line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
# <@LICENSE> |
2
|
|
|
|
|
|
|
# Licensed to the Apache Software Foundation (ASF) under one or more |
3
|
|
|
|
|
|
|
# contributor license agreements. See the NOTICE file distributed with |
4
|
|
|
|
|
|
|
# this work for additional information regarding copyright ownership. |
5
|
|
|
|
|
|
|
# The ASF licenses this file to you under the Apache License, Version 2.0 |
6
|
|
|
|
|
|
|
# (the "License"); you may not use this file except in compliance with |
7
|
|
|
|
|
|
|
# the License. You may obtain a copy of the License at: |
8
|
|
|
|
|
|
|
# |
9
|
|
|
|
|
|
|
# http://www.apache.org/licenses/LICENSE-2.0 |
10
|
|
|
|
|
|
|
# |
11
|
|
|
|
|
|
|
# Unless required by applicable law or agreed to in writing, software |
12
|
|
|
|
|
|
|
# distributed under the License is distributed on an "AS IS" BASIS, |
13
|
|
|
|
|
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
14
|
|
|
|
|
|
|
# See the License for the specific language governing permissions and |
15
|
|
|
|
|
|
|
# limitations under the License. |
16
|
|
|
|
|
|
|
# </@LICENSE> |
17
|
|
|
|
|
|
|
|
18
|
|
|
|
|
|
|
package Mail::SpamAssassin::BayesStore::DBM; |
19
|
|
|
|
|
|
|
|
20
|
21
|
|
|
21
|
|
156
|
use strict; |
|
21
|
|
|
|
|
56
|
|
|
21
|
|
|
|
|
685
|
|
21
|
21
|
|
|
21
|
|
153
|
use warnings; |
|
21
|
|
|
|
|
45
|
|
|
21
|
|
|
|
|
626
|
|
22
|
|
|
|
|
|
|
# use bytes; |
23
|
21
|
|
|
21
|
|
123
|
use re 'taint'; |
|
21
|
|
|
|
|
45
|
|
|
21
|
|
|
|
|
731
|
|
24
|
|
|
|
|
|
|
|
25
|
21
|
|
|
21
|
|
197
|
use Fcntl; |
|
21
|
|
|
|
|
48
|
|
|
21
|
|
|
|
|
6066
|
|
26
|
21
|
|
|
21
|
|
213
|
use Errno qw(EBADF); |
|
21
|
|
|
|
|
44
|
|
|
21
|
|
|
|
|
1125
|
|
27
|
21
|
|
|
21
|
|
142
|
use File::Basename; |
|
21
|
|
|
|
|
43
|
|
|
21
|
|
|
|
|
1459
|
|
28
|
21
|
|
|
21
|
|
149
|
use File::Spec; |
|
21
|
|
|
|
|
44
|
|
|
21
|
|
|
|
|
809
|
|
29
|
21
|
|
|
21
|
|
126
|
use File::Path; |
|
21
|
|
|
|
|
50
|
|
|
21
|
|
|
|
|
2124
|
|
30
|
|
|
|
|
|
|
|
31
|
|
|
|
|
|
|
BEGIN { |
32
|
21
|
|
|
|
|
115
|
eval { require Digest::SHA; import Digest::SHA qw(sha1); 1 } |
|
21
|
|
|
|
|
584
|
|
|
21
|
|
|
|
|
574
|
|
33
|
21
|
50
|
|
21
|
|
76
|
or do { require Digest::SHA1; import Digest::SHA1 qw(sha1) } |
|
0
|
|
|
|
|
0
|
|
|
0
|
|
|
|
|
0
|
|
34
|
|
|
|
|
|
|
} |
35
|
|
|
|
|
|
|
|
36
|
21
|
|
|
21
|
|
126
|
use Mail::SpamAssassin; |
|
21
|
|
|
|
|
51
|
|
|
21
|
|
|
|
|
641
|
|
37
|
21
|
|
|
21
|
|
121
|
use Mail::SpamAssassin::Util qw(untaint_var am_running_on_windows); |
|
21
|
|
|
|
|
54
|
|
|
21
|
|
|
|
|
1084
|
|
38
|
21
|
|
|
21
|
|
6999
|
use Mail::SpamAssassin::BayesStore; |
|
21
|
|
|
|
|
70
|
|
|
21
|
|
|
|
|
637
|
|
39
|
21
|
|
|
21
|
|
145
|
use Mail::SpamAssassin::Logger; |
|
21
|
|
|
|
|
63
|
|
|
21
|
|
|
|
|
1565
|
|
40
|
|
|
|
|
|
|
|
41
|
21
|
|
|
21
|
|
148
|
use constant MAGIC_RE => qr/^\015\001\007\011\003/; |
|
21
|
|
|
|
|
45
|
|
|
21
|
|
|
|
|
153203
|
|
42
|
|
|
|
|
|
|
|
43
|
|
|
|
|
|
|
our ( @DBNAMES, |
44
|
|
|
|
|
|
|
$NSPAM_MAGIC_TOKEN, $NHAM_MAGIC_TOKEN, $LAST_EXPIRE_MAGIC_TOKEN, $LAST_JOURNAL_SYNC_MAGIC_TOKEN, |
45
|
|
|
|
|
|
|
$NTOKENS_MAGIC_TOKEN, $OLDEST_TOKEN_AGE_MAGIC_TOKEN, $LAST_EXPIRE_REDUCE_MAGIC_TOKEN, |
46
|
|
|
|
|
|
|
$RUNNING_EXPIRE_MAGIC_TOKEN, $DB_VERSION_MAGIC_TOKEN, $LAST_ATIME_DELTA_MAGIC_TOKEN, |
47
|
|
|
|
|
|
|
$NEWEST_TOKEN_AGE_MAGIC_TOKEN |
48
|
|
|
|
|
|
|
); |
49
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
|
our @ISA = qw( Mail::SpamAssassin::BayesStore ); |
51
|
|
|
|
|
|
|
|
52
|
|
|
|
|
|
|
# db layout (quoting Matt): |
53
|
|
|
|
|
|
|
# |
54
|
|
|
|
|
|
|
# > need five db files though to make it real fast: |
55
|
|
|
|
|
|
|
# [probs] 1. ngood and nbad (two entries, so could be a flat file rather |
56
|
|
|
|
|
|
|
# than a db file). (now 2 entries in db_toks) |
57
|
|
|
|
|
|
|
# [toks] 2. good token -> number seen |
58
|
|
|
|
|
|
|
# [toks] 3. bad token -> number seen (both are packed into 1 entry in 1 db) |
59
|
|
|
|
|
|
|
# [probs] 4. Consolidated good token -> probability |
60
|
|
|
|
|
|
|
# [probs] 5. Consolidated bad token -> probability |
61
|
|
|
|
|
|
|
# > As you add new mails, you update the entry in 2 or 3, then regenerate |
62
|
|
|
|
|
|
|
# > the entry for that token in 4 or 5. |
63
|
|
|
|
|
|
|
# > Then as you test a new mail, you just need to pull the probability |
64
|
|
|
|
|
|
|
# > direct from 4 and 5, and generate the overall probability. A simple and |
65
|
|
|
|
|
|
|
# > very fast operation. |
66
|
|
|
|
|
|
|
# |
67
|
|
|
|
|
|
|
# jm: we use probs as overall probability. <0.5 = ham, >0.5 = spam |
68
|
|
|
|
|
|
|
# |
69
|
|
|
|
|
|
|
# update: probs is no longer maintained as a db, to keep on-disk and in-core |
70
|
|
|
|
|
|
|
# usage down. |
71
|
|
|
|
|
|
|
# |
72
|
|
|
|
|
|
|
# also, added a new one to support forgetting, auto-learning, and |
73
|
|
|
|
|
|
|
# auto-forgetting for refiled mails: |
74
|
|
|
|
|
|
|
# [seen] 6. a list of Message-IDs of messages already learnt from. values |
75
|
|
|
|
|
|
|
# are 's' for learnt-as-spam, 'h' for learnt-as-ham. |
76
|
|
|
|
|
|
|
# |
77
|
|
|
|
|
|
|
# and another, called [scancount] to model the scan-count for expiry. |
78
|
|
|
|
|
|
|
# This is not a database. Instead it increases by one byte for each |
79
|
|
|
|
|
|
|
# message scanned (note: scanned, not learned). |
80
|
|
|
|
|
|
|
|
81
|
|
|
|
|
|
|
@DBNAMES = qw(toks seen); |
82
|
|
|
|
|
|
|
|
83
|
|
|
|
|
|
|
# These are the magic tokens we use to track stuff in the DB. |
84
|
|
|
|
|
|
|
# The format is '^M^A^G^I^C' followed by any string you want. |
85
|
|
|
|
|
|
|
# None of the control chars will be in a real token. |
86
|
|
|
|
|
|
|
$DB_VERSION_MAGIC_TOKEN = "\015\001\007\011\003DBVERSION"; |
87
|
|
|
|
|
|
|
$LAST_ATIME_DELTA_MAGIC_TOKEN = "\015\001\007\011\003LASTATIMEDELTA"; |
88
|
|
|
|
|
|
|
$LAST_EXPIRE_MAGIC_TOKEN = "\015\001\007\011\003LASTEXPIRE"; |
89
|
|
|
|
|
|
|
$LAST_EXPIRE_REDUCE_MAGIC_TOKEN = "\015\001\007\011\003LASTEXPIREREDUCE"; |
90
|
|
|
|
|
|
|
$LAST_JOURNAL_SYNC_MAGIC_TOKEN = "\015\001\007\011\003LASTJOURNALSYNC"; |
91
|
|
|
|
|
|
|
$NEWEST_TOKEN_AGE_MAGIC_TOKEN = "\015\001\007\011\003NEWESTAGE"; |
92
|
|
|
|
|
|
|
$NHAM_MAGIC_TOKEN = "\015\001\007\011\003NHAM"; |
93
|
|
|
|
|
|
|
$NSPAM_MAGIC_TOKEN = "\015\001\007\011\003NSPAM"; |
94
|
|
|
|
|
|
|
$NTOKENS_MAGIC_TOKEN = "\015\001\007\011\003NTOKENS"; |
95
|
|
|
|
|
|
|
$OLDEST_TOKEN_AGE_MAGIC_TOKEN = "\015\001\007\011\003OLDESTAGE"; |
96
|
|
|
|
|
|
|
$RUNNING_EXPIRE_MAGIC_TOKEN = "\015\001\007\011\003RUNNINGEXPIRE"; |
97
|
|
|
|
|
|
|
|
98
|
|
|
|
|
|
|
sub HAS_DBM_MODULE { |
99
|
136
|
|
|
136
|
0
|
360
|
my ($self) = @_; |
100
|
136
|
100
|
|
|
|
482
|
if (exists($self->{has_dbm_module})) { |
101
|
79
|
|
|
|
|
307
|
return $self->{has_dbm_module}; |
102
|
|
|
|
|
|
|
} |
103
|
57
|
|
|
|
|
120
|
$self->{has_dbm_module} = eval { require DB_File; }; |
|
57
|
|
|
|
|
11255
|
|
104
|
|
|
|
|
|
|
} |
105
|
|
|
|
|
|
|
|
106
|
|
|
|
|
|
|
sub DBM_MODULE { |
107
|
136
|
|
|
136
|
0
|
552
|
return "DB_File"; |
108
|
|
|
|
|
|
|
} |
109
|
|
|
|
|
|
|
|
110
|
|
|
|
|
|
|
# Possible file extensions used by the kinds of database files DB_File |
111
|
|
|
|
|
|
|
# might create. We need these so we can create a new file and rename |
112
|
|
|
|
|
|
|
# it into place. |
113
|
|
|
|
|
|
|
sub DB_EXTENSIONS { |
114
|
0
|
|
|
0
|
0
|
0
|
return ('', '.db'); |
115
|
|
|
|
|
|
|
} |
116
|
|
|
|
|
|
|
|
117
|
|
|
|
|
|
|
########################################################################### |
118
|
|
|
|
|
|
|
|
119
|
|
|
|
|
|
|
sub new { |
120
|
63
|
|
|
63
|
1
|
205
|
my $class = shift; |
121
|
63
|
|
33
|
|
|
479
|
$class = ref($class) || $class; |
122
|
|
|
|
|
|
|
|
123
|
63
|
|
|
|
|
533
|
my $self = $class->SUPER::new(@_); |
124
|
|
|
|
|
|
|
|
125
|
63
|
|
|
|
|
274
|
$self->{supported_db_version} = 3; |
126
|
|
|
|
|
|
|
|
127
|
63
|
|
|
|
|
205
|
$self->{already_tied} = 0; |
128
|
63
|
|
|
|
|
174
|
$self->{is_locked} = 0; |
129
|
63
|
|
|
|
|
204
|
$self->{string_to_journal} = ''; |
130
|
|
|
|
|
|
|
|
131
|
63
|
|
|
|
|
1108
|
$self; |
132
|
|
|
|
|
|
|
} |
133
|
|
|
|
|
|
|
|
134
|
|
|
|
|
|
|
########################################################################### |
135
|
|
|
|
|
|
|
|
136
|
|
|
|
|
|
|
sub tie_db_readonly { |
137
|
148
|
|
|
148
|
1
|
357
|
my ($self) = @_; |
138
|
|
|
|
|
|
|
|
139
|
148
|
100
|
|
|
|
630
|
if (!$self->HAS_DBM_MODULE) { |
140
|
134
|
|
|
|
|
576
|
dbg("bayes: %s module not installed, cannot use bayes", $self->DBM_MODULE); |
141
|
134
|
|
|
|
|
696
|
return 0; |
142
|
|
|
|
|
|
|
} |
143
|
|
|
|
|
|
|
|
144
|
|
|
|
|
|
|
# return if we've already tied to the db's, using the same mode |
145
|
|
|
|
|
|
|
# (locked/unlocked) as before. |
146
|
14
|
100
|
66
|
|
|
87
|
return 1 if ($self->{already_tied} && $self->{is_locked} == 0); |
147
|
|
|
|
|
|
|
|
148
|
12
|
|
|
|
|
29
|
my $main = $self->{bayes}->{main}; |
149
|
12
|
50
|
|
|
|
40
|
if (!defined($main->{conf}->{bayes_path})) { |
150
|
0
|
|
|
|
|
0
|
dbg("bayes: bayes_path not defined"); |
151
|
0
|
|
|
|
|
0
|
return 0; |
152
|
|
|
|
|
|
|
} |
153
|
|
|
|
|
|
|
|
154
|
12
|
|
|
|
|
51
|
$self->read_db_configs(); |
155
|
|
|
|
|
|
|
|
156
|
12
|
|
|
|
|
57
|
my $path = $main->sed_path($main->{conf}->{bayes_path}); |
157
|
|
|
|
|
|
|
|
158
|
12
|
|
|
|
|
22
|
my $found = 0; |
159
|
12
|
|
|
|
|
48
|
for my $ext ($self->DB_EXTENSIONS) { |
160
|
20
|
100
|
|
|
|
413
|
if (-f $path.'_toks'.$ext) { |
161
|
4
|
|
|
|
|
23
|
$found = 1; |
162
|
4
|
|
|
|
|
16
|
last; |
163
|
|
|
|
|
|
|
} |
164
|
|
|
|
|
|
|
} |
165
|
|
|
|
|
|
|
|
166
|
12
|
100
|
|
|
|
48
|
if (!$found) { |
167
|
8
|
|
|
|
|
57
|
dbg("bayes: no dbs present, cannot tie DB R/O: %s", $path.'_toks'); |
168
|
8
|
|
|
|
|
76
|
return 0; |
169
|
|
|
|
|
|
|
} |
170
|
|
|
|
|
|
|
|
171
|
4
|
|
|
|
|
32
|
foreach my $dbname (@DBNAMES) { |
172
|
8
|
|
|
|
|
35
|
my $name = $path.'_'.$dbname; |
173
|
8
|
|
|
|
|
23
|
my $db_var = 'db_'.$dbname; |
174
|
8
|
|
|
|
|
46
|
dbg("bayes: tie-ing to DB file R/O $name"); |
175
|
|
|
|
|
|
|
|
176
|
|
|
|
|
|
|
# Bug 6901, [rt.cpan.org #83060] |
177
|
|
|
|
|
|
|
# DB_File: Repeated tie to the same hash with no untie causes corruption |
178
|
8
|
|
|
|
|
14
|
untie %{$self->{$db_var}}; # has no effect if the variable is not tied |
|
8
|
|
|
|
|
42
|
|
179
|
|
|
|
|
|
|
|
180
|
8
|
100
|
|
|
|
22
|
if (!tie %{$self->{$db_var}}, $self->DBM_MODULE, $name, O_RDONLY, |
|
8
|
|
|
|
|
51
|
|
181
|
|
|
|
|
|
|
(oct($main->{conf}->{bayes_file_mode}) & 0666)) |
182
|
|
|
|
|
|
|
{ |
183
|
|
|
|
|
|
|
# bug 2975: it's acceptable for the db_seen to not be present, |
184
|
|
|
|
|
|
|
# to allow it to be recycled. if that's the case, just create |
185
|
|
|
|
|
|
|
# a new, empty one. we don't need to lock it, since we won't |
186
|
|
|
|
|
|
|
# be writing to it; let the R/W api deal with that case. |
187
|
|
|
|
|
|
|
|
188
|
1
|
50
|
|
|
|
8
|
if ($dbname eq 'seen') { |
189
|
|
|
|
|
|
|
# Bug 6901, [rt.cpan.org #83060] |
190
|
1
|
|
|
|
|
7
|
untie %{$self->{$db_var}}; # has no effect if the variable is not tied |
|
1
|
|
|
|
|
5
|
|
191
|
1
|
|
|
|
|
6
|
tie %{$self->{$db_var}}, $self->DBM_MODULE, $name, O_RDWR|O_CREAT, |
192
|
1
|
50
|
|
|
|
2
|
(oct($main->{conf}->{bayes_file_mode}) & 0666) |
193
|
|
|
|
|
|
|
or goto failed_to_tie; |
194
|
|
|
|
|
|
|
} |
195
|
|
|
|
|
|
|
else { |
196
|
0
|
|
|
|
|
0
|
goto failed_to_tie; |
197
|
|
|
|
|
|
|
} |
198
|
|
|
|
|
|
|
} |
199
|
|
|
|
|
|
|
} |
200
|
|
|
|
|
|
|
|
201
|
4
|
|
|
|
|
62
|
$self->{db_version} = ($self->get_storage_variables())[6]; |
202
|
4
|
|
|
|
|
40
|
dbg("bayes: found bayes db version %s", $self->{db_version}); |
203
|
|
|
|
|
|
|
|
204
|
|
|
|
|
|
|
# If the DB version is one we don't understand, abort! |
205
|
4
|
50
|
|
|
|
31
|
if ($self->_check_db_version() != 0) { |
206
|
0
|
|
|
|
|
0
|
warn("bayes: bayes db version ".$self->{db_version}." is not able to be used, aborting!"); |
207
|
0
|
|
|
|
|
0
|
$self->untie_db(); |
208
|
0
|
|
|
|
|
0
|
return 0; |
209
|
|
|
|
|
|
|
} |
210
|
|
|
|
|
|
|
|
211
|
4
|
|
|
|
|
11
|
$self->{already_tied} = 1; |
212
|
4
|
|
|
|
|
24
|
return 1; |
213
|
|
|
|
|
|
|
|
214
|
0
|
|
|
|
|
0
|
failed_to_tie: |
215
|
|
|
|
|
|
|
warn "bayes: cannot open bayes databases ${path}_* R/O: tie failed: $!\n"; |
216
|
0
|
|
|
|
|
0
|
foreach my $dbname (@DBNAMES) { |
217
|
0
|
|
|
|
|
0
|
my $db_var = 'db_'.$dbname; |
218
|
0
|
0
|
|
|
|
0
|
next unless exists $self->{$db_var}; |
219
|
0
|
|
|
|
|
0
|
dbg("bayes: untie-ing DB file $dbname"); |
220
|
0
|
|
|
|
|
0
|
untie %{$self->{$db_var}}; |
|
0
|
|
|
|
|
0
|
|
221
|
|
|
|
|
|
|
} |
222
|
|
|
|
|
|
|
|
223
|
0
|
|
|
|
|
0
|
return 0; |
224
|
|
|
|
|
|
|
} |
225
|
|
|
|
|
|
|
|
226
|
|
|
|
|
|
|
# tie() to the databases, read-write and locked. Any callers of |
227
|
|
|
|
|
|
|
# this should ensure they call untie_db() afterwards! |
228
|
|
|
|
|
|
|
# |
229
