#!/usr/bin/php
<?php

define("RSS_INTERVAL", 300); // seconds between downloading RSS changelog
define("IGNORE_TIME", 60*15); // seconds we don't repeat links

define('TRACURL', 'trac.kohanaphp.com');

define("ircBotPath", str_replace("\\", "/", dirName(__FILE__)));
include(ircBotPath."/lib/ircBot.class.php");

include('lastRSS.inc.php');

if (version_compare(phpversion(),'4.3.0','<') || !defined("STDIN")) {
        define('STDIN',fopen("php://stdin","r"));
        define('STDOUT',fopen("php://stdout","r"));
        define('STDERR',fopen("php://stderr","r"));
        register_shutdown_function( create_function( '' , 'fclose(STDIN); fclose(STDOUT); fclose(STDERR); return true;' ) );
}


/************************  Configuration  ************************/

ircConf::set("botNick", "KohanaTrac");       /** Bot nick */
ircConf::set("botName", "Kohana Trac Bot");  /** Bot's IRC name */
ircConf::set("botIdent", "TracBot");      /** Bot's ident */

/** NickServ password, comment out if you don't have one.
  NOTE: this is also used when the bot NickName is taken to try ghost it
  with NickServ, only when set */
//ircConf::set("botNickServPass", "secret");

/** Servers to connect to - MUST BE ARRAY()!!! Used for rotation */
ircConf::set("servers", array("irc.freenode.net"));
ircConf::set("serverPort", 6667); /** Server port */
ircConf::set("stayAlive", true); /** Stay-Alive - Keep trying connections */

/** Some networks and people don't like bots */
ircConf::set("versionReply", "greg's fantastic trac bot");

/** Bot channel configuration, this is a new requirement, channel users need to
be configured for each channel, users list MUST be an array */
ircConf::setChannel("#kohana-dev", "+ntR", array('keyword'=>'secret'), array());
ircConf::setChannel("#kohana", "+ntR", array(), array());
ircConf::setChannel("#kohanatest", "+ntR", array(), array());

/** Log/Debug, when TRUE messages will written to the logfile */
define("ircDoLog", true);
/** Log raw IRC output when in debug mode */
define("ircDebug", false);

$irc = new ircBot;

/** ADD CALLBACKS
 * This is where the bot gets its functionality.
 * See comments & code below for writing callback functions */
$irc->addCallback(onChanMsg,   "channelResponses");


$irc->addCallback(onChanMsg,   "onMsgCheck");
$irc->addCallback(onOp,        "onOpCheck");
//$irc->addCallback(onMode,      "onModeCheck");
$irc->addCallback(onTopic,     "onTopicCheck");
$irc->addCallback(onNotice,    "onNoticeCheck");
$irc->addCallback(onPrivMsg,   "onPrivMsgCheck");

$irc->addCallback(onJoin,      "onEventCheck");
$irc->addCallback(onKick,      "onEventCheck");
$irc->addCallback(onPart,      "onEventCheck");
$irc->addCallback(onQuit,      "onEventCheck");
$irc->addCallback(onKill,      "onEventCheck");


/** Finally start our bot and connect to IRC :) */
$irc->init();

function channelResponses(&$bot, $nick, $ident, $host, $channel, $text) {
        // ignore CIA bot
        if (preg_match('/^CIA-\d+/', $nick)) return;


        if (preg_match("/^".$GLOBALS["ircConf"]["botNick"].".*suck/", $text)) {
                // someone said "KohanaTrac sucks"
                $bot->provMsg($channel, 'No, '.$nick.', you suck');
                return;
        }


        $direct_msg = false;
        if (preg_match("/^".$GLOBALS["ircConf"]["botNick"]."[:,\- ]/", $text)) {
                // this msg was directed at the bot
                $direct_msg = true;
                // replace our name (interferes with wiki links)
                $text = str_replace($GLOBALS["ircConf"]["botNick"],"",$text);
        }

