1616.1.5
by Martin Pool
Cleanup and document some check code |
1 |
# Copyright (C) 2005, 2006 by Canonical Ltd
|
1
by mbp at sourcefrog
import from baz patch-364 |
2 |
|
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.
|
|
7 |
||
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.
|
|
12 |
||
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
|
|
16 |
||
1335
by Martin Pool
doc |
17 |
# TODO: Check ancestries are correct for every revision: includes
|
18 |
# every committed so far, and in a reasonable order.
|
|
19 |
||
1347
by Martin Pool
- refactor check code into method object |
20 |
# TODO: Also check non-mainline revisions mentioned as parents.
|
21 |
||
22 |
# TODO: Check for extra files in the control directory.
|
|
23 |
||
1348
by Martin Pool
- more refactoring of check code |
24 |
# TODO: Check revision, inventory and entry objects have all
|
25 |
# required fields.
|
|
26 |
||
1185.16.101
by mbp at sourcefrog
todo |
27 |
# TODO: Get every revision in the revision-store even if they're not
|
28 |
# referenced by history and make sure they're all valid.
|
|
1347
by Martin Pool
- refactor check code into method object |
29 |
|
1616.1.5
by Martin Pool
Cleanup and document some check code |
30 |
# TODO: Perhaps have a way to record errors other than by raising exceptions;
|
31 |
# would perhaps be enough to accumulate exception objects in a list without
|
|
32 |
# raising them. If there's more than one exception it'd be good to see them
|
|
33 |
# all.
|
|
34 |
||
1104
by Martin Pool
- Add a simple UIFactory |
35 |
import bzrlib.ui |
1130
by Martin Pool
- check command writes output through logging not direct |
36 |
from bzrlib.trace import note, warning |
1092.2.20
by Robert Collins
symlink and weaves, whaddya know |
37 |
from bzrlib.osutils import rename, sha_string, fingerprint_file |
1346
by Martin Pool
- remove dead code from bzrlib.check |
38 |
from bzrlib.trace import mutter |
39 |
from bzrlib.errors import BzrCheckError, NoSuchRevision |
|
40 |
from bzrlib.inventory import ROOT_ID |
|
1
by mbp at sourcefrog
import from baz patch-364 |
41 |
|
1104
by Martin Pool
- Add a simple UIFactory |
42 |
|
1347
by Martin Pool
- refactor check code into method object |
43 |
class Check(object): |
44 |
"""Check a branch"""
