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 = "APIs-Google (+https://developers.google.com/webmasters/APIs-Google.html)";
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);
114 $response_code = curl_getinfo($ch, CURLINFO_RESPONSE_CODE);
115 if (FALSE === $response_code) {
116 trigger_error(curl_error($ch), E_USER_WARNING);
120 if ($response_code >= 400) {
121 trigger_error("HTTP reponse code $response_code", E_USER_WARNING);
129 * Get the contents from a URL.
131 private static function getUrlContents($url, $user_agent = NULL) {
132 $ch = curl_init($url);
133 curl_setopt_array($ch, [
134 CURLOPT_HEADER => FALSE,
135 CURLOPT_CONNECTTIMEOUT => Tweeper::$maxConnectionTimeout,
136 // Follow http redirects to get the real URL.
137 CURLOPT_FOLLOWLOCATION => TRUE,
138 CURLOPT_COOKIEFILE => "",
139 CURLOPT_RETURNTRANSFER => TRUE,
140 CURLOPT_HTTPHEADER => ['Accept-language: en'],
141 CURLOPT_USERAGENT => isset($user_agent) ? $user_agent : Tweeper::$userAgent,
143 $contents = Tweeper::curlExec($ch);
150 * Get the headers from a URL.
152 private static function getUrlInfo($url, $user_agent = NULL) {
153 $ch = curl_init($url);
154 curl_setopt_array($ch, [
155 CURLOPT_HEADER => TRUE,
156 CURLOPT_NOBODY => TRUE,
157 CURLOPT_CONNECTTIMEOUT => Tweeper::$maxConnectionTimeout,
158 // Follow http redirects to get the real URL.
159 CURLOPT_FOLLOWLOCATION => TRUE,
160 CURLOPT_RETURNTRANSFER => TRUE,
161 CURLOPT_USERAGENT => isset($user_agent) ? $user_agent : Tweeper::$userAgent,
164 $ret = Tweeper::curlExec($ch);
165 if (FALSE === $ret) {
170 $url_info = curl_getinfo($ch);
171 if (FALSE === $url_info) {
172 trigger_error(curl_error($ch), E_USER_WARNING);
180 * Generate an RSS <enclosure/> element.
182 public static function generateEnclosure($url) {
183 $supported_content_types = [
184 "application/octet-stream",
204 $url_info = Tweeper::getUrlInfo($url);
205 if (FALSE === $url_info) {
206 trigger_error("Failed to retrieve info for URL: " . $url, E_USER_WARNING);
210 $supported = in_array($url_info['content_type'], $supported_content_types);
212 trigger_error("Unsupported enclosure content type \"" . $url_info['content_type'] . "\" for URL: " . $url_info['url'], E_USER_WARNING);
216 // The RSS specification says that the enclosure element URL must be http.
217 // See http://sourceforge.net/p/feedvalidator/bugs/72/
218 $http_url = preg_replace("/^https/", "http", $url_info['url']);
220 // When the server does not provide a Content-Length header,
221 // curl_getinfo() would return a negative value for
222 // "download_content_length", however RSS recommends to use 0 when the
223 // enclosure's size cannot be determined.
224 // See: https://www.feedvalidator.org/docs/error/UseZeroForUnknown.html
225 $length = max($url_info['download_content_length'], 0);
227 $dom = new DOMDocument();
228 $enc = $dom->createElement('enclosure');
229 $enc->setAttribute('url', $http_url);
230 $enc->setAttribute('length', $length);
231 $enc->setAttribute('type', $url_info['content_type']);
237 * Mimic the message from libxml.c::php_libxml_ctx_error_level()
239 private static function logXmlError($error) {
242 switch ($error->level) {
243 case LIBXML_ERR_WARNING:
244 $output .= "Warning $error->code: ";
247 case LIBXML_ERR_ERROR:
248 $output .= "Error $error->code: ";
251 case LIBXML_ERR_FATAL:
252 $output .= "Fatal Error $error->code: ";
256 $output .= trim($error->message);
259 $output .= " in $error->file";
262 $output .= " in Entity,";
265 $output .= " line $error->line";
267 trigger_error($output, E_USER_WARNING);
271 * Convert json to XML.
273 private static function jsonToXml($json, $root_node_name) {
274 // Apparently the ObjectNormalizer used afterwards is not able to handle
275 // the stdClass object created by json_decode() with the default setting
276 // $assoc = false; so use $assoc = true.
277 $data = json_decode($json, $assoc = TRUE);
282 $encoder = new XmlEncoder();
283 $normalizer = new ObjectNormalizer();
284 $serializer = new Serializer([$normalizer], [$encoder]);
286 $serializer_options = [
287 'xml_encoding' => "UTF-8",
288 'xml_format_output' => TRUE,
289 'xml_root_node_name' => $root_node_name,
292 $xml_data = $serializer->serialize($data, 'xml', $serializer_options);
294 trigger_error("Cannot serialize data", E_USER_WARNING);
302 * Convert the Instagram content to XML.
