From f1ae0a40b774ad7260e7449986764c9104553e03 Mon Sep 17 00:00:00 2001
From: Jan Szejko <janek37@gmail.com>
Date: Wed, 25 Jan 2017 17:16:14 +0100
Subject: [PATCH] accept missing dc tags
---
librarian/pypdf.py | 7 +++++--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/librarian/pypdf.py b/librarian/pypdf.py
index aa4dc1d..920f1b9 100644
--- a/librarian/pypdf.py
+++ b/librarian/pypdf.py
@@ -83,7 +83,7 @@ class EduModule(Xmill):
def get_dc(self, element, dc_field, single=False):
values = map(lambda t: t.text, element.xpath("//dc:%s" % dc_field, namespaces={'dc': DCNS.uri}))
if single:
- return values[0]
+ return values[0] if len(values) else ''
return values
def handle_rdf__RDF(self, _):
@@ -114,7 +114,10 @@ class EduModule(Xmill):
@escape(True)
def get_description(self, element):
- return self.get_dc(element, 'description', single=True)
+ desc = self.get_dc(element, 'description', single=True)
+ if not desc:
+ print '!! no descripton'
+ return desc
def handle_utwor(self, element):
lines = [
--
2.20.1