<?php
/**
- * @file src/Util/ParseUrl.php
- * @brief Get informations about a given URL
+ * @copyright Copyright (C) 2010-2022, the Friendica project
+ *
+ * @license GNU AGPL version 3 or any later version
+ *
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU Affero General Public License as
+ * published by the Free Software Foundation, either version 3 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU Affero General Public License for more details.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with this program. If not, see <https://www.gnu.org/licenses/>.
+ *
*/
+
namespace Friendica\Util;
use DOMDocument;
use Friendica\Content\OEmbed;
use Friendica\Core\Hook;
use Friendica\Core\Logger;
+use Friendica\Database\Database;
use Friendica\Database\DBA;
-use Friendica\Object\Image;
+use Friendica\DI;
+use Friendica\Network\HTTPException;
+use Friendica\Network\HTTPClient\Client\HttpClientOptions;
/**
- * @brief Class with methods for extracting certain content from an url
+ * Get information about a given URL
+ *
+ * Class with methods for extracting certain content from an url
*/
class ParseUrl
{
+ const DEFAULT_EXPIRATION_FAILURE = 'now + 1 day';
+ const DEFAULT_EXPIRATION_SUCCESS = 'now + 3 months';
+
+ /**
+ * Maximum number of characters for the description
+ */
+ const MAX_DESC_COUNT = 250;
+
+ /**
+ * Minimum number of characters for the description
+ */
+ const MIN_DESC_COUNT = 100;
+
+ /**
+ * Fetch the content type of the given url
+ * @param string $url URL of the page
+ * @return array content type
+ */
+ public static function getContentType(string $url)
+ {
+ $curlResult = DI::httpClient()->head($url);
+
+ // Workaround for systems that can't handle a HEAD request
+ if (!$curlResult->isSuccess() && ($curlResult->getReturnCode() == 405)) {
+ $curlResult = DI::httpClient()->get($url, [HttpClientOptions::CONTENT_LENGTH => 1000000]);
+ }
+
+ if (!$curlResult->isSuccess()) {
+ return [];
+ }
+
+ $contenttype = $curlResult->getHeader('Content-Type')[0] ?? '';
+ if (empty($contenttype)) {
+ return ['application', 'octet-stream'];
+ }
+
+ return explode('/', current(explode(';', $contenttype)));
+ }
+
/**
- * @brief Search for chached embeddable data of an url otherwise fetch it
+ * Search for chached embeddable data of an url otherwise fetch it
*
* @param string $url The url of the page which should be scraped
- * @param bool $no_guessing If true the parse doens't search for
- * preview pictures
* @param bool $do_oembed The false option is used by the function fetch_oembed()
* to avoid endless loops
*
* @return array which contains needed data for embedding
- * string 'url' => The url of the parsed page
- * string 'type' => Content type
- * string 'title' => The title of the content
- * string 'text' => The description for the content
- * string 'image' => A preview image of the content (only available
- * if $no_geuessing = false
- * array'images' = Array of preview pictures
- * string 'keywords' => The tags which belong to the content
+ * string 'url' => The url of the parsed page
+ * string 'type' => Content type
+ * string 'title' => (optional) The title of the content
+ * string 'text' => (optional) The description for the content
+ * string 'image' => (optional) A preview image of the content
+ * array 'images' => (optional) Array of preview pictures
+ * string 'keywords' => (optional) The tags which belong to the content
*
- * @throws \Friendica\Network\HTTPException\InternalServerErrorException
+ * @throws HTTPException\InternalServerErrorException
* @see ParseUrl::getSiteinfo() for more information about scraping
* embeddable content
*/
- public static function getSiteinfoCached($url, $no_guessing = false, $do_oembed = true)
+ public static function getSiteinfoCached($url, $do_oembed = true): array
{
- if ($url == "") {
- return false;
+ if (empty($url)) {
+ return [
+ 'url' => '',
+ 'type' => 'error',
+ ];
}
+ $urlHash = hash('sha256', $url);
+
$parsed_url = DBA::selectFirst('parsed_url', ['content'],
- ['url' => Strings::normaliseLink($url), 'guessing' => !$no_guessing, 'oembed' => $do_oembed]
+ ['url_hash' => $urlHash, 'oembed' => $do_oembed]
);
if (!empty($parsed_url['content'])) {
$data = unserialize($parsed_url['content']);
return $data;
}
- $data = self::getSiteinfo($url, $no_guessing, $do_oembed);
+ $data = self::getSiteinfo($url, $do_oembed);
+
+ $expires = $data['expires'];
- DBA::insert(
+ unset($data['expires']);
+
+ DI::dba()->insert(
'parsed_url',
[
- 'url' => Strings::normaliseLink($url), 'guessing' => !$no_guessing,
- 'oembed' => $do_oembed, 'content' => serialize($data),
- 'created' => DateTimeFormat::utcNow()
+ 'url_hash' => $urlHash,
+ 'oembed' => $do_oembed,
+ 'url' => $url,
+ 'content' => serialize($data),
+ 'created' => DateTimeFormat::utcNow(),
+ 'expires' => $expires,
],
- true
+ Database::INSERT_UPDATE
);
return $data;
}
/**
- * @brief Parse a page for embeddable content information
+ * Parse a page for embeddable content information
*
* This method parses to url for meta data which can be used to embed
* the content. If available it prioritizes Open Graph meta tags.
