// Author: Keyvan Minoukadeh\r
// Copyright (c) 2013 Keyvan Minoukadeh\r
// License: AGPLv3\r
-// Version: 3.1\r
-// Date: 2013-03-05\r
+// Version: 3.2\r
+// Date: 2013-05-13\r
// More info: http://fivefilters.org/content-only/\r
// Help: http://help.fivefilters.org\r
\r
\r
// Usage\r
// -----\r
-// Request this file passing it your feed in the querystring: makefulltextfeed.php?url=mysite.org\r
-// The following options can be passed in the querystring:\r
-// * URL: url=[feed or website url] (required, should be URL-encoded - in php: urlencode($url))\r
-// * URL points to HTML (not feed): html=true (optional, by default it's automatically detected)\r
-// * API key: key=[api key] (optional, refer to config.php)\r
-// * Max entries to process: max=[max number of items] (optional)\r
+// Request this file passing it a web page or feed URL in the querystring: makefulltextfeed.php?url=example.org/article\r
+// For more request parameters, see http://help.fivefilters.org/customer/portal/articles/226660-usage\r
\r
-error_reporting(E_ALL ^ E_NOTICE);\r
+//error_reporting(E_ALL ^ E_NOTICE);\r
ini_set("display_errors", 1);\r
@set_time_limit(120);\r
\r
\r
// set include path\r
set_include_path(realpath(dirname(__FILE__).'/libraries').PATH_SEPARATOR.get_include_path());\r
-// Autoloading of classes allows us to include files only when they're\r
-// needed. If we've got a cached copy, for example, only Zend_Cache is loaded.\r
-function autoload($class_name) {\r
- static $dir = null;\r
- if ($dir === null) $dir = dirname(__FILE__).'/libraries/';\r
- static $mapping = array(\r
- // Include FeedCreator for RSS/Atom creation\r
- 'FeedWriter' => 'feedwriter/FeedWriter.php',\r
- 'FeedItem' => 'feedwriter/FeedItem.php',\r
- // Include ContentExtractor and Readability for identifying and extracting content from URLs\r
- 'ContentExtractor' => 'content-extractor/ContentExtractor.php',\r
- 'SiteConfig' => 'content-extractor/SiteConfig.php',\r
- 'Readability' => 'readability/Readability.php',\r
- // Include Humble HTTP Agent to allow parallel requests and response caching\r
- 'HumbleHttpAgent' => 'humble-http-agent/HumbleHttpAgent.php',\r
- 'SimplePie_HumbleHttpAgent' => 'humble-http-agent/SimplePie_HumbleHttpAgent.php',\r
- 'CookieJar' => 'humble-http-agent/CookieJar.php',\r
- // Include Zend Cache to improve performance (cache results)\r
- 'Zend_Cache' => 'Zend/Cache.php',\r
- // Language detect\r
- 'Text_LanguageDetect' => 'language-detect/LanguageDetect.php',\r
- // HTML5 Lib\r
- 'HTML5_Parser' => 'html5/Parser.php',\r
- // htmLawed - used if XSS filter is enabled (xss_filter)\r
- 'htmLawed' => 'htmLawed/htmLawed.php'\r
- );\r
- if (isset($mapping[$class_name])) {\r
- debug("** Loading class $class_name ({$mapping[$class_name]})");\r
- require $dir.$mapping[$class_name];\r
- return true;\r
- } else {\r
- return false;\r
- }\r
-}\r
-spl_autoload_register('autoload');\r
-require dirname(__FILE__).'/libraries/simplepie/autoloader.php';\r
+\r
+require_once dirname(__FILE__).'/makefulltextfeedHelpers.php';\r
\r
////////////////////////////////\r
// Load config file\r
////////////////////////////////\r
// Check if service is enabled\r
////////////////////////////////\r
-if (!$options->enabled) { \r
- die('The full-text RSS service is currently disabled'); \r
+if (!$options->enabled) {\r
+ die('The full-text RSS service is currently disabled');\r
}\r
\r
////////////////////////////////\r
////////////////////////////////\r
// Check for feed URL\r
////////////////////////////////\r
-if (!isset($_GET['url'])) { \r
- die('No URL supplied'); \r
+if (!isset($_GET['url'])) {\r
+ die('No URL supplied');\r
}\r
$url = trim($_GET['url']);\r
if (strtolower(substr($url, 0, 7)) == 'feed://') {\r
if (isset($_GET['links'])) $redirect .= '&links='.urlencode($_GET['links']);\r
if (isset($_GET['exc'])) $redirect .= '&exc='.urlencode($_GET['exc']);\r
if (isset($_GET['format'])) $redirect .= '&format='.urlencode($_GET['format']);\r
- if (isset($_GET['callback'])) $redirect .= '&callback='.urlencode($_GET['callback']); \r
+ if (isset($_GET['callback'])) $redirect .= '&callback='.urlencode($_GET['callback']);\r
if (isset($_GET['l'])) $redirect .= '&l='.urlencode($_GET['l']);\r
if (isset($_GET['xss'])) $redirect .= '&xss';\r
if (isset($_GET['use_extracted_title'])) $redirect .= '&use_extracted_title';\r
+ if (isset($_GET['content'])) $redirect .= '&content='.urlencode($_GET['content']);\r
+ if (isset($_GET['summary'])) $redirect .= '&summary='.urlencode($_GET['summary']);\r
if (isset($_GET['debug'])) $redirect .= '&debug';\r
if ($debug_mode) {\r
