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