From f1487a3cfa5cb73d71f5fb52e4f5f6809136bf1b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Mike=20F=C3=A4hrmann?= <mike_faehrmann@web.de>
Date: Sun, 24 Oct 2021 21:15:21 +0200
Subject: [PATCH] [kemonoparty:discord] improve 'inline' extraction (#1940)

- extract media.discordapp.*NET* URLs
- rewrite media.discordapp.net to cdn.discordapp.com
- use a more restricted set of characters for the URL path
---
 gallery_dl/extractor/kemonoparty.py | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/gallery_dl/extractor/kemonoparty.py b/gallery_dl/extractor/kemonoparty.py
index 8dc29c4b43..d5aad67248 100644
--- a/gallery_dl/extractor/kemonoparty.py
+++ b/gallery_dl/extractor/kemonoparty.py
@@ -237,7 +237,8 @@ def items(self):
         self._prepare_ddosguard_cookies()
 
         find_inline = re.compile(
-            r"https?://(?:cdn|media)\.discordapp.com/\S+").findall
+            r"https?://(?:cdn\.discordapp.com|media\.discordapp\.net)"
+            r"(/[A-Za-z0-9-._~:/?#\[\]@!$&'()*+,;%=]+)").findall
 
         posts = self.posts()
         max_posts = self.config("max-posts")
@@ -251,7 +252,8 @@ def items(self):
                 attachment["type"] = "attachment"
                 append(attachment)
             for path in find_inline(post["content"] or ""):
-                append({"path": path, "name": path, "type": "inline"})
+                append({"path": "https://cdn.discordapp.com" + path,
+                        "name": path, "type": "inline"})
 
             post["channel_name"] = self.channel_name
             post["date"] = text.parse_datetime(