debug('Redirecting to hide access key, follow URL below to continue');\r
\r
///////////////////////////////////////////////\r
// Set timezone.\r
-// Prevents warnings, but needs more testing - \r
+// Prevents warnings, but needs more testing -\r
// perhaps if timezone is set in php.ini we\r
// don't need to set it at all...\r
///////////////////////////////////////////////\r
}\r
$key_index = ($valid_key) ? (int)$_GET['key'] : 0;\r
if (!$valid_key && $options->key_required) {\r
- die('A valid key must be supplied'); \r
+ die('A valid key must be supplied');\r
}\r
if (!$valid_key && isset($_GET['key']) && $_GET['key'] != '') {\r
die('The entered key is invalid');\r
$favour_feed_titles = $options->favour_feed_titles;\r
}\r
\r
+///////////////////////////////////////////////\r
+// Include full content in output?\r
+///////////////////////////////////////////////\r
+if ($options->content === 'user') {\r
+ if (isset($_GET['content']) && $_GET['content'] === '0') {\r
+ $options->content = false;\r
+ } else {\r
+ $options->content = true;\r
+ }\r
+}\r
+\r
+///////////////////////////////////////////////\r
+// Include summaries in output?\r
+///////////////////////////////////////////////\r
+if ($options->summary === 'user') {\r
+ if (isset($_GET['summary']) && $_GET['summary'] === '1') {\r
+ $options->summary = true;\r
+ } else {\r
+ $options->summary = false;\r
+ }\r
+}\r
+\r
///////////////////////////////////////////////\r
// Exclude items if extraction fails\r
///////////////////////////////////////////////\r
$detect_language = $options->detect_language;\r
}\r
\r
-if ($detect_language >= 2) {\r
- $language_codes = array('albanian' => 'sq','arabic' => 'ar','azeri' => 'az','bengali' => 'bn','bulgarian' => 'bg',\r
- 'cebuano' => 'ceb', // ISO 639-2\r
- 'croatian' => 'hr','czech' => 'cs','danish' => 'da','dutch' => 'nl','english' => 'en','estonian' => 'et','farsi' => 'fa','finnish' => 'fi','french' => 'fr','german' => 'de','hausa' => 'ha',\r
- 'hawaiian' => 'haw', // ISO 639-2 \r
- 'hindi' => 'hi','hungarian' => 'hu','icelandic' => 'is','indonesian' => 'id','italian' => 'it','kazakh' => 'kk','kyrgyz' => 'ky','latin' => 'la','latvian' => 'lv','lithuanian' => 'lt','macedonian' => 'mk','mongolian' => 'mn','nepali' => 'ne','norwegian' => 'no','pashto' => 'ps',\r
- 'pidgin' => 'cpe', // ISO 639-2 \r
- 'polish' => 'pl','portuguese' => 'pt','romanian' => 'ro','russian' => 'ru','serbian' => 'sr','slovak' => 'sk','slovene' => 'sl','somali' => 'so','spanish' => 'es','swahili' => 'sw','swedish' => 'sv','tagalog' => 'tl','turkish' => 'tr','ukrainian' => 'uk','urdu' => 'ur','uzbek' => 'uz','vietnamese' => 'vi','welsh' => 'cy');\r
-}\r
$use_cld = extension_loaded('cld') && (version_compare(PHP_VERSION, '5.3.0') >= 0);\r
\r
/////////////////////////////////////\r
//////////////////////////////////\r
if ($options->caching) {\r
debug('Caching is enabled...');\r
- $cache_id = md5($max.$url.$valid_key.$links.$favour_feed_titles.$xss_filter.$exclude_on_fail.$format.$detect_language.(int)isset($_GET['pubsub']));\r
+ $cache_id = md5($max.$url.(int)$valid_key.$links.(int)$favour_feed_titles.(int)$options->content.(int)$options->summary.(int)$xss_filter.(int)$exclude_on_fail.$format.$detect_language.(int)isset($_GET['pubsub']));\r
$check_cache = true;\r
if ($options->apc && $options->smart_cache) {\r
apc_add("cache.$cache_id", 0, 10*60);\r
//////////////////////////////////\r
// Set up HTTP agent\r
//////////////////////////////////\r
+global $http;\r
$http = new HumbleHttpAgent();\r
$http->debug = $debug_mode;\r
$http->userAgentMap = $options->user_agents;\r
//////////////////////////////////\r
// Set up Content Extractor\r
//////////////////////////////////\r
+global $extractor;\r
$extractor = new ContentExtractor(dirname(__FILE__).'/site_config/custom', dirname(__FILE__).'/site_config/standard');\r
$extractor->debug = $debug_mode;\r
SiteConfig::$debug = $debug_mode;\r
$isDummyFeed = true;\r
unset($feed, $result);\r
// create single item dummy feed object\r
- class DummySingleItemFeed {\r
- public $item;\r
- function __construct($url) { $this->item = new DummySingleItem($url); }\r
- public function get_title() { return ''; }\r
- public function get_description() { return 'Content extracted from '.$this->item->url; }\r
- public function get_link() { return $this->item->url; }\r
- public function get_language() { return false; }\r
- public function get_image_url() { return false; }\r
- public function get_items($start=0, $max=1) { return array(0=>$this->item); }\r
- }\r
- class DummySingleItem {\r
- public $url;\r
- function __construct($url) { $this->url = $url; }\r