|
|
|
|
|
|
|
sub tie_db_writable { |
230
|
30
|
|
|
30
|
1
|
77
|
my ($self) = @_; |
231
|
|
|
|
|
|
|
|
232
|
30
|
100
|
|
|
|
125
|
if (!$self->HAS_DBM_MODULE) { |
233
|
2
|
|
|
|
|
7
|
dbg("bayes: %s module not installed, cannot use bayes", $self->DBM_MODULE); |
234
|
2
|
|
|
|
|
5
|
return 0; |
235
|
|
|
|
|
|
|
} |
236
|
|
|
|
|
|
|
|
237
|
|
|
|
|
|
|
# Useful shortcut ... |
238
|
28
|
|
|
|
|
82
|
my $main = $self->{bayes}->{main}; |
239
|
|
|
|
|
|
|
|
240
|
|
|
|
|
|
|
# if we've already tied the db's using the same mode |
241
|
|
|
|
|
|
|
# (locked/unlocked) as we want now, freshen the lock and return. |
242
|
28
|
100
|
66
|
|
|
98
|
if ($self->{already_tied} && $self->{is_locked} == 1) { |
243
|
2
|
|
|
|
|
14
|
$main->{locker}->refresh_lock($self->{locked_file}); |
244
|
2
|
|
|
|
|
9
|
return 1; |
245
|
|
|
|
|
|
|
} |
246
|
|
|
|
|
|
|
|
247
|
26
|
50
|
|
|
|
84
|
if (!defined($main->{conf}->{bayes_path})) { |
248
|
0
|
|
|
|
|
0
|
dbg("bayes: bayes_path not defined"); |
249
|
0
|
|
|
|
|
0
|
return 0; |
250
|
|
|
|
|
|
|
} |
251
|
|
|
|
|
|
|
|
252
|
26
|
|
|
|
|
110
|
$self->read_db_configs(); |
253
|
|
|
|
|
|
|
|
254
|
26
|
|
|
|
|
104
|
my $path = $main->sed_path($main->{conf}->{bayes_path}); |
255
|
|
|
|
|
|
|
|
256
|
26
|
|
|
|
|
51
|
my $found = 0; |
257
|
26
|
|
|
|
|
84
|
for my $ext ($self->DB_EXTENSIONS) { |
258
|
30
|
100
|
|
|
|
507
|
if (-f $path.'_toks'.$ext) { |
259
|
22
|
|
|
|
|
63
|
$found = 1; |
260
|
22
|
|
|
|
|
45
|
last; |
261
|
|
|
|
|
|
|
} |
262
|
|
|
|
|
|
|
} |
263
|
|
|
|
|
|
|
|
264
|
26
|
|
|
|
|
1265
|
my $parentdir = dirname($path); |
265
|
26
|
50
|
|
|
|
413
|
if (!-d $parentdir) { |
266
|
|
|
|
|
|
|
# run in an eval(); if mkpath has no perms, it calls die() |
267
|
0
|
|
|
|
|
0
|
eval { |
268
|
0
|
|
|
|
|
0
|
mkpath($parentdir, 0, (oct($main->{conf}->{bayes_file_mode}) & 0777)); |
269
|
|
|
|
|
|
|
}; |
270
|
|
|
|
|
|
|
} |
271
|
|
|
|
|
|
|
|
272
|
26
|
|
|
|
|
60
|
my $tout; |
273
|
26
|
50
|
|
|
|
76
|
if ($main->{learn_wait_for_lock}) { |
274
|
0
|
|
|
|
|
0
|
$tout = 300; # TODO: Dan to write better lock code |
275
|
|
|
|
|
|
|
} else { |
276
|
26
|
|
|
|
|
45
|
$tout = 10; |
277
|
|
|
|
|
|
|
} |
278
|
26
|
50
|
|
|
|
203
|
if ($main->{locker}->safe_lock($path, $tout, $main->{conf}->{bayes_file_mode})) |
279
|
|
|
|
|
|
|
{ |
280
|
26
|
|
|
|
|
67
|
$self->{locked_file} = $path; |
281
|
26
|
|
|
|
|
52
|
$self->{is_locked} = 1; |
282
|
|
|
|
|
|
|
} else { |
283
|
0
|
|
|
|
|
0
|
warn "bayes: cannot open bayes databases ${path}_* R/W: lock failed: $!\n"; |
284
|
0
|
|
|
|
|
0
|
return 0; |
285
|
|
|
|
|
|
|
} |
286
|
|
|
|
|
|
|
|
287
|
26
|
|
|
|
|
139
|
my $umask = umask 0; |
288
|
26
|
|
|
|
|
106
|
foreach my $dbname (@DBNAMES) { |
289
|
52
|
|
|
|
|
169
|
my $name = $path.'_'.$dbname; |
290
|
52
|
|
|
|
|
89
|
my $db_var = 'db_'.$dbname; |
291
|
52
|
|
|
|
|
182
|
dbg("bayes: tie-ing to DB file R/W $name"); |
292
|
|
|
|
|
|
|
|
293
|
52
|
50
|
|
|
|
168
|
($self->DBM_MODULE eq 'DB_File') and |
294
|
|
|
|
|
|
|
Mail::SpamAssassin::Util::avoid_db_file_locking_bug ($name); |
295
|
|
|
|
|
|
|
|
296
|
|
|
|
|
|
|
# Bug 6901, [rt.cpan.org #83060] |
297
|
52
|
|
|
|
|
74
|
untie %{$self->{$db_var}}; # has no effect if the variable is not tied |
|
52
|
|
|
|
|
159
|
|
298
|
52
|
|
|
|
|
146
|
tie %{$self->{$db_var}}, $self->DBM_MODULE, $name, O_RDWR|O_CREAT, |
299
|
52
|
50
|
|
|
|
74
|
(oct($main->{conf}->{bayes_file_mode}) & 0666) |
300
|
|
|
|
|
|
|
or goto failed_to_tie; |
301
|
|
|
|
|
|
|
} |
302
|
26
|
|
|
|
|
126
|
umask $umask; |
303
|
|
|
|
|
|
|
|
304
|
|
|
|
|
|
|
# set our cache to what version DB we're using |
305
|
26
|
|
|
|
|
147
|
$self->{db_version} = ($self->get_storage_variables())[6]; |
306
|
|
|
|
|
|
|
# don't bother printing this unless found since it would be bogus anyway |
307
|
26
|
100
|
|
|
|
131
|
dbg("bayes: found bayes db version %s", $self->{db_version}) if $found; |
308
|
|
|
|
|
|
|
|
309
|
|
|
|
|
|
|
# figure out if we can read the current DB and if we need to do a |
310
|
|
|
|
|
|
|
# DB version update and do it if necessary if either has a problem, |
311
|
|
|
|
|
|
|
# fail immediately |
312
|
|
|
|
|
|
|
# |
313
|
26
|
50
|
66
|
|
|
163
|
if ($found && !$self->_upgrade_db()) { |
|
|
100
|
|
|
|
|
|
314
|
0
|
|
|
|
|
0
|
$self->untie_db(); |
315
|
0
|
|
|
|
|
0
|
return 0; |
316
|
|
|
|
|
|
|
} |
317
|
|
|
|
|
|
|
elsif (!$found) { # new DB, make sure we know that ... |
318
|
4
|
|
|
|
|
29
|
$self->{db_version} = $self->{db_toks}->{$DB_VERSION_MAGIC_TOKEN} = $self->DB_VERSION; |
319
|
4
|
|
|
|
|
64
|
$self->{db_toks}->{$NTOKENS_MAGIC_TOKEN} = 0; # no tokens in the db ... |
320
|
4
|
|
|
|
|
24
|
dbg("bayes: new db, set db version %s and 0 tokens", $self->{db_version}); |
321
|
|
|
|
|
|
|
} |
322
|
|
|
|
|
|
|
|
323
|
26
|
|
|
|
|
50
|
$self->{already_tied} = 1; |
324
|
26
|
|
|
|
|
111
|
return 1; |
325
|
|
|
|
|
|
|
|
326
|
0
|
|
|
|
|
0
|
failed_to_tie: |
327
|
|
|
|
|
|
|
my $err = $!; |
328
|
0
|
|
|
|
|
0
|
umask $umask; |
329
|
|
|
|
|
|
|
|
330
|
0
|
|
|
|
|
0
|
foreach my $dbname (@DBNAMES) { |
331
|
0
|
|
|
|
|
0
|
my $db_var = 'db_'.$dbname; |
332
|
0
|
0
|
|
|
|
0
|
next unless exists $self->{$db_var}; |
333
|
0
|
|
|
|
|
0
|
dbg("bayes: untie-ing DB file $dbname"); |
334
|
0
|
|
|
|
|
0
|
untie %{$self->{$db_var}}; |
|
0
|
|
|
|
|
0
|
|
335
|
|
|
|
|
|
|
} |
336
|
|
|
|
|
|
|
|
337
|
0
|
0
|
|
|
|
0
|
if ($self->{is_locked}) { |
338
|
0
|
|
|
|
|
0
|
$self->{bayes}->{main}->{locker}->safe_unlock($self->{locked_file}); |
339
|
0
|
|
|
|
|
0
|
$self->{is_locked} = 0; |
340
|
|
|
|
|
|
|
} |
341
|
0
|
|
|
|
|
0
|
warn "bayes: cannot open bayes databases ${path}_* R/W: tie failed: $err\n"; |
342
|
0
|
|
|
|
|
0
|
return 0; |
343
|
|
|
|
|
|
|
} |
344
|
|
|
|
|
|
|
|
345
|
|
|
|
|
|
|
# Do we understand how to deal with this DB version? |
346
|
|
|
|
|
|
|
sub _check_db_version { |
347
|
26
|
|
|
26
|
|
53
|
my ($self) = @_; |
348
|
|
|
|
|
|
|
|
349
|
|
|
|
|
|
|
# return -1 if older, 0 if current, 1 if newer |
350
|
26
|
|
|
|
|
103
|
return $self->{db_version} <=> $self->DB_VERSION; |
351
|
|
|
|
|
|
|
} |
352
|
|
|
|
|
|
|
|
353
|
|
|
|
|
|
|
# Check to see if we need to upgrade the DB, and do so if necessary |
354
|
|
|
|
|
|
|
sub _upgrade_db { |
355
|
22
|
|
|
22
|
|
62
|
my ($self) = @_; |
356
|
|
|
|
|
|
|
|
357
|
22
|
|
|
|
|
81
|
my $verschk = $self->_check_db_version(); |
358
|
22
|
|
|
|
|
36
|
my $res = 0; # used later on for tie() checks |
359
|
22
|
|
|
|
|
31
|
my $umask; # used later for umask modifications |
360
|
|
|
|
|
|
|
|
361
|
|
|
|
|
|
|
# If the DB is the latest version, no problem. |
362
|
22
|
50
|
|
|
|
125
|
return 1 if ($verschk == 0); |
363
|
|
|
|
|
|
|
|
364
|
|
|
|
|
|
|
# If the DB is a newer version that we know what to do with ... abort! |
365
|
0
|
0
|
|
|
|
0
|
if ($verschk == 1) { |
366
|
0
|
|
|
|
|
0
|
warn("bayes: bayes db version ".$self->{db_version}." is newer than we understand, aborting!"); |
367
|
0
|
|
|
|
|
0
|
return 0; |
368
|
|
|
|
|
|
|
} |
369
|
|
|
|
|
|
|
|
370
|
|
|
|
|
|
|
# If the current DB version is lower than the new version, upgrade! |
371
|
|
|
|
|
|
|
# Do conversions in order so we can go 1 -> 3, make sure to update |
372
|
|
|
|
|
|
|
# $self->{db_version} along the way |
373
|
|
|
|
|
|
|
|
374
|
0
|
|
|
|
|
0
|
dbg("bayes: detected bayes db format %s, upgrading", $self->{db_version}); |
375
|
|
|
|
|
|
|
|
376
|
|
|
|
|
|
|
# since DB_File will not shrink a database (!!), we need to *create* |
377
|
|
|
|
|
|
|
# a new one instead. |
378
|
0
|
|
|
|
|
0
|
my $main = $self->{bayes}->{main}; |
379
|
0
|
|
|
|
|
0
|
my $path = $main->sed_path($main->{conf}->{bayes_path}); |
380
|
0
|
|
|
|
|
0
|
my $name = $path.'_toks'; |
381
|
|
|
|
|
|
|
|
382
|
|
|
|
|
|
|
# older version's journal files are likely not in the same format as the new ones, so remove it. |
383
|
0
|
|
|
|
|
0
|
my $jpath = $self->_get_journal_filename(); |
384
|
0
|
0
|
|
|
|
0
|
if (-f $jpath) { |
385
|
0
|
|
|
|
|
0
|
dbg("bayes: old journal file found, removing"); |
386
|
0
|
0
|
|
|
|
0
|
warn "bayes: couldn't remove $jpath: $!" if (!unlink $jpath); |
387
|
|
|
|
|
|
|
} |
388
|
|
|
|
|
|
|
|
389
|
0
|
0
|
|
|
|
0
|
if ($self->{db_version} < 2) { |
390
|
0
|
|
|
|
|
0
|
dbg("bayes: upgrading database format from v%s to v2", $self->{db_version}); |
391
|
0
|
|
|
|
|
0
|
$self->set_running_expire_tok(); |
392
|
|
|
|
|
|
|
|
393
|
0
|
|
|
|
|
0
|
my ($DB_NSPAM_MAGIC_TOKEN, $DB_NHAM_MAGIC_TOKEN, $DB_NTOKENS_MAGIC_TOKEN); |
394
|
0
|
|
|
|
|
0
|
my ($DB_OLDEST_TOKEN_AGE_MAGIC_TOKEN, $DB_LAST_EXPIRE_MAGIC_TOKEN); |
395
|
|
|
|
|
|
|
|
396
|
|
|
|
|
|
|
# Magic tokens for version 0, defined as '**[A-Z]+' |
397
|
0
|
0
|
|
|
|
0
|
if ($self->{db_version} == 0) { |
398
|
0
|
|
|
|
|
0
|
$DB_NSPAM_MAGIC_TOKEN = '**NSPAM'; |
399
|
0
|
|
|
|
|
0
|
$DB_NHAM_MAGIC_TOKEN = '**NHAM'; |
400
|
0
|
|
|
|
|
0
|
$DB_NTOKENS_MAGIC_TOKEN = '**NTOKENS'; |
401
|
|
|
|
|
|
|
#$DB_OLDEST_TOKEN_AGE_MAGIC_TOKEN = '**OLDESTAGE'; |
402
|
|
|
|
|
|
|
#$DB_LAST_EXPIRE_MAGIC_TOKEN = '**LASTEXPIRE'; |
403
|
|
|
|
|
|
|
#$DB_SCANCOUNT_BASE_MAGIC_TOKEN = '**SCANBASE'; |
404
|
|
|
|
|
|
|
#$DB_RUNNING_EXPIRE_MAGIC_TOKEN = '**RUNNINGEXPIRE'; |
405
|
|
|
|
|
|
|
} |
406
|
|
|
|
|
|
|
else { |
407
|
0
|
|
|
|
|
0
|
$DB_NSPAM_MAGIC_TOKEN = "\015\001\007\011\003NSPAM"; |
408
|
0
|
|
|
|
|
0
|
$DB_NHAM_MAGIC_TOKEN = "\015\001\007\011\003NHAM"; |
409
|
0
|
|
|
|
|
0
|
$DB_NTOKENS_MAGIC_TOKEN = "\015\001\007\011\003NTOKENS"; |
410
|
|
|
|
|
|
|
#$DB_OLDEST_TOKEN_AGE_MAGIC_TOKEN = "\015\001\007\011\003OLDESTAGE"; |
411
|
|
|
|
|
|
|
#$DB_LAST_EXPIRE_MAGIC_TOKEN = "\015\001\007\011\003LASTEXPIRE"; |
412
|
|
|
|
|
|
|
#$DB_SCANCOUNT_BASE_MAGIC_TOKEN = "\015\001\007\011\003SCANBASE"; |
413
|
|
|
|
|
|
|
#$DB_RUNNING_EXPIRE_MAGIC_TOKEN = "\015\001\007\011\003RUNNINGEXPIRE"; |
414
|
|
|
|
|
|
|
} |
415
|
|
|
|
|
|
|
|
416
|
|
|
|
|
|
|
# remember when we started ... |
417
|
0
|
|
|
|
|
0
|
my $started = time; |
418
|
0
|
|
|
|
|
0
|
my $newatime = $started; |
419
|
|
|
|
|
|
|
|
420
|
|
|
|
|
|
|
# use O_EXCL to avoid races (bonus paranoia, since we should be locked |
421
|
|
|
|
|
|
|
# anyway) |
422
|
0
|
|
|
|
|
0
|
my %new_toks; |
423
|
0
|
|
|
|
|
0
|
$umask = umask 0; |
424
|
|
|
|
|
|
|
|
425
|
|
|
|
|
|
|
$res = tie %new_toks, $self->DBM_MODULE, "${name}.new", |
426
|
|
|
|
|
|
|
O_RDWR|O_CREAT|O_EXCL, |
427
|
0
|
|
|
|
|
0
|
(oct($main->{conf}->{bayes_file_mode}) & 0666); |
428
|
0
|
|
|
|
|
0
|
umask $umask; |
429
|
0
|
0
|
|
|
|
0
|
return 0 unless $res; |
430
|
0
|
|
|
|
|
0
|
undef $res; |
431
|
|
|
|
|
|
|
|
432
|
|
|
|
|
|
|
# add the magic tokens to the new db. |
433
|
0
|
|
|
|
|
0
|
$new_toks{$NSPAM_MAGIC_TOKEN} = $self->{db_toks}->{$DB_NSPAM_MAGIC_TOKEN}; |
434
|
0
|
|
|
|
|
0
|
$new_toks{$NHAM_MAGIC_TOKEN} = $self->{db_toks}->{$DB_NHAM_MAGIC_TOKEN}; |
435
|
0
|
|
|
|
|
0
|
$new_toks{$NTOKENS_MAGIC_TOKEN} = $self->{db_toks}->{$DB_NTOKENS_MAGIC_TOKEN}; |
436
|
0
|
|
|
|
|
0
|
$new_toks{$DB_VERSION_MAGIC_TOKEN} = 2; # we're now a DB version 2 file |
437
|
0
|
|
|
|
|
0
|
$new_toks{$OLDEST_TOKEN_AGE_MAGIC_TOKEN} = $newatime; |
438
|
0
|
|
|
|
|
0
|
$new_toks{$LAST_EXPIRE_MAGIC_TOKEN} = $newatime; |
439
|
0
|
|
|
|
|
0
|
$new_toks{$NEWEST_TOKEN_AGE_MAGIC_TOKEN} = $newatime; |
440
|
0
|
|
|
|
|
0
|
$new_toks{$LAST_JOURNAL_SYNC_MAGIC_TOKEN} = $newatime; |
441
|
0
|
|
|
|
|
0
|
$new_toks{$LAST_ATIME_DELTA_MAGIC_TOKEN} = 0; |
442
|
0
|
|
|
|
|
0
|
$new_toks{$LAST_EXPIRE_REDUCE_MAGIC_TOKEN} = 0; |
443
|
|
|
|
|
|
|
|
444
|
|
|
|
|
|
|
# deal with the data tokens |
445
|
0
|
|
|
|
|
0
|
my ($tok, $packed); |
446
|
0
|
|
|
|
|
0
|
my $count = 0; |
447
|
0
|
|
|
|
|
0
|
while (($tok, $packed) = each %{$self->{db_toks}}) { |
|
0
|
|
|
|
|
0
|
|
448
|
0
|
0
|
|
|
|
0
|
next if ($tok =~ /^(?:\*\*[A-Z]+$|\015\001\007\011\003)/); # skip magic tokens |
449
|
|
|
|
|
|
|
|
450
|
0
|
|
|
|
|
0
|
my ($ts, $th, $atime) = $self->tok_unpack($packed); |
451
|
0
|
|
|
|
|
0
|
$new_toks{$tok} = $self->tok_pack($ts, $th, $newatime); |
452
|
|
|
|
|
|
|
|
453
|
|
|
|
|
|
|
# Refresh the lock every so often... |
454
|
0
|
0
|
|
|
|
0
|
if (($count++ % 1000) == 0) { |
455
|
0
|
|
|
|
|
0
|
$self->set_running_expire_tok(); |
456
|
|
|
|
|
|
|
} |
457
|
|
|
|
|
|
|
} |
458
|
|
|
|
|
|
|
|
459
|
|
|
|
|
|
|
|
460
|
|
|
|
|
|
|
# now untie so we can do renames |
461
|
0
|
|
|
|
|
0
|
untie %{$self->{db_toks}}; |
|
0
|
|
|
|
|
0
|
|
462
|
0
|
|
|
|
|
0
|
untie %new_toks; |
463
|
|
|
|
|
|
|
|
464
|
|
|
|
|
|
|
# This is the critical phase (moving files around), so don't allow |
465
|
|
|
|
|
|
|
# it to be interrupted. |
466
|
0
|
|
|
|
|
0
|
local $SIG{'INT'} = 'IGNORE'; |
467
|
0
|
|
|
|
|
0
|
local $SIG{'TERM'} = 'IGNORE'; |
468
|
0
|
0
|
|
|
|
0
|
local $SIG{'HUP'} = 'IGNORE' if !am_running_on_windows(); |
469
|
|
|
|
|
|
|
|
470
|
|
|
|
|
|
|
# older versions used scancount, so kill the stupid little file ... |
471
