src/Util/ParseUrl.php

   1 <?php
   2 /**
   3  * @copyright Copyright (C) 2010-2022, the Friendica project
   4  *
   5  * @license GNU AGPL version 3 or any later version
   6  *
   7  * This program is free software: you can redistribute it and/or modify
   8  * it under the terms of the GNU Affero General Public License as
   9  * published by the Free Software Foundation, either version 3 of the
  10  * License, or (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15  * GNU Affero General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU Affero General Public License
  18  * along with this program.  If not, see <https://www.gnu.org/licenses/>.
  19  *
  20  */
  21
  22 namespace Friendica\Util;
  23
  24 use DOMDocument;
  25 use DOMXPath;
  26 use Friendica\Content\OEmbed;
  27 use Friendica\Core\Hook;
  28 use Friendica\Core\Logger;
  29 use Friendica\Database\Database;
  30 use Friendica\Database\DBA;
  31 use Friendica\DI;
  32 use Friendica\Network\HTTPClient\Client\HttpClient;
  33 use Friendica\Network\HTTPException;
  34 use Friendica\Network\HTTPClient\Client\HttpClientOptions;
  35
  36 /**
  37  * Get information about a given URL
  38  *
  39  * Class with methods for extracting certain content from an url
  40  */
  41 class ParseUrl
  42 {
  43         const DEFAULT_EXPIRATION_FAILURE = 'now + 1 day';
  44         const DEFAULT_EXPIRATION_SUCCESS = 'now + 3 months';
  45
  46         /**
  47          * Maximum number of characters for the description
  48          */
  49         const MAX_DESC_COUNT = 250;
  50
  51         /**
  52          * Minimum number of characters for the description
  53          */
  54         const MIN_DESC_COUNT = 100;
  55
  56         /**
  57          * Fetch the content type of the given url
  58          * @param string $url    URL of the page
  59          * @param string $accept content-type to accept
  60          * @return array content type
  61          */
  62         public static function getContentType(string $url, string $accept = HttpClient::ACCEPT_DEFAULT)
  63         {
  64                 $curlResult = DI::httpClient()->head($url, [HttpClientOptions::ACCEPT_CONTENT => $accept]);
  65
  66                 // Workaround for systems that can't handle a HEAD request
  67                 if (!$curlResult->isSuccess() && ($curlResult->getReturnCode() == 405)) {
  68                         $curlResult = DI::httpClient()->get($url, [HttpClientOptions::CONTENT_LENGTH => 1000000, HttpClientOptions::ACCEPT_CONTENT => $accept]);
  69                 }
  70
  71                 if (!$curlResult->isSuccess()) {
  72                         return [];
  73                 }
  74
  75                 $contenttype =  $curlResult->getHeader('Content-Type')[0] ?? '';
  76                 if (empty($contenttype)) {
  77                         return ['application', 'octet-stream'];
  78                 }
  79
  80                 return explode('/', current(explode(';', $contenttype)));
  81         }
  82
  83         /**
  84          * Search for chached embeddable data of an url otherwise fetch it
  85          *
  86          * @param string $url         The url of the page which should be scraped
  87          * @param bool   $do_oembed   The false option is used by the function fetch_oembed()
  88          *                            to avoid endless loops
  89          *
  90          * @return array which contains needed data for embedding
  91          *    string 'url'      => The url of the parsed page
  92          *    string 'type'     => Content type
  93          *    string 'title'    => (optional) The title of the content
  94          *    string 'text'     => (optional) The description for the content
  95          *    string 'image'    => (optional) A preview image of the content
  96          *    array  'images'   => (optional) Array of preview pictures
  97          *    string 'keywords' => (optional) The tags which belong to the content
  98          *
  99          * @throws HTTPException\InternalServerErrorException
 100          * @see   ParseUrl::getSiteinfo() for more information about scraping
 101          * embeddable content
 102          */
 103         public static function getSiteinfoCached($url, $do_oembed = true): array
 104         {
 105                 if (empty($url)) {
 106                         return [
 107                                 'url' => '',
 108                                 'type' => 'error',
 109                         ];
 110                 }
 111
 112                 $urlHash = hash('sha256', $url);
 113
 114                 $parsed_url = DBA::selectFirst('parsed_url', ['content'],
 115                         ['url_hash' => $urlHash, 'oembed' => $do_oembed]
 116                 );
 117                 if (!empty($parsed_url['content'])) {
 118                         $data = unserialize($parsed_url['content']);
 119                         return $data;
 120                 }
 121
 122                 $data = self::getSiteinfo($url, $do_oembed);
 123
 124                 $expires = $data['expires'];
 125
 126                 unset($data['expires']);
 127
 128                 DI::dba()->insert(
 129                         'parsed_url',
 130                         [
 131                                 'url_hash' => $urlHash,
 132                                 'oembed'   => $do_oembed,
 133                                 'url'      => $url,
 134                                 'content'  => serialize($data),
 135                                 'created'  => DateTimeFormat::utcNow(),
 136                                 'expires'  => $expires,
 137                         ],
 138                         Database::INSERT_UPDATE
 139                 );
 140
 141                 return $data;
 142         }
 143
 144         /**
 145          * Parse a page for embeddable content information
 146          *
 147          * This method parses to url for meta data which can be used to embed
 148          * the content. If available it prioritizes Open Graph meta tags.
 149          * If this is not available it uses the twitter cards meta tags.
