From 524e5f30921731421b4da85d4043034c5cdcb2ac Mon Sep 17 00:00:00 2001
From: Sandro La Bruzzo <sandro.labruzzo@isti.cnr.it>
Date: Mon, 19 Apr 2021 15:17:25 +0200
Subject: [PATCH] Improved parallelization on transformation wf on hadoop

---
 .../dhp/transformation/TransformSparkJobNode.java      | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/transformation/TransformSparkJobNode.java b/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/transformation/TransformSparkJobNode.java
index 8eafaadca..c7ba700a0 100644
--- a/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/transformation/TransformSparkJobNode.java
+++ b/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/transformation/TransformSparkJobNode.java
@@ -149,4 +149,14 @@ public class TransformSparkJobNode {
 		}
 	}
 
+	/**
+	 * Calculates the number of partitions allocating at most @rpt records for a single transformation task.
+	 * @param totalInput
+	 * @param rpt
+	 * @return
+	 */
+	private static int getRepartitionNumber(long totalInput, Integer rpt) {
+		return (int) (totalInput / rpt);
+	}
+
 }