7 * Tweeper - a Twitter to RSS web scraper.
9 * Copyright (C) 2013-2020 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";
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).
56 * @param bool $verbose_output
57 * Enables showing non-fatal errors like XML parsing errors.
59 public function __construct($generate_enclosure = FALSE, $show_usernames = TRUE, $show_multimedia = TRUE, $verbose_output = TRUE) {
60 $this->generate_enclosure = $generate_enclosure;
61 $this->show_usernames = $show_usernames;
62 $this->show_multimedia = $show_multimedia;
63 $this->verbose_output = $verbose_output;
67 * Convert numeric Epoch to the date format expected in a RSS document.
69 public static function epochToRssDate($timestamp) {
70 if (!is_numeric($timestamp) || is_nan($timestamp)) {
74 return gmdate(DATE_RSS, $timestamp);
78 * Convert generic date string to the date format expected in a RSS document.
80 public static function strToRssDate($date) {
81 $timestamp = strtotime($date);
82 if (FALSE === $timestamp) {
86 return Tweeper::epochToRssDate($timestamp);
90 * Convert string to UpperCamelCase.
92 public static function toUpperCamelCase($str, $delim = ' ') {
93 $str_upper = ucwords($str, $delim);
94 $str_camel_case = str_replace($delim, '', $str_upper);
95 return $str_camel_case;
99 * Perform a cURL session multiple times when it fails with a timeout.
101 * @param resource $ch
102 * a cURL session handle.
104 private static function curlExec($ch) {
108 $ret = curl_exec($ch);
109 if (FALSE === $ret) {
110 trigger_error(curl_error($ch), E_USER_WARNING);
112 } while (curl_errno($ch) == CURLE_OPERATION_TIMEDOUT && ++$attempt < Tweeper::$maxConnectionRetries);
118 * Get the contents from a URL.
120 private static function getUrlContents($url, $user_agent = NULL) {
121 $ch = curl_init($url);
122 curl_setopt_array($ch, [
123 CURLOPT_HEADER => FALSE,
124 CURLOPT_CONNECTTIMEOUT => Tweeper::$maxConnectionTimeout,
125 // Follow http redirects to get the real URL.
126 CURLOPT_FOLLOWLOCATION => TRUE,
127 CURLOPT_COOKIEFILE => "",
128 CURLOPT_RETURNTRANSFER => TRUE,
129 CURLOPT_HTTPHEADER => ['Accept-language: en'],
130 CURLOPT_USERAGENT => isset($user_agent) ? $user_agent : Tweeper::$userAgent,
132 $contents = Tweeper::curlExec($ch);
139 * Get the headers from a URL.
141 private static function getUrlInfo($url, $user_agent = NULL) {
142 $ch = curl_init($url);
143 curl_setopt_array($ch, [
144 CURLOPT_HEADER => TRUE,
145 CURLOPT_NOBODY => TRUE,
146 CURLOPT_CONNECTTIMEOUT => Tweeper::$maxConnectionTimeout,
147 // Follow http redirects to get the real URL.
148 CURLOPT_FOLLOWLOCATION => TRUE,
149 CURLOPT_RETURNTRANSFER => TRUE,
150 CURLOPT_USERAGENT => isset($user_agent) ? $user_agent : Tweeper::$userAgent,
153 $ret = Tweeper::curlExec($ch);
154 if (FALSE === $ret) {
159 $url_info = curl_getinfo($ch);
160 if (FALSE === $url_info) {
161 trigger_error(curl_error($ch), E_USER_WARNING);
169 * Generate an RSS <enclosure/> element.
171 public static function generateEnclosure($url) {
172 $supported_content_types = [
173 "application/octet-stream",
193 $url_info = Tweeper::getUrlInfo($url);
194 if (FALSE === $url_info) {
195 trigger_error("Failed to retrieve info for URL: " . $url, E_USER_WARNING);
199 $supported = in_array($url_info['content_type'], $supported_content_types);
201 trigger_error("Unsupported enclosure content type \"" . $url_info['content_type'] . "\" for URL: " . $url_info['url'], E_USER_WARNING);
205 // The RSS specification says that the enclosure element URL must be http.
206 // See http://sourceforge.net/p/feedvalidator/bugs/72/
207 $http_url = preg_replace("/^https/", "http", $url_info['url']);
209 // When the server does not provide a Content-Length header,
210 // curl_getinfo() would return a negative value for
211 // "download_content_length", however RSS recommends to use 0 when the
212 // enclosure's size cannot be determined.
213 // See: https://www.feedvalidator.org/docs/error/UseZeroForUnknown.html
214 $length = max($url_info['download_content_length'], 0);
216 $dom = new DOMDocument();
217 $enc = $dom->createElement('enclosure');
218 $enc->setAttribute('url', $http_url);
219 $enc->setAttribute('length', $length);
220 $enc->setAttribute('type', $url_info['content_type']);
226 * Mimic the message from libxml.c::php_libxml_ctx_error_level()
228 private static function logXmlError($error) {
231 switch ($error->level) {
232 case LIBXML_ERR_WARNING:
233 $output .= "Warning $error->code: ";
236 case LIBXML_ERR_ERROR:
237 $output .= "Error $error->code: ";
240 case LIBXML_ERR_FATAL:
241 $output .= "Fatal Error $error->code: ";
245 $output .= trim($error->message);
248 $output .= " in $error->file";
251 $output .= " in Entity,";
254 $output .= " line $error->line";
256 trigger_error($output, E_USER_WARNING);
260 * Convert json to XML.