- public function get_permalink() { return $this->url; }\r
- public function get_title() { return null; }\r
- public function get_date($format='') { return false; }\r
- public function get_author($key=0) { return null; }\r
- public function get_authors() { return null; }\r
- public function get_description() { return ''; }\r
- public function get_enclosure($key=0, $prefer=null) { return null; }\r
- public function get_enclosures() { return null; }\r
- public function get_categories() { return null; }\r
- }\r
$feed = new DummySingleItemFeed($url);\r
}\r
\r
////////////////////////////////////////////\r
// Loop through feed items\r
////////////////////////////////////////////\r
-$items = $feed->get_items(0, $max); \r
+$items = $feed->get_items(0, $max);\r
// Request all feed items in parallel (if supported)\r
$urls_sanitized = array();\r
$urls = array();\r
$is_single_page = false;\r
if ($single_page_response = getSinglePage($item, $html, $effective_url)) {\r
$is_single_page = true;\r
- $html = $single_page_response['body'];\r
- // remove strange things\r
- $html = str_replace('</[>', '', $html); \r
- $html = convert_to_utf8($html, $single_page_response['headers']);\r
$effective_url = $single_page_response['effective_url'];\r
- debug("Retrieved single-page view from $effective_url");\r
+ // check if action defined for returned Content-Type\r
+ $mime_info = get_mime_action_info($single_page_response['headers']);\r
+ if (isset($mime_info['action'])) {\r
+ if ($mime_info['action'] == 'exclude') {\r
+ continue; // skip this feed item entry\r
+ } elseif ($mime_info['action'] == 'link') {\r
+ if ($mime_info['type'] == 'image') {\r
+ $html = "<a href=\"$effective_url\"><img src=\"$effective_url\" alt=\"{$mime_info['name']}\" /></a>";\r
+ } else {\r
+ $html = "<a href=\"$effective_url\">Download {$mime_info['name']}</a>";\r
+ }\r
+ $extracted_title = $mime_info['name'];\r
+ $do_content_extraction = false;\r
+ }\r
+ }\r
+ if ($do_content_extraction) {\r
+ $html = $single_page_response['body'];\r
+ // remove strange things\r
+ $html = str_replace('</[>', '', $html);\r
+ $html = convert_to_utf8($html, $single_page_response['headers']);\r
+ debug("Retrieved single-page view from $effective_url");\r
+ }\r
unset($single_page_response);\r
}\r
+ }\r
+ if ($do_content_extraction) {\r
debug('--------');\r
debug('Attempting to extract content');\r
$extract_result = $extractor->process($html, $effective_url);\r
$readability = $extractor->readability;\r
- $content_block = ($extract_result) ? $extractor->getContent() : null; \r
+ $content_block = ($extract_result) ? $extractor->getContent() : null;\r
$extracted_title = ($extract_result) ? $extractor->getTitle() : '';\r
// Deal with multi-page articles\r
//die('Next: '.$extractor->getNextPageUrl());\r
$is_multi_page = (!$is_single_page && $extract_result && $extractor->getNextPageUrl());\r
- if ($options->multipage && $is_multi_page) {\r
+ if ($options->multipage && $is_multi_page && $options->content) {\r
debug('--------');\r
debug('Attempting to process multi-page article');\r
$multi_page_urls = array();\r
// check it's not what we have already!\r
if (!in_array($next_page_url, $multi_page_urls)) {\r
// it's not, so let's attempt to fetch it\r
- $multi_page_urls[] = $next_page_url; \r
+ $multi_page_urls[] = $next_page_url;\r
$_prev_ref = $http->referer;\r
if (($response = $http->get($next_page_url, true)) && $response['status_code'] < 300) {\r
// make sure mime type is not something with a different action associated\r
// did we successfully deal with this multi-page article?\r
if (empty($multi_page_content)) {\r
debug('Failed to extract all parts of multi-page article, so not going to include them');\r
- $multi_page_content[] = $readability->dom->createElement('p')->innerHTML = '<em>This article appears to continue on subsequent pages which we could not extract</em>';\r
+ $_page = $readability->dom->createElement('p');\r
+ $_page->innerHTML = '<em>This article appears to continue on subsequent pages which we could not extract</em>';\r
+ $multi_page_content[] = $_page;\r
}\r
foreach ($multi_page_content as $_page) {\r
$_page = $content_block->ownerDocument->importNode($_page, true);\r
$content_block->appendChild($_page);\r
}\r
- unset($multi_page_urls, $multi_page_content, $page_mime_info, $next_page_url);\r
+ unset($multi_page_urls, $multi_page_content, $page_mime_info, $next_page_url, $_page);\r
}\r
}\r
// use extracted title for both feed and item title if we're using single-item dummy feed\r
} else {\r
$html = $content_block->ownerDocument->saveXML($content_block); // essentially outerHTML\r
