src/Util/ParseUrl.php

   1 <?php
   2 /**
   3  * @copyright Copyright (C) 2010-2023, 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\Content\Text\HTML;
  28 use Friendica\Protocol\HTTP\MediaType;
  29 use Friendica\Core\Hook;
  30 use Friendica\Core\Logger;
  31 use Friendica\Database\Database;
  32 use Friendica\Database\DBA;
  33 use Friendica\DI;
  34 use Friendica\Network\HTTPClient\Client\HttpClientAccept;
  35 use Friendica\Network\HTTPException;
  36 use Friendica\Network\HTTPClient\Client\HttpClientOptions;
  37
  38 /**
  39  * Get information about a given URL
  40  *
  41  * Class with methods for extracting certain content from an url
  42  */
  43 class ParseUrl
  44 {
  45         const DEFAULT_EXPIRATION_FAILURE = 'now + 1 day';
  46         const DEFAULT_EXPIRATION_SUCCESS = 'now + 3 months';
  47
  48         /**
  49          * Maximum number of characters for the description
  50          */
  51         const MAX_DESC_COUNT = 250;
  52
  53         /**
  54          * Minimum number of characters for the description
  55          */
  56         const MIN_DESC_COUNT = 100;
  57
  58         /**
  59          * Fetch the content type of the given url
  60          * @param string $url    URL of the page
  61          * @param string $accept content-type to accept
  62          * @param int    $timeout
  63          * @return array content type
  64          */
  65         public static function getContentType(string $url, string $accept = HttpClientAccept::DEFAULT, int $timeout = 0): array
  66         {
  67                 if (!empty($timeout)) {
  68                         $options = [HttpClientOptions::TIMEOUT => $timeout];
  69                 } else {
  70                         $options = [];
  71                 }
  72
  73                 $curlResult = DI::httpClient()->head($url, array_merge([HttpClientOptions::ACCEPT_CONTENT => $accept], $options));
  74
  75                 // Workaround for systems that can't handle a HEAD request. Don't retry on timeouts.
  76                 if (!$curlResult->isSuccess() && ($curlResult->getReturnCode() >= 400) && !in_array($curlResult->getReturnCode(), [408, 504])) {
  77                         $curlResult = DI::httpClient()->get($url, $accept, array_merge([HttpClientOptions::CONTENT_LENGTH => 1000000], $options));
  78                 }
  79
  80                 if (!$curlResult->isSuccess()) {
  81                         Logger::debug('Got HTTP Error', ['http error' => $curlResult->getReturnCode(), 'url' => $url]);
  82                         return [];
  83                 }
  84
  85                 $contenttype =  $curlResult->getHeader('Content-Type')[0] ?? '';
  86                 if (empty($contenttype)) {
  87                         return ['application', 'octet-stream'];
  88                 }
  89
  90                 return explode('/', current(explode(';', $contenttype)));
  91         }
  92
  93         /**
  94          * Search for chached embeddable data of an url otherwise fetch it
  95          *
  96          * @param string $url         The url of the page which should be scraped
  97          * @param bool   $do_oembed   The false option is used by the function fetch_oembed()
  98          *                            to avoid endless loops
  99          *
 100          * @return array which contains needed data for embedding
 101          *    string 'url'      => The url of the parsed page
 102          *    string 'type'     => Content type
 103          *    string 'title'    => (optional) The title of the content
 104          *    string 'text'     => (optional) The description for the content
 105          *    string 'image'    => (optional) A preview image of the content
 106          *    array  'images'   => (optional) Array of preview pictures
 107          *    string 'keywords' => (optional) The tags which belong to the content
 108          *
 109          * @throws HTTPException\InternalServerErrorException
 110          * @see   ParseUrl::getSiteinfo() for more information about scraping
 111          * embeddable content
 112          */
 113         public static function getSiteinfoCached(string $url, bool $do_oembed = true): array
 114         {
 115                 if (empty($url)) {
 116                         return [
 117                                 'url' => '',
 118                                 'type' => 'error',
 119                         ];
 120                 }
 121
 122                 $urlHash = hash('sha256', $url);
 123
 124                 $parsed_url = DBA::selectFirst('parsed_url', ['content'],
 125                         ['url_hash' => $urlHash, 'oembed' => $do_oembed]
 126                 );
 127                 if (!empty($parsed_url['content'])) {
 128                         $data = unserialize($parsed_url['content']);
 129                         return $data;
 130                 }
 131
 132                 $data = self::getSiteinfo($url, $do_oembed);
 133
 134                 $expires = $data['expires'];
 135
 136                 unset($data['expires']);
 137
 138                 DI::dba()->insert(
 139                         'parsed_url',
 140                         [
 141                                 'url_hash' => $urlHash,
 142                                 'oembed'   => $do_oembed,
 143                                 'url'      => $url,
 144                                 'content'  => serialize($data),
 145                                 'created'  => DateTimeFormat::utcNow(),
 146                                 'expires'  => $expires,
 147                         ],
 148                         Database::INSERT_UPDATE
 149                 );
 150
 151                 return $data;
 152         }
 153
 154         /**
 155          * Parse a page for embeddable content information
 156          *
 157          * This method parses to url for meta data which can be used to embed
 158          * the content. If available it prioritizes Open Graph meta tags.
 159          * If this is not available it uses the twitter cards meta tags.