262 private static function jsonToXml($json, $root_node_name) {
263 // Apparently the ObjectNormalizer used afterwards is not able to handle
264 // the stdClass object created by json_decode() with the default setting
265 // $assoc = false; so use $assoc = true.
266 $data = json_decode($json, $assoc = TRUE);
271 $encoder = new XmlEncoder();
272 $normalizer = new ObjectNormalizer();
273 $serializer = new Serializer([$normalizer], [$encoder]);
275 $serializer_options = [
276 'xml_encoding' => "UTF-8",
277 'xml_format_output' => TRUE,
278 'xml_root_node_name' => $root_node_name,
281 $xml_data = $serializer->serialize($data, 'xml', $serializer_options);
283 trigger_error("Cannot serialize data", E_USER_WARNING);
291 * Convert the Instagram content to XML.
293 private function getXmlInstagramCom($html) {
294 // Extract the json data from the html code.
295 $json_match_expr = '/window._sharedData = (.*);/';
296 $ret = preg_match($json_match_expr, $html, $matches);
298 trigger_error("Cannot match expression: $json_match_expr\n", E_USER_WARNING);
302 $data = json_decode($matches[1], $assoc = TRUE);
304 // The "qe" object contains elements which will result in invalid XML
305 // element names, so remove it.
308 // The "knobs" object contains elements with undefined namespaces, so
309 // remove it to silence an error message.
310 unset($data["knobs"]);
312 $json = json_encode($data);
314 return Tweeper::jsonToXml($json, 'instagram');
318 * Make the Facebook HTML processable.
320 private function preprocessHtmlFacebookCom($html) {
321 $html = str_replace('<!--', '', $html);
322 $html = str_replace('-->', '', $html);
327 * Convert the HTML retrieved from the site to XML.
329 private function htmlToXml($html, $host) {
330 $xmlDoc = new DOMDocument();
332 // Handle warnings and errors when loading invalid HTML.
333 $xml_errors_value = libxml_use_internal_errors(TRUE);
335 // If there is a host-specific method to get the XML data, use it!
336 $get_xml_host_method = 'getXml' . Tweeper::toUpperCamelCase($host, '.');
337 if (method_exists($this, $get_xml_host_method)) {
338 $xml_data = call_user_func_array([$this, $get_xml_host_method], [$html]);
339 $xmlDoc->loadXML($xml_data);
342 $xmlDoc->loadHTML($html);
345 if ($this->verbose_output) {
346 foreach (libxml_get_errors() as $xml_error) {
347 Tweeper::logXmlError($xml_error);
350 libxml_clear_errors();
351 libxml_use_internal_errors($xml_errors_value);
357 * Load a stylesheet if the web site is supported.
359 private function loadStylesheet($host) {
360 $stylesheet = "file://" . __DIR__ . "/rss_converter_" . $host . ".xsl";
361 if (FALSE === file_exists($stylesheet)) {
362 trigger_error("Conversion to RSS not supported for $host ($stylesheet not found)", E_USER_WARNING);
366 $stylesheet_contents = file_get_contents($stylesheet);
367 if (FALSE === $stylesheet_contents) {
368 trigger_error("Cannot open $stylesheet", E_USER_WARNING);
372 $xslDoc = new DOMDocument();
373 $xslDoc->loadXML($stylesheet_contents);
375 $xsltProcessor = new XSLTProcessor();
376 $xsltProcessor->registerPHPFunctions();
377 $xsltProcessor->setParameter('', 'generate-enclosure', $this->generate_enclosure);
378 $xsltProcessor->setParameter('', 'show-usernames', $this->show_usernames);
379 $xsltProcessor->setParameter('', 'show-multimedia', $this->show_multimedia);
380 $xsltProcessor->importStylesheet($xslDoc);
382 return $xsltProcessor;
386 * Convert the site content to RSS.
388 public function tweep($src_url, $host = NULL, $validate_scheme = TRUE) {
389 $url = parse_url($src_url);
390 if (FALSE === $url) {
391 trigger_error("Invalid URL: $src_url", E_USER_WARNING);
395 if (TRUE === $validate_scheme) {
396 $scheme = $url["scheme"];
397 if (!in_array($scheme, ["http", "https"])) {
398 trigger_error("unsupported scheme: $scheme", E_USER_WARNING);
403 // If the host is not given derive it from the URL.
404 if (NULL === $host) {
405 if (empty($url["host"])) {
406 trigger_error("Invalid host in URL: $src_url", E_USER_WARNING);
409 // Strip the leading www. to be more forgiving on input URLs.
410 $host = preg_replace('/^www\./', '', $url["host"]);
413 $xsltProcessor = $this->loadStylesheet($host);
414 if (NULL === $xsltProcessor) {
418 $html = Tweeper::getUrlContents($src_url);
419 if (FALSE === $html) {
420 trigger_error("Failed to retrieve $src_url", E_USER_WARNING);
424 $preprocess_html_host_method = 'preprocessHtml' . Tweeper::toUpperCamelCase($host, '.');
425 if (method_exists($this, $preprocess_html_host_method)) {
426 $html = call_user_func_array([$this, $preprocess_html_host_method], [$html]);
429 $xmlDoc = $this->htmlToXml($html, $host);
430 if (NULL === $xmlDoc) {
434 $output = $xsltProcessor->transformToXML($xmlDoc);
435 if (FALSE === $output) {
436 trigger_error('XSL transformation failed.', E_USER_WARNING);