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 ($page_url) {
 546                                 /*
 547                                  * According to the specifications someone could place a picture
 548                                  * URL into the content field as well. But this doesn't seem to
 549                                  * happen in the wild, so we don't cover it here.
 550                                  */
 551                                 if (!empty($image['url'])) {
 552                                         $image['url'] = self::completeUrl($image['url'], $page_url);
 553                                         $photodata = Images::getInfoFromURLCached($image['url']);
 554                                         if (!empty($photodata) && ($photodata[0] > 50) && ($photodata[1] > 50)) {
 555                                                 $image['src'] = $image['url'];
 556                                                 $image['width'] = $photodata[0];
 557                                                 $image['height'] = $photodata[1];
 558                                                 $image['contenttype'] = $photodata['mime'];
 559                                                 unset($image['url']);
 560                                                 ksort($image);
 561                                         } else {
 562                                                 $image = [];
 563                                         }
 564                                 } else {
 565                                         $image = [];
 566                                 }
 567                         });
 568
 569                         $siteinfo['images'] = array_values(array_filter($siteinfo['images']));
 570                 }
 571
 572                 foreach (['audio', 'video'] as $element) {
 573                         if (!empty($siteinfo[$element])) {
 574                                 array_walk($siteinfo[$element], function (&$media) use ($page_url, &$siteinfo) {
 575                                         $url = '';
 576                                         $embed = '';
 577                                         $content = '';
 578                                         $contenttype = '';
 579                                         foreach (['embed', 'content', 'url'] as $field) {
 580                                                 if (!empty($media[$field])) {
 581                                                         $media[$field] = self::completeUrl($media[$field], $page_url);
 582                                                         $type = self::getContentType($media[$field]);
 583                                                         if (($type[0] ?? '') == 'text') {
 584                                                                 if ($field == 'embed') {
 585                                                                         $embed = $media[$field];
 586                                                                 } else {
 587                                                                         $url = $media[$field];
 588                                                                 }
 589                                                         } elseif (!empty($type[0])) {
 590                                                                 $content = $media[$field];
 591                                                                 $contenttype = implode('/', $type);
 592                                                         }
 593                                                 }
 594                                                 unset($media[$field]);
 595                                         }
 596
 597                                         foreach (['image', 'preview'] as $field) {
 598                                                 if (!empty($media[$field])) {
 599                                                         $media[$field] = self::completeUrl($media[$field], $page_url);
 600                                                 }
 601                                         }
 602
 603                                         if (!empty($url)) {
 604                                                 $media['url'] = $url;
 605                                         }
 606                                         if (!empty($embed)) {
 607                                                 $media['embed'] = $embed;
 608                                                 if (empty($siteinfo['player']['embed'])) {
 609                                                         $siteinfo['player']['embed'] = $embed;
 610                                                 }
 611                                         }
 612                                         if (!empty($content)) {
 613                                                 $media['src'] = $content;
 614                                         }