 160          * As fallback it uses standard html elements with meta informations
 161          * like \<title\>Awesome Title\</title\> or
 162          * \<meta name="description" content="An awesome description"\>
 163          *
 164          * @param string $url         The url of the page which should be scraped
 165          * @param bool   $do_oembed   The false option is used by the function fetch_oembed()
 166          *                            to avoid endless loops
 167          * @param int    $count       Internal counter to avoid endless loops
 168          *
 169          * @return array which contains needed data for embedding
 170          *    string 'url'      => The url of the parsed page
 171          *    string 'type'     => Content type (error, link, photo, image, audio, video)
 172          *    string 'title'    => (optional) The title of the content
 173          *    string 'text'     => (optional) The description for the content
 174          *    string 'image'    => (optional) A preview image of the content
 175          *    array  'images'   => (optional) Array of preview pictures
 176          *    string 'keywords' => (optional) The tags which belong to the content
 177          *
 178          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 179          * @todo  https://developers.google.com/+/plugins/snippet/
 180          * @verbatim
 181          * <meta itemprop="name" content="Awesome title">
 182          * <meta itemprop="description" content="An awesome description">
 183          * <meta itemprop="image" content="http://maple.libertreeproject.org/images/tree-icon.png">
 184          *
 185          * <body itemscope itemtype="http://schema.org/Product">
 186          *   <h1 itemprop="name">Shiny Trinket</h1>
 187          *   <img itemprop="image" src="{image-url}" />
 188          *   <p itemprop="description">Shiny trinkets are shiny.</p>
 189          * </body>
 190          * @endverbatim
 191          */
 192         public static function getSiteinfo(string $url, bool $do_oembed = true, int $count = 1): array
 193         {
 194                 if (empty($url)) {
 195                         return [
 196                                 'url' => '',
 197                                 'type' => 'error',
 198                         ];
 199                 }
 200
 201                 // Check if the URL does contain a scheme
 202                 $scheme = parse_url($url, PHP_URL_SCHEME);
 203
 204                 if ($scheme == '') {
 205                         $url = 'http://' . ltrim($url, '/');
 206                 }
 207
 208                 $url = trim($url, "'\"");
 209
 210                 $url = Network::stripTrackingQueryParams($url);
 211
 212                 $siteinfo = [
 213                         'url' => $url,
 214                         'type' => 'link',
 215                         'expires' => DateTimeFormat::utc(self::DEFAULT_EXPIRATION_FAILURE),
 216                 ];
 217
 218                 if ($count > 10) {
 219                         Logger::warning('Endless loop detected', ['url' => $url]);
 220                         return $siteinfo;
 221                 }
 222
 223                 $type = self::getContentType($url);
 224                 Logger::info('Got content-type', ['content-type' => $type, 'url' => $url]);
 225                 if (!empty($type) && in_array($type[0], ['image', 'video', 'audio'])) {
 226                         $siteinfo['type'] = $type[0];
 227                         return $siteinfo;
 228                 }
 229
 230                 if ((count($type) >= 2) && (($type[0] != 'text') || ($type[1] != 'html'))) {
 231                         Logger::info('Unparseable content-type, quitting here, ', ['content-type' => $type, 'url' => $url]);
 232                         return $siteinfo;
 233                 }
 234
 235                 $curlResult = DI::httpClient()->get($url, HttpClientAccept::HTML, [HttpClientOptions::CONTENT_LENGTH => 1000000]);
 236                 if (!$curlResult->isSuccess() || empty($curlResult->getBody())) {
 237                         Logger::info('Empty body or error when fetching', ['url' => $url, 'success' => $curlResult->isSuccess(), 'code' => $curlResult->getReturnCode()]);
 238                         return $siteinfo;
 239                 }
 240
 241                 $siteinfo['expires'] = DateTimeFormat::utc(self::DEFAULT_EXPIRATION_SUCCESS);
 242
 243                 if ($cacheControlHeader = $curlResult->getHeader('Cache-Control')[0] ?? '') {
 244                         if (preg_match('/max-age=([0-9]+)/i', $cacheControlHeader, $matches)) {
 245                                 $maxAge = max(86400, (int)array_pop($matches));
 246                                 $siteinfo['expires'] = DateTimeFormat::utc("now + $maxAge seconds");
 247                         }
 248                 }
 249
 250                 $body = $curlResult->getBody();
 251
 252                 if ($do_oembed) {
 253                         $oembed_data = OEmbed::fetchURL($url, false, false);
 254
 255                         if (!empty($oembed_data->type)) {
 256                                 if (!in_array($oembed_data->type, ['error', 'rich', 'image', 'video', 'audio', ''])) {
 257                                         $siteinfo['type'] = $oembed_data->type;
 258                                 }
 259
 260                                 // See https://github.com/friendica/friendica/pull/5763#discussion_r217913178
 261                                 if ($siteinfo['type'] != 'photo') {
 262                                         if (!empty($oembed_data->title)) {
 263                                                 $siteinfo['title'] = trim($oembed_data->title);
 264                                         }
 265                                         if (!empty($oembed_data->description)) {
 266                                                 $siteinfo['text'] = trim($oembed_data->description);
 267                                         }
 268                                         if (!empty($oembed_data->author_name)) {
 269                                                 $siteinfo['author_name'] = trim($oembed_data->author_name);
 270                                         }
 271                                         if (!empty($oembed_data->author_url)) {
 272                                                 $siteinfo['author_url'] = trim($oembed_data->author_url);
 273                                         }
 274                                         if (!empty($oembed_data->provider_name)) {
 275                                                 $siteinfo['publisher_name'] = trim($oembed_data->provider_name);
 276                                         }
 277                                         if (!empty($oembed_data->provider_url)) {
 278                                                 $siteinfo['publisher_url'] = trim($oembed_data->provider_url);
 279                                         }
 280                                         if (!empty($oembed_data->thumbnail_url)) {
 281                                                 $siteinfo['image'] = $oembed_data->thumbnail_url;
 282                                         }
 283                                 }
 284                         }
 285                 }
 286
 287                 $charset = '';
 288                 try {
 289                         // Look for a charset, first in headers
 290                         $mediaType = MediaType::fromContentType($curlResult->getContentType());
 291                         if (isset($mediaType->parameters['charset'])) {
 292                                 $charset = $mediaType->parameters['charset'];
 293                         }
 294                 } catch(\InvalidArgumentException $e) {}
 295
 296                 $siteinfo['charset'] = $charset;
 297
 298                 if ($charset && strtoupper($charset) != 'UTF-8') {
 299                         // See https://github.com/friendica/friendica/issues/5470#issuecomment-418351211
 300                         $charset = str_ireplace('latin-1', 'latin1', $charset);
 301
 302                         Logger::info('detected charset', ['charset' => $charset]);
