From 30e0f60ac837e6265fd8fbe15edd2d9fc9e30b30 Mon Sep 17 00:00:00 2001 From: "miriam.baglioni" Date: Thu, 14 Apr 2022 08:50:37 +0200 Subject: [PATCH] [Enrichment Step] get rid of hive --- .../PrepareResultOrcidAssociationStep1.java | 10 +--------- .../orcidtoresultfromsemrel/oozie_app/workflow.xml | 2 +- 2 files changed, 2 insertions(+), 10 deletions(-) diff --git a/dhp-workflows/dhp-enrichment/src/main/java/eu/dnetlib/dhp/orcidtoresultfromsemrel/PrepareResultOrcidAssociationStep1.java b/dhp-workflows/dhp-enrichment/src/main/java/eu/dnetlib/dhp/orcidtoresultfromsemrel/PrepareResultOrcidAssociationStep1.java index d17d87fed..3a8ffc4e2 100644 --- a/dhp-workflows/dhp-enrichment/src/main/java/eu/dnetlib/dhp/orcidtoresultfromsemrel/PrepareResultOrcidAssociationStep1.java +++ b/dhp-workflows/dhp-enrichment/src/main/java/eu/dnetlib/dhp/orcidtoresultfromsemrel/PrepareResultOrcidAssociationStep1.java @@ -80,7 +80,7 @@ public class PrepareResultOrcidAssociationStep1 { conf, isSparkSessionManaged, spark -> { - removeOutputDir(spark, outputPath); + // removeOutputDir(spark, outputPath); prepareInfo( spark, inputPath, outputPath, resultType, resultClazz, allowedsemrel, allowedPids); }); @@ -96,14 +96,6 @@ public class PrepareResultOrcidAssociationStep1 { List allowedPids) { final String inputResultPath = inputPath + "/" + resultType; - readPath(spark, inputPath + "/relation", Relation.class) - .filter( - (FilterFunction) r -> !r.getDataInfo().getDeletedbyinference() - && allowedsemrel.contains(r.getRelClass().toLowerCase())) - .write() - .mode(SaveMode.Overwrite) - .option("compression", "gzip") - .json(outputPath + "/" + resultType + "/relationSubset"); Dataset relation = readPath(spark, outputPath + "/relationSubset", Relation.class); diff --git a/dhp-workflows/dhp-enrichment/src/main/resources/eu/dnetlib/dhp/wf/subworkflows/orcidtoresultfromsemrel/oozie_app/workflow.xml b/dhp-workflows/dhp-enrichment/src/main/resources/eu/dnetlib/dhp/wf/subworkflows/orcidtoresultfromsemrel/oozie_app/workflow.xml index 6bdce3cd5..605b9c885 100644 --- a/dhp-workflows/dhp-enrichment/src/main/resources/eu/dnetlib/dhp/wf/subworkflows/orcidtoresultfromsemrel/oozie_app/workflow.xml +++ b/dhp-workflows/dhp-enrichment/src/main/resources/eu/dnetlib/dhp/wf/subworkflows/orcidtoresultfromsemrel/oozie_app/workflow.xml @@ -14,7 +14,7 @@ - + Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]