<?php

/**

 * Zend Framework

 *

 * LICENSE

 *

 * This source file is subject to the new BSD license that is bundled

 * with this package in the file LICENSE.txt.

 * It is also available through the world-wide-web at this URL:

 * http://framework.zend.com/license/new-bsd

 * If you did not receive a copy of the license and are unable to

 * obtain it through the world-wide-web, please send an email

 * to license@zend.com so we can send you a copy immediately.

 *

 * @category   Zend

 * @package    Zend_Feed_Reader

 * @copyright  Copyright (c) 2005-2012 Zend Technologies USA Inc. (http://www.zend.com)

 * @license    http://framework.zend.com/license/new-bsd     New BSD License

 * @version    $Id: Reader.php 25275 2013-03-06 09:55:33Z frosch $

 */

/**

 * @see Zend_Feed

 */

require_once 'Zend/Feed.php';

/**

 * @see Zend_Feed_Reader_Feed_Rss

 */

require_once 'Zend/Feed/Reader/Feed/Rss.php';

/**

 * @see Zend_Feed_Reader_Feed_Atom

 */

require_once 'Zend/Feed/Reader/Feed/Atom.php';

/**

 * @see Zend_Feed_Reader_FeedSet

 */

require_once 'Zend/Feed/Reader/FeedSet.php';

/**

 * @category   Zend

 * @package    Zend_Feed_Reader

 * @copyright  Copyright (c) 2005-2012 Zend Technologies USA Inc. (http://www.zend.com)

 * @license    http://framework.zend.com/license/new-bsd     New BSD License

 */

class Zend_Feed_Reader

{

    /**

     * Namespace constants

     */

    const NAMESPACE_ATOM_03  = 'http://purl.org/atom/ns#';

    const NAMESPACE_ATOM_10  = 'http://www.w3.org/2005/Atom';

    const NAMESPACE_RDF      = 'http://www.w3.org/1999/02/22-rdf-syntax-ns#';

    const NAMESPACE_RSS_090  = 'http://my.netscape.com/rdf/simple/0.9/';

    const NAMESPACE_RSS_10   = 'http://purl.org/rss/1.0/';

    /**

     * Feed type constants

     */

    const TYPE_ANY              = 'any';

    const TYPE_ATOM_03          = 'atom-03';

    const TYPE_ATOM_10          = 'atom-10';

    const TYPE_ATOM_10_ENTRY    = 'atom-10-entry';

    const TYPE_ATOM_ANY         = 'atom';

    const TYPE_RSS_090          = 'rss-090';

    const TYPE_RSS_091          = 'rss-091';

    const TYPE_RSS_091_NETSCAPE = 'rss-091n';

    const TYPE_RSS_091_USERLAND = 'rss-091u';

    const TYPE_RSS_092          = 'rss-092';

    const TYPE_RSS_093          = 'rss-093';

    const TYPE_RSS_094          = 'rss-094';

    const TYPE_RSS_10           = 'rss-10';

    const TYPE_RSS_20           = 'rss-20';

    const TYPE_RSS_ANY          = 'rss';

    /**

     * Cache instance

     *

     * @var Zend_Cache_Core

     */

    protected static $_cache = null;

    /**

     * HTTP client object to use for retrieving feeds

     *

     * @var Zend_Http_Client

     */

    protected static $_httpClient = null;

    /**

     * Override HTTP PUT and DELETE request methods?

     *

     * @var boolean

     */

    protected static $_httpMethodOverride = false;

    protected static $_httpConditionalGet = false;

    protected static $_pluginLoader = null;

    protected static $_prefixPaths = array();

    protected static $_extensions = array(

        'feed' => array(

            'DublinCore_Feed',

            'Atom_Feed'

        ),

        'entry' => array(

            'Content_Entry',

            'DublinCore_Entry',

            'Atom_Entry'

        ),

        'core' => array(

            'DublinCore_Feed',

            'Atom_Feed',

            'Content_Entry',

            'DublinCore_Entry',

            'Atom_Entry'

        )

    );

    /**

     * Get the Feed cache

     *

     * @return Zend_Cache_Core

     */

    public static function getCache()

    {

        return self::$_cache;

    }

    /**

     * Set the feed cache

     *

     * @param Zend_Cache_Core $cache

     * @return void

     */

    public static function setCache(Zend_Cache_Core $cache)

    {

        self::$_cache = $cache;

    }

    /**

     * Set the HTTP client instance

     *

     * Sets the HTTP client object to use for retrieving the feeds.

     *

     * @param  Zend_Http_Client $httpClient

     * @return void

     */

    public static function setHttpClient(Zend_Http_Client $httpClient)

    {

        self::$_httpClient = $httpClient;

    }

    /**

     * Gets the HTTP client object. If none is set, a new Zend_Http_Client will be used.

