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
|
|
|
|
|
|
|
=head1 NAME |
19
|
|
|
|
|
|
|
|
20
|
|
|
|
|
|
|
Mail::SpamAssassin::Bayes - support for learning classifiers |
21
|
|
|
|
|
|
|
|
22
|
|
|
|
|
|
|
=head1 DESCRIPTION |
23
|
|
|
|
|
|
|
|
24
|
|
|
|
|
|
|
This is the general class used to train a learning classifier with new samples |
25
|
|
|
|
|
|
|
of spam and ham mail, and classify based on prior training. |
26
|
|
|
|
|
|
|
|
27
|
|
|
|
|
|
|
Prior to version 3.3.0, the default Bayes implementation was here; if you're |
28
|
|
|
|
|
|
|
looking for information on that, it has moved to |
29
|
|
|
|
|
|
|
C<Mail::SpamAssassin::Plugin::Bayes>. |
30
|
|
|
|
|
|
|
|
31
|
|
|
|
|
|
|
=cut |
32
|
|
|
|
|
|
|
|
33
|
|
|
|
|
|
|
package Mail::SpamAssassin::Bayes; |
34
|
|
|
|
|
|
|
|
35
|
22
|
|
|
22
|
|
157
|
use strict; |
|
22
|
|
|
|
|
46
|
|
|
22
|
|
|
|
|
767
|
|
36
|
22
|
|
|
22
|
|
150
|
use warnings; |
|
22
|
|
|
|
|
55
|
|
|
22
|
|
|
|
|
1045
|
|
37
|
|
|
|
|
|
|
# use bytes; |
38
|
22
|
|
|
22
|
|
149
|
use re 'taint'; |
|
22
|
|
|
|
|
49
|
|
|
22
|
|
|
|
|
1000
|
|
39
|
|
|
|
|
|
|
|
40
|
22
|
|
|
22
|
|
136
|
use Mail::SpamAssassin; |
|
22
|
|
|
|
|
63
|
|
|
22
|
|
|
|
|
615
|
|
41
|
22
|
|
|
22
|
|
140
|
use Mail::SpamAssassin::PerMsgStatus; |
|
22
|
|
|
|
|
55
|
|
|
22
|
|
|
|
|
686
|
|
42
|
22
|
|
|
22
|
|
117
|
use Mail::SpamAssassin::Logger; |
|
22
|
|
|
|
|
67
|
|
|
22
|
|
|
|
|
1828
|
|
43
|
22
|
|
|
22
|
|
144
|
use Mail::SpamAssassin::Util qw(untaint_var); |
|
22
|
|
|
|
|
56
|
|
|
22
|
|
|
|
|
15937
|
|
44
|
|
|
|
|
|
|
|
45
|
|
|
|
|
|
|
our @ISA = qw(); |
46
|
|
|
|
|
|
|
|
47
|
|
|
|
|
|
|
########################################################################### |
48
|
|
|
|
|
|
|
|
49
|
|
|
|
|
|
|
sub new { |
50
|
64
|
|
|
64
|
0
|
182
|
my $class = shift; |
51
|
64
|
|
33
|
|
|
401
|
$class = ref($class) || $class; |
52
|
|
|
|
|
|
|
|
53
|
64
|
|
|
|
|
178
|
my ($main) = @_; |
54
|
|
|
|
|
|
|
my $self = { |
55
|
|
|
|
|
|
|
'main' => $main, |
56
|
|
|
|
|
|
|
'conf' => $main->{conf}, |
57
|
64
|
|
|
|
|
348
|
'use_ignores' => 1, |
58
|
|
|
|
|
|
|
}; |
59
|
64
|
|
|
|
|
202
|
bless ($self, $class); |
60
|
|
|
|
|
|
|
|
61
|
64
|
|
|
|
|
423
|
$self->{main}->call_plugins("learner_new"); |
62
|
64
|
|
|
|
|
342
|
$self; |
63
|
|
|
|
|
|
|
} |
64
|
|
|
|
|
|
|
|
65
|
|
|
|
|
|
|
########################################################################### |
66
|
|
|
|
|
|
|
|
67
|
|
|
|
|
|
|
sub finish { |
68
|
40
|
|
|
40
|
0
|
105
|