* \<meta name="description" content="An awesome description"\>
*
* @param string $url The url of the page which should be scraped
- * @param bool $no_guessing If true the parse doens't search for
- * preview pictures
* @param bool $do_oembed The false option is used by the function fetch_oembed()
* to avoid endless loops
* @param int $count Internal counter to avoid endless loops
*
* @return array which contains needed data for embedding
- * string 'url' => The url of the parsed page
- * string 'type' => Content type
- * string 'title' => The title of the content
- * string 'text' => The description for the content
- * string 'image' => A preview image of the content (only available
- * if $no_geuessing = false
- * array'images' = Array of preview pictures
- * string 'keywords' => The tags which belong to the content
+ * string 'url' => The url of the parsed page
+ * string 'type' => Content type (error, link, photo, image, audio, video)
+ * string 'title' => (optional) The title of the content
+ * string 'text' => (optional) The description for the content
+ * string 'image' => (optional) A preview image of the content
+ * array 'images' => (optional) Array of preview pictures
+ * string 'keywords' => (optional) The tags which belong to the content
*
* @throws \Friendica\Network\HTTPException\InternalServerErrorException
* @todo https://developers.google.com/+/plugins/snippet/
* </body>
* @endverbatim
*/
- public static function getSiteinfo($url, $no_guessing = false, $do_oembed = true, $count = 1)
+ public static function getSiteinfo($url, $do_oembed = true, $count = 1)
{
- $a = \get_app();
-
- $siteinfo = [];
+ if (empty($url)) {
+ return [
+ 'url' => '',
+ 'type' => 'error',
+ ];
+ }
// Check if the URL does contain a scheme
$scheme = parse_url($url, PHP_URL_SCHEME);
if ($scheme == '') {
- $url = 'http://' . trim($url, '/');
+ $url = 'http://' . ltrim($url, '/');
}
- if ($count > 10) {
- Logger::log('Endless loop detected for ' . $url, Logger::DEBUG);
- return $siteinfo;
- }
-
- $url = trim($url, "'");
- $url = trim($url, '"');
+ $url = trim($url, "'\"");
$url = Network::stripTrackingQueryParams($url);
- $siteinfo['url'] = $url;
- $siteinfo['type'] = 'link';
+ $siteinfo = [
+ 'url' => $url,
+ 'type' => 'link',
+ 'expires' => DateTimeFormat::utc(self::DEFAULT_EXPIRATION_FAILURE),
+ ];
- $curlResult = Network::curl($url);
- if (!$curlResult->isSuccess()) {
+ if ($count > 10) {
+ Logger::notice('Endless loop detected', ['url' => $url]);
return $siteinfo;
}
- // If the file is too large then exit
- if (defaults($curlResult->getInfo(), 'download_content_length', 0) > 1000000) {
+ $type = self::getContentType($url);
+ Logger::info('Got content-type', ['content-type' => $type, 'url' => $url]);
+ if (!empty($type) && in_array($type[0], ['image', 'video', 'audio'])) {
+ $siteinfo['type'] = $type[0];
return $siteinfo;
}
- // If it isn't a HTML file then exit
- if (($curlResult->getContentType() != '') && !strstr(strtolower($curlResult->getContentType()), 'html')) {
+ if ((count($type) >= 2) && (($type[0] != 'text') || ($type[1] != 'html'))) {
+ Logger::info('Unparseable content-type, quitting here, ', ['content-type' => $type, 'url' => $url]);
return $siteinfo;
}
- $header = $curlResult->getHeader();
+ $curlResult = DI::httpClient()->get($url, [HttpClientOptions::CONTENT_LENGTH => 1000000]);
+ if (!$curlResult->isSuccess() || empty($curlResult->getBody())) {
+ Logger::info('Empty body or error when fetching', ['url' => $url, 'success' => $curlResult->isSuccess(), 'code' => $curlResult->getReturnCode()]);
+ return $siteinfo;
+ }
+
+ $siteinfo['expires'] = DateTimeFormat::utc(self::DEFAULT_EXPIRATION_SUCCESS);
+
+ if ($cacheControlHeader = $curlResult->getHeader('Cache-Control')[0] ?? '') {
+ if (preg_match('/max-age=([0-9]+)/i', $cacheControlHeader, $matches)) {
+ $maxAge = max(86400, (int)array_pop($matches));
+ $siteinfo['expires'] = DateTimeFormat::utc("now + $maxAge seconds");
+ }
+ }
+
$body = $curlResult->getBody();
if ($do_oembed) {
- $oembed_data = OEmbed::fetchURL($url);
+ $oembed_data = OEmbed::fetchURL($url, false, false);
if (!empty($oembed_data->type)) {
- if (!in_array($oembed_data->type, ['error', 'rich', ''])) {
