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\HttpClientAccept;
  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          * @param int    $timeout
  61          * @return array content type
  62          */
  63         public static function getContentType(string $url, string $accept = HttpClientAccept::DEFAULT, int $timeout = 0)
  64         {
  65                 if (!empty($timeout)) {
  66                         $options = [HttpClientOptions::TIMEOUT => $timeout];
  67                 } else {
  68                         $options = [];
  69                 }
  70
  71                 $curlResult = DI::httpClient()->head($url, array_merge([HttpClientOptions::ACCEPT_CONTENT => $accept], $options));
  72
  73                 // Workaround for systems that can't handle a HEAD request. Don't retry on timeouts.
  74                 if (!$curlResult->isSuccess() && ($curlResult->getReturnCode() >= 400) && !in_array($curlResult->getReturnCode(), [408, 504])) {
  75                         $curlResult = DI::httpClient()->get($url, $accept, array_merge([HttpClientOptions::CONTENT_LENGTH => 1000000], $options));
  76                 }
  77
  78                 if (!$curlResult->isSuccess()) {
  79                         Logger::debug('Got HTTP Error', ['http error' => $curlResult->getReturnCode(), 'url' => $url]);
  80                         return [];
  81                 }
  82
  83                 $contenttype =  $curlResult->getHeader('Content-Type')[0] ?? '';
  84                 if (empty($contenttype)) {
  85                         return ['application', 'octet-stream'];
  86                 }
  87
  88                 return explode('/', current(explode(';', $contenttype)));
  89         }
  90
  91         /**
  92          * Search for chached embeddable data of an url otherwise fetch it
  93          *
  94          * @param string $url         The url of the page which should be scraped
  95          * @param bool   $do_oembed   The false option is used by the function fetch_oembed()
  96          *                            to avoid endless loops
  97          *
  98          * @return array which contains needed data for embedding
  99          *    string 'url'      => The url of the parsed page
 100          *    string 'type'     => Content type
 101          *    string 'title'    => (optional) The title of the content
 102          *    string 'text'     => (optional) The description for the content
 103          *    string 'image'    => (optional) A preview image of the content
 104          *    array  'images'   => (optional) Array of preview pictures
 105          *    string 'keywords' => (optional) The tags which belong to the content
 106          *
 107          * @throws HTTPException\InternalServerErrorException
 108          * @see   ParseUrl::getSiteinfo() for more information about scraping
 109          * embeddable content
 110          */
 111         public static function getSiteinfoCached($url, $do_oembed = true): array
 112         {
 113                 if (empty($url)) {
 114                         return [
 115                                 'url' => '',
 116                                 'type' => 'error',
 117                         ];
 118                 }
 119
 120                 $urlHash = hash('sha256', $url);
 121
 122                 $parsed_url = DBA::selectFirst('parsed_url', ['content'],
 123                         ['url_hash' => $urlHash, 'oembed' => $do_oembed]
 124                 );
 125                 if (!empty($parsed_url['content'])) {
 126                         $data = unserialize($parsed_url['content']);
 127                         return $data;
 128                 }
 129
 130                 $data = self::getSiteinfo($url, $do_oembed);
 131
 132                 $expires = $data['expires'];
 133
 134                 unset($data['expires']);
 135
 136                 DI::dba()->insert(
 137                         'parsed_url',
 138                         [
 139                                 'url_hash' => $urlHash,
 140                                 'oembed'   => $do_oembed,
 141                                 'url'      => $url,
 142                                 'content'  => serialize($data),
 143                                 'created'  => DateTimeFormat::utcNow(),
 144                                 'expires'  => $expires,
 145                         ],
 146                         Database::INSERT_UPDATE
 147                 );
 148
 149                 return $data;
 150         }
 151
 152         /**
 153          * Parse a page for embeddable content information
 154          *
 155          * This method parses to url for meta data which can be used to embed
 156          * the content. If available it prioritizes Open Graph meta tags.
 157          * If this is not available it uses the twitter cards meta tags.
 158          * As fallback it uses standard html elements with meta informations
 159          * like \<title\>Awesome Title\</title\> or
 160          * \<meta name="description" content="An awesome description"\>
 161          *
 162          * @param string $url         The url of the page which should be scraped
 163          * @param bool   $do_oembed   The false option is used by the function fetch_oembed()
 164          *                            to avoid endless loops
 165          * @param int    $count       Internal counter to avoid endless loops
 166          *
 167          * @return array which contains needed data for embedding
 168          *    string 'url'      => The url of the parsed page
 169          *    string 'type'     => Content type (error, link, photo, image, audio, video)
 170          *    string 'title'    => (optional) The title of the content
 171          *    string 'text'     => (optional) The description for the content
 172          *    string 'image'    => (optional) A preview image of the content
 173          *    array  'images'   => (optional) Array of preview pictures
 174          *    string 'keywords' => (optional) The tags which belong to the content
 175          *
 176          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 177          * @todo  https://developers.google.com/+/plugins/snippet/
 178          * @verbatim
 179          * <meta itemprop="name" content="Awesome title">
 180          * <meta itemprop="description" content="An awesome description">
 181          * <meta itemprop="image" content="http://maple.libertreeproject.org/images/tree-icon.png">
 182          *
 183          * <body itemscope itemtype="http://schema.org/Product">
 184          *   <h1 itemprop="name">Shiny Trinket</h1>
 185          *   <img itemprop="image" src="{image-url}" />
 186          *   <p itemprop="description">Shiny trinkets are shiny.</p>
 187          * </body>
 188          * @endverbatim
 189          */
 190         public static function getSiteinfo($url, $do_oembed = true, $count = 1)
 191         {
 192                 if (empty($url)) {
 193                         return [
 194                                 'url' => '',
 195                                 'type' => 'error',
 196                         ];
 197                 }
 198
 199                 // Check if the URL does contain a scheme
 200                 $scheme = parse_url($url, PHP_URL_SCHEME);
 201
 202                 if ($scheme == '') {
 203                         $url = 'http://' . ltrim($url, '/');
 204                 }
 205
 206                 $url = trim($url, "'\"");
 207
 208                 $url = Network::stripTrackingQueryParams($url);
 209
 210                 $siteinfo = [
 211                         'url' => $url,
 212                         'type' => 'link',
