7 * Tweeper - a Twitter to RSS web scraper.
9 * Copyright (C) 2013-2018 Antonio Ospite <ao2@ao2.it>
11 * This program is free software: you can redistribute it and/or modify
12 * it under the terms of the GNU General Public License as published by
13 * the Free Software Foundation, either version 3 of the License, or
14 * (at your option) any later version.
16 * This program is distributed in the hope that it will be useful,
17 * but WITHOUT ANY WARRANTY; without even the implied warranty of
18 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
19 * GNU General Public License for more details.
21 * You should have received a copy of the GNU General Public License
22 * along with this program. If not, see <http://www.gnu.org/licenses/>.
28 use Symfony\Component\Serializer\Serializer;
29 use Symfony\Component\Serializer\Encoder\XmlEncoder;
30 use Symfony\Component\Serializer\Normalizer\ObjectNormalizer;
32 date_default_timezone_set('UTC');
35 * Scrape supported websites and perform conversion to RSS.
39 private static $userAgent = "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:60.0) Gecko/20100101 Firefox/60.0";
40 private static $maxConnectionTimeout = 5;
41 private static $maxConnectionRetries = 5;
44 * Create a new Tweeper object controlling optional settings.
46 * @param bool $generate_enclosure
47 * Enables the creation of <enclosure/> elements (disabled by default).
48 * @param bool $show_usernames
49 * Enables showing the username in front of the content for multi-user
50 * sites (enabled by default). Only some stylesheets supports this
51 * functionality (twitter, instagram, pump.io).
52 * @param bool $show_multimedia
53 * Enables showing multimedia content (images, videos) directly in the
54 * item description (enabled by default). Only some stylesheets supports
55 * this functionality (twitter, instagram, dilbert).
57 public function __construct($generate_enclosure = FALSE, $show_usernames = TRUE, $show_multimedia = TRUE) {
58 $this->generate_enclosure = $generate_enclosure;
59 $this->show_usernames = $show_usernames;
60 $this->show_multimedia = $show_multimedia;
64 * Convert numeric Epoch to the date format expected in a RSS document.
66 public static function epochToRssDate($timestamp) {
67 if (!is_numeric($timestamp) || is_nan($timestamp)) {
71 return gmdate(DATE_RSS, $timestamp);
75 * Convert generic date string to the date format expected in a RSS document.
77 public static function strToRssDate($date) {
78 $timestamp = strtotime($date);
79 if (FALSE === $timestamp) {
83 return Tweeper::epochToRssDate($timestamp);
87 * Convert string to UpperCamelCase.
89 public static function toUpperCamelCase($str, $delim = ' ') {
90 $str_upper = ucwords($str, $delim);
91 $str_camel_case = str_replace($delim, '', $str_upper);
92 return $str_camel_case;
96 * Perform a cURL session multiple times when it fails with a timeout.
99 * a cURL session handle.
101 private static function curlExec($ch) {
105 $ret = curl_exec($ch);
106 if (FALSE === $ret) {
107 trigger_error(curl_error($ch), E_USER_WARNING);
109 } while (curl_errno($ch) == CURLE_OPERATION_TIMEDOUT && ++$attempt < Tweeper::$maxConnectionRetries);
115 * Get the contents from a URL.
117 private static function getUrlContents($url) {
118 $ch = curl_init($url);
119 curl_setopt_array($ch, array(
120 CURLOPT_HEADER => FALSE,
121 CURLOPT_CONNECTTIMEOUT => Tweeper::$maxConnectionTimeout,
122 // Follow http redirects to get the real URL.
123 CURLOPT_FOLLOWLOCATION => TRUE,
124 CURLOPT_COOKIEFILE => "",
125 CURLOPT_RETURNTRANSFER => TRUE,
126 CURLOPT_SSL_VERIFYHOST => FALSE,
127 CURLOPT_SSL_VERIFYPEER => FALSE,
128 CURLOPT_HTTPHEADER => array('Accept-language: en'),
129 CURLOPT_USERAGENT => Tweeper::$userAgent,
131 $contents = Tweeper::curlExec($ch);
138 * Get the headers from a URL.
