diff --git a/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkCreateMergeRels.java b/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkCreateMergeRels.java index 70457fb4a0..6d625cd117 100644 --- a/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkCreateMergeRels.java +++ b/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkCreateMergeRels.java @@ -78,9 +78,9 @@ public class SparkCreateMergeRels extends AbstractSparkAction { final String isLookUpUrl = parser.get("isLookUpUrl"); final String actionSetId = parser.get("actionSetId"); int cut = Optional - .ofNullable(parser.get("cutConnectedComponent")) - .map(Integer::valueOf) - .orElse(0); + .ofNullable(parser.get("cutConnectedComponent")) + .map(Integer::valueOf) + .orElse(0); log.info("connected component cut: '{}'", cut); log.info("graphBasePath: '{}'", graphBasePath); log.info("isLookUpUrl: '{}'", isLookUpUrl); diff --git a/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkCreateSimRels.java b/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkCreateSimRels.java index a66ab431c7..3beb90e0b3 100644 --- a/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkCreateSimRels.java +++ b/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkCreateSimRels.java @@ -65,9 +65,9 @@ public class SparkCreateSimRels extends AbstractSparkAction { final String actionSetId = parser.get("actionSetId"); final String workingPath = parser.get("workingPath"); final int numPartitions = Optional - .ofNullable(parser.get("numPartitions")) - .map(Integer::valueOf) - .orElse(NUM_PARTITIONS); + .ofNullable(parser.get("numPartitions")) + .map(Integer::valueOf) + .orElse(NUM_PARTITIONS); log.info("numPartitions: '{}'", numPartitions); log.info("graphBasePath: '{}'", graphBasePath); diff --git a/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkPropagateRelation.java b/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkPropagateRelation.java index 1073adbead..03e6674e4d 100644 --- a/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkPropagateRelation.java +++ b/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkPropagateRelation.java @@ -3,7 +3,6 @@ package eu.dnetlib.dhp.oa.dedup; import static org.apache.spark.sql.functions.col; -import com.google.common.base.Joiner; import org.apache.commons.io.IOUtils; import org.apache.commons.lang3.StringUtils; import org.apache.spark.SparkConf; @@ -13,6 +12,8 @@ import org.apache.spark.sql.*; import org.slf4j.Logger; import org.slf4j.LoggerFactory; +import com.google.common.base.Joiner; + import eu.dnetlib.dhp.application.ArgumentApplicationParser; import eu.dnetlib.dhp.schema.common.ModelSupport; import eu.dnetlib.dhp.schema.oaf.DataInfo; @@ -99,17 +100,21 @@ public class SparkPropagateRelation extends AbstractSparkAction { getDeletedFn()); save( - distinctRelations(newRels - .union(updated) - .union(mergeRels) - .map((MapFunction) r -> r, Encoders.kryo(Relation.class))), + distinctRelations( + newRels + .union(updated) + .union(mergeRels) + .map((MapFunction) r -> r, Encoders.kryo(Relation.class))), outputRelationPath, SaveMode.Overwrite); } private Dataset distinctRelations(Dataset rels) { return rels .filter(getRelationFilterFunction()) - .groupByKey((MapFunction) r -> String.join(r.getSource(), r.getTarget(), r.getRelType(), r.getSubRelType(), r.getRelClass()), Encoders.STRING()) + .groupByKey( + (MapFunction) r -> String + .join(r.getSource(), r.getTarget(), r.getRelType(), r.getSubRelType(), r.getRelClass()), + Encoders.STRING()) .agg(new RelationAggregator().toColumn()) .map((MapFunction, Relation>) t -> t._2(), Encoders.bean(Relation.class)); } diff --git a/dhp-workflows/dhp-dedup-openaire/src/test/java/eu/dnetlib/dhp/oa/dedup/SparkDedupTest.java b/dhp-workflows/dhp-dedup-openaire/src/test/java/eu/dnetlib/dhp/oa/dedup/SparkDedupTest.java index 91e45447ca..fb5ebc0993 100644 --- a/dhp-workflows/dhp-dedup-openaire/src/test/java/eu/dnetlib/dhp/oa/dedup/SparkDedupTest.java +++ b/dhp-workflows/dhp-dedup-openaire/src/test/java/eu/dnetlib/dhp/oa/dedup/SparkDedupTest.java @@ -292,7 +292,8 @@ public class SparkDedupTest implements Serializable { FileUtils.deleteDirectory(new File(testOutputBasePath + "/" + testActionSetId + "/publication_mergerel")); FileUtils.deleteDirectory(new File(testOutputBasePath + "/" + testActionSetId + "/software_mergerel")); FileUtils.deleteDirectory(new File(testOutputBasePath + "/" + testActionSetId + "/dataset_mergerel")); - FileUtils.deleteDirectory(new File(testOutputBasePath + "/" + testActionSetId + "/otherresearchproduct_mergerel")); + FileUtils + .deleteDirectory(new File(testOutputBasePath + "/" + testActionSetId + "/otherresearchproduct_mergerel")); } @Test