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 (Linux; U; Android 4.3; en-us; SM-N900T Build/JSS15J)";
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 Twitter mobile date to the date format expected in a RSS document.
92 public static function twitterToRssDate($date) {
93 // Twitter uses relative timestamps in minutes for recent tweets.
94 if (preg_match('/^(\d+)m$/', $date, $matches)) {
95 $timestamp = strtotime("+" . $matches[1] . " min", time());
96 if (FALSE === $timestamp) {
102 * In case the time is specified put it after the date,
103 * to make it recognized by strptime().
105 if (preg_match('/(.*) - (.*)/', $date, $matches)) {
106 $date = $matches[2] . " " . $matches[1];
109 $timestamp = strtotime($date);
110 if (FALSE === $timestamp) {
115 * The twitter mobile UI usually only specifies the month and the day, so
116 * strtotime($date) may interpret the date as being in the future.
118 * If the date is in the future it is probably in the same day but in the
121 if ($timestamp > time()) {
122 $timestamp = strtotime('-1 years', $timestamp);
126 return Tweeper::epochToRssDate($timestamp);
130 * Convert string to UpperCamelCase.
132 public static function toUpperCamelCase($str, $delim = ' ') {
133 $str_upper = ucwords($str, $delim);
134 $str_camel_case = str_replace($delim, '', $str_upper);
135 return $str_camel_case;
139 * Perform a cURL session multiple times when it fails with a timeout.
141 * @param resource $ch
142 * a cURL session handle.
144 private static function curlExec($ch) {
148 $ret = curl_exec($ch);
149 if (FALSE === $ret) {
150 trigger_error(curl_error($ch), E_USER_WARNING);
152 } while (curl_errno($ch) == CURLE_OPERATION_TIMEDOUT && ++$attempt < Tweeper::$maxConnectionRetries);
158 * Get the contents from a URL.
160 private static function getUrlContents($url) {
161 $ch = curl_init($url);
162 curl_setopt_array($ch, [
163 CURLOPT_HEADER => FALSE,
164 CURLOPT_CONNECTTIMEOUT => Tweeper::$maxConnectionTimeout,
165 // Follow http redirects to get the real URL.
166 CURLOPT_FOLLOWLOCATION => TRUE,
167 CURLOPT_COOKIEFILE => "",
168 CURLOPT_RETURNTRANSFER => TRUE,
169 CURLOPT_HTTPHEADER => ['Accept-language: en'],
170 CURLOPT_USERAGENT => Tweeper::$userAgent,
172 $contents = Tweeper::curlExec($ch);
179 * Get the headers from a URL.
181 private static function getUrlInfo($url) {
182 $ch = curl_init($url);
183 curl_setopt_array($ch, [
184 CURLOPT_HEADER => TRUE,
185 CURLOPT_NOBODY => TRUE,
186 CURLOPT_CONNECTTIMEOUT => Tweeper::$maxConnectionTimeout,
187 // Follow http redirects to get the real URL.
188 CURLOPT_FOLLOWLOCATION => TRUE,
189 CURLOPT_RETURNTRANSFER => TRUE,
190 CURLOPT_USERAGENT => Tweeper::$userAgent,
193 $ret = Tweeper::curlExec($ch);
194 if (FALSE === $ret) {
199 $url_info = curl_getinfo($ch);
200 if (FALSE === $url_info) {
201 trigger_error(curl_error($ch), E_USER_WARNING);
209 * Generate an RSS <enclosure/> element.
211 public static function generateEnclosure($url) {
212 $supported_content_types = [
213 "application/octet-stream",
233 $url_info = Tweeper::getUrlInfo($url);
234 if (FALSE === $url_info) {
235 trigger_error("Failed to retrieve info for URL: " . $url, E_USER_WARNING);
239 $supported = in_array($url_info['content_type'], $supported_content_types);
241 trigger_error("Unsupported enclosure content type \"" . $url_info['content_type'] . "\" for URL: " . $url_info['url'], E_USER_WARNING);
245 // The RSS specification says that the enclosure element URL must be http.
246 // See http://sourceforge.net/p/feedvalidator/bugs/72/
247 $http_url = preg_replace("/^https/", "http", $url_info['url']);
249 // When the server does not provide a Content-Length header,
250 // curl_getinfo() would return a negative value for
251 // "download_content_length", however RSS recommends to use 0 when the
252 // enclosure's size cannot be determined.
253 // See: https://www.feedvalidator.org/docs/error/UseZeroForUnknown.html
254 $length = max($url_info['download_content_length'], 0);
256 $dom = new DOMDocument();
257 $enc = $dom->createElement('enclosure');
258 $enc->setAttribute('url', $http_url);
259 $enc->setAttribute('length', $length);
260 $enc->setAttribute('type', $url_info['content_type']);
266 * Mimic the message from libxml.c::php_libxml_ctx_error_level()
268 private static function logXmlError($error) {
271 switch ($error->level) {
272 case LIBXML_ERR_WARNING:
273 $output .= "Warning $error->code: ";
276 case LIBXML_ERR_ERROR:
277 $output .= "Error $error->code: ";
280 case LIBXML_ERR_FATAL:
281 $output .= "Fatal Error $error->code: ";
285 $output .= trim($error->message);
288 $output .= " in $error->file";
291 $output .= " in Entity,";
294 $output .= " line $error->line";
296 trigger_error($output, E_USER_WARNING);
300 * Convert json to XML.
