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\Core\Hook;
  28 use Friendica\Core\Logger;
  29 use Friendica\Database\Database;
  30 use Friendica\Database\DBA;
  31 use Friendica\DI;
  32 use Friendica\Network\HTTPClient\Client\HttpClientAccept;
  33 use Friendica\Network\HTTPException;
  34 use Friendica\Network\HTTPClient\Client\HttpClientOptions;
  35
  36 /**
  37  * Get information about a given URL
  38  *
  39  * Class with methods for extracting certain content from an url
  40  */
  41 class ParseUrl
  42 {
  43         const DEFAULT_EXPIRATION_FAILURE = 'now + 1 day';
  44         const DEFAULT_EXPIRATION_SUCCESS = 'now + 3 months';
  45
  46         /**
  47          * Maximum number of characters for the description
  48          */
  49         const MAX_DESC_COUNT = 250;
  50
  51         /**
  52          * Minimum number of characters for the description
  53          */
  54         const MIN_DESC_COUNT = 100;
  55
  56         /**
  57          * Fetch the content type of the given url
  58          * @param string $url    URL of the page
  59          * @param string $accept content-type to accept
  60          * @param int    $timeout
  61          * @return array content type
  62          */
  63         public static function getContentType(string $url, string $accept = HttpClientAccept::DEFAULT, int $timeout = 0): array
  64         {
  65                 if (!empty($timeout)) {
  66                         $options = [HttpClientOptions::TIMEOUT => $timeout];
  67                 } else {
  68                         $options = [];
  69                 }
  70
  71                 $curlResult = DI::httpClient()->head($url, array_merge([HttpClientOptions::ACCEPT_CONTENT => $accept], $options));
  72
  73                 // Workaround for systems that can't handle a HEAD request. Don't retry on timeouts.
  74                 if (!$curlResult->isSuccess() && ($curlResult->getReturnCode() >= 400) && !in_array($curlResult->getReturnCode(), [408, 504])) {
  75                         $curlResult = DI::httpClient()->get($url, $accept, array_merge([HttpClientOptions::CONTENT_LENGTH => 1000000], $options));
  76                 }
  77
  78                 if (!$curlResult->isSuccess()) {
  79                         Logger::debug('Got HTTP Error', ['http error' => $curlResult->getReturnCode(), 'url' => $url]);
  80                         return [];
  81                 }
  82
  83                 $contenttype =  $curlResult->getHeader('Content-Type')[0] ?? '';
  84                 if (empty($contenttype)) {
  85                         return ['application', 'octet-stream'];
  86                 }
  87
  88                 return explode('/', current(explode(';', $contenttype)));
  89         }
  90
  91         /**
  92          * Search for chached embeddable data of an url otherwise fetch it
  93          *
  94          * @param string $url         The url of the page which should be scraped
  95          * @param bool   $do_oembed   The false option is used by the function fetch_oembed()
  96          *                            to avoid endless loops
  97          *
  98          * @return array which contains needed data for embedding
  99          *    string 'url'      => The url of the parsed page
 100          *    string 'type'     => Content type
 101          *    string 'title'    => (optional) The title of the content
 102          *    string 'text'     => (optional) The description for the content
 103          *    string 'image'    => (optional) A preview image of the content
 104          *    array  'images'   => (optional) Array of preview pictures
 105          *    string 'keywords' => (optional) The tags which belong to the content
 106          *
 107          * @throws HTTPException\InternalServerErrorException
 108          * @see   ParseUrl::getSiteinfo() for more information about scraping
 109          * embeddable content
 110          */
 111         public static function getSiteinfoCached(string $url, bool $do_oembed = true): array
 112         {
 113                 if (empty($url)) {
 114                         return [
 115                                 'url' => '',
 116                                 'type' => 'error',
 117                         ];
 118                 }
 119
 120                 $urlHash = hash('sha256', $url);
 121
 122                 $parsed_url = DBA::selectFirst('parsed_url', ['content'],
 123                         ['url_hash' => $urlHash, 'oembed' => $do_oembed]
 124                 );
 125                 if (!empty($parsed_url['content'])) {
 126                         $data = unserialize($parsed_url['content']);
 127                         return $data;
 128                 }
 129
 130                 $data = self::getSiteinfo($url, $do_oembed);
 131
 132                 $expires = $data['expires'];
 133
 134                 unset($data['expires']);
 135
 136                 DI::dba()->insert(
 137                         'parsed_url',
 138                         [
 139                                 'url_hash' => $urlHash,
 140                                 'oembed'   => $do_oembed,
 141                                 'url'      => $url,
 142                                 'content'  => serialize($data),
 143                                 'created'  => DateTimeFormat::utcNow(),
 144                                 'expires'  => $expires,
 145                         ],
 146                         Database::INSERT_UPDATE
 147                 );
 148
 149                 return $data;
 150         }
 151
 152         /**
 153          * Parse a page for embeddable content information
 154          *
 155          * This method parses to url for meta data which can be used to embed
 156          * the content. If available it prioritizes Open Graph meta tags.
 157          * If this is not available it uses the twitter cards meta tags.
