D-Net
/
dhp-graph-dump
4
0
Fork 0
Code Issues Pull Requests 1 Projects Releases Wiki Activity

[SKG-IF] fixing issue in deserialization

This commit is contained in:
Miriam Baglioni 2024-03-14 10:11:45 +01:00
parent e8f19ad003
commit b176bbef1d
1 changed files with 11 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
dump/src/main/java/eu/dnetlib/dhp/oa/graph/dump/filterentities/SelectConnectedEntities.java
View File

@ -4,8 +4,10 @@ package eu.dnetlib.dhp.oa.graph.dump.filterentities;
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
import java.io.Serializable;
import java.util.Arrays;
import java.util.Optional;
import java.util.stream.Collectors;
import java.util.stream.Stream;
import org.apache.commons.io.IOUtils;
import org.apache.spark.SparkConf;
@ -105,7 +107,7 @@ public class SelectConnectedEntities implements Serializable {
.joinWith(relation, resultIds.col("id").equalTo(relation.col("source")))
.map((MapFunction<Tuple2<Identifiers, Relation>, Relation>) t2 -> t2._2(), Encoders.bean(Relation.class));
// write relations having sorce and target in the set
// write relations having source and target in the set
resultIds
.joinWith(resultSource, resultIds.col("id").equalTo(resultSource.col("target")))
.map((MapFunction<Tuple2<Identifiers, Relation>, Relation>) t2 -> t2._2(), Encoders.bean(Relation.class))
@ -153,11 +155,11 @@ public class SelectConnectedEntities implements Serializable {
.json(workingDir + "project");
// read the results and select all the distinct instance.hostedbykey
Dataset<String> hostedbyIds = spark.emptyDataset(Encoders.STRING());
Dataset<String> datasourceReferencedIds = spark.emptyDataset(Encoders.STRING());
for (EntityType entity : ModelSupport.entityTypes.keySet())
if (ModelSupport.isResult(entity)) {
Class<R> resultClazz = ModelSupport.entityTypes.get(entity);
hostedbyIds = hostedbyIds
datasourceReferencedIds = datasourceReferencedIds
.union(
Utils
.readPath(spark, workingDir + entity.name(), resultClazz)
@ -165,14 +167,15 @@ public class SelectConnectedEntities implements Serializable {
(FlatMapFunction<R, String>) r -> r
.getInstance()
.stream()
.map(i -> i.getHostedby().getKey())
.flatMap(i -> Stream.of(i.getHostedby().getKey(), i.getCollectedfrom().getKey()))
.collect(Collectors.toList())
.iterator(),
Encoders.STRING()));
}
datasourceReferencedIds = datasourceReferencedIds.distinct();
// join with the datasources and write the datasource in the join
hostedbyIds
.joinWith(datasources, hostedbyIds.col("value").equalTo(datasources.col("id")))
datasourceReferencedIds
.joinWith(datasources, datasourceReferencedIds.col("value").equalTo(datasources.col("id")))
.map((MapFunction<Tuple2<String, Datasource>, Datasource>) t2 -> t2._2(), Encoders.bean(Datasource.class))
.write()
.mode(SaveMode.Overwrite)
@ -196,20 +199,15 @@ public class SelectConnectedEntities implements Serializable {
// selecting relations between datasources and organizations in the selected set
Dataset<Datasource> datasourceSbs = Utils.readPath(spark, workingDir + "datasource", Datasource.class);
Dataset<Relation> dsSourceRels = datasourceSbs
.joinWith(relation, datasourceSbs.col("id").equalTo(relation.col("source")))
.joinWith(relation, datasourceSbs.col("id").as("dsId").equalTo(relation.col("source")))
.map((MapFunction<Tuple2<Datasource, Relation>, Relation>) t2 -> t2._2(), Encoders.bean(Relation.class));
dsSourceRels
.joinWith(organizationSbs, dsSourceRels.col("target").equalTo(organizations.col("id")))
.joinWith(organizationSbs, dsSourceRels.col("target").equalTo(organizations.col("id").as("orgId")))
.map((MapFunction<Tuple2<Relation, Organization>, Relation>) t2 -> t2._1(), Encoders.bean(Relation.class))
.write()
.mode(SaveMode.Append)
.option("compression", "gzip")
.json(workingDir + "relation");
/**
* DATASOURCE_PROVIDED_BY_ORGANIZATION(
* "isProvidedBy"),
*/
}
}