}\r
- unset($content_block);\r
+ //unset($content_block);\r
// post-processing cleanup\r
$html = preg_replace('!<p>[\s\h\v]*</p>!u', '', $html);\r
if ($links == 'remove') {\r
}\r
}\r
\r
- if ($valid_key && isset($_GET['pubsub'])) { // used only on fivefilters.org at the moment\r
- $newitem->addElement('guid', 'http://fivefilters.org/content-only/redirect.php?url='.urlencode($item->get_permalink()), array('isPermaLink'=>'false'));\r
+ if ($valid_key && isset($_GET['pubsub'])) { // used only on fivefilters.org at the moment\r
+ $newitem->addElement('guid', 'http://fivefilters.org/content-only/redirect.php?url='.urlencode($item->get_permalink()), array('isPermaLink'=>'false'));\r
+ } else {\r
+ $newitem->addElement('guid', $item->get_permalink(), array('isPermaLink'=>'true'));\r
+ }\r
+ // filter xss?\r
+ if ($xss_filter) {\r
+ debug('Filtering HTML to remove XSS');\r
+ $html = htmLawed::hl($html, array('safe'=>1, 'deny_attribute'=>'style', 'comment'=>1, 'cdata'=>1));\r
+ }\r
+\r
+ // add content\r
+ if ($options->summary === true) {\r
+ // get summary\r
+ $summary = '';\r
+ if (!$do_content_extraction) {\r
+ $summary = $html;\r
} else {\r
- $newitem->addElement('guid', $item->get_permalink(), array('isPermaLink'=>'true'));\r
- }\r
- // filter xss?\r
- if ($xss_filter) {\r
- debug('Filtering HTML to remove XSS');\r
- $html = htmLawed::hl($html, array('safe'=>1, 'deny_attribute'=>'style', 'comment'=>1, 'cdata'=>1));\r
- }\r
- $newitem->setDescription($html);\r
- \r
- // set date\r
- if ((int)$item->get_date('U') > 0) {\r
- $newitem->setDate((int)$item->get_date('U'));\r
- } elseif ($extractor->getDate()) {\r
- $newitem->setDate($extractor->getDate());\r
- }\r
- \r
- // add authors\r
- if ($authors = $item->get_authors()) {\r
- foreach ($authors as $author) {\r
- // for some feeds, SimplePie stores author's name as email, e.g. http://feeds.feedburner.com/nymag/intel\r
- if ($author->get_name() !== null) {\r
- $newitem->addElement('dc:creator', $author->get_name());\r
- } elseif ($author->get_email() !== null) {\r
- $newitem->addElement('dc:creator', $author->get_email());\r
+ // Try to get first few paragraphs\r
+ if (isset($content_block) && ($content_block instanceof DOMElement)) {\r
+ $_paras = $content_block->getElementsByTagName('p');\r
+ foreach ($_paras as $_para) {\r
+ $summary .= preg_replace("/[\n\r\t ]+/", ' ', $_para->textContent).' ';\r
+ if (strlen($summary) > 200) break;\r
}\r
+ } else {\r
+ $summary = $html;\r
}\r
- } elseif ($authors = $extractor->getAuthors()) {\r
- //TODO: make sure the list size is reasonable\r
- foreach ($authors as $author) {\r
- // TODO: xpath often selects authors from other articles linked from the page.\r
- // for now choose first item\r
- $newitem->addElement('dc:creator', $author);\r
- break;\r
+ }\r
+ unset($_paras, $_para);\r
+ $summary = get_excerpt($summary);\r
+ $newitem->setDescription($summary);\r
+ if ($options->content) $newitem->setElement('content:encoded', $html);\r
+ } else {\r
+ if ($options->content) $newitem->setDescription($html);\r
+ }\r
+\r
+ // set date\r
+ if ((int)$item->get_date('U') > 0) {\r
+ $newitem->setDate((int)$item->get_date('U'));\r
+ } elseif ($extractor->getDate()) {\r
+ $newitem->setDate($extractor->getDate());\r
+ }\r
+\r
+ // add authors\r
+ if ($authors = $item->get_authors()) {\r
+ foreach ($authors as $author) {\r
+ // for some feeds, SimplePie stores author's name as email, e.g. http://feeds.feedburner.com/nymag/intel\r
+ if ($author->get_name() !== null) {\r
+ $newitem->addElement('dc:creator', $author->get_name());\r
+ } elseif ($author->get_email() !== null) {\r
+ $newitem->addElement('dc:creator', $author->get_email());\r
}\r
}\r
- \r
- // add language\r
- if ($detect_language) {\r
- $language = $extractor->getLanguage();\r
- if (!$language) $language = $feed->get_language();\r
- if (($detect_language == 3 || (!$language && $detect_language == 2)) && $text_sample) {\r
- try {\r
- if ($use_cld) {\r
- // Use PHP-CLD extension\r
- $php_cld = 'CLD\detect'; // in quotes to prevent PHP 5.2 parse error\r
- $res = $php_cld($text_sample);\r
- if (is_array($res) && count($res) > 0) {\r
- $language = $res[0]['code'];\r
- } \r
- } else {\r
- //die('what');\r
- // Use PEAR's Text_LanguageDetect\r
- if (!isset($l)) {\r
- $l = new Text_LanguageDetect('libraries/language-detect/lang.dat', 'libraries/language-detect/unicode_blocks.dat');\r
- }\r
- $l_result = $l->detect($text_sample, 1);\r
- if (count($l_result) > 0) {\r
- $language = $language_codes[key($l_result)];\r
- }\r
+ } elseif ($authors = $extractor->getAuthors()) {\r
+ //TODO: make sure the list size is reasonable\r