        if (preg_match('/(?<!\w)r(\d+)(?!\w)/', $text, $matches) ||
          preg_match('/(?<!\w)\[(\d+)\](?!\w)/', $text, $matches)) {
                $bot->log('Got possibly changeset link: '.$matches[0]);
                if (false !== ($submatch = scrapeUrl('http://'.TRACURL.'/changeset/'.$matches[1], '/id="searchable"><p>(.*)<\/p>/ms'))) {
                        // have to call html_entity_decode() twice beacuse it puts stuff like &amp;#34;
                        $desc = html_entity_decode(strip_tags(html_entity_decode($submatch[1])));
                        // get rid of extra spaces, and newlines
                        $desc = trim( str_replace(array("\n","\r"), array("  ",""), $desc) );
                        if (strlen($desc) > 60) {
                                // trim if it's over 60, and show we trimmed it
                                $desc = substr($desc,0,60).'...';
                        }
                        if (!checkIgnore($matches[0]) || $direct_msg) {
                                $bot->privMsg($channel, 'Revision '.$matches[1].' at http://'.TRACURL.'/changeset/'.$matches[1].': '.$desc);
                        }
                }
        }
        if (preg_match('/(?<![\w\/])([A-Z][a-z]+)([A-Z0-9][a-z0-9]*)+/', $text, $matches)) {
                $bot->log('Got possibly wiki link: '.$matches[0].' checking http://'.TRACURL.'/wiki/'.$matches[0]." = ".checkUrl('http://'.TRACURL.'/wiki/'.$matches[0]));
                if (checkUrl('http://'.TRACURL.'/wiki/'.$matches[0]) == 200) {
                        if (!checkIgnore($matches[0]) || $direct_msg) {
                                $bot->privMsg($channel, 'Wiki page '.$matches[0].' at http://'.TRACURL.'/wiki/'.$matches[0]);
                        }
                }
        }
        if (preg_match('/(?<!\w)(?:#|bug |ticket )(\d{1,4})(?!\w)/', $text, $matches)) {
                $bot->log('Got possibly ticket link: '.$matches[0]);
                if (false !== ($submatch = scrapeUrl('http://'.TRACURL.'/ticket/'.$matches[1],
                  //'/<h2 class="summary">(.*)<\/h2>.*<h3 class="status">(.*)<\/h3>/msU') )) {
                  '/<span class="status">(.*)<\/span>.*<h2 class="summary">(.*)<\/h2>/msU') )) {
                        $bot->log("Its a link.. posting maybe");
                        $status = html_entity_decode(strip_tags(html_entity_decode($submatch[1])));
                        if (!checkIgnore($matches[0]) || $direct_msg) {
                                $bot->log("Not ignoring, posting for sure");
                                $bot->privMsg($channel, 'Ticket '.$matches[1].' at http://'.TRACURL.'/ticket/'.$matches[1].': '.$submatch[2].' '.$status);
                        }
                }
        }
}

function onMsgCheck(&$bot, $nick, $ident, $host, $channel, $text) {
        checkRss($bot);
}
function onOpCheck(&$bot, $nick, $ident, $host, $channel, $user) {
        checkRss($bot);
}
function onEventCheck(&$bot, $nick, $ident, $host, $channelOrQuit = null) {
        checkRss($bot);
}
function onTopicCheck(&$bot, $nick, $ident, $host, $channel, $topic) {
        checkRss($bot);
}
function onNoticeCheck(&$bot, $nick, $ident, $host, $dest, $text) {
        checkRss($bot);
}
function onPrivMsgCheck(&$bot, $nick, $ident, $host, $us, $text) {
        checkRss($bot);
}

function checkRss(&$bot) {
        static $lastchecked;
        static $lastitem;

        if (!isset($lastchecked) || ($lastchecked < time() - RSS_INTERVAL)) {
                $bot->log('Checking RSS feed..');
                $rss = new lastRSS;
                $rss->cache_dir = '';
                $rss->cp = 'US-ASCII';
                //$rssurl = 'http://'.TRACURL.'/timeline?milestone=on&ticket=on&changeset=on&wiki=on&max=50&daysback=1&format=rss';
                $rssurl = 'http://'.TRACURL.'/timeline?milestone=on&ticket=on&changeset=on&wiki=on&max=50&daysback=1&format=rss';
                if ($rs = $rss->get($rssurl)) {
                        if (!isset($lastitem)) {
                                $lastitem = (isset($rs['items'][0]) ? md5($rs['items'][0]['title']) : '');
                        } else {
                                $messages = array();

                                foreach ($rs['items'] as $item) {
                                        if (md5($item['title']) == $lastitem) {
                                                // don't show any more
                                                break;
                                        }