|
0
|
|
|
|
|
0
|
my $msgc = $path.'_msgcount'; |
472
|
0
|
0
|
|
|
|
0
|
if (-f $msgc) { |
473
|
0
|
|
|
|
|
0
|
dbg("bayes: old msgcount file found, removing"); |
474
|
0
|
0
|
|
|
|
0
|
if (!unlink $msgc) { |
475
|
0
|
|
|
|
|
0
|
warn "bayes: couldn't remove $msgc: $!"; |
476
|
|
|
|
|
|
|
} |
477
|
|
|
|
|
|
|
} |
478
|
|
|
|
|
|
|
|
479
|
|
|
|
|
|
|
# now rename in the new one. Try several extensions |
480
|
0
|
|
|
|
|
0
|
for my $ext ($self->DB_EXTENSIONS) { |
481
|
0
|
|
|
|
|
0
|
my $newf = $name.'.new'.$ext; |
482
|
0
|
|
|
|
|
0
|
my $oldf = $name.$ext; |
483
|
0
|
0
|
|
|
|
0
|
next unless (-f $newf); |
484
|
0
|
0
|
|
|
|
0
|
if (!rename ($newf, $oldf)) { |
485
|
0
|
|
|
|
|
0
|
warn "bayes: rename $newf to $oldf failed: $!\n"; |
486
|
0
|
|
|
|
|
0
|
return 0; |
487
|
|
|
|
|
|
|
} |
488
|
|
|
|
|
|
|
} |
489
|
|
|
|
|
|
|
|
490
|
|
|
|
|
|
|
# re-tie to the new db in read-write mode ... |
491
|
0
|
|
|
|
|
0
|
$umask = umask 0; |
492
|
|
|
|
|
|
|
# Bug 6901, [rt.cpan.org #83060] |
493
|
0
|
|
|
|
|
0
|
untie %{$self->{db_toks}}; # has no effect if the variable is not tied |
|
0
|
|
|
|
|
0
|
|
494
|
0
|
|
|
|
|
0
|
$res = tie %{$self->{db_toks}}, $self->DBM_MODULE, $name, O_RDWR|O_CREAT, |
495
|
0
|
|
|
|
|
0
|
(oct($main->{conf}->{bayes_file_mode}) & 0666); |
496
|
0
|
|
|
|
|
0
|
umask $umask; |
497
|
0
|
0
|
|
|
|
0
|
return 0 unless $res; |
498
|
0
|
|
|
|
|
0
|
undef $res; |
499
|
|
|
|
|
|
|
|
500
|
|
|
|
|
|
|
dbg("bayes: upgraded database format from v%s to v2 in %d seconds", |
501
|
0
|
|
|
|
|
0
|
$self->{db_version}, time - $started); |
502
|
0
|
|
|
|
|
0
|
$self->{db_version} = 2; # need this for other functions which check |
503
|
|
|
|
|
|
|
} |
504
|
|
|
|
|
|
|
|
505
|
|
|
|
|
|
|
# Version 3 of the database converts all existing tokens to SHA1 hashes |
506
|
0
|
0
|
|
|
|
0
|
if ($self->{db_version} == 2) { |
507
|
0
|
|
|
|
|
0
|
dbg("bayes: upgrading database format from v%s to v3", $self->{db_version}); |
508
|
0
|
|
|
|
|
0
|
$self->set_running_expire_tok(); |
509
|
|
|
|
|
|
|
|
510
|
0
|
|
|
|
|
0
|
my $DB_NSPAM_MAGIC_TOKEN = "\015\001\007\011\003NSPAM"; |
511
|
0
|
|
|
|
|
0
|
my $DB_NHAM_MAGIC_TOKEN = "\015\001\007\011\003NHAM"; |
512
|
0
|
|
|
|
|
0
|
my $DB_NTOKENS_MAGIC_TOKEN = "\015\001\007\011\003NTOKENS"; |
513
|
0
|
|
|
|
|
0
|
my $DB_OLDEST_TOKEN_AGE_MAGIC_TOKEN = "\015\001\007\011\003OLDESTAGE"; |
514
|
0
|
|
|
|
|
0
|
my $DB_LAST_EXPIRE_MAGIC_TOKEN = "\015\001\007\011\003LASTEXPIRE"; |
515
|
0
|
|
|
|
|
0
|
my $DB_NEWEST_TOKEN_AGE_MAGIC_TOKEN = "\015\001\007\011\003NEWESTAGE"; |
516
|
0
|
|
|
|
|
0
|
my $DB_LAST_JOURNAL_SYNC_MAGIC_TOKEN = "\015\001\007\011\003LASTJOURNALSYNC"; |
517
|
0
|
|
|
|
|
0
|
my $DB_LAST_ATIME_DELTA_MAGIC_TOKEN = "\015\001\007\011\003LASTATIMEDELTA"; |
518
|
0
|
|
|
|
|
0
|
my $DB_LAST_EXPIRE_REDUCE_MAGIC_TOKEN = "\015\001\007\011\003LASTEXPIREREDUCE"; |
519
|
|
|
|
|
|
|
|
520
|
|
|
|
|
|
|
# remember when we started ... |
521
|
0
|
|
|
|
|
0
|
my $started = time; |
522
|
|
|
|
|
|
|
|
523
|
|
|
|
|
|
|
# use O_EXCL to avoid races (bonus paranoia, since we should be locked |
524
|
|
|
|
|
|
|
# anyway) |
525
|
0
|
|
|
|
|
0
|
my %new_toks; |
526
|
0
|
|
|
|
|
0
|
$umask = umask 0; |
527
|
|
|
|
|
|
|
$res = tie %new_toks, $self->DBM_MODULE, "${name}.new", O_RDWR|O_CREAT|O_EXCL, |
528
|
0
|
|
|
|
|
0
|
(oct($main->{conf}->{bayes_file_mode}) & 0666); |
529
|
0
|
|
|
|
|
0
|
umask $umask; |
530
|
0
|
0
|
|
|
|
0
|
return 0 unless $res; |
531
|
0
|
|
|
|
|
0
|
undef $res; |
532
|
|
|
|
|
|
|
|
533
|
|
|
|
|
|
|
# add the magic tokens to the new db. |
534
|
0
|
|
|
|
|
0
|
$new_toks{$NSPAM_MAGIC_TOKEN} = $self->{db_toks}->{$DB_NSPAM_MAGIC_TOKEN}; |
535
|
0
|
|
|
|
|
0
|
$new_toks{$NHAM_MAGIC_TOKEN} = $self->{db_toks}->{$DB_NHAM_MAGIC_TOKEN}; |
536
|
0
|
|
|
|
|
0
|
$new_toks{$NTOKENS_MAGIC_TOKEN} = $self->{db_toks}->{$DB_NTOKENS_MAGIC_TOKEN}; |
537
|
0
|
|
|
|
|
0
|
$new_toks{$DB_VERSION_MAGIC_TOKEN} = 3; # we're now a DB version 3 file |
538
|
0
|
|
|
|
|
0
|
$new_toks{$OLDEST_TOKEN_AGE_MAGIC_TOKEN} = $self->{db_toks}->{$DB_OLDEST_TOKEN_AGE_MAGIC_TOKEN}; |
539
|
0
|
|
|
|
|
0
|
$new_toks{$LAST_EXPIRE_MAGIC_TOKEN} = $self->{db_toks}->{$DB_LAST_EXPIRE_MAGIC_TOKEN}; |
540
|
0
|
|
|
|
|
0
|
$new_toks{$NEWEST_TOKEN_AGE_MAGIC_TOKEN} = $self->{db_toks}->{$DB_NEWEST_TOKEN_AGE_MAGIC_TOKEN}; |
541
|
0
|
|
|
|
|
0
|
$new_toks{$LAST_JOURNAL_SYNC_MAGIC_TOKEN} = $self->{db_toks}->{$DB_LAST_JOURNAL_SYNC_MAGIC_TOKEN}; |
542
|
0
|
|
|
|
|
0
|
$new_toks{$LAST_ATIME_DELTA_MAGIC_TOKEN} = $self->{db_toks}->{$DB_LAST_ATIME_DELTA_MAGIC_TOKEN}; |
543
|
0
|
|
|
|
|
0
|
$new_toks{$LAST_EXPIRE_REDUCE_MAGIC_TOKEN} =$self->{db_toks}->{$DB_LAST_EXPIRE_REDUCE_MAGIC_TOKEN}; |
544
|
|
|
|
|
|
|
|
545
|
|
|
|
|
|
|
# deal with the data tokens |
546
|
0
|
|
|
|
|
0
|
my $count = 0; |
547
|
0
|
|
|
|
|
0
|
while (my ($tok, $packed) = each %{$self->{db_toks}}) { |
|
0
|
|
|
|
|
0
|
|
548
|
0
|
0
|
|
|
|
0
|
next if ($tok =~ /^\015\001\007\011\003/); # skip magic tokens |
549
|
0
|
|
|
|
|
0
|
my $tok_hash = substr(sha1($tok), -5); |
550
|
0
|
|
|
|
|
0
|
$new_toks{$tok_hash} = $packed; |
551
|
|
|
|
|
|
|
|
552
|
|
|
|
|
|
|
# Refresh the lock every so often... |
553
|
0
|
0
|
|
|
|
0
|
if (($count++ % 1000) == 0) { |
554
|
0
|
|
|
|
|
0
|
$self->set_running_expire_tok(); |
555
|
|
|
|
|
|
|
} |
556
|
|
|
|
|
|
|
} |
557
|
|
|
|
|
|
|
|
558
|
|
|
|
|
|
|
# now untie so we can do renames |
559
|
0
|
|
|
|
|
0
|
untie %{$self->{db_toks}}; |
|
0
|
|
|
|
|
0
|
|
560
|
0
|
|
|
|
|
0
|
untie %new_toks; |
561
|
|
|
|
|
|
|
|
562
|
|
|
|
|
|
|
# This is the critical phase (moving files around), so don't allow |
563
|
|
|
|
|
|
|
# it to be interrupted. |
564
|
0
|
|
|
|
|
0
|
local $SIG{'INT'} = 'IGNORE'; |
565
|
0
|
|
|
|
|
0
|
local $SIG{'TERM'} = 'IGNORE'; |
566
|
0
|
0
|
|
|
|
0
|
local $SIG{'HUP'} = 'IGNORE' if !am_running_on_windows(); |
567
|
|
|
|
|
|
|
|
568
|
|
|
|
|
|
|
# now rename in the new one. Try several extensions |
569
|
0
|
|
|
|
|
0
|
for my $ext ($self->DB_EXTENSIONS) { |
570
|
0
|
|
|
|
|
0
|
my $newf = $name.'.new'.$ext; |
571
|
0
|
|
|
|
|
0
|
my $oldf = $name.$ext; |
572
|
0
|
0
|
|
|
|
0
|
next unless (-f $newf); |
573
|
0
|
0
|
|
|
|
0
|
if (!rename($newf, $oldf)) { |
574
|
0
|
|
|
|
|
0
|
warn "bayes: rename $newf to $oldf failed: $!\n"; |
575
|
0
|
|
|
|
|
0
|
return 0; |
576
|
|
|
|
|
|
|
} |
577
|
|
|
|
|
|
|
} |
578
|
|
|
|
|
|
|
|
579
|
|
|
|
|
|
|
# re-tie to the new db in read-write mode ... |
580
|
0
|
|
|
|
|
0
|
$umask = umask 0; |
581
|
|
|
|
|
|
|
# Bug 6901, [rt.cpan.org #83060] |
582
|
0
|
|
|
|
|
0
|
untie %{$self->{db_toks}}; # has no effect if the variable is not tied |
|
0
|
|
|
|
|
0
|
|
583
|
0
|
|
|
|
|
0
|
$res = tie %{$self->{db_toks}}, $self->DBM_MODULE, $name, O_RDWR|O_CREAT, |
584
|
0
|
|
|
|
|
0
|
(oct ($main->{conf}->{bayes_file_mode}) & 0666); |
585
|
0
|
|
|
|
|
0
|
umask $umask; |
586
|
0
|
0
|
|
|
|
0
|
return 0 unless $res; |
587
|
0
|
|
|
|
|
0
|
undef $res; |
588
|
|
|
|
|
|
|
|
589
|
|
|
|
|
|
|
dbg("bayes: upgraded database format from v%s to v3 in %d seconds", |
590
|
0
|
|
|
|
|
0
|
$self->{db_version}, time - $started); |
591
|
|
|
|
|
|
|
|
592
|
0
|
|
|
|
|
0
|
$self->{db_version} = 3; # need this for other functions which check |
593
|
|
|
|
|
|
|
} |
594
|
|
|
|
|
|
|
|
595
|
|
|
|
|
|
|
# if ($self->{db_version} == 3) { |
596
|
|
|
|
|
|
|
# ... |
597
|
|
|
|
|
|
|
# $self->{db_version} = 4; # need this for other functions which check |
598
|
|
|
|
|
|
|
# } |
599
|
|
|
|
|
|
|
# ... and so on. |
600
|
|
|
|
|
|
|
|
601
|
0
|
|
|
|
|
0
|
return 1; |
602
|
|
|
|
|
|
|
} |
603
|
|
|
|
|
|
|
|
604
|
|
|
|
|
|
|
########################################################################### |
605
|
|
|
|
|
|
|
|
606
|
|
|
|
|
|
|
sub untie_db { |
607
|
69
|
|
|
69
|
1
|
170
|
my $self = shift; |
608
|
|
|
|
|
|
|
|
609
|
69
|
100
|
|
|
|
367
|
return if (!$self->{already_tied}); |
610
|
|
|
|
|
|
|
|
611
|
30
|
|
|
|
|
119
|
dbg("bayes: untie-ing"); |
612
|
|
|
|
|
|
|
|
613
|
30
|
|
|
|
|
88
|
foreach my $dbname (@DBNAMES) { |
614
|
60
|
|
|
|
|
141
|
my $db_var = 'db_'.$dbname; |
615
|
|
|
|
|
|
|
|
616
|
60
|
100
|
|
|
|
143
|
if (exists $self->{$db_var}) { |
617
|
|
|
|
|
|
|
# dbg("bayes: untie-ing $db_var"); |
618
|
56
|
|
|
|
|
71
|
untie %{$self->{$db_var}}; |
|
56
|
|
|
|
|
872
|
|
619
|
56
|
|
|
|
|
232
|
delete $self->{$db_var}; |
620
|
|
|
|
|
|
|
} |
621
|
|
|
|
|
|
|
} |
622
|
|
|
|
|
|
|
|
623
|
30
|
100
|
|
|
|
92
|
if ($self->{is_locked}) { |
624
|
26
|
|
|
|
|
79
|
dbg("bayes: files locked, now unlocking lock"); |
625
|
26
|
|
|
|
|
204
|
$self->{bayes}->{main}->{locker}->safe_unlock ($self->{locked_file}); |
626
|
26
|
|
|
|
|
60
|
$self->{is_locked} = 0; |
627
|
|
|
|
|
|
|
} |
628
|
|
|
|
|
|
|
|
629
|
30
|
|
|
|
|
47
|
$self->{already_tied} = 0; |
630
|
30
|
|
|
|
|
74
|
$self->{db_version} = undef; |
631
|
|
|
|
|
|
|
} |
632
|
|
|
|
|
|
|
|
633
|
|
|
|
|
|
|
########################################################################### |
634
|
|
|
|
|
|
|
|
635
|
|
|
|
|
|
|
sub calculate_expire_delta { |
636
|
0
|
|
|
0
|
1
|
0
|
my ($self, $newest_atime, $start, $max_expire_mult) = @_; |
637
|
|
|
|
|
|
|
|
638
|
0
|
|
|
|
|
0
|
my %delta; # use a hash since an array is going to be very sparse |
639
|
|
|
|
|
|
|
|
640
|
|
|
|
|
|
|
# do the first pass, figure out atime delta |
641
|
0
|
|
|
|
|
0
|
my ($tok, $packed); |
642
|
0
|
|
|
|
|
0
|
while (($tok, $packed) = each %{$self->{db_toks}}) { |
|
0
|
|
|
|
|
0
|
|
643
|
0
|
0
|
|
|
|
0
|
next if ($tok =~ MAGIC_RE); # skip magic tokens |
644
|
|
|
|
|
|
|
|
645
|
0
|
|
|
|
|
0
|
my ($ts, $th, $atime) = $self->tok_unpack ($packed); |
646
|
|
|
|
|
|
|
|
647
|
|
|
|
|
|
|
# Go through from $start * 1 to $start * 512, mark how many tokens |
648
|
|
|
|
|
|
|
# we would expire |
649
|
0
|
|
|
|
|
0
|
my $token_age = $newest_atime - $atime; |
650
|
0
|
|
|
|
|
0
|
for (my $i = 1; $i <= $max_expire_mult; $i<<=1) { |
651
|
0
|
0
|
|
|
|
0
|
if ($token_age >= $start * $i) { |
652
|
0
|
|
|
|
|
0
|
$delta{$i}++; |
653
|
|
|
|
|
|
|
} |
654
|
|
|
|
|
|
|
else { |
655
|
|
|
|
|
|
|
# If the token age is less than the expire delta, it'll be |
656
|
|
|
|
|
|
|
# less for all upcoming checks too, so abort early. |
657
|
0
|
|
|
|
|
0
|
last; |
658
|
|
|
|
|
|
|
} |
659
|
|
|
|
|
|
|
} |
660
|
|
|
|
|
|
|
} |
661
|
0
|
|
|
|
|
0
|
return %delta; |
662
|
|
|
|
|
|
|
} |
663
|
|
|
|
|
|
|
|
664
|
|
|
|
|
|
|
########################################################################### |
665
|
|
|
|
|
|
|
|
666
|
|
|
|
|
|
|
sub token_expiration { |
667
|
0
|
|
|
0
|
1
|
0
|
my ($self, $opts, $newdelta, @vars) = @_; |
668
|
|
|
|
|
|
|
|
669
|
0
|
|
|
|
|
0
|
my $deleted = 0; |
670
|
0
|
|
|
|
|
0
|
my $kept = 0; |
671
|
0
|
|
|
|
|
0
|
my $num_hapaxes = 0; |
672
|
0
|
|
|
|
|
0
|
my $num_lowfreq = 0; |
673
|
|
|
|
|
|
|
|
674
|
|
|
|
|
|
|
# since DB_File will not shrink a database (!!), we need to *create* |
675
|
|
|
|
|
|
|
# a new one instead. |
676
|
0
|
|
|
|
|
0
|
my $main = $self->{bayes}->{main}; |
677
|
0
|
|
|
|
|
0
|
my $path = $main->sed_path($main->{conf}->{bayes_path}); |
678
|
|
|
|
|
|
|
|
679
|
|
|
|
|
|
|
# use a temporary PID-based suffix just in case another one was |
680
|
|
|
|
|
|
|
# created previously by an interrupted expire |
681
|
0
|
|
|
|
|
0
|
my $tmpsuffix = "expire$$"; |
682
|
0
|
|
|
|
|
0
|
my $tmpdbname = $path.'_toks.'.$tmpsuffix; |
683
|
|
|
|
|
|
|
|
684
|
|
|
|
|
|
|
# clean out any leftover db copies from previous runs |
685
|
0
|
|
|
|
|
0
|
for my $ext ($self->DB_EXTENSIONS) { unlink ($tmpdbname.$ext); } |
|
0
|
|
|
|
|
0
|
|
686
|
|
|
|
|
|
|
|
687
|
|
|
|
|
|
|
# use O_EXCL to avoid races (bonus paranoia, since we should be locked |
688
|
|
|
|
|
|
|
# anyway) |
689
|
0
|
|
|
|
|
0
|
my %new_toks; |
690
|
0
|
|
|
|
|
0
|
my $umask = umask 0; |
691
|
|
|
|
|
|
|
tie %new_toks, $self->DBM_MODULE, $tmpdbname, O_RDWR|O_CREAT|O_EXCL, |
692
|
0
|
|
|
|
|
0
|
(oct ($main->{conf}->{bayes_file_mode}) & 0666); |
693
|
0
|
|
|
|
|
0
|
umask $umask; |
694
|
0
|
|
|
|
|
0
|
my $oldest; |
695
|
|
|
|
|
|
|
|
696
|
0
|
|
|
|
|
0
|
my $showdots = $opts->{showdots}; |
697
|
0
|
0
|
|
|
|
0
|
if ($showdots) { print STDERR "\n"; } |
|
0
|
|
|
|
|
0
|
|
698
|
|
|
|
|
|
|
|
699
|
|
|
|
|
|
|
# We've chosen a new atime delta if we've gotten here, so record it |
700
|
|
|
|
|
|
|
# for posterity. |
701
|
0
|
|
|
|
|
0
|
$new_toks{$LAST_ATIME_DELTA_MAGIC_TOKEN} = $newdelta; |
702
|
|
|
|
|
|
|
|
703
|
|
|
|
|
|
|
# Figure out how old is too old... |
704
|
0
|
|
|
|
|
0
|
my $too_old = $vars[10] - $newdelta; # tooold = newest - delta |
705
|
|
|
|
|
|
|
|
706
|
|
|
|
|
|
|
# Go ahead and do the move to new db/expire run now ... |
707
|
0
|
|
|
|
|
0
|
my ($tok, $packed); |
708
|
0
|
|
|
|
|
0
|
while (($tok, $packed) = each %{$self->{db_toks}}) { |
|
0
|
|
|
|
|
0
|
|
709
|
0
|
0
|
|
|
|
0
|