+ if (!in_array($oembed_data->type, ['error', 'rich', 'image', 'video', 'audio', ''])) {
$siteinfo['type'] = $oembed_data->type;
}
// See https://github.com/friendica/friendica/pull/5763#discussion_r217913178
if ($siteinfo['type'] != 'photo') {
- if (isset($oembed_data->title)) {
+ if (!empty($oembed_data->title)) {
$siteinfo['title'] = trim($oembed_data->title);
}
- if (isset($oembed_data->description)) {
+ if (!empty($oembed_data->description)) {
$siteinfo['text'] = trim($oembed_data->description);
}
- if (isset($oembed_data->thumbnail_url)) {
+ if (!empty($oembed_data->author_name)) {
+ $siteinfo['author_name'] = trim($oembed_data->author_name);
+ }
+ if (!empty($oembed_data->author_url)) {
+ $siteinfo['author_url'] = trim($oembed_data->author_url);
+ }
+ if (!empty($oembed_data->provider_name)) {
+ $siteinfo['publisher_name'] = trim($oembed_data->provider_name);
+ }
+ if (!empty($oembed_data->provider_url)) {
+ $siteinfo['publisher_url'] = trim($oembed_data->provider_url);
+ }
+ if (!empty($oembed_data->thumbnail_url)) {
$siteinfo['image'] = $oembed_data->thumbnail_url;
}
}
}
}
- // Fetch the first mentioned charset. Can be in body or header
$charset = '';
- if (preg_match('/charset=(.*?)[\'"\s\n]/', $header, $matches)) {
+ // Look for a charset, first in headers
+ // Expected form: Content-Type: text/html; charset=ISO-8859-4
+ if (preg_match('/charset=([a-z0-9-_.\/]+)/i', $curlResult->getContentType(), $matches)) {
$charset = trim(trim(trim(array_pop($matches)), ';,'));
}
+ // Then in body that gets precedence
+ // Expected forms:
+ // - <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
+ // - <meta charset="utf-8">
+ // - <meta charset=utf-8>
+ // - <meta charSet="utf-8">
+ // We escape <style> and <script> tags since they can contain irrelevant charset information
+ // (see https://github.com/friendica/friendica/issues/9251#issuecomment-698636806)
+ Strings::performWithEscapedBlocks($body, '#<(?:style|script).*?</(?:style|script)>#ism', function ($body) use (&$charset) {
+ if (preg_match('/charset=["\']?([a-z0-9-_.\/]+)/i', $body, $matches)) {
+ $charset = trim(trim(trim(array_pop($matches)), ';,'));
+ }
+ });
+
+ $siteinfo['charset'] = $charset;
+
if ($charset && strtoupper($charset) != 'UTF-8') {
// See https://github.com/friendica/friendica/issues/5470#issuecomment-418351211
$charset = str_ireplace('latin-1', 'latin1', $charset);
- Logger::log('detected charset ' . $charset, Logger::DEBUG);
+ Logger::info('detected charset', ['charset' => $charset]);
$body = iconv($charset, 'UTF-8//TRANSLIT', $body);
}
@$doc->loadHTML($body);
XML::deleteNode($doc, 'style');
- XML::deleteNode($doc, 'script');
XML::deleteNode($doc, 'option');
XML::deleteNode($doc, 'h1');
XML::deleteNode($doc, 'h2');
}
}
if ($content != '') {
- $siteinfo = self::getSiteinfo($content, $no_guessing, $do_oembed, ++$count);
+ $siteinfo = self::getSiteinfo($content, $do_oembed, ++$count);
return $siteinfo;
}
}
case 'twitter:image:src':
$siteinfo['image'] = $meta_tag['content'];
break;
- case 'twitter:card':
- // Detect photo pages
- if ($meta_tag['content'] == 'summary_large_image') {
- $siteinfo['type'] = 'photo';
- }
- break;
case 'twitter:description':
$siteinfo['text'] = trim($meta_tag['content']);
break;
case 'twitter:title':
$siteinfo['title'] = trim($meta_tag['content']);
break;
+ case 'twitter:player':
+ $siteinfo['player']['embed'] = trim($meta_tag['content']);
+ break;
+ case 'twitter:player:stream':
+ $siteinfo['player']['stream'] = trim($meta_tag['content']);
+ break;
+ case 'twitter:player:width':
+ $siteinfo['player']['width'] = intval($meta_tag['content']);
+ break;
+ case 'twitter:player:height':
+ $siteinfo['player']['height'] = intval($meta_tag['content']);
+ break;
case 'dc.title':
$siteinfo['title'] = trim($meta_tag['content']);
break;
case 'dc.description':
$siteinfo['text'] = trim($meta_tag['content']);
break;
+ case 'dc.creator':
+ $siteinfo['publisher_name'] = trim($meta_tag['content']);
+ break;
case 'keywords':
$keywords = explode(',', $meta_tag['content']);
break;
case 'og:image':
$siteinfo['image'] = $meta_tag['content'];
break;
+ case 'og:image:url':
+ $siteinfo['image'] = $meta_tag['content'];
