forked from D-Net/dnet-hadoop
82 lines
2.3 KiB
Java
82 lines
2.3 KiB
Java
|
|
package eu.dnetlib.dhp.ircdl_extention;
|
|
|
|
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkHiveSession;
|
|
|
|
import java.util.Optional;
|
|
|
|
import org.apache.commons.io.IOUtils;
|
|
import org.apache.spark.SparkConf;
|
|
import org.apache.spark.api.java.function.FilterFunction;
|
|
import org.apache.spark.sql.Dataset;
|
|
import org.apache.spark.sql.SaveMode;
|
|
import org.apache.spark.sql.SparkSession;
|
|
|
|
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
|
|
import eu.dnetlib.dhp.ircdl_extention.model.Result;
|
|
|
|
public class PrepareDataciteSpark {
|
|
|
|
public static void main(String[] args) throws Exception {
|
|
|
|
String jsonConfiguration = IOUtils
|
|
.toString(
|
|
PrepareDataciteSpark.class
|
|
.getResourceAsStream(
|
|
"/eu/dnetlib/dhp/ircdl_extention/prepare_parameters.json"));
|
|
|
|
final ArgumentApplicationParser parser = new ArgumentApplicationParser(jsonConfiguration);
|
|
|
|
parser.parseArgument(args);
|
|
|
|
Boolean isSparkSessionManaged = Optional
|
|
.ofNullable(parser.get("isSparkSessionManaged"))
|
|
.map(Boolean::valueOf)
|
|
.orElse(Boolean.TRUE);
|
|
|
|
final String inputPath = parser.get("inputPath");
|
|
|
|
final String outputPath = parser.get("outputPath");
|
|
|
|
SparkConf conf = new SparkConf();
|
|
conf.set("hive.metastore.uris", "thrift://iis-cdh5-test-m3.ocean.icm.edu.pl:9083");
|
|
|
|
runWithSparkHiveSession(
|
|
conf,
|
|
isSparkSessionManaged,
|
|
spark -> {
|
|
Utils.removeOutputDir(spark, outputPath);
|
|
exec(spark, outputPath, inputPath);
|
|
});
|
|
}
|
|
|
|
private static void exec(SparkSession spark, String output_path, String input_path) {
|
|
|
|
Dataset<Result> datacite = Utils
|
|
.readPath(
|
|
spark, input_path, Result.class)
|
|
.filter(
|
|
(FilterFunction<Result>) r -> r.getId().startsWith("50|datacite"));
|
|
|
|
datacite.write().option("compression", "gzip").mode(SaveMode.Overwrite).json(output_path + "allDatacite");
|
|
getProviderResult(output_path, datacite, "Zenodo");
|
|
getProviderResult(output_path, datacite, "Figshare");
|
|
getProviderResult(output_path, datacite, "Dryad");
|
|
|
|
}
|
|
|
|
private static void getProviderResult(String output_path, Dataset<Result> datacite, String provider) {
|
|
datacite
|
|
.filter(
|
|
(FilterFunction<Result>) r -> r
|
|
.getPid()
|
|
.stream()
|
|
.anyMatch(p -> p.getKey().equals("doi") && p.getValue().contains(provider.toLowerCase())))
|
|
.write()
|
|
.option("compression", "gzip")
|
|
.mode(SaveMode.Overwrite)
|
|
.json(output_path + provider);
|
|
}
|
|
|
|
}
|