 615                                         if (!empty($contenttype)) {
 616                                                 $media['contenttype'] = $contenttype;
 617                                         }
 618                                         if (empty($url) && empty($content) && empty($embed)) {
 619                                                 $media = [];
 620                                         }
 621                                         ksort($media);
 622                                 });
 623
 624                                 $siteinfo[$element] = array_values(array_filter($siteinfo[$element]));
 625                         }
 626                         if (empty($siteinfo[$element])) {
 627                                 unset($siteinfo[$element]);
 628                         }
 629                 }
 630                 return $siteinfo;
 631         }
 632
 633         /**
 634          * Convert tags from CSV to an array
 635          *
 636          * @param string $string Tags
 637          * @return array with formatted Hashtags
 638          */
 639         public static function convertTagsToArray(string $string): array
 640         {
 641                 $arr_tags = str_getcsv($string);
 642                 if (count($arr_tags)) {
 643                         // add the # sign to every tag
 644                         array_walk($arr_tags, ['self', 'arrAddHashes']);
 645
 646                         return $arr_tags;
 647                 }
 648                 return [];
 649         }
 650
 651         /**
 652          * Add a hasht sign to a string
 653          *
 654          * This method is used as callback function
 655          *
 656          * @param string $tag The pure tag name
 657          * @param int    $k   Counter for internal use
 658          * @return void
 659          */
 660         private static function arrAddHashes(string &$tag, int $k)
 661         {
 662                 $tag = '#' . $tag;
 663         }
 664
 665         /**
 666          * Add a scheme to an url
 667          *
 668          * The src attribute of some html elements (e.g. images)
 669          * can miss the scheme so we need to add the correct
 670          * scheme
 671          *
 672          * @param string $url    The url which possibly does have
 673          *                       a missing scheme (a link to an image)
 674          * @param string $scheme The url with a correct scheme
 675          *                       (e.g. the url from the webpage which does contain the image)
 676          *
 677          * @return string The url with a scheme
 678          */
 679         private static function completeUrl(string $url, string $scheme): string
 680         {
 681                 $urlarr = parse_url($url);
 682
 683                 // If the url does allready have an scheme
 684                 // we can stop the process here
 685                 if (isset($urlarr['scheme'])) {
 686                         return $url;
 687                 }
 688
 689                 $schemearr = parse_url($scheme);
 690
 691                 $complete = $schemearr['scheme'] . '://' . $schemearr['host'];
 692
 693                 if (!empty($schemearr['port'])) {
 694                         $complete .= ':' . $schemearr['port'];
 695                 }
 696
 697                 if (!empty($urlarr['path'])) {
 698                         if (strpos($urlarr['path'], '/') !== 0) {
 699                                 $complete .= '/';
 700                         }
 701
 702                         $complete .= $urlarr['path'];
 703                 }
 704
 705                 if (!empty($urlarr['query'])) {
 706                         $complete .= '?' . $urlarr['query'];
 707                 }
 708
 709                 if (!empty($urlarr['fragment'])) {
 710                         $complete .= '#' . $urlarr['fragment'];
 711                 }
 712
 713                 return $complete;
 714         }
 715
 716         /**
 717          * Parse the Json-Ld parts of a web page
 718          *
 719          * @param array $siteinfo
 720          * @param array $jsonld
 721          * @return array siteinfo
 722          */
 723         private static function parseParts(array $siteinfo, array $jsonld): array
 724         {
 725                 if (!empty($jsonld['@graph']) && is_array($jsonld['@graph'])) {
 726                         foreach ($jsonld['@graph'] as $part) {
 727                                 if (!empty($part) && is_array($part)) {
 728                                         $siteinfo = self::parseParts($siteinfo, $part);
 729                                 }
 730                         }
 731                 } elseif (!empty($jsonld['@type'])) {
 732                         $siteinfo = self::parseJsonLd($siteinfo, $jsonld);
 733                 } elseif (!empty($jsonld)) {
 734                         $keys = array_keys($jsonld);
