utils/web.py: Attempt to fix handling of RSS feeds using broken entit…

…y references.
tz18 · Nov 26, 2013 · 42b8a06 · 42b8a06
1 parent ff5d83e
commit 42b8a06
Showing 1 changed file with 8 additions and 1 deletion.
diff --git a/src/utils/web.py b/src/utils/web.py
@@ -199,7 +199,14 @@ def handle_data(self, data):
         self.data.append(data)
 
     def handle_entityref(self, data):
-        self.data.append(unichr(htmlentitydefs.name2codepoint[data]))
+        if data in htmlentitydefs.name2codepoint:
+            self.data.append(unichr(htmlentitydefs.name2codepoint[data]))
+        elif sys.version_info[0] >= 3 and isinstance(data, bytes):
+            self.data.append(data.decode())
+        elif sys.version_info[0] < 3 and isinstance(data, str):
+            self.data.append(data.decode('utf8', errors='replace'))
+        else:
+            self.data.append(data)
 
     def getText(self):
         text = ''.join(self.data).strip()