 213                         'expires' => DateTimeFormat::utc(self::DEFAULT_EXPIRATION_FAILURE),
 214                 ];
 215
 216                 if ($count > 10) {
 217                         Logger::notice('Endless loop detected', ['url' => $url]);
 218                         return $siteinfo;
 219                 }
 220
 221                 $type = self::getContentType($url);
 222                 Logger::info('Got content-type', ['content-type' => $type, 'url' => $url]);
 223                 if (!empty($type) && in_array($type[0], ['image', 'video', 'audio'])) {
 224                         $siteinfo['type'] = $type[0];
 225                         return $siteinfo;
 226                 }
 227
 228                 if ((count($type) >= 2) && (($type[0] != 'text') || ($type[1] != 'html'))) {
 229                         Logger::info('Unparseable content-type, quitting here, ', ['content-type' => $type, 'url' => $url]);
 230                         return $siteinfo;
 231                 }
 232
 233                 $curlResult = DI::httpClient()->get($url, HttpClientAccept::HTML, [HttpClientOptions::CONTENT_LENGTH => 1000000]);
 234                 if (!$curlResult->isSuccess() || empty($curlResult->getBody())) {
 235                         Logger::info('Empty body or error when fetching', ['url' => $url, 'success' => $curlResult->isSuccess(), 'code' => $curlResult->getReturnCode()]);
 236                         return $siteinfo;
 237                 }
 238
 239                 $siteinfo['expires'] = DateTimeFormat::utc(self::DEFAULT_EXPIRATION_SUCCESS);
 240
 241                 if ($cacheControlHeader = $curlResult->getHeader('Cache-Control')[0] ?? '') {
 242                         if (preg_match('/max-age=([0-9]+)/i', $cacheControlHeader, $matches)) {
 243                                 $maxAge = max(86400, (int)array_pop($matches));
 244                                 $siteinfo['expires'] = DateTimeFormat::utc("now + $maxAge seconds");
 245                         }
 246                 }
 247
 248                 $body = $curlResult->getBody();
 249
 250                 if ($do_oembed) {
 251                         $oembed_data = OEmbed::fetchURL($url, false, false);
 252
 253                         if (!empty($oembed_data->type)) {
 254                                 if (!in_array($oembed_data->type, ['error', 'rich', 'image', 'video', 'audio', ''])) {
 255                                         $siteinfo['type'] = $oembed_data->type;
 256                                 }
 257
 258                                 // See https://github.com/friendica/friendica/pull/5763#discussion_r217913178
 259                                 if ($siteinfo['type'] != 'photo') {
 260                                         if (!empty($oembed_data->title)) {
 261                                                 $siteinfo['title'] = trim($oembed_data->title);
 262                                         }
 263                                         if (!empty($oembed_data->description)) {
 264                                                 $siteinfo['text'] = trim($oembed_data->description);
 265                                         }
 266                                         if (!empty($oembed_data->author_name)) {
 267                                                 $siteinfo['author_name'] = trim($oembed_data->author_name);
 268                                         }
 269                                         if (!empty($oembed_data->author_url)) {
 270                                                 $siteinfo['author_url'] = trim($oembed_data->author_url);
 271                                         }
 272                                         if (!empty($oembed_data->provider_name)) {
 273                                                 $siteinfo['publisher_name'] = trim($oembed_data->provider_name);
 274                                         }
 275                                         if (!empty($oembed_data->provider_url)) {
 276                                                 $siteinfo['publisher_url'] = trim($oembed_data->provider_url);
 277                                         }
 278                                         if (!empty($oembed_data->thumbnail_url)) {
 279                                                 $siteinfo['image'] = $oembed_data->thumbnail_url;
 280                                         }
 281                                 }
 282                         }
 283                 }
 284
 285                 $charset = '';
 286                 // Look for a charset, first in headers
 287                 // Expected form: Content-Type: text/html; charset=ISO-8859-4
 288                 if (preg_match('/charset=([a-z0-9-_.\/]+)/i', $curlResult->getContentType(), $matches)) {
 289                         $charset = trim(trim(trim(array_pop($matches)), ';,'));
 290                 }
 291
 292                 // Then in body that gets precedence
 293                 // Expected forms:
 294                 // - <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
 295                 // - <meta charset="utf-8">
 296                 // - <meta charset=utf-8>
 297                 // - <meta charSet="utf-8">
 298                 // We escape <style> and <script> tags since they can contain irrelevant charset information
 299                 // (see https://github.com/friendica/friendica/issues/9251#issuecomment-698636806)
 300                 Strings::performWithEscapedBlocks($body, '#<(?:style|script).*?</(?:style|script)>#ism', function ($body) use (&$charset) {
 301                         if (preg_match('/charset=["\']?([a-z0-9-_.\/]+)/i', $body, $matches)) {
 302                                 $charset = trim(trim(trim(array_pop($matches)), ';,'));
 303                         }
 304                 });
 305
 306                 $siteinfo['charset'] = $charset;
 307
 308                 if ($charset && strtoupper($charset) != 'UTF-8') {
 309                         // See https://github.com/friendica/friendica/issues/5470#issuecomment-418351211
 310                         $charset = str_ireplace('latin-1', 'latin1', $charset);
 311
 312                         Logger::info('detected charset', ['charset' => $charset]);
 313                         $body = iconv($charset, 'UTF-8//TRANSLIT', $body);
 314                 }
 315
 316                 $body = mb_convert_encoding($body, 'HTML-ENTITIES', 'UTF-8');
 317
 318                 $doc = new DOMDocument();
 319                 @$doc->loadHTML($body);
 320
 321                 XML::deleteNode($doc, 'style');
 322                 XML::deleteNode($doc, 'option');
 323                 XML::deleteNode($doc, 'h1');
 324                 XML::deleteNode($doc, 'h2');
 325                 XML::deleteNode($doc, 'h3');
 326                 XML::deleteNode($doc, 'h4');
 327                 XML::deleteNode($doc, 'h5');
 328                 XML::deleteNode($doc, 'h6');
 329                 XML::deleteNode($doc, 'ol');
 330                 XML::deleteNode($doc, 'ul');
 331
 332                 $xpath = new DOMXPath($doc);
 333
 334                 $list = $xpath->query('//meta[@content]');
 335                 foreach ($list as $node) {
 336                         $meta_tag = [];
 337                         if ($node->attributes->length) {
 338                                 foreach ($node->attributes as $attribute) {
 339                                         $meta_tag[$attribute->name] = $attribute->value;
 340                                 }
 341                         }
 342
 343                         if (@$meta_tag['http-equiv'] == 'refresh') {
 344                                 $path = $meta_tag['content'];
 345                                 $pathinfo = explode(';', $path);