 158          * As fallback it uses standard html elements with meta informations
 159          * like \<title\>Awesome Title\</title\> or
 160          * \<meta name="description" content="An awesome description"\>
 161          *
 162          * @param string $url         The url of the page which should be scraped
 163          * @param bool   $do_oembed   The false option is used by the function fetch_oembed()
 164          *                            to avoid endless loops
 165          * @param int    $count       Internal counter to avoid endless loops
 166          *
 167          * @return array which contains needed data for embedding
 168          *    string 'url'      => The url of the parsed page
 169          *    string 'type'     => Content type (error, link, photo, image, audio, video)
 170          *    string 'title'    => (optional) The title of the content
 171          *    string 'text'     => (optional) The description for the content
 172          *    string 'image'    => (optional) A preview image of the content
 173          *    array  'images'   => (optional) Array of preview pictures
 174          *    string 'keywords' => (optional) The tags which belong to the content
 175          *
 176          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 177          * @todo  https://developers.google.com/+/plugins/snippet/
 178          * @verbatim
 179          * <meta itemprop="name" content="Awesome title">
 180          * <meta itemprop="description" content="An awesome description">
 181          * <meta itemprop="image" content="http://maple.libertreeproject.org/images/tree-icon.png">
 182          *
 183          * <body itemscope itemtype="http://schema.org/Product">
 184          *   <h1 itemprop="name">Shiny Trinket</h1>
 185          *   <img itemprop="image" src="{image-url}" />
 186          *   <p itemprop="description">Shiny trinkets are shiny.</p>
 187          * </body>
 188          * @endverbatim
 189          */
 190         public static function getSiteinfo(string $url, bool $do_oembed = true, int $count = 1): array
 191         {
 192                 if (empty($url)) {
 193                         return [
 194                                 'url' => '',
 195                                 'type' => 'error',
 196                         ];
 197                 }
 198
 199                 // Check if the URL does contain a scheme
 200                 $scheme = parse_url($url, PHP_URL_SCHEME);
 201
 202                 if ($scheme == '') {
 203                         $url = 'http://' . ltrim($url, '/');
 204                 }
 205
 206                 $url = trim($url, "'\"");
 207
 208                 $url = Network::stripTrackingQueryParams($url);
 209
 210                 $siteinfo = [
 211                         'url' => $url,
 212                         'type' => 'link',
 213                         'expires' => DateTimeFormat::utc(self::DEFAULT_EXPIRATION_FAILURE),
 214                 ];
 215
 216                 if ($count > 10) {
 217                         Logger::warning('Endless loop detected', ['url' => $url]);
 218                         return $siteinfo;
 219                 }
 220
 221                 $type = self::getContentType($url);
 222                 Logger::info('Got content-type', ['content-type' => $type, 'url' => $url]);
 223                 if (!empty($type) && in_array($type[0], ['image', 'video', 'audio'])) {
 224                         $siteinfo['type'] = $type[0];
 225                         return $siteinfo;
 226                 }
 227
 228                 if ((count($type) >= 2) && (($type[0] != 'text') || ($type[1] != 'html'))) {
 229                         Logger::info('Unparseable content-type, quitting here, ', ['content-type' => $type, 'url' => $url]);
 230                         return $siteinfo;
 231                 }
 232
 233                 $curlResult = DI::httpClient()->get($url, HttpClientAccept::HTML, [HttpClientOptions::CONTENT_LENGTH => 1000000]);
 234                 if (!$curlResult->isSuccess() || empty($curlResult->getBody())) {
 235                         Logger::info('Empty body or error when fetching', ['url' => $url, 'success' => $curlResult->isSuccess(), 'code' => $curlResult->getReturnCode()]);
 236                         return $siteinfo;
 237                 }
 238
 239                 $siteinfo['expires'] = DateTimeFormat::utc(self::DEFAULT_EXPIRATION_SUCCESS);
 240
 241                 if ($cacheControlHeader = $curlResult->getHeader('Cache-Control')[0] ?? '') {
 242                         if (preg_match('/max-age=([0-9]+)/i', $cacheControlHeader, $matches)) {
 243                                 $maxAge = max(86400, (int)array_pop($matches));
 244                                 $siteinfo['expires'] = DateTimeFormat::utc("now + $maxAge seconds");
 245                         }
 246                 }
 247
 248                 $body = $curlResult->getBody();
 249
 250                 if ($do_oembed) {
 251                         $oembed_data = OEmbed::fetchURL($url, false, false);
 252
 253                         if (!empty($oembed_data->type)) {
 254                                 if (!in_array($oembed_data->type, ['error', 'rich', 'image', 'video', 'audio', ''])) {
 255                                         $siteinfo['type'] = $oembed_data->type;
 256                                 }
 257
 258                                 // See https://github.com/friendica/friendica/pull/5763#discussion_r217913178
 259                                 if ($siteinfo['type'] != 'photo') {
 260                                         if (!empty($oembed_data->title)) {
 261                                                 $siteinfo['title'] = trim($oembed_data->title);
 262                                         }
 263                                         if (!empty($oembed_data->description)) {
 264                                                 $siteinfo['text'] = trim($oembed_data->description);
 265                                         }
 266                                         if (!empty($oembed_data->author_name)) {
 267                                                 $siteinfo['author_name'] = trim($oembed_data->author_name);
 268                                         }
 269                                         if (!empty($oembed_data->author_url)) {
 270                                                 $siteinfo['author_url'] = trim($oembed_data->author_url);
 271                                         }
 272                                         if (!empty($oembed_data->provider_name)) {
 273                                                 $siteinfo['publisher_name'] = trim($oembed_data->provider_name);
 274                                         }
 275                                         if (!empty($oembed_data->provider_url)) {
 276                                                 $siteinfo['publisher_url'] = trim($oembed_data->provider_url);
 277                                         }
 278                                         if (!empty($oembed_data->thumbnail_url)) {
 279                                                 $siteinfo['image'] = $oembed_data->thumbnail_url;
 280                                         }
 281                                 }
 282                         }
 283                 }
 284
 285                 $charset = '';
 286                 // Look for a charset, first in headers
 287                 // Expected form: Content-Type: text/html; charset=ISO-8859-4
 288                 if (preg_match('/charset=([a-z0-9-_.\/]+)/i', $curlResult->getContentType(), $matches)) {
 289                         $charset = trim(trim(trim(array_pop($matches)), ';,'));
 290                 } else {
 291                         // Then in body that gets precedence
 292                         // Expected forms:
 293                         // - <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
 294                         // - <meta charset="utf-8">
 295                         // - <meta charset=utf-8>
 296                         // - <meta charSet="utf-8">
 297                         // We escape <style> and <script> tags since they can contain irrelevant charset information
 298                         // (see https://github.com/friendica/friendica/issues/9251#issuecomment-698636806)