 303                         $body = iconv($charset, 'UTF-8//TRANSLIT', $body);
 304                 }
 305
 306                 $body = mb_convert_encoding($body, 'HTML-ENTITIES', 'UTF-8');
 307
 308                 if (empty($body)) {
 309                         return $siteinfo;
 310                 }
 311
 312                 $doc = new DOMDocument();
 313                 @$doc->loadHTML($body);
 314
 315                 $siteinfo['charset'] = HTML::extractCharset($doc) ?? $siteinfo['charset'];
 316
 317                 XML::deleteNode($doc, 'style');
 318                 XML::deleteNode($doc, 'option');
 319                 XML::deleteNode($doc, 'h1');
 320                 XML::deleteNode($doc, 'h2');
 321                 XML::deleteNode($doc, 'h3');
 322                 XML::deleteNode($doc, 'h4');
 323                 XML::deleteNode($doc, 'h5');
 324                 XML::deleteNode($doc, 'h6');
 325                 XML::deleteNode($doc, 'ol');
 326                 XML::deleteNode($doc, 'ul');
 327
 328                 $xpath = new DOMXPath($doc);
 329
 330                 $list = $xpath->query('//meta[@content]');
 331                 foreach ($list as $node) {
 332                         $meta_tag = [];
 333                         if ($node->attributes->length) {
 334                                 foreach ($node->attributes as $attribute) {
 335                                         $meta_tag[$attribute->name] = $attribute->value;
 336                                 }
 337                         }
 338
 339                         if (@$meta_tag['http-equiv'] == 'refresh') {
 340                                 $path = $meta_tag['content'];
 341                                 $pathinfo = explode(';', $path);
 342                                 $content = '';
 343                                 foreach ($pathinfo as $value) {
 344                                         if (substr(strtolower($value), 0, 4) == 'url=') {
 345                                                 $content = substr($value, 4);
 346                                         }
 347                                 }
 348                                 if ($content != '') {
 349                                         $siteinfo = self::getSiteinfo($content, $do_oembed, ++$count);
 350                                         return $siteinfo;
 351                                 }
 352                         }
 353                 }
 354
 355                 $list = $xpath->query('//title');
 356                 if ($list->length > 0) {
 357                         $siteinfo['title'] = trim($list->item(0)->nodeValue);
 358                 }
 359
 360                 $list = $xpath->query('//meta[@name]');
 361                 foreach ($list as $node) {
 362                         $meta_tag = [];
 363                         if ($node->attributes->length) {
 364                                 foreach ($node->attributes as $attribute) {
 365                                         $meta_tag[$attribute->name] = $attribute->value;
 366                                 }
 367                         }
 368
 369                         if (empty($meta_tag['content'])) {
 370                                 continue;
 371                         }
 372
 373                         $meta_tag['content'] = trim(html_entity_decode($meta_tag['content'], ENT_QUOTES, 'UTF-8'));
 374
 375                         switch (strtolower($meta_tag['name'])) {
 376                                 case 'fulltitle':
 377                                         $siteinfo['title'] = trim($meta_tag['content']);
 378                                         break;
 379                                 case 'description':
 380                                         $siteinfo['text'] = trim($meta_tag['content']);
 381                                         break;
 382                                 case 'thumbnail':
 383                                         $siteinfo['image'] = $meta_tag['content'];
 384                                         break;
 385                                 case 'twitter:image':
 386                                         $siteinfo['image'] = $meta_tag['content'];
 387                                         break;
 388                                 case 'twitter:image:src':
 389                                         $siteinfo['image'] = $meta_tag['content'];
 390                                         break;
 391                                 case 'twitter:description':
 392                                         $siteinfo['text'] = trim($meta_tag['content']);
 393                                         break;
 394                                 case 'twitter:title':
 395                                         $siteinfo['title'] = trim($meta_tag['content']);
 396                                         break;
 397                                 case 'twitter:player':
 398                                         $siteinfo['player']['embed'] = trim($meta_tag['content']);
 399                                         break;
 400                                 case 'twitter:player:stream':
 401                                         $siteinfo['player']['stream'] = trim($meta_tag['content']);
 402                                         break;
 403                                 case 'twitter:player:width':
 404                                         $siteinfo['player']['width'] = intval($meta_tag['content']);
 405                                         break;
 406                                 case 'twitter:player:height':
 407                                         $siteinfo['player']['height'] = intval($meta_tag['content']);
 408                                         break;
 409                                 case 'dc.title':
 410                                         $siteinfo['title'] = trim($meta_tag['content']);
 411                                         break;
 412                                 case 'dc.description':
 413                                         $siteinfo['text'] = trim($meta_tag['content']);
 414                                         break;
 415                                 case 'dc.creator':
 416                                         $siteinfo['publisher_name'] = trim($meta_tag['content']);
 417                                         break;
 418                                 case 'keywords':
 419                                         $keywords = explode(',', $meta_tag['content']);
 420                                         break;
 421                                 case 'news_keywords':
 422                                         $keywords = explode(',', $meta_tag['content']);
 423                                         break;
 424                         }
 425                 }
 426
 427                 if (isset($keywords)) {
 428                         $siteinfo['keywords'] = [];
 429                         foreach ($keywords as $keyword) {
 430                                 if (!in_array(trim($keyword), $siteinfo['keywords'])) {
 431                                         $siteinfo['keywords'][] = trim($keyword);
 432                                 }
 433                         }
 434                 }
 435
 436                 $list = $xpath->query('//meta[@property]');
 437                 foreach ($list as $node) {
 438                         $meta_tag = [];
 439                         if ($node->attributes->length) {
 440                                 foreach ($node->attributes as $attribute) {
 441                                         $meta_tag[$attribute->name] = $attribute->value;
 442                                 }
 443                         }
 444
 445                         if (!empty($meta_tag['content'])) {
 446                                 $meta_tag['content'] = trim(html_entity_decode($meta_tag['content'], ENT_QUOTES, 'UTF-8'));
 447
 448                                 switch (strtolower($meta_tag['property'])) {
 449                                         case 'og:image':
 450                                                 $siteinfo['image'] = $meta_tag['content'];
 451                                                 break;