     *

     * @return Zend_Http_Client_Abstract

     */

    public static function getHttpClient()

    {

        if (!self::$_httpClient instanceof Zend_Http_Client) {

            /**

             * @see Zend_Http_Client

             */

            require_once 'Zend/Http/Client.php';

            self::$_httpClient = new Zend_Http_Client();

        }

        return self::$_httpClient;

    }

    /**

     * Toggle using POST instead of PUT and DELETE HTTP methods

     *

     * Some feed implementations do not accept PUT and DELETE HTTP

     * methods, or they can't be used because of proxies or other

     * measures. This allows turning on using POST where PUT and

     * DELETE would normally be used; in addition, an

     * X-Method-Override header will be sent with a value of PUT or

     * DELETE as appropriate.

     *

     * @param  boolean $override Whether to override PUT and DELETE.

     * @return void

     */

    public static function setHttpMethodOverride($override = true)

    {

        self::$_httpMethodOverride = $override;

    }

    /**

     * Get the HTTP override state

     *

     * @return boolean

     */

    public static function getHttpMethodOverride()

    {

        return self::$_httpMethodOverride;

    }

    /**

     * Set the flag indicating whether or not to use HTTP conditional GET

     *

     * @param  bool $bool

     * @return void

     */

    public static function useHttpConditionalGet($bool = true)

    {

        self::$_httpConditionalGet = $bool;

    }

    /**

     * Import a feed by providing a URL

     *

     * @param  string $url The URL to the feed

     * @param  string $etag OPTIONAL Last received ETag for this resource

     * @param  string $lastModified OPTIONAL Last-Modified value for this resource

     * @return Zend_Feed_Reader_FeedInterface

     */

    public static function import($uri, $etag = null, $lastModified = null)

    {

        $cache       = self::getCache();

        $feed        = null;

        $responseXml = '';

        $client      = self::getHttpClient();

        $client->resetParameters();

        $client->setHeaders('If-None-Match', null);

        $client->setHeaders('If-Modified-Since', null);

        $client->setUri($uri);

        $cacheId = 'Zend_Feed_Reader_' . md5($uri);

        if (self::$_httpConditionalGet && $cache) {

            $data = $cache->load($cacheId);

            if ($data) {

                if ($etag === null) {

                    $etag = $cache->load($cacheId.'_etag');

                }

                if ($lastModified === null) {

                    $lastModified = $cache->load($cacheId.'_lastmodified');

                }

                if ($etag) {

                    $client->setHeaders('If-None-Match', $etag);

                }

                if ($lastModified) {

                    $client->setHeaders('If-Modified-Since', $lastModified);

                }

            }

            $response = $client->request('GET');

            if ($response->getStatus() !== 200 && $response->getStatus() !== 304) {

                require_once 'Zend/Feed/Exception.php';

                throw new Zend_Feed_Exception('Feed failed to load, got response code ' . $response->getStatus());

            }

            if ($response->getStatus() == 304) {

                $responseXml = $data;

            } else {

                $responseXml = $response->getBody();

                $cache->save($responseXml, $cacheId);

                if ($response->getHeader('ETag')) {

                    $cache->save($response->getHeader('ETag'), $cacheId.'_etag');

                }

                if ($response->getHeader('Last-Modified')) {

                    $cache->save($response->getHeader('Last-Modified'), $cacheId.'_lastmodified');

                }

            }

            if (empty($responseXml)) {

                require_once 'Zend/Feed/Exception.php';

                throw new Zend_Feed_Exception('Feed failed to load, got empty response body');

            }

            return self::importString($responseXml);

        } elseif ($cache) {

            $data = $cache->load($cacheId);

            if ($data !== false) {

                return self::importString($data);

            }

            $response = $client->request('GET');

            if ($response->getStatus() !== 200) {

                require_once 'Zend/Feed/Exception.php';

                throw new Zend_Feed_Exception('Feed failed to load, got response code ' . $response->getStatus());

            }

            $responseXml = $response->getBody();

            $cache->save($responseXml, $cacheId);

            if (empty($responseXml)) {

                require_once 'Zend/Feed/Exception.php';

                throw new Zend_Feed_Exception('Feed failed to load, got empty response body');

            }

            return self::importString($responseXml);

        } else {

            $response = $client->request('GET');

            if ($response->getStatus() !== 200) {

                require_once 'Zend/Feed/Exception.php';

                throw new Zend_Feed_Exception('Feed failed to load, got response code ' . $response->getStatus());

            }

            $responseXml = $response->getBody();

            if (empty($responseXml)) {

                require_once 'Zend/Feed/Exception.php';

                throw new Zend_Feed_Exception('Feed failed to load, got empty response body');