 346                                 $content = '';
 347                                 foreach ($pathinfo as $value) {
 348                                         if (substr(strtolower($value), 0, 4) == 'url=') {
 349                                                 $content = substr($value, 4);
 350                                         }
 351                                 }
 352                                 if ($content != '') {
 353                                         $siteinfo = self::getSiteinfo($content, $do_oembed, ++$count);
 354                                         return $siteinfo;
 355                                 }
 356                         }
 357                 }
 358
 359                 $list = $xpath->query('//title');
 360                 if ($list->length > 0) {
 361                         $siteinfo['title'] = trim($list->item(0)->nodeValue);
 362                 }
 363
 364                 $list = $xpath->query('//meta[@name]');
 365                 foreach ($list as $node) {
 366                         $meta_tag = [];
 367                         if ($node->attributes->length) {
 368                                 foreach ($node->attributes as $attribute) {
 369                                         $meta_tag[$attribute->name] = $attribute->value;
 370                                 }
 371                         }
 372
 373                         if (empty($meta_tag['content'])) {
 374                                 continue;
 375                         }
 376
 377                         $meta_tag['content'] = trim(html_entity_decode($meta_tag['content'], ENT_QUOTES, 'UTF-8'));
 378
 379                         switch (strtolower($meta_tag['name'])) {
 380                                 case 'fulltitle':
 381                                         $siteinfo['title'] = trim($meta_tag['content']);
 382                                         break;
 383                                 case 'description':
 384                                         $siteinfo['text'] = trim($meta_tag['content']);
 385                                         break;
 386                                 case 'thumbnail':
 387                                         $siteinfo['image'] = $meta_tag['content'];
 388                                         break;
 389                                 case 'twitter:image':
 390                                         $siteinfo['image'] = $meta_tag['content'];
 391                                         break;
 392                                 case 'twitter:image:src':
 393                                         $siteinfo['image'] = $meta_tag['content'];
 394                                         break;
 395                                 case 'twitter:description':
 396                                         $siteinfo['text'] = trim($meta_tag['content']);
 397                                         break;
 398                                 case 'twitter:title':
 399                                         $siteinfo['title'] = trim($meta_tag['content']);
 400                                         break;
 401                                 case 'twitter:player':
 402                                         $siteinfo['player']['embed'] = trim($meta_tag['content']);
 403                                         break;
 404                                 case 'twitter:player:stream':
 405                                         $siteinfo['player']['stream'] = trim($meta_tag['content']);
 406                                         break;
 407                                 case 'twitter:player:width':
 408                                         $siteinfo['player']['width'] = intval($meta_tag['content']);
 409                                         break;
 410                                 case 'twitter:player:height':
 411                                         $siteinfo['player']['height'] = intval($meta_tag['content']);
 412                                         break;
 413                                 case 'dc.title':
 414                                         $siteinfo['title'] = trim($meta_tag['content']);
 415                                         break;
 416                                 case 'dc.description':
 417                                         $siteinfo['text'] = trim($meta_tag['content']);
 418                                         break;
 419                                 case 'dc.creator':
 420                                         $siteinfo['publisher_name'] = trim($meta_tag['content']);
 421                                         break;
 422                                 case 'keywords':
 423                                         $keywords = explode(',', $meta_tag['content']);
 424                                         break;
 425                                 case 'news_keywords':
 426                                         $keywords = explode(',', $meta_tag['content']);
 427                                         break;
 428                         }
 429                 }
 430
 431                 if (isset($keywords)) {
 432                         $siteinfo['keywords'] = [];
 433                         foreach ($keywords as $keyword) {
 434                                 if (!in_array(trim($keyword), $siteinfo['keywords'])) {
 435                                         $siteinfo['keywords'][] = trim($keyword);
 436                                 }
 437                         }
 438                 }
 439
 440                 $list = $xpath->query('//meta[@property]');
 441                 foreach ($list as $node) {
 442                         $meta_tag = [];
 443                         if ($node->attributes->length) {
 444                                 foreach ($node->attributes as $attribute) {
 445                                         $meta_tag[$attribute->name] = $attribute->value;
 446                                 }
 447                         }
 448
 449                         if (!empty($meta_tag['content'])) {
 450                                 $meta_tag['content'] = trim(html_entity_decode($meta_tag['content'], ENT_QUOTES, 'UTF-8'));
 451
 452                                 switch (strtolower($meta_tag['property'])) {
 453                                         case 'og:image':
 454                                                 $siteinfo['image'] = $meta_tag['content'];
 455                                                 break;
 456                                         case 'og:image:url':
 457                                                 $siteinfo['image'] = $meta_tag['content'];
 458                                                 break;
 459                                         case 'og:image:secure_url':
 460                                                 $siteinfo['image'] = $meta_tag['content'];
 461                                                 break;
 462                                         case 'og:title':
 463                                                 $siteinfo['title'] = trim($meta_tag['content']);
 464                                                 break;
 465                                         case 'og:description':
 466                                                 $siteinfo['text'] = trim($meta_tag['content']);
 467                                                 break;
 468                                         case 'og:site_name':
 469                                                 $siteinfo['publisher_name'] = trim($meta_tag['content']);
 470                                                 break;
 471                                         case 'og:locale':
 472                                                 $siteinfo['language'] = trim($meta_tag['content']);
