From c0db1e52aeffabf1ad12e81c6271e9bfc009177e Mon Sep 17 00:00:00 2001
From: Timmy Welch <timmy@narnian.us>
Date: Tue, 12 Mar 2024 18:20:12 -0700
Subject: [PATCH] Make cleanup_html produce text that is more compliant with
 markdown

---
 comictalker/talker_utils.py | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/comictalker/talker_utils.py b/comictalker/talker_utils.py
index 8453f97..233d948 100644
--- a/comictalker/talker_utils.py
+++ b/comictalker/talker_utils.py
@@ -47,9 +47,10 @@ def cleanup_html(string: str | None, remove_html_tables: bool = False) -> str:
 
     # put in our own
     string = re.sub(r"<br>|</li>", "\n", string, flags=re.IGNORECASE)
+    string = re.sub(r"<li>", "* ", string, flags=re.IGNORECASE)
     string = re.sub(r"</p>", "\n\n", string, flags=re.IGNORECASE)
-    string = re.sub(r"<h([1-6])>", "*", string, flags=re.IGNORECASE)
-    string = re.sub(r"</h[1-6]>", "*\n", string, flags=re.IGNORECASE)
+    string = re.sub(r"<h([1-6])>", lambda m: '#'*int(m.group(1))+' ', string, flags=re.IGNORECASE)
+    string = re.sub(r"</h[1-6]>", "\n", string, flags=re.IGNORECASE)
 
     # remove the tables
     p = re.compile(r"<table[^<]*?>.*?</table>")