 299                         Strings::performWithEscapedBlocks($body, '#<(?:style|script).*?</(?:style|script)>#ism', function ($body) use (&$charset) {
 300                                 if (preg_match('/charset=["\']?([a-z0-9-_.\/]+)/i', $body, $matches)) {
 301                                         $charset = trim(trim(trim(array_pop($matches)), ';,'));
 302                                 }
 303                         });
 304                 }
 305
 306                 $siteinfo['charset'] = $charset;
 307
 308                 if ($charset && strtoupper($charset) != 'UTF-8') {
 309                         // See https://github.com/friendica/friendica/issues/5470#issuecomment-418351211
 310                         $charset = str_ireplace('latin-1', 'latin1', $charset);
 311
 312                         Logger::info('detected charset', ['charset' => $charset]);
 313                         $body = iconv($charset, 'UTF-8//TRANSLIT', $body);
 314                 }
 315
 316                 $body = mb_convert_encoding($body, 'HTML-ENTITIES', 'UTF-8');
 317
 318                 if (empty($body)) {
 319                         return $siteinfo;
 320                 }
 321
 322                 $doc = new DOMDocument();
 323                 @$doc->loadHTML($body);
 324
 325                 XML::deleteNode($doc, 'style');
 326                 XML::deleteNode($doc, 'option');
 327                 XML::deleteNode($doc, 'h1');
 328                 XML::deleteNode($doc, 'h2');
 329                 XML::deleteNode($doc, 'h3');
 330                 XML::deleteNode($doc, 'h4');
 331                 XML::deleteNode($doc, 'h5');
 332                 XML::deleteNode($doc, 'h6');
 333                 XML::deleteNode($doc, 'ol');
 334                 XML::deleteNode($doc, 'ul');
 335
 336                 $xpath = new DOMXPath($doc);
 337
 338                 $list = $xpath->query('//meta[@content]');
 339                 foreach ($list as $node) {
 340                         $meta_tag = [];
 341                         if ($node->attributes->length) {
 342                                 foreach ($node->attributes as $attribute) {
 343                                         $meta_tag[$attribute->name] = $attribute->value;
 344                                 }
 345                         }
 346
 347                         if (@$meta_tag['http-equiv'] == 'refresh') {
 348                                 $path = $meta_tag['content'];
 349                                 $pathinfo = explode(';', $path);
 350                                 $content = '';
 351                                 foreach ($pathinfo as $value) {
 352                                         if (substr(strtolower($value), 0, 4) == 'url=') {
 353                                                 $content = substr($value, 4);
 354                                         }
 355                                 }
 356                                 if ($content != '') {
 357                                         $siteinfo = self::getSiteinfo($content, $do_oembed, ++$count);
 358                                         return $siteinfo;
 359                                 }
 360                         }
 361                 }
 362
 363                 $list = $xpath->query('//title');
 364                 if ($list->length > 0) {
 365                         $siteinfo['title'] = trim($list->item(0)->nodeValue);
 366                 }
 367
 368                 $list = $xpath->query('//meta[@name]');
 369                 foreach ($list as $node) {
 370                         $meta_tag = [];
 371                         if ($node->attributes->length) {
 372                                 foreach ($node->attributes as $attribute) {
 373                                         $meta_tag[$attribute->name] = $attribute->value;
 374                                 }
 375                         }
 376
 377                         if (empty($meta_tag['content'])) {
 378                                 continue;
 379                         }
 380
 381                         $meta_tag['content'] = trim(html_entity_decode($meta_tag['content'], ENT_QUOTES, 'UTF-8'));
 382
 383                         switch (strtolower($meta_tag['name'])) {
 384                                 case 'fulltitle':
 385                                         $siteinfo['title'] = trim($meta_tag['content']);
 386                                         break;
 387                                 case 'description':
 388                                         $siteinfo['text'] = trim($meta_tag['content']);
 389                                         break;
 390                                 case 'thumbnail':
 391                                         $siteinfo['image'] = $meta_tag['content'];
 392                                         break;
 393                                 case 'twitter:image':
 394                                         $siteinfo['image'] = $meta_tag['content'];
 395                                         break;
 396                                 case 'twitter:image:src':
 397                                         $siteinfo['image'] = $meta_tag['content'];
 398                                         break;
 399                                 case 'twitter:description':
 400                                         $siteinfo['text'] = trim($meta_tag['content']);
 401                                         break;
 402                                 case 'twitter:title':
 403                                         $siteinfo['title'] = trim($meta_tag['content']);
 404                                         break;
 405                                 case 'twitter:player':
 406                                         $siteinfo['player']['embed'] = trim($meta_tag['content']);
 407                                         break;
 408                                 case 'twitter:player:stream':
 409                                         $siteinfo['player']['stream'] = trim($meta_tag['content']);
 410                                         break;
 411                                 case 'twitter:player:width':
 412                                         $siteinfo['player']['width'] = intval($meta_tag['content']);
 413                                         break;
 414                                 case 'twitter:player:height':
 415                                         $siteinfo['player']['height'] = intval($meta_tag['content']);
 416                                         break;
 417                                 case 'dc.title':
 418                                         $siteinfo['title'] = trim($meta_tag['content']);
 419                                         break;
 420                                 case 'dc.description':
 421                                         $siteinfo['text'] = trim($meta_tag['content']);
 422                                         break;
 423                                 case 'dc.creator':
 424                                         $siteinfo['publisher_name'] = trim($meta_tag['content']);
 425                                         break;
 426                                 case 'keywords':
 427                                         $keywords = explode(',', $meta_tag['content']);
 428                                         break;
 429                                 case 'news_keywords':
 430                                         $keywords = explode(',', $meta_tag['content']);
 431                                         break;
 432                         }
 433                 }
 434
 435                 if (isset($keywords)) {
 436                         $siteinfo['keywords'] = [];
 437                         foreach ($keywords as $keyword) {
 438                                 if (!in_array(trim($keyword), $siteinfo['keywords'])) {
 439                                         $siteinfo['keywords'][] = trim($keyword);
 440                                 }
 441                         }
 442                 }
 443
 444                 $list = $xpath->query('//meta[@property]');
 445                 foreach ($list as $node) {
 446                         $meta_tag = [];
 447                         if ($node->attributes->length) {
 448                                 foreach ($node->attributes as $attribute) {
 449                                         $meta_tag[$attribute->name] = $attribute->value;