 473                                                 break;
 474                                         case 'og:type':
 475                                                 $siteinfo['pagetype'] = trim($meta_tag['content']);
 476                                                 break;
 477                                         case 'twitter:description':
 478                                                 $siteinfo['text'] = trim($meta_tag['content']);
 479                                                 break;
 480                                         case 'twitter:title':
 481                                                 $siteinfo['title'] = trim($meta_tag['content']);
 482                                                 break;
 483                                         case 'twitter:image':
 484                                                 $siteinfo['image'] = $meta_tag['content'];
 485                                                 break;
 486                                 }
 487                         }
 488                 }
 489
 490                 $list = $xpath->query("//script[@type='application/ld+json']");
 491                 foreach ($list as $node) {
 492                         if (!empty($node->nodeValue)) {
 493                                 if ($jsonld = json_decode($node->nodeValue, true)) {
 494                                         $siteinfo = self::parseParts($siteinfo, $jsonld);
 495                                 }
 496                         }
 497                 }
 498
 499                 if (!empty($siteinfo['player']['stream'])) {
 500                         // Only add player data to media arrays if there is no duplicate
 501                         $content_urls = array_merge(array_column($siteinfo['audio'] ?? [], 'content'), array_column($siteinfo['video'] ?? [], 'content'));
 502                         if (!in_array($siteinfo['player']['stream'], $content_urls)) {
 503                                 $contenttype = self::getContentType($siteinfo['player']['stream']);
 504                                 if (!empty($contenttype[0]) && in_array($contenttype[0], ['audio', 'video'])) {
 505                                         $media = ['content' => $siteinfo['player']['stream']];
 506
 507                                         if (!empty($siteinfo['player']['embed'])) {
 508                                                 $media['embed'] = $siteinfo['player']['embed'];
 509                                         }
 510
 511                                         $siteinfo[$contenttype[0]][] = $media;
 512                                 }
 513                         }
 514                 }
 515
 516                 if (!empty($siteinfo['image'])) {
 517                         $siteinfo['images'] = $siteinfo['images'] ?? [];
 518                         array_unshift($siteinfo['images'], ['url' => $siteinfo['image']]);
 519                         unset($siteinfo['image']);
 520                 }
 521
 522                 $siteinfo = self::checkMedia($url, $siteinfo);
 523
 524                 if (!empty($siteinfo['text']) && mb_strlen($siteinfo['text']) > self::MAX_DESC_COUNT) {
 525                         $siteinfo['text'] = mb_substr($siteinfo['text'], 0, self::MAX_DESC_COUNT) . '…';
 526                         $pos = mb_strrpos($siteinfo['text'], '.');
 527                         if ($pos > self::MIN_DESC_COUNT) {
 528                                 $siteinfo['text'] = mb_substr($siteinfo['text'], 0, $pos + 1);
 529                         }
 530                 }
 531
 532                 Logger::info('Siteinfo fetched', ['url' => $url, 'siteinfo' => $siteinfo]);
 533
 534                 Hook::callAll('getsiteinfo', $siteinfo);
 535
 536                 ksort($siteinfo);
 537
 538                 return $siteinfo;
 539         }
 540
 541         /**
 542          * Check the attached media elements.
 543          * Fix existing data and add missing data.
 544          *
 545          * @param string $page_url
 546          * @param array $siteinfo
 547          * @return array
 548          */
 549         private static function checkMedia(string $page_url, array $siteinfo) : array
 550         {
 551                 if (!empty($siteinfo['images'])) {
 552                         array_walk($siteinfo['images'], function (&$image) use ($page_url) {
 553                                 // According to the specifications someone could place a picture url into the content field as well.
 554                                 // But this doesn't seem to happen in the wild, so we don't cover it here.
 555                                 if (!empty($image['url'])) {
 556                                         $image['url'] = self::completeUrl($image['url'], $page_url);
 557                                         $photodata = Images::getInfoFromURLCached($image['url']);
 558                                         if (!empty($photodata) && ($photodata[0] > 50) && ($photodata[1] > 50)) {
 559                                                 $image['src'] = $image['url'];
 560                                                 $image['width'] = $photodata[0];
 561                                                 $image['height'] = $photodata[1];
 562                                                 $image['contenttype'] = $photodata['mime'];
 563                                                 unset($image['url']);
 564                                                 ksort($image);
 565                                         } else {
 566                                                 $image = [];
 567                                         }
 568                                 } else {
 569                                         $image = [];
 570                                 }
 571                         });
 572
 573                         $siteinfo['images'] = array_values(array_filter($siteinfo['images']));
 574                 }
 575
 576                 foreach (['audio', 'video'] as $element) {
 577                         if (!empty($siteinfo[$element])) {
 578                                 array_walk($siteinfo[$element], function (&$media) use ($page_url, &$siteinfo) {
 579                                         $url = '';
 580                                         $embed = '';
 581                                         $content = '';
 582                                         $contenttype = '';
 583                                         foreach (['embed', 'content', 'url'] as $field) {
 584                                                 if (!empty($media[$field])) {
 585                                                         $media[$field] = self::completeUrl($media[$field], $page_url);
 586                                                         $type = self::getContentType($media[$field]);
 587                                                         if (($type[0] ?? '') == 'text') {
 588                                                                 if ($field == 'embed') {
 589                                                                         $embed = $media[$field];
 590                                                                 } else {
 591                                                                         $url = $media[$field];
 592                                                                 }
 593                                                         } elseif (!empty($type[0])) {
 594                                                                 $content = $media[$field];
 595                                                                 $contenttype = implode('/', $type);
 596                                                         }