next if ($tok =~ MAGIC_RE); # skip magic tokens |
710
|
|
|
|
|
|
|
|
711
|
0
|
|
|
|
|
0
|
my ($ts, $th, $atime) = $self->tok_unpack ($packed); |
712
|
|
|
|
|
|
|
|
713
|
0
|
0
|
|
|
|
0
|
if ($atime < $too_old) { |
714
|
0
|
|
|
|
|
0
|
$deleted++; |
715
|
|
|
|
|
|
|
} |
716
|
|
|
|
|
|
|
else { |
717
|
|
|
|
|
|
|
# if token atime > newest, reset to newest ... |
718
|
0
|
0
|
|
|
|
0
|
if ($atime > $vars[10]) { |
719
|
0
|
|
|
|
|
0
|
$atime = $vars[10]; |
720
|
|
|
|
|
|
|
} |
721
|
|
|
|
|
|
|
|
722
|
0
|
|
|
|
|
0
|
$new_toks{$tok} = $self->tok_pack ($ts, $th, $atime); $kept++; |
|
0
|
|
|
|
|
0
|
|
723
|
0
|
0
|
0
|
|
|
0
|
if (!defined($oldest) || $atime < $oldest) { $oldest = $atime; } |
|
0
|
|
|
|
|
0
|
|
724
|
0
|
0
|
0
|
|
|
0
|
if ($ts + $th == 1) { |
|
|
0
|
|
|
|
|
|
725
|
0
|
|
|
|
|
0
|
$num_hapaxes++; |
726
|
|
|
|
|
|
|
} elsif ($ts < 8 && $th < 8) { |
727
|
0
|
|
|
|
|
0
|
$num_lowfreq++; |
728
|
|
|
|
|
|
|
} |
729
|
|
|
|
|
|
|
} |
730
|
|
|
|
|
|
|
|
731
|
0
|
0
|
|
|
|
0
|
if ((($kept + $deleted) % 1000) == 0) { |
732
|
0
|
0
|
|
|
|
0
|
if ($showdots) { print STDERR "."; } |
|
0
|
|
|
|
|
0
|
|
733
|
0
|
|
|
|
|
0
|
$self->set_running_expire_tok(); |
734
|
|
|
|
|
|
|
} |
735
|
|
|
|
|
|
|
} |
736
|
|
|
|
|
|
|
|
737
|
|
|
|
|
|
|
# and add the magic tokens. don't add the expire_running token. |
738
|
0
|
|
|
|
|
0
|
$new_toks{$DB_VERSION_MAGIC_TOKEN} = $self->DB_VERSION; |
739
|
|
|
|
|
|
|
|
740
|
|
|
|
|
|
|
# We haven't changed messages of each type seen, so just copy over. |
741
|
0
|
|
|
|
|
0
|
$new_toks{$NSPAM_MAGIC_TOKEN} = $vars[1]; |
742
|
0
|
|
|
|
|
0
|
$new_toks{$NHAM_MAGIC_TOKEN} = $vars[2]; |
743
|
|
|
|
|
|
|
|
744
|
|
|
|
|
|
|
# We magically haven't removed the newest token, so just copy that value over. |
745
|
0
|
|
|
|
|
0
|
$new_toks{$NEWEST_TOKEN_AGE_MAGIC_TOKEN} = $vars[10]; |
746
|
|
|
|
|
|
|
|
747
|
|
|
|
|
|
|
# The rest of these have been modified, so replace as necessary. |
748
|
0
|
|
|
|
|
0
|
$new_toks{$NTOKENS_MAGIC_TOKEN} = $kept; |
749
|
0
|
|
|
|
|
0
|
$new_toks{$LAST_EXPIRE_MAGIC_TOKEN} = time(); |
750
|
0
|
|
|
|
|
0
|
$new_toks{$OLDEST_TOKEN_AGE_MAGIC_TOKEN} = $oldest; |
751
|
0
|
|
|
|
|
0
|
$new_toks{$LAST_EXPIRE_REDUCE_MAGIC_TOKEN} = $deleted; |
752
|
|
|
|
|
|
|
|
753
|
|
|
|
|
|
|
# Sanity check: if we expired too many tokens, abort! |
754
|
0
|
0
|
|
|
|
0
|
if ($kept < 100000) { |
755
|
0
|
|
|
|
|
0
|
dbg("bayes: token expiration would expire too many tokens, aborting"); |
756
|
|
|
|
|
|
|
# set the magic tokens appropriately |
757
|
|
|
|
|
|
|
# make sure the next expire run does a first pass |
758
|
0
|
|
|
|
|
0
|
$self->{db_toks}->{$LAST_EXPIRE_MAGIC_TOKEN} = time(); |
759
|
0
|
|
|
|
|
0
|
$self->{db_toks}->{$LAST_EXPIRE_REDUCE_MAGIC_TOKEN} = 0; |
760
|
0
|
|
|
|
|
0
|
$self->{db_toks}->{$LAST_ATIME_DELTA_MAGIC_TOKEN} = 0; |
761
|
|
|
|
|
|
|
|
762
|
|
|
|
|
|
|
# remove the new DB |
763
|
0
|
|
|
|
|
0
|
untie %new_toks; |
764
|
0
|
|
|
|
|
0
|
for my $ext ($self->DB_EXTENSIONS) { unlink ($tmpdbname.$ext); } |
|
0
|
|
|
|
|
0
|
|
765
|
|
|
|
|
|
|
|
766
|
|
|
|
|
|
|
# reset the results for the return |
767
|
0
|
|
|
|
|
0
|
$kept = $vars[3]; |
768
|
0
|
|
|
|
|
0
|
$deleted = 0; |
769
|
0
|
|
|
|
|
0
|
$num_hapaxes = 0; |
770
|
0
|
|
|
|
|
0
|
$num_lowfreq = 0; |
771
|
|
|
|
|
|
|
} |
772
|
|
|
|
|
|
|
else { |
773
|
|
|
|
|
|
|
# now untie so we can do renames |
774
|
0
|
|
|
|
|
0
|
untie %{$self->{db_toks}}; |
|
0
|
|
|
|
|
0
|
|
775
|
0
|
|
|
|
|
0
|
untie %new_toks; |
776
|
|
|
|
|
|
|
|
777
|
|
|
|
|
|
|
# This is the critical phase (moving files around), so don't allow |
778
|
|
|
|
|
|
|
# it to be interrupted. Scope the signal changes. |
779
|
|
|
|
|
|
|
{ |
780
|
0
|
|
|
|
|
0
|
local $SIG{'INT'} = 'IGNORE'; |
|
0
|
|
|
|
|
0
|
|
781
|
0
|
|
|
|
|
0
|
local $SIG{'TERM'} = 'IGNORE'; |
782
|
0
|
0
|
|
|
|
0
|
local $SIG{'HUP'} = 'IGNORE' if !am_running_on_windows(); |
783
|
|
|
|
|
|
|
|
784
|
|
|
|
|
|
|
# now rename in the new one. Try several extensions |
785
|
0
|
|
|
|
|
0
|
for my $ext ($self->DB_EXTENSIONS) { |
786
|
0
|
|
|
|
|
0
|
my $newf = $tmpdbname.$ext; |
787
|
0
|
|
|
|
|
0
|
my $oldf = $path.'_toks'.$ext; |
788
|
0
|
0
|
|
|
|
0
|
next unless (-f $newf); |
789
|
0
|
0
|
|
|
|
0
|
if (!rename ($newf, $oldf)) { |
790
|
0
|
|
|
|
|
0
|
warn "bayes: rename $newf to $oldf failed: $!\n"; |
791
|
|
|
|
|
|
|
} |
792
|
|
|
|
|
|
|
} |
793
|
|
|
|
|
|
|
} |
794
|
|
|
|
|
|
|
} |
795
|
|
|
|
|
|
|
|
796
|
|
|
|
|
|
|
# Call untie_db() so we unlock correctly. |
797
|
0
|
|
|
|
|
0
|
$self->untie_db(); |
798
|
|
|
|
|
|
|
|
799
|
0
|
|
|
|
|
0
|
return ($kept, $deleted, $num_hapaxes, $num_lowfreq); |
800
|
|
|
|
|
|
|
} |
801
|
|
|
|
|
|
|
|
802
|
|
|
|
|
|
|
########################################################################### |
803
|
|
|
|
|
|
|
|
804
|
|
|
|
|
|
|
# Is a sync due? |
805
|
|
|
|
|
|
|
sub sync_due { |
806
|
10
|
|
|
10
|
1
|
34
|
my ($self) = @_; |
807
|
|
|
|
|
|
|
|
808
|
|
|
|
|
|
|
# don't bother doing old db versions |
809
|
10
|
50
|
|
|
|
56
|
return 0 if ($self->{db_version} < $self->DB_VERSION); |
810
|
|
|
|
|
|
|
|
811
|
10
|
|
|
|
|
29
|
my $conf = $self->{bayes}->{main}->{conf}; |
812
|
10
|
50
|
|
|
|
40
|
return 0 if ($conf->{bayes_journal_max_size} == 0); |
813
|
|
|
|
|
|
|
|
814
|
10
|
|
|
|
|
52
|
my @vars = $self->get_storage_variables(); |
815
|
10
|
|
|
|
|
61
|
dbg("bayes: DB journal sync: last sync: %s", $vars[7]); |
816
|
|
|
|
|
|
|
|
817
|
|
|
|
|
|
|
## Ok, should we do a sync? |
818
|
|
|
|
|
|
|
|
819
|
|
|
|
|
|
|
# Not if the journal file doesn't exist, it's not a file, or it's 0 |
820
|
|
|
|
|
|
|
# bytes long. |
821
|
10
|
100
|
66
|
|
|
52
|
return 0 unless (stat($self->_get_journal_filename()) && -f _); |
822
|
|
|
|
|
|
|
|
823
|
|
|
|
|
|
|
# Yes if the file size is larger than the specified maximum size. |
824
|
4
|
50
|
|
|
|
27
|
return 1 if (-s _ > $conf->{bayes_journal_max_size}); |
825
|
|
|
|
|
|
|
|
826
|
|
|
|
|
|
|
# Yes there has been a sync before, and if it's been at least a day |
827
|
|
|
|
|
|
|
# since that sync. |
828
|
4
|
50
|
33
|
|
|
18
|
return 1 if (($vars[7] > 0) && (time - $vars[7] > 86400)); |
829
|
|
|
|
|
|
|
|
830
|
|
|
|
|
|
|
# No, I guess not. |
831
|
4
|
|
|
|
|
22
|
return 0; |
832
|
|
|
|
|
|
|
} |
833
|
|
|
|
|
|
|
|
834
|
|
|
|
|
|
|
########################################################################### |
835
|
|
|
|
|
|
|
# db_seen reading APIs |
836
|
|
|
|
|
|
|
|
837
|
|
|
|
|
|
|
sub seen_get { |
838
|
24
|
|
|
24
|
1
|
59
|
my ($self, $msgid) = @_; |
839
|
24
|
|
|
|
|
374
|
$self->{db_seen}->{$msgid}; |
840
|
|
|
|
|
|
|
} |
841
|
|
|
|
|
|
|
|
842
|
|
|
|
|
|
|
sub seen_put { |
843
|
6
|
|
|
6
|
1
|
35
|
my ($self, $msgid, $seen) = @_; |
844
|
|
|
|
|
|
|
|
845
|
6
|
100
|
|
|
|
27
|
if ($self->{bayes}->{main}->{learn_to_journal}) { |
846
|
2
|
|
|
|
|
17
|
$self->defer_update ("m $seen $msgid"); |
847
|
|
|
|
|
|
|
} |
848
|
|
|
|
|
|
|
else { |
849
|
4
|
|
|
|
|
25
|
$self->_seen_put_direct($msgid, $seen); |
850
|
|
|
|
|
|
|
} |
851
|
|
|
|
|
|
|
} |
852
|
|
|
|
|
|
|
sub _seen_put_direct { |
853
|
6
|
|
|
6
|
|
23
|
my ($self, $msgid, $seen) = @_; |
854
|
6
|
|
|
|
|
203
|
$self->{db_seen}->{$msgid} = $seen; |
855
|
|
|
|
|
|
|
} |
856
|
|
|
|
|
|
|
|
857
|
|
|
|
|
|
|
sub seen_delete { |
858
|
4
|
|
|
4
|
1
|
29
|
my ($self, $msgid) = @_; |
859
|
|
|
|
|
|
|
|
860
|
4
|
50
|
|
|
|
24
|
if ($self->{bayes}->{main}->{learn_to_journal}) { |
861
|
0
|
|
|
|
|
0
|
$self->defer_update ("m f $msgid"); |
862
|
|
|
|
|
|
|
} |
863
|
|
|
|
|
|
|
else { |
864
|
4
|
|
|
|
|
23
|
$self->_seen_delete_direct($msgid); |
865
|
|
|
|
|
|
|
} |
866
|
|
|
|
|
|
|
} |
867
|
|
|
|
|
|
|
sub _seen_delete_direct { |
868
|
4
|
|
|
4
|
|
24
|
my ($self, $msgid) = @_; |
869
|
4
|
|
|
|
|
101
|
delete $self->{db_seen}->{$msgid}; |
870
|
|
|
|
|
|
|
} |
871
|
|
|
|
|
|
|
|
872
|
|
|
|
|
|
|
########################################################################### |
873
|
|
|
|
|
|
|
# db reading APIs |
874
|
|
|
|
|
|
|
|
875
|
|
|
|
|
|
|
sub tok_get { |
876
|
3878
|
|
|
3878
|
1
|
5720
|
my ($self, $tok) = @_; |
877
|
3878
|
|
|
|
|
9960
|
$self->tok_unpack ($self->{db_toks}->{$tok}); |
878
|
|
|
|
|
|
|
} |
879
|
|
|
|
|
|
|
|
880
|
|
|
|
|
|
|
sub tok_get_all { |
881
|
6
|
|
|
6
|
1
|
271
|
my ($self, @tokens) = @_; |
882
|
|
|
|
|
|
|
|
883
|
6
|
|
|
|
|
26
|
my @tokensdata; |
884
|
6
|
|
|
|
|
23
|
foreach my $token (@tokens) { |
885
|
1658
|
|
|
|
|
4786
|
my ($tok_spam, $tok_ham, $atime) = $self->tok_unpack($self->{db_toks}->{$token}); |
886
|
1658
|
|
|
|
|
5816
|
push(@tokensdata, [$token, $tok_spam, $tok_ham, $atime]); |
887
|
|
|
|
|
|
|
} |
888
|
6
|
|
|
|
|
94
|
return \@tokensdata; |
889
|
|
|
|
|
|
|
} |
890
|
|
|
|
|
|
|
|
891
|
|
|
|
|
|
|
# return the magic tokens in a specific order: |
892
|
|
|
|
|
|
|
# 0: scan count base |
893
|
|
|
|
|
|
|
# 1: number of spam |
894
|
|
|
|
|
|
|
# 2: number of ham |
895
|
|
|
|
|
|
|
# 3: number of tokens in db |
896
|
|
|
|
|
|
|
# 4: last expire atime |
897
|
|
|
|
|
|
|
# 5: oldest token in db atime |
898
|
|
|
|
|
|
|
# 6: db version value |
899
|
|
|
|
|
|
|
# 7: last journal sync |
900
|
|
|
|
|
|
|
# 8: last atime delta |
901
|
|
|
|
|
|
|
# 9: last expire reduction count |
902
|
|
|
|
|
|
|
# 10: newest token in db atime |
903
|
|
|
|
|
|
|
# |
904
|
|
|
|
|
|
|
sub get_storage_variables { |
905
|
64
|
|
|
64
|
1
|
139
|
my ($self) = @_; |
906
|
64
|
|
|
|
|
100
|
my @values; |
907
|
|
|
|
|
|
|
|
908
|
64
|
|
|
|
|
1117
|
my $db_ver = $self->{db_toks}->{$DB_VERSION_MAGIC_TOKEN}; |
909
|
|
|
|
|
|
|
|
910
|
64
|
100
|
66
|
|
|
568
|
if (!$db_ver || $db_ver =~ /\D/) { $db_ver = 0; } |
|
4
|
|
|
|
|
11
|
|
911
|
|
|
|
|
|
|
|
912
|
64
|
100
|
|
|
|
204
|
if ($db_ver >= 2) { |
|
|
50
|
|
|
|
|
|
|
|
0
|
|
|
|
|
|
913
|
60
|
|
|
|
|
136
|
my $DB2_LAST_ATIME_DELTA_MAGIC_TOKEN = "\015\001\007\011\003LASTATIMEDELTA"; |
914
|
60
|
|
|
|
|
109
|
my $DB2_LAST_EXPIRE_MAGIC_TOKEN = "\015\001\007\011\003LASTEXPIRE"; |
915
|
60
|
|
|
|
|
103
|
my $DB2_LAST_EXPIRE_REDUCE_MAGIC_TOKEN = "\015\001\007\011\003LASTEXPIREREDUCE"; |
916
|
60
|
|
|
|
|
106
|
my $DB2_LAST_JOURNAL_SYNC_MAGIC_TOKEN = "\015\001\007\011\003LASTJOURNALSYNC"; |
917
|
60
|
|
|
|
|
106
|
my $DB2_NEWEST_TOKEN_AGE_MAGIC_TOKEN = "\015\001\007\011\003NEWESTAGE"; |
918
|
60
|
|
|
|
|
113
|
my $DB2_NHAM_MAGIC_TOKEN = "\015\001\007\011\003NHAM"; |
919
|
60
|
|
|
|
|
105
|
my $DB2_NSPAM_MAGIC_TOKEN = "\015\001\007\011\003NSPAM"; |
920
|
60
|
|
|
|
|
102
|
my $DB2_NTOKENS_MAGIC_TOKEN = "\015\001\007\011\003NTOKENS"; |
921
|
60
|
|
|
|
|
105
|
my $DB2_OLDEST_TOKEN_AGE_MAGIC_TOKEN = "\015\001\007\011\003OLDESTAGE"; |
922
|
60
|
|
|
|
|
99
|
my $DB2_RUNNING_EXPIRE_MAGIC_TOKEN = "\015\001\007\011\003RUNNINGEXPIRE"; |
923
|
|
|
|
|
|
|
|
924
|
|
|
|
|
|
|
@values = ( |
925
|
|
|
|
|
|
|
0, |
926
|
|
|
|
|
|
|
$self->{db_toks}->{$DB2_NSPAM_MAGIC_TOKEN}, |
927
|
|
|
|
|
|
|
$self->{db_toks}->{$DB2_NHAM_MAGIC_TOKEN}, |
928
|
|
|
|
|
|
|
$self->{db_toks}->{$DB2_NTOKENS_MAGIC_TOKEN}, |
929
|
|
|
|
|
|
|
$self->{db_toks}->{$DB2_LAST_EXPIRE_MAGIC_TOKEN}, |
930
|
|
|
|
|
|
|
$self->{db_toks}->{$DB2_OLDEST_TOKEN_AGE_MAGIC_TOKEN}, |
931
|
|
|
|
|
|
|
$db_ver, |
932
|
|
|
|
|
|
|
$self->{db_toks}->{$DB2_LAST_JOURNAL_SYNC_MAGIC_TOKEN}, |
933
|
|
|
|
|
|
|
$self->{db_toks}->{$DB2_LAST_ATIME_DELTA_MAGIC_TOKEN}, |
934
|
|
|
|
|
|
|
$self->{db_toks}->{$DB2_LAST_EXPIRE_REDUCE_MAGIC_TOKEN}, |
935
|
60
|
|
|
|
|
4351
|
$self->{db_toks}->{$DB2_NEWEST_TOKEN_AGE_MAGIC_TOKEN}, |
936
|
|
|
|
|
|
|
); |
937
|
|
|
|
|
|
|
} |
938
|
|
|
|
|
|
|
elsif ($db_ver == 0) { |
939
|
4
|
|
|
|
|
18
|
my $DB0_NSPAM_MAGIC_TOKEN = '**NSPAM'; |
940
|
4
|
|
|
|
|
15
|
my $DB0_NHAM_MAGIC_TOKEN = '**NHAM'; |
941
|
4
|
|
|
|
|
11
|
my $DB0_OLDEST_TOKEN_AGE_MAGIC_TOKEN = '**OLDESTAGE'; |
942
|
4
|
|
|
|
|
10
|
my $DB0_LAST_EXPIRE_MAGIC_TOKEN = '**LASTEXPIRE'; |
943
|
4
|
|
|
|
|
9
|
my $DB0_NTOKENS_MAGIC_TOKEN = '**NTOKENS'; |
944
|
4
|
|
|
|
|
8
|
my $DB0_SCANCOUNT_BASE_MAGIC_TOKEN = '**SCANBASE'; |
945
|
|
|
|
|
|
|
|
946
|
|
|
|
|
|
|
@values = ( |
947
|
|
|
|
|
|
|
$self->{db_toks}->{$DB0_SCANCOUNT_BASE_MAGIC_TOKEN}, |
948
|
|
|
|
|
|
|
$self->{db_toks}->{$DB0_NSPAM_MAGIC_TOKEN}, |
949
|
|
|
|
|
|
|
$self->{db_toks}->{$DB0_NHAM_MAGIC_TOKEN}, |
950
|
|
|
|
|
|
|
$self->{db_toks}->{$DB0_NTOKENS_MAGIC_TOKEN}, |
951
|
|
|
|
|
|
|
$self->{db_toks}->{$DB0_LAST_EXPIRE_MAGIC_TOKEN}, |
952
|
4
|
|
|
|
|
71
|
$self->{db_toks}->{$DB0_OLDEST_TOKEN_AGE_MAGIC_TOKEN}, |
953
|
|
|
|
|
|
|
0, |
954
|
|
|
|
|
|
|
0, |
955
|
|
|
|
|
|
|
0, |
956
|
|
|
|
|
|
|
0, |
957
|
|
|
|
|
|
|
0, |
958
|
|
|
|
|
|
|
); |
959
|
|
|
|
|
|
|
} |
960
|
|
|
|
|
|
|
elsif ($db_ver == 1) { |
961
|
0
|
|
|
|
|
0
|
my $DB1_NSPAM_MAGIC_TOKEN = "\015\001\007\011\003NSPAM"; |
962
|
0
|
|
|
|
|
0
|
my $DB1_NHAM_MAGIC_TOKEN = "\015\001\007\011\003NHAM"; |
963
|
0
|
|
|
|
|
0
|