 735                         $numeric_keys = true;
 736                         foreach ($keys as $key) {
 737                                 if (!is_int($key)) {
 738                                         $numeric_keys = false;
 739                                 }
 740                         }
 741                         if ($numeric_keys) {
 742                                 foreach ($jsonld as $part) {
 743                                         if (!empty($part) && is_array($part)) {
 744                                                 $siteinfo = self::parseParts($siteinfo, $part);
 745                                         }
 746                                 }
 747                         }
 748                 }
 749
 750                 array_walk_recursive($siteinfo, function (&$element) {
 751                         if (is_string($element)) {
 752                                 $element = trim(strip_tags(html_entity_decode($element, ENT_COMPAT, 'UTF-8')));
 753                         }
 754                 });
 755
 756                 return $siteinfo;
 757         }
 758
 759         /**
 760          * Improve the siteinfo with information from the provided JSON-LD information
 761          * @see https://jsonld.com/
 762          * @see https://schema.org/
 763          *
 764          * @param array $siteinfo
 765          * @param array $jsonld
 766          * @return array siteinfo
 767          */
 768         private static function parseJsonLd(array $siteinfo, array $jsonld): array
 769         {
 770                 $type = JsonLD::fetchElement($jsonld, '@type');
 771                 if (empty($type)) {
 772                         Logger::info('Empty type', ['url' => $siteinfo['url']]);
 773                         return $siteinfo;
 774                 }
 775
 776                 // Silently ignore some types that aren't processed
 777                 if (in_array($type, ['SiteNavigationElement', 'JobPosting', 'CreativeWork', 'MusicAlbum',
 778                         'WPHeader', 'WPSideBar', 'WPFooter', 'LegalService', 'MusicRecording',
 779                         'ItemList', 'BreadcrumbList', 'Blog', 'Dataset', 'Product'])) {
 780                         return $siteinfo;
 781                 }
 782
 783                 switch ($type) {
 784                         case 'Article':
 785                         case 'AdvertiserContentArticle':
 786                         case 'NewsArticle':
 787                         case 'Report':
 788                         case 'SatiricalArticle':
 789                         case 'ScholarlyArticle':
 790                         case 'SocialMediaPosting':
 791                         case 'TechArticle':
 792                         case 'ReportageNewsArticle':
 793                         case 'SocialMediaPosting':
 794                         case 'BlogPosting':
 795                         case 'LiveBlogPosting':
 796                         case 'DiscussionForumPosting':
 797                                 return self::parseJsonLdArticle($siteinfo, $jsonld);
 798                         case 'WebPage':
 799                         case 'AboutPage':
 800                         case 'CheckoutPage':
 801                         case 'CollectionPage':
 802                         case 'ContactPage':
 803                         case 'FAQPage':
 804                         case 'ItemPage':
 805                         case 'MedicalWebPage':
 806                         case 'ProfilePage':
 807                         case 'QAPage':
 808                         case 'RealEstateListing':
 809                         case 'SearchResultsPage':
 810                         case 'MediaGallery':
 811                         case 'ImageGallery':
 812                         case 'VideoGallery':
 813                         case 'RadioEpisode':
 814                         case 'Event':
 815                                 return self::parseJsonLdWebPage($siteinfo, $jsonld);
 816                         case 'WebSite':
 817                                 return self::parseJsonLdWebSite($siteinfo, $jsonld);
 818                         case 'Organization':
 819                         case 'Airline':
 820                         case 'Consortium':
 821                         case 'Corporation':
 822                         case 'EducationalOrganization':
 823                         case 'FundingScheme':
 824                         case 'GovernmentOrganization':
 825                         case 'LibrarySystem':
 826                         case 'LocalBusiness':
 827                         case 'MedicalOrganization':
 828                         case 'NGO':
 829                         case 'NewsMediaOrganization':
 830                         case 'Project':
 831                         case 'SportsOrganization':
 832                         case 'WorkersUnion':
 833                                 return self::parseJsonLdWebOrganization($siteinfo, $jsonld);