 452                                         case 'og:image:url':
 453                                                 $siteinfo['image'] = $meta_tag['content'];
 454                                                 break;
 455                                         case 'og:image:secure_url':
 456                                                 $siteinfo['image'] = $meta_tag['content'];
 457                                                 break;
 458                                         case 'og:title':
 459                                                 $siteinfo['title'] = trim($meta_tag['content']);
 460                                                 break;
 461                                         case 'og:description':
 462                                                 $siteinfo['text'] = trim($meta_tag['content']);
 463                                                 break;
 464                                         case 'og:site_name':
 465                                                 $siteinfo['publisher_name'] = trim($meta_tag['content']);
 466                                                 break;
 467                                         case 'og:locale':
 468                                                 $siteinfo['language'] = trim($meta_tag['content']);
 469                                                 break;
 470                                         case 'og:type':
 471                                                 $siteinfo['pagetype'] = trim($meta_tag['content']);
 472                                                 break;
 473                                         case 'twitter:description':
 474                                                 $siteinfo['text'] = trim($meta_tag['content']);
 475                                                 break;
 476                                         case 'twitter:title':
 477                                                 $siteinfo['title'] = trim($meta_tag['content']);
 478                                                 break;
 479                                         case 'twitter:image':
 480                                                 $siteinfo['image'] = $meta_tag['content'];
 481                                                 break;
 482                                 }
 483                         }
 484                 }
 485
 486                 $list = $xpath->query("//script[@type='application/ld+json']");
 487                 foreach ($list as $node) {
 488                         if (!empty($node->nodeValue)) {
 489                                 if ($jsonld = json_decode($node->nodeValue, true)) {
 490                                         $siteinfo = self::parseParts($siteinfo, $jsonld);
 491                                 }
 492                         }
 493                 }
 494
 495                 if (!empty($siteinfo['player']['stream'])) {
 496                         // Only add player data to media arrays if there is no duplicate
 497                         $content_urls = array_merge(array_column($siteinfo['audio'] ?? [], 'content'), array_column($siteinfo['video'] ?? [], 'content'));
 498                         if (!in_array($siteinfo['player']['stream'], $content_urls)) {
 499                                 $contenttype = self::getContentType($siteinfo['player']['stream']);
 500                                 if (!empty($contenttype[0]) && in_array($contenttype[0], ['audio', 'video'])) {
 501                                         $media = ['content' => $siteinfo['player']['stream']];
 502
 503                                         if (!empty($siteinfo['player']['embed'])) {
 504                                                 $media['embed'] = $siteinfo['player']['embed'];
 505                                         }
 506
 507                                         $siteinfo[$contenttype[0]][] = $media;
 508                                 }
 509                         }
 510                 }
 511
 512                 if (!empty($siteinfo['image'])) {
 513                         $siteinfo['images'] = $siteinfo['images'] ?? [];
 514                         array_unshift($siteinfo['images'], ['url' => $siteinfo['image']]);
 515                         unset($siteinfo['image']);
 516                 }
 517
 518                 $siteinfo = self::checkMedia($url, $siteinfo);
 519
 520                 if (!empty($siteinfo['text']) && mb_strlen($siteinfo['text']) > self::MAX_DESC_COUNT) {
 521                         $siteinfo['text'] = mb_substr($siteinfo['text'], 0, self::MAX_DESC_COUNT) . '…';
 522                         $pos = mb_strrpos($siteinfo['text'], '.');
 523                         if ($pos > self::MIN_DESC_COUNT) {
 524                                 $siteinfo['text'] = mb_substr($siteinfo['text'], 0, $pos + 1);
 525                         }
 526                 }
 527
 528                 Logger::info('Siteinfo fetched', ['url' => $url, 'siteinfo' => $siteinfo]);
 529
 530                 Hook::callAll('getsiteinfo', $siteinfo);
 531
 532                 ksort($siteinfo);
 533
 534                 return $siteinfo;
 535         }
 536
 537         /**
 538          * Check the attached media elements.
 539          * Fix existing data and add missing data.
 540          *
 541          * @param string $page_url
 542          * @param array $siteinfo
 543          * @return array
 544          */
 545         private static function checkMedia(string $page_url, array $siteinfo) : array
 546         {
 547                 if (!empty($siteinfo['images'])) {
 548                         array_walk($siteinfo['images'], function (&$image) use ($page_url) {
 549                                 /*
 550                                  * According to the specifications someone could place a picture
 551                                  * URL into the content field as well. But this doesn't seem to
 552                                  * happen in the wild, so we don't cover it here.
 553                                  */
 554                                 if (!empty($image['url'])) {
 555                                         $image['url'] = self::completeUrl($image['url'], $page_url);
 556                                         $photodata = Images::getInfoFromURLCached($image['url']);
 557                                         if (($photodata) && ($photodata[0] > 50) && ($photodata[1] > 50)) {
 558                                                 $image['src'] = $image['url'];
 559                                                 $image['width'] = $photodata[0];
 560                                                 $image['height'] = $photodata[1];
 561                                                 $image['contenttype'] = $photodata['mime'];
 562                                                 $image['blurhash'] = $photodata['blurhash'] ?? null;
 563                                                 unset($image['url']);
 564                                                 ksort($image);
 565                                         } else {
 566                                                 $image = [];
 567                                         }
 568                                 } else {
 569                                         $image = [];
 570                                 }
 571                         });
 572
 573                         $siteinfo['images'] = array_values(array_filter($siteinfo['images']));
 574                 }
 575
 576                 foreach (['audio', 'video'] as $element) {
 577                         if (!empty($siteinfo[$element])) {
 578                                 array_walk($siteinfo[$element], function (&$media) use ($page_url, &$siteinfo) {
 579                                         $url = '';
 580                                         $embed = '';
 581                                         $content = '';
 582                                         $contenttype = '';
 583                                         foreach (['embed', 'content', 'url'] as $field) {
 584                                                 if (!empty($media[$field])) {
 585                                                         $media[$field] = self::completeUrl($media[$field], $page_url);