            }

            $reader = self::importString($responseXml);

            $reader->setOriginalSourceUri($uri);

            return $reader;

        }

    }

    /**

     * Import a feed by providing a Zend_Feed_Abstract object

     *

     * @param  Zend_Feed_Abstract $feed A fully instantiated Zend_Feed object

     * @return Zend_Feed_Reader_FeedInterface

     */

    public static function importFeed(Zend_Feed_Abstract $feed)

    {

        $dom  = $feed->getDOM()->ownerDocument;

        $type = self::detectType($dom);

        self::_registerCoreExtensions();

        if (substr($type, 0, 3) == 'rss') {

            $reader = new Zend_Feed_Reader_Feed_Rss($dom, $type);

        } else {

            $reader = new Zend_Feed_Reader_Feed_Atom($dom, $type);

        }

        return $reader;

    }

    /**

     * Import a feed froma string

     *

     * @param  string $string

     * @return Zend_Feed_Reader_FeedInterface

     */

    public static function importString($string)

    {

        $libxml_errflag = libxml_use_internal_errors(true);

        $oldValue = libxml_disable_entity_loader(true);

        $dom = new DOMDocument;

        $status = $dom->loadXML($string);

        foreach ($dom->childNodes as $child) {

            if ($child->nodeType === XML_DOCUMENT_TYPE_NODE) {

                require_once 'Zend/Feed/Exception.php';

                throw new Zend_Feed_Exception(

                    'Invalid XML: Detected use of illegal DOCTYPE'

                );

            }

        }

        libxml_disable_entity_loader($oldValue);

        libxml_use_internal_errors($libxml_errflag);

        if (!$status) {

            // Build error message

            $error = libxml_get_last_error();

            if ($error && $error->message) {

                $errormsg = "DOMDocument cannot parse XML: {$error->message}";

            } else {

                $errormsg = "DOMDocument cannot parse XML: Please check the XML document's validity";

            }

            require_once 'Zend/Feed/Exception.php';

            throw new Zend_Feed_Exception($errormsg);

        }

        $type = self::detectType($dom);

        self::_registerCoreExtensions();

        if (substr($type, 0, 3) == 'rss') {

            $reader = new Zend_Feed_Reader_Feed_Rss($dom, $type);

        } elseif (substr($type, 8, 5) == 'entry') {

            $reader = new Zend_Feed_Reader_Entry_Atom($dom->documentElement, 0, Zend_Feed_Reader::TYPE_ATOM_10);

        } elseif (substr($type, 0, 4) == 'atom') {

            $reader = new Zend_Feed_Reader_Feed_Atom($dom, $type);

        } else {

            require_once 'Zend/Feed/Exception.php';

            throw new Zend_Feed_Exception('The URI used does not point to a '

            . 'valid Atom, RSS or RDF feed that Zend_Feed_Reader can parse.');

        }

        return $reader;

    }

    /**

     * Imports a feed from a file located at $filename.

     *

     * @param  string $filename

     * @throws Zend_Feed_Exception

     * @return Zend_Feed_Reader_FeedInterface

     */

    public static function importFile($filename)

    {

        @ini_set('track_errors', 1);

        $feed = @file_get_contents($filename);

        @ini_restore('track_errors');

        if ($feed === false) {

            /**

             * @see Zend_Feed_Exception

             */

            require_once 'Zend/Feed/Exception.php';

            throw new Zend_Feed_Exception("File could not be loaded: $php_errormsg");

        }

        return self::importString($feed);

    }

    public static function findFeedLinks($uri)

    {

        // Get the HTTP response from $uri and save the contents

        $client = self::getHttpClient();

        $client->setUri($uri);

        $response = $client->request();

        if ($response->getStatus() !== 200) {

            /**

             * @see Zend_Feed_Exception

             */

            require_once 'Zend/Feed/Exception.php';

            throw new Zend_Feed_Exception("Failed to access $uri, got response code " . $response->getStatus());

        }

        $responseHtml = $response->getBody();

        $libxml_errflag = libxml_use_internal_errors(true);

        $oldValue = libxml_disable_entity_loader(true);

        $dom = new DOMDocument;

        $status = $dom->loadHTML($responseHtml);

        libxml_disable_entity_loader($oldValue);

        libxml_use_internal_errors($libxml_errflag);

        if (!$status) {

            // Build error message

            $error = libxml_get_last_error();

            if ($error && $error->message) {

                $errormsg = "DOMDocument cannot parse HTML: {$error->message}";

            } else {

                $errormsg = "DOMDocument cannot parse HTML: Please check the XML document's validity";

            }