100 lines
3.0 KiB
Java
100 lines
3.0 KiB
Java
|
|
package eu.dnetlib.dhp.skgif;
|
|
|
|
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
|
|
|
|
import java.io.Serializable;
|
|
import java.util.*;
|
|
|
|
import org.apache.commons.io.IOUtils;
|
|
import org.apache.spark.SparkConf;
|
|
import org.apache.spark.sql.*;
|
|
import org.apache.spark.sql.types.StructType;
|
|
import org.slf4j.Logger;
|
|
import org.slf4j.LoggerFactory;
|
|
|
|
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
|
|
import eu.dnetlib.dhp.oa.graph.dump.Utils;
|
|
import eu.dnetlib.dhp.schema.oaf.Relation;
|
|
|
|
/**
|
|
* @author miriam.baglioni
|
|
* @Date 04/09/23
|
|
*/
|
|
public class PrepareResultRelation implements Serializable {
|
|
private static final Logger log = LoggerFactory.getLogger(PrepareResultRelation.class);
|
|
|
|
public static void main(String[] args) throws Exception {
|
|
String jsonConfiguration = IOUtils
|
|
.toString(
|
|
PrepareResultRelation.class
|
|
.getResourceAsStream(
|
|
"/eu/dnetlib/dhp/oa/graph/dump/result_relation_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 inputPath = parser.get("sourcePath");
|
|
log.info("inputPath: {}", inputPath);
|
|
|
|
final String outputPath = parser.get("outputPath");
|
|
log.info("outputPath: {}", outputPath);
|
|
|
|
SparkConf conf = new SparkConf();
|
|
|
|
runWithSparkSession(
|
|
conf,
|
|
isSparkSessionManaged,
|
|
spark -> {
|
|
Utils.removeOutputDir(spark, outputPath);
|
|
prepareResultRelationList(spark, inputPath, outputPath);
|
|
});
|
|
}
|
|
|
|
private static void prepareResultRelationList(SparkSession spark, String inputPath, String outputPath) {
|
|
final StructType structureSchema = new StructType()
|
|
.fromDDL("`id` STRING, `dataInfo` STRUCT<`deletedbyinference`:BOOLEAN,`invisible`:BOOLEAN>");
|
|
Dataset<Relation> relation = spark
|
|
.read()
|
|
.json(inputPath)
|
|
.as(Encoders.bean(Relation.class))
|
|
.filter("dataInfo.deletedbyinference != true and dataInfo.invisible != true")
|
|
.filter(
|
|
"relClass == 'hasAuthorInstitution' or relClass == 'outcome' or " +
|
|
"relClass == 'IsSupplementedBy' or relClass == 'IsDocumentedBy' or relClass == 'IsPartOf' " +
|
|
"relClass == IsNewVersionOf");
|
|
|
|
org.apache.spark.sql.Dataset<Row> df = spark.createDataFrame(new ArrayList<Row>(), structureSchema);
|
|
List<String> entities = Arrays
|
|
.asList(
|
|
"publication", "dataset", "otherresearchproduct", "software");
|
|
for (String e : entities)
|
|
df = df
|
|
.union(
|
|
spark
|
|
.read()
|
|
.schema(structureSchema)
|
|
.json(inputPath + "/" + e)
|
|
.filter("dataInfo.deletedbyinference != true and dataInfo.invisible != true"));
|
|
|
|
relation
|
|
.joinWith(df, relation.col("source").equalTo(df.col("id")))
|
|
.select(
|
|
new Column("id").as("resultId"),
|
|
new Column("target"),
|
|
new Column("relClass"))
|
|
.write()
|
|
.mode(SaveMode.Overwrite)
|
|
.option("compression", "gzip")
|
|
.json(outputPath);
|
|
|
|
}
|
|
|
|
}
|