302 private static function jsonToXml($json, $root_node_name) {
303 // Apparently the ObjectNormalizer used afterwards is not able to handle
304 // the stdClass object created by json_decode() with the default setting
305 // $assoc = false; so use $assoc = true.
306 $data = json_decode($json, $assoc = TRUE);
311 $encoder = new XmlEncoder();
312 $normalizer = new ObjectNormalizer();
313 $serializer = new Serializer([$normalizer], [$encoder]);
315 $serializer_options = [
316 'xml_encoding' => "UTF-8",
317 'xml_format_output' => TRUE,
318 'xml_root_node_name' => $root_node_name,
321 $xml_data = $serializer->serialize($data, 'xml', $serializer_options);
323 trigger_error("Cannot serialize data", E_USER_WARNING);
331 * Convert the Instagram content to XML.
333 private function getXmlInstagramCom($html) {
334 // Extract the json data from the html code.
335 $json_match_expr = '/window._sharedData = (.*);/';
336 $ret = preg_match($json_match_expr, $html, $matches);
338 trigger_error("Cannot match expression: $json_match_expr\n", E_USER_WARNING);
342 $data = json_decode($matches[1], $assoc = TRUE);
344 // The "qe" object contains elements which will result in invalid XML
345 // element names, so remove it.
348 // The "knobs" object contains elements with undefined namespaces, so
349 // remove it to silence an error message.
350 unset($data["knobs"]);
352 $json = json_encode($data);
354 return Tweeper::jsonToXml($json, 'instagram');
358 * Make the Facebook HTML processable.
360 private function preprocessHtmlFacebookCom($html) {
361 $html = str_replace('<!--', '', $html);
362 $html = str_replace('-->', '', $html);
367 * Convert the HTML retrieved from the site to XML.
369 private function htmlToXml($html, $host) {
370 $xmlDoc = new DOMDocument();
372 // Handle warnings and errors when loading invalid HTML.
373 $xml_errors_value = libxml_use_internal_errors(TRUE);
375 // If there is a host-specific method to get the XML data, use it!
376 $get_xml_host_method = 'getXml' . Tweeper::toUpperCamelCase($host, '.');
377 if (method_exists($this, $get_xml_host_method)) {
378 $xml_data = call_user_func_array([$this, $get_xml_host_method], [$html]);
379 $xmlDoc->loadXML($xml_data);
382 $xmlDoc->loadHTML($html);
385 if ($this->verbose_output) {
386 foreach (libxml_get_errors() as $xml_error) {
387 Tweeper::logXmlError($xml_error);
390 libxml_clear_errors();
391 libxml_use_internal_errors($xml_errors_value);
397 * Load a stylesheet if the web site is supported.
399 private function loadStylesheet($host) {
400 $stylesheet = "file://" . __DIR__ . "/rss_converter_" . $host . ".xsl";
401 if (FALSE === file_exists($stylesheet)) {
402 trigger_error("Conversion to RSS not supported for $host ($stylesheet not found)", E_USER_WARNING);
406 $stylesheet_contents = Tweeper::getUrlContents($stylesheet);
407 if (FALSE === $stylesheet_contents) {
408 trigger_error("Cannot open $stylesheet", E_USER_WARNING);
412 $xslDoc = new DOMDocument();
413 $xslDoc->loadXML($stylesheet_contents);
415 $xsltProcessor = new XSLTProcessor();
416 $xsltProcessor->registerPHPFunctions();
417 $xsltProcessor->setParameter('', 'generate-enclosure', $this->generate_enclosure);
418 $xsltProcessor->setParameter('', 'show-usernames', $this->show_usernames);
419 $xsltProcessor->setParameter('', 'show-multimedia', $this->show_multimedia);
420 $xsltProcessor->importStylesheet($xslDoc);
422 return $xsltProcessor;
426 * Convert the site content to RSS.
428 public function tweep($src_url, $host = NULL, $validate_scheme = TRUE) {
429 $url = parse_url($src_url);
430 if (FALSE === $url) {
431 trigger_error("Invalid URL: $src_url", E_USER_WARNING);
435 if (TRUE === $validate_scheme) {
436 $scheme = $url["scheme"];
437 if (!in_array($scheme, ["http", "https"])) {
438 trigger_error("unsupported scheme: $scheme", E_USER_WARNING);
443 // If the host is not given derive it from the URL.
444 if (NULL === $host) {
445 if (empty($url["host"])) {
446 trigger_error("Invalid host in URL: $src_url", E_USER_WARNING);
449 // Strip the leading www. to be more forgiving on input URLs.
450 $host = preg_replace('/^www\./', '', $url["host"]);
453 $xsltProcessor = $this->loadStylesheet($host);
454 if (NULL === $xsltProcessor) {
458 $html = Tweeper::getUrlContents($src_url);
459 if (FALSE === $html) {
460 trigger_error("Failed to retrieve $src_url", E_USER_WARNING);
464 $preprocess_html_host_method = 'preprocessHtml' . Tweeper::toUpperCamelCase($host, '.');
465 if (method_exists($this, $preprocess_html_host_method)) {
466 $html = call_user_func_array([$this, $preprocess_html_host_method], [$html]);
469 $xmlDoc = $this->htmlToXml($html, $host);
470 if (NULL === $xmlDoc) {
474 $output = $xsltProcessor->transformToXML($xmlDoc);
475 if (FALSE === $output) {
476 trigger_error('XSL transformation failed.', E_USER_WARNING);