my $DB1_OLDEST_TOKEN_AGE_MAGIC_TOKEN = "\015\001\007\011\003OLDESTAGE"; |
964
|
0
|
|
|
|
|
0
|
my $DB1_LAST_EXPIRE_MAGIC_TOKEN = "\015\001\007\011\003LASTEXPIRE"; |
965
|
0
|
|
|
|
|
0
|
my $DB1_NTOKENS_MAGIC_TOKEN = "\015\001\007\011\003NTOKENS"; |
966
|
0
|
|
|
|
|
0
|
my $DB1_SCANCOUNT_BASE_MAGIC_TOKEN = "\015\001\007\011\003SCANBASE"; |
967
|
|
|
|
|
|
|
|
968
|
|
|
|
|
|
|
@values = ( |
969
|
|
|
|
|
|
|
$self->{db_toks}->{$DB1_SCANCOUNT_BASE_MAGIC_TOKEN}, |
970
|
|
|
|
|
|
|
$self->{db_toks}->{$DB1_NSPAM_MAGIC_TOKEN}, |
971
|
|
|
|
|
|
|
$self->{db_toks}->{$DB1_NHAM_MAGIC_TOKEN}, |
972
|
|
|
|
|
|
|
$self->{db_toks}->{$DB1_NTOKENS_MAGIC_TOKEN}, |
973
|
|
|
|
|
|
|
$self->{db_toks}->{$DB1_LAST_EXPIRE_MAGIC_TOKEN}, |
974
|
0
|
|
|
|
|
0
|
$self->{db_toks}->{$DB1_OLDEST_TOKEN_AGE_MAGIC_TOKEN}, |
975
|
|
|
|
|
|
|
1, |
976
|
|
|
|
|
|
|
0, |
977
|
|
|
|
|
|
|
0, |
978
|
|
|
|
|
|
|
0, |
979
|
|
|
|
|
|
|
0, |
980
|
|
|
|
|
|
|
); |
981
|
|
|
|
|
|
|
} |
982
|
|
|
|
|
|
|
|
983
|
64
|
|
|
|
|
353
|
foreach (@values) { |
984
|
704
|
100
|
66
|
|
|
1581
|
if (!$_ || $_ =~ /\D/) { |
985
|
414
|
|
|
|
|
490
|
$_ = 0; |
986
|
|
|
|
|
|
|
} |
987
|
|
|
|
|
|
|
} |
988
|
|
|
|
|
|
|
|
989
|
64
|
|
|
|
|
361
|
return @values; |
990
|
|
|
|
|
|
|
} |
991
|
|
|
|
|
|
|
|
992
|
|
|
|
|
|
|
sub dump_db_toks { |
993
|
0
|
|
|
0
|
1
|
0
|
my ($self, $template, $regex, @vars) = @_; |
994
|
|
|
|
|
|
|
|
995
|
0
|
|
|
|
|
0
|
while (my ($tok, $tokvalue) = each %{$self->{db_toks}}) { |
|
0
|
|
|
|
|
0
|
|
996
|
0
|
0
|
|
|
|
0
|
next if ($tok =~ MAGIC_RE); # skip magic tokens |
997
|
0
|
0
|
0
|
|
|
0
|
next if (defined $regex && ($tok !~ /$regex/o)); |
998
|
|
|
|
|
|
|
|
999
|
|
|
|
|
|
|
# We have the value already, so just unpack it. |
1000
|
0
|
|
|
|
|
0
|
my ($ts, $th, $atime) = $self->tok_unpack ($tokvalue); |
1001
|
|
|
|
|
|
|
|
1002
|
0
|
|
|
|
|
0
|
my $prob = $self->{bayes}->_compute_prob_for_token($tok, $vars[1], $vars[2], $ts, $th); |
1003
|
0
|
|
0
|
|
|
0
|
$prob ||= 0.5; |
1004
|
|
|
|
|
|
|
|
1005
|
0
|
|
|
|
|
0
|
my $encoded_tok = unpack("H*",$tok); |
1006
|
0
|
|
|
|
|
0
|
printf $template,$prob,$ts,$th,$atime,$encoded_tok; |
1007
|
|
|
|
|
|
|
} |
1008
|
|
|
|
|
|
|
} |
1009
|
|
|
|
|
|
|
|
1010
|
|
|
|
|
|
|
sub set_last_expire { |
1011
|
0
|
|
|
0
|
1
|
0
|
my ($self, $time) = @_; |
1012
|
0
|
|
|
|
|
0
|
$self->{db_toks}->{$LAST_EXPIRE_MAGIC_TOKEN} = time(); |
1013
|
|
|
|
|
|
|
} |
1014
|
|
|
|
|
|
|
|
1015
|
|
|
|
|
|
|
## Don't bother using get_magic_tokens here. This token should only |
1016
|
|
|
|
|
|
|
## ever exist when we're running expire, so we don't want to convert it if |
1017
|
|
|
|
|
|
|
## it's there and we're not expiring ... |
1018
|
|
|
|
|
|
|
sub get_running_expire_tok { |
1019
|
10
|
|
|
10
|
1
|
25
|
my ($self) = @_; |
1020
|
10
|
|
|
|
|
251
|
my $running = $self->{db_toks}->{$RUNNING_EXPIRE_MAGIC_TOKEN}; |
1021
|
10
|
50
|
33
|
|
|
69
|
if (!$running || $running =~ /\D/) { return; } |
|
10
|
|
|
|
|
49
|
|
1022
|
0
|
|
|
|
|
0
|
return $running; |
1023
|
|
|
|
|
|
|
} |
1024
|
|
|
|
|
|
|
|
1025
|
|
|
|
|
|
|
sub set_running_expire_tok { |
1026
|
2
|
|
|
2
|
1
|
9
|
my ($self) = @_; |
1027
|
|
|
|
|
|
|
|
1028
|
|
|
|
|
|
|
# update the lock and running expire magic token |
1029
|
2
|
|
|
|
|
17
|
$self->{bayes}->{main}->{locker}->refresh_lock ($self->{locked_file}); |
1030
|
2
|
|
|
|
|
51
|
$self->{db_toks}->{$RUNNING_EXPIRE_MAGIC_TOKEN} = time(); |
1031
|
|
|
|
|
|
|
} |
1032
|
|
|
|
|
|
|
|
1033
|
|
|
|
|
|
|
sub remove_running_expire_tok { |
1034
|
0
|
|
|
0
|
1
|
0
|
my ($self) = @_; |
1035
|
0
|
|
|
|
|
0
|
delete $self->{db_toks}->{$RUNNING_EXPIRE_MAGIC_TOKEN}; |
1036
|
|
|
|
|
|
|
} |
1037
|
|
|
|
|
|
|
|
1038
|
|
|
|
|
|
|
########################################################################### |
1039
|
|
|
|
|
|
|
|
1040
|
|
|
|
|
|
|
# db abstraction: allow deferred writes, since we will be frequently |
1041
|
|
|
|
|
|
|
# writing while checking. |
1042
|
|
|
|
|
|
|
|
1043
|
|
|
|
|
|
|
sub tok_count_change { |
1044
|
0
|
|
|
0
|
1
|
0
|
my ($self, $ds, $dh, $tok, $atime) = @_; |
1045
|
|
|
|
|
|
|
|
1046
|
0
|
0
|
|
|
|
0
|
$atime = 0 unless defined $atime; |
1047
|
|
|
|
|
|
|
|
1048
|
0
|
0
|
|
|
|
0
|
if ($self->{bayes}->{main}->{learn_to_journal}) { |
1049
|
|
|
|
|
|
|
# we can't store the SHA1 binary value in the journal, so convert it |
1050
|
|
|
|
|
|
|
# to a printable value that can be converted back later |
1051
|
0
|
|
|
|
|
0
|
my $encoded_tok = unpack("H*",$tok); |
1052
|
0
|
|
|
|
|
0
|
$self->defer_update ("c $ds $dh $atime $encoded_tok"); |
1053
|
|
|
|
|
|
|
} else { |
1054
|
0
|
|
|
|
|
0
|
$self->tok_sync_counters ($ds, $dh, $atime, $tok); |
1055
|
|
|
|
|
|
|
} |
1056
|
|
|
|
|
|
|
} |
1057
|
|
|
|
|
|
|
|
1058
|
|
|
|
|
|
|
sub multi_tok_count_change { |
1059
|
10
|
|
|
10
|
1
|
35
|
my ($self, $ds, $dh, $tokens, $atime) = @_; |
1060
|
|
|
|
|
|
|
|
1061
|
10
|
100
|
|
|
|
32
|
$atime = 0 unless defined $atime; |
1062
|
|
|
|
|
|
|
|
1063
|
10
|
|
|
|
|
24
|
foreach my $tok (keys %{$tokens}) { |
|
10
|
|
|
|
|
248
|
|
1064
|
2770
|
100
|
|
|
|
7134
|
if ($self->{bayes}->{main}->{learn_to_journal}) { |
1065
|
|
|
|
|
|
|
# we can't store the SHA1 binary value in the journal, so convert it |
1066
|
|
|
|
|
|
|
# to a printable value that can be converted back later |
1067
|
554
|
|
|
|
|
804
|
my $encoded_tok = unpack("H*",$tok); |
1068
|
554
|
|
|
|
|
1084
|
$self->defer_update ("c $ds $dh $atime $encoded_tok"); |
1069
|
|
|
|
|
|
|
} else { |
1070
|
2216
|
|
|
|
|
4394
|
$self->tok_sync_counters ($ds, $dh, $atime, $tok); |
1071
|
|
|
|
|
|
|
} |
1072
|
|
|
|
|
|
|
} |
1073
|
|
|
|
|
|
|
} |
1074
|
|
|
|
|
|
|
|
1075
|
|
|
|
|
|
|
sub nspam_nham_get { |
1076
|
10
|
|
|
10
|
1
|
25
|
my ($self) = @_; |
1077
|
10
|
|
|
|
|
36
|
my @vars = $self->get_storage_variables(); |
1078
|
10
|
|
|
|
|
52
|
($vars[1], $vars[2]); |
1079
|
|
|
|
|
|
|
} |
1080
|
|
|
|
|
|
|
|
1081
|
|
|
|
|
|
|
sub nspam_nham_change { |
1082
|
10
|
|
|
10
|
1
|
35
|
my ($self, $ds, $dh) = @_; |
1083
|
|
|
|
|
|
|
|
1084
|
10
|
100
|
|
|
|
43
|
if ($self->{bayes}->{main}->{learn_to_journal}) { |
1085
|
2
|
|
|
|
|
75
|
$self->defer_update ("n $ds $dh"); |
1086
|
|
|
|
|
|
|
} else { |
1087
|
8
|
|
|
|
|
64
|
$self->tok_sync_nspam_nham ($ds, $dh); |
1088
|
|
|
|
|
|
|
} |
1089
|
|
|
|
|
|
|
} |
1090
|
|
|
|
|
|
|
|
1091
|
|
|
|
|
|
|
sub tok_touch { |
1092
|
0
|
|
|
0
|
1
|
0
|
my ($self, $tok, $atime) = @_; |
1093
|
|
|
|
|
|
|
# we can't store the SHA1 binary value in the journal, so convert it |
1094
|
|
|
|
|
|
|
# to a printable value that can be converted back later |
1095
|
0
|
|
|
|
|
0
|
my $encoded_tok = unpack("H*", $tok); |
1096
|
0
|
|
|
|
|
0
|
$self->defer_update ("t $atime $encoded_tok"); |
1097
|
|
|
|
|
|
|
} |
1098
|
|
|
|
|
|
|
|
1099
|
|
|
|
|
|
|
sub tok_touch_all { |
1100
|
4
|
|
|
4
|
1
|
21
|
my ($self, $tokens, $atime) = @_; |
1101
|
|
|
|
|
|
|
|
1102
|
4
|
|
|
|
|
11
|
foreach my $token (@{$tokens}) { |
|
4
|
|
|
|
|
22
|
|
1103
|
|
|
|
|
|
|
# we can't store the SHA1 binary value in the journal, so convert it |
1104
|
|
|
|
|
|
|
# to a printable value that can be converted back later |
1105
|
272
|
|
|
|
|
463
|
my $encoded_tok = unpack("H*", $token); |
1106
|
272
|
|
|
|
|
506
|
$self->defer_update ("t $atime $encoded_tok"); |
1107
|
|
|
|
|
|
|
} |
1108
|
|
|
|
|
|
|
} |
1109
|
|
|
|
|
|
|
|
1110
|
|
|
|
|
|
|
sub defer_update { |
1111
|
830
|
|
|
830
|
0
|
1054
|
my ($self, $str) = @_; |
1112
|
830
|
|
|
|
|
1393
|
$self->{string_to_journal} .= "$str\n"; |
1113
|
|
|
|
|
|
|
} |
1114
|
|
|
|
|
|
|
|
1115
|
|
|
|
|
|
|
########################################################################### |
1116
|
|
|
|
|
|
|
|
1117
|
|
|
|
|
|
|
sub cleanup { |
1118
|
14
|
|
|
14
|
1
|
49
|
my ($self) = @_; |
1119
|
|
|
|
|
|
|
|
1120
|
14
|
|
|
|
|
58
|
my $nbytes = length ($self->{string_to_journal}); |
1121
|
14
|
100
|
|
|
|
68
|
return if ($nbytes == 0); |
1122
|
|
|
|
|
|
|
|
1123
|
6
|
|
|
|
|
54
|
my $path = $self->_get_journal_filename(); |
1124
|
|
|
|
|
|
|
|
1125
|
|
|
|
|
|
|
# use append mode, write atomically, then close, so simultaneous updates are |
1126
|
|
|
|
|
|
|
# not lost |
1127
|
6
|
|
|
|
|
27
|
my $conf = $self->{bayes}->{main}->{conf}; |
1128
|
|
|
|
|
|
|
|
1129
|
|
|
|
|
|
|
# set the umask to the inverse of what we want ... |
1130
|
6
|
|
|
|
|
58
|
my $umask = umask(0777 - (oct ($conf->{bayes_file_mode}) & 0666)); |
1131
|
|
|
|
|
|
|
|
1132
|
6
|
50
|
|
|
|
514
|
if (!open (OUT, ">>".$path)) { |
1133
|
0
|
|
|
|
|
0
|
warn "bayes: cannot write to $path, bayes db update ignored: $!\n"; |
1134
|
0
|
|
|
|
|
0
|
umask $umask; # reset umask |
1135
|
0
|
|
|
|
|
0
|
return; |
1136
|
|
|
|
|
|
|
} |
1137
|
6
|
|
|
|
|
42
|
umask $umask; # reset umask |
1138
|
|
|
|
|
|
|
|
1139
|
|
|
|
|
|
|
# do not use print() here, it will break up the buffer if it's >8192 bytes, |
1140
|
|
|
|
|
|
|
# which could result in two sets of tokens getting mixed up and their |
1141
|
|
|
|
|
|
|
# touches missed. |
1142
|
6
|
|
|
|
|
22
|
my $write_failure = 0; |
1143
|
6
|
|
|
|
|
24
|
my $original_point = tell OUT; |
1144
|
6
|
50
|
|
|
|
26
|
$original_point >= 0 or die "Can't obtain file position: $!"; |
1145
|
6
|
|
|
|
|
11
|
my $len; |
1146
|
6
|
|
|
|
|
17
|
do { |
1147
|
6
|
|
|
|
|
259
|
$len = syswrite (OUT, $self->{string_to_journal}, $nbytes); |
1148
|
|
|
|
|
|
|
|
1149
|
|
|
|
|
|
|
# argh, write failure, give up |
1150
|
6
|
50
|
33
|
|
|
79
|
if (!defined $len || $len < 0) { |
1151
|
0
|
|
|
|
|
0
|
my $err = ''; |
1152
|
0
|
0
|
|
|
|
0
|
if (!defined $len) { |
1153
|
0
|
|
|
|
|
0
|
$len = 0; |
1154
|
0
|
|
|
|
|
0
|
$err = " ($!)"; |
1155
|
|
|
|
|
|
|
} |
1156
|
0
|
|
|
|
|
0
|
warn "bayes: write failed to Bayes journal $path ($len of $nbytes)!$err\n"; |
1157
|
0
|
|
|
|
|
0
|
last; |
1158
|
|
|
|
|
|
|
} |
1159
|
|
|
|
|
|
|
|
1160
|
|
|
|
|
|
|
# This shouldn't happen, but could if the fs is full... |
1161
|
6
|
50
|
|
|
|
42
|
if ($len != $nbytes) { |
1162
|
0
|
|
|
|
|
0
|
warn "bayes: partial write to bayes journal $path ($len of $nbytes), recovering\n"; |
1163
|
|
|
|
|
|
|
|
1164
|
|
|
|
|
|
|
# we want to be atomic, so revert the journal file back to where |
1165
|
|
|
|
|
|
|
# we know it's "good". if we can't truncate the journal, or we've |
1166
|
|
|
|
|
|
|
# tried 5 times to do the write, abort! |
1167
|
0
|
0
|
0
|
|
|
0
|
if (!truncate(OUT, $original_point) || ($write_failure++ > 4)) { |
1168
|
0
|
|
|
|
|
0
|
warn "bayes: cannot write to bayes journal $path, aborting!\n"; |
1169
|
0
|
|
|
|
|
0
|
last; |
1170
|
|
|
|
|
|
|
} |
1171
|
|
|
|
|
|
|
|
1172
|
|
|
|
|
|
|
# if the fs is full, let's give the system a break |
1173
|
0
|
|
|
|
|
0
|
sleep 1; |
1174
|
|
|
|
|
|
|
} |
1175
|
|
|
|
|
|
|
} while ($len != $nbytes); |
1176
|
|
|
|
|
|
|
|
1177
|
6
|
50
|
|
|
|
93
|
if (!close OUT) { |
1178
|
0
|
|
|
|
|
0
|
warn "bayes: cannot write to $path, bayes db update ignored\n"; |
1179
|
|
|
|
|
|
|
} |
1180
|
|
|
|
|
|
|
|
1181
|
6
|
|
|
|
|
36
|
$self->{string_to_journal} = ''; |
1182
|
|
|
|
|
|
|
} |
1183
|
|
|
|
|
|
|
|
1184
|
|
|
|
|
|
|
# Return a qr'd RE to match a token with the correct format's magic token |
1185
|
|
|
|
|
|
|
sub get_magic_re { |
1186
|
538
|
|
|
538
|
1
|
784
|
my ($self) = @_; |
1187
|
|
|
|
|
|
|
|
1188
|
538
|
50
|
66
|
|
|
1766
|
if (!defined $self->{db_version} || $self->{db_version} >= 1) { |
1189
|
538
|
|
|
|
|
1124
|
return MAGIC_RE; |
1190
|
|
|
|
|
|
|
} |
1191
|
|
|
|
|
|
|
|
1192
|
|
|
|
|
|
|
# When in doubt, assume v0 |
1193
|
0
|
|
|
|
|
0
|
return qr/^\*\*[A-Z]+$/; |
1194
|
|
|
|
|
|
|
} |
1195
|
|
|
|
|
|
|
|
1196
|
|
|
|
|
|
|
# provide a more generalized public interface into the journal sync |
1197
|
|
|
|
|
|
|
|
1198
|
|
|
|
|
|
|
sub sync { |
1199
|
2
|
|
|
2
|
1
|
10
|
my ($self, $opts) = @_; |
1200
|
|
|
|
|
|
|
|
1201
|
2
|
|
|
|
|
40
|
return $self->_sync_journal($opts); |
1202
|
|
|
|
|
|
|
} |
1203
|
|
|
|
|
|
|
|
1204
|
|
|
|
|
|
|
########################################################################### |
1205
|
|
|
|
|
|
|
# And this method reads the journal and applies the changes in one |
1206
|
|
|
|
|
|
|
# (locked) transaction. |
1207
|
|
|
|
|
|
|
|
1208
|
|
|
|
|
|
|
sub _sync_journal { |
1209
|
2
|
|
|
2
|
|
9
|
my ($self, $opts) = @_; |
1210
|
2
|
|
|
|
|
4
|
my $ret = 0; |
1211
|
|
|
|
|
|
|
|
1212
|
2
|
|
|
|
|
15
|
my $path = $self->_get_journal_filename(); |
1213
|
|
|
|
|
|
|
|
1214
|
|
|
|
|
|
|
# if $path doesn't exist, or it's not a file, or is 0 bytes in length, return |
1215
|
2
|
50
|
33
|
|
|
69
|
if (!stat($path) || !-f _ || -z _) { |
|
|
|
33
|
|
|
|
|
1216
|
0
|
|
|
|
|
0
|
return 0; |
1217
|
|
|
|
|
|
|
} |
1218
|
|
|
|
|
|
|
|
1219
|
2
|
|
|
|
|
14
|
my $eval_stat; |
1220
|
|
|
|
|
|
|
eval { |
1221
|
2
|
|
|
|
|
14
|
local $SIG{'__DIE__'}; # do not run user die() traps in here |
1222
|
2
|
50
|
|
|
|
20
|