 586                                                         $type = self::getContentType($media[$field]);
 587                                                         if (($type[0] ?? '') == 'text') {
 588                                                                 if ($field == 'embed') {
 589                                                                         $embed = $media[$field];
 590                                                                 } else {
 591                                                                         $url = $media[$field];
 592                                                                 }
 593                                                         } elseif (!empty($type[0])) {
 594                                                                 $content = $media[$field];
 595                                                                 $contenttype = implode('/', $type);
 596                                                         }
 597                                                 }
 598                                                 unset($media[$field]);
 599                                         }
 600
 601                                         foreach (['image', 'preview'] as $field) {
 602                                                 if (!empty($media[$field])) {
 603                                                         $media[$field] = self::completeUrl($media[$field], $page_url);
 604                                                 }
 605                                         }
 606
 607                                         if (!empty($url)) {
 608                                                 $media['url'] = $url;
 609                                         }
 610                                         if (!empty($embed)) {
 611                                                 $media['embed'] = $embed;
 612                                                 if (empty($siteinfo['player']['embed'])) {
 613                                                         $siteinfo['player']['embed'] = $embed;
 614                                                 }
 615                                         }
 616                                         if (!empty($content)) {
 617                                                 $media['src'] = $content;
 618                                         }
 619                                         if (!empty($contenttype)) {
 620                                                 $media['contenttype'] = $contenttype;
 621                                         }
 622                                         if (empty($url) && empty($content) && empty($embed)) {
 623                                                 $media = [];
 624                                         }
 625                                         ksort($media);
 626                                 });
 627
 628                                 $siteinfo[$element] = array_values(array_filter($siteinfo[$element]));
 629                         }
 630                         if (empty($siteinfo[$element])) {
 631                                 unset($siteinfo[$element]);
 632                         }
 633                 }
 634                 return $siteinfo;
 635         }
 636
 637         /**
 638          * Convert tags from CSV to an array
 639          *
 640          * @param string $string Tags
 641          *
 642          * @return array with formatted Hashtags
 643          */
 644         public static function convertTagsToArray(string $string): array
 645         {
 646                 $arr_tags = str_getcsv($string);
 647                 if (count($arr_tags)) {
 648                         // add the # sign to every tag
 649                         array_walk($arr_tags, [self::class, 'arrAddHashes']);
 650
 651                         return $arr_tags;
 652                 }
 653                 return [];
 654         }
 655
 656         /**
 657          * Add a hasht sign to a string
 658          *
 659          * This method is used as callback function
 660          *
 661          * @param string $tag The pure tag name
 662          * @param int    $k   Counter for internal use
 663          *
 664          * @return void
 665          */
 666         private static function arrAddHashes(string &$tag, int $k)
 667         {
 668                 $tag = '#' . $tag;
 669         }
 670
 671         /**
 672          * Add a scheme to an url
 673          *
 674          * The src attribute of some html elements (e.g. images)
 675          * can miss the scheme so we need to add the correct
 676          * scheme
 677          *
 678          * @param string $url    The url which possibly does have
 679          *                       a missing scheme (a link to an image)
 680          * @param string $scheme The url with a correct scheme
 681          *                       (e.g. the url from the webpage which does contain the image)
 682          *
 683          * @return string The url with a scheme
 684          */
 685         private static function completeUrl(string $url, string $scheme): string
 686         {
 687                 $urlarr = parse_url($url);
 688
 689                 // If the url does already have an scheme
 690                 // we can stop the process here
 691                 if (isset($urlarr['scheme'])) {
 692                         return $url;
 693                 }
 694
 695                 $schemearr = parse_url($scheme);
 696
 697                 $complete = $schemearr['scheme'] . '://' . $schemearr['host'];
 698
 699                 if (!empty($schemearr['port'])) {
 700                         $complete .= ':' . $schemearr['port'];
 701                 }
 702
 703                 if (!empty($urlarr['path'])) {
 704                         if (strpos($urlarr['path'], '/') !== 0) {
 705                                 $complete .= '/';
 706                         }
 707
 708                         $complete .= $urlarr['path'];
 709                 }
 710
 711                 if (!empty($urlarr['query'])) {
 712                         $complete .= '?' . $urlarr['query'];
 713                 }
 714
 715                 if (!empty($urlarr['fragment'])) {
 716                         $complete .= '#' . $urlarr['fragment'];
 717                 }
 718
 719                 return $complete;
 720         }
 721
 722         /**
 723          * Parse the Json-Ld parts of a web page
 724          *
 725          * @param array $siteinfo
 726          * @param array $jsonld
 727          *
 728          * @return array siteinfo
 729          */
 730         private static function parseParts(array $siteinfo, array $jsonld): array
 731         {
 732                 if (!empty($jsonld['@graph']) && is_array($jsonld['@graph'])) {
 733                         foreach ($jsonld['@graph'] as $part) {
 734                                 if (!empty($part) && is_array($part)) {
 735                                         $siteinfo = self::parseParts($siteinfo, $part);
 736                                 }
 737                         }
 738                 } elseif (!empty($jsonld['@type'])) {
 739                         $siteinfo = self::parseJsonLd($siteinfo, $jsonld);
 740                 } elseif (!empty($jsonld)) {
 741                         $keys = array_keys($jsonld);
 742                         $numeric_keys = true;
 743                         foreach ($keys as $key) {
 744                                 if (!is_int($key)) {
 745                                         $numeric_keys = false;
 746                                 }
 747                         }
 748                         if ($numeric_keys) {
 749                                 foreach ($jsonld as $part) {
 750                                         if (!empty($part) && is_array($part)) {
 751                                                 $siteinfo = self::parseParts($siteinfo, $part);
 752                                         }
 753                                 }
 754                         }
 755                 }
 756
 757                 array_walk_recursive($siteinfo, function (&$element) {
 758                         if (is_string($element)) {