+ foreach ($authors as $author) {\r
+ // TODO: xpath often selects authors from other articles linked from the page.\r
+ // for now choose first item\r
+ $newitem->addElement('dc:creator', $author);\r
+ break;\r
+ }\r
+ }\r
+\r
+ // add language\r
+ if ($detect_language) {\r
+ $language = $extractor->getLanguage();\r
+ if (!$language) $language = $feed->get_language();\r
+ if (($detect_language == 3 || (!$language && $detect_language == 2)) && $text_sample) {\r
+ try {\r
+ if ($use_cld) {\r
+ // Use PHP-CLD extension\r
+ $php_cld = 'CLD\detect'; // in quotes to prevent PHP 5.2 parse error\r
+ $res = $php_cld($text_sample);\r
+ if (is_array($res) && count($res) > 0) {\r
+ $language = $res[0]['code'];\r
+ }\r
+ } else {\r
+ //die('what');\r
+ // Use PEAR's Text_LanguageDetect\r
+ if (!isset($l)) {\r
+ $l = new Text_LanguageDetect();\r
+ $l->setNameMode(2); // return ISO 639-1 codes (e.g. "en")\r
+ }\r
+ $l_result = $l->detect($text_sample, 1);\r
+ if (count($l_result) > 0) {\r
+ $language = key($l_result);\r
}\r
- } catch (Exception $e) {\r
- //die('error: '.$e); \r
- // do nothing\r
}\r
- }\r
- if ($language && (strlen($language) < 7)) { \r
- $newitem->addElement('dc:language', $language);\r
+ } catch (Exception $e) {\r
+ //die('error: '.$e);\r
+ // do nothing\r
}\r
}\r
- \r
- // add MIME type (if it appeared in our exclusions lists)\r
- if (isset($mime_info['mime'])) $newitem->addElement('dc:format', $mime_info['mime']);\r
- // add effective URL (URL after redirects)\r
- if (isset($effective_url)) {\r
- //TODO: ensure $effective_url is valid witout - sometimes it causes problems, e.g.\r
- //http://www.siasat.pk/forum/showthread.php?108883-Pakistan-Chowk-by-Rana-Mubashir-\96-25th-March-2012-Special-Program-from-Liari-(Karachi)\r
- //temporary measure: use utf8_encode()\r
- $newitem->addElement('dc:identifier', remove_url_cruft(utf8_encode($effective_url)));\r
- } else {\r
- $newitem->addElement('dc:identifier', remove_url_cruft($item->get_permalink()));\r
+ if ($language && (strlen($language) < 7)) {\r
+ $newitem->addElement('dc:language', $language);\r
}\r
- \r
- // add categories\r
- if ($categories = $item->get_categories()) {\r
- foreach ($categories as $category) {\r
- if ($category->get_label() !== null) {\r
- $newitem->addElement('category', $category->get_label());\r
- }\r
+ }\r
+\r
+ // add MIME type (if it appeared in our exclusions lists)\r
+ if (isset($mime_info['mime'])) $newitem->addElement('dc:format', $mime_info['mime']);\r
+ // add effective URL (URL after redirects)\r
+ if (isset($effective_url)) {\r
+ //TODO: ensure $effective_url is valid witout - sometimes it causes problems, e.g.\r
+ //http://www.siasat.pk/forum/showthread.php?108883-Pakistan-Chowk-by-Rana-Mubashir-�-25th-March-2012-Special-Program-from-Liari-(Karachi)\r
+ //temporary measure: use utf8_encode()\r
+ $newitem->addElement('dc:identifier', remove_url_cruft(utf8_encode($effective_url)));\r
+ } else {\r
+ $newitem->addElement('dc:identifier', remove_url_cruft($item->get_permalink()));\r
+ }\r
+\r
+ // add categories\r
+ if ($categories = $item->get_categories()) {\r
+ foreach ($categories as $category) {\r
+ if ($category->get_label() !== null) {\r
+ $newitem->addElement('category', $category->get_label());\r
}\r
}\r
- \r
- // check for enclosures\r
- if ($options->keep_enclosures) {\r
- if ($enclosures = $item->get_enclosures()) {\r
- foreach ($enclosures as $enclosure) {\r
- // thumbnails\r
- foreach ((array)$enclosure->get_thumbnails() as $thumbnail) {\r
- $newitem->addElement('media:thumbnail', '', array('url'=>$thumbnail));\r
- }\r
- if (!$enclosure->get_link()) continue;\r
- $enc = array();\r
- // Media RSS spec ($enc): http://search.yahoo.com/mrss\r
- // SimplePie methods ($enclosure): http://simplepie.org/wiki/reference/start#methods4\r
- $enc['url'] = $enclosure->get_link();\r
- if ($enclosure->get_length()) $enc['fileSize'] = $enclosure->get_length();\r
- if ($enclosure->get_type()) $enc['type'] = $enclosure->get_type();\r
- if ($enclosure->get_medium()) $enc['medium'] = $enclosure->get_medium();\r
- if ($enclosure->get_expression()) $enc['expression'] = $enclosure->get_expression();\r
- if ($enclosure->get_bitrate()) $enc['bitrate'] = $enclosure->get_bitrate();\r
- if ($enclosure->get_framerate()) $enc['framerate'] = $enclosure->get_framerate();\r
- if ($enclosure->get_sampling_rate()) $enc['samplingrate'] = $enclosure->get_sampling_rate();\r
- if ($enclosure->get_channels()) $enc['channels'] = $enclosure->get_channels();\r
- if ($enclosure->get_duration()) $enc['duration'] = $enclosure->get_duration();\r
- if ($enclosure->get_height()) $enc['height'] = $enclosure->get_height();\r
- if ($enclosure->get_width()) $enc['width'] = $enclosure->get_width();\r
- if ($enclosure->get_language()) $enc['lang'] = $enclosure->get_language();\r
- $newitem->addElement('media:content', '', $enc);\r
+ }\r
+\r
+ // check for enclosures\r
+ if ($options->keep_enclosures) {\r
+ if ($enclosures = $item->get_enclosures()) {\r
+ foreach ($enclosures as $enclosure) {\r
+ // thumbnails\r
+ foreach ((array)$enclosure->get_thumbnails() as $thumbnail) {\r
+ $newitem->addElement('media:thumbnail', '', array('url'=>$thumbnail));\r
}\r
+ if (!$enclosure->get_link()) continue;\r
+ $enc = array();\r
+ // Media RSS spec ($enc): http://search.yahoo.com/mrss\r
+ // SimplePie methods ($enclosure): http://simplepie.org/wiki/reference/start#methods4\r
+ $enc['url'] = $enclosure->get_link();\r
+ if ($enclosure->get_length()) $enc['fileSize'] = $enclosure->get_length();\r
+ if ($enclosure->get_type()) $enc['type'] = $enclosure->get_type();\r
+ if ($enclosure->get_medium()) $enc['medium'] = $enclosure->get_medium();\r
+ if ($enclosure->get_expression()) $enc['expression'] = $enclosure->get_expression();\r
+ if ($enclosure->get_bitrate()) $enc['bitrate'] = $enclosure->get_bitrate();\r
+ if ($enclosure->get_framerate()) $enc['framerate'] = $enclosure->get_framerate();\r
+ if ($enclosure->get_sampling_rate()) $enc['samplingrate'] = $enclosure->get_sampling_rate();\r
+ if ($enclosure->get_channels()) $enc['channels'] = $enclosure->get_channels();\r
+ if ($enclosure->get_duration()) $enc['duration'] = $enclosure->get_duration();\r
+ if ($enclosure->get_height()) $enc['height'] = $enclosure->get_height();\r
+ if ($enclosure->get_width()) $enc['width'] = $enclosure->get_width();\r
+ if ($enclosure->get_language()) $enc['lang'] = $enclosure->get_language();\r
+ $newitem->addElement('media:content', '', $enc);\r
}\r
}\r
- /* } */\r
+ }\r
$output->addItem($newitem);\r
unset($html);\r
$item_count++;\r
}\r
if ($add_to_cache) {\r
ob_start();\r
- $output->genarateFeed();\r
+ $output->genarateFeed(false);\r
$output = ob_get_contents();\r
ob_end_clean();\r
if ($html_only && $item_count == 0) {\r
}\r
echo $output;\r
} else {\r
- $output->genarateFeed();\r
+ $output->genarateFeed(false);\r
}\r
if ($callback) echo ');';\r
}\r
\r
-///////////////////////////////\r
-// HELPER FUNCTIONS\r
-///////////////////////////////\r
-\r
-function url_allowed($url) {\r
- global $options;\r
- if (!empty($options->allowed_urls)) {\r
- $allowed = false;\r
- foreach ($options->allowed_urls as $allowurl) {\r
- if (stristr($url, $allowurl) !== false) {\r
- $allowed = true;\r
- break;\r
- }\r
- }\r
- if (!$allowed) return false;\r
- } else {\r
- foreach ($options->blocked_urls as $blockurl) {\r
- if (stristr($url, $blockurl) !== false) {\r
- return false;\r
- }\r
- }\r
- }\r
- return true;\r
-}\r
-\r
-//////////////////////////////////////////////\r
-// Convert $html to UTF8\r
-// (uses HTTP headers and HTML to find encoding)\r
-// adapted from http://stackoverflow.com/questions/910793/php-detect-encoding-and-make-everything-utf-8\r
-//////////////////////////////////////////////\r
-function convert_to_utf8($html, $header=null)\r
-{\r
- $encoding = null;\r
- if ($html || $header) {\r
- if (is_array($header)) $header = implode("\n", $header);\r
- if (!$header || !preg_match_all('/^Content-Type:\s+([^;]+)(?:;\s*charset=["\']?([^;"\'\n]*))?/im', $header, $match, PREG_SET_ORDER)) {\r
- // error parsing the response\r
- debug('Could not find Content-Type header in HTTP response');\r
- } else {\r
- $match = end($match); // get last matched element (in case of redirects)\r
- if (isset($match[2])) $encoding = trim($match[2], "\"' \r\n\0\x0B\t");\r
- }\r
- // TODO: check to see if encoding is supported (can we convert it?)\r
- // If it's not, result will be empty string.\r
- // For now we'll check for invalid encoding types returned by some sites, e.g. 'none'\r
- // Problem URL: http://facta.co.jp/blog/archives/20111026001026.html\r
- if (!$encoding || $encoding == 'none') {\r
- // search for encoding in HTML - only look at the first 50000 characters\r
- // Why 50000? See, for example, http://www.lemonde.fr/festival-de-cannes/article/2012/05/23/deux-cretes-en-goguette-sur-la-croisette_1705732_766360.html\r
- // TODO: improve this so it looks at smaller chunks first\r
- $html_head = substr($html, 0, 50000);\r
- if (preg_match('/^<\?xml\s+version=(?:"[^"]*"|\'[^\']*\')\s+encoding=("[^"]*"|\'[^\']*\')/s', $html_head, $match)) {\r
- $encoding = trim($match[1], '"\'');\r
