From e0d1e381bf536d1c077546526c21eab909444193 Mon Sep 17 00:00:00 2001
From: Bryan Newbold <bnewbold@archive.org>
Date: Sun, 15 Apr 2018 05:44:59 +0000
Subject: NLineInputFormat requires RawProtocol

Should make this a command line argument or something. Want one in
HADOOP, the other for local/tests/inline/etc.
---
 mapreduce/extraction_cdx_grobid.py | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/mapreduce/extraction_cdx_grobid.py b/mapreduce/extraction_cdx_grobid.py
index 6690f49..e23950c 100755
--- a/mapreduce/extraction_cdx_grobid.py
+++ b/mapreduce/extraction_cdx_grobid.py
@@ -40,7 +40,8 @@ sentry_client = raven.Client()
 class MRExtractCdxGrobid(MRJob):
 
     # CDX lines in; JSON status out
-    HADOOP_INPUT_FORMAT = 'org.apache.hadoop.mapred.lib.NLineInputFormat'
+    #HADOOP_INPUT_FORMAT = 'org.apache.hadoop.mapred.lib.NLineInputFormat'
+    #INPUT_PROTOCOL = mrjob.protocol.RawProtocol
     INPUT_PROTOCOL = mrjob.protocol.RawValueProtocol
     OUTPUT_PROTOCOL = mrjob.protocol.JSONValueProtocol
 
-- 
cgit v1.2.3