 834                         case 'Person':
 835                         case 'Patient':
 836                         case 'PerformingGroup':
 837                         case 'DanceGroup';
 838                         case 'MusicGroup':
 839                         case 'TheaterGroup':
 840                                 return self::parseJsonLdWebPerson($siteinfo, $jsonld);
 841                         case 'AudioObject':
 842                         case 'Audio':
 843                                 return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'audio');
 844                         case 'VideoObject':
 845                                 return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'video');
 846                         case 'ImageObject':
 847                                 return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'images');
 848                         default:
 849                                 Logger::info('Unknown type', ['type' => $type, 'url' => $siteinfo['url']]);
 850                                 return $siteinfo;
 851                 }
 852         }
 853
 854         /**
 855          * Fetch author and publisher data
 856          *
 857          * @param array $siteinfo
 858          * @param array $jsonld
 859          * @return array siteinfo
 860          */
 861         private static function parseJsonLdAuthor(array $siteinfo, array $jsonld): array
 862         {
 863                 $jsonldinfo = [];
 864
 865                 if (!empty($jsonld['publisher']) && is_array($jsonld['publisher'])) {
 866                         $content = JsonLD::fetchElement($jsonld, 'publisher', 'name');
 867                         if (!empty($content) && is_string($content)) {
 868                                 $jsonldinfo['publisher_name'] = trim($content);
 869                         }
 870
 871                         $content = JsonLD::fetchElement($jsonld, 'publisher', 'url');
 872                         if (!empty($content) && is_string($content)) {
 873                                 $jsonldinfo['publisher_url'] = trim($content);
 874                         }
 875
 876                         $brand = JsonLD::fetchElement($jsonld, 'publisher', 'brand', '@type', 'Organization');
 877                         if (!empty($brand) && is_array($brand)) {
 878                                 $content = JsonLD::fetchElement($brand, 'name');
 879                                 if (!empty($content) && is_string($content)) {
 880                                         $jsonldinfo['publisher_name'] = trim($content);
 881                                 }
 882
 883                                 $content = JsonLD::fetchElement($brand, 'url');
 884                                 if (!empty($content) && is_string($content)) {
 885                                         $jsonldinfo['publisher_url'] = trim($content);
 886                                 }
 887
 888                                 $content = JsonLD::fetchElement($brand, 'logo', 'url');
 889                                 if (!empty($content) && is_string($content)) {
 890                                         $jsonldinfo['publisher_img'] = trim($content);
 891                                 }
 892                         }
 893
 894                         $logo = JsonLD::fetchElement($jsonld, 'publisher', 'logo');
 895                         if (!empty($logo) && is_array($logo)) {
 896                                 $content = JsonLD::fetchElement($logo, 'url');
 897                                 if (!empty($content) && is_string($content)) {
 898                                         $jsonldinfo['publisher_img'] = trim($content);
 899                                 }
 900                         }
 901                 } elseif (!empty($jsonld['publisher']) && is_string($jsonld['publisher'])) {
 902                         $jsonldinfo['publisher_name'] = trim($jsonld['publisher']);
 903                 }
 904
 905                 if (!empty($jsonld['author']) && is_array($jsonld['author'])) {
 906                         $content = JsonLD::fetchElement($jsonld, 'author', 'name');
 907                         if (!empty($content) && is_string($content)) {
 908                                 $jsonldinfo['author_name'] = trim($content);
 909                         }
 910
 911                         $content = JsonLD::fetchElement($jsonld, 'author', 'sameAs');
 912                         if (!empty($content) && is_string($content)) {
 913                                 $jsonldinfo['author_url'] = trim($content);
 914                         }
 915
 916                         $content = JsonLD::fetchElement($jsonld, 'author', 'url');
 917                         if (!empty($content) && is_string($content)) {
 918                                 $jsonldinfo['author_url'] = trim($content);
 919                         }
 920