 759                                 $element = trim(strip_tags(html_entity_decode($element, ENT_COMPAT, 'UTF-8')));
 760                         }
 761                 });
 762
 763                 return $siteinfo;
 764         }
 765
 766         /**
 767          * Improve the siteinfo with information from the provided JSON-LD information
 768          * @see https://jsonld.com/
 769          * @see https://schema.org/
 770          *
 771          * @param array $siteinfo
 772          * @param array $jsonld
 773          *
 774          * @return array siteinfo
 775          */
 776         private static function parseJsonLd(array $siteinfo, array $jsonld): array
 777         {
 778                 $type = JsonLD::fetchElement($jsonld, '@type');
 779                 if (empty($type)) {
 780                         Logger::info('Empty type', ['url' => $siteinfo['url']]);
 781                         return $siteinfo;
 782                 }
 783
 784                 // Silently ignore some types that aren't processed
 785                 if (in_array($type, ['SiteNavigationElement', 'JobPosting', 'CreativeWork', 'MusicAlbum',
 786                         'WPHeader', 'WPSideBar', 'WPFooter', 'LegalService', 'MusicRecording',
 787                         'ItemList', 'BreadcrumbList', 'Blog', 'Dataset', 'Product'])) {
 788                         return $siteinfo;
 789                 }
 790
 791                 switch ($type) {
 792                         case 'Article':
 793                         case 'AdvertiserContentArticle':
 794                         case 'NewsArticle':
 795                         case 'Report':
 796                         case 'SatiricalArticle':
 797                         case 'ScholarlyArticle':
 798                         case 'SocialMediaPosting':
 799                         case 'TechArticle':
 800                         case 'ReportageNewsArticle':
 801                         case 'SocialMediaPosting':
 802                         case 'BlogPosting':
 803                         case 'LiveBlogPosting':
 804                         case 'DiscussionForumPosting':
 805                                 return self::parseJsonLdArticle($siteinfo, $jsonld);
 806                         case 'WebPage':
 807                         case 'AboutPage':
 808                         case 'CheckoutPage':
 809                         case 'CollectionPage':
 810                         case 'ContactPage':
 811                         case 'FAQPage':
 812                         case 'ItemPage':
 813                         case 'MedicalWebPage':
 814                         case 'ProfilePage':
 815                         case 'QAPage':
 816                         case 'RealEstateListing':
 817                         case 'SearchResultsPage':
 818                         case 'MediaGallery':
 819                         case 'ImageGallery':
 820                         case 'VideoGallery':
 821                         case 'RadioEpisode':
 822                         case 'Event':
 823                                 return self::parseJsonLdWebPage($siteinfo, $jsonld);
 824                         case 'WebSite':
 825                                 return self::parseJsonLdWebSite($siteinfo, $jsonld);
 826                         case 'Organization':
 827                         case 'Airline':
 828                         case 'Consortium':
 829                         case 'Corporation':
 830                         case 'EducationalOrganization':
 831                         case 'FundingScheme':
 832                         case 'GovernmentOrganization':
 833                         case 'LibrarySystem':
 834                         case 'LocalBusiness':
 835                         case 'MedicalOrganization':
 836                         case 'NGO':
 837                         case 'NewsMediaOrganization':
 838                         case 'Project':
 839                         case 'SportsOrganization':
 840                         case 'WorkersUnion':
 841                                 return self::parseJsonLdWebOrganization($siteinfo, $jsonld);
 842                         case 'Person':
 843                         case 'Patient':
 844                         case 'PerformingGroup':
 845                         case 'DanceGroup';
 846                         case 'MusicGroup':
 847                         case 'TheaterGroup':
 848                                 return self::parseJsonLdWebPerson($siteinfo, $jsonld);
 849                         case 'AudioObject':
 850                         case 'Audio':
 851                                 return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'audio');
 852                         case 'VideoObject':
 853                                 return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'video');
 854                         case 'ImageObject':
 855                                 return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'images');
 856                         default:
 857                                 Logger::info('Unknown type', ['type' => $type, 'url' => $siteinfo['url']]);
 858                                 return $siteinfo;
 859                 }
 860         }
 861
 862         /**
 863          * Fetch author and publisher data
 864          *
 865          * @param array $siteinfo
 866          * @param array $jsonld
 867          *
 868          * @return array siteinfo
 869          */
 870         private static function parseJsonLdAuthor(array $siteinfo, array $jsonld): array
 871         {
 872                 $jsonldinfo = [];
 873
 874                 if (!empty($jsonld['publisher']) && is_array($jsonld['publisher'])) {
 875                         $content = JsonLD::fetchElement($jsonld, 'publisher', 'name');
 876                         if (!empty($content) && is_string($content)) {
 877                                 $jsonldinfo['publisher_name'] = trim($content);
 878                         }
 879
 880                         $content = JsonLD::fetchElement($jsonld, 'publisher', 'url');
 881                         if (!empty($content) && is_string($content)) {
 882                                 $jsonldinfo['publisher_url'] = trim($content);
 883                         }
 884
 885                         $brand = JsonLD::fetchElement($jsonld, 'publisher', 'brand', '@type', 'Organization');
 886                         if (!empty($brand) && is_array($brand)) {
 887                                 $content = JsonLD::fetchElement($brand, 'name');
 888                                 if (!empty($content) && is_string($content)) {
 889                                         $jsonldinfo['publisher_name'] = trim($content);
 890                                 }
 891
 892                                 $content = JsonLD::fetchElement($brand, 'url');
 893                                 if (!empty($content) && is_string($content)) {
 894                                         $jsonldinfo['publisher_url'] = trim($content);
 895                                 }
 896
 897                                 $content = JsonLD::fetchElement($brand, 'logo', 'url');
 898                                 if (!empty($content) && is_string($content)) {
 899                                         $jsonldinfo['publisher_img'] = trim($content);
 900                                 }
 901                         }
 902
 903                         $logo = JsonLD::fetchElement($jsonld, 'publisher', 'logo');
 904                         if (!empty($logo) && is_array($logo)) {
 905                                 $content = JsonLD::fetchElement($logo, 'url');
 906                                 if (!empty($content) && is_string($content)) {
 907                                         $jsonldinfo['publisher_img'] = trim($content);
 908                                 }
 909                         }
 910                 } elseif (!empty($jsonld['publisher']) && is_string($jsonld['publisher'])) {
 911                         $jsonldinfo['publisher_name'] = trim($jsonld['publisher']);
 912                 }
 913
 914                 if (!empty($jsonld['author']) && is_array($jsonld['author'])) {