 450                                 }
 451                         }
 452
 453                         if (!empty($meta_tag['content'])) {
 454                                 $meta_tag['content'] = trim(html_entity_decode($meta_tag['content'], ENT_QUOTES, 'UTF-8'));
 455
 456                                 switch (strtolower($meta_tag['property'])) {
 457                                         case 'og:image':
 458                                                 $siteinfo['image'] = $meta_tag['content'];
 459                                                 break;
 460                                         case 'og:image:url':
 461                                                 $siteinfo['image'] = $meta_tag['content'];
 462                                                 break;
 463                                         case 'og:image:secure_url':
 464                                                 $siteinfo['image'] = $meta_tag['content'];
 465                                                 break;
 466                                         case 'og:title':
 467                                                 $siteinfo['title'] = trim($meta_tag['content']);
 468                                                 break;
 469                                         case 'og:description':
 470                                                 $siteinfo['text'] = trim($meta_tag['content']);
 471                                                 break;
 472                                         case 'og:site_name':
 473                                                 $siteinfo['publisher_name'] = trim($meta_tag['content']);
 474                                                 break;
 475                                         case 'og:locale':
 476                                                 $siteinfo['language'] = trim($meta_tag['content']);
 477                                                 break;
 478                                         case 'og:type':
 479                                                 $siteinfo['pagetype'] = trim($meta_tag['content']);
 480                                                 break;
 481                                         case 'twitter:description':
 482                                                 $siteinfo['text'] = trim($meta_tag['content']);
 483                                                 break;
 484                                         case 'twitter:title':
 485                                                 $siteinfo['title'] = trim($meta_tag['content']);
 486                                                 break;
 487                                         case 'twitter:image':
 488                                                 $siteinfo['image'] = $meta_tag['content'];
 489                                                 break;
 490                                 }
 491                         }
 492                 }
 493
 494                 $list = $xpath->query("//script[@type='application/ld+json']");
 495                 foreach ($list as $node) {
 496                         if (!empty($node->nodeValue)) {
 497                                 if ($jsonld = json_decode($node->nodeValue, true)) {
 498                                         $siteinfo = self::parseParts($siteinfo, $jsonld);
 499                                 }
 500                         }
 501                 }
 502
 503                 if (!empty($siteinfo['player']['stream'])) {
 504                         // Only add player data to media arrays if there is no duplicate
 505                         $content_urls = array_merge(array_column($siteinfo['audio'] ?? [], 'content'), array_column($siteinfo['video'] ?? [], 'content'));
 506                         if (!in_array($siteinfo['player']['stream'], $content_urls)) {
 507                                 $contenttype = self::getContentType($siteinfo['player']['stream']);
 508                                 if (!empty($contenttype[0]) && in_array($contenttype[0], ['audio', 'video'])) {
 509                                         $media = ['content' => $siteinfo['player']['stream']];
 510
 511                                         if (!empty($siteinfo['player']['embed'])) {
 512                                                 $media['embed'] = $siteinfo['player']['embed'];
 513                                         }
 514
 515                                         $siteinfo[$contenttype[0]][] = $media;
 516                                 }
 517                         }
 518                 }
 519
 520                 if (!empty($siteinfo['image'])) {
 521                         $siteinfo['images'] = $siteinfo['images'] ?? [];
 522                         array_unshift($siteinfo['images'], ['url' => $siteinfo['image']]);
 523                         unset($siteinfo['image']);
 524                 }
 525
 526                 $siteinfo = self::checkMedia($url, $siteinfo);
 527
 528                 if (!empty($siteinfo['text']) && mb_strlen($siteinfo['text']) > self::MAX_DESC_COUNT) {
 529                         $siteinfo['text'] = mb_substr($siteinfo['text'], 0, self::MAX_DESC_COUNT) . '…';
 530                         $pos = mb_strrpos($siteinfo['text'], '.');
 531                         if ($pos > self::MIN_DESC_COUNT) {
 532                                 $siteinfo['text'] = mb_substr($siteinfo['text'], 0, $pos + 1);
 533                         }
 534                 }
 535
 536                 Logger::info('Siteinfo fetched', ['url' => $url, 'siteinfo' => $siteinfo]);
 537
 538                 Hook::callAll('getsiteinfo', $siteinfo);
 539
 540                 ksort($siteinfo);
 541
 542                 return $siteinfo;
 543         }
 544
 545         /**
 546          * Check the attached media elements.
 547          * Fix existing data and add missing data.
 548          *
 549          * @param string $page_url
 550          * @param array $siteinfo
 551          * @return array
 552          */
 553         private static function checkMedia(string $page_url, array $siteinfo) : array
 554         {
 555                 if (!empty($siteinfo['images'])) {
 556                         array_walk($siteinfo['images'], function (&$image) use ($page_url) {
 557                                 /*
 558                                  * According to the specifications someone could place a picture
 559                                  * URL into the content field as well. But this doesn't seem to
 560                                  * happen in the wild, so we don't cover it here.
 561                                  */
 562                                 if (!empty($image['url'])) {
 563                                         $image['url'] = self::completeUrl($image['url'], $page_url);
 564                                         $photodata = Images::getInfoFromURLCached($image['url']);
 565                                         if (($photodata) && ($photodata[0] > 50) && ($photodata[1] > 50)) {
 566                                                 $image['src'] = $image['url'];
 567                                                 $image['width'] = $photodata[0];
 568                                                 $image['height'] = $photodata[1];
 569                                                 $image['contenttype'] = $photodata['mime'];
 570                                                 $image['blurhash'] = $photodata['blurhash'] ?? null;
 571                                                 unset($image['url']);
 572                                                 ksort($image);
 573                                         } else {
 574                                                 $image = [];
 575                                         }
 576                                 } else {
 577                                         $image = [];
 578                                 }
 579                         });
 580
 581                         $siteinfo['images'] = array_values(array_filter($siteinfo['images']));
 582                 }
 583
 584                 foreach (['audio', 'video'] as $element) {
 585                         if (!empty($siteinfo[$element])) {
 586                                 array_walk($siteinfo[$element], function (&$media) use ($page_url, &$siteinfo) {
 587                                         $url = '';