+ break;
+ case 'og:image:secure_url':
+ $siteinfo['image'] = $meta_tag['content'];
+ break;
case 'og:title':
$siteinfo['title'] = trim($meta_tag['content']);
break;
case 'og:description':
$siteinfo['text'] = trim($meta_tag['content']);
break;
+ case 'og:site_name':
+ $siteinfo['publisher_name'] = trim($meta_tag['content']);
+ break;
+ case 'og:locale':
+ $siteinfo['language'] = trim($meta_tag['content']);
+ break;
+ case 'og:type':
+ $siteinfo['pagetype'] = trim($meta_tag['content']);
+ break;
+ case 'twitter:description':
+ $siteinfo['text'] = trim($meta_tag['content']);
+ break;
+ case 'twitter:title':
+ $siteinfo['title'] = trim($meta_tag['content']);
+ break;
+ case 'twitter:image':
+ $siteinfo['image'] = $meta_tag['content'];
+ break;
}
}
}
- // Prevent to have a photo type without an image
- if ((empty($siteinfo['image']) || !empty($siteinfo['text'])) && ($siteinfo['type'] == 'photo')) {
- $siteinfo['type'] = 'link';
- }
-
- if (empty($siteinfo['image']) && !$no_guessing) {
- $list = $xpath->query('//img[@src]');
- foreach ($list as $node) {
- $img_tag = [];
- if ($node->attributes->length) {
- foreach ($node->attributes as $attribute) {
- $img_tag[$attribute->name] = $attribute->value;
- }
+ $list = $xpath->query("//script[@type='application/ld+json']");
+ foreach ($list as $node) {
+ if (!empty($node->nodeValue)) {
+ if ($jsonld = json_decode($node->nodeValue, true)) {
+ $siteinfo = self::parseParts($siteinfo, $jsonld);
}
+ }
+ }
- $src = self::completeUrl($img_tag['src'], $url);
- $photodata = Image::getInfoFromURL($src);
+ if (!empty($siteinfo['player']['stream'])) {
+ // Only add player data to media arrays if there is no duplicate
+ $content_urls = array_merge(array_column($siteinfo['audio'] ?? [], 'content'), array_column($siteinfo['video'] ?? [], 'content'));
+ if (!in_array($siteinfo['player']['stream'], $content_urls)) {
+ $contenttype = self::getContentType($siteinfo['player']['stream']);
+ if (!empty($contenttype[0]) && in_array($contenttype[0], ['audio', 'video'])) {
+ $media = ['content' => $siteinfo['player']['stream']];
- if (($photodata) && ($photodata[0] > 150) && ($photodata[1] > 150)) {
- if ($photodata[0] > 300) {
- $photodata[1] = round($photodata[1] * (300 / $photodata[0]));
- $photodata[0] = 300;
+ if (!empty($siteinfo['player']['embed'])) {
+ $media['embed'] = $siteinfo['player']['embed'];
}
- if ($photodata[1] > 300) {
- $photodata[0] = round($photodata[0] * (300 / $photodata[1]));
- $photodata[1] = 300;
- }
- $siteinfo['images'][] = [
- 'src' => $src,
- 'width' => $photodata[0],
- 'height' => $photodata[1]
- ];
+
+ $siteinfo[$contenttype[0]][] = $media;
}
}
- } elseif (!empty($siteinfo['image'])) {
- $src = self::completeUrl($siteinfo['image'], $url);
+ }
+ if (!empty($siteinfo['image'])) {
+ $siteinfo['images'] = $siteinfo['images'] ?? [];
+ array_unshift($siteinfo['images'], ['url' => $siteinfo['image']]);
unset($siteinfo['image']);
+ }
- $photodata = Image::getInfoFromURL($src);
+ $siteinfo = self::checkMedia($url, $siteinfo);
- if (($photodata) && ($photodata[0] > 10) && ($photodata[1] > 10)) {
- $siteinfo['images'][] = ['src' => $src,
- 'width' => $photodata[0],
- 'height' => $photodata[1]];
+ if (!empty($siteinfo['text']) && mb_strlen($siteinfo['text']) > self::MAX_DESC_COUNT) {
+ $siteinfo['text'] = mb_substr($siteinfo['text'], 0, self::MAX_DESC_COUNT) . '…';
+ $pos = mb_strrpos($siteinfo['text'], '.');
+ if ($pos > self::MIN_DESC_COUNT) {
+ $siteinfo['text'] = mb_substr($siteinfo['text'], 0, $pos + 1);
}
}
- if ((@$siteinfo['text'] == '') && (@$siteinfo['title'] != '') && !$no_guessing) {
- $text = '';
+ Logger::info('Siteinfo fetched', ['url' => $url, 'siteinfo' => $siteinfo]);
- $list = $xpath->query('//div[@class="article"]');
- foreach ($list as $node) {
- if (strlen($node->nodeValue) > 40) {
- $text .= ' ' . trim($node->nodeValue);
- }
- }
+ Hook::callAll('getsiteinfo', $siteinfo);
- if ($text == '') {
- $list = $xpath->query('//div[@class="content"]');
- foreach ($list as $node) {
- if (strlen($node->nodeValue) > 40) {
- $text .= ' ' . trim($node->nodeValue);
+ ksort($siteinfo);
+
+ return $siteinfo;
+ }
+
+ /**
+ * Check the attached media elements.