- } elseif (preg_match('/<meta\s+http-equiv=["\']?Content-Type["\']? content=["\'][^;]+;\s*charset=["\']?([^;"\'>]+)/i', $html_head, $match)) {\r
- $encoding = trim($match[1]);\r
- } elseif (preg_match_all('/<meta\s+([^>]+)>/i', $html_head, $match)) {\r
- foreach ($match[1] as $_test) {\r
- if (preg_match('/charset=["\']?([^"\']+)/i', $_test, $_m)) {\r
- $encoding = trim($_m[1]);\r
- break;\r
- }\r
- }\r
- }\r
- }\r
- if (isset($encoding)) $encoding = trim($encoding);\r
- // trim is important here!\r
- if (!$encoding || (strtolower($encoding) == 'iso-8859-1')) {\r
- // replace MS Word smart qutoes\r
- $trans = array();\r
- $trans[chr(130)] = '‚'; // Single Low-9 Quotation Mark\r
- $trans[chr(131)] = 'ƒ'; // Latin Small Letter F With Hook\r
- $trans[chr(132)] = '„'; // Double Low-9 Quotation Mark\r
- $trans[chr(133)] = '…'; // Horizontal Ellipsis\r
- $trans[chr(134)] = '†'; // Dagger\r
- $trans[chr(135)] = '‡'; // Double Dagger\r
- $trans[chr(136)] = 'ˆ'; // Modifier Letter Circumflex Accent\r
- $trans[chr(137)] = '‰'; // Per Mille Sign\r
- $trans[chr(138)] = 'Š'; // Latin Capital Letter S With Caron\r
- $trans[chr(139)] = '‹'; // Single Left-Pointing Angle Quotation Mark\r
- $trans[chr(140)] = 'Œ'; // Latin Capital Ligature OE\r
- $trans[chr(145)] = '‘'; // Left Single Quotation Mark\r
- $trans[chr(146)] = '’'; // Right Single Quotation Mark\r
- $trans[chr(147)] = '“'; // Left Double Quotation Mark\r
- $trans[chr(148)] = '”'; // Right Double Quotation Mark\r
- $trans[chr(149)] = '•'; // Bullet\r
- $trans[chr(150)] = '–'; // En Dash\r
- $trans[chr(151)] = '—'; // Em Dash\r
- $trans[chr(152)] = '˜'; // Small Tilde\r
- $trans[chr(153)] = '™'; // Trade Mark Sign\r
- $trans[chr(154)] = 'š'; // Latin Small Letter S With Caron\r
- $trans[chr(155)] = '›'; // Single Right-Pointing Angle Quotation Mark\r
- $trans[chr(156)] = 'œ'; // Latin Small Ligature OE\r
- $trans[chr(159)] = 'Ÿ'; // Latin Capital Letter Y With Diaeresis\r
- $html = strtr($html, $trans);\r
- }\r
- if (!$encoding) {\r
- debug('No character encoding found, so treating as UTF-8');\r
- $encoding = 'utf-8';\r
- } else {\r
- debug('Character encoding: '.$encoding);\r
- if (strtolower($encoding) != 'utf-8') {\r
- debug('Converting to UTF-8');\r
- $html = SimplePie_Misc::change_encoding($html, $encoding, 'utf-8');\r
- /*\r
- if (function_exists('iconv')) {\r
- // iconv appears to handle certain character encodings better than mb_convert_encoding\r
- $html = iconv($encoding, 'utf-8', $html);\r
- } else {\r
- $html = mb_convert_encoding($html, 'utf-8', $encoding);\r
- }\r
- */\r
- }\r
- }\r
- }\r
- return $html;\r
-}\r
-\r
-function makeAbsolute($base, $elem) {\r
- $base = new SimplePie_IRI($base);\r
- // remove '//' in URL path (used to prevent URLs from resolving properly)\r
- // TODO: check if this is still the case\r
- if (isset($base->path)) $base->path = preg_replace('!//+!', '/', $base->path);\r
- foreach(array('a'=>'href', 'img'=>'src') as $tag => $attr) {\r
- $elems = $elem->getElementsByTagName($tag);\r
- for ($i = $elems->length-1; $i >= 0; $i--) {\r
- $e = $elems->item($i);\r
- //$e->parentNode->replaceChild($articleContent->ownerDocument->createTextNode($e->textContent), $e);\r
- makeAbsoluteAttr($base, $e, $attr);\r
- }\r
- if (strtolower($elem->tagName) == $tag) makeAbsoluteAttr($base, $elem, $attr);\r
- }\r
-}\r
-function makeAbsoluteAttr($base, $e, $attr) {\r
- if ($e->hasAttribute($attr)) {\r
- // Trim leading and trailing white space. I don't really like this but \r
- // unfortunately it does appear on some sites. e.g. <img src=" /path/to/image.jpg" />\r
- $url = trim(str_replace('%20', ' ', $e->getAttribute($attr)));\r
- $url = str_replace(' ', '%20', $url);\r
- if (!preg_match('!https?://!i', $url)) {\r
- if ($absolute = SimplePie_IRI::absolutize($base, $url)) {\r
- $e->setAttribute($attr, $absolute);\r
- }\r
- }\r
- }\r
-}\r
-function makeAbsoluteStr($base, $url) {\r
- $base = new SimplePie_IRI($base);\r
- // remove '//' in URL path (causes URLs not to resolve properly)\r
- if (isset($base->path)) $base->path = preg_replace('!//+!', '/', $base->path);\r
- if (preg_match('!^https?://!i', $url)) {\r
- // already absolute\r
- return $url;\r
- } else {\r
- if ($absolute = SimplePie_IRI::absolutize($base, $url)) {\r
- return $absolute;\r
- }\r
- return false;\r
- }\r
-}\r
-// returns single page response, or false if not found\r
-function getSinglePage($item, $html, $url) {\r
- global $http, $extractor;\r
- debug('Looking for site config files to see if single page link exists');\r
- $site_config = $extractor->buildSiteConfig($url, $html);\r
- $splink = null;\r
- if (!empty($site_config->single_page_link)) {\r