 915                         $content = JsonLD::fetchElement($jsonld, 'author', 'name');
 916                         if (!empty($content) && is_string($content)) {
 917                                 $jsonldinfo['author_name'] = trim($content);
 918                         }
 919
 920                         $content = JsonLD::fetchElement($jsonld, 'author', 'sameAs');
 921                         if (!empty($content) && is_string($content)) {
 922                                 $jsonldinfo['author_url'] = trim($content);
 923                         }
 924
 925                         $content = JsonLD::fetchElement($jsonld, 'author', 'url');
 926                         if (!empty($content) && is_string($content)) {
 927                                 $jsonldinfo['author_url'] = trim($content);
 928                         }
 929
 930                         $logo = JsonLD::fetchElement($jsonld, 'author', 'logo');
 931                         if (!empty($logo) && is_array($logo)) {
 932                                 $content = JsonLD::fetchElement($logo, 'url');
 933                                 if (!empty($content) && is_string($content)) {
 934                                         $jsonldinfo['author_img'] = trim($content);
 935                                 }
 936                         }
 937                 } elseif (!empty($jsonld['author']) && is_string($jsonld['author'])) {
 938                         $jsonldinfo['author_name'] = trim($jsonld['author']);
 939                 }
 940
 941                 Logger::info('Fetched Author information', ['fetched' => $jsonldinfo]);
 942
 943                 return array_merge($siteinfo, $jsonldinfo);
 944         }
 945
 946         /**
 947          * Fetch data from the provided JSON-LD Article type
 948          * @see https://schema.org/Article
 949          *
 950          * @param array $siteinfo
 951          * @param array $jsonld
 952          *
 953          * @return array siteinfo
 954          */
 955         private static function parseJsonLdArticle(array $siteinfo, array $jsonld): array
 956         {
 957                 $jsonldinfo = [];
 958
 959                 $content = JsonLD::fetchElement($jsonld, 'headline');
 960                 if (!empty($content) && is_string($content)) {
 961                         $jsonldinfo['title'] = trim($content);
 962                 }
 963
 964                 $content = JsonLD::fetchElement($jsonld, 'alternativeHeadline');
 965                 if (!empty($content) && is_string($content) && (($jsonldinfo['title'] ?? '') != trim($content))) {
 966                         $jsonldinfo['alternative_title'] = trim($content);
 967                 }
 968
 969                 $content = JsonLD::fetchElement($jsonld, 'description');
 970                 if (!empty($content) && is_string($content)) {
 971                         $jsonldinfo['text'] = trim($content);
 972                 }
 973
 974                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
 975                 if (!empty($content)) {
 976                         $jsonldinfo['image'] = trim($content);
 977                 }
 978
 979                 $content = JsonLD::fetchElement($jsonld, 'image', 'url', '@type', 'ImageObject');
 980                 if (!empty($content) && is_string($content)) {
 981                         $jsonldinfo['image'] = trim($content);
 982                 }
 983
 984                 if (!empty($jsonld['keywords']) && !is_array($jsonld['keywords'])) {
 985                         $content = JsonLD::fetchElement($jsonld, 'keywords');
 986                         if (!empty($content)) {
 987                                 $siteinfo['keywords'] = [];
 988                                 $keywords = explode(',', $content);
 989                                 foreach ($keywords as $keyword) {
 990                                         $siteinfo['keywords'][] = trim($keyword);
 991                                 }
 992                         }
 993                 } elseif (!empty($jsonld['keywords'])) {
 994                         $content = JsonLD::fetchElementArray($jsonld, 'keywords');
 995                         if (!empty($content) && is_array($content)) {
 996                                 $jsonldinfo['keywords'] = $content;
 997                         }
 998                 }
 999
1000                 $content = JsonLD::fetchElement($jsonld, 'datePublished');
1001                 if (!empty($content) && is_string($content)) {
1002                         $jsonldinfo['published'] = DateTimeFormat::utc($content);
1003                 }
1004
1005                 $content = JsonLD::fetchElement($jsonld, 'dateModified');
1006                 if (!empty($content) && is_string($content)) {
1007                         $jsonldinfo['modified'] = DateTimeFormat::utc($content);
1008                 }
1009
1010                 $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
1011
1012                 Logger::info('Fetched article information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1013
1014                 return array_merge($siteinfo, $jsonldinfo);
1015         }
1016
1017         /**
1018          * Fetch data from the provided JSON-LD WebPage type
1019          * @see https://schema.org/WebPage
1020          *
1021          * @param array $siteinfo
1022          * @param array $jsonld
1023          *
1024          * @return array siteinfo
1025          */
1026         private static function parseJsonLdWebPage(array $siteinfo, array $jsonld): array
1027         {
1028                 $jsonldinfo = [];
1029
1030                 $content = JsonLD::fetchElement($jsonld, 'name');
1031                 if (!empty($content)) {
1032                         $jsonldinfo['title'] = trim($content);
1033                 }
1034
1035                 $content = JsonLD::fetchElement($jsonld, 'description');
1036                 if (!empty($content) && is_string($content)) {
1037                         $jsonldinfo['text'] = trim($content);
1038                 }
1039
1040                 $content = JsonLD::fetchElement($jsonld, 'image');
1041                 if (!empty($content) && is_string($content)) {
1042                         $jsonldinfo['image'] = trim($content);
1043                 }
1044
1045                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
1046                 if (!empty($content) && is_string($content)) {
1047                         $jsonldinfo['image'] = trim($content);
1048                 }
1049
1050                 $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
1051
1052                 Logger::info('Fetched WebPage information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1053
1054                 return array_merge($siteinfo, $jsonldinfo);
1055         }
1056
1057         /**
1058          * Fetch data from the provided JSON-LD WebSite type
1059          * @see https://schema.org/WebSite
1060          *
1061          * @param array $siteinfo
1062          * @param array $jsonld
1063          *
1064          * @return array siteinfo
1065          */
1066         private static function parseJsonLdWebSite(array $siteinfo, array $jsonld): array
1067         {
1068                 $jsonldinfo = [];
1069
1070                 $content = JsonLD::fetchElement($jsonld, 'name');
1071                 if (!empty($content) && is_string($content)) {
1072                         $jsonldinfo['publisher_name'] = trim($content);
1073                 }
1074
1075                 $content = JsonLD::fetchElement($jsonld, 'description');
1076                 if (!empty($content) && is_string($content)) {
1077                         $jsonldinfo['publisher_description'] = trim($content);
1078                 }
1079
1080                 $content = JsonLD::fetchElement($jsonld, 'url');
1081                 if (!empty($content) && is_string($content)) {
1082                         $jsonldinfo['publisher_url'] = trim($content);
1083                 }
1084
1085                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
1086                 if (!empty($content) && is_string($content)) {
1087                         $jsonldinfo['image'] = trim($content);
1088                 }
1089
1090                 $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