 597                                                 }
 598                                                 unset($media[$field]);
 599                                         }
 600
 601                                         foreach (['image', 'preview'] as $field) {
 602                                                 if (!empty($media[$field])) {
 603                                                         $media[$field] = self::completeUrl($media[$field], $page_url);
 604                                                 }
 605                                         }
 606
 607                                         if (!empty($url)) {
 608                                                 $media['url'] = $url;
 609                                         }
 610                                         if (!empty($embed)) {
 611                                                 $media['embed'] = $embed;
 612                                                 if (empty($siteinfo['player']['embed'])) {
 613                                                         $siteinfo['player']['embed'] = $embed;
 614                                                 }
 615                                         }
 616                                         if (!empty($content)) {
 617                                                 $media['src'] = $content;
 618                                         }
 619                                         if (!empty($contenttype)) {
 620                                                 $media['contenttype'] = $contenttype;
 621                                         }
 622                                         if (empty($url) && empty($content) && empty($embed)) {
 623                                                 $media = [];
 624                                         }
 625                                         ksort($media);
 626                                 });
 627
 628                                 $siteinfo[$element] = array_values(array_filter($siteinfo[$element]));
 629                         }
 630                         if (empty($siteinfo[$element])) {
 631                                 unset($siteinfo[$element]);
 632                         }
 633                 }
 634                 return $siteinfo;
 635         }
 636
 637         /**
 638          * Convert tags from CSV to an array
 639          *
 640          * @param string $string Tags
 641          * @return array with formatted Hashtags
 642          */
 643         public static function convertTagsToArray($string)
 644         {
 645                 $arr_tags = str_getcsv($string);
 646                 if (count($arr_tags)) {
 647                         // add the # sign to every tag
 648                         array_walk($arr_tags, ["self", "arrAddHashes"]);
 649
 650                         return $arr_tags;
 651                 }
 652         }
 653
 654         /**
 655          * Add a hasht sign to a string
 656          *
 657          * This method is used as callback function
 658          *
 659          * @param string $tag The pure tag name
 660          * @param int    $k   Counter for internal use
 661          * @return void
 662          */
 663         private static function arrAddHashes(&$tag, $k)
 664         {
 665                 $tag = "#" . $tag;
 666         }
 667
 668         /**
 669          * Add a scheme to an url
 670          *
 671          * The src attribute of some html elements (e.g. images)
 672          * can miss the scheme so we need to add the correct
 673          * scheme
 674          *
 675          * @param string $url    The url which possibly does have
 676          *                       a missing scheme (a link to an image)
 677          * @param string $scheme The url with a correct scheme
 678          *                       (e.g. the url from the webpage which does contain the image)
 679          *
 680          * @return string The url with a scheme
 681          */
 682         private static function completeUrl($url, $scheme)
 683         {
 684                 $urlarr = parse_url($url);
 685
 686                 // If the url does allready have an scheme
 687                 // we can stop the process here
 688                 if (isset($urlarr["scheme"])) {
 689                         return($url);
 690                 }
 691
 692                 $schemearr = parse_url($scheme);
 693
 694                 $complete = $schemearr["scheme"]."://".$schemearr["host"];
 695
 696                 if (!empty($schemearr["port"])) {
 697                         $complete .= ":".$schemearr["port"];
 698                 }
 699
 700                 if (!empty($urlarr["path"])) {
 701                         if (strpos($urlarr["path"], "/") !== 0) {
 702                                 $complete .= "/";
 703                         }
 704
 705                         $complete .= $urlarr["path"];
 706                 }
 707
 708                 if (!empty($urlarr["query"])) {
 709                         $complete .= "?".$urlarr["query"];
 710                 }
 711
 712                 if (!empty($urlarr["fragment"])) {
 713                         $complete .= "#".$urlarr["fragment"];
 714                 }
 715
 716                 return($complete);
 717         }
 718
 719         /**
 720          * Parse the Json-Ld parts of a web page
 721          *
 722          * @param array $siteinfo
 723          * @param array $jsonld
 724          * @return array siteinfo
 725          */
 726         private static function parseParts(array $siteinfo, array $jsonld)
 727         {
 728                 if (!empty($jsonld['@graph']) && is_array($jsonld['@graph'])) {
 729                         foreach ($jsonld['@graph'] as $part) {
 730                                 if (!empty($part) && is_array($part)) {
 731                                         $siteinfo = self::parseParts($siteinfo, $part);
 732                                 }
 733                         }
 734                 } elseif (!empty($jsonld['@type'])) {
 735                         $siteinfo = self::parseJsonLd($siteinfo, $jsonld);
 736                 } elseif (!empty($jsonld)) {
 737                         $keys = array_keys($jsonld);
 738                         $numeric_keys = true;
 739                         foreach ($keys as $key) {
 740                                 if (!is_int($key)) {
 741                                         $numeric_keys = false;
 742                                 }
 743                         }
 744                         if ($numeric_keys) {
 745                                 foreach ($jsonld as $part) {
 746                                         if (!empty($part) && is_array($part)) {
 747                                                 $siteinfo = self::parseParts($siteinfo, $part);
 748                                         }
 749                                 }
 750                         }
 751                 }
 752
 753                 array_walk_recursive($siteinfo, function (&$element) {
 754                         if (is_string($element)) {
 755                                 $element = trim(strip_tags(html_entity_decode($element, ENT_COMPAT, 'UTF-8')));
 756                         }
 757                 });
 758
 759                 return $siteinfo;
 760         }
 761
 762         /**
 763          * Improve the siteinfo with information from the provided JSON-LD information
 764          * @see https://jsonld.com/
 765          * @see https://schema.org/
 766          *
 767          * @param array $siteinfo
 768          * @param array $jsonld
 769          * @return array siteinfo
 770          */