 921                         $logo = JsonLD::fetchElement($jsonld, 'author', 'logo');
 922                         if (!empty($logo) && is_array($logo)) {
 923                                 $content = JsonLD::fetchElement($logo, 'url');
 924                                 if (!empty($content) && is_string($content)) {
 925                                         $jsonldinfo['author_img'] = trim($content);
 926                                 }
 927                         }
 928                 } elseif (!empty($jsonld['author']) && is_string($jsonld['author'])) {
 929                         $jsonldinfo['author_name'] = trim($jsonld['author']);
 930                 }
 931
 932                 Logger::info('Fetched Author information', ['fetched' => $jsonldinfo]);
 933
 934                 return array_merge($siteinfo, $jsonldinfo);
 935         }
 936
 937         /**
 938          * Fetch data from the provided JSON-LD Article type
 939          * @see https://schema.org/Article
 940          *
 941          * @param array $siteinfo
 942          * @param array $jsonld
 943          * @return array siteinfo
 944          */
 945         private static function parseJsonLdArticle(array $siteinfo, array $jsonld): array
 946         {
 947                 $jsonldinfo = [];
 948
 949                 $content = JsonLD::fetchElement($jsonld, 'headline');
 950                 if (!empty($content) && is_string($content)) {
 951                         $jsonldinfo['title'] = trim($content);
 952                 }
 953
 954                 $content = JsonLD::fetchElement($jsonld, 'alternativeHeadline');
 955                 if (!empty($content) && is_string($content) && (($jsonldinfo['title'] ?? '') != trim($content))) {
 956                         $jsonldinfo['alternative_title'] = trim($content);
 957                 }
 958
 959                 $content = JsonLD::fetchElement($jsonld, 'description');
 960                 if (!empty($content) && is_string($content)) {
 961                         $jsonldinfo['text'] = trim($content);
 962                 }
 963
 964                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
 965                 if (!empty($content)) {
 966                         $jsonldinfo['image'] = trim($content);
 967                 }
 968
 969                 $content = JsonLD::fetchElement($jsonld, 'image', 'url', '@type', 'ImageObject');
 970                 if (!empty($content) && is_string($content)) {
 971                         $jsonldinfo['image'] = trim($content);
 972                 }
 973
 974                 if (!empty($jsonld['keywords']) && !is_array($jsonld['keywords'])) {
 975                         $content = JsonLD::fetchElement($jsonld, 'keywords');
 976                         if (!empty($content)) {
 977                                 $siteinfo['keywords'] = [];
 978                                 $keywords = explode(',', $content);
 979                                 foreach ($keywords as $keyword) {
 980                                         $siteinfo['keywords'][] = trim($keyword);
 981                                 }
 982                         }
 983                 } elseif (!empty($jsonld['keywords'])) {
 984                         $content = JsonLD::fetchElementArray($jsonld, 'keywords');
 985                         if (!empty($content) && is_array($content)) {
 986                                 $jsonldinfo['keywords'] = $content;
 987                         }
 988                 }
 989
 990                 $content = JsonLD::fetchElement($jsonld, 'datePublished');
 991                 if (!empty($content) && is_string($content)) {
 992                         $jsonldinfo['published'] = DateTimeFormat::utc($content);
 993                 }
 994
 995                 $content = JsonLD::fetchElement($jsonld, 'dateModified');
 996                 if (!empty($content) && is_string($content)) {
 997                         $jsonldinfo['modified'] = DateTimeFormat::utc($content);
 998                 }
 999
1000                 $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
1001
1002                 Logger::info('Fetched article information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1003
1004                 return array_merge($siteinfo, $jsonldinfo);
1005         }
1006
1007         /**
1008          * Fetch data from the provided JSON-LD WebPage type
1009          * @see https://schema.org/WebPage
1010          *
1011          * @param array $siteinfo
1012          * @param array $jsonld
1013          * @return array siteinfo
1014          */
1015         private static function parseJsonLdWebPage(array $siteinfo, array $jsonld): array
1016         {
1017                 $jsonldinfo = [];
1018
1019                 $content = JsonLD::fetchElement($jsonld, 'name');
1020                 if (!empty($content)) {
1021                         $jsonldinfo['title'] = trim($content);
1022                 }
1023