1091
1092                 Logger::info('Fetched WebSite information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1093                 return array_merge($siteinfo, $jsonldinfo);
1094         }
1095
1096         /**
1097          * Fetch data from the provided JSON-LD Organization type
1098          * @see https://schema.org/Organization
1099          *
1100          * @param array $siteinfo
1101          * @param array $jsonld
1102          *
1103          * @return array siteinfo
1104          */
1105         private static function parseJsonLdWebOrganization(array $siteinfo, array $jsonld): array
1106         {
1107                 $jsonldinfo = [];
1108
1109                 $content = JsonLD::fetchElement($jsonld, 'name');
1110                 if (!empty($content) && is_string($content)) {
1111                         $jsonldinfo['publisher_name'] = trim($content);
1112                 }
1113
1114                 $content = JsonLD::fetchElement($jsonld, 'description');
1115                 if (!empty($content) && is_string($content)) {
1116                         $jsonldinfo['publisher_description'] = trim($content);
1117                 }
1118
1119                 $content = JsonLD::fetchElement($jsonld, 'url');
1120                 if (!empty($content) && is_string($content)) {
1121                         $jsonldinfo['publisher_url'] = trim($content);
1122                 }
1123
1124                 $content = JsonLD::fetchElement($jsonld, 'logo', 'url', '@type', 'ImageObject');
1125                 if (!empty($content) && is_string($content)) {
1126                         $jsonldinfo['publisher_img'] = trim($content);
1127                 } elseif (!empty($content) && is_array($content)) {
1128                         $jsonldinfo['publisher_img'] = trim($content[0]);
1129                 }
1130
1131                 $content = JsonLD::fetchElement($jsonld, 'brand', 'name', '@type', 'Organization');
1132                 if (!empty($content) && is_string($content)) {
1133                         $jsonldinfo['publisher_name'] = trim($content);
1134                 }
1135
1136                 $content = JsonLD::fetchElement($jsonld, 'brand', 'url', '@type', 'Organization');
1137                 if (!empty($content) && is_string($content)) {
1138                         $jsonldinfo['publisher_url'] = trim($content);
1139                 }
1140
1141                 Logger::info('Fetched Organization information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1142                 return array_merge($siteinfo, $jsonldinfo);
1143         }
1144
1145         /**
1146          * Fetch data from the provided JSON-LD Person type
1147          * @see https://schema.org/Person
1148          *
1149          * @param array $siteinfo
1150          * @param array $jsonld
1151          *
1152          * @return array siteinfo
1153          */
1154         private static function parseJsonLdWebPerson(array $siteinfo, array $jsonld): array
1155         {
1156                 $jsonldinfo = [];
1157
1158                 $content = JsonLD::fetchElement($jsonld, 'name');
1159                 if (!empty($content) && is_string($content)) {
1160                         $jsonldinfo['author_name'] = trim($content);
1161                 }
1162
1163                 $content = JsonLD::fetchElement($jsonld, 'description');
1164                 if (!empty($content) && is_string($content)) {
1165                         $jsonldinfo['author_description'] = trim($content);
1166                 }
1167
1168                 $content = JsonLD::fetchElement($jsonld, 'sameAs');
1169                 if (!empty($content) && is_string($content)) {
1170                         $jsonldinfo['author_url'] = trim($content);
1171                 }
1172
1173                 $content = JsonLD::fetchElement($jsonld, 'url');
1174                 if (!empty($content) && is_string($content)) {
1175                         $jsonldinfo['author_url'] = trim($content);
1176                 }
1177
1178                 $content = JsonLD::fetchElement($jsonld, 'image', 'url', '@type', 'ImageObject');
1179                 if (!empty($content) && !is_string($content)) {
1180                         Logger::notice('Unexpected return value for the author image', ['content' => $content]);
1181                 }
1182
1183                 if (!empty($content) && is_string($content)) {
1184                         $jsonldinfo['author_img'] = trim($content);
1185                 }
1186
1187                 Logger::info('Fetched Person information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1188                 return array_merge($siteinfo, $jsonldinfo);
1189         }
1190
1191         /**
1192          * Fetch data from the provided JSON-LD MediaObject type
1193          * @see https://schema.org/MediaObject
1194          *
1195          * @param array $siteinfo
1196          * @param array $jsonld
1197          *
1198          * @return array siteinfo
1199          */
1200         private static function parseJsonLdMediaObject(array $siteinfo, array $jsonld, string $name): array
1201         {
1202                 $media = [];
1203
1204                 $content = JsonLD::fetchElement($jsonld, 'caption');
1205                 if (!empty($content) && is_string($content)) {
1206                         $media['caption'] = trim($content);
1207                 }
1208
1209                 $content = JsonLD::fetchElement($jsonld, 'url');
1210                 if (!empty($content) && is_string($content)) {
1211                         $media['url'] = trim($content);
1212                 }
1213
1214                 $content = JsonLD::fetchElement($jsonld, 'mainEntityOfPage');
1215                 if (!empty($content) && is_string($content)) {
1216                         $media['main'] = Strings::compareLink($content, $siteinfo['url']);
1217                 }
1218
1219                 $content = JsonLD::fetchElement($jsonld, 'description');
1220                 if (!empty($content) && is_string($content)) {
1221                         $media['description'] = trim($content);
1222                 }
1223
1224                 $content = JsonLD::fetchElement($jsonld, 'name');
1225                 if (!empty($content) && (($media['description'] ?? '') != trim($content))) {
1226                         $media['name'] = trim($content);
1227                 }
1228
1229                 $content = JsonLD::fetchElement($jsonld, 'contentUrl');
1230                 if (!empty($content) && is_string($content)) {
1231                         $media['content'] = trim($content);
1232                 }
1233
1234                 $content = JsonLD::fetchElement($jsonld, 'embedUrl');
1235                 if (!empty($content) && is_string($content)) {
1236                         $media['embed'] = trim($content);
1237                 }
1238
1239                 $content = JsonLD::fetchElement($jsonld, 'height');
1240                 if (!empty($content) && is_string($content)) {
1241                         $media['height'] = trim($content);
1242                 }
1243
1244                 $content = JsonLD::fetchElement($jsonld, 'width');
1245                 if (!empty($content) && is_string($content)) {
1246                         $media['width'] = trim($content);
1247                 }
1248
1249                 $content = JsonLD::fetchElement($jsonld, 'image');
1250                 if (!empty($content) && is_string($content)) {
1251                         $media['image'] = trim($content);
1252                 }
1253
1254                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
1255                 if (!empty($content) && (($media['image'] ?? '') != trim($content))) {
1256                         if (!empty($media['image'])) {
1257                                 $media['preview'] = trim($content);
1258                         } else {
1259                                 $media['image'] = trim($content);
1260                         }
1261                 }
1262
1263                 Logger::info('Fetched Media information', ['url' => $siteinfo['url'], 'fetched' => $media]);
1264                 $siteinfo[$name][] = $media;
1265                 return $siteinfo;
1266         }
1267 }