304 private function getXmlInstagramCom($html) {
305 // Extract the json data from the html code.
306 $json_match_expr = '/window._sharedData = (.*);/';
307 $ret = preg_match($json_match_expr, $html, $matches);
309 trigger_error("Cannot match expression: $json_match_expr\n", E_USER_WARNING);
313 $data = json_decode($matches[1], $assoc = TRUE);
315 // The "qe" object contains elements which will result in invalid XML
316 // element names, so remove it.
319 // The "knobs" object contains elements with undefined namespaces, so
320 // remove it to silence an error message.
321 unset($data["knobs"]);
323 $json = json_encode($data);
325 return Tweeper::jsonToXml($json, 'instagram');
329 * Make the Facebook HTML processable.
331 private function preprocessHtmlFacebookCom($html) {
332 $html = str_replace('<!--', '', $html);
333 $html = str_replace('-->', '', $html);
338 * Convert the HTML retrieved from the site to XML.
340 private function htmlToXml($html, $host) {
341 $xmlDoc = new DOMDocument();
343 // Handle warnings and errors when loading invalid HTML.
344 $xml_errors_value = libxml_use_internal_errors(TRUE);
346 // If there is a host-specific method to get the XML data, use it!
347 $get_xml_host_method = 'getXml' . Tweeper::toUpperCamelCase($host, '.');
348 if (method_exists($this, $get_xml_host_method)) {
349 $xml_data = call_user_func_array([$this, $get_xml_host_method], [$html]);
350 if (NULL === $xml_data) {
353 $xmlDoc->loadXML($xml_data);
356 $xmlDoc->loadHTML($html);
359 if ($this->verbose_output) {
360 foreach (libxml_get_errors() as $xml_error) {
361 Tweeper::logXmlError($xml_error);
364 libxml_clear_errors();
365 libxml_use_internal_errors($xml_errors_value);
371 * Load a stylesheet if the web site is supported.
373 private function loadStylesheet($host) {
374 $stylesheet = "file://" . __DIR__ . "/rss_converter_" . $host . ".xsl";
375 if (FALSE === file_exists($stylesheet)) {
376 trigger_error("Conversion to RSS not supported for $host ($stylesheet not found)", E_USER_WARNING);
380 $stylesheet_contents = file_get_contents($stylesheet);
381 if (FALSE === $stylesheet_contents) {
382 trigger_error("Cannot open $stylesheet", E_USER_WARNING);
386 $xslDoc = new DOMDocument();
387 $xslDoc->loadXML($stylesheet_contents);
389 $xsltProcessor = new XSLTProcessor();
390 $xsltProcessor->registerPHPFunctions();
391 $xsltProcessor->setParameter('', 'generate-enclosure', $this->generate_enclosure);
392 $xsltProcessor->setParameter('', 'show-usernames', $this->show_usernames);
393 $xsltProcessor->setParameter('', 'show-multimedia', $this->show_multimedia);
394 $xsltProcessor->importStylesheet($xslDoc);
396 return $xsltProcessor;
400 * Convert the site content to RSS.
402 public function tweep($src_url, $host = NULL, $validate_scheme = TRUE) {
403 $url = parse_url($src_url);
404 if (FALSE === $url) {
405 trigger_error("Invalid URL: $src_url", E_USER_WARNING);
409 if (TRUE === $validate_scheme) {
410 $scheme = $url["scheme"];
411 if (!in_array($scheme, ["http", "https"])) {
412 trigger_error("unsupported scheme: $scheme", E_USER_WARNING);
417 // If the host is not given derive it from the URL.
418 if (NULL === $host) {
419 if (empty($url["host"])) {
420 trigger_error("Invalid host in URL: $src_url", E_USER_WARNING);
423 // Strip the leading www. to be more forgiving on input URLs.
424 $host = preg_replace('/^www\./', '', $url["host"]);
427 $xsltProcessor = $this->loadStylesheet($host);
428 if (NULL === $xsltProcessor) {
432 $html = Tweeper::getUrlContents($src_url);
433 if (FALSE === $html) {
434 trigger_error("Failed to retrieve $src_url", E_USER_WARNING);
438 $preprocess_html_host_method = 'preprocessHtml' . Tweeper::toUpperCamelCase($host, '.');
439 if (method_exists($this, $preprocess_html_host_method)) {
440 $html = call_user_func_array([$this, $preprocess_html_host_method], [$html]);
443 $xmlDoc = $this->htmlToXml($html, $host);
444 if (NULL === $xmlDoc) {
448 $output = $xsltProcessor->transformToXML($xmlDoc);
449 if (FALSE === $output) {
450 trigger_error('XSL transformation failed.', E_USER_WARNING);