2021-11-11 16:11:25 +01:00
|
|
|
|
|
|
|
package eu.dnetlib.dhp.actionmanager.createunresolvedentities;
|
|
|
|
|
2021-12-22 15:33:05 +01:00
|
|
|
import static eu.dnetlib.dhp.actionmanager.bipmodel.Constants.*;
|
|
|
|
import static eu.dnetlib.dhp.actionmanager.bipmodel.Constants.UPDATE_CLASS_NAME;
|
2021-11-11 16:11:25 +01:00
|
|
|
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
|
|
|
|
|
|
|
|
import java.io.Serializable;
|
2021-12-22 16:03:41 +01:00
|
|
|
import java.util.Arrays;
|
2021-11-11 16:11:25 +01:00
|
|
|
import java.util.List;
|
|
|
|
import java.util.Optional;
|
|
|
|
import java.util.stream.Collectors;
|
|
|
|
|
2021-12-22 16:03:41 +01:00
|
|
|
import eu.dnetlib.dhp.schema.oaf.Instance;
|
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.JavaRDD;
|
|
|
|
import org.apache.spark.api.java.JavaSparkContext;
|
|
|
|
import org.apache.spark.api.java.function.MapFunction;
|
|
|
|
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 com.fasterxml.jackson.databind.ObjectMapper;
|
|
|
|
|
2021-12-22 15:33:05 +01:00
|
|
|
import eu.dnetlib.dhp.actionmanager.bipmodel.BipDeserialize;
|
|
|
|
import eu.dnetlib.dhp.actionmanager.bipmodel.BipScore;
|
2021-11-11 16:11:25 +01:00
|
|
|
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
|
|
|
|
import eu.dnetlib.dhp.common.HdfsSupport;
|
|
|
|
import eu.dnetlib.dhp.schema.common.ModelConstants;
|
|
|
|
import eu.dnetlib.dhp.schema.oaf.KeyValue;
|
|
|
|
import eu.dnetlib.dhp.schema.oaf.Measure;
|
|
|
|
import eu.dnetlib.dhp.schema.oaf.Result;
|
|
|
|
import eu.dnetlib.dhp.schema.oaf.utils.OafMapperUtils;
|
2021-11-15 09:58:48 +01:00
|
|
|
import eu.dnetlib.dhp.utils.DHPUtils;
|
2021-11-11 16:11:25 +01:00
|
|
|
|
|
|
|
public class PrepareBipFinder implements Serializable {
|
|
|
|
|
|
|
|
private static final Logger log = LoggerFactory.getLogger(PrepareBipFinder.class);
|
|
|
|
private static final ObjectMapper OBJECT_MAPPER = new ObjectMapper();
|
|
|
|
|
2021-12-22 16:03:41 +01:00
|
|
|
public static void main(String[] args) throws Exception {
|
2021-11-11 16:11:25 +01:00
|
|
|
|
|
|
|
String jsonConfiguration = IOUtils
|
|
|
|
.toString(
|
|
|
|
PrepareBipFinder.class
|
|
|
|
.getResourceAsStream(
|
2021-11-12 10:18:01 +01:00
|
|
|
"/eu/dnetlib/dhp/actionmanager/createunresolvedentities/prepare_parameters.json"));
|
2021-11-11 16:11:25 +01:00
|
|
|
|
|
|
|
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);
|
|
|
|
|
|
|
|
final 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 -> {
|
|
|
|
HdfsSupport.remove(outputPath, spark.sparkContext().hadoopConfiguration());
|
|
|
|
prepareResults(spark, sourcePath, outputPath);
|
|
|
|
});
|
|
|
|
}
|
|
|
|
|
2021-12-22 16:03:41 +01:00
|
|
|
private static void prepareResults(SparkSession spark, String inputPath, String outputPath) {
|
2021-11-11 16:11:25 +01:00
|
|
|
|
|
|
|
final JavaSparkContext sc = JavaSparkContext.fromSparkContext(spark.sparkContext());
|
|
|
|
|
|
|
|
JavaRDD<BipDeserialize> bipDeserializeJavaRDD = sc
|
|
|
|
.textFile(inputPath)
|
|
|
|
.map(item -> OBJECT_MAPPER.readValue(item, BipDeserialize.class));
|
|
|
|
|
|
|
|
spark
|
|
|
|
.createDataset(bipDeserializeJavaRDD.flatMap(entry -> entry.keySet().stream().map(key -> {
|
|
|
|
BipScore bs = new BipScore();
|
|
|
|
bs.setId(key);
|
|
|
|
bs.setScoreList(entry.get(key));
|
2021-12-22 15:33:05 +01:00
|
|
|
|
2021-11-11 16:11:25 +01:00
|
|
|
return bs;
|
|
|
|
}).collect(Collectors.toList()).iterator()).rdd(), Encoders.bean(BipScore.class))
|
|
|
|
.map((MapFunction<BipScore, Result>) v -> {
|
|
|
|
Result r = new Result();
|
|
|
|
|
2021-11-12 10:56:58 +01:00
|
|
|
r.setId(DHPUtils.generateUnresolvedIdentifier(v.getId(), DOI));
|
2021-12-22 16:03:41 +01:00
|
|
|
Instance inst = new Instance();
|
|
|
|
inst.setMeasures(getMeasure(v));
|
|
|
|
r.setInstance(Arrays.asList(inst));
|
2021-11-11 16:11:25 +01:00
|
|
|
return r;
|
|
|
|
}, Encoders.bean(Result.class))
|
|
|
|
.write()
|
|
|
|
.mode(SaveMode.Overwrite)
|
|
|
|
.option("compression", "gzip")
|
|
|
|
.json(outputPath + "/bip");
|
|
|
|
}
|
|
|
|
|
|
|
|
private static List<Measure> getMeasure(BipScore value) {
|
|
|
|
return value
|
|
|
|
.getScoreList()
|
|
|
|
.stream()
|
|
|
|
.map(score -> {
|
|
|
|
Measure m = new Measure();
|
|
|
|
m.setId(score.getId());
|
|
|
|
m
|
|
|
|
.setUnit(
|
|
|
|
score
|
|
|
|
.getUnit()
|
|
|
|
.stream()
|
|
|
|
.map(unit -> {
|
|
|
|
KeyValue kv = new KeyValue();
|
|
|
|
kv.setValue(unit.getValue());
|
|
|
|
kv.setKey(unit.getKey());
|
|
|
|
kv
|
|
|
|
.setDataInfo(
|
|
|
|
OafMapperUtils
|
|
|
|
.dataInfo(
|
|
|
|
false,
|
|
|
|
UPDATE_DATA_INFO_TYPE,
|
|
|
|
true,
|
|
|
|
false,
|
|
|
|
OafMapperUtils
|
|
|
|
.qualifier(
|
|
|
|
UPDATE_MEASURE_BIP_CLASS_ID,
|
|
|
|
UPDATE_CLASS_NAME,
|
|
|
|
ModelConstants.DNET_PROVENANCE_ACTIONS,
|
|
|
|
ModelConstants.DNET_PROVENANCE_ACTIONS),
|
|
|
|
""));
|
|
|
|
return kv;
|
|
|
|
})
|
|
|
|
.collect(Collectors.toList()));
|
|
|
|
return m;
|
|
|
|
})
|
|
|
|
.collect(Collectors.toList());
|
|
|
|
}
|
|
|
|
}
|