From 6c92ee4c0b137c28abd03ed72190210da8a1e72b Mon Sep 17 00:00:00 2001
From: Bryan Newbold <bnewbold@archive.org>
Date: Tue, 21 Aug 2018 16:28:53 -0700
Subject: extraction: do want content, not text

XML can have non-unicode characters? Who knew.
---
 mapreduce/extraction_cdx_grobid.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'mapreduce')

diff --git a/mapreduce/extraction_cdx_grobid.py b/mapreduce/extraction_cdx_grobid.py
index 227a026..040538c 100755
--- a/mapreduce/extraction_cdx_grobid.py
+++ b/mapreduce/extraction_cdx_grobid.py
@@ -163,7 +163,7 @@ class MRExtractCdxGrobid(MRJob):
                 extra=grobid_response.text)
 
         info['grobid0:status'] = {'status': 'partial'}
-        info['grobid0:tei_xml'] = grobid_response.text
+        info['grobid0:tei_xml'] = grobid_response.content
 
         # Convert TEI XML to JSON
         try:
-- 
cgit v1.2.3