#!/usr/bin/perl

use utf8;
use Irssi;
use 5.10.0;
use vars qw($VERSION %IRSSI);
use Encode qw/encode decode/;
use HTML::Entities;
use LWP::Simple;
use LWP::UserAgent;
# use warnings;
# use strict;

# print join ("\n",@INC);

$VERSION = '0.01';
%IRSSI = (
		authors	=> '',
		contact	=> '',
		name	=> 'pagetitle',
		description	=> 'head website title in #ubuntu-cn.',
		license	=> '',
);

our $domain_suffix;
open FILE, "<", "/home/alvin/.irssi/scripts/domain_suffix.txt";
$domain_suffix = <FILE>;
chomp $domain_suffix;
while (<FILE>) {
	chomp $_;
	$domain_suffix .= "|$_";
}
close FILE;

our $time_stamp;
our $flag;
our @nick_array; # store: "nick time" seperated with a space
our $last;


$time_stamp = time - 10;
$flag = 0;

our $ua;
our $bot = 1;	# bot online, 1; else 0
$ua = LWP::UserAgent->new;
$ua->agent('Mozilla/5.0 (X11; Linux i686; rv:12.0) Gecko/20100101 Firefox/12.0 Iceweasel/12.0');
$ua->timeout(5);
$ua->max_size(10240/2);  # some sitei's title is after 1024bytes

sub check_bot {
	my $server;
	eval {$server = Irssi::server_find_tag("FREENODE") };
	return if $server->{tag} !~ /freenode/i;
	my $chan = $server->channel_find("#ubuntu-cn");
	eval {
		$bot = 0;
		foreach ($chan->nicks()) {
			if ($_->{host} ~~ /unaffiliated\/kves$/) {
				$bot = 1;
				last;
			}
		}
	};
	return $bot;
}

sub delete_flood {
	my $t;
	my @tmp = @nick_array;
	@nick_array = undef;
	foreach (@tmp) {
		$_ =~ / (\S+)$/;
		$t = $1;
		push @nick_array, $_ if (time - $t <= 90);
	}
}

sub check_flood {
	my ($nick) = @_;
	my $count;
	my ($n, $t);	#nick, time
	my $rt;

	$rt = 0;
	$count = 0;
	foreach (@nick_array) {
		$_ =~ /^(\S+) (\S+)$/;
		$n = $1;
		$t = $2;
		$count++ if (($n eq $nick) and (time - $t <= 90));
	}
	$rt = 1 if ($count >= 4);

	return $rt;
}

sub fetch_header {
	Irssi::signal_continue(@_);
	my ($server, $msg, $nick, $address, $target) = @_;
	my $url;
	my $base_url;	#true base url
	my $base_url2;	#maybe fake, like url shorter
	my $title;
	my $meta;
	my $enc;
	my $resp;
	my $content;
	my @escape;
	my $tmp;

	return if $server->{tag} !~ /freenode/i;

	if ($nick =~ /(\^k\^|\[ub\]|kk|sevk)/i) {
		# print "Fetch Title:    ^k^'s statement";
		return;
	}
	return if ($target !~ '#ubuntu-cn');

	{
		$msg =~ /([a-zA-Z0-9][\x21-\x7E]*?(\.\b($domain_suffix)\b)+[\x21-\x7E]*)/i;
		# $msg =~ /([\x21-\x7E]*?(\.\b($domain_suffix)\b)+[\x21-\x7E]*)/i;
		$url = $1;
		unless (defined $url) {
			$msg =~ /([a-zA-Z0-9]?[\x21-\x7E]*\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}\/[\x21-\x7E]*)/i;
			$url = $1;
			return if $url =~ /[^.]*192\.168\./;
			return if $url =~ /[^.]*127\.0\./;
		}
		return unless (defined $url);
	}
	return if ($url =~ /@.*?\.\b($domain_suffix)\b/i);	# e-mail

	&check_bot;
	return if ($bot == 1 and $url =~ /^http/i);

	$url = "http://" . $url unless ($url ~~ /^http/i or $url ~~ /^ftp/i);
	{
		$url =~ /([^\/.]+?\.\b($domain_suffix)\b[^\/]*)/i;
		$base_url2 = $1;
	}

	push @nick_array, "$nick " . time;
	if (&check_flood($nick) == 1) {
		print "Fetch Title:    flood detected.";
		return;
	}
	&delete_flood;

	my $retry = 3;
	do {
		$resp = $ua->get($url);
		$retry--;
	} while ((not $resp->is_success) and $retry);
	if ($resp->is_success ne 1) {
		print "Fetch Title:    not success    $url    ", $resp->status_line;
		return;
	}

	{
		$resp->base =~ /([^\/.]+?\.\b($domain_suffix)\b[^\/]*)/i;
		$base_url = $1;
		if (not $base_url) {
			$url =~ /([^\/.]+?\.\b($domain_suffix)\b[^\/]*)/i;
			$base_url = $1;
		}
		if ($base_url ne $base_url2) {
			$base_url = "$base_url <~ $base_url2"
		}
	}

	($content, $enc) = split ";", $resp->header("Content-Type");
	if ($content ne 'text/html') {
		return;
	}
	{
		$enc =~ s/.*?=(.*)/$1/;
	}

	{
		$resp->decoded_content(charset => "$enc") =~ m/<title>(.*?)<\/title>/gsmi;
		if ((not defined $1) or $1 ~~ /^\s+$/) {
			# print "Fetch Title:    $url    can't get title";
			return;
		}
		$title = $1;
	}
	$title =~ s/(\n|\r)//gmi;
	$title =~ s/\s+/ /gmi;
	$title =~ s/^\s*(.+?)\s*$/$1/gmi;
	if ($title =~ /\b(porn|sex|pussy|cock|eroti)/img) {
		print "Fetch Title:    forbidden words";
		# $title = "..xxXXxx..";
		return;
	}


	$title = "*HTTPS* " . $title if ($resp->base =~ /^https/i or $url =~ /^https/i);
	@escape = ($title =~ m/(&#?[a-zA-Z0-9]+;)/gm);
	foreach (@escape) {
		$tmp = decode_entities($_);
		eval { $tmp = decode("iso-8859-1", $tmp); };
		$title =~ s/$_/$tmp/g;
	}

	$title = substr($title, 0, 96) . " ..." if (length $title > 100);
print "Fetch Title:    \$bot = $bot    URL: $url    base: $base_url    ENCl: $enc    TITLE: $title";

	return if ($title eq "" or not defined $title);

	if ($flag == 0) {
		$flag = 1;
		if (time - $time_stamp > 120) {
			$server->command("msg $target Title: $title (@ $base_url)");
			$time_stamp = time;
			$last = $title;
		}
		elsif (time - $time_stamp > 10 and $title ne $last) {
			$server->command("msg $target Title: $title (@ $base_url)");
			$time_stamp = time;
			$last = $title;
		}
		$flag = 0;
	}
}

Irssi::signal_add_last('message public', 'fetch_header');
# Irssi::signal_add_last('message join', 'check_bot');
# Irssi::signal_add_last('message part', 'check_bot');
# Irssi::signal_add_last('message quit', 'check_bot');
# Irssi::signal_add_last('message kick', 'check_bot');
# Irssi::signal_add_last('message nick', 'check_bot');
Irssi::signal_add_last('channel sync', 'check_bot');