- $splink = $site_config->single_page_link;\r
- } elseif (!empty($site_config->single_page_link_in_feed)) {\r
- // single page link xpath is targeted at feed\r
- $splink = $site_config->single_page_link_in_feed;\r
- // so let's replace HTML with feed item description\r
- $html = $item->get_description();\r
- }\r
- if (isset($splink)) {\r
- // Build DOM tree from HTML\r
- $readability = new Readability($html, $url);\r
- $xpath = new DOMXPath($readability->dom);\r
- // Loop through single_page_link xpath expressions\r
- $single_page_url = null;\r
- foreach ($splink as $pattern) {\r
- $elems = @$xpath->evaluate($pattern, $readability->dom);\r
- if (is_string($elems)) {\r
- $single_page_url = trim($elems);\r
- break;\r
- } elseif ($elems instanceof DOMNodeList && $elems->length > 0) {\r
- foreach ($elems as $item) {\r
- if ($item instanceof DOMElement && $item->hasAttribute('href')) {\r
- $single_page_url = $item->getAttribute('href');\r
- break 2;\r
- } elseif ($item instanceof DOMAttr && $item->value) {\r
- $single_page_url = $item->value;\r
- break 2;\r
- }\r
- }\r
- }\r
- }\r
- // If we've got URL, resolve against $url\r
- if (isset($single_page_url) && ($single_page_url = makeAbsoluteStr($url, $single_page_url))) {\r
- // check it's not what we have already!\r
- if ($single_page_url != $url) {\r
- // it's not, so let's try to fetch it...\r
- $_prev_ref = $http->referer;\r
- $http->referer = $single_page_url;\r
- if (($response = $http->get($single_page_url, true)) && $response['status_code'] < 300) {\r
- $http->referer = $_prev_ref;\r
- return $response;\r
- }\r
- $http->referer = $_prev_ref;\r
- }\r
- }\r
- }\r
- return false;\r
-}\r
-\r
-// based on content-type http header, decide what to do\r
-// param: HTTP headers string\r
-// return: array with keys: 'mime', 'type', 'subtype', 'action', 'name'\r
-// e.g. array('mime'=>'image/jpeg', 'type'=>'image', 'subtype'=>'jpeg', 'action'=>'link', 'name'=>'Image')\r
-function get_mime_action_info($headers) {\r
- global $options;\r
- // check if action defined for returned Content-Type\r
- $info = array();\r
- if (preg_match('!^Content-Type:\s*(([-\w]+)/([-\w\+]+))!im', $headers, $match)) {\r
- // look for full mime type (e.g. image/jpeg) or just type (e.g. image)\r
- // match[1] = full mime type, e.g. image/jpeg\r
- // match[2] = first part, e.g. image\r
- // match[3] = last part, e.g. jpeg\r
- $info['mime'] = strtolower(trim($match[1]));\r
- $info['type'] = strtolower(trim($match[2]));\r
- $info['subtype'] = strtolower(trim($match[3]));\r
- foreach (array($info['mime'], $info['type']) as $_mime) {\r
- if (isset($options->content_type_exc[$_mime])) {\r
- $info['action'] = $options->content_type_exc[$_mime]['action'];\r
- $info['name'] = $options->content_type_exc[$_mime]['name'];\r
- break;\r
- }\r
- }\r
- }\r
- return $info;\r
-}\r
-\r
-function remove_url_cruft($url) {\r
- // remove google analytics for the time being\r
- // regex adapted from http://navitronic.co.uk/2010/12/removing-google-analytics-cruft-from-urls/\r
- // https://gist.github.com/758177\r
- return preg_replace('/(\?|\&)utm_[a-z]+=[^\&]+/', '', $url);\r
-}\r
-\r
-function make_substitutions($string) {\r
- if ($string == '') return $string;\r
- global $item, $effective_url;\r
- $string = str_replace('{url}', htmlspecialchars($item->get_permalink()), $string);\r
- $string = str_replace('{effective-url}', htmlspecialchars($effective_url), $string);\r
- return $string;\r
-}\r
-\r
-function get_cache() {\r
- global $options, $valid_key;\r
- static $cache = null;\r
- if ($cache === null) {\r
- $frontendOptions = array(\r
- 'lifetime' => 10*60, // cache lifetime of 10 minutes\r
- 'automatic_serialization' => false,\r
- 'write_control' => false,\r
- 'automatic_cleaning_factor' => $options->cache_cleanup,\r
- 'ignore_user_abort' => false\r
- );\r
- $backendOptions = array(\r
- 'cache_dir' => ($valid_key) ? $options->cache_dir.'/rss-with-key/' : $options->cache_dir.'/rss/', // directory where to put the cache files\r
- 'file_locking' => false,\r
- 'read_control' => true,\r
- 'read_control_type' => 'strlen',\r
- 'hashed_directory_level' => $options->cache_directory_level,\r
- 'hashed_directory_perm' => 0777,\r
- 'cache_file_perm' => 0664,\r
- 'file_name_prefix' => 'ff'\r
- );\r
- // getting a Zend_Cache_Core object\r
- $cache = Zend_Cache::factory('Core', 'File', $frontendOptions, $backendOptions);\r
- }\r
- return $cache;\r
-}\r
-\r
-function debug($msg) {\r
- global $debug_mode;\r
- if ($debug_mode) {\r
- echo '* ',$msg,"\n";\r
- ob_flush();\r
- flush();\r
- }\r
-}
\ No newline at end of file