 150          * As fallback it uses standard html elements with meta informations
 151          * like \<title\>Awesome Title\</title\> or
 152          * \<meta name="description" content="An awesome description"\>
 153          *
 154          * @param string $url         The url of the page which should be scraped
 155          * @param bool   $do_oembed   The false option is used by the function fetch_oembed()
 156          *                            to avoid endless loops
 157          * @param int    $count       Internal counter to avoid endless loops
 158          *
 159          * @return array which contains needed data for embedding
 160          *    string 'url'      => The url of the parsed page
 161          *    string 'type'     => Content type (error, link, photo, image, audio, video)
 162          *    string 'title'    => (optional) The title of the content
 163          *    string 'text'     => (optional) The description for the content
 164          *    string 'image'    => (optional) A preview image of the content
 165          *    array  'images'   => (optional) Array of preview pictures
 166          *    string 'keywords' => (optional) The tags which belong to the content
 167          *
 168          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 169          * @todo  https://developers.google.com/+/plugins/snippet/
 170          * @verbatim
 171          * <meta itemprop="name" content="Awesome title">
 172          * <meta itemprop="description" content="An awesome description">
 173          * <meta itemprop="image" content="http://maple.libertreeproject.org/images/tree-icon.png">
 174          *
 175          * <body itemscope itemtype="http://schema.org/Product">
 176          *   <h1 itemprop="name">Shiny Trinket</h1>
 177          *   <img itemprop="image" src="{image-url}" />
 178          *   <p itemprop="description">Shiny trinkets are shiny.</p>
 179          * </body>
 180          * @endverbatim
 181          */
 182         public static function getSiteinfo($url, $do_oembed = true, $count = 1)
 183         {
 184                 if (empty($url)) {
 185                         return [
 186                                 'url' => '',
 187                                 'type' => 'error',
 188                         ];
 189                 }
 190
 191                 // Check if the URL does contain a scheme
 192                 $scheme = parse_url($url, PHP_URL_SCHEME);
 193
 194                 if ($scheme == '') {
 195                         $url = 'http://' . ltrim($url, '/');
 196                 }
 197
 198                 $url = trim($url, "'\"");
 199
 200                 $url = Network::stripTrackingQueryParams($url);
 201
 202                 $siteinfo = [
 203                         'url' => $url,
 204                         'type' => 'link',
 205                         'expires' => DateTimeFormat::utc(self::DEFAULT_EXPIRATION_FAILURE),
 206                 ];
 207
 208                 if ($count > 10) {
 209                         Logger::notice('Endless loop detected', ['url' => $url]);
 210                         return $siteinfo;
 211                 }
 212
 213                 $type = self::getContentType($url);
 214                 Logger::info('Got content-type', ['content-type' => $type, 'url' => $url]);
 215                 if (!empty($type) && in_array($type[0], ['image', 'video', 'audio'])) {
 216                         $siteinfo['type'] = $type[0];
 217                         return $siteinfo;
 218                 }
 219
 220                 if ((count($type) >= 2) && (($type[0] != 'text') || ($type[1] != 'html'))) {
 221                         Logger::info('Unparseable content-type, quitting here, ', ['content-type' => $type, 'url' => $url]);
 222                         return $siteinfo;
 223                 }
 224
 225                 $curlResult = DI::httpClient()->get($url, [HttpClientOptions::CONTENT_LENGTH => 1000000, HttpClientOptions::ACCEPT_CONTENT => HttpClient::ACCEPT_HTML]);
 226                 if (!$curlResult->isSuccess() || empty($curlResult->getBody())) {
 227                         Logger::info('Empty body or error when fetching', ['url' => $url, 'success' => $curlResult->isSuccess(), 'code' => $curlResult->getReturnCode()]);
 228                         return $siteinfo;
 229                 }
 230
 231                 $siteinfo['expires'] = DateTimeFormat::utc(self::DEFAULT_EXPIRATION_SUCCESS);
 232
 233                 if ($cacheControlHeader = $curlResult->getHeader('Cache-Control')[0] ?? '') {
 234                         if (preg_match('/max-age=([0-9]+)/i', $cacheControlHeader, $matches)) {
 235                                 $maxAge = max(86400, (int)array_pop($matches));
 236                                 $siteinfo['expires'] = DateTimeFormat::utc("now + $maxAge seconds");
 237                         }
 238                 }
 239
 240                 $body = $curlResult->getBody();
 241
 242                 if ($do_oembed) {
 243                         $oembed_data = OEmbed::fetchURL($url, false, false);
 244
 245                         if (!empty($oembed_data->type)) {
 246                                 if (!in_array($oembed_data->type, ['error', 'rich', 'image', 'video', 'audio', ''])) {
 247                                         $siteinfo['type'] = $oembed_data->type;
 248                                 }
 249
 250                                 // See https://github.com/friendica/friendica/pull/5763#discussion_r217913178
 251                                 if ($siteinfo['type'] != 'photo') {
 252                                         if (!empty($oembed_data->title)) {
 253                                                 $siteinfo['title'] = trim($oembed_data->title);
 254                                         }
 255                                         if (!empty($oembed_data->description)) {
 256                                                 $siteinfo['text'] = trim($oembed_data->description);
 257                                         }
 258                                         if (!empty($oembed_data->author_name)) {
 259                                                 $siteinfo['author_name'] = trim($oembed_data->author_name);
 260                                         }
 261                                         if (!empty($oembed_data->author_url)) {
 262                                                 $siteinfo['author_url'] = trim($oembed_data->author_url);
 263                                         }
 264                                         if (!empty($oembed_data->provider_name)) {
 265                                                 $siteinfo['publisher_name'] = trim($oembed_data->provider_name);
 266                                         }
 267                                         if (!empty($oembed_data->provider_url)) {
 268                                                 $siteinfo['publisher_url'] = trim($oembed_data->provider_url);
 269                                         }
 270                                         if (!empty($oembed_data->thumbnail_url)) {
 271                                                 $siteinfo['image'] = $oembed_data->thumbnail_url;
 272                                         }
 273                                 }
 274                         }
 275                 }
 276
 277                 $charset = '';
 278                 // Look for a charset, first in headers
 279                 // Expected form: Content-Type: text/html; charset=ISO-8859-4
 280                 if (preg_match('/charset=([a-z0-9-_.\/]+)/i', $curlResult->getContentType(), $matches)) {
