X-Git-Url: https://git.mxchange.org/?a=blobdiff_plain;f=src%2FContent%2FPageInfo.php;h=d3b38336b5d790baeaec023fbb22fa7266f2397a;hb=68046573a4171dcbd5db75a991487bafa4b4370f;hp=212082f9f3664e26ff53fa87d92bb7ad15c2b7fb;hpb=972b65ba33f65be3fee4b6201304702c51b22ed3;p=friendica.git

diff --git a/src/Content/PageInfo.php b/src/Content/PageInfo.php
index 212082f9f3..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;
 		}
@@ -129,25 +129,16 @@ class PageInfo
 		}
 
 		// Escape some bad characters
-		$data['url'] = str_replace(['[', ']'], ['&#91;', '&#93;'], htmlentities($data['url'], ENT_QUOTES, 'UTF-8', false));
-		$data['title'] = str_replace(['[', ']'], ['&#91;', '&#93;'], htmlentities($data['title'], ENT_QUOTES, 'UTF-8', false));
+		$text = "[attachment";
 
-		$text = "[attachment type='" . $data['type'] . "'";
-
-		if (empty($data['text'])) {
-			$data['text'] = $data['title'];
+		foreach (['type', 'url', 'title', 'alternative_title', 'publisher_name', 'publisher_url', 'publisher_img', 'author_name', 'author_url', 'author_img'] as $field) {
+			if (!empty($data[$field])) {
+				$text .= " " . $field . "='" . str_replace(['[', ']'], ['&#91;', '&#93;'], htmlentities($data[$field], ENT_QUOTES, 'UTF-8', false)) . "'";
+			}
 		}
 
 		if (empty($data['text'])) {
-			$data['text'] = $data['url'];
-		}
-
-		if (!empty($data['url'])) {
-			$text .= " url='" . $data['url'] . "'";
-		}
-
-		if (!empty($data['title'])) {
-			$text .= " title='" . $data['title'] . "'";
+			$data['text'] = '';
 		}
 
 		// Only embedd a picture link when it seems to be a valid picture ("width" is set)
@@ -160,10 +151,18 @@ class PageInfo
 				$text .= " image='" . $preview . "'";
 			} else {
 				$text .= " preview='" . $preview . "'";
+
+				if (empty($data['text'])) {
+					$data['text'] = $data['title'];
+				}
+
+				if (empty($data['text'])) {
+					$data['text'] = $data['url'];
+				}
 			}
 		}
 
-		$text .= ']' . $data['text'] . '[/attachment]';
+		$text .= ']' . str_replace(['[', ']'], ['&#91;', '&#93;'], $data['text']) . '[/attachment]';
 
 		$hashtags = '';
 		if (!empty($data['keywords'])) {
@@ -188,7 +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;
@@ -247,21 +246,26 @@ 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) {
-			preg_match('~(?<=\W|^)(?<![=\]])(https?://.+)$~is', $body, $matches);
+			preg_match(Strings::autoLinkRegEx(), $body, $matches);
 			if ($matches && !Strings::endsWith($body, $matches[1])) {
 				unset($matches);
 			}
@@ -272,22 +276,36 @@ class PageInfo
 
 	/**
 	 * Remove the provided URL from the body if it is at the end of it.
-	 * Keep the link label if it isn't the full URL.
+	 * Keep the link label if it isn't the full URL or a shortened version of it.
 	 *
 	 * @param string $body
 	 * @param string $url
-	 * @return string|string[]|null
+	 * @return string
 	 */
 	protected static function stripTrailingUrlFromBody(string $body, string $url)
 	{
 		$quotedUrl = preg_quote($url, '#');
-		$body = preg_replace("#(?:
+		$body = preg_replace_callback("#(?:
 			\[url]$quotedUrl\[/url]|
 			\[url=$quotedUrl]$quotedUrl\[/url]|
 			\[url=$quotedUrl]([^[]*?)\[/url]|
 			$quotedUrl
-		)$#isx", '$1', $body);
+		)$#isx", function ($match) use ($url) {
+			// Stripping URLs with no label
+			if (empty($match[1])) {
+				return '';
+			}
 
-		return $body;
+			// Stripping link labels that include a shortened version of the URL
+			$trimMatch = trim($match[1], '.…');
+			if (!empty($trimMatch) && strpos($url, $trimMatch) !== false) {
+				return '';
+			}
+
+			// Keep all other labels
+			return $match[1];
+		}, $body);
+
+		return rtrim($body);
 	}
 }