if ($self->tie_db_writable()) { |
1223
|
2
|
|
|
|
|
22
|
$ret = $self->_sync_journal_trapped($opts, $path); |
1224
|
|
|
|
|
|
|
} |
1225
|
2
|
|
|
|
|
17
|
1; |
1226
|
2
|
50
|
|
|
|
6
|
} or do { |
1227
|
0
|
0
|
|
|
|
0
|
$eval_stat = $@ ne '' ? $@ : "errno=$!"; chomp $eval_stat; |
|
0
|
|
|
|
|
0
|
|
1228
|
|
|
|
|
|
|
}; |
1229
|
|
|
|
|
|
|
|
1230
|
|
|
|
|
|
|
# ok, untie from write-mode if we can |
1231
|
2
|
50
|
|
|
|
17
|
if (!$self->{bayes}->{main}->{learn_caller_will_untie}) { |
1232
|
2
|
|
|
|
|
12
|
$self->untie_db(); |
1233
|
|
|
|
|
|
|
} |
1234
|
|
|
|
|
|
|
|
1235
|
|
|
|
|
|
|
# handle any errors that may have occurred |
1236
|
2
|
50
|
|
|
|
12
|
if (defined $eval_stat) { |
1237
|
0
|
|
|
|
|
0
|
warn "bayes: $eval_stat\n"; |
1238
|
0
|
|
|
|
|
0
|
return 0; |
1239
|
|
|
|
|
|
|
} |
1240
|
|
|
|
|
|
|
|
1241
|
2
|
|
|
|
|
14
|
$ret; |
1242
|
|
|
|
|
|
|
} |
1243
|
|
|
|
|
|
|
|
1244
|
|
|
|
|
|
|
sub _sync_journal_trapped { |
1245
|
2
|
|
|
2
|
|
12
|
my ($self, $opts, $path) = @_; |
1246
|
|
|
|
|
|
|
|
1247
|
|
|
|
|
|
|
# Flag that we're doing work |
1248
|
2
|
|
|
|
|
25
|
$self->set_running_expire_tok(); |
1249
|
|
|
|
|
|
|
|
1250
|
2
|
|
|
|
|
12
|
my $started = time(); |
1251
|
2
|
|
|
|
|
7
|
my $count = 0; |
1252
|
2
|
|
|
|
|
4
|
my $total_count = 0; |
1253
|
2
|
|
|
|
|
9
|
my %tokens; |
1254
|
2
|
|
|
|
|
7
|
my $showdots = $opts->{showdots}; |
1255
|
2
|
|
|
|
|
7
|
my $retirepath = $path.".old"; |
1256
|
|
|
|
|
|
|
|
1257
|
|
|
|
|
|
|
# if $path doesn't exist, or it's not a file, or is 0 bytes in length, |
1258
|
|
|
|
|
|
|
# return we have to check again since the file may have been removed |
1259
|
|
|
|
|
|
|
# by a recent bayes db upgrade ... |
1260
|
2
|
50
|
33
|
|
|
65
|
if (!stat($path) || !-f _ || -z _) { |
|
|
|
33
|
|
|
|
|
1261
|
0
|
|
|
|
|
0
|
return 0; |
1262
|
|
|
|
|
|
|
} |
1263
|
|
|
|
|
|
|
|
1264
|
2
|
50
|
|
|
|
30
|
if (!-r $path) { # will we be able to read the file? |
1265
|
0
|
|
|
|
|
0
|
warn "bayes: bad permissions on journal, can't read: $path\n"; |
1266
|
0
|
|
|
|
|
0
|
return 0; |
1267
|
|
|
|
|
|
|
} |
1268
|
|
|
|
|
|
|
|
1269
|
|
|
|
|
|
|
# This is the critical phase (moving files around), so don't allow |
1270
|
|
|
|
|
|
|
# it to be interrupted. |
1271
|
|
|
|
|
|
|
{ |
1272
|
2
|
|
|
|
|
5
|
local $SIG{'INT'} = 'IGNORE'; |
|
2
|
|
|
|
|
62
|
|
1273
|
2
|
|
|
|
|
45
|
local $SIG{'TERM'} = 'IGNORE'; |
1274
|
2
|
50
|
|
|
|
14
|
local $SIG{'HUP'} = 'IGNORE' if !am_running_on_windows(); |
1275
|
|
|
|
|
|
|
|
1276
|
|
|
|
|
|
|
# retire the journal, so we can update the db files from it in peace. |
1277
|
|
|
|
|
|
|
# TODO: use locking here |
1278
|
2
|
50
|
|
|
|
107
|
if (!rename ($path, $retirepath)) { |
1279
|
0
|
|
|
|
|
0
|
warn "bayes: failed rename $path to $retirepath\n"; |
1280
|
0
|
|
|
|
|
0
|
return 0; |
1281
|
|
|
|
|
|
|
} |
1282
|
|
|
|
|
|
|
|
1283
|
|
|
|
|
|
|
# now read the retired journal |
1284
|
2
|
|
|
|
|
13
|
local *JOURNAL; |
1285
|
2
|
50
|
|
|
|
75
|
if (!open (JOURNAL, "<$retirepath")) { |
1286
|
0
|
|
|
|
|
0
|
warn "bayes: cannot open read $retirepath\n"; |
1287
|
0
|
|
|
|
|
0
|
return 0; |
1288
|
|
|
|
|
|
|
} |
1289
|
|
|
|
|
|
|
|
1290
|
|
|
|
|
|
|
|
1291
|
|
|
|
|
|
|
# Read the journal |
1292
|
2
|
|
|
|
|
129
|
for ($!=0; defined($_=<JOURNAL>); $!=0) { |
1293
|
558
|
|
|
|
|
829
|
$total_count++; |
1294
|
|
|
|
|
|
|
|
1295
|
558
|
50
|
|
|
|
2835
|
if (/^t (\d+) (.+)$/) { # Token timestamp update, cache resultant entries |
|
|
100
|
|
|
|
|
|
|
|
100
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
1296
|
0
|
|
|
|
|
0
|
my $tok = pack("H*",$2); |
1297
|
0
|
0
|
0
|
|
|
0
|
$tokens{$tok} = $1+0 if (!exists $tokens{$tok} || $1+0 > $tokens{$tok}); |
1298
|
|
|
|
|
|
|
} elsif (/^c (-?\d+) (-?\d+) (\d+) (.+)$/) { # Add/full token update |
1299
|
554
|
|
|
|
|
1747
|
my $tok = pack("H*",$4); |
1300
|
554
|
|
|
|
|
2144
|
$self->tok_sync_counters ($1+0, $2+0, $3+0, $tok); |
1301
|
554
|
|
|
|
|
2939
|
$count++; |
1302
|
|
|
|
|
|
|
} elsif (/^n (-?\d+) (-?\d+)$/) { # update ham/spam count |
1303
|
2
|
|
|
|
|
20
|
$self->tok_sync_nspam_nham ($1+0, $2+0); |
1304
|
2
|
|
|
|
|
32
|
$count++; |
1305
|
|
|
|
|
|
|
} elsif (/^m ([hsf]) (.+)$/) { # update msgid seen database |
1306
|
2
|
50
|
|
|
|
21
|
if ($1 eq "f") { |
1307
|
0
|
|
|
|
|
0
|
$self->_seen_delete_direct($2); |
1308
|
|
|
|
|
|
|
} |
1309
|
|
|
|
|
|
|
else { |
1310
|
2
|
|
|
|
|
13
|
$self->_seen_put_direct($2,$1); |
1311
|
|
|
|
|
|
|
} |
1312
|
2
|
|
|
|
|
31
|
$count++; |
1313
|
|
|
|
|
|
|
} else { |
1314
|
0
|
|
|
|
|
0
|
warn "bayes: gibberish entry found in journal: $_"; |
1315
|
|
|
|
|
|
|
} |
1316
|
|
|
|
|
|
|
} |
1317
|
2
|
0
|
33
|
|
|
34
|
defined $_ || $!==0 or |
|
|
50
|
|
|
|
|
|
1318
|
|
|
|
|
|
|
$!==EBADF ? dbg("bayes: error reading journal file: $!") |
1319
|
|
|
|
|
|
|
: die "error reading journal file: $!"; |
1320
|
2
|
50
|
|
|
|
45
|
close(JOURNAL) or die "Can't close journal file: $!"; |
1321
|
|
|
|
|
|
|
|
1322
|
|
|
|
|
|
|
# Now that we've determined what tokens we need to update and their |
1323
|
|
|
|
|
|
|
# final values, update the DB. Should be much smaller than the full |
1324
|
|
|
|
|
|
|
# journal entries. |
1325
|
2
|
|
|
|
|
25
|
while (my ($k,$v) = each %tokens) { |
1326
|
0
|
|
|
|
|
0
|
$self->tok_touch_token ($v, $k); |
1327
|
|
|
|
|
|
|
|
1328
|
0
|
0
|
|
|
|
0
|
if ((++$count % 1000) == 0) { |
1329
|
0
|
0
|
|
|
|
0
|
if ($showdots) { print STDERR "."; } |
|
0
|
|
|
|
|
0
|
|
1330
|
0
|
|
|
|
|
0
|
$self->set_running_expire_tok(); |
1331
|
|
|
|
|
|
|
} |
1332
|
|
|
|
|
|
|
} |
1333
|
|
|
|
|
|
|
|
1334
|
2
|
50
|
|
|
|
11
|
if ($showdots) { print STDERR "\n"; } |
|
0
|
|
|
|
|
0
|
|
1335
|
|
|
|
|
|
|
|
1336
|
|
|
|
|
|
|
# we're all done, so unlink the old journal file |
1337
|
2
|
50
|
|
|
|
186
|
unlink ($retirepath) || warn "bayes: can't unlink $retirepath: $!\n"; |
1338
|
|
|
|
|
|
|
|
1339
|
2
|
|
|
|
|
56
|
$self->{db_toks}->{$LAST_JOURNAL_SYNC_MAGIC_TOKEN} = $started; |
1340
|
|
|
|
|
|
|
|
1341
|
2
|
|
|
|
|
47
|
my $done = time(); |
1342
|
2
|
|
|
|
|
19
|
my $msg = ("bayes: synced databases from journal in " . |
1343
|
|
|
|
|
|
|
($done - $started) . |
1344
|
|
|
|
|
|
|
" seconds: $count unique entries ($total_count total entries)"); |
1345
|
|
|
|
|
|
|
|
1346
|
2
|
50
|
|
|
|
12
|
if ($opts->{verbose}) { |
1347
|
0
|
|
|
|
|
0
|
print $msg,"\n"; |
1348
|
|
|
|
|
|
|
} else { |
1349
|
2
|
|
|
|
|
31
|
dbg($msg); |
1350
|
|
|
|
|
|
|
} |
1351
|
|
|
|
|
|
|
} |
1352
|
|
|
|
|
|
|
|
1353
|
|
|
|
|
|
|
# else, that's the lot, we're synced. return |
1354
|
2
|
|
|
|
|
13
|
return 1; |
1355
|
|
|
|
|
|
|
} |
1356
|
|
|
|
|
|
|
|
1357
|
|
|
|
|
|
|
sub tok_touch_token { |
1358
|
0
|
|
|
0
|
0
|
0
|
my ($self, $atime, $tok) = @_; |
1359
|
0
|
|
|
|
|
0
|
my ($ts, $th, $oldatime) = $self->tok_get ($tok); |
1360
|
|
|
|
|
|
|
|
1361
|
|
|
|
|
|
|
# If the new atime is < the old atime, ignore the update |
1362
|
|
|
|
|
|
|
# We figure that we'll never want to lower a token atime, so abort if |
1363
|
|
|
|
|
|
|
# we try. (journal out of sync, etc.) |
1364
|
0
|
0
|
|
|
|
0
|
return if ($oldatime >= $atime); |
1365
|
|
|
|
|
|
|
|
1366
|
0
|
|
|
|
|
0
|
$self->tok_put ($tok, $ts, $th, $atime); |
1367
|
|
|
|
|
|
|
} |
1368
|
|
|
|
|
|
|
|
1369
|
|
|
|
|
|
|
sub tok_sync_counters { |
1370
|
2770
|
|
|
2770
|
0
|
4835
|
my ($self, $ds, $dh, $atime, $tok) = @_; |
1371
|
2770
|
|
|
|
|
4654
|
my ($ts, $th, $oldatime) = $self->tok_get ($tok); |
1372
|
2770
|
50
|
|
|
|
4950
|
$ts += $ds; if ($ts < 0) { $ts = 0; } |
|
2770
|
|
|
|
|
4453
|
|
|
0
|
|
|
|
|
0
|
|
1373
|
2770
|
50
|
|
|
|
2855
|
$th += $dh; if ($th < 0) { $th = 0; } |
|
2770
|
|
|
|
|
3958
|
|
|
0
|
|
|
|
|
0
|
|
1374
|
|
|
|
|
|
|
|
1375
|
|
|
|
|
|
|
# Don't roll the atime of tokens backwards ... |
1376
|
2770
|
100
|
|
|
|
4550
|
$atime = $oldatime if ($oldatime > $atime); |
1377
|
|
|
|
|
|
|
|
1378
|
2770
|
|
|
|
|
4663
|
$self->tok_put ($tok, $ts, $th, $atime); |
1379
|
|
|
|
|
|
|
} |
1380
|
|
|
|
|
|
|
|
1381
|
|
|
|
|
|
|
sub tok_put { |
1382
|
2770
|
|
|
2770
|
0
|
4287
|
my ($self, $tok, $ts, $th, $atime) = @_; |
1383
|
2770
|
|
100
|
|
|
6742
|
$ts ||= 0; |
1384
|
2770
|
|
100
|
|
|
6740
|
$th ||= 0; |
1385
|
|
|
|
|
|
|
|
1386
|
|
|
|
|
|
|
# Ignore magic tokens, the don't go in this way ... |
1387
|
2770
|
50
|
|
|
|
5028
|
return if ($tok =~ MAGIC_RE); |
1388
|
|
|
|
|
|
|
|
1389
|
|
|
|
|
|
|
# use defined() rather than exists(); the latter is not supported |
1390
|
|
|
|
|
|
|
# by NDBM_File, believe it or not. Using defined() did not |
1391
|
|
|
|
|
|
|
# indicate any noticeable speed hit in my testing. (Mar 31 2003 jm) |
1392
|
2770
|
|
|
|
|
11072
|
my $exists_already = defined $self->{db_toks}->{$tok}; |
1393
|
|
|
|
|
|
|
|
1394
|
2770
|
100
|
100
|
|
|
8161
|
if ($ts == 0 && $th == 0) { |
1395
|
1108
|
50
|
|
|
|
1819
|
return if (!$exists_already); # If the token doesn't exist, just return |
1396
|
1108
|
|
|
|
|
21302
|
$self->{db_toks}->{$NTOKENS_MAGIC_TOKEN}--; |
1397
|
1108
|
|
|
|
|
20315
|
delete $self->{db_toks}->{$tok}; |
1398
|
|
|
|
|
|
|
} else { |
1399
|
1662
|
50
|
|
|
|
2602
|
if (!$exists_already) { # If the token doesn't exist, raise the token count |
1400
|
1662
|
|
|
|
|
29911
|
$self->{db_toks}->{$NTOKENS_MAGIC_TOKEN}++; |
1401
|
|
|
|
|
|
|
} |
1402
|
|
|
|
|
|
|
|
1403
|
1662
|
|
|
|
|
6304
|
$self->{db_toks}->{$tok} = $self->tok_pack ($ts, $th, $atime); |
1404
|
|
|
|
|
|
|
|
1405
|
1662
|
|
|
|
|
16457
|
my $newmagic = $self->{db_toks}->{$NEWEST_TOKEN_AGE_MAGIC_TOKEN}; |
1406
|
1662
|
100
|
66
|
|
|
7780
|
if (!defined ($newmagic) || $atime > $newmagic) { |
1407
|
4
|
|
|
|
|
56
|
$self->{db_toks}->{$NEWEST_TOKEN_AGE_MAGIC_TOKEN} = $atime; |
1408
|
|
|
|
|
|
|
} |
1409
|
|
|
|
|
|
|
|
1410
|
|
|
|
|
|
|
# Make sure to check for either !defined or "" ... Apparently |
1411
|
|
|
|
|
|
|
# sometimes the DB module doesn't return the value correctly. :( |
1412
|
1662
|
|
|
|
|
15046
|
my $oldmagic = $self->{db_toks}->{$OLDEST_TOKEN_AGE_MAGIC_TOKEN}; |
1413
|
1662
|
100
|
66
|
|
|
10898
|
if (!defined ($oldmagic) || $oldmagic eq "" || $atime < $oldmagic) { |
|
|
|
66
|
|
|
|
|
1414
|
4
|
|
|
|
|
57
|
$self->{db_toks}->{$OLDEST_TOKEN_AGE_MAGIC_TOKEN} = $atime; |
1415
|
|
|
|
|
|
|
} |
1416
|
|
|
|
|
|
|
} |
1417
|
|
|
|
|
|
|
} |
1418
|
|
|
|
|
|
|
|
1419
|
|
|
|
|
|
|
sub tok_sync_nspam_nham { |
1420
|
10
|
|
|
10
|
0
|
29
|
my ($self, $ds, $dh) = @_; |
1421
|
10
|
|
|
|
|
41
|
my ($ns, $nh) = ($self->get_storage_variables())[1,2]; |
1422
|
10
|
100
|
|
|
|
41
|
if ($ds) { $ns += $ds; } if ($ns < 0) { $ns = 0; } |
|
6
|
50
|
|
|
|
14
|
|
|
10
|
|
|
|
|
33
|
|
|
0
|
|
|
|
|
0
|
|
1423
|
10
|
100
|
|
|
|
31
|
if ($dh) { $nh += $dh; } if ($nh < 0) { $nh = 0; } |
|
4
|
50
|
|
|
|
14
|
|
|
10
|
|
|
|
|
31
|
|
|
0
|
|
|
|
|
0
|
|
1424
|
10
|
|
|
|
|
280
|
$self->{db_toks}->{$NSPAM_MAGIC_TOKEN} = $ns; |
1425
|
10
|
|
|
|
|
205
|
$self->{db_toks}->{$NHAM_MAGIC_TOKEN} = $nh; |
1426
|
|
|
|
|
|
|
} |
1427
|
|
|
|
|
|
|
|
1428
|
|
|
|
|
|
|
########################################################################### |
1429
|
|
|
|
|
|
|
|
1430
|
|
|
|
|
|
|
sub _get_journal_filename { |
1431
|
18
|
|
|
18
|
|
48
|
my ($self) = @_; |
1432
|
|
|
|
|
|
|
|
1433
|
18
|
|
|
|
|
45
|
my $main = $self->{bayes}->{main}; |
1434
|
18
|
|
|
|
|
134
|
return $main->sed_path($main->{conf}->{bayes_path}."_journal"); |
1435
|
|
|
|
|
|
|
} |
1436
|
|
|
|
|
|
|
|
1437
|
|
|
|
|
|
|
########################################################################### |
1438
|
|
|
|
|
|
|
|
1439
|
|
|
|
|
|
|
# this is called directly from sa-learn(1). |
1440
|
|
|
|
|
|
|
sub perform_upgrade { |
1441
|
0
|
|
|
0
|
1
|
0
|
my ($self, $opts) = @_; |
1442
|
0
|
|
|
|
|
0
|
my $ret = 0; |
1443
|
|
|
|
|
|
|
|
1444
|
0
|
|
|
|
|
0
|
my $eval_stat; |
1445
|
|
|
|
|
|
|
eval { |
1446
|
0
|
|
|
|
|
0
|
local $SIG{'__DIE__'}; # do not run user die() traps in here |
1447
|
|
|
|
|
|
|
|
1448
|
21
|
|
|
21
|
|
223
|
use File::Basename; |
|
21
|
|
|
|
|
87
|
|
|
21
|
|
|
|
|
50867
|
|
1449
|
|
|
|
|
|
|
|
1450
|
|
|
|
|
|
|
# bayes directory |
1451
|
0
|
|
|
|
|
0
|
my $main = $self->{bayes}->{main}; |
1452
|
0
|
|
|
|
|
0
|
my $path = $main->sed_path($main->{conf}->{bayes_path}); |
1453
|
|
|
|
|
|
|
|
1454
|
|
|
|
|
|
|
# prevent dirname() from tainting the result, it assumes $1 is not tainted |
1455
|
0
|
|
|
|
|
0
|
local($1,$2,$3); # Bug 6310; perl #67962 (fixed in perl 5.12/5.13) |
1456
|
0
|
|
|
|
|
0
|
my $dir = dirname($path); |
1457
|
|
|
|
|
|
|
|
1458
|
|
|
|
|
|
|
# make temporary copy since old dbm and new dbm may have same name |
1459
|
0
|
0
|
|
|
|
0
|
opendir(DIR, $dir) or die "bayes: can't opendir $dir: $!"; |
1460
|
0
|
|
|
|
|
0
|
my @files = grep { /^bayes_(?:seen|toks)(?:\.\w+)?$/ } readdir(DIR); |
|
0
|
|
|
|
|
0
|
|
1461
|
0
|
0
|
|
|
|
0
|
closedir(DIR) or die "bayes: can't close directory $dir: $!"; |
1462
|
0
|
0
|
0
|
|
|
0
|
if (@files < 2 || !grep(/bayes_seen/,@files) || !grep(/bayes_toks/,@files)) |
|
|
|
0
|
|
|
|
|
1463
|
|
|
|
|
|
|
{ |
1464
|
0
|
|
|
|
|
0
|
die "bayes: unable to find bayes_toks and bayes_seen, stopping\n"; |
1465
|
|
|
|
|
|
|
} |
1466
|
|
|
|
|
|
|
# untaint @files (already safe after grep) |
1467
|
0
|
|
|
|
|
0
|
untaint_var(\@files); |
1468
|
|
|
|
|
|
|
|
1469
|
0
|
|
|
|
|
0
|
for (@files) { |
1470
|
0
|
|
|
|
|
0
|
my $src = "$dir/$_"; |
1471
|
0
|
|
|
|
|
0
|
my $dst = "$dir/old_$_"; |
1472
|
0
|
0
|
|
|
|
0
|
eval q{ |
1473
|
|
|
|
|
|
|
use File::Copy; |
1474
|
|
|
|
|
|
|
copy($src, $dst); |
1475
|
|
|
|
|
|
|
} || die "bayes: can't copy $src to $dst: $!\n"; |
1476
|
|
|
|
|
|
|
} |
1477
|
|
|
|
|
|
|
|
1478
|
|
|
|
|
|
|
# delete previous to make way for import |
1479
|
0
|
|
|
|
|
0
|
for (@files) { unlink("$dir/$_"); } |
|
0
|
|
|
|
|
0
|
|
1480
|
|
|
|
|
|
|
|
1481
|
|
|
|
|
|
|
# import |
1482
|
0
|
0
|
|
|
|
0
|
if ($self->tie_db_writable()) { |
1483
|
|
|
|
|
|
|
$ret += $self->upgrade_old_dbm_files_trapped("$dir/old_bayes_seen", |
1484
|
0
|
|
|
|
|
0
|
$self->{db_seen}); |
1485
|
|
|
|
|
|
|
$ret += $self->upgrade_old_dbm_files_trapped("$dir/old_bayes_toks", |
1486
|
0
|
|
|
|
|
0
|
$self->{db_toks}); |
1487
|
|
|
|
|
|
|
} |
1488
|
|
|
|
|
|
|
|
1489
|
0
|
0
|
|
|
|
0
|
if ($ret == 2) { |
1490
|
0
|
|
|
|
|
0
|
print "import successful, original files saved with \"old\" prefix\n"; |
1491
|
|
|
|
|
|
|
} |
1492
|
|
|
|
|
|
|
else { |
1493
|
0
|
|
|
|
|
0
|
print "import failed, original files saved with \"old\" prefix\n"; |
1494
|
|
|
|
|
|
|
} |
1495
|
0
|
|
|
|
|
0
|
1; |
1496
|
0
|
0
|
|
|
|
0
|
} or do { |
1497
|
0
|
0
|
|
|
|
0
|
$eval_stat = $@ ne '' ? $@ : "errno=$!"; chomp $eval_stat; |
|
0
|
|
|
|
|
0
|
|
1498
|
|
|
|
|
|
|
}; |
1499
|
|
|
|
|
|
|
|
1500
|
0
|
|
|
|
|
0
|
$self->untie_db(); |
1501
|
|
|
|
|
|
|
|
1502
|
|
|
|
|
|
|
# if we died, untie the dbm files |
1503
|
0
|
0
|
|
|
|
0
|
if (defined $eval_stat) { |
1504
|
0
|
|
|
|
|
0
|
warn "bayes: perform_upgrade: $eval_stat\n"; |
1505
|
0
|
|
|
|
|
0
|
return 0; |
1506
|
|
|
|
|
|
|
} |
1507
|
0
|
|
|
|
|
0
|
$ret; |
1508
|
|
|
|
|
|
|
} |
1509
|
|
|
|
|
|
|
|
1510
|
|
|
|
|
|
|
sub upgrade_old_dbm_files_trapped { |
1511
|
0
|
|
|
0
|
0
|
0
|
my ($self, $filename, $output) = @_; |
1512
|
|
|
|
|
|
|
|
1513
|
0
|
|
|
|
|
0
|
my $count; |
1514
|
|
|
|
|
|
|
my %in; |
1515
|
|
|
|
|
|
|
|
1516
|
0
|
|
|
|
|
0
|
print "upgrading to DB_File, please be patient: $filename\n"; |
1517
|
|
|
|
|
|
|
|
1518
|
|
|
|
|
|
|
# try each type of file until we find one with > 0 entries |
1519
|
0
|
|
|
|
|
0
|
for my $dbm ('DB_File', 'GDBM_File', 'NDBM_File', 'SDBM_File') { |
1520
|
0
|
|
|
|
|
0
|
$count = 0; |
1521
|
|
|
|
|
|
|
# wrap in eval so it doesn't run in general use. This accesses db |
1522
|
|
|
|
|
|
|
# modules directly. |
1523
|
|
|
|
|
|
|
# Note: (bug 2390), the 'use' needs to be on the same line as the eval |
1524
|
|
|
|
|
|
|
# for RPM dependency checks to work properly. It's lame, but... |
1525
|
0
|
|
|
|
|
0
|
my $eval_stat; |
1526
|
|
|
|
|
|
|
eval 'use ' . $dbm . '; |
1527
|
|
|
|
|
|
|
tie %in, "' . $dbm . '", $filename, O_RDONLY, 0600; |
1528
|
|
|
|
|
|
|
%{ $output } = %in; |
1529
|
|
|
|
|
|
|
$count = scalar keys %{ $output }; |
1530
|
|
|
|
|
|
|
untie %in; |
1531
|
|
|
|
|
|
|
1; |
1532
|
0
|
0
|
|
|
|
0
|
' or do { |
1533
|
0
|
0
|
|
|
|
0
|
$eval_stat = $@ ne '' ? $@ : "errno=$!"; chomp $eval_stat; |
|
0
|
|
|
|
|
0
|
|
1534
|
|
|
|
|
|
|
}; |
1535
|
0
|
0
|
|
|
|
0
|
if (defined $eval_stat) { |
|
|
0
|
|
|
|
|
|
1536
|
0
|
|
|
|
|
0
|
print "$dbm: $dbm module not installed(?), nothing copied: $eval_stat\n"; |
1537
|
0
|
|
|
|
|
0
|
dbg("bayes: error was: $eval_stat"); |
1538
|
|
|
|
|
|
|
} |
1539
|
|
|
|
|
|
|
elsif ($count == 0) { |
1540
|
0
|
|
|
|
|
0
|
print "$dbm: no database of that kind found, nothing copied\n"; |
1541
|
|
|
|
|
|
|
} |
1542
|
|
|
|
|
|
|
else { |
1543
|
0
|
|
|
|
|
0
|
print "$dbm: copied $count entries\n"; |
1544
|
0
|
|
|
|
|
0
|
return 1; |
1545
|
|
|
|
|
|
|
} |
1546
|
|
|
|
|
|
|
} |
1547
|
|
|
|
|
|
|
|
1548
|
0
|
|
|
|
|
0
|
return 0; |
1549
|
|
|
|
|
|
|
} |
1550
|
|
|
|
|
|
|
|
1551
|
|
|
|
|
|
|
sub clear_database { |
1552
|
2
|
|
|
2
|
1
|
10
|
my ($self) = @_; |
1553
|
|
|
|
|
|
|
|
1554
|
2
|
50
|
|
|
|
17
|
return 0 unless ($self->tie_db_writable()); |
1555
|
|
|
|
|
|
|
|
1556
|
2
|
|
|
|
|
11
|
dbg("bayes: untie-ing in preparation for removal."); |
1557
|
|
|
|
|
|
|
|
1558
|
2
|
|
|
|
|
15
|
foreach my $dbname (@DBNAMES) { |
1559
|
4
|
|
|
|
|
17
|
my $db_var = 'db_'.$dbname; |
1560
|
|
|
|
|
|
|
|
1561
|
4
|
50
|
|
|
|
16
|
if (exists $self->{$db_var}) { |
1562
|
|
|
|
|
|
|
# dbg("bayes: untie-ing $db_var"); |
1563
|
4
|
|
|
|
|
7
|
untie %{$self->{$db_var}}; |
|
4
|
|
|
|
|
61
|
|
1564
|
4
|
|
|
|
|
76
|
delete $self->{$db_var}; |
1565
|
|
|
|
|
|
|
} |
1566
|
|
|
|
|
|
|
} |
1567
|
|
|
|
|
|
|
|
1568
|
2
|
|
|
|
|
18
|
my $path = $self->{bayes}->{main}->sed_path($self->{bayes}->{main}->{conf}->{bayes_path}); |
1569
|
|
|
|
|
|
|
|
1570
|
2
|
|
|
|
|
7
|
foreach my $dbname (@DBNAMES, 'journal') { |
1571
|
6
|
|
|
|
|
20
|
foreach my $ext ($self->DB_EXTENSIONS) { |
1572
|
12
|
|
|
|
|
35
|
my $name = $path.'_'.$dbname.$ext; |
1573
|
12
|
|
|
|
|
647
|
my $ret = unlink $name; |
1574
|
12
|
100
|
|
|
|
69
|
dbg("bayes: clear_database: %s %s", |
1575
|
|
|
|
|
|
|
$ret ? 'removed' : 'tried to remove', $name); |
1576
|
|
|
|
|
|
|
} |
1577
|
|
|
|
|
|
|
} |
1578
|
|
|
|
|
|
|
|
1579
|
|
|
|
|
|
|
# the journal file needs to be done separately since it has no extension |
1580
|
2
|
|
|
|
|
9
|
foreach my $dbname ('journal') { |
1581
|
2
|
|
|
|
|
9
|
my $name = $path.'_'.$dbname; |
1582
|
2
|
|
|
|
|
78
|
my $ret = unlink $name; |
1583
|
2
|
50
|
|
|
|
20
|
dbg("bayes: clear_database: %s %s", |
1584
|
|
|
|
|
|
|
$ret ? 'removed' : 'tried to remove', $name); |
1585
|
|
|
|
|
|
|
} |
1586
|
|
|
|
|
|
|
|
1587
|
2
|
|
|
|
|
10
|
$self->untie_db(); |
1588
|
|
|
|
|
|
|
|
1589
|
2
|
|
|
|
|
12
|
return 1; |
1590
|
|
|
|
|
|
|
} |
1591
|
|
|
|
|
|
|
|
1592
|
|
|
|
|
|
|
sub backup_database { |
1593
|
0
|
|
|
0
|
1
|
0
|
my ($self) = @_; |
1594
|
|
|
|
|
|
|
|
1595
|
|
|
|
|
|
|
# we tie writable because we want the upgrade code to kick in if needed |
1596
|
0
|
0
|
|
|
|
0
|
return 0 unless ($self->tie_db_writable()); |
1597
|
|
|
|
|
|
|
|
1598
|
0
|
|
|
|
|
0
|
my @vars = $self->get_storage_variables(); |
1599
|
|
|
|
|
|
|
|
1600
|
0
|
|
|
|
|
0
|
print "v\t$vars[6]\tdb_version # this must be the first line!!!\n"; |
1601
|
0
|
|
|
|
|
0
|
print "v\t$vars[1]\tnum_spam\n"; |
1602
|
0
|
|
|
|
|
0
|
print "v\t$vars[2]\tnum_nonspam\n"; |
1603
|
|
|
|
|
|
|
|
1604
|
0
|
|
|
|
|
0
|
while (my ($tok, $packed) = each %{$self->{db_toks}}) { |
|
0
|
|
|
|
|
0
|
|
1605
|
0
|
0
|
|
|
|
0
|
next if ($tok =~ MAGIC_RE); # skip magic tokens |
1606
|
|
|
|
|
|
|
|
1607
|
0
|
|
|
|
|
0
|
my ($ts, $th, $atime) = $self->tok_unpack($packed); |
1608
|
0
|
|
|
|
|
0
|
my $encoded_token = unpack("H*",$tok); |
1609
|
0
|
|
|
|
|
0
|
print "t\t$ts\t$th\t$atime\t$encoded_token\n"; |
1610
|
|
|
|
|
|
|
} |
1611
|
|
|
|
|
|
|
|
1612
|
0
|
|
|
|
|
0
|
while (my ($msgid, $flag) = each %{$self->{db_seen}}) { |
|
0
|
|
|
|
|
0
|
|
1613
|
0
|
|
|
|
|
0
|
print "s\t$flag\t$msgid\n"; |
1614
|
|
|
|
|
|
|
} |
1615
|
|
|
|
|
|
|
|
1616
|
0
|
|
|
|
|
0
|
$self->untie_db(); |
1617
|
|
|
|
|
|
|
|
1618
|
0
|
|
|
|
|
0
|
return 1; |
1619
|
|
|
|
|
|
|
} |
1620
|
|
|
|
|
|
|
|
1621
|
|
|
|
|
|
|
sub restore_database { |
1622
|
0
|
|
|
0
|
1
|
0
|
my ($self, $filename, $showdots) = @_; |
1623
|
|
|
|
|
|
|
|
1624
|
0
|
|
|
|
|
0
|
local *DUMPFILE; |
1625
|
0
|
0
|
|
|
|
0
|
if (!open(DUMPFILE, '<', $filename)) { |
1626
|
0
|
|
|
|
|
0
|
dbg("bayes: unable to open backup file $filename: $!"); |
1627
|
0
|
|
|
|
|
0
|
return 0; |
1628
|
|
|
|
|
|
|
} |
1629
|
|
|
|
|
|
|
|
1630
|
0
|
0
|
|
|
|
0
|
if (!$self->tie_db_writable()) { |
1631
|
0
|
|
|
|
|
0
|
dbg("bayes: failed to tie db writable"); |
1632
|
0
|
|
|
|
|
0
|
return 0; |
1633
|
|
|
|
|
|
|
} |
1634
|
|
|
|
|
|
|
|
1635
|
0
|
|
|
|
|
0
|
my $main = $self->{bayes}->{main}; |
1636
|
0
|
|
|
|
|
0
|
my $path = $main->sed_path($main->{conf}->{bayes_path}); |
1637
|
|
|
|
|
|
|
|
1638
|
|
|
|
|
|
|
# use a temporary PID-based suffix just in case another one was |
1639
|
|
|
|
|
|
|
# created previously by an interrupted expire |
1640
|
0
|
|
|
|
|
0
|
my $tmpsuffix = "convert$$"; |
1641
|
0
|
|
|
|
|
0
|
my $tmptoksdbname = $path.'_toks.'.$tmpsuffix; |
1642
|
0
|
|
|
|
|
0
|
my $tmpseendbname = $path.'_seen.'.$tmpsuffix; |
1643
|
0
|
|
|
|
|
0
|
my $toksdbname = $path.'_toks'; |
1644
|
0
|
|
|
|
|
0
|
my $seendbname = $path.'_seen'; |
1645
|
|
|
|
|
|
|
|
1646
|
0
|
|
|
|
|
0
|
my %new_toks; |
1647
|
|
|
|
|
|
|
my %new_seen; |
1648
|
0
|
|
|
|
|
0
|
my $umask = umask 0; |
1649
|
0
|
0
|
|
|
|
0
|
unless (tie %new_toks, $self->DBM_MODULE, $tmptoksdbname, O_RDWR|O_CREAT|O_EXCL, |
1650
|
|
|
|
|
|
|
(oct ($main->{conf}->{bayes_file_mode}) & 0666)) { |
1651
|
0
|
|
|
|
|
0
|
dbg("bayes: failed to tie temp toks db: $!"); |
1652
|
0
|
|
|
|
|
0
|
$self->untie_db(); |
1653
|
0
|
|
|
|
|
0
|
umask $umask; |
1654
|
0
|
|
|
|
|
0
|
return 0; |
1655
|
|
|
|
|
|
|
} |
1656
|
0
|
0
|
|
|
|
0
|
unless (tie %new_seen, $self->DBM_MODULE, $tmpseendbname, O_RDWR|O_CREAT|O_EXCL, |
1657
|
|
|
|
|
|
|
(oct ($main->{conf}->{bayes_file_mode}) & 0666)) { |
1658
|
0
|
|
|
|
|
0
|
dbg("bayes: failed to tie temp seen db: $!"); |
1659
|
0
|
|
|
|
|
0
|
untie %new_toks; |
1660
|
0
|
|
|
|
|
0
|
$self->_unlink_file($tmptoksdbname); |
1661
|
0
|
|
|
|
|
0
|
$self->untie_db(); |
1662
|
0
|
|
|
|
|
0
|
umask $umask; |
1663
|
0
|
|
|
|
|
0
|
return 0; |
1664
|
|
|
|
|
|
|
} |
1665
|
0
|
|
|
|
|
0
|
umask $umask; |
1666
|
|
|
|
|
|
|
|
1667
|
0
|
|
|
|
|
0
|
my $line_count = 0; |
1668
|
0
|
|
|
|
|
0
|
my $db_version; |
1669
|
0
|
|
|
|
|
0
|
my $token_count = 0; |
1670
|
0
|
|
|
|
|
0
|
my $num_spam; |
1671
|
|
|
|
|
|
|
my $num_ham; |
1672
|
0
|
|
|
|
|
0
|
my $error_p = 0; |
1673
|
0
|
|
|
|
|
0
|
my $newest_token_age = 0; |
1674
|
|
|
|
|
|
|
# Kinda wierd I know, but we need a nice big value and we know there will be |
1675
|
|
|
|
|
|
|
# no tokens > time() since we reset atime if > time(), so use that with a |
1676
|
|
|
|
|
|
|
# little buffer just in case. |
1677
|
0
|
|
|
|
|
0
|
my $oldest_token_age = time() + 100000; |
1678
|
|
|
|
|
|
|
|
1679
|
0
|
|
|
|
|
0
|
my $line = <DUMPFILE>; |
1680
|
0
|
0
|
|
|
|
0
|
defined $line or die "Error reading dump file: $!"; |
1681
|
0
|
|
|
|
|
0
|
$line_count++; |
1682
|
|
|
|
|
|
|
|
1683
|
|
|
|
|
|
|
# We require the database version line to be the first in the file so we can |
1684
|
|
|
|
|
|
|
# figure out how to properly deal with the file. If it is not the first |
1685
|
|
|
|
|
|
|
# line then fail |
1686
|
0
|
0
|
|
|
|
0
|
if ($line =~ m/^v\s+(\d+)\s+db_version/) { |
1687
|
0
|
|
|
|
|
0
|
$db_version = $1; |
1688
|
|
|
|
|
|
|
} |
1689
|
|
|
|
|
|
|
else { |
1690
|
0
|
|
|
|
|
0
|
dbg("bayes: database version must be the first line in the backup file, correct and re-run"); |
1691
|
0
|
|
|
|
|
0
|
untie %new_toks; |
1692
|
0
|
|
|
|
|
0
|
untie %new_seen; |
1693
|
0
|
|
|
|
|
0
|
$self->_unlink_file($tmptoksdbname); |
1694
|
0
|
|
|
|
|
0
|
$self->_unlink_file($tmpseendbname); |
1695
|
0
|
|
|
|
|
0
|
$self->untie_db(); |
1696
|
0
|
|
|
|
|
0
|
return 0; |
1697
|
|
|
|
|
|
|
} |
1698
|
|
|
|
|
|
|
|
1699
|
0
|
0
|
0
|
|
|
0
|
unless ($db_version == 2 || $db_version == 3) { |
1700
|
0
|
|
|
|
|
0
|
warn("bayes: database version $db_version is unsupported, must be version 2 or 3"); |
1701
|
0
|
|
|
|
|
0
|
untie %new_toks; |
1702
|
0
|
|
|
|
|
0
|
untie %new_seen; |
1703
|
0
|
|
|
|
|
0
|
$self->_unlink_file($tmptoksdbname); |
1704
|
0
|
|
|
|
|
0
|
$self->_unlink_file($tmpseendbname); |
1705
|
0
|
|
|
|
|
0
|
$self->untie_db(); |
1706
|
0
|
|
|
|
|
0
|
return 0; |
1707
|
|
|
|
|
|
|
} |
1708
|
|
|
|
|
|
|
|
1709
|
0
|
|
|
|
|
0
|
for ($!=0; defined($line=<DUMPFILE>); $!=0) { |
1710
|
0
|
|
|
|
|
0
|
chomp($line); |
1711
|
0
|
|
|
|
|
0
|
$line_count++; |
1712
|
|
|
|
|
|
|
|
1713
|
0
|
0
|
|
|
|
0
|
if ($line_count % 1000 == 0) { |
1714
|
0
|
0
|
|
|
|
0
|
print STDERR "." if ($showdots); |
1715
|
|
|
|
|
|
|
} |
1716
|
|
|
|
|
|
|
|
1717
|
0
|
0
|
|
|
|
0
|
if ($line =~ /^v\s+/) { # variable line |
|
|
0
|
|
|
|
|
|
|
|
0
|
|
|
|
|
|
1718
|
0
|
|
|
|
|
0
|
my @parsed_line = split(/\s+/, $line, 3); |
1719
|
0
|
|
|
|
|
0
|
my $value = $parsed_line[1] + 0; |
1720
|
0
|
0
|
|
|
|
0
|
if ($parsed_line[2] eq 'num_spam') { |
|
|
0
|
|
|
|
|
|
1721
|
0
|
|
|
|
|
0
|
$num_spam = $value; |
1722
|
|
|
|
|
|
|
} |
1723
|
|
|
|
|
|
|
elsif ($parsed_line[2] eq 'num_nonspam') { |
1724
|
0
|
|
|
|
|
0
|
$num_ham = $value; |
1725
|
|
|
|
|
|
|
} |
1726
|
|
|
|
|
|
|
else { |
1727