 281                         $charset = trim(trim(trim(array_pop($matches)), ';,'));
 282                 }
 283
 284                 // Then in body that gets precedence
 285                 // Expected forms:
 286                 // - <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
 287                 // - <meta charset="utf-8">
 288                 // - <meta charset=utf-8>
 289                 // - <meta charSet="utf-8">
 290                 // We escape <style> and <script> tags since they can contain irrelevant charset information
 291                 // (see https://github.com/friendica/friendica/issues/9251#issuecomment-698636806)
 292                 Strings::performWithEscapedBlocks($body, '#<(?:style|script).*?</(?:style|script)>#ism', function ($body) use (&$charset) {
 293                         if (preg_match('/charset=["\']?([a-z0-9-_.\/]+)/i', $body, $matches)) {
 294                                 $charset = trim(trim(trim(array_pop($matches)), ';,'));
 295                         }
 296                 });
 297
 298                 $siteinfo['charset'] = $charset;
 299
 300                 if ($charset && strtoupper($charset) != 'UTF-8') {
 301                         // See https://github.com/friendica/friendica/issues/5470#issuecomment-418351211
 302                         $charset = str_ireplace('latin-1', 'latin1', $charset);
 303
 304                         Logger::info('detected charset', ['charset' => $charset]);
 305                         $body = iconv($charset, 'UTF-8//TRANSLIT', $body);
 306                 }
 307
 308                 $body = mb_convert_encoding($body, 'HTML-ENTITIES', 'UTF-8');
 309
 310                 $doc = new DOMDocument();
 311                 @$doc->loadHTML($body);
 312
 313                 XML::deleteNode($doc, 'style');
 314                 XML::deleteNode($doc, 'option');
 315                 XML::deleteNode($doc, 'h1');
 316                 XML::deleteNode($doc, 'h2');
 317                 XML::deleteNode($doc, 'h3');
 318                 XML::deleteNode($doc, 'h4');
 319                 XML::deleteNode($doc, 'h5');
 320                 XML::deleteNode($doc, 'h6');
 321                 XML::deleteNode($doc, 'ol');
 322                 XML::deleteNode($doc, 'ul');
 323
 324                 $xpath = new DOMXPath($doc);
 325
 326                 $list = $xpath->query('//meta[@content]');
 327                 foreach ($list as $node) {
 328                         $meta_tag = [];
 329                         if ($node->attributes->length) {
 330                                 foreach ($node->attributes as $attribute) {
 331                                         $meta_tag[$attribute->name] = $attribute->value;
 332                                 }
 333                         }
 334
 335                         if (@$meta_tag['http-equiv'] == 'refresh') {
 336                                 $path = $meta_tag['content'];
 337                                 $pathinfo = explode(';', $path);
 338                                 $content = '';
 339                                 foreach ($pathinfo as $value) {
 340                                         if (substr(strtolower($value), 0, 4) == 'url=') {
 341                                                 $content = substr($value, 4);
 342                                         }
 343                                 }
 344                                 if ($content != '') {
 345                                         $siteinfo = self::getSiteinfo($content, $do_oembed, ++$count);
 346                                         return $siteinfo;
 347                                 }
 348                         }
 349                 }
 350
 351                 $list = $xpath->query('//title');
 352                 if ($list->length > 0) {
 353                         $siteinfo['title'] = trim($list->item(0)->nodeValue);
 354                 }
 355
 356                 $list = $xpath->query('//meta[@name]');
 357                 foreach ($list as $node) {
 358                         $meta_tag = [];
 359                         if ($node->attributes->length) {
 360                                 foreach ($node->attributes as $attribute) {
 361                                         $meta_tag[$attribute->name] = $attribute->value;
 362                                 }
 363                         }
 364
 365                         if (empty($meta_tag['content'])) {
 366                                 continue;
 367                         }
 368
 369                         $meta_tag['content'] = trim(html_entity_decode($meta_tag['content'], ENT_QUOTES, 'UTF-8'));
 370
 371                         switch (strtolower($meta_tag['name'])) {
 372                                 case 'fulltitle':
 373                                         $siteinfo['title'] = trim($meta_tag['content']);
 374                                         break;
 375                                 case 'description':
 376                                         $siteinfo['text'] = trim($meta_tag['content']);
 377                                         break;
 378                                 case 'thumbnail':
 379                                         $siteinfo['image'] = $meta_tag['content'];
 380                                         break;
 381                                 case 'twitter:image':
 382                                         $siteinfo['image'] = $meta_tag['content'];
 383                                         break;
 384                                 case 'twitter:image:src':
 385                                         $siteinfo['image'] = $meta_tag['content'];
 386                                         break;
 387                                 case 'twitter:description':
 388                                         $siteinfo['text'] = trim($meta_tag['content']);
 389                                         break;
 390                                 case 'twitter:title':
 391                                         $siteinfo['title'] = trim($meta_tag['content']);
 392                                         break;
 393                                 case 'twitter:player':
 394                                         $siteinfo['player']['embed'] = trim($meta_tag['content']);
 395                                         break;
 396                                 case 'twitter:player:stream':
 397                                         $siteinfo['player']['stream'] = trim($meta_tag['content']);
 398                                         break;
 399                                 case 'twitter:player:width':
 400                                         $siteinfo['player']['width'] = intval($meta_tag['content']);
 401                                         break;
 402                                 case 'twitter:player:height':
 403                                         $siteinfo['player']['height'] = intval($meta_tag['content']);
 404                                         break;
 405                                 case 'dc.title':
 406                                         $siteinfo['title'] = trim($meta_tag['content']);
 407                                         break;
 408                                 case 'dc.description':
 409                                         $siteinfo['text'] = trim($meta_tag['content']);
 410                                         break;
 411                                 case 'dc.creator':
 412                                         $siteinfo['publisher_name'] = trim($meta_tag['content']);
 413                                         break;
 414                                 case 'keywords':
