From e350b818e647a2d111c252594cbbdfad617b8538 Mon Sep 17 00:00:00 2001
From: TheSiahxyz <164138827+TheSiahxyz@users.noreply.github.com>
Date: Tue, 30 Jun 2026 10:36:57 +0900
Subject: modified newsboat/fulltext.py

---
 ar/.config/newsboat/fulltext.py | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

(limited to 'ar')

diff --git a/ar/.config/newsboat/fulltext.py b/ar/.config/newsboat/fulltext.py
index 4f567e5..33896b1 100755
--- a/ar/.config/newsboat/fulltext.py
+++ b/ar/.config/newsboat/fulltext.py
@@ -90,8 +90,11 @@ def extract(url):
         with open(cp, "r", encoding="utf-8", errors="replace") as f:
             return f.read()
     try:
+        # Force UTF-8 decoding: some sites (e.g. CNBC) serve UTF-8 but declare
+        # no charset in the HTTP header or an early <meta>, so rdrview falls back
+        # to Latin-1/CP1252 and mangles multibyte chars (em-dash -> "â€"").
         out = subprocess.run(
-            ["rdrview", "-H", url],
+            ["rdrview", "-E", "UTF-8", "-H", url],
             capture_output=True, text=True, timeout=TIMEOUT,
         )
     except (FileNotFoundError, subprocess.TimeoutExpired):
-- 
cgit v1.2.3