garbage_inventories()
298
self.pb.update('Checking unused inventories.', 1, 3)
299
if not self.garbage_inventories:
300
self.pb.note('Inventory ok.')
302
self.pb.update('Backing up inventory...', 0, 0)
303
self.repo.control_weaves.copy(self.inventory, 'inventory.backup', self.transaction)
304
self.pb.note('Backup Inventory created.')
305
# asking for '' should never return a non-empty weave
306
new_inventory_vf = self.repo.control_weaves.get_empty('inventory.new',
309
# we have topological order of revisions and non ghost parents ready.
310
self._setup_steps(len(self.revisions))
311
for rev_id in TopoSorter(self.revisions.get_graph().items()).iter_topo_order():
312
parents = self.revisions.get_parents(rev_id)
313
# double check this really is in topological order.
314
unavailable = [p for p in parents if p not in new_inventory_vf]
315
assert len(unavailable) == 0
316
# this entry has all the non ghost parents in the inventory
318
self._reweave_step('adding inventories')
319
# ugly but needed, weaves are just way tooooo slow else.
320
new_inventory_vf.add_lines(rev_id, parents, self.inventory.get_lines(rev_id))
322
# if this worked, the set of new_inventory_vf.names should equal
324
assert set(new_inventory_vf.versions()) == set(self.revisions.versions())
325
self.pb.update('Writing weave')
326
self.repo.control_weaves.copy(new_inventory_vf, 'inventory', self.transaction)
327
self.repo.control_weaves.delete('inventory.new', self.transaction)
328
self.inventory = None
329
self.pb.note('Inventory regenerated.')
331
def _check_garbage_inventories(self):
332
"""Check for garbage inventories which we cannot trust
334
We cant trust them because their pre-requisite file data may not
335
be present - all we know is that their revision was not installed.
337
inventories = set(self.inventory.versions())
338
revisions = set(self.revisions.versions())
339
garbage = inventories.difference(revisions)
340
self.garbage_inventories = len(garbage)
341
for revision_id in garbage:
342
mutter('Garbage inventory {%s} found.', revision_id)
added
removed
bzrlib/reconcile.py