 415                                         $keywords = explode(',', $meta_tag['content']);
 416                                         break;
 417                                 case 'news_keywords':
 418                                         $keywords = explode(',', $meta_tag['content']);
 419                                         break;
 420                         }
 421                 }
 422
 423                 if (isset($keywords)) {
 424                         $siteinfo['keywords'] = [];
 425                         foreach ($keywords as $keyword) {
 426                                 if (!in_array(trim($keyword), $siteinfo['keywords'])) {
 427                                         $siteinfo['keywords'][] = trim($keyword);
 428                                 }
 429                         }
 430                 }
 431
 432                 $list = $xpath->query('//meta[@property]');
 433                 foreach ($list as $node) {
 434                         $meta_tag = [];
 435                         if ($node->attributes->length) {
 436                                 foreach ($node->attributes as $attribute) {
 437                                         $meta_tag[$attribute->name] = $attribute->value;
 438                                 }
 439                         }
 440
 441                         if (!empty($meta_tag['content'])) {
 442                                 $meta_tag['content'] = trim(html_entity_decode($meta_tag['content'], ENT_QUOTES, 'UTF-8'));
 443
 444                                 switch (strtolower($meta_tag['property'])) {
 445                                         case 'og:image':
 446                                                 $siteinfo['image'] = $meta_tag['content'];
 447                                                 break;
 448                                         case 'og:image:url':
 449                                                 $siteinfo['image'] = $meta_tag['content'];
 450                                                 break;
 451                                         case 'og:image:secure_url':
 452                                                 $siteinfo['image'] = $meta_tag['content'];
 453                                                 break;
 454                                         case 'og:title':
 455                                                 $siteinfo['title'] = trim($meta_tag['content']);
 456                                                 break;
 457                                         case 'og:description':
 458                                                 $siteinfo['text'] = trim($meta_tag['content']);
 459                                                 break;
 460                                         case 'og:site_name':
 461                                                 $siteinfo['publisher_name'] = trim($meta_tag['content']);
 462                                                 break;
 463                                         case 'og:locale':
 464                                                 $siteinfo['language'] = trim($meta_tag['content']);
 465                                                 break;
 466                                         case 'og:type':
 467                                                 $siteinfo['pagetype'] = trim($meta_tag['content']);
 468                                                 break;
 469                                         case 'twitter:description':
 470                                                 $siteinfo['text'] = trim($meta_tag['content']);
 471                                                 break;
 472                                         case 'twitter:title':
 473                                                 $siteinfo['title'] = trim($meta_tag['content']);
 474                                                 break;
 475                                         case 'twitter:image':
 476                                                 $siteinfo['image'] = $meta_tag['content'];
 477                                                 break;
 478                                 }
 479                         }
 480                 }
 481
 482                 $list = $xpath->query("//script[@type='application/ld+json']");
 483                 foreach ($list as $node) {
 484                         if (!empty($node->nodeValue)) {
 485                                 if ($jsonld = json_decode($node->nodeValue, true)) {
 486                                         $siteinfo = self::parseParts($siteinfo, $jsonld);
 487                                 }
 488                         }
 489                 }
 490
 491                 if (!empty($siteinfo['player']['stream'])) {
 492                         // Only add player data to media arrays if there is no duplicate
 493                         $content_urls = array_merge(array_column($siteinfo['audio'] ?? [], 'content'), array_column($siteinfo['video'] ?? [], 'content'));
 494                         if (!in_array($siteinfo['player']['stream'], $content_urls)) {
 495                                 $contenttype = self::getContentType($siteinfo['player']['stream']);
 496                                 if (!empty($contenttype[0]) && in_array($contenttype[0], ['audio', 'video'])) {
 497                                         $media = ['content' => $siteinfo['player']['stream']];
 498
 499                                         if (!empty($siteinfo['player']['embed'])) {
 500                                                 $media['embed'] = $siteinfo['player']['embed'];
 501                                         }
 502
 503                                         $siteinfo[$contenttype[0]][] = $media;
 504                                 }
 505                         }
 506                 }
 507
 508                 if (!empty($siteinfo['image'])) {
 509                         $siteinfo['images'] = $siteinfo['images'] ?? [];
 510                         array_unshift($siteinfo['images'], ['url' => $siteinfo['image']]);
 511                         unset($siteinfo['image']);
 512                 }
 513
 514                 $siteinfo = self::checkMedia($url, $siteinfo);
 515
 516                 if (!empty($siteinfo['text']) && mb_strlen($siteinfo['text']) > self::MAX_DESC_COUNT) {
 517                         $siteinfo['text'] = mb_substr($siteinfo['text'], 0, self::MAX_DESC_COUNT) . '…';
 518                         $pos = mb_strrpos($siteinfo['text'], '.');
 519                         if ($pos > self::MIN_DESC_COUNT) {
 520                                 $siteinfo['text'] = mb_substr($siteinfo['text'], 0, $pos + 1);
 521                         }
 522                 }
 523
 524                 Logger::info('Siteinfo fetched', ['url' => $url, 'siteinfo' => $siteinfo]);
 525
 526                 Hook::callAll('getsiteinfo', $siteinfo);
 527
 528                 ksort($siteinfo);
 529
 530                 return $siteinfo;
 531         }
 532
 533         /**
 534          * Check the attached media elements.
 535          * Fix existing data and add missing data.
 536          *
 537          * @param string $page_url
 538          * @param array $siteinfo
 539          * @return array
 540          */
 541         private static function checkMedia(string $page_url, array $siteinfo) : array
 542         {
 543                 if (!empty($siteinfo['images'])) {
 544                         array_walk($siteinfo['images'], function (&$image) use ($page_url) {
 545                                 // According to the specifications someone could place a picture url into the content field as well.
 546                                 // But this doesn't seem to happen in the wild, so we don't cover it here.