+ * Fix existing data and add missing data.
+ *
+ * @param string $page_url
+ * @param array $siteinfo
+ * @return array
+ */
+ private static function checkMedia(string $page_url, array $siteinfo) : array
+ {
+ if (!empty($siteinfo['images'])) {
+ array_walk($siteinfo['images'], function (&$image) use ($page_url) {
+ // According to the specifications someone could place a picture url into the content field as well.
+ // But this doesn't seem to happen in the wild, so we don't cover it here.
+ if (!empty($image['url'])) {
+ $image['url'] = self::completeUrl($image['url'], $page_url);
+ $photodata = Images::getInfoFromURLCached($image['url']);
+ if (!empty($photodata) && ($photodata[0] > 50) && ($photodata[1] > 50)) {
+ $image['src'] = $image['url'];
+ $image['width'] = $photodata[0];
+ $image['height'] = $photodata[1];
+ $image['contenttype'] = $photodata['mime'];
+ unset($image['url']);
+ ksort($image);
+ } else {
+ $image = [];
}
+ } else {
+ $image = [];
}
- }
+ });
- // If none text was found then take the paragraph content
- if ($text == '') {
- $list = $xpath->query('//p');
- foreach ($list as $node) {
- if (strlen($node->nodeValue) > 40) {
- $text .= ' ' . trim($node->nodeValue);
+ $siteinfo['images'] = array_values(array_filter($siteinfo['images']));
+ }
+
+ foreach (['audio', 'video'] as $element) {
+ if (!empty($siteinfo[$element])) {
+ array_walk($siteinfo[$element], function (&$media) use ($page_url, &$siteinfo) {
+ $url = '';
+ $embed = '';
+ $content = '';
+ $contenttype = '';
+ foreach (['embed', 'content', 'url'] as $field) {
+ if (!empty($media[$field])) {
+ $media[$field] = self::completeUrl($media[$field], $page_url);
+ $type = self::getContentType($media[$field]);
+ if (($type[0] ?? '') == 'text') {
+ if ($field == 'embed') {
+ $embed = $media[$field];
+ } else {
+ $url = $media[$field];
+ }
+ } elseif (!empty($type[0])) {
+ $content = $media[$field];
+ $contenttype = implode('/', $type);
+ }
+ }
+ unset($media[$field]);
}
- }
- }
- if ($text != '') {
- $text = trim(str_replace(["\n", "\r"], [' ', ' '], $text));
+ foreach (['image', 'preview'] as $field) {
+ if (!empty($media[$field])) {
+ $media[$field] = self::completeUrl($media[$field], $page_url);
+ }
+ }
- while (strpos($text, ' ')) {
- $text = trim(str_replace(' ', ' ', $text));
- }
+ if (!empty($url)) {
+ $media['url'] = $url;
+ }
+ if (!empty($embed)) {
+ $media['embed'] = $embed;
+ if (empty($siteinfo['player']['embed'])) {
+ $siteinfo['player']['embed'] = $embed;
+ }
+ }
+ if (!empty($content)) {
+ $media['src'] = $content;
+ }
+ if (!empty($contenttype)) {
+ $media['contenttype'] = $contenttype;
+ }
+ if (empty($url) && empty($content) && empty($embed)) {
+ $media = [];
+ }
+ ksort($media);
+ });
- $siteinfo['text'] = trim(html_entity_decode(substr($text, 0, 350), ENT_QUOTES, 'UTF-8') . '...');
+ $siteinfo[$element] = array_values(array_filter($siteinfo[$element]));
+ }
+ if (empty($siteinfo[$element])) {
+ unset($siteinfo[$element]);
}
}
-
- Logger::log('Siteinfo for ' . $url . ' ' . print_r($siteinfo, true), Logger::DEBUG);
-
- Hook::callAll('getsiteinfo', $siteinfo);
-
return $siteinfo;
}
/**
- * @brief Convert tags from CSV to an array
+ * Convert tags from CSV to an array
*
* @param string $string Tags
* @return array with formatted Hashtags
}
/**
- * @brief Add a hasht sign to a string
+ * Add a hasht sign to a string
*
- * This method is used as callback function
+ * This method is used as callback function
*
* @param string $tag The pure tag name
* @param int $k Counter for internal use
}
/**
- * @brief Add a scheme to an url
+ * Add a scheme to an url
*
* The src attribute of some html elements (e.g. images)
* can miss the scheme so we need to add the correct
return($complete);
}
+
+ /**
+ * Parse the Json-Ld parts of a web page
+ *
+ * @param array $siteinfo
+ * @param array $jsonld
+ * @return array siteinfo
+ */
+ private static function parseParts(array $siteinfo, array $jsonld)
+ {
+ if (!empty($jsonld['@graph']) && is_array($jsonld['@graph'])) {
+ foreach ($jsonld['@graph'] as $part) {
+ if (!empty($part) && is_array($part)) {
+ $siteinfo = self::parseParts($siteinfo, $part);
+ }
+ }
+ } elseif (!empty($jsonld['@type'])) {
+ $siteinfo = self::parseJsonLd($siteinfo, $jsonld);
+ } elseif (!empty($jsonld)) {
+ $keys = array_keys($jsonld);
+ $numeric_keys = true;
+ foreach ($keys as $key) {
+ if (!is_int($key)) {
+ $numeric_keys = false;
+ }
+ }
+ if ($numeric_keys) {
+ foreach ($jsonld as $part) {
+ if (!empty($part) && is_array($part)) {
+ $siteinfo = self::parseParts($siteinfo, $part);
+ }
+ }
+ }
+ }
+
+ array_walk_recursive($siteinfo, function (&$element) {
