2021-11-11 16:11:25 +01:00
|
|
|
|
|
|
|
package eu.dnetlib.dhp.actionmanager.createunresolvedentities;
|
|
|
|
|
2021-12-22 16:44:44 +01:00
|
|
|
import static eu.dnetlib.dhp.actionmanager.Constants.*;
|
2021-11-11 16:11:25 +01:00
|
|
|
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
|
|
|
|
|
|
|
|
import java.io.Serializable;
|
2022-02-01 08:38:41 +01:00
|
|
|
import java.util.Optional;
|
2021-11-11 16:11:25 +01:00
|
|
|
|
|
|
|
import org.apache.commons.io.IOUtils;
|
|
|
|
import org.apache.spark.SparkConf;
|
|
|
|
import org.apache.spark.api.java.function.MapFunction;
|
|
|
|
import org.apache.spark.api.java.function.MapGroupsFunction;
|
|
|
|
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;
|
|
|
|
|
|
|
|
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
|
|
|
|
import eu.dnetlib.dhp.schema.oaf.Result;
|
|
|
|
|
|
|
|
public class SparkSaveUnresolved implements Serializable {
|
2021-12-22 16:03:41 +01:00
|
|
|
private static final Logger log = LoggerFactory.getLogger(SparkSaveUnresolved.class);
|
2021-11-11 16:11:25 +01:00
|
|
|
|
|
|
|
public static void main(String[] args) throws Exception {
|
|
|
|
|
|
|
|
String jsonConfiguration = IOUtils
|
|
|
|
.toString(
|
|
|
|
PrepareFOSSparkJob.class
|
|
|
|
.getResourceAsStream(
|
|
|
|
"/eu/dnetlib/dhp/actionmanager/createunresolvedentities/produce_unresolved_parameters.json"));
|
|
|
|
|
|
|
|
final ArgumentApplicationParser parser = new ArgumentApplicationParser(jsonConfiguration);
|
|
|
|
|
|
|
|
parser.parseArgument(args);
|
|
|
|
|
|
|
|
Boolean isSparkSessionManaged = isSparkSessionManaged(parser);
|
|
|
|
log.info("isSparkSessionManaged: {}", isSparkSessionManaged);
|
|
|
|
|
|
|
|
String sourcePath = parser.get("sourcePath");
|
|
|
|
log.info("sourcePath: {}", sourcePath);
|
|
|
|
|
|
|
|
final String outputPath = parser.get("outputPath");
|
|
|
|
log.info("outputPath: {}", outputPath);
|
|
|
|
|
|
|
|
SparkConf conf = new SparkConf();
|
|
|
|
runWithSparkSession(
|
|
|
|
conf,
|
|
|
|
isSparkSessionManaged,
|
|
|
|
spark -> {
|
|
|
|
saveUnresolved(
|
|
|
|
spark,
|
|
|
|
sourcePath,
|
|
|
|
|
|
|
|
outputPath);
|
|
|
|
});
|
|
|
|
}
|
|
|
|
|
|
|
|
private static void saveUnresolved(SparkSession spark, String sourcePath, String outputPath) {
|
|
|
|
|
|
|
|
spark
|
|
|
|
.read()
|
|
|
|
.textFile(sourcePath + "/*")
|
|
|
|
.map(
|
|
|
|
(MapFunction<String, Result>) l -> OBJECT_MAPPER.readValue(l, Result.class),
|
|
|
|
Encoders.bean(Result.class))
|
2021-12-22 16:03:41 +01:00
|
|
|
.groupByKey((MapFunction<Result, String>) Result::getId, Encoders.STRING())
|
2021-11-11 16:11:25 +01:00
|
|
|
.mapGroups((MapGroupsFunction<String, Result, Result>) (k, it) -> {
|
|
|
|
Result ret = it.next();
|
2022-02-01 08:38:41 +01:00
|
|
|
it.forEachRemaining(r -> {
|
|
|
|
if (r.getInstance() != null) {
|
|
|
|
ret.setInstance(r.getInstance());
|
|
|
|
}
|
|
|
|
if (r.getSubject() != null) {
|
|
|
|
if (ret.getSubject() != null)
|
|
|
|
ret.getSubject().addAll(r.getSubject());
|
|
|
|
else
|
|
|
|
ret.setSubject(r.getSubject());
|
|
|
|
}
|
|
|
|
|
|
|
|
// ret.mergeFrom(r)
|
|
|
|
});
|
2021-11-11 16:11:25 +01:00
|
|
|
return ret;
|
|
|
|
}, Encoders.bean(Result.class))
|
|
|
|
.write()
|
|
|
|
.mode(SaveMode.Overwrite)
|
|
|
|
.option("compression", "gzip")
|
|
|
|
.json(outputPath);
|
|
|
|
}
|
|
|
|
|
|
|
|
}
|