 547                                 if (!empty($image['url'])) {
 548                                         $image['url'] = self::completeUrl($image['url'], $page_url);
 549                                         $photodata = Images::getInfoFromURLCached($image['url']);
 550                                         if (!empty($photodata) && ($photodata[0] > 50) && ($photodata[1] > 50)) {
 551                                                 $image['src'] = $image['url'];
 552                                                 $image['width'] = $photodata[0];
 553                                                 $image['height'] = $photodata[1];
 554                                                 $image['contenttype'] = $photodata['mime'];
 555                                                 unset($image['url']);
 556                                                 ksort($image);
 557                                         } else {
 558                                                 $image = [];
 559                                         }
 560                                 } else {
 561                                         $image = [];
 562                                 }
 563                         });
 564
 565                         $siteinfo['images'] = array_values(array_filter($siteinfo['images']));
 566                 }
 567
 568                 foreach (['audio', 'video'] as $element) {
 569                         if (!empty($siteinfo[$element])) {
 570                                 array_walk($siteinfo[$element], function (&$media) use ($page_url, &$siteinfo) {
 571                                         $url = '';
 572                                         $embed = '';
 573                                         $content = '';
 574                                         $contenttype = '';
 575                                         foreach (['embed', 'content', 'url'] as $field) {
 576                                                 if (!empty($media[$field])) {
 577                                                         $media[$field] = self::completeUrl($media[$field], $page_url);
 578                                                         $type = self::getContentType($media[$field]);
 579                                                         if (($type[0] ?? '') == 'text') {
 580                                                                 if ($field == 'embed') {
 581                                                                         $embed = $media[$field];
 582                                                                 } else {
 583                                                                         $url = $media[$field];
 584                                                                 }
 585                                                         } elseif (!empty($type[0])) {
 586                                                                 $content = $media[$field];
 587                                                                 $contenttype = implode('/', $type);
 588                                                         }
 589                                                 }
 590                                                 unset($media[$field]);
 591                                         }
 592
 593                                         foreach (['image', 'preview'] as $field) {
 594                                                 if (!empty($media[$field])) {
 595                                                         $media[$field] = self::completeUrl($media[$field], $page_url);
 596                                                 }
 597                                         }
 598
 599                                         if (!empty($url)) {
 600                                                 $media['url'] = $url;
 601                                         }
 602                                         if (!empty($embed)) {
 603                                                 $media['embed'] = $embed;
 604                                                 if (empty($siteinfo['player']['embed'])) {
 605                                                         $siteinfo['player']['embed'] = $embed;
 606                                                 }
 607                                         }
 608                                         if (!empty($content)) {
 609                                                 $media['src'] = $content;
 610                                         }
 611                                         if (!empty($contenttype)) {
 612                                                 $media['contenttype'] = $contenttype;
 613                                         }
 614                                         if (empty($url) && empty($content) && empty($embed)) {
 615                                                 $media = [];
 616                                         }
 617                                         ksort($media);
 618                                 });
 619
 620                                 $siteinfo[$element] = array_values(array_filter($siteinfo[$element]));
 621                         }
 622                         if (empty($siteinfo[$element])) {
 623                                 unset($siteinfo[$element]);
 624                         }
 625                 }
 626                 return $siteinfo;
 627         }
 628
 629         /**
 630          * Convert tags from CSV to an array
 631          *
 632          * @param string $string Tags
 633          * @return array with formatted Hashtags
 634          */
 635         public static function convertTagsToArray($string)
 636         {
 637                 $arr_tags = str_getcsv($string);
 638                 if (count($arr_tags)) {
 639                         // add the # sign to every tag
 640                         array_walk($arr_tags, ["self", "arrAddHashes"]);
 641
 642                         return $arr_tags;
 643                 }
 644         }
 645
 646         /**
 647          * Add a hasht sign to a string
 648          *
 649          * This method is used as callback function
 650          *
 651          * @param string $tag The pure tag name
 652          * @param int    $k   Counter for internal use
 653          * @return void
 654          */
 655         private static function arrAddHashes(&$tag, $k)
 656         {
 657                 $tag = "#" . $tag;
 658         }
 659
 660         /**
 661          * Add a scheme to an url
 662          *
 663          * The src attribute of some html elements (e.g. images)
 664          * can miss the scheme so we need to add the correct
 665          * scheme
 666          *
 667          * @param string $url    The url which possibly does have
 668          *                       a missing scheme (a link to an image)
 669          * @param string $scheme The url with a correct scheme
 670          *                       (e.g. the url from the webpage which does contain the image)
 671          *
 672          * @return string The url with a scheme
 673          */
 674         private static function completeUrl($url, $scheme)
 675         {
 676                 $urlarr = parse_url($url);
 677
 678                 // If the url does allready have an scheme
 679                 // we can stop the process here
 680                 if (isset($urlarr["scheme"])) {
 681                         return($url);
 682                 }
 683
 684                 $schemearr = parse_url($scheme);
 685
 686                 $complete = $schemearr["scheme"]."://".$schemearr["host"];
 687
 688                 if (!empty($schemearr["port"])) {
 689                         $complete .= ":".$schemearr["port"];
 690                 }
 691
 692                 if (!empty($urlarr["path"])) {
 693                         if (strpos($urlarr["path"], "/") !== 0) {
 694                                 $complete .= "/";
 695                         }
 696
 697                         $complete .= $urlarr["path"];
 698                 }
 699
 700                 if (!empty($urlarr["query"])) {
 701                         $complete .= "?".$urlarr["query"];
 702                 }
 703
 704                 if (!empty($urlarr["fragment"])) {
 705                         $complete .= "#".$urlarr["fragment"];
 706                 }
 707
 708                 return($complete);
 709         }
 710
 711         /**
 712          * Parse the Json-Ld parts of a web page
 713          *
 714          * @param array $siteinfo
 715          * @param array $jsonld
 716          * @return array siteinfo
 717          */
 718         private static function parseParts(array $siteinfo, array $jsonld)
 719         {
 720                 if (!empty($jsonld['@graph']) && is_array($jsonld['@graph'])) {
 721                         foreach ($jsonld['@graph'] as $part) {
 722                                 if (!empty($part) && is_array($part)) {
 723                                         $siteinfo = self::parseParts($siteinfo, $part);
 724                                 }
 725                         }
 726                 } elseif (!empty($jsonld['@type'])) {
 727                         $siteinfo = self::parseJsonLd($siteinfo, $jsonld);
 728                 } elseif (!empty($jsonld)) {
 729                         $keys = array_keys($jsonld);
 730                         $numeric_keys = true;
 731                         foreach ($keys as $key) {
 732                                 if (!is_int($key)) {
 733                                         $numeric_keys = false;
 734                                 }
 735                         }
 736                         if ($numeric_keys) {
 737                                 foreach ($jsonld as $part) {
 738                                         if (!empty($part) && is_array($part)) {
 739                                                 $siteinfo = self::parseParts($siteinfo, $part);
 740                                         }
 741                                 }
 742                         }
 743                 }
 744
 745                 array_walk_recursive($siteinfo, function (&$element) {
 746                         if (is_string($element)) {
 747                                 $element = trim(strip_tags(html_entity_decode($element, ENT_COMPAT, 'UTF-8')));
 748                         }
 749                 });
 750
 751                 return $siteinfo;
 752         }
 753
 754         /**
 755          * Improve the siteinfo with information from the provided JSON-LD information
 756          * @see https://jsonld.com/
 757          * @see https://schema.org/
 758          *
 759          * @param array $siteinfo
 760          * @param array $jsonld
 761          * @return array siteinfo
 762          */
 763         private static function parseJsonLd(array $siteinfo, array $jsonld)
 764         {
 765                 $type = JsonLD::fetchElement($jsonld, '@type');
 766                 if (empty($type)) {
 767                         Logger::info('Empty type', ['url' => $siteinfo['url']]);
 768                         return $siteinfo;
 769                 }
 770
 771                 // Silently ignore some types that aren't processed
 772                 if (in_array($type, ['SiteNavigationElement', 'JobPosting', 'CreativeWork', 'MusicAlbum',
 773                         'WPHeader', 'WPSideBar', 'WPFooter', 'LegalService', 'MusicRecording',
 774                         'ItemList', 'BreadcrumbList', 'Blog', 'Dataset', 'Product'])) {
 775                         return $siteinfo;
 776                 }
 777
 778                 switch ($type) {
 779                         case 'Article':
 780                         case 'AdvertiserContentArticle':
 781                         case 'NewsArticle':
 782                         case 'Report':
 783                         case 'SatiricalArticle':
 784                         case 'ScholarlyArticle':
 785                         case 'SocialMediaPosting':
 786                         case 'TechArticle':
 787                         case 'ReportageNewsArticle':
 788                         case 'SocialMediaPosting':
 789                         case 'BlogPosting':
 790                         case 'LiveBlogPosting':
 791                         case 'DiscussionForumPosting':
 792                                 return self::parseJsonLdArticle($siteinfo, $jsonld);
 793                         case 'WebPage':
 794                         case 'AboutPage':
 795                         case 'CheckoutPage':
 796                         case 'CollectionPage':
 797                         case 'ContactPage':
 798                         case 'FAQPage':
 799                         case 'ItemPage':
 800                         case 'MedicalWebPage':
 801                         case 'ProfilePage':
 802                         case 'QAPage':
 803                         case 'RealEstateListing':
 804                         case 'SearchResultsPage':
 805                         case 'MediaGallery':
 806                         case 'ImageGallery':
 807                         case 'VideoGallery':
 808                         case 'RadioEpisode':
 809                         case 'Event':
 810                                 return self::parseJsonLdWebPage($siteinfo, $jsonld);
 811                         case 'WebSite':
 812                                 return self::parseJsonLdWebSite($siteinfo, $jsonld);
 813                         case 'Organization':
 814                         case 'Airline':
 815                         case 'Consortium':
 816                         case 'Corporation':
 817                         case 'EducationalOrganization':
 818                         case 'FundingScheme':
 819                         case 'GovernmentOrganization':
 820                         case 'LibrarySystem':
 821                         case 'LocalBusiness':
 822                         case 'MedicalOrganization':
 823                         case 'NGO':
 824                         case 'NewsMediaOrganization':
 825                         case 'Project':
 826                         case 'SportsOrganization':
 827                         case 'WorkersUnion':
 828                                 return self::parseJsonLdWebOrganization($siteinfo, $jsonld);
 829                         case 'Person':
 830                         case 'Patient':
 831                         case 'PerformingGroup':
 832                         case 'DanceGroup';
 833                         case 'MusicGroup':
 834                         case 'TheaterGroup':
 835                                 return self::parseJsonLdWebPerson($siteinfo, $jsonld);
 836                         case 'AudioObject':
 837                         case 'Audio':
 838                                 return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'audio');
 839                         case 'VideoObject':
 840                                 return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'video');
 841                         case 'ImageObject':
 842                                 return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'images');
 843                         default:
 844                                 Logger::info('Unknown type', ['type' => $type, 'url' => $siteinfo['url']]);
 845                                 return $siteinfo;
 846                 }
 847         }
 848
 849         /**
 850          * Fetch author and publisher data
 851          *
 852          * @param array $siteinfo
 853          * @param array $jsonld
 854          * @return array siteinfo
 855          */
 856         private static function parseJsonLdAuthor(array $siteinfo, array $jsonld)
 857         {
 858                 $jsonldinfo = [];
 859
 860                 if (!empty($jsonld['publisher']) && is_array($jsonld['publisher'])) {
 861                         $content = JsonLD::fetchElement($jsonld, 'publisher', 'name');
 862                         if (!empty($content) && is_string($content)) {
 863                                 $jsonldinfo['publisher_name'] = trim($content);
 864                         }
 865
 866                         $content = JsonLD::fetchElement($jsonld, 'publisher', 'url');
 867                         if (!empty($content) && is_string($content)) {
 868                                 $jsonldinfo['publisher_url'] = trim($content);
 869                         }
 870
 871                         $brand = JsonLD::fetchElement($jsonld, 'publisher', 'brand', '@type', 'Organization');
 872                         if (!empty($brand) && is_array($brand)) {
 873                                 $content = JsonLD::fetchElement($brand, 'name');
 874                                 if (!empty($content) && is_string($content)) {
 875                                         $jsonldinfo['publisher_name'] = trim($content);
 876                                 }
 877
 878                                 $content = JsonLD::fetchElement($brand, 'url');
