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 = 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]
146
assert len(unavailable) == 0
147
# this entry has all the non ghost parents in the inventory
149
self._reweave_step('adding inventories')
150
# ugly but needed, weaves are just way tooooo slow else.
151
if isinstance(new_inventory, WeaveFile):
152
Weave.add_lines(new_inventory, rev_id, parents, self.inventory.get_lines(rev_id))
154
new_inventory.add_lines(rev_id, parents, self.inventory.get_lines(rev_id))
156
if isinstance(new_inventory, WeaveFile):
157
new_inventory._save()
158
# if this worked, the set of new_inventory.names should equal
160
assert set(new_inventory.versions()) == self.pending
161
self.pb.update('Writing weave')
162
self.repo.control_weaves.copy(new_inventory, 'inventory', self.repo.get_transaction())
163
self.repo.control_weaves.delete('inventory.new', self.repo.get_transaction())
164
self.inventory = None
165
self.pb.note('Inventory regenerated.')
167
def _setup_steps(self, new_total):
168
"""Setup the markers we need to control the progress bar."""
169
self.total = new_total
172
def _graph_revision(self, rev_id):
173
"""Load a revision into the revision graph."""
174
# pick a random revision
175
# analyse revision id rev_id and put it in the stack.
176
self._reweave_step('loading revisions')
177
rev = self.repo.get_revision_reconcile(rev_id)
178
assert rev.revision_id == rev_id
180
for parent in rev.parent_ids:
181
if self._parent_is_available(parent):
182
parents.append(parent)
184
mutter('found ghost %s', parent)
185
self._rev_graph[rev_id] = parents
186
if set(self.inventory.get_parents(rev_id)) != set(parents):
187
self.inconsistent_parents += 1
188
mutter('Inconsistent inventory parents: id {%s} '
189
'inventory claims %r, '
190
'available parents are %r, '
191
'unavailable parents are %r',
193
set(self.inventory.get_parents(rev_id)),
195
set(rev.parent_ids).difference(set(parents)))
197
def _check_garbage_inventories(self):
198
"""Check for garbage inventories which we cannot trust
200
We cant trust them because their pre-requisite file data may not
201
be present - all we know is that their revision was not installed.
203
inventories = set(self.inventory.versions())
204
revisions = set(self._rev_graph.keys())
205
garbage = inventories.difference(revisions)
206
self.garbage_inventories = len(garbage)
207
for revision_id in garbage:
208
mutter('Garbage inventory {%s} found.', revision_id)
210
def _parent_is_available(self, parent):
211
"""True if parent is a fully available revision
213
A fully available revision has a inventory and a revision object in the
216
return (parent in self._rev_graph or
217
(parent in self.inventory and self.repo.has_revision(parent)))
219
def _reweave_step(self, message):
220
"""Mark a single step of regeneration complete."""
221
self.pb.update(message, self.count, self.total)
225
class KnitReconciler(RepoReconciler):
226
"""Reconciler that reconciles a knit format repository.
228
This will detect garbage inventories and remove them.
230
Inconsistent parentage is checked for in the revision weave.
233
def _reconcile_steps(self):
234
"""Perform the steps to reconcile this repository."""
236
self._reinsert_revisions()
239
def _load_indexes(self):
240
"""Load indexes for the reconciliation."""
241
self.transaction = self.repo.get_transaction()
242
self.pb.update('Reading indexes.', 0, 2)
243
self.inventory = self.repo.get_inventory_weave()
244
self.pb.update('Reading indexes.', 1, 2)
245
self.revisions = self.repo._revision_store.get_revision_file(self.transaction)
246
self.pb.update('Reading indexes.', 2, 2)
248
def _gc_inventory(self):
249
"""Remove inventories that are not referenced from the revision store."""
250
self.pb.update('Checking unused inventories.', 0, 1)
251
self._check_garbage_inventories()
252
self.pb.update('Checking unused inventories.', 1, 3)
253
if not self.garbage_inventories:
254
self.pb.note('Inventory ok.')
256
self.pb.update('Backing up inventory...', 0, 0)
257
self.repo.control_weaves.copy(self.inventory, 'inventory.backup', self.transaction)
258
self.pb.note('Backup Inventory created.')
259
# asking for '' should never return a non-empty weave
260
new_inventory = self.repo.control_weaves.get_empty('inventory.new',
263
# we have topological order of revisions and non ghost parents ready.
264
self._setup_steps(len(self._rev_graph))
265
for rev_id in TopoSorter(self._rev_graph.items()).iter_topo_order():
266
parents = self._rev_graph[rev_id]
267
# double check this really is in topological order.
268
unavailable = [p for p in parents if p not in new_inventory]
269
assert len(unavailable) == 0
270
# this entry has all the non ghost parents in the inventory
272
self._reweave_step('adding inventories')
273
# ugly but needed, weaves are just way tooooo slow else.
274
new_inventory.add_lines(rev_id, parents, self.inventory.get_lines(rev_id))
276
# if this worked, the set of new_inventory.names should equal
278
assert set(new_inventory.versions()) == set(self.revisions.versions())
279
self.pb.update('Writing weave')
280
self.repo.control_weaves.copy(new_inventory, 'inventory', self.transaction)
281
self.repo.control_weaves.delete('inventory.new', self.transaction)
282
self.inventory = None
283
self.pb.note('Inventory regenerated.')
285
def _reinsert_revisions(self):
286
"""Correct the revision history for revisions in the revision knit."""
287
# the total set of revisions to process
288
self.pending = set(self.revisions.versions())
290
# mapping from revision_id to parents
292
# errors that we detect
293
self.inconsistent_parents = 0
294
# we need the revision id of each revision and its available parents list
295
self._setup_steps(len(self.pending))
296
for rev_id in self.pending:
297
# put a revision into the graph.
298
self._graph_revision(rev_id)
300
if not self.inconsistent_parents:
301
self.pb.note('Revision history accurate.')
303
self._setup_steps(len(self._rev_graph))
304
for rev_id, parents in self._rev_graph.items():
305
if parents != self.revisions.get_parents(rev_id):
306
self.revisions.fix_parents(rev_id, parents)
307
self._reweave_step('Fixing parents')
308
self.pb.note('Ancestry corrected.')
310
def _graph_revision(self, rev_id):
311
"""Load a revision into the revision graph."""
312
# pick a random revision
313
# analyse revision id rev_id and put it in the stack.
314
self._reweave_step('loading revisions')
315
rev = self.repo._revision_store.get_revision(rev_id, self.transaction)
316
assert rev.revision_id == rev_id
318
for parent in rev.parent_ids:
319
if self.revisions.has_version(parent):
320
parents.append(parent)
322
mutter('found ghost %s', parent)
323
self._rev_graph[rev_id] = parents
324
if set(self.inventory.get_parents(rev_id)) != set(parents):
325
self.inconsistent_parents += 1
326
mutter('Inconsistent inventory parents: id {%s} '
327
'inventory claims %r, '
328
'available parents are %r, '
329
'unavailable parents are %r',
331
set(self.inventory.get_parents(rev_id)),
333
set(rev.parent_ids).difference(set(parents)))
335
def _check_garbage_inventories(self):
336
"""Check for garbage inventories which we cannot trust
338
We cant trust them because their pre-requisite file data may not
339
be present - all we know is that their revision was not installed.
341
inventories = set(self.inventory.versions())
342
revisions = set(self.revisions.versions())
343
garbage = inventories.difference(revisions)
344
self.garbage_inventories = len(garbage)
345
for revision_id in garbage:
346
mutter('Garbage inventory {%s} found.', revision_id)
added
removed
bzrlib/reconcile.py