1024                 $content = JsonLD::fetchElement($jsonld, 'description');
1025                 if (!empty($content)) {
1026                         $jsonldinfo['text'] = trim($content);
1027                 }
1028
1029                 $content = JsonLD::fetchElement($jsonld, 'image');
1030                 if (!empty($content)) {
1031                         $jsonldinfo['image'] = trim($content);
1032                 }
1033
1034                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
1035                 if (!empty($content)) {
1036                         $jsonldinfo['image'] = trim($content);
1037                 }
1038
1039                 $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
1040
1041                 Logger::info('Fetched WebPage information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1042
1043                 return array_merge($siteinfo, $jsonldinfo);
1044         }
1045
1046         /**
1047          * Fetch data from the provided JSON-LD WebSite type
1048          * @see https://schema.org/WebSite
1049          *
1050          * @param array $siteinfo
1051          * @param array $jsonld
1052          * @return array siteinfo
1053          */
1054         private static function parseJsonLdWebSite(array $siteinfo, array $jsonld): array
1055         {
1056                 $jsonldinfo = [];
1057
1058                 $content = JsonLD::fetchElement($jsonld, 'name');
1059                 if (!empty($content)) {
1060                         $jsonldinfo['publisher_name'] = trim($content);
1061                 }
1062
1063                 $content = JsonLD::fetchElement($jsonld, 'description');
1064                 if (!empty($content)) {
1065                         $jsonldinfo['publisher_description'] = trim($content);
1066                 }
1067
1068                 $content = JsonLD::fetchElement($jsonld, 'url');
1069                 if (!empty($content)) {
1070                         $jsonldinfo['publisher_url'] = trim($content);
1071                 }
1072
1073                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
1074                 if (!empty($content)) {
1075                         $jsonldinfo['image'] = trim($content);
1076                 }
1077
1078                 $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
1079
1080                 Logger::info('Fetched WebSite information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1081                 return array_merge($siteinfo, $jsonldinfo);
1082         }
1083
1084         /**
1085          * Fetch data from the provided JSON-LD Organization type
1086          * @see https://schema.org/Organization
1087          *
1088          * @param array $siteinfo
1089          * @param array $jsonld
1090          * @return array siteinfo
1091          */
1092         private static function parseJsonLdWebOrganization(array $siteinfo, array $jsonld): array
1093         {
1094                 $jsonldinfo = [];
1095
1096                 $content = JsonLD::fetchElement($jsonld, 'name');
1097                 if (!empty($content)) {
1098                         $jsonldinfo['publisher_name'] = trim($content);
1099                 }
1100
1101                 $content = JsonLD::fetchElement($jsonld, 'description');
1102                 if (!empty($content)) {
1103                         $jsonldinfo['publisher_description'] = trim($content);
1104                 }
1105
1106                 $content = JsonLD::fetchElement($jsonld, 'url');
1107                 if (!empty($content)) {
1108                         $jsonldinfo['publisher_url'] = trim($content);
1109                 }
1110
1111                 $content = JsonLD::fetchElement($jsonld, 'logo', 'url', '@type', 'ImageObject');
1112                 if (!empty($content)) {
1113                         $jsonldinfo['publisher_img'] = trim($content);
1114                 }
1115
1116                 $content = JsonLD::fetchElement($jsonld, 'brand', 'name', '@type', 'Organization');
1117                 if (!empty($content)) {
1118                         $jsonldinfo['publisher_name'] = trim($content);
1119                 }
1120
1121                 $content = JsonLD::fetchElement($jsonld, 'brand', 'url', '@type', 'Organization');
1122                 if (!empty($content)) {
1123                         $jsonldinfo['publisher_url'] = trim($content);
1124                 }
1125
1126                 Logger::info('Fetched Organization information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1127                 return array_merge($siteinfo, $jsonldinfo);
1128         }
1129
1130         /**
1131          * Fetch data from the provided JSON-LD Person type
1132          * @see https://schema.org/Person
1133          *
1134          * @param array $siteinfo
1135          * @param array $jsonld
1136          * @return array siteinfo
1137          */
1138         private static function parseJsonLdWebPerson(array $siteinfo, array $jsonld): array
