Merge pull request #11531 from annando/display-polls

[friendica.git] / src / Content / PageInfo.php

diff --git a/src/Content/PageInfo.php b/src/Content/PageInfo.php
index 39bd35f73af68511db77c60a441c2005305247ce..41ecb3d7fa35a0ec207747a7c174fe9707e51b39 100644 (file)
--- a/src/Content/PageInfo.php
+++ b/src/Content/PageInfo.php
@@ -1,6 +1,6 @@
 <?php
 /**
- * @copyright Copyright (C) 2020, Friendica
+ * @copyright Copyright (C) 2010-2022, 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;
                }
@@ -129,17 +129,12 @@ 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['url'])) {
-                       $text .= " url='" . $data['url'] . "'";
-               }
-
-               if (!empty($data['title'])) {
-                       $text .= " title='" . $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($data['text'])) {
@@ -160,14 +155,14 @@ class PageInfo
                                if (empty($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'])) {
@@ -192,7 +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;
@@ -251,21 +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);
                        }
@@ -292,12 +292,13 @@ class PageInfo
                        $quotedUrl
                )$#isx", function ($match) use ($url) {
                        // Stripping URLs with no label
-                       if (!isset($match[1])) {
+                       if (empty($match[1])) {
                                return '';
                        }
 
                        // 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 '';
                        }