 771         private static function parseJsonLd(array $siteinfo, array $jsonld)
 772         {
 773                 $type = JsonLD::fetchElement($jsonld, '@type');
 774                 if (empty($type)) {
 775                         Logger::info('Empty type', ['url' => $siteinfo['url']]);
 776                         return $siteinfo;
 777                 }
 778
 779                 // Silently ignore some types that aren't processed
 780                 if (in_array($type, ['SiteNavigationElement', 'JobPosting', 'CreativeWork', 'MusicAlbum',
 781                         'WPHeader', 'WPSideBar', 'WPFooter', 'LegalService', 'MusicRecording',
 782                         'ItemList', 'BreadcrumbList', 'Blog', 'Dataset', 'Product'])) {
 783                         return $siteinfo;
 784                 }
 785
 786                 switch ($type) {
 787                         case 'Article':
 788                         case 'AdvertiserContentArticle':
 789                         case 'NewsArticle':
 790                         case 'Report':
 791                         case 'SatiricalArticle':
 792                         case 'ScholarlyArticle':
 793                         case 'SocialMediaPosting':
 794                         case 'TechArticle':
 795                         case 'ReportageNewsArticle':
 796                         case 'SocialMediaPosting':
 797                         case 'BlogPosting':
 798                         case 'LiveBlogPosting':
 799                         case 'DiscussionForumPosting':
 800                                 return self::parseJsonLdArticle($siteinfo, $jsonld);
 801                         case 'WebPage':
 802                         case 'AboutPage':
 803                         case 'CheckoutPage':
 804                         case 'CollectionPage':
 805                         case 'ContactPage':
 806                         case 'FAQPage':
 807                         case 'ItemPage':
 808                         case 'MedicalWebPage':
 809                         case 'ProfilePage':
 810                         case 'QAPage':
 811                         case 'RealEstateListing':
 812                         case 'SearchResultsPage':
 813                         case 'MediaGallery':
 814                         case 'ImageGallery':
 815                         case 'VideoGallery':
 816                         case 'RadioEpisode':
 817                         case 'Event':
 818                                 return self::parseJsonLdWebPage($siteinfo, $jsonld);
 819                         case 'WebSite':
 820                                 return self::parseJsonLdWebSite($siteinfo, $jsonld);
 821                         case 'Organization':
 822                         case 'Airline':
 823                         case 'Consortium':
 824                         case 'Corporation':
 825                         case 'EducationalOrganization':
 826                         case 'FundingScheme':
 827                         case 'GovernmentOrganization':
 828                         case 'LibrarySystem':
 829                         case 'LocalBusiness':
 830                         case 'MedicalOrganization':
 831                         case 'NGO':
 832                         case 'NewsMediaOrganization':
 833                         case 'Project':
 834                         case 'SportsOrganization':
 835                         case 'WorkersUnion':
 836                                 return self::parseJsonLdWebOrganization($siteinfo, $jsonld);
 837                         case 'Person':
 838                         case 'Patient':
 839                         case 'PerformingGroup':
 840                         case 'DanceGroup';
 841                         case 'MusicGroup':
 842                         case 'TheaterGroup':
 843                                 return self::parseJsonLdWebPerson($siteinfo, $jsonld);
 844                         case 'AudioObject':
 845                         case 'Audio':
 846                                 return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'audio');
 847                         case 'VideoObject':
 848                                 return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'video');
 849                         case 'ImageObject':
 850                                 return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'images');
 851                         default:
 852                                 Logger::info('Unknown type', ['type' => $type, 'url' => $siteinfo['url']]);
 853                                 return $siteinfo;
 854                 }
 855         }
 856
 857         /**
 858          * Fetch author and publisher data
 859          *
 860          * @param array $siteinfo
 861          * @param array $jsonld
 862          * @return array siteinfo
 863          */
 864         private static function parseJsonLdAuthor(array $siteinfo, array $jsonld)
 865         {
 866                 $jsonldinfo = [];
 867
 868                 if (!empty($jsonld['publisher']) && is_array($jsonld['publisher'])) {
 869                         $content = JsonLD::fetchElement($jsonld, 'publisher', 'name');
 870                         if (!empty($content) && is_string($content)) {
 871                                 $jsonldinfo['publisher_name'] = trim($content);
 872                         }
 873
 874                         $content = JsonLD::fetchElement($jsonld, 'publisher', 'url');
 875                         if (!empty($content) && is_string($content)) {
 876                                 $jsonldinfo['publisher_url'] = trim($content);
 877                         }
 878
 879                         $brand = JsonLD::fetchElement($jsonld, 'publisher', 'brand', '@type', 'Organization');
 880                         if (!empty($brand) && is_array($brand)) {
 881                                 $content = JsonLD::fetchElement($brand, 'name');
 882                                 if (!empty($content) && is_string($content)) {
 883                                         $jsonldinfo['publisher_name'] = trim($content);
 884                                 }
 885
 886                                 $content = JsonLD::fetchElement($brand, 'url');
 887                                 if (!empty($content) && is_string($content)) {
 888                                         $jsonldinfo['publisher_url'] = trim($content);
 889                                 }
 890
 891                                 $content = JsonLD::fetchElement($brand, 'logo', 'url');
 892                                 if (!empty($content) && is_string($content)) {
 893                                         $jsonldinfo['publisher_img'] = trim($content);
 894                                 }
 895                         }
 896
 897                         $logo = JsonLD::fetchElement($jsonld, 'publisher', 'logo');
 898                         if (!empty($logo) && is_array($logo)) {
 899                                 $content = JsonLD::fetchElement($logo, 'url');
 900                                 if (!empty($content) && is_string($content)) {
 901                                         $jsonldinfo['publisher_img'] = trim($content);
 902                                 }
 903                         }
 904                 } elseif (!empty($jsonld['publisher']) && is_string($jsonld['publisher'])) {
 905                         $jsonldinfo['publisher_name'] = trim($jsonld['publisher']);
 906                 }
 907
 908                 if (!empty($jsonld['author']) && is_array($jsonld['author'])) {
 909                         $content = JsonLD::fetchElement($jsonld, 'author', 'name');
 910                         if (!empty($content) && is_string($content)) {