 588                                         $embed = '';
 589                                         $content = '';
 590                                         $contenttype = '';
 591                                         foreach (['embed', 'content', 'url'] as $field) {
 592                                                 if (!empty($media[$field])) {
 593                                                         $media[$field] = self::completeUrl($media[$field], $page_url);
 594                                                         $type = self::getContentType($media[$field]);
 595                                                         if (($type[0] ?? '') == 'text') {
 596                                                                 if ($field == 'embed') {
 597                                                                         $embed = $media[$field];
 598                                                                 } else {
 599                                                                         $url = $media[$field];
 600                                                                 }
 601                                                         } elseif (!empty($type[0])) {
 602                                                                 $content = $media[$field];
 603                                                                 $contenttype = implode('/', $type);
 604                                                         }
 605                                                 }
 606                                                 unset($media[$field]);
 607                                         }
 608
 609                                         foreach (['image', 'preview'] as $field) {
 610                                                 if (!empty($media[$field])) {
 611                                                         $media[$field] = self::completeUrl($media[$field], $page_url);
 612                                                 }
 613                                         }
 614
 615                                         if (!empty($url)) {
 616                                                 $media['url'] = $url;
 617                                         }
 618                                         if (!empty($embed)) {
 619                                                 $media['embed'] = $embed;
 620                                                 if (empty($siteinfo['player']['embed'])) {
 621                                                         $siteinfo['player']['embed'] = $embed;
 622                                                 }
 623                                         }
 624                                         if (!empty($content)) {
 625                                                 $media['src'] = $content;
 626                                         }
 627                                         if (!empty($contenttype)) {
 628                                                 $media['contenttype'] = $contenttype;
 629                                         }
 630                                         if (empty($url) && empty($content) && empty($embed)) {
 631                                                 $media = [];
 632                                         }
 633                                         ksort($media);
 634                                 });
 635
 636                                 $siteinfo[$element] = array_values(array_filter($siteinfo[$element]));
 637                         }
 638                         if (empty($siteinfo[$element])) {
 639                                 unset($siteinfo[$element]);
 640                         }
 641                 }
 642                 return $siteinfo;
 643         }
 644
 645         /**
 646          * Convert tags from CSV to an array
 647          *
 648          * @param string $string Tags
 649          *
 650          * @return array with formatted Hashtags
 651          */
 652         public static function convertTagsToArray(string $string): array
 653         {
 654                 $arr_tags = str_getcsv($string);
 655                 if (count($arr_tags)) {
 656                         // add the # sign to every tag
 657                         array_walk($arr_tags, ['self', 'arrAddHashes']);
 658
 659                         return $arr_tags;
 660                 }
 661                 return [];
 662         }
 663
 664         /**
 665          * Add a hasht sign to a string
 666          *
 667          * This method is used as callback function
 668          *
 669          * @param string $tag The pure tag name
 670          * @param int    $k   Counter for internal use
 671          *
 672          * @return void
 673          */
 674         private static function arrAddHashes(string &$tag, int $k)
 675         {
 676                 $tag = '#' . $tag;
 677         }
 678
 679         /**
 680          * Add a scheme to an url
 681          *
 682          * The src attribute of some html elements (e.g. images)
 683          * can miss the scheme so we need to add the correct
 684          * scheme
 685          *
 686          * @param string $url    The url which possibly does have
 687          *                       a missing scheme (a link to an image)
 688          * @param string $scheme The url with a correct scheme
 689          *                       (e.g. the url from the webpage which does contain the image)
 690          *
 691          * @return string The url with a scheme
 692          */
 693         private static function completeUrl(string $url, string $scheme): string
 694         {
 695                 $urlarr = parse_url($url);
 696
 697                 // If the url does allready have an scheme
 698                 // we can stop the process here
 699                 if (isset($urlarr['scheme'])) {
 700                         return $url;
 701                 }
 702
 703                 $schemearr = parse_url($scheme);
 704
 705                 $complete = $schemearr['scheme'] . '://' . $schemearr['host'];
 706
 707                 if (!empty($schemearr['port'])) {
 708                         $complete .= ':' . $schemearr['port'];
 709                 }
 710
 711                 if (!empty($urlarr['path'])) {
 712                         if (strpos($urlarr['path'], '/') !== 0) {
 713                                 $complete .= '/';
 714                         }
 715
 716                         $complete .= $urlarr['path'];
 717                 }
 718
 719                 if (!empty($urlarr['query'])) {
 720                         $complete .= '?' . $urlarr['query'];
 721                 }
 722
 723                 if (!empty($urlarr['fragment'])) {
 724                         $complete .= '#' . $urlarr['fragment'];
 725                 }
 726
 727                 return $complete;
 728         }
 729
 730         /**
 731          * Parse the Json-Ld parts of a web page
 732          *
 733          * @param array $siteinfo
 734          * @param array $jsonld
 735          *
 736          * @return array siteinfo
 737          */
 738         private static function parseParts(array $siteinfo, array $jsonld): array
 739         {
 740                 if (!empty($jsonld['@graph']) && is_array($jsonld['@graph'])) {
 741                         foreach ($jsonld['@graph'] as $part) {
 742                                 if (!empty($part) && is_array($part)) {
 743                                         $siteinfo = self::parseParts($siteinfo, $part);
 744                                 }
 745                         }
 746                 } elseif (!empty($jsonld['@type'])) {
 747                         $siteinfo = self::parseJsonLd($siteinfo, $jsonld);
 748                 } elseif (!empty($jsonld)) {
 749                         $keys = array_keys($jsonld);
 750                         $numeric_keys = true;
 751                         foreach ($keys as $key) {
 752                                 if (!is_int($key)) {
 753                                         $numeric_keys = false;
 754                                 }
 755                         }
 756                         if ($numeric_keys) {
 757                                 foreach ($jsonld as $part) {
 758                                         if (!empty($part) && is_array($part)) {