                                        $title = preg_replace(
                                                array(
                                                        '/(?<!\w)r(\d+)(?!\w)/',
                                                        '/(?<!\w)\[(\d+)\](?!\w)/',
                                                        '/(?<!\w|&)#(\d+)(?!\w)/',
                                                ),

                                                array(
                                                        'http://'.TRACURL.'/changeset/$1',
                                                        'http://'.TRACURL.'/changeset/$1',
                                                        'http://'.TRACURL.'/ticket/$1',
                                                ),
                                                $item['title']);
                                        $title = preg_replace_callback('/([A-Z][a-z]+)([A-Z0-9][a-z0-9]+)*/', 'replace_wiki_url', $title);

                                        if (substr($title,0,9) == 'Changeset') {
                                                // changeset titles include a description
                                                $messages[] = $title.' ('.$item['pubDate'].')';
                                        } else {
                                                // have to call html_entity_decode() twice beacuse it puts stuff like &amp;#34;
                                                $desc = html_entity_decode(strip_tags(html_entity_decode($item['description'])));
                                                // get rid of extra spaces, and newlines
                                                $desc = trim( str_replace(array("\n","\r"), array("  ",""), $desc) );
                                                if (strlen($desc) > 60) {
                                                        // trim if it's over 60, and show we trimmed it
                                                        $desc = substr($desc,0,60).'...';
                                                }

                                                $messages[] = $title.': '.$desc.' ('.$item['pubDate'].')';
                                        }
                                }
                                if (count($messages) > 0) {
                                        // send them backwards, because we processed the rss forwards
                                        foreach (array_reverse($messages) as $message) {
                                                $bot->privMsg('#kohana-dev', $message);
                                        }
                                }

                                // set the lastitem to the first entry now
                                $lastitem = md5($rs['items'][0]['title']);
                        }
                } else {
                        // set $lastitem to something, so when we do get something, it will print it
                        $lastitem = "nadda";
                }
                $lastchecked = time();
        }
}

function checkUrl($url) {
        $u = parse_url($url);
        if (empty($u['port'])) {
                switch($u['scheme']) {
                        case 'http': $u['port'] = 80; break;
                        case 'https': $u['port'] = 443; break;
                        default: return false;
                }
        }
        if (!($s = fsockopen($u['host'],$u['port']))) {
                return false;
        }
        $request = "HEAD ".$u['path']." HTTP/1.1\r\n";
        $request .= "Host: ".$u['host']."\r\n";
        $request .= "Connection: Close\r\n\r\n";
        fwrite($s, $request);
        $data = '';
        while (!feof($s)) {
                $data .= fgets($s, 1024);
        }
        fclose($s);
        if (preg_match('/HTTP\/1.[01] (\d+)/', $data, $matches)) {
                return $matches[1];
        }
        return false;
}

function replace_wiki_url($matches) {
        if (checkUrl('http://'.TRACURL.'/wiki/'.$matches[0]) == 200) {
                return 'http://'.TRACURL.'/wiki/'.$matches[0];
        } else {
                return $matches[0];
        }
}


function scrapeUrl($url, $regex) {
        $u = parse_url($url);
        if (empty($u['port'])) {
                switch($u['scheme']) {
                        case 'http': $u['port'] = 80; break;
                        case 'https': $u['port'] = 443; break;
                        default: return false;
                }
        }
        if (!($s = fsockopen($u['host'],$u['port']))) {
                return false;
        }
        $request = "GET ".$u['path']." HTTP/1.1\r\n";
        $request .= "Host: ".$u['host']."\r\n";
        $request .= "Connection: Close\r\n\r\n";
        fwrite($s, $request);
        $data = '';
        while (!feof($s)) {
                $data .= fgets($s, 1024);
        }
        fclose($s);
        if (preg_match($regex, $data, $matches)) {
                return $matches;
        }
        return false;
}


// return true if we said this link within IGNORE_TIME seconds
function checkIgnore($link) {
        global $irc;

        static $ignoring;
        if (!isset($ignoring)) $ignoring = array();

        foreach ($ignoring as $key=>$time) {
                if ($time < time() - IGNORE_TIME) {
                        $irc->log("No longer ignoring \"$key\"");
                        unset($ignoring[$key]);
                }
        }

        if (!isset($ignoring[$link])) {
                $ignoring[$link] = time();
                $irc->log("Now ignoring \"$link\" for ".IGNORE_TIME."s");
                return false;
        }
        $irc->log("Currently ignoring \"$link\"");
        return true;
}