+ if (is_string($element)) {
+ $element = trim(strip_tags(html_entity_decode($element, ENT_COMPAT, 'UTF-8')));
+ }
+ });
+
+ return $siteinfo;
+ }
+
+ /**
+ * Improve the siteinfo with information from the provided JSON-LD information
+ * @see https://jsonld.com/
+ * @see https://schema.org/
+ *
+ * @param array $siteinfo
+ * @param array $jsonld
+ * @return array siteinfo
+ */
+ private static function parseJsonLd(array $siteinfo, array $jsonld)
+ {
+ $type = JsonLD::fetchElement($jsonld, '@type');
+ if (empty($type)) {
+ Logger::info('Empty type', ['url' => $siteinfo['url']]);
+ return $siteinfo;
+ }
+
+ // Silently ignore some types that aren't processed
+ if (in_array($type, ['SiteNavigationElement', 'JobPosting', 'CreativeWork', 'MusicAlbum',
+ 'WPHeader', 'WPSideBar', 'WPFooter', 'LegalService', 'MusicRecording',
+ 'ItemList', 'BreadcrumbList', 'Blog', 'Dataset', 'Product'])) {
+ return $siteinfo;
+ }
+
+ switch ($type) {
+ case 'Article':
+ case 'AdvertiserContentArticle':
+ case 'NewsArticle':
+ case 'Report':
+ case 'SatiricalArticle':
+ case 'ScholarlyArticle':
+ case 'SocialMediaPosting':
+ case 'TechArticle':
+ case 'ReportageNewsArticle':
+ case 'SocialMediaPosting':
+ case 'BlogPosting':
+ case 'LiveBlogPosting':
+ case 'DiscussionForumPosting':
+ return self::parseJsonLdArticle($siteinfo, $jsonld);
+ case 'WebPage':
+ case 'AboutPage':
+ case 'CheckoutPage':
+ case 'CollectionPage':
+ case 'ContactPage':
+ case 'FAQPage':
+ case 'ItemPage':
+ case 'MedicalWebPage':
+ case 'ProfilePage':
+ case 'QAPage':
+ case 'RealEstateListing':
+ case 'SearchResultsPage':
+ case 'MediaGallery':
+ case 'ImageGallery':
+ case 'VideoGallery':
+ case 'RadioEpisode':
+ case 'Event':
+ return self::parseJsonLdWebPage($siteinfo, $jsonld);
+ case 'WebSite':
+ return self::parseJsonLdWebSite($siteinfo, $jsonld);
+ case 'Organization':
+ case 'Airline':
+ case 'Consortium':
+ case 'Corporation':
+ case 'EducationalOrganization':
+ case 'FundingScheme':
+ case 'GovernmentOrganization':
+ case 'LibrarySystem':
+ case 'LocalBusiness':
+ case 'MedicalOrganization':
+ case 'NGO':
+ case 'NewsMediaOrganization':
+ case 'Project':
+ case 'SportsOrganization':
+ case 'WorkersUnion':
+ return self::parseJsonLdWebOrganization($siteinfo, $jsonld);
+ case 'Person':
+ case 'Patient':
+ case 'PerformingGroup':
+ case 'DanceGroup';
+ case 'MusicGroup':
+ case 'TheaterGroup':
+ return self::parseJsonLdWebPerson($siteinfo, $jsonld);
+ case 'AudioObject':
+ case 'Audio':
+ return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'audio');
+ case 'VideoObject':
+ return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'video');
+ case 'ImageObject':
+ return self::parseJsonLdMediaObject($siteinfo, $jsonld, 'images');
+ default:
+ Logger::info('Unknown type', ['type' => $type, 'url' => $siteinfo['url']]);
+ return $siteinfo;
+ }
+ }
+
+ /**
+ * Fetch author and publisher data
+ *
+ * @param array $siteinfo
+ * @param array $jsonld
+ * @return array siteinfo
+ */
+ private static function parseJsonLdAuthor(array $siteinfo, array $jsonld)
+ {
+ $jsonldinfo = [];
+
+ if (!empty($jsonld['publisher']) && is_array($jsonld['publisher'])) {
+ $content = JsonLD::fetchElement($jsonld, 'publisher', 'name');
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['publisher_name'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'publisher', 'url');
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['publisher_url'] = trim($content);
+ }
+
+ $brand = JsonLD::fetchElement($jsonld, 'publisher', 'brand', '@type', 'Organization');
+ if (!empty($brand) && is_array($brand)) {
+ $content = JsonLD::fetchElement($brand, 'name');
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['publisher_name'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($brand, 'url');
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['publisher_url'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($brand, 'logo', 'url');
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['publisher_img'] = trim($content);
+ }
+ }
+
+ $logo = JsonLD::fetchElement($jsonld, 'publisher', 'logo');
+ if (!empty($logo) && is_array($logo)) {
+ $content = JsonLD::fetchElement($logo, 'url');
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['publisher_img'] = trim($content);
+ }
+ }
+ } elseif (!empty($jsonld['publisher']) && is_string($jsonld['publisher'])) {
+ $jsonldinfo['publisher_name'] = trim($jsonld['publisher']);
+ }
+
+ if (!empty($jsonld['author']) && is_array($jsonld['author'])) {
+ $content = JsonLD::fetchElement($jsonld, 'author', 'name');
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['author_name'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'author', 'sameAs');