 759                                                 $siteinfo = self::parseParts($siteinfo, $part);
 760                                         }
 761                                 }
 762                         }
 763                 }
 764
 765                 array_walk_recursive($siteinfo, function (&$element) {
 766                         if (is_string($element)) {
 767                                 $element = trim(strip_tags(html_entity_decode($element, ENT_COMPAT, 'UTF-8')));
 768                         }
 769                 });
 770
 771                 return $siteinfo;
 772         }
 773
 774         /**
 775          * Improve the siteinfo with information from the provided JSON-LD information
 776          * @see https://jsonld.com/
 777          * @see https://schema.org/
 778          *
 779          * @param array $siteinfo
 780          * @param array $jsonld
 781          *
 782          * @return array siteinfo
 783          */
 784         private static function parseJsonLd(array $siteinfo, array $jsonld): array
 785         {
 786                 $type = JsonLD::fetchElement($jsonld, '@type');
 787                 if (empty($type)) {
 788                         Logger::info('Empty type', ['url' => $siteinfo['url']]);
 789                         return $siteinfo;
 790                 }
 791
 792                 // Silently ignore some types that aren't processed
 793                 if (in_array($type, ['SiteNavigationElement', 'JobPosting', 'CreativeWork', 'MusicAlbum',
 794                         'WPHeader', 'WPSideBar', 'WPFooter', 'LegalService', 'MusicRecording',
 795                         'ItemList', 'BreadcrumbList', 'Blog', 'Dataset', 'Product'])) {
 796                         return $siteinfo;
 797                 }
 798
 799                 switch ($type) {
 800                         case 'Article':
 801                         case 'AdvertiserContentArticle':
 802                         case 'NewsArticle':
 803                         case 'Report':
 804                         case 'SatiricalArticle':
 805                         case 'ScholarlyArticle':
 806                         case 'SocialMediaPosting':
 807                         case 'TechArticle':
 808                         case 'ReportageNewsArticle':
 809                         case 'SocialMediaPosting':
 810                         case 'BlogPosting':
 811                         case 'LiveBlogPosting':
 812                         case 'DiscussionForumPosting':
 813                                 return self::parseJsonLdArticle($siteinfo, $jsonld);
 814                         case 'WebPage':
 815                         case 'AboutPage':
 816                         case 'CheckoutPage':
 817                         case 'CollectionPage':
 818                         case 'ContactPage':
 819                         case 'FAQPage':
 820                         case 'ItemPage':
 821                         case 'MedicalWebPage':
 822                         case 'ProfilePage':
 823                         case 'QAPage':
 824                         case 'RealEstateListing':
 825                         case 'SearchResultsPage':
 826                         case 'MediaGallery':
 827                         case 'ImageGallery':
 828                         case 'VideoGallery':
 829                         case 'RadioEpisode':
 830                         case 'Event':
 831                                 return self::parseJsonLdWebPage($siteinfo, $jsonld);
 832                         case 'WebSite':
 833                                 return self::parseJsonLdWebSite($siteinfo, $jsonld);
 834                         case 'Organization':
 835                         case 'Airline':
 836                         case 'Consortium':
 837                         case 'Corporation':
 838                         case 'EducationalOrganization':
 839                         case 'FundingScheme':
 840                         case 'GovernmentOrganization':
 841                         case 'LibrarySystem':
 842                         case 'LocalBusiness':
 843                         case 'MedicalOrganization':
 844                         case 'NGO':
 845                         case 'NewsMediaOrganization':
 846                         case 'Project':
 847                         case 'SportsOrganization':
 848                         case 'WorkersUnion':
 849                                 return self::parseJsonLdWebOrganization($siteinfo, $jsonld);
 850                         case 'Person':
 851                         case 'Patient':
 852                         case 'PerformingGroup':
 853                         case 'DanceGroup';
 854                         case 'MusicGroup':
 855                         case 'TheaterGroup':
 856                                 return self::parseJsonLdWebPerson($siteinfo, $jsonld);
 857                         case 'AudioObject':
 858                         case 'Audio':
 859                                 return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'audio');
 860                         case 'VideoObject':
 861                                 return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'video');
 862                         case 'ImageObject':
 863                                 return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'images');
 864                         default:
 865                                 Logger::info('Unknown type', ['type' => $type, 'url' => $siteinfo['url']]);
 866                                 return $siteinfo;
 867                 }
 868         }
 869
 870         /**
 871          * Fetch author and publisher data
 872          *
 873          * @param array $siteinfo
 874          * @param array $jsonld
 875          *
 876          * @return array siteinfo
 877          */
 878         private static function parseJsonLdAuthor(array $siteinfo, array $jsonld): array
 879         {
 880                 $jsonldinfo = [];
 881
 882                 if (!empty($jsonld['publisher']) && is_array($jsonld['publisher'])) {
 883                         $content = JsonLD::fetchElement($jsonld, 'publisher', 'name');
 884                         if (!empty($content) && is_string($content)) {
 885                                 $jsonldinfo['publisher_name'] = trim($content);
 886                         }
 887
 888                         $content = JsonLD::fetchElement($jsonld, 'publisher', 'url');
 889                         if (!empty($content) && is_string($content)) {
 890                                 $jsonldinfo['publisher_url'] = trim($content);
 891                         }
 892
 893                         $brand = JsonLD::fetchElement($jsonld, 'publisher', 'brand', '@type', 'Organization');
 894                         if (!empty($brand) && is_array($brand)) {
 895                                 $content = JsonLD::fetchElement($brand, 'name');
 896                                 if (!empty($content) && is_string($content)) {
 897                                         $jsonldinfo['publisher_name'] = trim($content);
 898                                 }
 899
 900                                 $content = JsonLD::fetchElement($brand, 'url');
 901                                 if (!empty($content) && is_string($content)) {
 902                                         $jsonldinfo['publisher_url'] = trim($content);
 903                                 }
 904
 905                                 $content = JsonLD::fetchElement($brand, 'logo', 'url');
 906                                 if (!empty($content) && is_string($content)) {
 907                                         $jsonldinfo['publisher_img'] = trim($content);
 908                                 }
 909                         }
 910
 911                         $logo = JsonLD::fetchElement($jsonld, 'publisher', 'logo');
 912                         if (!empty($logo) && is_array($logo)) {
 913                                 $content = JsonLD::fetchElement($logo, 'url');
 914                                 if (!empty($content) && is_string($content)) {
 915                                         $jsonldinfo['publisher_img'] = trim($content);
 916                                 }
 917                         }
 918                 } elseif (!empty($jsonld['publisher']) && is_string($jsonld['publisher'])) {
