346 lines
12 KiB
Java
346 lines
12 KiB
Java
|
|
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.function.Predicate;
|
|
import java.util.stream.Collectors;
|
|
import java.util.stream.Stream;
|
|
|
|
import org.apache.commons.io.IOUtils;
|
|
import org.apache.spark.SparkConf;
|
|
import org.apache.spark.api.java.function.FilterFunction;
|
|
import org.apache.spark.api.java.function.FlatMapFunction;
|
|
import org.apache.spark.api.java.function.MapFunction;
|
|
import org.apache.spark.sql.*;
|
|
import org.apache.spark.sql.Dataset;
|
|
import org.apache.spark.sql.types.DataType;
|
|
import org.apache.spark.sql.types.DataTypes;
|
|
import org.apache.spark.sql.types.StructType;
|
|
import org.slf4j.Logger;
|
|
import org.slf4j.LoggerFactory;
|
|
|
|
import com.fasterxml.jackson.core.JsonProcessingException;
|
|
import com.fasterxml.jackson.databind.ObjectMapper;
|
|
|
|
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
|
|
import eu.dnetlib.dhp.oa.graph.dump.skgif.EmitFromEntities;
|
|
import eu.dnetlib.dhp.oa.graph.dump.skgif.Utils;
|
|
import eu.dnetlib.dhp.schema.common.EntityType;
|
|
import eu.dnetlib.dhp.schema.common.ModelSupport;
|
|
import eu.dnetlib.dhp.schema.oaf.*;
|
|
import scala.Tuple2;
|
|
|
|
/**
|
|
* @author miriam.baglioni
|
|
* @Date 12/03/24
|
|
*/
|
|
public class SelectConnectedEntities implements Serializable {
|
|
private static final Logger log = LoggerFactory.getLogger(EmitFromEntities.class);
|
|
|
|
public static void main(String[] args) throws Exception {
|
|
String jsonConfiguration = IOUtils
|
|
.toString(
|
|
SelectConnectedEntities.class
|
|
.getResourceAsStream(
|
|
"/eu/dnetlib/dhp/oa/graph/dump/select_connected_entities_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 filterPath = parser.get("filterPath");
|
|
log.info("filterPath: {}", filterPath);
|
|
final String workingDir = parser.get("workingDir");
|
|
log.info("workingDir: {}", workingDir);
|
|
SparkConf conf = new SparkConf();
|
|
|
|
runWithSparkSession(
|
|
conf,
|
|
isSparkSessionManaged,
|
|
spark -> {
|
|
|
|
selectConnectedEntities2(spark, inputPath, filterPath, workingDir);
|
|
|
|
});
|
|
}
|
|
|
|
private static void selectConnectedEntities2(SparkSession spark, String inputPath, String filterPath,
|
|
String workingDir) {
|
|
final StructType rp = new StructType()
|
|
.add(
|
|
"dataInfo", new StructType()
|
|
.add("deletedbyinference", DataTypes.BooleanType))
|
|
.add("source", DataTypes.StringType)
|
|
.add("target", DataTypes.StringType);
|
|
|
|
Dataset<String> resultIds = spark.emptyDataset(Encoders.STRING());
|
|
for (EntityType entity : ModelSupport.entityTypes.keySet())
|
|
if (ModelSupport.isResult(entity))
|
|
resultIds = resultIds
|
|
.union(
|
|
spark
|
|
.read()
|
|
.parquet(filterPath + entity.name() + "_ids")
|
|
.select("id")
|
|
.as(Encoders.STRING()));
|
|
|
|
Dataset<Row> relation = spark
|
|
.read()
|
|
.schema(rp)
|
|
.json(inputPath + "relation")
|
|
.filter(("datainfo.deletedbyinference != true"))
|
|
.drop("datainfo");
|
|
|
|
Dataset<Row> matchingRels = relation
|
|
.join(
|
|
resultIds, relation
|
|
.col("source")
|
|
.equalTo(resultIds.col("value")),
|
|
"leftsemi")
|
|
.select("target")
|
|
.distinct();
|
|
|
|
Dataset<Row> organization = spark
|
|
.read()
|
|
.schema(Encoders.bean(Organization.class).schema())
|
|
.json(inputPath + "organization")
|
|
.filter("datainfo.deletedbyinference != true ");
|
|
|
|
Dataset<Project> projects = Utils
|
|
.readPath(spark, inputPath + "project", Project.class)
|
|
.filter((FilterFunction<Project>) p -> !p.getDataInfo().getDeletedbyinference())
|
|
.filter(
|
|
(FilterFunction<Project>) p -> Optional.ofNullable(p.getFundingtree()).isPresent() &&
|
|
p.getFundingtree().size() > 0 &&
|
|
Utils
|
|
.getFunderName(p.getFundingtree().get(0).getValue())
|
|
.equalsIgnoreCase("European Commission"));
|
|
|
|
organization
|
|
.join(matchingRels, organization.col("id").equalTo(matchingRels.col("target")), "leftsemi")
|
|
.write()
|
|
.mode(SaveMode.Overwrite)
|
|
.option("compression", "gzip")
|
|
