2 # Copyright (C) 2014-2018 all contributors <meta@public-inbox.org>
3 # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
5 # Used for training spam (via SpamAssassin) and removing messages from a
7 my $usage = "$0 (spam|ham) < /path/to/message";
10 use PublicInbox::Config;
11 use PublicInbox::InboxWritable;
12 use PublicInbox::MIME;
13 use PublicInbox::Address;
14 use PublicInbox::Spamcheck::Spamc;
15 my $train = shift or die "usage: $usage\n";
16 if ($train !~ /\A(?:ham|spam|rm)\z/) {
17 die "`$train' not recognized.\nusage: $usage\n";
20 my $spamc = PublicInbox::Spamcheck::Spamc->new;
21 my $pi_config = PublicInbox::Config->new;
23 my $mime = PublicInbox::MIME->new(eval {
25 my $data = scalar <STDIN>;
26 $data =~ s/\A[\r\n]*From [^\r\n]*\r?\n//s;
30 if ($train eq 'ham') {
31 $spamc->hamlearn(\$data);
32 } elsif ($train eq 'spam') {
33 $spamc->spamlearn(\$data);
35 die "spamc failed with: $?\n" if $?;
44 foreach my $h (qw(Cc To)) {
45 my $val = $mime->header($h) or next;
46 foreach my $email (PublicInbox::Address::emails($val)) {
47 $dests{lc($email)} = 1;
51 if ($train eq 'spam') {
52 $pi_config->each_inbox(sub {
54 $ibx = PublicInbox::InboxWritable->new($ibx);
55 my $im = $ibx->importer(0);
56 $im->remove($mime, 'spam');
61 require PublicInbox::MDA if $train eq "ham";
63 # n.b. message may be cross-posted to multiple public-inboxes
64 foreach my $recipient (keys %dests) {
65 my $dst = $pi_config->lookup($recipient) or next;
66 # We do not touch GIT_COMMITTER_* env here so we can track
67 # who trained the message.
68 $dst->{name} = $ENV{GIT_COMMITTER_NAME} || $dst->{name};
69 $dst->{-primary_address} = $ENV{GIT_COMMITTER_EMAIL} || $recipient;
70 $dst = PublicInbox::InboxWritable->new($dst);
71 my $im = $dst->importer(0);
73 if ($train eq "spam" || $train eq "rm") {
74 # This needs to be idempotent, as my inotify trainer
75 # may train for each cross-posted message, and this
76 # script already learns for every list in
77 # ~/.public-inbox/config
78 $im->remove($mime, $train);
79 } else { # $train eq "ham"
80 # no checking for spam here, we assume the message has
81 # been reviewed by a human at this point:
82 PublicInbox::MDA->set_list_headers($mime, $dst);
84 # Ham messages are trained when they're marked into
85 # a SEEN state, so this is idempotent: