From aeb01fa353a028650c01ee417368154d9c45c220 Mon Sep 17 00:00:00 2001
From: Claudio Atzori <claudio.atzori@isti.cnr.it>
Date: Tue, 17 Mar 2020 11:57:24 +0100
Subject: [PATCH] reading from newline delimited json textfiles instead of
 sequence files

---
 .../main/java/eu/dnetlib/dhp/graph/GraphJoiner.java    | 10 ++++------
 1 file changed, 4 insertions(+), 6 deletions(-)

diff --git a/dhp-workflows/dhp-graph-provision/src/main/java/eu/dnetlib/dhp/graph/GraphJoiner.java b/dhp-workflows/dhp-graph-provision/src/main/java/eu/dnetlib/dhp/graph/GraphJoiner.java
index 29492cb90..d260e0551 100644
--- a/dhp-workflows/dhp-graph-provision/src/main/java/eu/dnetlib/dhp/graph/GraphJoiner.java
+++ b/dhp-workflows/dhp-graph-provision/src/main/java/eu/dnetlib/dhp/graph/GraphJoiner.java
@@ -226,9 +226,8 @@ public class GraphJoiner implements Serializable {
      * @return the JavaPairRDD<String, TypedRow> indexed by entity identifier
      */
     private JavaPairRDD<String, TypedRow> readPathEntity(final JavaSparkContext sc, final String inputPath, final String type) {
-        return sc.sequenceFile(inputPath + "/" + type, Text.class, Text.class)
-                .mapToPair((PairFunction<Tuple2<Text, Text>, String, TypedRow>) item -> {
-                    final String s = item._2().toString();
+        return sc.textFile(inputPath + "/" + type)
+                .mapToPair((PairFunction<String, String, TypedRow>) s -> {
                     final DocumentContext json = JsonPath.parse(s);
                     final String id = json.read("$.id");
                     return new Tuple2<>(id, new TypedRow()
@@ -247,9 +246,8 @@ public class GraphJoiner implements Serializable {
      * @return the JavaRDD<TypedRow> containing all the relationships
      */
     private JavaRDD<TypedRow> readPathRelation(final JavaSparkContext sc, final String inputPath) {
-        return sc.sequenceFile(inputPath + "/relation", Text.class, Text.class)
-                .map(item -> {
-                    final String s = item._2().toString();
+        return sc.textFile(inputPath + "/relation")
+                .map(s -> {
                     final DocumentContext json = JsonPath.parse(s);
                     return new TypedRow()
                             .setSourceId(json.read("$.source"))