2 # Copyright (C) 2014-2020 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|rm> </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 $?;
42 sub remove_or_add ($$$$) {
43 my ($ibx, $train, $mime, $addr) = @_;
45 # We do not touch GIT_COMMITTER_* env here so we can track
46 # who trained the message.
47 $ibx->{name} = $ENV{GIT_COMMITTER_NAME} // $ibx->{name};
48 $ibx->{-primary_address} = $ENV{GIT_COMMITTER_EMAIL} // $addr;
49 $ibx = PublicInbox::InboxWritable->new($ibx);
50 my $im = $ibx->importer(0);
53 # This needs to be idempotent, as my inotify trainer
54 # may train for each cross-posted message, and this
55 # script already learns for every list in
56 # ~/.public-inbox/config
57 $im->remove($mime, $train);
58 } elsif ($train eq "ham") {
59 # no checking for spam here, we assume the message has
60 # been reviewed by a human at this point:
61 PublicInbox::MDA->set_list_headers($mime, $ibx);
63 # Ham messages are trained when they're marked into
64 # a SEEN state, so this is idempotent:
70 # spam is removed from all known inboxes since it is often Bcc:-ed
71 if ($train eq 'spam') {
72 $pi_config->each_inbox(sub {
74 $ibx = PublicInbox::InboxWritable->new($ibx);
75 my $im = $ibx->importer(0);
76 $im->remove($mime, 'spam');
80 require PublicInbox::MDA;
83 my %dests; # address => <PublicInbox::Inbox|0(false)>
84 for ($mime->header('Cc'), $mime->header('To')) {
85 foreach my $addr (PublicInbox::Address::emails($_)) {
87 $dests{$addr} //= $pi_config->lookup($addr) // 0;
91 # n.b. message may be cross-posted to multiple public-inboxes
93 while (my ($addr, $ibx) = each %dests) {
94 next unless ref($ibx); # $ibx may be 0
95 next if $seen{"$ibx"}++;
96 remove_or_add($ibx, $train, $mime, $addr);
98 my $dests = PublicInbox::MDA->inboxes_for_list_id($pi_config, $mime);
99 for my $ibx (@$dests) {
100 next if !$seen{"$ibx"}++;
101 remove_or_add($ibx, $train, $mime, $ibx->{-primary_address});