- * Scrape supported websites and perform conversion to RSS.
- */
-class Tweeper {
-
- private static $userAgent = "Mozilla/5.0 (Windows NT 6.1; rv:22.0) Gecko/20130405 Firefox/22.0";
-
- /**
- * Constructor sets up {@link $generate_enclosure}.
- */
- public function __construct($generate_enclosure = FALSE) {
- $this->generate_enclosure = $generate_enclosure;
- }
-
- /**
- * Convert numeric Epoch to the date format expected in a RSS document.
- */
- public static function epoch_to_gmdate($timestamp) {
- if (!is_numeric($timestamp) || is_nan($timestamp)) {
- $timestamp = 0;
- }
-
- return gmdate('D, d M Y H:i:s', $timestamp) . ' GMT';
- }
-
- /**
- * Convert generic date string to the date format expected in a RSS document.
- */
- public static function str_to_gmdate($date) {
- $timestamp = strtotime($date);
- if (FALSE === $timestamp) {
- $timestamp = 0;
- }
-
- return Tweeper::epoch_to_gmdate($timestamp);
- }
-
- /**
- * Get the contents from a URL.
- */
- private static function get_contents($url) {
- $ch = curl_init($url);
- curl_setopt_array($ch, array(
- CURLOPT_HEADER => FALSE,
- // Follow http redirects to get the real URL.
- CURLOPT_FOLLOWLOCATION => TRUE,
- CURLOPT_RETURNTRANSFER => TRUE,
- CURLOPT_SSL_VERIFYHOST => FALSE,
- CURLOPT_SSL_VERIFYPEER => FALSE,
- CURLOPT_HTTPHEADER => array('Accept-language: en'),
- CURLOPT_USERAGENT => Tweeper::$userAgent,
- ));
- $contents = curl_exec($ch);
- curl_close($ch);
-
- return $contents;
- }
-
- /**
- * Get the headers from a URL.
- */
- private static function get_info($url) {
- $ch = curl_init($url);
- curl_setopt_array($ch, array(
- CURLOPT_HEADER => TRUE,
- CURLOPT_NOBODY => TRUE,
- // Follow http redirects to get the real URL.
- CURLOPT_FOLLOWLOCATION => TRUE,
- CURLOPT_RETURNTRANSFER => TRUE,
- CURLOPT_SSL_VERIFYHOST => FALSE,
- CURLOPT_SSL_VERIFYPEER => FALSE,
- CURLOPT_USERAGENT => Tweeper::$userAgent,
- ));
- curl_exec($ch);
- $url_info = curl_getinfo($ch);
- curl_close($ch);
-
- return $url_info;
- }
-
- /**
- * Generate an RSS <enclosure/> element.
- */
- public static function generate_enclosure($url) {
- $supported_content_types = array(
- "application/ogg",
- "audio/aac",
- "audio/mp4",
- "audio/mpeg",
- "audio/ogg",
- "audio/vorbis",
- "audio/wav",
- "audio/webm",
- "audio/x-midi",
- "image/gif",
- "image/jpeg",
- "video/avi",
- "video/mp4",
- "video/mpeg",
- "video/ogg",
- );
-
- // The RSS specification says that the enclosure element URL must be http.
- // See http://sourceforge.net/p/feedvalidator/bugs/72/
- $http_url = preg_replace("/^https/", "http", $url);
-
- $url_info = Tweeper::get_info($http_url);
-
- $supported = in_array($url_info['content_type'], $supported_content_types);
- if (!$supported) {
- error_log("Unsupported enclosure content type \"" . $url_info['content_type'] . "\" for URL: " . $url_info['url']);
- return '';
- }
-
- $dom = new DomDocument();
- $enc = $dom->createElement('enclosure');
- $enc->setAttribute('url', $url_info['url']);
- $enc->setAttribute('length', $url_info['download_content_length']);
- $enc->setAttribute('type', $url_info['content_type']);
-
- $dom->appendChild($enc);
-
- return $dom->saveXML($enc);
- }
-
- /**
- * Mimic the message from libxml.c::php_libxml_ctx_error_level()
- */
- private function log_xml_error($error) {
- $output = "";
-
- switch ($error->level) {
- case LIBXML_ERR_WARNING:
- $output .= "Warning $error->code: ";
- break;
-
- case LIBXML_ERR_ERROR:
- $output .= "Error $error->code: ";
- break;
-
- case LIBXML_ERR_FATAL:
- $output .= "Fatal Error $error->code: ";
- break;
- }
-
- $output .= trim($error->message);
-
- if ($error->file) {
- $output .= " in $error->file";
- }
- else {
- $output .= " in Entity,";
- }
-
- $output .= " line $error->line";
-
- error_log($output);
- }
-
- /**
- * Load a stylesheet if the web site is supported.