 879                                 if (!empty($content) && is_string($content)) {
 880                                         $jsonldinfo['publisher_url'] = trim($content);
 881                                 }
 882
 883                                 $content = JsonLD::fetchElement($brand, 'logo', 'url');
 884                                 if (!empty($content) && is_string($content)) {
 885                                         $jsonldinfo['publisher_img'] = trim($content);
 886                                 }
 887                         }
 888
 889                         $logo = JsonLD::fetchElement($jsonld, 'publisher', 'logo');
 890                         if (!empty($logo) && is_array($logo)) {
 891                                 $content = JsonLD::fetchElement($logo, 'url');
 892                                 if (!empty($content) && is_string($content)) {
 893                                         $jsonldinfo['publisher_img'] = trim($content);
 894                                 }
 895                         }
 896                 } elseif (!empty($jsonld['publisher']) && is_string($jsonld['publisher'])) {
 897                         $jsonldinfo['publisher_name'] = trim($jsonld['publisher']);
 898                 }
 899
 900                 if (!empty($jsonld['author']) && is_array($jsonld['author'])) {
 901                         $content = JsonLD::fetchElement($jsonld, 'author', 'name');
 902                         if (!empty($content) && is_string($content)) {
 903                                 $jsonldinfo['author_name'] = trim($content);
 904                         }
 905
 906                         $content = JsonLD::fetchElement($jsonld, 'author', 'sameAs');
 907                         if (!empty($content) && is_string($content)) {
 908                                 $jsonldinfo['author_url'] = trim($content);
 909                         }
 910
 911                         $content = JsonLD::fetchElement($jsonld, 'author', 'url');
 912                         if (!empty($content) && is_string($content)) {
 913                                 $jsonldinfo['author_url'] = trim($content);
 914                         }
 915
 916                         $logo = JsonLD::fetchElement($jsonld, 'author', 'logo');
 917                         if (!empty($logo) && is_array($logo)) {
 918                                 $content = JsonLD::fetchElement($logo, 'url');
 919                                 if (!empty($content) && is_string($content)) {
 920                                         $jsonldinfo['author_img'] = trim($content);
 921                                 }
 922                         }
 923                 } elseif (!empty($jsonld['author']) && is_string($jsonld['author'])) {
 924                         $jsonldinfo['author_name'] = trim($jsonld['author']);
 925                 }
 926
 927                 Logger::info('Fetched Author information', ['fetched' => $jsonldinfo]);
 928
 929                 return array_merge($siteinfo, $jsonldinfo);
 930         }
 931
 932         /**
 933          * Fetch data from the provided JSON-LD Article type
 934          * @see https://schema.org/Article
 935          *
 936          * @param array $siteinfo
 937          * @param array $jsonld
 938          * @return array siteinfo
 939          */
 940         private static function parseJsonLdArticle(array $siteinfo, array $jsonld)
 941         {
 942                 $jsonldinfo = [];
 943
 944                 $content = JsonLD::fetchElement($jsonld, 'headline');
 945                 if (!empty($content) && is_string($content)) {
 946                         $jsonldinfo['title'] = trim($content);
 947                 }
 948
 949                 $content = JsonLD::fetchElement($jsonld, 'alternativeHeadline');
 950                 if (!empty($content) && is_string($content) && (($jsonldinfo['title'] ?? '') != trim($content))) {
 951                         $jsonldinfo['alternative_title'] = trim($content);
 952                 }
 953
 954                 $content = JsonLD::fetchElement($jsonld, 'description');
 955                 if (!empty($content) && is_string($content)) {
 956                         $jsonldinfo['text'] = trim($content);
 957                 }
 958
 959                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
 960                 if (!empty($content)) {
 961                         $jsonldinfo['image'] = trim($content);
 962                 }
 963
 964                 $content = JsonLD::fetchElement($jsonld, 'image', 'url', '@type', 'ImageObject');
 965                 if (!empty($content) && is_string($content)) {
 966                         $jsonldinfo['image'] = trim($content);
 967                 }
 968
 969                 if (!empty($jsonld['keywords']) && !is_array($jsonld['keywords'])) {
 970                         $content = JsonLD::fetchElement($jsonld, 'keywords');
 971                         if (!empty($content)) {
 972                                 $siteinfo['keywords'] = [];
 973                                 $keywords = explode(',', $content);
 974                                 foreach ($keywords as $keyword) {
 975                                         $siteinfo['keywords'][] = trim($keyword);
 976                                 }
 977                         }
 978                 } elseif (!empty($jsonld['keywords'])) {
 979                         $content = JsonLD::fetchElementArray($jsonld, 'keywords');
 980                         if (!empty($content) && is_array($content)) {
 981                                 $jsonldinfo['keywords'] = $content;
 982                         }
 983                 }
 984
 985                 $content = JsonLD::fetchElement($jsonld, 'datePublished');
 986                 if (!empty($content) && is_string($content)) {
 987                         $jsonldinfo['published'] = DateTimeFormat::utc($content);
 988                 }
 989
 990                 $content = JsonLD::fetchElement($jsonld, 'dateModified');
 991                 if (!empty($content) && is_string($content)) {
 992                         $jsonldinfo['modified'] = DateTimeFormat::utc($content);
 993                 }
 994
 995                 $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
 996
 997                 Logger::info('Fetched article information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
 998
 999                 return array_merge($siteinfo, $jsonldinfo);
1000         }
1001
1002         /**
1003          * Fetch data from the provided JSON-LD WebPage type
1004          * @see https://schema.org/WebPage
1005          *
1006          * @param array $siteinfo
1007          * @param array $jsonld
1008          * @return array siteinfo
1009          */
1010         private static function parseJsonLdWebPage(array $siteinfo, array $jsonld)
1011         {
1012                 $jsonldinfo = [];
1013
1014                 $content = JsonLD::fetchElement($jsonld, 'name');
1015                 if (!empty($content)) {
1016                         $jsonldinfo['title'] = trim($content);
1017                 }
1018
1019                 $content = JsonLD::fetchElement($jsonld, 'description');
1020                 if (!empty($content)) {
1021                         $jsonldinfo['text'] = trim($content);
1022                 }
1023
1024                 $content = JsonLD::fetchElement($jsonld, 'image');
1025                 if (!empty($content)) {
1026                         $jsonldinfo['image'] = trim($content);
1027                 }
1028
1029                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
1030                 if (!empty($content)) {
1031                         $jsonldinfo['image'] = trim($content);
1032                 }
1033
1034                 $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
1035
1036                 Logger::info('Fetched WebPage information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1037
1038                 return array_merge($siteinfo, $jsonldinfo);
1039         }
1040
1041         /**
1042          * Fetch data from the provided JSON-LD WebSite type
1043          * @see https://schema.org/WebSite
1044          *
1045          * @param array $siteinfo
1046          * @param array $jsonld
1047          * @return array siteinfo
1048          */
1049         private static function parseJsonLdWebSite(array $siteinfo, array $jsonld)
1050         {
1051                 $jsonldinfo = [];
1052
