I have the following scrip condition for my one of my queues, to
prevent auto-replies to mail declared spam by spam assassin:
–cut here–
if ($self->TransactionObj->Type eq ‘Create’) {
$RT::Logger->error(“Got a create transacation…”);
my $co = $self->TransactionObj->ContentObj;
return 0 unless $co;
$RT::Logger->error("…checking if via email.");
return 1 unless $co->GetHeader(‘Received’);
$RT::Logger->error(“Looks like a juicy email”);
check if email is flagged as spam. assumes header set by mail
server
my $is_spam = $co->GetHeader(‘X-Spam-Flag’) || ‘nope’;
my $rv = $is_spam eq ‘YES’ ? 0 : 1;
$RT::Logger->error(“spam determination = $is_spam (autoreply =
$rv)”);
return $rv;
} else {
return 0; # do not run action
}
–cut here–
Is there some other (faster) preferred way to pull just the headers?
I have one message I’m trying to debug why we sent an autoreply to a
message that clearly was declared spam. It got to the step “checking
if via email” in my scrip, then seems to have returned true when it
didn’t find a Received header.
I’m wondering if it is just timing out (no error in the log) since it
took over 30 seconds to get the content object:
Feb 19 11:32:49 rt RT: Got a create transacation… ((eval 15255):2)
Feb 19 11:33:21 rt RT: …checking if via email. ((eval 15255):5)
Actually, quite often I get my scrip terminate at this step, even
though we almost always get tickets via email instead of at the UI.
Any ideas?