my $self = shift; |
69
|
|
|
|
|
|
|
# we don't need to do the plugin; Mail::SpamAssassin::finish() does |
70
|
|
|
|
|
|
|
# that for us |
71
|
40
|
|
|
|
|
94
|
%{$self} = (); |
|
40
|
|
|
|
|
125
|
|
72
|
|
|
|
|
|
|
} |
73
|
|
|
|
|
|
|
|
74
|
|
|
|
|
|
|
########################################################################### |
75
|
|
|
|
|
|
|
|
76
|
|
|
|
|
|
|
# force the Bayes dbs to be closed, if they haven't already been; called |
77
|
|
|
|
|
|
|
# at the end of scan operation, or when switching between user IDs, |
78
|
|
|
|
|
|
|
# or when C<Mail::SpamAssassin::finish_learner()> is called. |
79
|
|
|
|
|
|
|
# |
80
|
|
|
|
|
|
|
sub force_close { |
81
|
6
|
|
|
6
|
0
|
13
|
my $self = shift; |
82
|
6
|
|
|
|
|
9
|
my $quiet = shift; |
83
|
6
|
|
|
|
|
19
|
$self->{main}->call_plugins("learner_close", { quiet => $quiet }); |
84
|
|
|
|
|
|
|
} |
85
|
|
|
|
|
|
|
|
86
|
|
|
|
|
|
|
########################################################################### |
87
|
|
|
|
|
|
|
|
88
|
|
|
|
|
|
|
sub ignore_message { |
89
|
14
|
|
|
14
|
0
|
50
|
my ($self,$PMS) = @_; |
90
|
|
|
|
|
|
|
|
91
|
14
|
50
|
|
|
|
46
|
return 0 unless $self->{use_ignores}; |
92
|
|
|
|
|
|
|
|
93
|
14
|
|
|
|
|
127
|
my $ig_from = $self->{main}->call_plugins ("check_wb_list", |
94
|
|
|
|
|
|
|
{ permsgstatus => $PMS, type => 'from', list => 'bayes_ignore_from' }); |
95
|
14
|
|
|
|
|
107
|
my $ig_to = $self->{main}->call_plugins ("check_wb_list", |
96
|
|
|
|
|
|
|
{ permsgstatus => $PMS, type => 'to', list => 'bayes_ignore_to' }); |
97
|
|
|
|
|
|
|
|
98
|
14
|
|
33
|
|
|
126
|
my $ignore = $ig_from || $ig_to; |
99
|
14
|
50
|
|
|
|
50
|
dbg("bayes: not using bayes, bayes_ignore_from or _to rule") if $ignore; |
100
|
14
|
|
|
|
|
46
|
return $ignore; |
101
|
|
|
|
|
|
|
} |
102
|
|
|
|
|
|
|
|
103
|
|
|
|
|
|
|
########################################################################### |
104
|
|
|
|
|
|
|
|
105
|
|
|
|
|
|
|
sub learn { |
106
|
10
|
|
|
10
|
0
|
2112
|
my ($self, $isspam, $msg, $id) = @_; |
107
|
10
|
50
|
|
|
|
55
|
return unless $self->{conf}->{use_learner}; |
108
|
10
|
50
|
|
|
|
33
|
return unless defined $msg; |
109
|
|
|
|
|
|
|
|
110
|
10
|
50
|
|
|
|
36
|
if( $self->{use_ignores} ) # Remove test when PerMsgStatus available. |
111
|
|
|
|
|
|
|
{ |
112
|
|
|
|
|
|
|
# DMK, koppel@ece.lsu.edu: Hoping that the ultimate fix to bug 2263 will |
113
|
|
|
|
|
|
|
# make it unnecessary to construct a PerMsgStatus here. |
114
|
10
|
|
|
|
|
90
|
