From dd55ad92e6fef9b4ece185ddea262cba6becb39f Mon Sep 17 00:00:00 2001
From: Bob Mottram <bob@libreserver.org>
Date: Wed, 29 Nov 2023 11:37:44 +0000
Subject: [PATCH] Remove zero length strings from feeds

---
 newswire.py | 2 ++
 utils.py    | 6 ++++++
 2 files changed, 8 insertions(+)

diff --git a/newswire.py b/newswire.py
index 2e1c48c5a..34dcac4b1 100644
--- a/newswire.py
+++ b/newswire.py
@@ -19,6 +19,7 @@ from datetime import timezone
 from collections import OrderedDict
 from utils import valid_post_date
 from categories import set_hashtag_category
+from utils import remove_zero_length_strings
 from utils import date_from_string_format
 from utils import acct_handle_dir
 from utils import remove_eol
@@ -1373,6 +1374,7 @@ def get_rss(base_dir: str, domain: str, session, url: str,
                         timeout=timeout_sec,
                         allow_redirects=True)
         if result:
+            result.text = remove_zero_length_strings(result.text)
             if int(len(result.text) / 1024) >= max_feed_size_kb:
                 print('WARN: feed is too large: ' + url)
             elif not contains_invalid_chars(result.text):
diff --git a/utils.py b/utils.py
index bb90553a7..705168ab5 100644
--- a/utils.py
+++ b/utils.py
@@ -46,6 +46,12 @@ INVALID_ACTOR_URL_CHARACTERS = (
 )
 
 
+def remove_zero_length_strings(text: str) -> str:
+    """removes zero length strings from text
+    """
+    return text.replace('​', '')
+
+
 def _utc_mktime(utc_tuple):
     """Returns number of seconds elapsed since epoch
     Note that no timezone are taken into consideration.