 911                                 $jsonldinfo['author_name'] = trim($content);
 912                         }
 913
 914                         $content = JsonLD::fetchElement($jsonld, 'author', 'sameAs');
 915                         if (!empty($content) && is_string($content)) {
 916                                 $jsonldinfo['author_url'] = trim($content);
 917                         }
 918
 919                         $content = JsonLD::fetchElement($jsonld, 'author', 'url');
 920                         if (!empty($content) && is_string($content)) {
 921                                 $jsonldinfo['author_url'] = trim($content);
 922                         }
 923
 924                         $logo = JsonLD::fetchElement($jsonld, 'author', 'logo');
 925                         if (!empty($logo) && is_array($logo)) {
 926                                 $content = JsonLD::fetchElement($logo, 'url');
 927                                 if (!empty($content) && is_string($content)) {
 928                                         $jsonldinfo['author_img'] = trim($content);
 929                                 }
 930                         }
 931                 } elseif (!empty($jsonld['author']) && is_string($jsonld['author'])) {
 932                         $jsonldinfo['author_name'] = trim($jsonld['author']);
 933                 }
 934
 935                 Logger::info('Fetched Author information', ['fetched' => $jsonldinfo]);
 936
 937                 return array_merge($siteinfo, $jsonldinfo);
 938         }
 939
 940         /**
 941          * Fetch data from the provided JSON-LD Article type
 942          * @see https://schema.org/Article
 943          *
 944          * @param array $siteinfo
 945          * @param array $jsonld
 946          * @return array siteinfo
 947          */
 948         private static function parseJsonLdArticle(array $siteinfo, array $jsonld)
 949         {
 950                 $jsonldinfo = [];
 951
 952                 $content = JsonLD::fetchElement($jsonld, 'headline');
 953                 if (!empty($content) && is_string($content)) {
 954                         $jsonldinfo['title'] = trim($content);
 955                 }
 956
 957                 $content = JsonLD::fetchElement($jsonld, 'alternativeHeadline');
 958                 if (!empty($content) && is_string($content) && (($jsonldinfo['title'] ?? '') != trim($content))) {
 959                         $jsonldinfo['alternative_title'] = trim($content);
 960                 }
 961
 962                 $content = JsonLD::fetchElement($jsonld, 'description');
 963                 if (!empty($content) && is_string($content)) {
 964                         $jsonldinfo['text'] = trim($content);
 965                 }
 966
 967                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
 968                 if (!empty($content)) {
 969                         $jsonldinfo['image'] = trim($content);
 970                 }
 971
 972                 $content = JsonLD::fetchElement($jsonld, 'image', 'url', '@type', 'ImageObject');
 973                 if (!empty($content) && is_string($content)) {
 974                         $jsonldinfo['image'] = trim($content);
 975                 }
 976
 977                 if (!empty($jsonld['keywords']) && !is_array($jsonld['keywords'])) {
 978                         $content = JsonLD::fetchElement($jsonld, 'keywords');
 979                         if (!empty($content)) {
 980                                 $siteinfo['keywords'] = [];
 981                                 $keywords = explode(',', $content);
 982                                 foreach ($keywords as $keyword) {
 983                                         $siteinfo['keywords'][] = trim($keyword);
 984                                 }
 985                         }
 986                 } elseif (!empty($jsonld['keywords'])) {
 987                         $content = JsonLD::fetchElementArray($jsonld, 'keywords');
 988                         if (!empty($content) && is_array($content)) {
 989                                 $jsonldinfo['keywords'] = $content;
 990                         }
 991                 }
 992
 993                 $content = JsonLD::fetchElement($jsonld, 'datePublished');
 994                 if (!empty($content) && is_string($content)) {
 995                         $jsonldinfo['published'] = DateTimeFormat::utc($content);
 996                 }
 997
 998                 $content = JsonLD::fetchElement($jsonld, 'dateModified');
 999                 if (!empty($content) && is_string($content)) {
1000                         $jsonldinfo['modified'] = DateTimeFormat::utc($content);
1001                 }
1002
1003                 $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
1004
1005                 Logger::info('Fetched article information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1006
1007                 return array_merge($siteinfo, $jsonldinfo);
1008         }
1009
1010         /**
1011          * Fetch data from the provided JSON-LD WebPage type
1012          * @see https://schema.org/WebPage
1013          *
1014          * @param array $siteinfo
1015          * @param array $jsonld
1016          * @return array siteinfo
1017          */
1018         private static function parseJsonLdWebPage(array $siteinfo, array $jsonld)
1019         {
1020                 $jsonldinfo = [];
1021
1022                 $content = JsonLD::fetchElement($jsonld, 'name');
1023                 if (!empty($content)) {
1024                         $jsonldinfo['title'] = trim($content);
1025                 }
1026
1027                 $content = JsonLD::fetchElement($jsonld, 'description');
1028                 if (!empty($content)) {
1029                         $jsonldinfo['text'] = trim($content);
1030                 }
1031
1032                 $content = JsonLD::fetchElement($jsonld, 'image');
1033                 if (!empty($content)) {
1034                         $jsonldinfo['image'] = trim($content);
1035                 }
1036
1037                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
1038                 if (!empty($content)) {
1039                         $jsonldinfo['image'] = trim($content);
1040                 }
1041
1042                 $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
1043
1044                 Logger::info('Fetched WebPage information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1045
1046                 return array_merge($siteinfo, $jsonldinfo);
1047         }
1048
1049         /**
1050          * Fetch data from the provided JSON-LD WebSite type
1051          * @see https://schema.org/WebSite
1052          *
1053          * @param array $siteinfo
1054          * @param array $jsonld
1055          * @return array siteinfo
1056          */
1057         private static function parseJsonLdWebSite(array $siteinfo, array $jsonld)
1058         {
1059                 $jsonldinfo = [];
1060
1061                 $content = JsonLD::fetchElement($jsonld, 'name');
1062                 if (!empty($content)) {
1063                         $jsonldinfo['publisher_name'] = trim($content);
1064                 }
1065
1066                 $content = JsonLD::fetchElement($jsonld, 'description');
1067                 if (!empty($content)) {
1068                         $jsonldinfo['publisher_description'] = trim($content);
1069                 }
1070
1071                 $content = JsonLD::fetchElement($jsonld, 'url');
1072                 if (!empty($content)) {
1073                         $jsonldinfo['publisher_url'] = trim($content);
1074                 }
1075
1076                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
1077                 if (!empty($content)) {
