97 lines
3.1 KiB
Java
97 lines
3.1 KiB
Java
|
|
package eu.dnetlib.dhp.oa.graph.dump.eosc;
|
|
|
|
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
|
|
|
|
import java.io.Serializable;
|
|
import java.util.ArrayList;
|
|
import java.util.Optional;
|
|
|
|
import javax.rmi.CORBA.Util;
|
|
|
|
import org.apache.commons.io.IOUtils;
|
|
import org.apache.spark.SparkConf;
|
|
import org.apache.spark.api.java.function.MapFunction;
|
|
import org.apache.spark.api.java.function.MapGroupsFunction;
|
|
import org.apache.spark.sql.*;
|
|
import org.slf4j.Logger;
|
|
import org.slf4j.LoggerFactory;
|
|
|
|
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
|
|
import eu.dnetlib.dhp.eosc.model.Relation;
|
|
import eu.dnetlib.dhp.eosc.model.Result;
|
|
import scala.Tuple2;
|
|
|
|
/**
|
|
* @author miriam.baglioni
|
|
* @Date 02/02/23
|
|
*/
|
|
public class SparkExtendResultWithRelation implements Serializable {
|
|
|
|
private static final Logger log = LoggerFactory.getLogger(SparkExtendResultWithRelation.class);
|
|
|
|
public static void main(String[] args) throws Exception {
|
|
String jsonConfiguration = IOUtils
|
|
.toString(
|
|
SparkExtendResultWithRelation.class
|
|
.getResourceAsStream(
|
|
"/eu/dnetlib/dhp/oa/graph/dump/eosc_input_extendwithrelation_parameters.json"));
|
|
|
|
final ArgumentApplicationParser parser = new ArgumentApplicationParser(jsonConfiguration);
|
|
parser.parseArgument(args);
|
|
|
|
Boolean isSparkSessionManaged = Optional
|
|
.ofNullable(parser.get("isSparkSessionManaged"))
|
|
.map(Boolean::valueOf)
|
|
.orElse(Boolean.TRUE);
|
|
log.info("isSparkSessionManaged: {}", isSparkSessionManaged);
|
|
|
|
final String resultPath = parser.get("resultPath");
|
|
log.info("resultPath: {}", resultPath);
|
|
|
|
final String relationPath = parser.get("relationPath");
|
|
log.info("relationPath: {}", relationPath);
|
|
|
|
final String outputPath = parser.get("outputPath");
|
|
log.info("outputPath: {}", outputPath);
|
|
|
|
SparkConf conf = new SparkConf();
|
|
|
|
runWithSparkSession(
|
|
conf,
|
|
isSparkSessionManaged,
|
|
spark -> {
|
|
Utils.removeOutputDir(spark, outputPath);
|
|
extendResultWithRelation(spark, resultPath, relationPath, outputPath);
|
|
|
|
});
|
|
|
|
}
|
|
|
|
private static void extendResultWithRelation(SparkSession spark, String resultPath, String relationPath,
|
|
String outputPath) {
|
|
Dataset<Relation> relationDataset = Utils.readPath(spark, relationPath, Relation.class);
|
|
Dataset<Result> resultDataset = Utils.readPath(spark, resultPath, Result.class);
|
|
|
|
resultDataset
|
|
.joinWith(relationDataset, resultDataset.col("id").equalTo(relationDataset.col("source")), "left")
|
|
.groupByKey((MapFunction<Tuple2<Result, Relation>, String>) t2 -> t2._1().getId(), Encoders.STRING())
|
|
.mapGroups((MapGroupsFunction<String, Tuple2<Result, Relation>, Result>) (k, it) -> {
|
|
Tuple2<Result, Relation> first = it.next();
|
|
Result r = first._1();
|
|
if (Optional.ofNullable(first._2()).isPresent()) {
|
|
if (r.getRelations() == null) {
|
|
r.setRelations(new ArrayList<>());
|
|
}
|
|
r.getRelations().add(first._2());
|
|
it.forEachRemaining(t2 -> r.getRelations().add(t2._2()));
|
|
}
|
|
return r;
|
|
}, Encoders.bean(Result.class))
|
|
.write()
|
|
.option("compression", "gzip")
|
|
.mode(SaveMode.Overwrite)
|
|
.json(outputPath);
|
|
}
|
|
}
|