From d442e25cbcc34f9d78252a2f71763c9da1e65e96 Mon Sep 17 00:00:00 2001 From: miconis Date: Mon, 12 Apr 2021 15:56:22 +0200 Subject: [PATCH] bug fix: ids in self mergerels are not marked deletedbyinference=true --- .../src/main/java/eu/dnetlib/dhp/oa/dedup/SparkUpdateEntity.java | 1 + 1 file changed, 1 insertion(+) diff --git a/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkUpdateEntity.java b/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkUpdateEntity.java index cda4137ba..03709c8fe 100644 --- a/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkUpdateEntity.java +++ b/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkUpdateEntity.java @@ -91,6 +91,7 @@ public class SparkUpdateEntity extends AbstractSparkAction { final JavaPairRDD mergedIds = rel .where("relClass == 'merges'") + .where("source != target") .select(rel.col("target")) .distinct() .toJavaRDD()