From 75e2e0cb3ab86f4a1231a141c724c19fdb4839c9 Mon Sep 17 00:00:00 2001
From: Nathan Breit <nabreit@gmail.com>
Date: Thu, 18 Dec 2014 18:48:22 -0800
Subject: [PATCH] Defaulting to utf-8 when chardet returns None

On articles like this one chardet returns None:
http://news.zing.vn/nhip-song-tre/thay-giao-gay-sot-tung-bo-luat-tinh-yeu/a291427.html
This causes exceptions later on when encoding.lower() is called
---
 readability/encoding.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/readability/encoding.py b/readability/encoding.py
index a72c34d..fb4761d 100644
--- a/readability/encoding.py
+++ b/readability/encoding.py
@@ -26,7 +26,7 @@ def get_encoding(page):
     if not text.strip() or len(text) < 10:
         return enc # can't guess
     res = chardet.detect(text)
-    enc = res['encoding']
+    enc = res['encoding'] or 'utf-8'
     #print '->', enc, "%.2f" % res['confidence']
     enc = custom_decode(enc)
     return enc
@@ -45,4 +45,4 @@ def custom_decode(encoding):
     if encoding in alternates:
         return alternates[encoding]
     else:
-        return encoding
\ No newline at end of file
+        return encoding