From 0fbec69b82f2f730b1f1470b2f04a252f46255e5 Mon Sep 17 00:00:00 2001
From: Claudio Atzori <claudio.atzori@isti.cnr.it>
Date: Mon, 30 Mar 2020 19:48:41 +0200
Subject: [PATCH] use oozie prepare statement to cleanup working directories

---
 .../dnetlib/dhp/oa/provision/SparkXmlRecordBuilderJob.java   | 5 -----
 .../eu/dnetlib/dhp/oa/provision/oozie_app/workflow.xml       | 4 ++++
 2 files changed, 4 insertions(+), 5 deletions(-)

diff --git a/dhp-workflows/dhp-graph-provision/src/main/java/eu/dnetlib/dhp/oa/provision/SparkXmlRecordBuilderJob.java b/dhp-workflows/dhp-graph-provision/src/main/java/eu/dnetlib/dhp/oa/provision/SparkXmlRecordBuilderJob.java
index a84cda53a..0a898c0fc 100644
--- a/dhp-workflows/dhp-graph-provision/src/main/java/eu/dnetlib/dhp/oa/provision/SparkXmlRecordBuilderJob.java
+++ b/dhp-workflows/dhp-graph-provision/src/main/java/eu/dnetlib/dhp/oa/provision/SparkXmlRecordBuilderJob.java
@@ -29,14 +29,9 @@ public class SparkXmlRecordBuilderJob {
             final String otherDsTypeId = parser.get("otherDsTypeId");
 
             final FileSystem fs = FileSystem.get(spark.sparkContext().hadoopConfiguration());
-            if (fs.exists(new Path(outputPath))) {
-                fs.delete(new Path(outputPath), true);
-                fs.mkdirs(new Path(outputPath));
-            }
 
             new GraphJoiner(spark, ContextMapper.fromIS(isLookupUrl), otherDsTypeId, inputPath, outputPath)
                     .adjacencyLists();
-                    //.asXML();
         }
     }
 
diff --git a/dhp-workflows/dhp-graph-provision/src/main/resources/eu/dnetlib/dhp/oa/provision/oozie_app/workflow.xml b/dhp-workflows/dhp-graph-provision/src/main/resources/eu/dnetlib/dhp/oa/provision/oozie_app/workflow.xml
index 2eeab2906..a28174cce 100644
--- a/dhp-workflows/dhp-graph-provision/src/main/resources/eu/dnetlib/dhp/oa/provision/oozie_app/workflow.xml
+++ b/dhp-workflows/dhp-graph-provision/src/main/resources/eu/dnetlib/dhp/oa/provision/oozie_app/workflow.xml
@@ -58,6 +58,10 @@
 
     <action name="adjancency_lists">
         <spark xmlns="uri:oozie:spark-action:0.2">
+            <prepare>
+                <delete path="${outputPath}"/>
+                <mkdir path="${outputPath}"/>
+            </prepare>
             <master>yarn</master>
             <mode>cluster</mode>
             <name>build_adjacency_lists</name>