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['author_url'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'author', 'url');
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['author_url'] = trim($content);
+ }
+
+ $logo = JsonLD::fetchElement($jsonld, 'author', 'logo');
+ if (!empty($logo) && is_array($logo)) {
+ $content = JsonLD::fetchElement($logo, 'url');
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['author_img'] = trim($content);
+ }
+ }
+ } elseif (!empty($jsonld['author']) && is_string($jsonld['author'])) {
+ $jsonldinfo['author_name'] = trim($jsonld['author']);
+ }
+
+ Logger::info('Fetched Author information', ['fetched' => $jsonldinfo]);
+
+ return array_merge($siteinfo, $jsonldinfo);
+ }
+
+ /**
+ * Fetch data from the provided JSON-LD Article type
+ * @see https://schema.org/Article
+ *
+ * @param array $siteinfo
+ * @param array $jsonld
+ * @return array siteinfo
+ */
+ private static function parseJsonLdArticle(array $siteinfo, array $jsonld)
+ {
+ $jsonldinfo = [];
+
+ $content = JsonLD::fetchElement($jsonld, 'headline');
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['title'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'alternativeHeadline');
+ if (!empty($content) && is_string($content) && (($jsonldinfo['title'] ?? '') != trim($content))) {
+ $jsonldinfo['alternative_title'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'description');
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['text'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
+ if (!empty($content)) {
+ $jsonldinfo['image'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'image', 'url', '@type', 'ImageObject');
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['image'] = trim($content);
+ }
+
+ if (!empty($jsonld['keywords']) && !is_array($jsonld['keywords'])) {
+ $content = JsonLD::fetchElement($jsonld, 'keywords');
+ if (!empty($content)) {
+ $siteinfo['keywords'] = [];
+ $keywords = explode(',', $content);
+ foreach ($keywords as $keyword) {
+ $siteinfo['keywords'][] = trim($keyword);
+ }
+ }
+ } elseif (!empty($jsonld['keywords'])) {
+ $content = JsonLD::fetchElementArray($jsonld, 'keywords');
+ if (!empty($content) && is_array($content)) {
+ $jsonldinfo['keywords'] = $content;
+ }
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'datePublished');
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['published'] = DateTimeFormat::utc($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'dateModified');
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['modified'] = DateTimeFormat::utc($content);
+ }
+
+ $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
+
+ Logger::info('Fetched article information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
+
+ return array_merge($siteinfo, $jsonldinfo);
+ }
+
+ /**
+ * Fetch data from the provided JSON-LD WebPage type
+ * @see https://schema.org/WebPage
+ *
+ * @param array $siteinfo
+ * @param array $jsonld
+ * @return array siteinfo
+ */
+ private static function parseJsonLdWebPage(array $siteinfo, array $jsonld)
+ {
+ $jsonldinfo = [];
+
+ $content = JsonLD::fetchElement($jsonld, 'name');
+ if (!empty($content)) {
+ $jsonldinfo['title'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'description');
+ if (!empty($content)) {
+ $jsonldinfo['text'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'image');
+ if (!empty($content)) {
+ $jsonldinfo['image'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
+ if (!empty($content)) {
+ $jsonldinfo['image'] = trim($content);
+ }
+
+ $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
+
+ Logger::info('Fetched WebPage information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
+
+ return array_merge($siteinfo, $jsonldinfo);
+ }
+
+ /**
+ * Fetch data from the provided JSON-LD WebSite type
+ * @see https://schema.org/WebSite
+ *
+ * @param array $siteinfo
+ * @param array $jsonld
+ * @return array siteinfo
+ */
+ private static function parseJsonLdWebSite(array $siteinfo, array $jsonld)
+ {
+ $jsonldinfo = [];
+
+ $content = JsonLD::fetchElement($jsonld, 'name');
+ if (!empty($content)) {
+ $jsonldinfo['publisher_name'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'description');
+ if (!empty($content)) {
+ $jsonldinfo['publisher_description'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'url');
+ if (!empty($content)) {
+ $jsonldinfo['publisher_url'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
+ if (!empty($content)) {
+ $jsonldinfo['image'] = trim($content);
+ }
+
+ $jsonldinfo = self::parseJsonLdAuthor($jsonldinfo, $jsonld);
