From 51fd5cfb5a37df4c3f1bdcf77f37420b4b965601 Mon Sep 17 00:00:00 2001
From: Bryan Newbold <bnewbold@archive.org>
Date: Fri, 26 Jun 2020 09:56:29 -0700
Subject: handle None 'metadata' field correctly

---
 python/sandcrawler/pdfextract.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'python')

diff --git a/python/sandcrawler/pdfextract.py b/python/sandcrawler/pdfextract.py
index 3701172..5f9b898 100644
--- a/python/sandcrawler/pdfextract.py
+++ b/python/sandcrawler/pdfextract.py
@@ -78,7 +78,7 @@ class PdfExtractResult:
             return PdfExtractResult(
                 sha1hex=record['sha1hex'],
                 status=record['status'],
-                error_msg=record.get('metadata', {}).get('error_msg'),
+                error_msg=(record.get('metadata') or {}).get('error_msg'),
             )
         else:
             pdf_extra = dict()
-- 
cgit v1.2.3