#!/usr/local/bin/perl -w
# $Id: spam_filter.pm,v 1.1 2006/08/23 04:47:10 candy Exp candy $

use Socket; # inet_ntoa
use strict;

sub resolve($) {
	my ($host) = @_;
	my ($name, $aliases, $addrtype, $length, @addrs) = gethostbyname($host);
	my $remote = "";
	if (defined($addrtype) && $addrtype == 2) {
		$remote = inet_ntoa($addrs[0]);
	}
	return $remote;
}

#
sub rbl_match($$$) {
	my ($rbl, $ip, $host) = @_;
	my $len = length($rbl);
	my $match = 0;
	if (defined($ip)) {
		$match ||= ($rbl eq $ip)
			|| ($rbl =~ /\.$/ && substr($ip, 0, $len) eq $rbl)
	}
	if (defined($host)) {
		$host =~ tr/A-Z/a-z/;
		$rbl =~ tr/A-Z/a-z/;
		$match ||= ($rbl eq $host)
			|| ($rbl =~ /^\./ && substr($host, length($host) - $len, $len) eq $rbl);
	}
	return $match;
}

sub get_msg_links($) {
	my ($msg) = @_;
	my @list = ();
	while ($msg =~ /http:\/\/([a-zA-Z0-9\.\/\-+#_?~&%=^\@:;]+)/) {
		my $host = $1;
		$msg = $';
		$host =~ s/\/.*//;
		push(@list, $host);
	}
	return @list;
}

# URL based spam filter
# usage:
# ($block, $reason) = spam_filter($msg, $url, $remote_ip, $remote_host);
# if ($block) {
#	printf(STDERR "client %s(%s) denid by %s\n", $remore_ip, $remote_host, $reason);
# }
#
sub spam_filter($$$$) {
	my ($msg, $url, $ip, $host) = @_;
	my $block = 0;
	my $reason = "";
	my @hosts = get_msg_links($msg);
	if (defined($url) && $url ne "") {
		my $h = $url;
		$h =~ s/http:\/\///;
		$h =~ s/\/.*//;
		push(@hosts, $h);
	}
	if (open(RBL, ".htrbl")) {
		my $lbuf;
		while ($block == 0 && ($lbuf = <RBL>)) {
			chomp($lbuf);
			$lbuf =~ s/#.*//;
			$lbuf =~ s/[ \t]*$//;
			if ($lbuf ne "") {
				if (rbl_match($lbuf, $ip, $host)) {
					$reason = "rbl " . $lbuf;
					$block = 1;
				}
				my $i = 0;
				while ($block == 0 && $i < $#hosts + 1) {
					my $link_host = $hosts[$i++];
					my $link_ip = resolve($link_host);
					if (rbl_match($lbuf, $link_ip, $link_host)) {
						$reason = "bad link to " . $link_ip . "(" . $link_host . ")";
						$block = 2;
					}
				}
			}
		}
		close(RBL);
	}
	if ($block == 0 && $msg =~
	/sex/
	) {
		$reason = "contents filter";
		$block = 1;
	}
	return ($block, $reason);
}

# my ($msg, $url, $remote_ip, $remote_host) = ($ARGV[0], $ARGV[1], $ARGV[2], $ARGV[3]);
# printf("%s %s %s %s\n", $msg, $url, $remote_ip, $remote_host);
# my ($block, $reason) = spam_filter($msg, $url, $remote_ip, $remote_host);
#if ($block) {
#	printf(STDERR "client %s(%s) %s by %s\n", $ARGV[2], $ARGV[3], "denied", $reason);
#}

1;