|
0
|
|
|
|
|
0
|
dbg("bayes: restore_database: skipping unknown line: $line"); |
1728
|
|
|
|
|
|
|
} |
1729
|
|
|
|
|
|
|
} |
1730
|
|
|
|
|
|
|
elsif ($line =~ /^t\s+/) { # token line |
1731
|
0
|
|
|
|
|
0
|
my @parsed_line = split(/\s+/, $line, 5); |
1732
|
0
|
|
|
|
|
0
|
my $spam_count = $parsed_line[1] + 0; |
1733
|
0
|
|
|
|
|
0
|
my $ham_count = $parsed_line[2] + 0; |
1734
|
0
|
|
|
|
|
0
|
my $atime = $parsed_line[3] + 0; |
1735
|
0
|
|
|
|
|
0
|
my $token = $parsed_line[4]; |
1736
|
|
|
|
|
|
|
|
1737
|
0
|
|
|
|
|
0
|
my $token_warn_p = 0; |
1738
|
0
|
|
|
|
|
0
|
my @warnings; |
1739
|
|
|
|
|
|
|
|
1740
|
0
|
0
|
|
|
|
0
|
if ($spam_count < 0) { |
1741
|
0
|
|
|
|
|
0
|
$spam_count = 0; |
1742
|
0
|
|
|
|
|
0
|
push(@warnings, 'spam count < 0, resetting'); |
1743
|
0
|
|
|
|
|
0
|
$token_warn_p = 1; |
1744
|
|
|
|
|
|
|
} |
1745
|
0
|
0
|
|
|
|
0
|
if ($ham_count < 0) { |
1746
|
0
|
|
|
|
|
0
|
$ham_count = 0; |
1747
|
0
|
|
|
|
|
0
|
push(@warnings, 'ham count < 0, resetting'); |
1748
|
0
|
|
|
|
|
0
|
$token_warn_p = 1; |
1749
|
|
|
|
|
|
|
} |
1750
|
|
|
|
|
|
|
|
1751
|
0
|
0
|
0
|
|
|
0
|
if ($spam_count == 0 && $ham_count == 0) { |
1752
|
0
|
|
|
|
|
0
|
dbg("bayes: token has zero spam and ham count, skipping"); |
1753
|
0
|
|
|
|
|
0
|
next; |
1754
|
|
|
|
|
|
|
} |
1755
|
|
|
|
|
|
|
|
1756
|
0
|
0
|
|
|
|
0
|
if ($atime > time()) { |
1757
|
0
|
|
|
|
|
0
|
$atime = time(); |
1758
|
0
|
|
|
|
|
0
|
push(@warnings, 'atime > current time, resetting'); |
1759
|
0
|
|
|
|
|
0
|
$token_warn_p = 1; |
1760
|
|
|
|
|
|
|
} |
1761
|
|
|
|
|
|
|
|
1762
|
0
|
0
|
|
|
|
0
|
if ($token_warn_p) { |
1763
|
0
|
|
|
|
|
0
|
dbg("bayes: token (%s) has the following warnings:\n%s", |
1764
|
|
|
|
|
|
|
$token, join("\n",@warnings)); |
1765
|
|
|
|
|
|
|
} |
1766
|
|
|
|
|
|
|
|
1767
|
|
|
|
|
|
|
# database versions < 3 did not encode their token values |
1768
|
0
|
0
|
|
|
|
0
|
if ($db_version < 3) { |
1769
|
0
|
|
|
|
|
0
|
$token = substr(sha1($token), -5); |
1770
|
|
|
|
|
|
|
} |
1771
|
|
|
|
|
|
|
else { |
1772
|
|
|
|
|
|
|
# turn unpacked binary token back into binary value |
1773
|
0
|
|
|
|
|
0
|
$token = pack("H*",$token); |
1774
|
|
|
|
|
|
|
} |
1775
|
|
|
|
|
|
|
|
1776
|
0
|
|
|
|
|
0
|
$new_toks{$token} = $self->tok_pack($spam_count, $ham_count, $atime); |
1777
|
0
|
0
|
|
|
|
0
|
if ($atime < $oldest_token_age) { |
1778
|
0
|
|
|
|
|
0
|
$oldest_token_age = $atime; |
1779
|
|
|
|
|
|
|
} |
1780
|
0
|
0
|
|
|
|
0
|
if ($atime > $newest_token_age) { |
1781
|
0
|
|
|
|
|
0
|
$newest_token_age = $atime; |
1782
|
|
|
|
|
|
|
} |
1783
|
0
|
|
|
|
|
0
|
$token_count++; |
1784
|
|
|
|
|
|
|
} |
1785
|
|
|
|
|
|
|
elsif ($line =~ /^s\s+/) { # seen line |
1786
|
0
|
|
|
|
|
0
|
my @parsed_line = split(/\s+/, $line, 3); |
1787
|
0
|
|
|
|
|
0
|
my $flag = $parsed_line[1]; |
1788
|
0
|
|
|
|
|
0
|
my $msgid = $parsed_line[2]; |
1789
|
|
|
|
|
|
|
|
1790
|
0
|
0
|
0
|
|
|
0
|
unless ($flag eq 'h' || $flag eq 's') { |
1791
|
0
|
|
|
|
|
0
|
dbg("bayes: unknown seen flag ($flag) for line: $line, skipping"); |
1792
|
0
|
|
|
|
|
0
|
next; |
1793
|
|
|
|
|
|
|
} |
1794
|
|
|
|
|
|
|
|
1795
|
0
|
0
|
|
|
|
0
|
unless ($msgid) { |
1796
|
0
|
|
|
|
|
0
|
dbg("bayes: blank msgid for line: $line, skipping"); |
1797
|
0
|
|
|
|
|
0
|
next; |
1798
|
|
|
|
|
|
|
} |
1799
|
|
|
|
|
|
|
|
1800
|
0
|
|
|
|
|
0
|
$new_seen{$msgid} = $flag; |
1801
|
|
|
|
|
|
|
} |
1802
|
|
|
|
|
|
|
else { |
1803
|
0
|
|
|
|
|
0
|
dbg("bayes: skipping unknown line: $line"); |
1804
|
0
|
|
|
|
|
0
|
next; |
1805
|
|
|
|
|
|
|
} |
1806
|
|
|
|
|
|
|
} |
1807
|
0
|
0
|
0
|
|
|
0
|
defined $line || $!==0 or die "Error reading dump file: $!"; |
1808
|
0
|
0
|
|
|
|
0
|
close(DUMPFILE) or die "Can't close dump file: $!"; |
1809
|
|
|
|
|
|
|
|
1810
|
0
|
0
|
|
|
|
0
|
print STDERR "\n" if ($showdots); |
1811
|
|
|
|
|
|
|
|
1812
|
0
|
0
|
|
|
|
0
|
unless (defined($num_spam)) { |
1813
|
0
|
|
|
|
|
0
|
dbg("bayes: unable to find num spam, please check file"); |
1814
|
0
|
|
|
|
|
0
|
$error_p = 1; |
1815
|
|
|
|
|
|
|
} |
1816
|
|
|
|
|
|
|
|
1817
|
0
|
0
|
|
|
|
0
|
unless (defined($num_ham)) { |
1818
|
0
|
|
|
|
|
0
|
dbg("bayes: unable to find num ham, please check file"); |
1819
|
0
|
|
|
|
|
0
|
$error_p = 1; |
1820
|
|
|
|
|
|
|
} |
1821
|
|
|
|
|
|
|
|
1822
|
0
|
0
|
|
|
|
0
|
if ($error_p) { |
1823
|
0
|
|
|
|
|
0
|
dbg("bayes: error(s) while attempting to load $filename, correct and re-run"); |
1824
|
|
|
|
|
|
|
|
1825
|
0
|
|
|
|
|
0
|
untie %new_toks; |
1826
|
0
|
|
|
|
|
0
|
untie %new_seen; |
1827
|
0
|
|
|
|
|
0
|
$self->_unlink_file($tmptoksdbname); |
1828
|
0
|
|
|
|
|
0
|
$self->_unlink_file($tmpseendbname); |
1829
|
0
|
|
|
|
|
0
|
$self->untie_db(); |
1830
|
0
|
|
|
|
|
0
|
return 0; |
1831
|
|
|
|
|
|
|
} |
1832
|
|
|
|
|
|
|
|
1833
|
|
|
|
|
|
|
# set the calculated magic tokens |
1834
|
0
|
|
|
|
|
0
|
$new_toks{$DB_VERSION_MAGIC_TOKEN} = $self->DB_VERSION(); |
1835
|
0
|
|
|
|
|
0
|
$new_toks{$NTOKENS_MAGIC_TOKEN} = $token_count; |
1836
|
0
|
|
|
|
|
0
|
$new_toks{$NSPAM_MAGIC_TOKEN} = $num_spam; |
1837
|
0
|
|
|
|
|
0
|
$new_toks{$NHAM_MAGIC_TOKEN} = $num_ham; |
1838
|
0
|
|
|
|
|
0
|
$new_toks{$NEWEST_TOKEN_AGE_MAGIC_TOKEN} = $newest_token_age; |
1839
|
0
|
|
|
|
|
0
|
$new_toks{$OLDEST_TOKEN_AGE_MAGIC_TOKEN} = $oldest_token_age; |
1840
|
|
|
|
|
|
|
|
1841
|
|
|
|
|
|
|
# go ahead and zero out these, chances are good that they are bogus anyway. |
1842
|
0
|
|
|
|
|
0
|
$new_toks{$LAST_EXPIRE_MAGIC_TOKEN} = 0; |
1843
|
0
|
|
|
|
|
0
|
$new_toks{$LAST_JOURNAL_SYNC_MAGIC_TOKEN} = 0; |
1844
|
0
|
|
|
|
|
0
|
$new_toks{$LAST_ATIME_DELTA_MAGIC_TOKEN} = 0; |
1845
|
0
|
|
|
|
|
0
|
$new_toks{$LAST_EXPIRE_REDUCE_MAGIC_TOKEN} = 0; |
1846
|
|
|
|
|
|
|
|
1847
|
0
|
|
|
|
|
0
|
local $SIG{'INT'} = 'IGNORE'; |
1848
|
0
|
|
|
|
|
0
|
local $SIG{'TERM'} = 'IGNORE'; |
1849
|
0
|
0
|
|
|
|
0
|
local $SIG{'HUP'} = 'IGNORE' if !am_running_on_windows(); |
1850
|
|
|
|
|
|
|
|
1851
|
0
|
|
|
|
|
0
|
untie %new_toks; |
1852
|
0
|
|
|
|
|
0
|
untie %new_seen; |
1853
|
0
|
|
|
|
|
0
|
$self->untie_db(); |
1854
|
|
|
|
|
|
|
|
1855
|
|
|
|
|
|
|
# Here is where something can go horribly wrong and screw up the bayes |
1856
|
|
|
|
|
|
|
# database files. If we are able to copy one and not the other then it |
1857
|
|
|
|
|
|
|
# will leave the database in an inconsistent state. Since this is an |
1858
|
|
|
|
|
|
|
# edge case, and they're trying to replace the DB anyway we should be ok. |
1859
|
0
|
0
|
|
|
|
0
|
unless ($self->_rename_file($tmptoksdbname, $toksdbname)) { |
1860
|
0
|
|
|
|
|
0
|
dbg("bayes: error while renaming $tmptoksdbname to $toksdbname: $!"); |
1861
|
0
|
|
|
|
|
0
|
return 0; |
1862
|
|
|
|
|
|
|
} |
1863
|
0
|
0
|
|
|
|
0
|
unless ($self->_rename_file($tmpseendbname, $seendbname)) { |
1864
|
0
|
|
|
|
|
0
|
dbg("bayes: error while renaming $tmpseendbname to $seendbname: $!"); |
1865
|
0
|
|
|
|
|
0
|
dbg("bayes: database now in inconsistent state"); |
1866
|
0
|
|
|
|
|
0
|
return 0; |
1867
|
|
|
|
|
|
|
} |
1868
|
|
|
|
|
|
|
|
1869
|
0
|
|
|
|
|
0
|
dbg("bayes: parsed $line_count lines"); |
1870
|
0
|
|
|
|
|
0
|
dbg("bayes: created database with $token_count tokens based on $num_spam spam messages and $num_ham ham messages"); |
1871
|
|
|
|
|
|
|
|
1872
|
0
|
|
|
|
|
0
|
return 1; |
1873
|
|
|
|
|
|
|
} |
1874
|
|
|
|
|
|
|
|
1875
|
|
|
|
|
|
|
########################################################################### |
1876
|
|
|
|
|
|
|
|
1877
|
|
|
|
|
|
|
# token marshalling format for db_toks. |
1878
|
|
|
|
|
|
|
|
1879
|
|
|
|
|
|
|
# Since we may have many entries with few hits, especially thousands of hapaxes |
1880
|
|
|
|
|
|
|
# (1-occurrence entries), use a flexible entry format, instead of simply "2 |
1881
|
|
|
|
|
|
|
# packed ints", to keep the memory and disk space usage down. In my |
1882
|
|
|
|
|
|
|
# 18k-message test corpus, only 8.9% have >= 8 hits in either counter, so we |
1883
|
|
|
|
|
|
|
# can use a 1-byte representation for the other 91% of low-hitting entries |
1884
|
|
|
|
|
|
|
# and save masses of space. |
1885
|
|
|
|
|
|
|
|
1886
|
|
|
|
|
|
|
# This looks like: XXSSSHHH (XX = format bits, SSS = 3 spam-count bits, HHH = 3 |
1887
|
|
|
|
|
|
|
# ham-count bits). If XX in the first byte is 11, it's packed as this 1-byte |
1888
|
|
|
|
|
|
|
# representation; otherwise, if XX in the first byte is 00, it's packed as |
1889
|
|
|
|
|
|
|
# "CLL", ie. 1 byte and 2 32-bit "longs" in perl pack format. |
1890
|
|
|
|
|
|
|
|
1891
|
|
|
|
|
|
|
# Savings: roughly halves size of toks db, at the cost of a ~10% slowdown. |
1892
|
|
|
|
|
|
|
|
1893
|
21
|
|
|
21
|
|
233
|
use constant FORMAT_FLAG => 0xc0; # 11000000 |
|
21
|
|
|
|
|
52
|
|
|
21
|
|
|
|
|
1447
|
|
1894
|
21
|
|
|
21
|
|
163
|
use constant ONE_BYTE_FORMAT => 0xc0; # 11000000 |
|
21
|
|
|
|
|
43
|
|
|
21
|
|
|
|
|
1148
|
|
1895
|
21
|
|
|
21
|
|
147
|
use constant TWO_LONGS_FORMAT => 0x00; # 00000000 |
|
21
|
|
|
|
|
55
|
|
|
21
|
|
|
|
|
1148
|
|
1896
|
|
|
|
|
|
|
|
1897
|
21
|
|
|
21
|
|
146
|
use constant ONE_BYTE_SSS_BITS => 0x38; # 00111000 |
|
21
|
|
|
|
|
60
|
|
|
21
|
|
|
|
|
1016
|
|
1898
|
21
|
|
|
21
|
|
151
|
use constant ONE_BYTE_HHH_BITS => 0x07; # 00000111 |
|
21
|
|
|
|
|
63
|
|
|
21
|
|
|
|
|
10096
|
|
1899
|
|
|
|
|
|
|
|
1900
|
|
|
|
|
|
|
sub tok_unpack { |
1901
|
5536
|
|
|
5536
|
0
|
53150
|
my ($self, $value) = @_; |
1902
|
5536
|
|
100
|
|
|
17242
|
$value ||= 0; |
1903
|
|
|
|
|
|
|
|
1904
|
5536
|
|
|
|
|
6497
|
my ($packed, $atime); |
1905
|
5536
|
50
|
|
|
|
10610
|
if ($self->{db_version} >= 1) { |
|
|
0
|
|
|
|
|
|
1906
|
5536
|
|
|
|
|
12196
|
($packed, $atime) = unpack("CV", $value); |
1907
|
|
|
|
|
|
|
} |
1908
|
|
|
|
|
|
|
elsif ($self->{db_version} == 0) { |
1909
|
0
|
|
|
|
|
0
|
($packed, $atime) = unpack("CS", $value); |
1910
|
|
|
|
|
|
|
} |
1911
|
|
|
|
|
|
|
|
1912
|
5536
|
100
|
|
|
|
11476
|
if (($packed & FORMAT_FLAG) == ONE_BYTE_FORMAT) { |
|
|
50
|
|
|
|
|
|
1913
|
3138
|
|
50
|
|
|
9826
|
return (($packed & ONE_BYTE_SSS_BITS) >> 3, |
1914
|
|
|
|
|
|
|
$packed & ONE_BYTE_HHH_BITS, |
1915
|
|
|
|
|
|
|
$atime || 0); |
1916
|
|
|
|
|
|
|
} |
1917
|
|
|
|
|
|
|
elsif (($packed & FORMAT_FLAG) == TWO_LONGS_FORMAT) { |
1918
|
2398
|
|
|
|
|
2975
|
my ($packed, $ts, $th, $atime); |
1919
|
2398
|
50
|
|
|
|
3643
|
if ($self->{db_version} >= 1) { |
|
|
0
|
|
|
|
|
|
1920
|
2398
|
|
|
|
|
3920
|
($packed, $ts, $th, $atime) = unpack("CVVV", $value); |
1921
|
|
|
|
|
|
|
} |
1922
|
|
|
|
|
|
|
elsif ($self->{db_version} == 0) { |
1923
|
0
|
|
|
|
|
0
|
($packed, $ts, $th, $atime) = unpack("CLLS", $value); |
1924
|
|
|
|
|
|
|
} |
1925
|
2398
|
|
100
|
|
|
12972
|
return ($ts || 0, $th || 0, $atime || 0); |
|
|
|
100
|
|
|
|
|
|
|
|
100
|
|
|
|
|
1926
|
|
|
|
|
|
|
} |
1927
|
|
|
|
|
|
|
# other formats would go here... |
1928
|
|
|
|
|
|
|
else { |
1929
|
0
|
|
|
|
|
0
|
warn "bayes: unknown packing format for bayes db, please re-learn: $packed"; |
1930
|
0
|
|
|
|
|
0
|
return (0, 0, 0); |
1931
|
|
|
|
|
|
|
} |
1932
|
|
|
|
|
|
|
} |
1933
|
|
|
|
|
|
|
|
1934
|
|
|
|
|
|
|
sub tok_pack { |
1935
|
1662
|
|
|
1662
|
0
|
2877
|
my ($self, $ts, $th, $atime) = @_; |
1936
|
1662
|
|
100
|
|
|
3801
|
$ts ||= 0; $th ||= 0; $atime ||= 0; |
|
1662
|
|
100
|
|
|
4200
|
|
|
1662
|
|
50
|
|
|
2305
|
|
1937
|
1662
|
50
|
33
|
|
|
4059
|
if ($ts < 8 && $th < 8) { |
1938
|
1662
|
|
|
|
|
29573
|
return pack ("CV", ONE_BYTE_FORMAT | ($ts << 3) | $th, $atime); |
1939
|
|
|
|
|
|
|
} else { |
1940
|
0
|
|
|
|
|
0
|
return pack ("CVVV", TWO_LONGS_FORMAT, $ts, $th, $atime); |
1941
|
|
|
|
|
|
|
} |
1942
|
|
|
|
|
|
|
} |
1943
|
|
|
|
|
|
|
|
1944
|
|
|
|
|
|
|
########################################################################### |
1945
|
|
|
|
|
|
|
|
1946
|
|
|
|
|
|
|
sub db_readable { |
1947
|
16
|
|
|
16
|
1
|
38
|
my ($self) = @_; |
1948
|
16
|
|
|
|
|
68
|
return $self->{already_tied}; |
1949
|
|
|
|
|
|
|
} |
1950
|
|
|
|
|
|
|
|
1951
|
|
|
|
|
|
|
sub db_writable { |
1952
|
0
|
|
|
0
|
1
|
|
my ($self) = @_; |
1953
|
0
|
|
0
|
|
|
|
return $self->{already_tied} && $self->{is_locked}; |
1954
|
|
|
|
|
|
|
} |
1955
|
|
|
|
|
|
|
|
1956
|
|
|
|
|
|
|
########################################################################### |
1957
|
|
|
|
|
|
|
|
1958
|
|
|
|
|
|
|
sub _unlink_file { |
1959
|
0
|
|
|
0
|
|
|
my ($self, $filename) = @_; |
1960
|
|
|
|
|
|
|
|
1961
|
0
|
|
|
|
|
|
unlink $filename; |
1962
|
|
|
|
|
|
|
} |
1963
|
|
|
|
|
|
|
|
1964
|
|
|
|
|
|
|
sub _rename_file { |
1965
|
0
|
|
|
0
|
|
|
my ($self, $sourcefilename, $targetfilename) = @_; |
1966
|
|
|
|
|
|
|
|
1967
|
0
|
0
|
|
|
|
|
return 0 unless (rename($sourcefilename, $targetfilename)); |
1968
|
|
|
|
|
|
|
|
1969
|
0
|
|
|
|
|
|
return 1; |
1970
|
|
|
|
|
|
|
} |
1971
|
|
|
|
|
|
|
|
1972
|
0
|
|
|
0
|
0
|
|
sub sa_die { Mail::SpamAssassin::sa_die(@_); } |
1973
|
|
|
|
|
|
|
|
1974
|
|
|
|
|
|
|
1; |