+
+ Logger::info('Fetched WebSite information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
+ return array_merge($siteinfo, $jsonldinfo);
+ }
+
+ /**
+ * Fetch data from the provided JSON-LD Organization type
+ * @see https://schema.org/Organization
+ *
+ * @param array $siteinfo
+ * @param array $jsonld
+ * @return array siteinfo
+ */
+ private static function parseJsonLdWebOrganization(array $siteinfo, array $jsonld)
+ {
+ $jsonldinfo = [];
+
+ $content = JsonLD::fetchElement($jsonld, 'name');
+ if (!empty($content)) {
+ $jsonldinfo['publisher_name'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'description');
+ if (!empty($content)) {
+ $jsonldinfo['publisher_description'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'url');
+ if (!empty($content)) {
+ $jsonldinfo['publisher_url'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'logo', 'url', '@type', 'ImageObject');
+ if (!empty($content)) {
+ $jsonldinfo['publisher_img'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'brand', 'name', '@type', 'Organization');
+ if (!empty($content)) {
+ $jsonldinfo['publisher_name'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'brand', 'url', '@type', 'Organization');
+ if (!empty($content)) {
+ $jsonldinfo['publisher_url'] = trim($content);
+ }
+
+ Logger::info('Fetched Organization information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
+ return array_merge($siteinfo, $jsonldinfo);
+ }
+
+ /**
+ * Fetch data from the provided JSON-LD Person type
+ * @see https://schema.org/Person
+ *
+ * @param array $siteinfo
+ * @param array $jsonld
+ * @return array siteinfo
+ */
+ private static function parseJsonLdWebPerson(array $siteinfo, array $jsonld)
+ {
+ $jsonldinfo = [];
+
+ $content = JsonLD::fetchElement($jsonld, 'name');
+ if (!empty($content)) {
+ $jsonldinfo['author_name'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'description');
+ if (!empty($content)) {
+ $jsonldinfo['author_description'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'sameAs');
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['author_url'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'url');
+ if (!empty($content)) {
+ $jsonldinfo['author_url'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'image', 'url', '@type', 'ImageObject');
+ if (!empty($content) && !is_string($content)) {
+ Logger::notice('Unexpected return value for the author image', ['content' => $content]);
+ }
+
+ if (!empty($content) && is_string($content)) {
+ $jsonldinfo['author_img'] = trim($content);
+ }
+
+ Logger::info('Fetched Person information', ['url' => $siteinfo['url'], 'fetched' => $jsonldinfo]);
+ return array_merge($siteinfo, $jsonldinfo);
+ }
+
+ /**
+ * Fetch data from the provided JSON-LD MediaObject type
+ * @see https://schema.org/MediaObject
+ *
+ * @param array $siteinfo
+ * @param array $jsonld
+ * @return array siteinfo
+ */
+ private static function parseJsonLdMediaObject(array $siteinfo, array $jsonld, string $name)
+ {
+ $media = [];
+
+ $content = JsonLD::fetchElement($jsonld, 'caption');
+ if (!empty($content)) {
+ $media['caption'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'url');
+ if (!empty($content)) {
+ $media['url'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'mainEntityOfPage');
+ if (!empty($content)) {
+ $media['main'] = Strings::compareLink($content, $siteinfo['url']);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'description');
+ if (!empty($content)) {
+ $media['description'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'name');
+ if (!empty($content) && (($media['description'] ?? '') != trim($content))) {
+ $media['name'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'contentUrl');
+ if (!empty($content)) {
+ $media['content'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'embedUrl');
+ if (!empty($content)) {
+ $media['embed'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'height');
+ if (!empty($content)) {
+ $media['height'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'width');
+ if (!empty($content)) {
+ $media['width'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'image');
+ if (!empty($content)) {
+ $media['image'] = trim($content);
+ }
+
+ $content = JsonLD::fetchElement($jsonld, 'thumbnailUrl');
+ if (!empty($content) && (($media['image'] ?? '') != trim($content))) {
+ if (!empty($media['image'])) {
+ $media['preview'] = trim($content);
+ } else {
+ $media['image'] = trim($content);
+ }
+ }
+
+ Logger::info('Fetched Media information', ['url' => $siteinfo['url'], 'fetched' => $media]);
+ $siteinfo[$name][] = $media;
+ return $siteinfo;
+ }
}