2020-04-30 20:22:07 +02:00
|
|
|
|
2020-04-30 18:25:33 +02:00
|
|
|
package eu.dnetlib.dhp.blacklist;
|
|
|
|
|
2020-04-30 20:22:07 +02:00
|
|
|
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkHiveSession;
|
|
|
|
|
|
|
|
import java.util.Optional;
|
|
|
|
|
2020-04-30 18:25:33 +02:00
|
|
|
import org.apache.commons.io.IOUtils;
|
|
|
|
import org.apache.spark.SparkConf;
|
|
|
|
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;
|
|
|
|
|
2020-04-30 20:22:07 +02:00
|
|
|
import com.fasterxml.jackson.databind.ObjectMapper;
|
2020-04-30 18:25:33 +02:00
|
|
|
|
2020-04-30 20:22:07 +02:00
|
|
|
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
|
|
|
|
import eu.dnetlib.dhp.schema.oaf.Relation;
|
2020-04-30 18:25:33 +02:00
|
|
|
|
|
|
|
public class PrepareMergedRelationJob {
|
|
|
|
|
2020-04-30 20:22:07 +02:00
|
|
|
private static final Logger log = LoggerFactory.getLogger(PrepareMergedRelationJob.class);
|
|
|
|
private static final ObjectMapper OBJECT_MAPPER = new ObjectMapper();
|
|
|
|
|
|
|
|
public static void main(String[] args) throws Exception {
|
|
|
|
|
|
|
|
String jsonConfiguration = IOUtils
|
|
|
|
.toString(
|
|
|
|
PrepareMergedRelationJob.class
|
|
|
|
.getResourceAsStream(
|
|
|
|
"/eu/dnetlib/dhp/blacklist/input_preparerelation_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("sourcePath");
|
|
|
|
log.info("inputPath: {}", inputPath);
|
|
|
|
|
|
|
|
String outputPath = parser.get("outputPath");
|
|
|
|
log.info("outputPath: {} ", outputPath);
|
|
|
|
|
|
|
|
SparkConf conf = new SparkConf();
|
|
|
|
conf.set("hive.metastore.uris", parser.get("hive_metastore_uris"));
|
|
|
|
|
|
|
|
runWithSparkHiveSession(
|
|
|
|
conf,
|
|
|
|
isSparkSessionManaged,
|
|
|
|
spark -> {
|
|
|
|
selectMergesRelations(
|
|
|
|
spark,
|
|
|
|
inputPath,
|
|
|
|
outputPath);
|
|
|
|
});
|
|
|
|
}
|
|
|
|
|
|
|
|
private static void selectMergesRelations(SparkSession spark, String inputPath, String outputPath) {
|
|
|
|
|
|
|
|
Dataset<Relation> relation = readRelations(spark, inputPath);
|
2020-05-08 12:54:52 +02:00
|
|
|
|
2020-05-11 09:40:58 +02:00
|
|
|
relation
|
|
|
|
.filter("relclass = 'merges' and datainfo.deletedbyinference=false")
|
|
|
|
.write()
|
|
|
|
.mode(SaveMode.Overwrite)
|
|
|
|
.option("compression", "gzip")
|
|
|
|
.json(outputPath);
|
2020-04-30 20:22:07 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
public static org.apache.spark.sql.Dataset<Relation> readRelations(
|
|
|
|
SparkSession spark, String inputPath) {
|
|
|
|
return spark
|
|
|
|
.read()
|
|
|
|
.textFile(inputPath)
|
|
|
|
.map(
|
|
|
|
(MapFunction<String, Relation>) value -> OBJECT_MAPPER.readValue(value, Relation.class),
|
|
|
|
Encoders.bean(Relation.class));
|
|
|
|
}
|
2020-04-30 18:25:33 +02:00
|
|
|
}
|