From: Eric Wong <e@80x24.org>
Date: Sat, 3 Oct 2015 10:18:07 +0000 (+0000)
Subject: mda: support a 'filter=scrub' option for external lists
X-Git-Tag: v1.0.0~797
X-Git-Url: http://www.git.stargrave.org/?p=public-inbox.git;a=commitdiff_plain;h=9bc8cf2e9866393470cdb76a2af39d63f4246204

mda: support a 'filter=scrub' option for external lists

For list where we are not the primary archival entry point,
defaulting to filter=scrub makes sense since their list
conventions may be more tolerant of HTML and other crap
than we are.
---

diff --git a/lib/PublicInbox/Config.pm b/lib/PublicInbox/Config.pm
index 315d788b..0d73a867 100644
--- a/lib/PublicInbox/Config.pm
+++ b/lib/PublicInbox/Config.pm
@@ -61,7 +61,7 @@ sub lookup {
 	defined $pfx or return;
 
 	my %rv;
-	foreach my $k (qw(mainrepo address)) {
+	foreach my $k (qw(mainrepo address filter)) {
 		my $v = $self->{"$pfx.$k"};
 		$rv{$k} = $v if defined $v;
 	}
diff --git a/public-inbox-mda b/public-inbox-mda
index 1a9469ba..df8ca38c 100755
--- a/public-inbox-mda
+++ b/public-inbox-mda
@@ -38,7 +38,18 @@ if (PublicInbox::MDA->precheck($filter, $dst->{address}) &&
 	$filtered = undef;
 	$filter->simple($msg);
 
-	if (PublicInbox::Filter->run($msg, $filter)) {
+	my $filter_arg;
+	my $fcfg = $dst->{filter};
+	if (!defined $fcfg || $filter eq 'reject') {
+		$filter_arg = $filter;
+	} elsif ($fcfg eq 'scrub') {
+		$filter_arg = undef; # the default for legacy versions
+	} else {
+		warn "publicinbox.$dst->{listname}.filter=$fcfg invalid\n";
+		warn "must be either 'scrub' or 'reject' (the default)\n";
+	}
+
+	if (PublicInbox::Filter->run($msg, $filter_arg)) {
 		# run spamc again on the HTML-free message
 		if (do_spamc($msg, \$filtered)) {
 			$msg = Email::MIME->new(\$filtered);