|
|
1449
by Robert Collins
teach check about ghosts |
45 |
|
1616.1.5
by Martin Pool
Cleanup and document some check code |
46 |
# The Check object interacts with InventoryEntry.check, etc.
|
47 |
||
1347
by Martin Pool
- refactor check code into method object |
48 |
def __init__(self, branch): |
49 |
self.branch = branch |
|
1185.67.2
by Aaron Bentley
Renamed Branch.storage to Branch.repository |
50 |
self.repository = branch.repository |
1383
by Martin Pool
- untabify only |
51 |
self.checked_text_cnt = 0 |
52 |
self.checked_rev_cnt = 0 |
|
1449
by Robert Collins
teach check about ghosts |
53 |
self.ghosts = [] |
1365
by Martin Pool
- try to avoid checking texts repeatedly |
54 |
self.repeated_text_cnt = 0 |
1449
by Robert Collins
teach check about ghosts |
55 |
self.missing_parent_links = {} |
1348
by Martin Pool
- more refactoring of check code |
56 |
self.missing_inventory_sha_cnt = 0 |
57 |
self.missing_revision_cnt = 0 |
|
1616.1.5
by Martin Pool
Cleanup and document some check code |
58 |
# maps (file-id, version) -> sha1; used by InventoryFile._check
|
1365
by Martin Pool
- try to avoid checking texts repeatedly |
59 |
self.checked_texts = {} |
1185.50.28
by John Arbash Meinel
Lots of updates for 'bzr check' |
60 |
self.checked_weaves = {} |
676
by Martin Pool
- lock branch while checking |
61 |
|
1449
by Robert Collins
teach check about ghosts |
62 |
def check(self): |
63 |
self.branch.lock_read() |
|
1594.1.3
by Robert Collins
Fixup pb usage to use nested_progress_bar. |
64 |
self.progress = bzrlib.ui.ui_factory.nested_progress_bar() |
1449
by Robert Collins
teach check about ghosts |
65 |
try: |
1510
by Robert Collins
Merge from mpool, adjusting check to retain HTTP support. |
66 |
self.progress.update('retrieving inventory', 0, 0) |
67 |
# do not put in init, as it should be done with progess,
|
|
68 |
# and inside the lock.
|
|
1185.67.2
by Aaron Bentley
Renamed Branch.storage to Branch.repository |
69 |
self.inventory_weave = self.branch.repository.get_inventory_weave() |
1449
by Robert Collins
teach check about ghosts |
70 |
self.history = self.branch.revision_history() |
71 |
if not len(self.history): |
|
72 |
# nothing to see here
|
|
73 |
return
|
|
1510
by Robert Collins
Merge from mpool, adjusting check to retain HTTP support. |
74 |
self.plan_revisions() |
75 |
revno = 0 |
|
1185.50.28
by John Arbash Meinel
Lots of updates for 'bzr check' |
76 |
self.check_weaves() |
1510
by Robert Collins
Merge from mpool, adjusting check to retain HTTP support. |
77 |
while revno < len(self.planned_revisions): |
78 |
rev_id = self.planned_revisions[revno] |
|
79 |
self.progress.update('checking revision', revno, |
|
1449
by Robert Collins
teach check about ghosts |
80 |
len(self.planned_revisions)) |
1510
by Robert Collins
Merge from mpool, adjusting check to retain HTTP support. |
81 |
revno += 1 |
1449
by Robert Collins
teach check about ghosts |
82 |
self.check_one_rev(rev_id) |
1185.35.34
by Aaron Bentley
Made bzr check for stored revisions missing from ancestry |
83 |
finally: |
1594.1.3
by Robert Collins
Fixup pb usage to use nested_progress_bar. |
84 |
self.progress.finished() |
1449
by Robert Collins
teach check about ghosts |
85 |
self.branch.unlock() |
86 |
||
1510
by Robert Collins
Merge from mpool, adjusting check to retain HTTP support. |
87 |
def plan_revisions(self): |
1185.67.2
by Aaron Bentley
Renamed Branch.storage to Branch.repository |
88 |
repository = self.branch.repository |
1563.2.22
by Robert Collins
Move responsibility for repository.has_revision into RevisionStore |
89 |
self.planned_revisions = set(repository.all_revision_ids()) |
90 |
self.progress.clear() |
|
1563.2.35
by Robert Collins
cleanup deprecation warnings and finish conversion so the inventory is knit based too. |
91 |
inventoried = set(self.inventory_weave.versions()) |
1510
by Robert Collins
Merge from mpool, adjusting check to retain HTTP support. |
92 |
awol = self.planned_revisions - inventoried |
93 |
if len(awol) > 0: |
|
94 |
raise BzrCheckError('Stored revisions missing from inventory' |
|
95 |
'{%s}' % ','.join([f for f in awol])) |
|
96 |
self.planned_revisions = list(self.planned_revisions) |
|
97 |
||
1449
by Robert Collins
teach check about ghosts |
98 |
def report_results(self, verbose): |
1534.4.26
by Robert Collins
Move working tree initialisation out from Branch.initialize, deprecated Branch.initialize to Branch.create. |
99 |
note('checked branch %s format %s', |
1383
by Martin Pool
- untabify only |
100 |
self.branch.base, |
1534.4.44
by Robert Collins
Make a new BzrDir format that uses a versioned branch format in a branch/ subdirectory. |
101 |
self.branch._format) |
1348
by Martin Pool
- more refactoring of check code |
102 |
|
1365
by Martin Pool
- try to avoid checking texts repeatedly |
103 |
note('%6d revisions', self.checked_rev_cnt) |
104 |
note('%6d unique file texts', self.checked_text_cnt) |
|
105 |
note('%6d repeated file texts', self.repeated_text_cnt) |
|
1185.50.28
by John Arbash Meinel
Lots of updates for 'bzr check' |
106 |
note('%6d weaves', len(self.checked_weaves)) |
1348
by Martin Pool
- more refactoring of check code |
107 |
if self.missing_inventory_sha_cnt: |
1449
by Robert Collins
teach check about ghosts |
108 |
note('%6d revisions are missing inventory_sha1', |
1383
by Martin Pool
- untabify only |
109 |
self.missing_inventory_sha_cnt) |
1348
by Martin Pool
- more refactoring of check code |
110 |
if self.missing_revision_cnt: |
1449
by Robert Collins
teach check about ghosts |
111 |
note('%6d revisions are mentioned but not present', |
1383
by Martin Pool
- untabify only |
112 |
self.missing_revision_cnt) |
1449
by Robert Collins
teach check about ghosts |
113 |
if len(self.ghosts): |
114 |
note('%6d ghost revisions', len(self.ghosts)) |
|
115 |
if verbose: |
|
116 |
for ghost in self.ghosts: |
|
117 |
note(' %s', ghost) |
|
118 |
if len(self.missing_parent_links): |
|
119 |
note('%6d revisions missing parents in ancestry', |
|
120 |
len(self.missing_parent_links)) |
|
121 |
if verbose: |
|
122 |
for link, linkers in self.missing_parent_links.items(): |
|
123 |
note(' %s should be in the ancestry for:', link) |
|
124 |
for linker in linkers: |
|
125 |
note(' * %s', linker) |
|
126 |
||
127 |
def check_one_rev(self, rev_id): |
|
1383
by Martin Pool
- untabify only |
128 |
"""Check one revision.