1053                 $content = JsonLD::fetchElement($jsonld, 'name');
1054                 if (!empty($content)) {
1055                         $jsonldinfo['publisher_name'] = trim($content);
1056                 }
1057
1058                 $content = JsonLD::fetchElement($jsonld, 'description');
1059                 if (!empty($content)) {
1060                         $jsonldinfo['publisher_description'] = trim($content);
1061                 }
1062
1063                 $content = JsonLD::fetchElement($jsonld, 'url');
1064                 if (!empty($content)) {
1065                         $jsonldinfo['publisher_url'] = trim($content);
1066                 }
1067
1068                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
1069                 if (!empty($content)) {
1070                         $jsonldinfo['image'] = trim($content);
1071                 }
1072
1073                 $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
1074
1075                 Logger::info('Fetched WebSite information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1076                 return array_merge($siteinfo, $jsonldinfo);
1077         }
1078
1079         /**
1080          * Fetch data from the provided JSON-LD Organization type
1081          * @see https://schema.org/Organization
1082          *
1083          * @param array $siteinfo
1084          * @param array $jsonld
1085          * @return array siteinfo
1086          */
1087         private static function parseJsonLdWebOrganization(array $siteinfo, array $jsonld)
1088         {
1089                 $jsonldinfo = [];
1090
1091                 $content = JsonLD::fetchElement($jsonld, 'name');
1092                 if (!empty($content)) {
1093                         $jsonldinfo['publisher_name'] = trim($content);
1094                 }
1095
1096                 $content = JsonLD::fetchElement($jsonld, 'description');
1097                 if (!empty($content)) {
1098                         $jsonldinfo['publisher_description'] = trim($content);
1099                 }
1100
1101                 $content = JsonLD::fetchElement($jsonld, 'url');
1102                 if (!empty($content)) {
1103                         $jsonldinfo['publisher_url'] = trim($content);
1104                 }
1105
1106                 $content = JsonLD::fetchElement($jsonld, 'logo', 'url', '@type', 'ImageObject');
1107                 if (!empty($content)) {
1108                         $jsonldinfo['publisher_img'] = trim($content);
1109                 }
1110
1111                 $content = JsonLD::fetchElement($jsonld, 'brand', 'name', '@type', 'Organization');
1112                 if (!empty($content)) {
1113                         $jsonldinfo['publisher_name'] = trim($content);
1114                 }
1115
1116                 $content = JsonLD::fetchElement($jsonld, 'brand', 'url', '@type', 'Organization');
1117                 if (!empty($content)) {
1118                         $jsonldinfo['publisher_url'] = trim($content);
1119                 }
1120
1121                 Logger::info('Fetched Organization information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1122                 return array_merge($siteinfo, $jsonldinfo);
1123         }
1124
1125         /**
1126          * Fetch data from the provided JSON-LD Person type
1127          * @see https://schema.org/Person
1128          *
1129          * @param array $siteinfo
1130          * @param array $jsonld
1131          * @return array siteinfo
1132          */
1133         private static function parseJsonLdWebPerson(array $siteinfo, array $jsonld)
1134         {
1135                 $jsonldinfo = [];
1136
1137                 $content = JsonLD::fetchElement($jsonld, 'name');
1138                 if (!empty($content)) {
1139                         $jsonldinfo['author_name'] = trim($content);
1140                 }
1141
1142                 $content = JsonLD::fetchElement($jsonld, 'description');
1143                 if (!empty($content)) {
1144                         $jsonldinfo['author_description'] = trim($content);
1145                 }
1146
1147                 $content = JsonLD::fetchElement($jsonld, 'sameAs');
1148                 if (!empty($content) && is_string($content)) {
1149                         $jsonldinfo['author_url'] = trim($content);
1150                 }
1151
1152                 $content = JsonLD::fetchElement($jsonld, 'url');
1153                 if (!empty($content)) {
1154                         $jsonldinfo['author_url'] = trim($content);
1155                 }
1156
1157                 $content = JsonLD::fetchElement($jsonld, 'image', 'url', '@type', 'ImageObject');
1158                 if (!empty($content) && !is_string($content)) {
1159                         Logger::notice('Unexpected return value for the author image', ['content' => $content]);
1160                 }
1161
1162                 if (!empty($content) && is_string($content)) {
1163                         $jsonldinfo['author_img'] = trim($content);
1164                 }
1165
1166                 Logger::info('Fetched Person information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1167                 return array_merge($siteinfo, $jsonldinfo);
1168         }
1169
1170         /**
1171          * Fetch data from the provided JSON-LD MediaObject type
1172          * @see https://schema.org/MediaObject
1173          *
1174          * @param array $siteinfo
1175          * @param array $jsonld
1176          * @return array siteinfo
1177          */
1178         private static function parseJsonLdMediaObject(array $siteinfo, array $jsonld, string $name)
1179         {
1180                 $media = [];
1181
1182                 $content = JsonLD::fetchElement($jsonld, 'caption');
1183                 if (!empty($content)) {
1184                         $media['caption'] = trim($content);
1185                 }
1186
1187                 $content = JsonLD::fetchElement($jsonld, 'url');
1188                 if (!empty($content)) {
1189                         $media['url'] = trim($content);
1190                 }
1191
1192                 $content = JsonLD::fetchElement($jsonld, 'mainEntityOfPage');
1193                 if (!empty($content)) {
1194                         $media['main'] = Strings::compareLink($content, $siteinfo['url']);
1195                 }
1196
1197                 $content = JsonLD::fetchElement($jsonld, 'description');
1198                 if (!empty($content)) {
1199                         $media['description'] = trim($content);
1200                 }
1201
1202                 $content = JsonLD::fetchElement($jsonld, 'name');
1203                 if (!empty($content) && (($media['description'] ?? '') != trim($content))) {
1204                         $media['name'] = trim($content);
1205                 }
1206
1207                 $content = JsonLD::fetchElement($jsonld, 'contentUrl');
1208                 if (!empty($content)) {
1209                         $media['content'] = trim($content);
1210                 }
1211
1212                 $content = JsonLD::fetchElement($jsonld, 'embedUrl');
1213                 if (!empty($content)) {
1214                         $media['embed'] = trim($content);
1215                 }
1216
1217                 $content = JsonLD::fetchElement($jsonld, 'height');
1218                 if (!empty($content)) {
1219                         $media['height'] = trim($content);
1220                 }
1221
1222                 $content = JsonLD::fetchElement($jsonld, 'width');
1223                 if (!empty($content)) {
1224                         $media['width'] = trim($content);
1225                 }
1226
1227                 $content = JsonLD::fetchElement($jsonld, 'image');
1228                 if (!empty($content)) {
1229                         $media['image'] = trim($content);
1230                 }
1231
1232                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
1233                 if (!empty($content) && (($media['image'] ?? '') != trim($content))) {
1234                         if (!empty($media['image'])) {
1235                                 $media['preview'] = trim($content);
1236                         } else {
1237                                 $media['image'] = trim($content);
1238                         }
1239                 }
1240
1241                 Logger::info('Fetched Media information', ['url' => $siteinfo['url'], 'fetched' => $media]);
1242                 $siteinfo[$name][] = $media;
1243                 return $siteinfo;
1244         }
1245 }