140 private static function getUrlInfo($url) {
141 $ch = curl_init($url);
142 curl_setopt_array($ch, array(
143 CURLOPT_HEADER => TRUE,
144 CURLOPT_NOBODY => TRUE,
145 CURLOPT_CONNECTTIMEOUT => Tweeper::$maxConnectionTimeout,
146 // Follow http redirects to get the real URL.
147 CURLOPT_FOLLOWLOCATION => TRUE,
148 CURLOPT_RETURNTRANSFER => TRUE,
149 CURLOPT_SSL_VERIFYHOST => FALSE,
150 CURLOPT_SSL_VERIFYPEER => FALSE,
151 CURLOPT_USERAGENT => Tweeper::$userAgent,
154 $ret = Tweeper::curlExec($ch);
155 if (FALSE === $ret) {
160 $url_info = curl_getinfo($ch);
161 if (FALSE === $url_info) {
162 trigger_error(curl_error($ch), E_USER_WARNING);
170 * Generate an RSS <enclosure/> element.
172 public static function generateEnclosure($url) {
173 $supported_content_types = array(
174 "application/octet-stream",
194 $url_info = Tweeper::getUrlInfo($url);
195 if (FALSE === $url_info) {
196 trigger_error("Failed to retrieve info for URL: " . $url, E_USER_WARNING);
200 $supported = in_array($url_info['content_type'], $supported_content_types);
202 trigger_error("Unsupported enclosure content type \"" . $url_info['content_type'] . "\" for URL: " . $url_info['url'], E_USER_WARNING);
206 // The RSS specification says that the enclosure element URL must be http.
207 // See http://sourceforge.net/p/feedvalidator/bugs/72/
208 $http_url = preg_replace("/^https/", "http", $url_info['url']);
210 // When the server does not provide a Content-Length header,
211 // curl_getinfo() would return a negative value for
212 // "download_content_length", however RSS recommends to use 0 when the
213 // enclosure's size cannot be determined.
214 // See: https://www.feedvalidator.org/docs/error/UseZeroForUnknown.html
215 $length = max($url_info['download_content_length'], 0);
217 $dom = new DOMDocument();
218 $enc = $dom->createElement('enclosure');
219 $enc->setAttribute('url', $http_url);
220 $enc->setAttribute('length', $length);
221 $enc->setAttribute('type', $url_info['content_type']);
227 * Mimic the message from libxml.c::php_libxml_ctx_error_level()
229 private static function logXmlError($error) {
232 switch ($error->level) {
233 case LIBXML_ERR_WARNING:
234 $output .= "Warning $error->code: ";
237 case LIBXML_ERR_ERROR:
238 $output .= "Error $error->code: ";
241 case LIBXML_ERR_FATAL:
242 $output .= "Fatal Error $error->code: ";
246 $output .= trim($error->message);
249 $output .= " in $error->file";
252 $output .= " in Entity,";
255 $output .= " line $error->line";
257 trigger_error($output, E_USER_WARNING);
261 * Convert json to XML.
263 private static function jsonToXml($json, $root_node_name) {
264 // Apparently the ObjectNormalizer used afterwards is not able to handle
265 // the stdClass object created by json_decode() with the default setting
266 // $assoc = false; so use $assoc = true.
267 $data = json_decode($json, $assoc = TRUE);
272 $encoder = new XmlEncoder();
273 $normalizer = new ObjectNormalizer();
274 $serializer = new Serializer(array($normalizer), array($encoder));
276 $serializer_options = array(
277 'xml_encoding' => "UTF-8",
278 'xml_format_output' => TRUE,
279 'xml_root_node_name' => $root_node_name,
282 $xml_data = $serializer->serialize($data, 'xml', $serializer_options);
284 trigger_error("Cannot serialize data", E_USER_WARNING);
292 * Convert the Instagram content to XML.
294 private function getXmlInstagramCom($html) {
295 // Extract the json data from the html code.
296 $json_match_expr = '/window._sharedData = (.*);/';
297 $ret = preg_match($json_match_expr, $html, $matches);
299 trigger_error("Cannot match expression: $json_match_expr\n", E_USER_WARNING);
303 $data = json_decode($matches[1], $assoc = TRUE);
305 // The "qe" object contains elements which will result in invalid XML
306 // element names, so remove it.