- */
- private function load_stylesheet($host) {
- $stylesheet = "file://" . __DIR__ . "/rss_converter_" . $host . ".xsl";
- if (FALSE === file_exists($stylesheet)) {
- trigger_error("Conversion to RSS not supported for $host ($stylesheet not found)", E_USER_ERROR);
- return NULL;
- }
-
- $stylesheet_contents = $this->get_contents($stylesheet);
-
- $xslDoc = new DOMDocument();
- $xslDoc->loadXML($stylesheet_contents);
-
- $xsltProcessor = new XSLTProcessor();
- $xsltProcessor->registerPHPFunctions();
- $xsltProcessor->setParameter('', 'generateEnclosure', $this->generate_enclosure);
- $xsltProcessor->importStylesheet($xslDoc);
-
- return $xsltProcessor;
- }
-
- /**
- * Convert json to xml.
- */
- private function json_to_xml($json, $root_node_name) {
- // Apparently the ObjectNormalizer used afterwards is not able to handle
- // the stdClass object created by json_decode() with the default setting
- // $assoc = false; so use $assoc = true.
- $data = json_decode($json, $assoc = TRUE);
- if (!$data) {
- return NULL;
- }
-
- $encoder = new XmlEncoder();
- $normalizer = new ObjectNormalizer();
- $serializer = new Serializer(array($normalizer), array($encoder));
-
- $serializer_options = array(
- 'xml_encoding' => "UTF-8",
- 'xml_format_output' => TRUE,
- 'xml_root_node_name' => $root_node_name,
- );
-
- $xml_data = $serializer->serialize($data, 'xml', $serializer_options);
- if (!$xml_data) {
- trigger_error("Cannot serialize data", E_USER_ERROR);
- return NULL;
- }
-
- return $xml_data;
- }
-
- /**
- * Convert the Instagram content to XML.
- */
- private function get_xml_instagram_com($html) {
- // Extract the json data from the html code.
- $json_match_expr = '/window._sharedData = (.*);/';
- $ret = preg_match($json_match_expr, $html, $matches);
- if ($ret !== 1) {
- trigger_error("Cannot match expression: $json_match_expr\n", E_USER_ERROR);
- return NULL;
- }
-
- return $this->json_to_xml($matches[1], 'instagram');
- }
-
- /**
- * Make the Facebook HTML processable.
- */
- private function preprocess_html_facebook_com($html) {
- $html = str_replace('<!--', '', $html);
- $html = str_replace('-->', '', $html);
- return $html;
- }
-
- /**
- * Convert the HTML retrieved from the site to XML.
- */
- private function html_to_xml($html, $host) {
- $xmlDoc = new DOMDocument();
-
- // Handle warnings and errors when loading invalid HTML.
- $xml_errors_value = libxml_use_internal_errors(TRUE);
-
- // If there is a host-specific method to get the xml data, use it!
- $get_xml_host_method = 'get_xml_' . str_replace(".", "_", $host);
- if (method_exists($this, $get_xml_host_method)) {
- $xml_data = call_user_func_array(array($this, $get_xml_host_method), array($html));
- $xmlDoc->loadXML($xml_data);
- }
- else {
- $xmlDoc->loadHTML($html);
- }
-
- foreach (libxml_get_errors() as $xml_error) {
- $this->log_xml_error($xml_error);
- }
- libxml_clear_errors();
- libxml_use_internal_errors($xml_errors_value);
-
- return $xmlDoc;
- }
-
- /**
- * Convert the site content to RSS.
- */
- public function tweep($src_url) {
- $url = parse_url($src_url);
- if (FALSE === $url || empty($url["host"])) {
- trigger_error("Invalid URL: $src_url", E_USER_ERROR);
- return NULL;
- }
-
- // Strip the leading www. to be more forgiving on input URLs.
- $host = preg_replace('/^www\./', '', $url["host"]);
-
- $xsltProcessor = $this->load_stylesheet($host);
- if (NULL === $xsltProcessor) {
- return NULL;
- }
-
- $html = $this->get_contents($src_url);
- if (FALSE === $html) {
- return NULL;
- }
-
- $preprocess_html_host_method = 'preprocess_html_' . str_replace(".", "_", $host);
- if (method_exists($this, $preprocess_html_host_method)) {
- $html = call_user_func_array(array($this, $preprocess_html_host_method), array($html));
- }
-
- $xmlDoc = $this->html_to_xml($html, $host);
- if (NULL === $xmlDoc) {
- return NULL;
- }
-
- $output = $xsltProcessor->transformToXML($xmlDoc);
-
- if (FALSE === $output) {
- trigger_error('XSL transformation failed.', E_USER_ERROR);
- return NULL;
- }
- return $output;
- }
-
-}
-
-/**
projects / tweeper.git / blobdiff