X-Git-Url: https://git.mxchange.org/?a=blobdiff_plain;f=src%2FContent%2FPageInfo.php;h=d3b38336b5d790baeaec023fbb22fa7266f2397a;hb=68046573a4171dcbd5db75a991487bafa4b4370f;hp=776dc15df53ca0a2dabc6741479dbe8394b3c4aa;hpb=0a3d50a270620c15c901dcfc96be808d4a7c1278;p=friendica.git

diff --git a/src/Content/PageInfo.php b/src/Content/PageInfo.php
index 776dc15df5..d3b38336b5 100644
--- a/src/Content/PageInfo.php
+++ b/src/Content/PageInfo.php
@@ -1,6 +1,6 @@
 <?php
 /**
- * @copyright Copyright (C) 2020, Friendica
+ * @copyright Copyright (C) 2010-2021, the Friendica project
  *
  * @license GNU AGPL version 3 or any later version
  *
@@ -73,7 +73,7 @@ class PageInfo
 			// Additional link attachments are prepended before the existing [attachment] tag
 			$body = substr_replace($body, "\n[bookmark=" . $data['url'] . ']' . $linkTitle . "[/bookmark]\n", $existingAttachmentPos, 0);
 		} else {
-			$footer = PageInfo::getFooterFromData($data, $no_photos);
+			$footer = self::getFooterFromData($data, $no_photos);
 			$body = self::stripTrailingUrlFromBody($body, $data['url']);
 			$body .= "\n" . $footer;
 		}
@@ -155,7 +155,7 @@ class PageInfo
 				if (empty($data['text'])) {
 					$data['text'] = $data['title'];
 				}
-		
+
 				if (empty($data['text'])) {
 					$data['text'] = $data['url'];
 				}
@@ -187,7 +187,7 @@ class PageInfo
 	 */
 	public static function queryUrl(string $url, string $photo = '', bool $keywords = false, string $keyword_denylist = '')
 	{
-		$data = ParseUrl::getSiteinfoCached($url, true);
+		$data = ParseUrl::getSiteinfoCached($url);
 
 		if ($photo != '') {
 			$data['images'][0]['src'] = $photo;
@@ -246,17 +246,22 @@ class PageInfo
 	 * @param bool   $searchNakedUrls Whether we should pick a naked URL (outside of BBCode tags) as a last resort
 	 * @return string|null
 	 */
-	protected static function getRelevantUrlFromBody(string $body, bool $searchNakedUrls = false)
+	public static function getRelevantUrlFromBody(string $body, bool $searchNakedUrls = false)
 	{
 		$URLSearchString = 'https?://[^\[\]]*';
 
 		// Fix for Mastodon where the mentions are in a different format
 		$body = preg_replace("~\[url=($URLSearchString)]([#!@])(.*?)\[/url]~is", '$2[url=$1]$3[/url]', $body);
 
-		preg_match("~(?<![!#@])\[url]($URLSearchString)\[/url]$~is", $body, $matches);
+		// Remove all hashtags and mentions
+		$body = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '', $body);
+
+		// Search for pure links
+		preg_match("/\[url\](https?:.*?)\[\/url\]/ism", $body, $matches);
 
 		if (!$matches) {
-			preg_match("~(?<![!#@])\[url=($URLSearchString)].*\[/url]$~is", $body, $matches);
+			// Search for links with descriptions
+			preg_match("/\[url\=(https?:.*?)\].*?\[\/url\]/ism", $body, $matches);
 		}
 
 		if (!$matches && $searchNakedUrls) {
@@ -292,7 +297,8 @@ class PageInfo
 			}
 
 			// Stripping link labels that include a shortened version of the URL
-			if (strpos($url, trim($match[1], '.…')) !== false) {
+			$trimMatch = trim($match[1], '.…');
+			if (!empty($trimMatch) && strpos($url, $trimMatch) !== false) {
 				return '';
 			}