309 // The "knobs" object contains elements with undefined namespaces, so
310 // remove it to silence an error message.
311 unset($data["knobs"]);
313 $json = json_encode($data);
315 return Tweeper::jsonToXml($json, 'instagram');
319 * Make the Facebook HTML processable.
321 private function preprocessHtmlFacebookCom($html) {
322 $html = str_replace('<!--', '', $html);
323 $html = str_replace('-->', '', $html);
328 * Convert the HTML retrieved from the site to XML.
330 private function htmlToXml($html, $host) {
331 $xmlDoc = new DOMDocument();
333 // Handle warnings and errors when loading invalid HTML.
334 $xml_errors_value = libxml_use_internal_errors(TRUE);
336 // If there is a host-specific method to get the XML data, use it!
337 $get_xml_host_method = 'getXml' . Tweeper::toUpperCamelCase($host, '.');
338 if (method_exists($this, $get_xml_host_method)) {
339 $xml_data = call_user_func_array(array($this, $get_xml_host_method), array($html));
340 $xmlDoc->loadXML($xml_data);
343 $xmlDoc->loadHTML($html);
346 foreach (libxml_get_errors() as $xml_error) {
347 Tweeper::logXmlError($xml_error);
349 libxml_clear_errors();
350 libxml_use_internal_errors($xml_errors_value);
356 * Load a stylesheet if the web site is supported.
358 private function loadStylesheet($host) {
359 $stylesheet = "file://" . __DIR__ . "/rss_converter_" . $host . ".xsl";
360 if (FALSE === file_exists($stylesheet)) {
361 trigger_error("Conversion to RSS not supported for $host ($stylesheet not found)", E_USER_WARNING);
365 $stylesheet_contents = Tweeper::getUrlContents($stylesheet);
366 if (FALSE === $stylesheet_contents) {
367 trigger_error("Cannot open $stylesheet", E_USER_WARNING);
371 $xslDoc = new DOMDocument();
372 $xslDoc->loadXML($stylesheet_contents);
374 $xsltProcessor = new XSLTProcessor();
375 $xsltProcessor->registerPHPFunctions();
376 $xsltProcessor->setParameter('', 'generate-enclosure', $this->generate_enclosure);
377 $xsltProcessor->setParameter('', 'show-usernames', $this->show_usernames);
378 $xsltProcessor->setParameter('', 'show-multimedia', $this->show_multimedia);
379 $xsltProcessor->importStylesheet($xslDoc);
381 return $xsltProcessor;
385 * Convert the site content to RSS.
387 public function tweep($src_url, $host = NULL, $validate_scheme = TRUE) {
388 $url = parse_url($src_url);
389 if (FALSE === $url) {
390 trigger_error("Invalid URL: $src_url", E_USER_WARNING);
394 if (TRUE === $validate_scheme) {
395 $scheme = $url["scheme"];
396 if (!in_array($scheme, array("http", "https"))) {
397 trigger_error("unsupported scheme: $scheme", E_USER_WARNING);
402 // If the host is not given derive it from the URL.
403 if (NULL === $host) {
404 if (empty($url["host"])) {
405 trigger_error("Invalid host in URL: $src_url", E_USER_WARNING);
408 // Strip the leading www. to be more forgiving on input URLs.
409 $host = preg_replace('/^www\./', '', $url["host"]);
412 $xsltProcessor = $this->loadStylesheet($host);
413 if (NULL === $xsltProcessor) {
417 $html = Tweeper::getUrlContents($src_url);
418 if (FALSE === $html) {
419 trigger_error("Failed to retrieve $src_url", E_USER_WARNING);
423 $preprocess_html_host_method = 'preprocessHtml' . Tweeper::toUpperCamelCase($host, '.');
424 if (method_exists($this, $preprocess_html_host_method)) {
425 $html = call_user_func_array(array($this, $preprocess_html_host_method), array($html));
428 $xmlDoc = $this->htmlToXml($html, $host);
429 if (NULL === $xmlDoc) {
433 $output = $xsltProcessor->transformToXML($xmlDoc);
434 if (FALSE === $output) {
435 trigger_error('XSL transformation failed.', E_USER_WARNING);