.json(workingDir + "organization");
|
|
|
|
projects
|
|
.join(matchingRels, projects.col("id").equalTo(matchingRels.col("target")), "leftsemi")
|
|
.write()
|
|
.mode(SaveMode.Overwrite)
|
|
.option("compression", "gzip")
|
|
.json(workingDir + "/project");
|
|
|
|
Dataset<Row> datasources = spark
|
|
.read()
|
|
.schema(Encoders.bean(Datasource.class).schema())
|
|
.json(inputPath + "datasource")
|
|
.filter("datainfo.deletedbyinference != true");
|
|
final Dataset<String> datasourceReferencedIds = getDatasourceReferenceIdDataset(spark, workingDir);
|
|
datasources
|
|
.join(
|
|
datasourceReferencedIds, datasourceReferencedIds.col("value").equalTo(datasources.col("id")),
|
|
"left_semi")
|
|
.write()
|
|
.mode(SaveMode.Overwrite)
|
|
.option("compression", "gzip")
|
|
.json(workingDir + "datasource");
|
|
|
|
}
|
|
|
|
// private static <R extends Result> void selectConnectedEntities(SparkSession spark, String inputPath,
|
|
// String filterPath,
|
|
// String workingDir) throws JsonProcessingException {
|
|
//
|
|
// Dataset<String> resultIds = spark.emptyDataset(Encoders.STRING());
|
|
// for (EntityType entity : ModelSupport.entityTypes.keySet())
|
|
// if (ModelSupport.isResult(entity))
|
|
// resultIds = resultIds
|
|
// .union(
|
|
// spark
|
|
// .read()
|
|
// .parquet(filterPath + entity.name() + "_ids")
|
|
// .select("id")
|
|
// .as(Encoders.STRING()));
|
|
//
|
|
// Dataset<Relation> relation = Utils
|
|
// .readPath(spark, inputPath + "relation", Relation.class)
|
|
// .filter((FilterFunction<Relation>) r -> !r.getDataInfo().getDeletedbyinference());
|
|
// Dataset<Organization> organizations = Utils
|
|
// .readPath(spark, inputPath + "organization", Organization.class)
|
|
// .filter((FilterFunction<Organization>) o -> !o.getDataInfo().getDeletedbyinference());
|
|
// Dataset<Project> projects = Utils
|
|
// .readPath(spark, inputPath + "project", Project.class)
|
|
// .filter((FilterFunction<Project>) p -> !p.getDataInfo().getDeletedbyinference())
|
|
// .filter(
|
|
// (FilterFunction<Project>) p -> Optional.ofNullable(p.getFundingtree()).isPresent() &&
|
|
// p.getFundingtree().size() > 0 &&
|
|
// Utils
|
|
// .getFunderName(p.getFundingtree().get(0).getValue())
|
|
// .equalsIgnoreCase("European Commission"));
|
|
//
|
|
// Dataset<Datasource> datasources = Utils
|
|
// .readPath(spark, inputPath + "datasource", Datasource.class)
|
|
// .filter((FilterFunction<Datasource>) d -> !d.getDataInfo().getDeletedbyinference());
|
|
//
|
|
// // select relations having source in the set of identifiers selected for eosc
|
|
// Dataset<Row> relationSource = relation
|
|
// .join(resultIds, resultIds.col("value").equalTo(relation.col("source")), "left_semi");
|
|
// relationSource
|
|
// .join(resultIds, resultIds.col("value").equalTo(relation.col("target")), "left_semi")
|
|
// .write()
|
|
// .option("compression", "gzip")
|
|
// .mode(SaveMode.Overwrite)
|
|
// .json(workingDir + "resultrelation");
|
|
////
|
|
//// // write relations between results and organizations
|
|
// relationSource
|
|
// .joinWith(organizations, relation.col("target").equalTo(organizations.col("id")), "left_semi")
|
|
// .write()
|
|
// .mode(SaveMode.Overwrite)
|
|
// .option("compression", "gzip")
|
|
// .json(workingDir + "organizaitonrelation");
|
|
//
|
|
// relationSource
|
|
// .joinWith(projects, relation.col("target").equalTo(projects.col("id")), "left_semi")
|
|
// .write()
|
|
// .mode(SaveMode.Overwrite)
|
|
// .option("compression", "gzip")
|
|
// .json(workingDir + "projectrelation");
|
|
//
|
|
// // write organizations linked to results in the set
|
|
//
|
|
// organizations
|
|
// .join(relationSource, relationSource.col("target").equalTo(organizations.col("id")), "left_semi")
|
|
// .write()
|
|
// .mode(SaveMode.Overwrite)
|
|
// .option("compression", "gzip")
|
|
// .json(workingDir + "organization");
|
|
//
|
|
// // write projects linked to results in the set
|
|
// projects
|
|
// .join(relationSource, relationSource.col("target").equalTo(projects.col("id")))
|
|
// .write()
|
|