1139         {
1140                 $jsonldinfo = [];
1141
1142                 $content = JsonLD::fetchElement($jsonld, 'name');
1143                 if (!empty($content)) {
1144                         $jsonldinfo['author_name'] = trim($content);
1145                 }
1146
1147                 $content = JsonLD::fetchElement($jsonld, 'description');
1148                 if (!empty($content)) {
1149                         $jsonldinfo['author_description'] = trim($content);
1150                 }
1151
1152                 $content = JsonLD::fetchElement($jsonld, 'sameAs');
1153                 if (!empty($content) && is_string($content)) {
1154                         $jsonldinfo['author_url'] = trim($content);
1155                 }
1156
1157                 $content = JsonLD::fetchElement($jsonld, 'url');
1158                 if (!empty($content)) {
1159                         $jsonldinfo['author_url'] = trim($content);
1160                 }
1161
1162                 $content = JsonLD::fetchElement($jsonld, 'image', 'url', '@type', 'ImageObject');
1163                 if (!empty($content) && !is_string($content)) {
1164                         Logger::notice('Unexpected return value for the author image', ['content' => $content]);
1165                 }
1166
1167                 if (!empty($content) && is_string($content)) {
1168                         $jsonldinfo['author_img'] = trim($content);
1169                 }
1170
1171                 Logger::info('Fetched Person information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1172                 return array_merge($siteinfo, $jsonldinfo);
1173         }
1174
1175         /**
1176          * Fetch data from the provided JSON-LD MediaObject type
1177          * @see https://schema.org/MediaObject
1178          *
1179          * @param array $siteinfo
1180          * @param array $jsonld
1181          * @return array siteinfo
1182          */
1183         private static function parseJsonLdMediaObject(array $siteinfo, array $jsonld, string $name): array
1184         {
1185                 $media = [];
1186
1187                 $content = JsonLD::fetchElement($jsonld, 'caption');
1188                 if (!empty($content)) {
1189                         $media['caption'] = trim($content);
1190                 }
1191
1192                 $content = JsonLD::fetchElement($jsonld, 'url');
1193                 if (!empty($content)) {
1194                         $media['url'] = trim($content);
1195                 }
1196
1197                 $content = JsonLD::fetchElement($jsonld, 'mainEntityOfPage');
1198                 if (!empty($content)) {
1199                         $media['main'] = Strings::compareLink($content, $siteinfo['url']);
1200                 }
1201
1202                 $content = JsonLD::fetchElement($jsonld, 'description');
1203                 if (!empty($content)) {
1204                         $media['description'] = trim($content);
1205                 }
1206
1207                 $content = JsonLD::fetchElement($jsonld, 'name');
1208                 if (!empty($content) && (($media['description'] ?? '') != trim($content))) {
1209                         $media['name'] = trim($content);
1210                 }
1211
1212                 $content = JsonLD::fetchElement($jsonld, 'contentUrl');
1213                 if (!empty($content)) {
1214                         $media['content'] = trim($content);
1215                 }
1216
1217                 $content = JsonLD::fetchElement($jsonld, 'embedUrl');
1218                 if (!empty($content)) {
1219                         $media['embed'] = trim($content);
1220                 }
1221
1222                 $content = JsonLD::fetchElement($jsonld, 'height');
1223                 if (!empty($content)) {
1224                         $media['height'] = trim($content);
1225                 }
1226
1227                 $content = JsonLD::fetchElement($jsonld, 'width');
1228                 if (!empty($content)) {
1229                         $media['width'] = trim($content);
1230                 }
1231
1232                 $content = JsonLD::fetchElement($jsonld, 'image');
1233                 if (!empty($content)) {
1234                         $media['image'] = trim($content);
1235                 }
1236
1237                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
1238                 if (!empty($content) && (($media['image'] ?? '') != trim($content))) {
1239                         if (!empty($media['image'])) {
1240                                 $media['preview'] = trim($content);
1241                         } else {
1242                                 $media['image'] = trim($content);
1243                         }
1244                 }
1245
1246                 Logger::info('Fetched Media information', ['url' => $siteinfo['url'], 'fetched' => $media]);
1247                 $siteinfo[$name][] = $media;
1248                 return $siteinfo;
1249         }
1250 }