From 3ba74649f6f457759eff62147c8bc700e21ad198 Mon Sep 17 00:00:00 2001
From: pictuga <contact@pictuga.com>
Date: Tue, 10 Sep 2013 15:25:55 +0200
Subject: [PATCH] Test if linked pages are text documents

Useful for feeds such as HackerNews
---
 morss.py | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/morss.py b/morss.py
index 1f48c4b..d8e8739 100644
--- a/morss.py
+++ b/morss.py
@@ -351,6 +351,11 @@ def Fill(item, cache, feedurl='/', fast=False):
 		cache.set(item.link, 'error-http')
 		return True
 
+	if con.info().maintype != 'text':
+		log('non-text page')
+		cache.set(item.link, 'error-type')
+		return True
+
 	out = readability.Document(data, url=con.url).summary(True)
 
 	if countWord(out) > max(count_content, count_desc) > 0: