From e03ef53dd5b684b08ff7c948a09a0f56a63cb965 Mon Sep 17 00:00:00 2001
From: Chad Estioco <chadestioco@gmail.com>
Date: Tue, 23 Jul 2019 08:56:13 +0800
Subject: [PATCH] Use urlparse in get_domain.

---
 microdata.py | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/microdata.py b/microdata.py
index c865006..e015796 100755
--- a/microdata.py
+++ b/microdata.py
@@ -4,6 +4,7 @@
 import html5lib
 
 from collections import defaultdict
+from urlparse import urlparse
 
 
 try:
@@ -137,8 +138,9 @@ def get_domain(url_string):
         """
         Get the domain _including_ the protocol specified, if any.
         """
-        if "://" in url_string:
-            return "/".join(url_string.split("/")[0:3])
+        parsed = urlparse(url_string)
+        if parsed.scheme:
+            return "/".join(parsed.scheme, "", parsed.netloc)
         else:
             return url_string.split("/")[0]