|
129 |
||
130 |
rev_id - the one to check
|
|
131 |
||
132 |
last_rev_id - the previous one on the mainline, if any.
|
|
133 |
"""
|
|
134 |
||
1185.31.4
by John Arbash Meinel
Fixing mutter() calls to not have to do string processing. |
135 |
# mutter(' revision {%s}', rev_id)
|
1383
by Martin Pool
- untabify only |
136 |
branch = self.branch |
1449
by Robert Collins
teach check about ghosts |
137 |
try: |
138 |
rev_history_position = self.history.index(rev_id) |
|
139 |
except ValueError: |
|
140 |
rev_history_position = None |
|
141 |
last_rev_id = None |
|
142 |
if rev_history_position: |
|
1185.67.2
by Aaron Bentley
Renamed Branch.storage to Branch.repository |
143 |
rev = branch.repository.get_revision(rev_id) |
1449
by Robert Collins
teach check about ghosts |
144 |
if rev_history_position > 0: |
145 |
last_rev_id = self.history[rev_history_position - 1] |
|
146 |
else: |
|
1185.67.2
by Aaron Bentley
Renamed Branch.storage to Branch.repository |
147 |
rev = branch.repository.get_revision(rev_id) |
1449
by Robert Collins
teach check about ghosts |
148 |
|
1383
by Martin Pool
- untabify only |
149 |
if rev.revision_id != rev_id: |
150 |
raise BzrCheckError('wrong internal revision id in revision {%s}' |
|
151 |
% rev_id) |
|
152 |
||
153 |
# check the previous history entry is a parent of this entry
|
|
154 |
if rev.parent_ids: |
|
1449
by Robert Collins
teach check about ghosts |
155 |
if last_rev_id is not None: |
156 |
for parent_id in rev.parent_ids: |
|
157 |
if parent_id == last_rev_id: |
|
158 |
break
|
|
159 |
else: |
|
160 |
raise BzrCheckError("previous revision {%s} not listed among " |
|
161 |
"parents of {%s}" |
|
162 |
% (last_rev_id, rev_id)) |
|
163 |
for parent in rev.parent_ids: |
|
164 |
if not parent in self.planned_revisions: |
|
165 |
missing_links = self.missing_parent_links.get(parent, []) |
|
166 |
missing_links.append(rev_id) |
|
167 |
self.missing_parent_links[parent] = missing_links |
|
1510
by Robert Collins
Merge from mpool, adjusting check to retain HTTP support. |
168 |
# list based so somewhat slow,
|
169 |
# TODO have a planned_revisions list and set.
|
|
1587.1.4
by Robert Collins
Quick, urgent fix for common_ancestor performance. |
170 |
if self.branch.repository.has_revision(parent): |
1185.67.2
by Aaron Bentley
Renamed Branch.storage to Branch.repository |
171 |
missing_ancestry = self.repository.get_ancestry(parent) |
1449
by Robert Collins
teach check about ghosts |
172 |
for missing in missing_ancestry: |
173 |
if (missing is not None |
|
174 |
and missing not in self.planned_revisions): |
|
175 |
self.planned_revisions.append(missing) |
|
176 |
else: |
|
177 |
self.ghosts.append(rev_id) |
|
1383
by Martin Pool
- untabify only |
178 |
elif last_rev_id: |
179 |
raise BzrCheckError("revision {%s} has no parents listed " |
|
180 |
"but preceded by {%s}" |
|
181 |
% (rev_id, last_rev_id)) |
|
182 |
||
183 |
if rev.inventory_sha1: |
|
1185.67.2
by Aaron Bentley
Renamed Branch.storage to Branch.repository |
184 |
inv_sha1 = branch.repository.get_inventory_sha1(rev_id) |
1383
by Martin Pool
- untabify only |
185 |
if inv_sha1 != rev.inventory_sha1: |
186 |
raise BzrCheckError('Inventory sha1 hash doesn\'t match' |
|
187 |
' value in revision {%s}' % rev_id) |
|
188 |
else: |
|
189 |
missing_inventory_sha_cnt += 1 |
|
1185.31.4
by John Arbash Meinel
Fixing mutter() calls to not have to do string processing. |
190 |
mutter("no inventory_sha1 on revision {%s}", rev_id) |
1383
by Martin Pool
- untabify only |
191 |
self._check_revision_tree(rev_id) |
1362
by Martin Pool
- keep track of number of checked revisions |
192 |
self.checked_rev_cnt += 1 |
1349
by Martin Pool
- more refactoring of check code |
193 |
|
1185.50.28
by John Arbash Meinel
Lots of updates for 'bzr check' |
194 |
def check_weaves(self): |
195 |
"""Check all the weaves we can get our hands on.
|
|
196 |
"""
|
|
197 |
n_weaves = 1 |
|
198 |
weave_ids = [] |
|
1185.65.17
by Robert Collins
Merge from integration, mode-changes are broken. |
199 |
if self.branch.repository.weave_store.listable(): |
200 |
weave_ids = list(self.branch.repository.weave_store) |
|
1185.50.28
by John Arbash Meinel
Lots of updates for 'bzr check' |
201 |
n_weaves = len(weave_ids) |
202 |
self.progress.update('checking weave', 0, n_weaves) |
|
203 |
self.inventory_weave.check(progress_bar=self.progress) |
|
204 |
for i, weave_id in enumerate(weave_ids): |
|
205 |
self.progress.update('checking weave', i, n_weaves) |
|
1185.65.17
by Robert Collins
Merge from integration, mode-changes are broken. |
206 |
w = self.branch.repository.weave_store.get_weave(weave_id, |
207 |
self.branch.repository.get_transaction()) |
|
1185.50.28
by John Arbash Meinel
Lots of updates for 'bzr check' |
208 |
# No progress here, because it looks ugly.
|
209 |
w.check() |
|
210 |
self.checked_weaves[weave_id] = True |
|
211 |
||
1349
by Martin Pool
- more refactoring of check code |
212 |
def _check_revision_tree(self, rev_id): |
1185.67.2
by Aaron Bentley
Renamed Branch.storage to Branch.repository |
213 |
tree = self.branch.repository.revision_tree(rev_id) |
1383
by Martin Pool
- untabify only |
214 |
inv = tree.inventory |
215 |
seen_ids = {} |
|
216 |
for file_id in inv: |
|
217 |
if file_id in seen_ids: |
|
218 |
raise BzrCheckError('duplicated file_id {%s} ' |
|
219 |
'in inventory for revision {%s}' |
|
220 |
% (file_id, rev_id)) |
|
221 |
seen_ids[file_id] = True |
|
222 |
for file_id in inv: |
|
1092.2.20
by Robert Collins
symlink and weaves, whaddya know |
223 |
ie = inv[file_id] |
224 |
ie.check(self, rev_id, inv, tree) |
|
1383
by Martin Pool
- untabify only |
225 |
seen_names = {} |
226 |
for path, ie in inv.iter_entries(): |
|
227 |
if path in seen_names: |
|
228 |
raise BzrCheckError('duplicated path %s ' |
|
229 |
'in inventory for revision {%s}' |
|
230 |
% (path, rev_id)) |
|
231 |
seen_names[path] = True |
|
1349
by Martin Pool
- more refactoring of check code |
232 |
|
1347
by Martin Pool
- refactor check code into method object |
233 |
|
1449
by Robert Collins
teach check about ghosts |
234 |
def check(branch, verbose): |
1347
by Martin Pool
- refactor check code into method object |
235 |
"""Run consistency checks on a branch."""
|
1449
by Robert Collins
teach check about ghosts |
236 |
checker = Check(branch) |
237 |
checker.check() |
|
238 |
checker.report_results(verbose) |