 919                         $jsonldinfo['publisher_name'] = trim($jsonld['publisher']);
 920                 }
 921
 922                 if (!empty($jsonld['author']) && is_array($jsonld['author'])) {
 923                         $content = JsonLD::fetchElement($jsonld, 'author', 'name');
 924                         if (!empty($content) && is_string($content)) {
 925                                 $jsonldinfo['author_name'] = trim($content);
 926                         }
 927
 928                         $content = JsonLD::fetchElement($jsonld, 'author', 'sameAs');
 929                         if (!empty($content) && is_string($content)) {
 930                                 $jsonldinfo['author_url'] = trim($content);
 931                         }
 932
 933                         $content = JsonLD::fetchElement($jsonld, 'author', 'url');
 934                         if (!empty($content) && is_string($content)) {
 935                                 $jsonldinfo['author_url'] = trim($content);
 936                         }
 937
 938                         $logo = JsonLD::fetchElement($jsonld, 'author', 'logo');
 939                         if (!empty($logo) && is_array($logo)) {
 940                                 $content = JsonLD::fetchElement($logo, 'url');
 941                                 if (!empty($content) && is_string($content)) {
 942                                         $jsonldinfo['author_img'] = trim($content);
 943                                 }
 944                         }
 945                 } elseif (!empty($jsonld['author']) && is_string($jsonld['author'])) {
 946                         $jsonldinfo['author_name'] = trim($jsonld['author']);
 947                 }
 948
 949                 Logger::info('Fetched Author information', ['fetched' => $jsonldinfo]);
 950
 951                 return array_merge($siteinfo, $jsonldinfo);
 952         }
 953
 954         /**
 955          * Fetch data from the provided JSON-LD Article type
 956          * @see https://schema.org/Article
 957          *
 958          * @param array $siteinfo
 959          * @param array $jsonld
 960          *
 961          * @return array siteinfo
 962          */
 963         private static function parseJsonLdArticle(array $siteinfo, array $jsonld): array
 964         {
 965                 $jsonldinfo = [];
 966
 967                 $content = JsonLD::fetchElement($jsonld, 'headline');
 968                 if (!empty($content) && is_string($content)) {
 969                         $jsonldinfo['title'] = trim($content);
 970                 }
 971
 972                 $content = JsonLD::fetchElement($jsonld, 'alternativeHeadline');
 973                 if (!empty($content) && is_string($content) && (($jsonldinfo['title'] ?? '') != trim($content))) {
 974                         $jsonldinfo['alternative_title'] = trim($content);
 975                 }
 976
 977                 $content = JsonLD::fetchElement($jsonld, 'description');
 978                 if (!empty($content) && is_string($content)) {
 979                         $jsonldinfo['text'] = trim($content);
 980                 }
 981
 982                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
 983                 if (!empty($content)) {
 984                         $jsonldinfo['image'] = trim($content);
 985                 }
 986
 987                 $content = JsonLD::fetchElement($jsonld, 'image', 'url', '@type', 'ImageObject');
 988                 if (!empty($content) && is_string($content)) {
 989                         $jsonldinfo['image'] = trim($content);
 990                 }
 991
 992                 if (!empty($jsonld['keywords']) && !is_array($jsonld['keywords'])) {
 993                         $content = JsonLD::fetchElement($jsonld, 'keywords');
 994                         if (!empty($content)) {
 995                                 $siteinfo['keywords'] = [];
 996                                 $keywords = explode(',', $content);
 997                                 foreach ($keywords as $keyword) {
 998                                         $siteinfo['keywords'][] = trim($keyword);
 999                                 }
1000                         }
1001                 } elseif (!empty($jsonld['keywords'])) {
1002                         $content = JsonLD::fetchElementArray($jsonld, 'keywords');
1003                         if (!empty($content) && is_array($content)) {
1004                                 $jsonldinfo['keywords'] = $content;
1005                         }
1006                 }
1007
1008                 $content = JsonLD::fetchElement($jsonld, 'datePublished');
1009                 if (!empty($content) && is_string($content)) {
1010                         $jsonldinfo['published'] = DateTimeFormat::utc($content);
1011                 }
1012
1013                 $content = JsonLD::fetchElement($jsonld, 'dateModified');
1014                 if (!empty($content) && is_string($content)) {
1015                         $jsonldinfo['modified'] = DateTimeFormat::utc($content);
1016                 }
1017
1018                 $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
1019
1020                 Logger::info('Fetched article information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1021
1022                 return array_merge($siteinfo, $jsonldinfo);
1023         }
1024
1025         /**
1026          * Fetch data from the provided JSON-LD WebPage type
1027          * @see https://schema.org/WebPage
1028          *
1029          * @param array $siteinfo
1030          * @param array $jsonld
1031          *
1032          * @return array siteinfo
1033          */
1034         private static function parseJsonLdWebPage(array $siteinfo, array $jsonld): array
1035         {
1036                 $jsonldinfo = [];
1037
1038                 $content = JsonLD::fetchElement($jsonld, 'name');
1039                 if (!empty($content)) {
1040                         $jsonldinfo['title'] = trim($content);
1041                 }
1042
1043                 $content = JsonLD::fetchElement($jsonld, 'description');
1044                 if (!empty($content) && is_string($content)) {
1045                         $jsonldinfo['text'] = trim($content);
1046                 }
1047
1048                 $content = JsonLD::fetchElement($jsonld, 'image');
1049                 if (!empty($content) && is_string($content)) {
1050                         $jsonldinfo['image'] = trim($content);
1051                 }
1052
1053                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
1054                 if (!empty($content) && is_string($content)) {
1055                         $jsonldinfo['image'] = trim($content);
1056                 }
1057
1058                 $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
1059
1060                 Logger::info('Fetched WebPage information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1061
1062                 return array_merge($siteinfo, $jsonldinfo);
1063         }
1064
1065         /**
1066          * Fetch data from the provided JSON-LD WebSite type
1067          * @see https://schema.org/WebSite
1068          *
1069          * @param array $siteinfo
1070          * @param array $jsonld
1071          *
1072          * @return array siteinfo
1073          */
1074         private static function parseJsonLdWebSite(array $siteinfo, array $jsonld): array
1075         {
1076                 $jsonldinfo = [];
1077
1078                 $content = JsonLD::fetchElement($jsonld, 'name');
1079                 if (!empty($content) && is_string($content)) {
1080                         $jsonldinfo['publisher_name'] = trim($content);
1081                 }
1082
1083                 $content = JsonLD::fetchElement($jsonld, 'description');
1084                 if (!empty($content) && is_string($content)) {
1085                         $jsonldinfo['publisher_description'] = trim($content);
1086                 }
1087
1088                 $content = JsonLD::fetchElement($jsonld, 'url');
1089                 if (!empty($content) && is_string($content)) {
1090                         $jsonldinfo['publisher_url'] = trim($content);
1091                 }
1092
1093                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
1094                 if (!empty($content) && is_string($content)) {