// .mode(SaveMode.Overwrite)
|
|
// .option("compression", "gzip")
|
|
// .json(workingDir + "project");
|
|
//
|
|
// // read the results and select all the distinct instance.hostedbykey
|
|
// final Dataset<String> datasourceReferencedIds = getDatasourceReferenceIdDataset(spark, workingDir);
|
|
// // join with the datasources and write the datasource in the join
|
|
// datasources
|
|
// .joinWith(
|
|
// datasourceReferencedIds, datasourceReferencedIds.col("value").equalTo(datasources.col("id")),
|
|
// "left_semi")
|
|
// .write()
|
|
// .mode(SaveMode.Overwrite)
|
|
// .option("compression", "gzip")
|
|
// .json(workingDir + "datasource");
|
|
//
|
|
// // selecting relations between organizations and projects in the selected set
|
|
// StructType tp = StructType.fromDDL("`id` STRING");
|
|
// Dataset<Row> organizationSbs = spark
|
|
// .read()
|
|
// .schema(tp)
|
|
// .json(workingDir + "organization")
|
|
// .select("id");
|
|
//
|
|
// Dataset<Row> projectSbs = spark
|
|
// .read()
|
|
// .schema(tp)
|
|
// .json(workingDir + "project")
|
|
// .select("id");
|
|
////
|
|
// Dataset<Row> tmpRel;
|
|
// tmpRel = relation
|
|
// .join(
|
|
// organizationSbs, organizationSbs
|
|
// .col("id")
|
|
// .equalTo(relation.col("source")),
|
|
// "left_semi");
|
|
// tmpRel
|
|
// .join(projectSbs, tmpRel.col("target").equalTo(projectSbs.col("id")), "left_semi")
|
|
// .write()
|
|
// .mode(SaveMode.Overwrite)
|
|
// .option("compression", "gzip")
|
|
// .json(workingDir + "orgprojelation");
|
|
//
|
|
// // selecting relations between datasources and organizations in the selected set
|
|
// Dataset<Row> datasourceSbs = spark
|
|
// .read()
|
|
// .schema(tp)
|
|
// .json(workingDir + "datasource")
|
|
// .select("id");
|
|
//
|
|
// tmpRel = relation
|
|
// .join(datasourceSbs, datasourceSbs.col("id").equalTo(relation.col("source")), "left_semi");
|
|
// tmpRel
|
|
// .join(organizationSbs, tmpRel.col("target").equalTo(organizationSbs.col("id")), "left_semi")
|
|
// .write()
|
|
// .mode(SaveMode.Overwrite)
|
|
// .option("compression", "gzip")
|
|
// .json(workingDir + "datsorgrelation");
|
|
//
|
|
// Utils
|
|
// .readPath(spark, workingDir + "resultrelation", Relation.class)
|
|
// .union(Utils.readPath(spark, workingDir + "organizaitonrelation", Relation.class))
|
|
// .union(Utils.readPath(spark, workingDir + "projectrelation", Relation.class))
|
|
// .union(Utils.readPath(spark, workingDir + "orgprojelation", Relation.class))
|
|
// .union(Utils.readPath(spark, workingDir + "datsorgrelation", Relation.class))
|
|
// .write()
|
|
// .mode(SaveMode.Overwrite)
|
|
// .option("compression", "gzip")
|
|
// .json(workingDir + "relation");
|
|
//
|
|
// }
|
|
|
|
private static <R extends Result> Dataset<String> getDatasourceReferenceIdDataset(SparkSession spark,
|
|
String workingDir) {
|
|
Dataset<String> datasourceReferencedIds = spark.emptyDataset(Encoders.STRING());
|
|
|
|
for (EntityType entity : ModelSupport.entityTypes.keySet())
|
|
if (ModelSupport.isResult(entity)) {
|
|
Class<R> resultClazz = ModelSupport.entityTypes.get(entity);
|
|
datasourceReferencedIds = datasourceReferencedIds
|
|
.union(
|
|
Utils
|
|
.readPath(spark, workingDir + entity.name(), resultClazz)
|
|
.flatMap(
|
|
(FlatMapFunction<R, String>) r -> r
|
|
.getInstance()
|
|
.stream()
|
|
.filter(i -> i.getHostedby() != null && i.getHostedby().getKey() != null)
|
|
.map(i -> i.getHostedby().getKey())
|
|
.collect(Collectors.toList())
|
|
.iterator(),
|
|
Encoders.STRING()));
|
|
datasourceReferencedIds = datasourceReferencedIds
|
|
.union(
|
|
Utils
|
|
.readPath(spark, workingDir + entity.name(), resultClazz)
|
|
.flatMap(
|
|
(FlatMapFunction<R, String>) r -> r
|
|
.getInstance()
|
|
.stream()
|
|
.filter(i -> i.getCollectedfrom() != null && i.getCollectedfrom().getKey() != null)
|
|
.map(i -> i.getCollectedfrom().getKey())
|
|
.collect(Collectors.toList())
|
|
.iterator(),
|
|
Encoders.STRING()));
|
|
}
|
|
datasourceReferencedIds = datasourceReferencedIds.distinct();
|
|
return datasourceReferencedIds;
|
|
}
|
|
}
|