my $PMS = new Mail::SpamAssassin::PerMsgStatus $self->{main}, $msg; |
115
|
10
|
|
|
|
|
56
|
my $ignore = $self->ignore_message($PMS); |
116
|
10
|
|
|
|
|
77
|
$PMS->finish(); |
117
|
10
|
50
|
|
|
|
68
|
return 0 if $ignore; |
118
|
|
|
|
|
|
|
} |
119
|
|
|
|
|
|
|
|
120
|
10
|
|
|
|
|
81
|
return $self->{main}->call_plugins("learn_message", { isspam => $isspam, msg => $msg, id => $id }); |
121
|
|
|
|
|
|
|
} |
122
|
|
|
|
|
|
|
|
123
|
|
|
|
|
|
|
########################################################################### |
124
|
|
|
|
|
|
|
|
125
|
|
|
|
|
|
|
sub forget { |
126
|
4
|
|
|
4
|
0
|
16
|
my ($self, $msg, $id) = @_; |
127
|
4
|
50
|
|
|
|
21
|
return unless $self->{conf}->{use_learner}; |
128
|
4
|
50
|
|
|
|
14
|
return unless defined $msg; |
129
|
4
|
|
|
|
|
26
|
return $self->{main}->call_plugins("forget_message", { msg => $msg, id => $id }); |
130
|
|
|
|
|
|
|
} |
131
|
|
|
|
|
|
|
|
132
|
|
|
|
|
|
|
########################################################################### |
133
|
|
|
|
|
|
|
|
134
|
|
|
|
|
|
|
sub sync { |
135
|
2
|
|
|
2
|
0
|
7
|
my ($self, $sync, $expire, $opts) = @_; |
136
|
2
|
50
|
|
|
|
25
|
return 0 unless $self->{conf}->{use_learner}; |
137
|
|
|
|
|
|
|
|
138
|
2
|
50
|
|
|
|
11
|
if ($sync) { |
139
|
2
|
|
|
|
|
10
|
$self->{main}->call_plugins("learner_sync", $opts ); |
140
|
|
|
|
|
|
|
} |
141
|
2
|
50
|
|
|
|
12
|
if ($expire) { |
142
|
0
|
|
|
|
|
0
|
$self->{main}->call_plugins("learner_expire_old_training", $opts ); |
143
|
|
|
|
|
|
|
} |
144
|
|
|
|
|
|
|
|
145
|
2
|
|
|
|
|
11
|
return 0; |
146
|
|
|
|
|
|
|
} |
147
|
|
|
|
|
|
|
|
148
|
|
|
|
|
|
|
########################################################################### |
149
|
|
|
|
|
|
|
|
150
|
|
|
|
|
|
|
sub is_scan_available { |
151
|
144
|
|
|
144
|
0
|
333
|
my $self = shift; |
152
|
144
|
50
|
|
|
|
596
|
return 0 unless $self->{conf}->{use_learner}; |
153
|
144
|
|
|
|
|
518
|
return $self->{main}->call_plugins("learner_is_scan_available"); |
154
|
|
|
|
|
|
|
} |
155
|
|
|
|
|
|
|
|
156
|
|
|
|
|
|
|
########################################################################### |
157
|
|
|
|
|
|
|
|
158
|
|
|
|
|
|
|
sub dump_bayes_db { |
159
|
0
|
|
|
0
|
0
|
|
my($self, $magic, $toks, $regex) = @_; |
160
|
0
|
0
|
|
|
|
|
return 0 unless $self->{conf}->{use_learner}; |
161
|
0
|
|
|
|
|
|
return $self->{main}->call_plugins("learner_dump_database", { |
162
|
|
|
|
|
|
|
magic => $magic, toks => $toks, regex => $regex }); |
163
|
|
|
|
|
|
|
} |
164
|
|
|
|
|
|
|
|
165
|
|
|
|
|
|
|
1; |