1095                         $jsonldinfo['image'] = trim($content);
1096                 }
1097
1098                 $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
1099
1100                 Logger::info('Fetched WebSite information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1101                 return array_merge($siteinfo, $jsonldinfo);
1102         }
1103
1104         /**
1105          * Fetch data from the provided JSON-LD Organization type
1106          * @see https://schema.org/Organization
1107          *
1108          * @param array $siteinfo
1109          * @param array $jsonld
1110          *
1111          * @return array siteinfo
1112          */
1113         private static function parseJsonLdWebOrganization(array $siteinfo, array $jsonld): array
1114         {
1115                 $jsonldinfo = [];
1116
1117                 $content = JsonLD::fetchElement($jsonld, 'name');
1118                 if (!empty($content) && is_string($content)) {
1119                         $jsonldinfo['publisher_name'] = trim($content);
1120                 }
1121
1122                 $content = JsonLD::fetchElement($jsonld, 'description');
1123                 if (!empty($content) && is_string($content)) {
1124                         $jsonldinfo['publisher_description'] = trim($content);
1125                 }
1126
1127                 $content = JsonLD::fetchElement($jsonld, 'url');
1128                 if (!empty($content) && is_string($content)) {
1129                         $jsonldinfo['publisher_url'] = trim($content);
1130                 }
1131
1132                 $content = JsonLD::fetchElement($jsonld, 'logo', 'url', '@type', 'ImageObject');
1133                 if (!empty($content) && is_string($content)) {
1134                         $jsonldinfo['publisher_img'] = trim($content);
1135                 } elseif (!empty($content) && is_array($content)) {
1136                         $jsonldinfo['publisher_img'] = trim($content[0]);
1137                 }
1138
1139                 $content = JsonLD::fetchElement($jsonld, 'brand', 'name', '@type', 'Organization');
1140                 if (!empty($content) && is_string($content)) {
1141                         $jsonldinfo['publisher_name'] = trim($content);
1142                 }
1143
1144                 $content = JsonLD::fetchElement($jsonld, 'brand', 'url', '@type', 'Organization');
1145                 if (!empty($content) && is_string($content)) {
1146                         $jsonldinfo['publisher_url'] = trim($content);
1147                 }
1148
1149                 Logger::info('Fetched Organization information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1150                 return array_merge($siteinfo, $jsonldinfo);
1151         }
1152
1153         /**
1154          * Fetch data from the provided JSON-LD Person type
1155          * @see https://schema.org/Person
1156          *
1157          * @param array $siteinfo
1158          * @param array $jsonld
1159          *
1160          * @return array siteinfo
1161          */
1162         private static function parseJsonLdWebPerson(array $siteinfo, array $jsonld): array
1163         {
1164                 $jsonldinfo = [];
1165
1166                 $content = JsonLD::fetchElement($jsonld, 'name');
1167                 if (!empty($content) && is_string($content)) {
1168                         $jsonldinfo['author_name'] = trim($content);
1169                 }
1170
1171                 $content = JsonLD::fetchElement($jsonld, 'description');
1172                 if (!empty($content) && is_string($content)) {
1173                         $jsonldinfo['author_description'] = trim($content);
1174                 }
1175
1176                 $content = JsonLD::fetchElement($jsonld, 'sameAs');
1177                 if (!empty($content) && is_string($content)) {
1178                         $jsonldinfo['author_url'] = trim($content);
1179                 }
1180
1181                 $content = JsonLD::fetchElement($jsonld, 'url');
1182                 if (!empty($content) && is_string($content)) {
1183                         $jsonldinfo['author_url'] = trim($content);
1184                 }
1185
1186                 $content = JsonLD::fetchElement($jsonld, 'image', 'url', '@type', 'ImageObject');
1187                 if (!empty($content) && !is_string($content)) {
1188                         Logger::notice('Unexpected return value for the author image', ['content' => $content]);
1189                 }
1190
1191                 if (!empty($content) && is_string($content)) {
1192                         $jsonldinfo['author_img'] = trim($content);
1193                 }
1194
1195                 Logger::info('Fetched Person information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
1196                 return array_merge($siteinfo, $jsonldinfo);
1197         }
1198
1199         /**
1200          * Fetch data from the provided JSON-LD MediaObject type
1201          * @see https://schema.org/MediaObject
1202          *
1203          * @param array $siteinfo
1204          * @param array $jsonld
1205          *
1206          * @return array siteinfo
1207          */
1208         private static function parseJsonLdMediaObject(array $siteinfo, array $jsonld, string $name): array
1209         {
1210                 $media = [];
1211
1212                 $content = JsonLD::fetchElement($jsonld, 'caption');
1213                 if (!empty($content) && is_string($content)) {
1214                         $media['caption'] = trim($content);
1215                 }
1216
1217                 $content = JsonLD::fetchElement($jsonld, 'url');
1218                 if (!empty($content) && is_string($content)) {
1219                         $media['url'] = trim($content);
1220                 }
1221
1222                 $content = JsonLD::fetchElement($jsonld, 'mainEntityOfPage');
1223                 if (!empty($content) && is_string($content)) {
1224                         $media['main'] = Strings::compareLink($content, $siteinfo['url']);
1225                 }
1226
1227                 $content = JsonLD::fetchElement($jsonld, 'description');
1228                 if (!empty($content) && is_string($content)) {
1229                         $media['description'] = trim($content);
1230                 }
1231
1232                 $content = JsonLD::fetchElement($jsonld, 'name');
1233                 if (!empty($content) && (($media['description'] ?? '') != trim($content))) {
1234                         $media['name'] = trim($content);
1235                 }
1236
1237                 $content = JsonLD::fetchElement($jsonld, 'contentUrl');
1238                 if (!empty($content) && is_string($content)) {
1239                         $media['content'] = trim($content);
1240                 }
1241
1242                 $content = JsonLD::fetchElement($jsonld, 'embedUrl');
1243                 if (!empty($content) && is_string($content)) {
1244                         $media['embed'] = trim($content);
1245                 }
1246
1247                 $content = JsonLD::fetchElement($jsonld, 'height');
1248                 if (!empty($content) && is_string($content)) {
1249                         $media['height'] = trim($content);
1250                 }
1251
1252                 $content = JsonLD::fetchElement($jsonld, 'width');
1253                 if (!empty($content) && is_string($content)) {
1254                         $media['width'] = trim($content);
1255                 }
1256
1257                 $content = JsonLD::fetchElement($jsonld, 'image');
1258                 if (!empty($content) && is_string($content)) {
1259                         $media['image'] = trim($content);
1260                 }
1261
1262                 $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
1263                 if (!empty($content) && (($media['image'] ?? '') != trim($content))) {
1264                         if (!empty($media['image'])) {
1265                                 $media['preview'] = trim($content);
1266                         } else {
1267                                 $media['image'] = trim($content);
1268                         }
1269                 }
1270
1271                 Logger::info('Fetched Media information', ['url' => $siteinfo['url'], 'fetched' => $media]);
1272                 $siteinfo[$name][] = $media;
1273                 return $siteinfo;
1274         }
1275 }