2021-02-26 10:19:28 +01:00
|
|
|
|
2021-02-10 11:51:50 +01:00
|
|
|
package eu.dnetlib.dhp.oa.dedup;
|
|
|
|
|
2021-02-26 10:19:28 +01:00
|
|
|
import java.io.IOException;
|
|
|
|
|
2021-02-10 11:51:50 +01:00
|
|
|
import org.apache.commons.io.IOUtils;
|
|
|
|
import org.apache.spark.SparkConf;
|
|
|
|
import org.apache.spark.api.java.JavaRDD;
|
2021-08-11 12:13:22 +02:00
|
|
|
import org.apache.spark.sql.Encoders;
|
|
|
|
import org.apache.spark.sql.SaveMode;
|
|
|
|
import org.apache.spark.sql.SparkSession;
|
2021-02-10 11:51:50 +01:00
|
|
|
import org.slf4j.Logger;
|
|
|
|
import org.slf4j.LoggerFactory;
|
|
|
|
|
2021-02-26 10:19:28 +01:00
|
|
|
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
|
|
|
|
import eu.dnetlib.dhp.schema.common.ModelSupport;
|
2021-08-11 12:13:22 +02:00
|
|
|
import eu.dnetlib.dhp.schema.oaf.Relation;
|
2021-02-26 10:19:28 +01:00
|
|
|
import eu.dnetlib.dhp.utils.ISLookupClientFactory;
|
|
|
|
import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpService;
|
2021-02-10 11:51:50 +01:00
|
|
|
|
|
|
|
public class SparkCopyRelationsNoOpenorgs extends AbstractSparkAction {
|
|
|
|
|
2021-03-19 16:57:40 +01:00
|
|
|
private static final Logger log = LoggerFactory.getLogger(SparkCopyRelationsNoOpenorgs.class);
|
2021-02-26 10:19:28 +01:00
|
|
|
|
|
|
|
public SparkCopyRelationsNoOpenorgs(ArgumentApplicationParser parser, SparkSession spark) {
|
|
|
|
super(parser, spark);
|
|
|
|
}
|
|
|
|
|
|
|
|
public static void main(String[] args) throws Exception {
|
|
|
|
ArgumentApplicationParser parser = new ArgumentApplicationParser(
|
|
|
|
IOUtils
|
|
|
|
.toString(
|
|
|
|
SparkCopyRelationsNoOpenorgs.class
|
|
|
|
.getResourceAsStream(
|
|
|
|
"/eu/dnetlib/dhp/oa/dedup/updateEntity_parameters.json")));
|
|
|
|
parser.parseArgument(args);
|
|
|
|
|
|
|
|
SparkConf conf = new SparkConf();
|
|
|
|
conf.set("spark.serializer", "org.apache.spark.serializer.KryoSerializer");
|
|
|
|
conf.registerKryoClasses(ModelSupport.getOafModelClasses());
|
|
|
|
|
2021-03-19 16:57:40 +01:00
|
|
|
new SparkCopyRelationsNoOpenorgs(parser, getSparkSession(conf))
|
2021-02-26 10:19:28 +01:00
|
|
|
.run(ISLookupClientFactory.getLookUpService(parser.get("isLookUpUrl")));
|
|
|
|
}
|
|
|
|
|
|
|
|
public void run(ISLookUpService isLookUpService) throws IOException {
|
|
|
|
|
|
|
|
final String graphBasePath = parser.get("graphBasePath");
|
|
|
|
final String workingPath = parser.get("workingPath");
|
|
|
|
final String dedupGraphPath = parser.get("dedupGraphPath");
|
|
|
|
|
|
|
|
log.info("graphBasePath: '{}'", graphBasePath);
|
|
|
|
log.info("workingPath: '{}'", workingPath);
|
|
|
|
log.info("dedupGraphPath: '{}'", dedupGraphPath);
|
|
|
|
|
|
|
|
final String relationPath = DedupUtility.createEntityPath(graphBasePath, "relation");
|
|
|
|
final String outputPath = DedupUtility.createEntityPath(dedupGraphPath, "relation");
|
|
|
|
|
|
|
|
JavaRDD<Relation> simRels = spark
|
|
|
|
.read()
|
|
|
|
.textFile(relationPath)
|
|
|
|
.map(patchRelFn(), Encoders.bean(Relation.class))
|
|
|
|
.toJavaRDD()
|
2021-03-19 16:57:40 +01:00
|
|
|
.filter(x -> !isOpenorgs(x));
|
|
|
|
|
|
|
|
log.info("Number of non-Openorgs relations collected: {}", simRels.count());
|
2021-02-26 10:19:28 +01:00
|
|
|
|
|
|
|
spark
|
|
|
|
.createDataset(simRels.rdd(), Encoders.bean(Relation.class))
|
|
|
|
.write()
|
|
|
|
.mode(SaveMode.Overwrite)
|
|
|
|
.json(outputPath);
|
|
|
|
|
|
|
|
}
|
|
|
|
|
2021-02-10 11:51:50 +01:00
|
|
|
}
|