2022-08-08 14:10:47 +02:00
|
|
|
|
2022-07-22 16:23:12 +02:00
|
|
|
package eu.dnetlib.dhp.oa.graph.clean.country;
|
|
|
|
|
2022-08-08 14:10:47 +02:00
|
|
|
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
|
|
|
|
|
|
|
|
import java.io.Serializable;
|
|
|
|
import java.util.List;
|
|
|
|
import java.util.Optional;
|
|
|
|
|
2022-07-22 16:23:12 +02:00
|
|
|
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.MapFunction;
|
|
|
|
import org.apache.spark.sql.Dataset;
|
|
|
|
import org.apache.spark.sql.Encoders;
|
|
|
|
import org.apache.spark.sql.SaveMode;
|
|
|
|
import org.apache.spark.sql.SparkSession;
|
|
|
|
import org.slf4j.Logger;
|
|
|
|
import org.slf4j.LoggerFactory;
|
|
|
|
|
2022-08-08 14:10:47 +02:00
|
|
|
import com.fasterxml.jackson.databind.ObjectMapper;
|
2022-07-22 16:23:12 +02:00
|
|
|
|
2022-08-08 14:10:47 +02:00
|
|
|
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
|
|
|
|
import eu.dnetlib.dhp.oa.graph.clean.CleanContextSparkJob;
|
|
|
|
import eu.dnetlib.dhp.schema.common.ModelConstants;
|
|
|
|
import eu.dnetlib.dhp.schema.oaf.*;
|
|
|
|
import scala.Tuple2;
|
2022-07-22 16:23:12 +02:00
|
|
|
|
|
|
|
/**
|
|
|
|
* @author miriam.baglioni
|
|
|
|
* @Date 22/07/22
|
|
|
|
*/
|
|
|
|
public class GetDatasourceFromCountry implements Serializable {
|
2022-08-08 14:10:47 +02:00
|
|
|
private static final Logger log = LoggerFactory.getLogger(GetDatasourceFromCountry.class);
|
|
|
|
|
|
|
|
private static final ObjectMapper OBJECT_MAPPER = new ObjectMapper();
|
|
|
|
|
|
|
|
public static void main(String[] args) throws Exception {
|
|
|
|
|
|
|
|
String jsonConfiguration = IOUtils
|
|
|
|
.toString(
|
|
|
|
GetDatasourceFromCountry.class
|
|
|
|
.getResourceAsStream(
|
|
|
|
"/eu/dnetlib/dhp/oa/graph/input_datasource_country_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);
|
|
|
|
|
|
|
|
String inputPath = parser.get("inputPath");
|
|
|
|
log.info("inputPath: {}", inputPath);
|
|
|
|
|
|
|
|
String workingPath = parser.get("workingPath");
|
|
|
|
log.info("workingPath: {}", workingPath);
|
|
|
|
|
|
|
|
String country = parser.get("country");
|
|
|
|
log.info("country: {}", country);
|
|
|
|
|
|
|
|
SparkConf conf = new SparkConf();
|
|
|
|
runWithSparkSession(
|
|
|
|
conf,
|
|
|
|
isSparkSessionManaged,
|
|
|
|
spark -> {
|
|
|
|
getDatasourceFromCountry(spark, country, inputPath, workingPath);
|
|
|
|
});
|
|
|
|
}
|
|
|
|
|
|
|
|
private static void getDatasourceFromCountry(SparkSession spark, String country, String inputPath,
|
|
|
|
String workingPath) {
|
|
|
|
|
|
|
|
Dataset<Organization> organization = spark
|
|
|
|
.read()
|
|
|
|
.textFile(inputPath + "/organization")
|
|
|
|
.map(
|
|
|
|
(MapFunction<String, Organization>) value -> OBJECT_MAPPER.readValue(value, Organization.class),
|
|
|
|
Encoders.bean(Organization.class))
|
|
|
|
.filter(
|
|
|
|
(FilterFunction<Organization>) o -> !o.getDataInfo().getDeletedbyinference() &&
|
|
|
|
o.getCountry().getClassid().length() > 0 &&
|
|
|
|
o.getCountry().getClassid().equals(country));
|
|
|
|
|
|
|
|
// filtering of the relations taking the non deleted by inference and those with IsProvidedBy as relclass
|
|
|
|
Dataset<Relation> relation = spark
|
|
|
|
.read()
|
|
|
|
.textFile(inputPath + "/relation")
|
|
|
|
.map(
|
|
|
|
(MapFunction<String, Relation>) value -> OBJECT_MAPPER.readValue(value, Relation.class),
|
|
|
|
Encoders.bean(Relation.class))
|
|
|
|
.filter(
|
|
|
|
(FilterFunction<Relation>) rel -> rel.getRelClass().equalsIgnoreCase(ModelConstants.IS_PROVIDED_BY) &&
|
|
|
|
!rel.getDataInfo().getDeletedbyinference());
|
|
|
|
|
|
|
|
organization
|
2022-10-11 10:10:20 +02:00
|
|
|
.joinWith(relation, organization.col("id").equalTo(relation.col("target")))
|
2022-08-08 14:10:47 +02:00
|
|
|
.map((MapFunction<Tuple2<Organization, Relation>, String>) t2 -> t2._2().getSource(), Encoders.STRING())
|
|
|
|
.write()
|
|
|
|
.mode(SaveMode.Overwrite)
|
|
|
|
.option("compression", "gzip")
|
|
|
|
.json(workingPath);
|
|
|
|
|
|
|
|
}
|
2022-07-22 16:23:12 +02:00
|
|
|
}
|