1078                         $jsonldinfo['image'] = trim($content);
1079                 }
1080
1081                 $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
1082
1083                 Logger::info('Fetched WebSite information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1084                 return array_merge($siteinfo, $jsonldinfo);
1085         }
1086
1087         /**
1088          * Fetch data from the provided JSON-LD Organization type
1089          * @see https://schema.org/Organization
1090          *
1091          * @param array $siteinfo
1092          * @param array $jsonld
1093          * @return array siteinfo
1094          */
1095         private static function parseJsonLdWebOrganization(array $siteinfo, array $jsonld)
1096         {
1097                 $jsonldinfo = [];
1098
1099                 $content = JsonLD::fetchElement($jsonld, 'name');
1100                 if (!empty($content)) {
1101                         $jsonldinfo['publisher_name'] = trim($content);
1102                 }
1103
1104                 $content = JsonLD::fetchElement($jsonld, 'description');
1105                 if (!empty($content)) {
1106                         $jsonldinfo['publisher_description'] = trim($content);
1107                 }
1108
1109                 $content = JsonLD::fetchElement($jsonld, 'url');
1110                 if (!empty($content)) {
1111                         $jsonldinfo['publisher_url'] = trim($content);
1112                 }
1113
1114                 $content = JsonLD::fetchElement($jsonld, 'logo', 'url', '@type', 'ImageObject');
1115                 if (!empty($content)) {
1116                         $jsonldinfo['publisher_img'] = trim($content);
1117                 }
1118
1119                 $content = JsonLD::fetchElement($jsonld, 'brand', 'name', '@type', 'Organization');
1120                 if (!empty($content)) {
1121                         $jsonldinfo['publisher_name'] = trim($content);
1122                 }
1123
1124                 $content = JsonLD::fetchElement($jsonld, 'brand', 'url', '@type', 'Organization');
1125                 if (!empty($content)) {
1126                         $jsonldinfo['publisher_url'] = trim($content);
1127                 }
1128
1129                 Logger::info('Fetched Organization information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1130                 return array_merge($siteinfo, $jsonldinfo);
1131         }
1132
1133         /**
1134          * Fetch data from the provided JSON-LD Person type
1135          * @see https://schema.org/Person
1136          *
1137          * @param array $siteinfo
1138          * @param array $jsonld
1139          * @return array siteinfo
1140          */
1141         private static function parseJsonLdWebPerson(array $siteinfo, array $jsonld)
1142         {
1143                 $jsonldinfo = [];
1144
1145                 $content = JsonLD::fetchElement($jsonld, 'name');
1146                 if (!empty($content)) {
1147                         $jsonldinfo['author_name'] = trim($content);
1148                 }
1149
1150                 $content = JsonLD::fetchElement($jsonld, 'description');
1151                 if (!empty($content)) {
1152                         $jsonldinfo['author_description'] = trim($content);
1153                 }
1154
1155                 $content = JsonLD::fetchElement($jsonld, 'sameAs');
1156                 if (!empty($content) && is_string($content)) {
1157                         $jsonldinfo['author_url'] = trim($content);
1158                 }
1159
1160                 $content = JsonLD::fetchElement($jsonld, 'url');
1161                 if (!empty($content)) {
1162                         $jsonldinfo['author_url'] = trim($content);
1163                 }
1164
1165                 $content = JsonLD::fetchElement($jsonld, 'image', 'url', '@type', 'ImageObject');
1166                 if (!empty($content) && !is_string($content)) {
1167                         Logger::notice('Unexpected return value for the author image', ['content' => $content]);
1168                 }
1169
1170                 if (!empty($content) && is_string($content)) {
1171                         $jsonldinfo['author_img'] = trim($content);
1172                 }
1173
1174                 Logger::info('Fetched Person information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1175                 return array_merge($siteinfo, $jsonldinfo);
1176         }
1177
1178         /**
1179          * Fetch data from the provided JSON-LD MediaObject type
1180          * @see https://schema.org/MediaObject
1181          *
1182          * @param array $siteinfo
1183          * @param array $jsonld
1184          * @return array siteinfo
1185          */
1186         private static function parseJsonLdMediaObject(array $siteinfo, array $jsonld, string $name)
1187         {
1188                 $media = [];
1189
1190                 $content = JsonLD::fetchElement($jsonld, 'caption');
1191                 if (!empty($content)) {
1192                         $media['caption'] = trim($content);
1193                 }
1194
1195                 $content = JsonLD::fetchElement($jsonld, 'url');
1196                 if (!empty($content)) {
1197                         $media['url'] = trim($content);
1198                 }
1199
1200                 $content = JsonLD::fetchElement($jsonld, 'mainEntityOfPage');
1201                 if (!empty($content)) {
1202                         $media['main'] = Strings::compareLink($content, $siteinfo['url']);
1203                 }
1204
1205                 $content = JsonLD::fetchElement($jsonld, 'description');
1206                 if (!empty($content)) {
1207                         $media['description'] = trim($content);
1208                 }
1209
1210                 $content = JsonLD::fetchElement($jsonld, 'name');
1211                 if (!empty($content) && (($media['description'] ?? '') != trim($content))) {
1212                         $media['name'] = trim($content);
1213                 }
1214
1215                 $content = JsonLD::fetchElement($jsonld, 'contentUrl');
1216                 if (!empty($content)) {
1217                         $media['content'] = trim($content);
1218                 }
1219
1220                 $content = JsonLD::fetchElement($jsonld, 'embedUrl');
1221                 if (!empty($content)) {
1222                         $media['embed'] = trim($content);
1223                 }
1224
1225                 $content = JsonLD::fetchElement($jsonld, 'height');
1226                 if (!empty($content)) {
1227                         $media['height'] = trim($content);
1228                 }
1229
1230                 $content = JsonLD::fetchElement($jsonld, 'width');
1231                 if (!empty($content)) {
1232                         $media['width'] = trim($content);
1233                 }
1234
1235                 $content = JsonLD::fetchElement($jsonld, 'image');
1236                 if (!empty($content)) {
1237                         $media['image'] = trim($content);
1238                 }
1239
1240                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
1241                 if (!empty($content) && (($media['image'] ?? '') != trim($content))) {
1242                         if (!empty($media['image'])) {
1243                                 $media['preview'] = trim($content);
1244                         } else {
1245                                 $media['image'] = trim($content);
1246                         }
1247                 }
1248
1249                 Logger::info('Fetched Media information', ['url' => $siteinfo['url'], 'fetched' => $media]);
1250                 $siteinfo[$name][] = $media;
1251                 return $siteinfo;
1252         }
1253 }