From 58c9b47d433b5ef68d5fd8fa510e2bd37aff60ba Mon Sep 17 00:00:00 2001
From: Nick Sweeting <git@nicksweeting.com>
Date: Mon, 25 Mar 2019 16:27:50 -0400
Subject: [PATCH] fix rss parsing when items have newlines between them

---
 archivebox/parse.py | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/archivebox/parse.py b/archivebox/parse.py
index ce6b0358..baaa447e 100644
--- a/archivebox/parse.py
+++ b/archivebox/parse.py
@@ -154,7 +154,8 @@ def parse_rss_export(rss_file):
     """Parse RSS XML-format files into links"""
 
     rss_file.seek(0)
-    items = rss_file.read().split('</item>\n<item>')
+    items = rss_file.read().split('<item>')
+    items = items[1:] if items else []
     for item in items:
         # example item:
         # <item>
@@ -166,7 +167,7 @@ def parse_rss_export(rss_file):
         # </item>
 
         trailing_removed = item.split('</item>', 1)[0]
-        leading_removed = trailing_removed.split('<item>', 1)[-1]
+        leading_removed = trailing_removed.split('<item>', 1)[-1].strip()
         rows = leading_removed.split('\n')
 
         def get_row(key):