Merge pull request #12679 from nupplaphil/feat/reduce_config_dependency

[friendica.git] / src / Util / ParseUrl.php

diff --git a/src/Util/ParseUrl.php b/src/Util/ParseUrl.php
index d0b168278563e24be39d6a79c01939846e7e1d98..9d19a4ebac9e02aade552216f4bbfc23f9727f21 100644 (file)
--- a/src/Util/ParseUrl.php
+++ b/src/Util/ParseUrl.php
@@ -1,6 +1,6 @@
 <?php
 /**
- * @copyright Copyright (C) 2010-2022, the Friendica project
+ * @copyright Copyright (C) 2010-2023, the Friendica project
  *
  * @license GNU AGPL version 3 or any later version
  *
@@ -24,6 +24,8 @@ namespace Friendica\Util;
 use DOMDocument;
 use DOMXPath;
 use Friendica\Content\OEmbed;
+use Friendica\Content\Text\HTML;
+use Friendica\Protocol\HTTP\MediaType;
 use Friendica\Core\Hook;
 use Friendica\Core\Logger;
 use Friendica\Database\Database;
@@ -214,7 +216,7 @@ class ParseUrl
                ];
 
                if ($count > 10) {
-                       Logger::notice('Endless loop detected', ['url' => $url]);
+                       Logger::warning('Endless loop detected', ['url' => $url]);
                        return $siteinfo;
                }
 
@@ -283,25 +285,13 @@ class ParseUrl
                }
 
                $charset = '';
-               // 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)), ';,'));
+               try {
+                       // Look for a charset, first in headers
+                       $mediaType = MediaType::fromContentType($curlResult->getContentType());
+                       if (isset($mediaType->parameters['charset'])) {
+                               $charset = $mediaType->parameters['charset'];
                        }
-               });
+               } catch(\InvalidArgumentException $e) {}
 
                $siteinfo['charset'] = $charset;
 
@@ -322,6 +312,8 @@ class ParseUrl
                $doc = new DOMDocument();
                @$doc->loadHTML($body);
 
+               $siteinfo['charset'] = HTML::extractCharset($doc) ?? $siteinfo['charset'];
+
                XML::deleteNode($doc, 'style');
                XML::deleteNode($doc, 'option');
                XML::deleteNode($doc, 'h1');
@@ -567,6 +559,7 @@ class ParseUrl
                                                $image['width'] = $photodata[0];
                                                $image['height'] = $photodata[1];
                                                $image['contenttype'] = $photodata['mime'];
+                                               $image['blurhash'] = $photodata['blurhash'] ?? null;
                                                unset($image['url']);
                                                ksort($image);
                                        } else {
@@ -1131,6 +1124,8 @@ class ParseUrl
                $content = JsonLD::fetchElement($jsonld, 'logo', 'url', '@type', 'ImageObject');
                if (!empty($content) && is_string($content)) {
                        $jsonldinfo['publisher_img'] = trim($content);
+               } elseif (!empty($content) && is_array($content)) {
+                       $jsonldinfo['publisher_img'] = trim($content[0]);
                }
 
                $content = JsonLD::fetchElement($jsonld, 'brand', 'name', '@type', 'Organization');