From cc38809764c63dbed24300cdcf68d9f667f0dfad Mon Sep 17 00:00:00 2001 From: Evan Prodromou Date: Mon, 16 Feb 2009 16:34:07 -0500 Subject: First version of blogspam.net plugin Added a plugin for blogspam.net. The service is kind of aggressive, so by default I turn a lot of the tests off. But worth a look. --- plugins/BlogspamNetPlugin.php | 144 ++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 144 insertions(+) create mode 100644 plugins/BlogspamNetPlugin.php diff --git a/plugins/BlogspamNetPlugin.php b/plugins/BlogspamNetPlugin.php new file mode 100644 index 000000000..d9372bcd5 --- /dev/null +++ b/plugins/BlogspamNetPlugin.php @@ -0,0 +1,144 @@ +. + * + * @category Plugin + * @package Laconica + * @author Evan Prodromou + * @copyright 2009 Control Yourself, Inc. + * @license http://www.fsf.org/licensing/licenses/agpl-3.0.html GNU Affero General Public License version 3.0 + * @link http://laconi.ca/ + */ + +if (!defined('LACONICA')) { + exit(1); +} + +define('BLOGSPAMNETPLUGIN_VERSION', '0.1'); + +/** + * Plugin to check submitted notices with blogspam.net + * + * When new notices are saved, we check their text with blogspam.net (or + * a compatible service). + * + * Blogspam.net is supposed to catch blog comment spam, and I found that + * some of its tests (min/max size, bayesian match) gave a lot of false positives. + * So, I've turned those tests off by default. This may not get as many + * hits, but it's better than nothing. + * + * @category Plugin + * @package Laconica + * @author Evan Prodromou + * @license http://www.fsf.org/licensing/licenses/agpl-3.0.html GNU Affero General Public License version 3.0 + * @link http://laconi.ca/ + * + * @see Event + */ + +class BlogspamNetPlugin extends Plugin +{ + var $baseUrl = 'http://test.blogspam.net:8888/'; + + function __construct($url=null) + { + parent::__construct(); + if ($url) { + $this->baseUrl = $url; + } + } + + function onStartNoticeSave($notice) + { + $args = $this->testArgs($notice); + common_debug("Blogspamnet args = " . print_r($args, TRUE)); + $request = xmlrpc_encode_request('testComment', array($args)); + $context = stream_context_create(array('http' => array('method' => "POST", + 'header' => + "Content-Type: text/xml\r\n". + "User-Agent: " . $this->userAgent(), + 'content' => $request))); + $file = file_get_contents($this->baseUrl, false, $context); + $response = xmlrpc_decode($file); + if (xmlrpc_is_fault($response)) { + throw new ServerException("$response[faultString] ($response[faultCode])", 500); + } else { + common_debug("Blogspamnet results = " . $response); + if (preg_match('/^ERROR(:(.*))?$/', $response, $match)) { + throw new ServerException(sprintf(_("Error from %s: %s"), $this->baseUrl, $match[2]), 500); + } else if (preg_match('/^SPAM(:(.*))?$/', $response, $match)) { + throw new ClientException(sprintf(_("Spam checker results: %s"), $match[2]), 400); + } else if (preg_match('/^OK$/', $response)) { + // don't do anything + } else { + throw new ServerException(sprintf(_("Unexpected response from %s: %s"), $this->baseUrl, $response), 500); + } + } + return true; + } + + function testArgs($notice) + { + $args = array(); + $args['comment'] = $notice->content; + $args['ip'] = $this->getClientIP(); + + if (isset($_SERVER) && array_key_exists('HTTP_USER_AGENT', $_SERVER)) { + $args['agent'] = $_SERVER['HTTP_USER_AGENT']; + } + + $profile = $notice->getProfile(); + + if ($profile && $profile->homepage) { + $args['link'] = $profile->homepage; + } + + if ($profile && $profile->fullname) { + $args['name'] = $profile->fullname; + } else { + $args['name'] = $profile->nickname; + } + + $args['site'] = common_root_url(); + $args['version'] = $this->userAgent(); + + $args['options'] = "max-size=140,min-size=0,min-words=0,exclude=bayasian"; + + return $args; + } + + function getClientIP() + { + if (isset($_SERVER) && array_key_exists('REQUEST_METHOD', $_SERVER)) { + // Note: order matters here; use proxy-forwarded stuff first + foreach (array('HTTP_X_FORWARDED_FOR', 'CLIENT-IP', 'REMOTE_ADDR') as $k) { + if (isset($_SERVER[$k])) { + return $_SERVER[$k]; + } + } + } + return '127.0.0.1'; + } + + function userAgent() + { + return 'BlogspamNetPlugin/'.BLOGSPAMNETPLUGIN_VERSION . ' Laconica/' . LACONICA_VERSION; + } +} -- cgit v1.2.3-54-g00ecf