2020-12-01 14:31:34 +01:00
|
|
|
|
2020-11-30 20:14:23 +01:00
|
|
|
package eu.dnetlib.dhp.actionmanager.bipfinder;
|
|
|
|
|
2021-12-22 15:08:33 +01:00
|
|
|
import static eu.dnetlib.dhp.actionmanager.common.Constants.*;
|
2020-12-01 14:31:34 +01:00
|
|
|
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
|
|
|
|
|
|
|
|
import java.io.Serializable;
|
|
|
|
import java.util.List;
|
|
|
|
import java.util.Optional;
|
|
|
|
import java.util.stream.Collectors;
|
|
|
|
|
2020-11-30 20:14:23 +01:00
|
|
|
import org.apache.commons.io.IOUtils;
|
|
|
|
import org.apache.hadoop.io.Text;
|
|
|
|
import org.apache.hadoop.mapred.SequenceFileOutputFormat;
|
|
|
|
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.Dataset;
|
|
|
|
import org.apache.spark.sql.Encoders;
|
|
|
|
import org.apache.spark.sql.SparkSession;
|
|
|
|
import org.slf4j.Logger;
|
|
|
|
import org.slf4j.LoggerFactory;
|
|
|
|
|
2020-12-01 14:31:34 +01:00
|
|
|
import com.fasterxml.jackson.databind.ObjectMapper;
|
2020-11-30 20:14:23 +01:00
|
|
|
|
2021-12-22 15:08:33 +01:00
|
|
|
import eu.dnetlib.dhp.actionmanager.common.BipDeserialize;
|
|
|
|
import eu.dnetlib.dhp.actionmanager.common.BipScore;
|
2020-12-01 14:31:34 +01:00
|
|
|
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
|
|
|
|
import eu.dnetlib.dhp.common.HdfsSupport;
|
|
|
|
import eu.dnetlib.dhp.schema.action.AtomicAction;
|
2021-12-14 11:12:17 +01:00
|
|
|
import eu.dnetlib.dhp.schema.common.ModelConstants;
|
2020-12-01 14:31:34 +01:00
|
|
|
import eu.dnetlib.dhp.schema.oaf.*;
|
|
|
|
import eu.dnetlib.dhp.schema.oaf.KeyValue;
|
2021-12-14 11:12:17 +01:00
|
|
|
import eu.dnetlib.dhp.schema.oaf.utils.OafMapperUtils;
|
2020-12-01 14:31:34 +01:00
|
|
|
import scala.Tuple2;
|
2020-11-30 20:14:23 +01:00
|
|
|
|
2020-12-01 18:36:29 +01:00
|
|
|
/**
|
|
|
|
* created the Atomic Action for each tipe of results
|
|
|
|
*/
|
2020-11-30 20:14:23 +01:00
|
|
|
public class SparkAtomicActionScoreJob implements Serializable {
|
|
|
|
|
2021-05-14 10:58:12 +02:00
|
|
|
private static final String DOI = "doi";
|
2020-12-01 14:31:34 +01:00
|
|
|
private static final Logger log = LoggerFactory.getLogger(SparkAtomicActionScoreJob.class);
|
|
|
|
private static final ObjectMapper OBJECT_MAPPER = new ObjectMapper();
|
2020-11-30 20:14:23 +01:00
|
|
|
|
2020-12-01 14:31:34 +01:00
|
|
|
public static <I extends Result> void main(String[] args) throws Exception {
|
2020-11-30 20:14:23 +01:00
|
|
|
|
2020-12-01 14:31:34 +01:00
|
|
|
String jsonConfiguration = IOUtils
|
|
|
|
.toString(
|
|
|
|
SparkAtomicActionScoreJob.class
|
|
|
|
.getResourceAsStream(
|
2021-12-14 14:51:14 +01:00
|
|
|
"/eu/dnetlib/dhp/actionmanager/bipfinder/input_actionset_parameter.json"));
|
2020-12-01 14:31:34 +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 inputPath = parser.get("inputPath");
|
|
|
|
log.info("inputPath {}: ", inputPath);
|
|
|
|
|
|
|
|
final String outputPath = parser.get("outputPath");
|
|
|
|
log.info("outputPath {}: ", outputPath);
|
|
|
|
|
|
|
|
SparkConf conf = new SparkConf();
|
|
|
|
|
|
|
|
runWithSparkSession(
|
|
|
|
conf,
|
|
|
|
isSparkSessionManaged,
|
|
|
|
spark -> {
|
|
|
|
removeOutputDir(spark, outputPath);
|
2021-12-14 14:51:14 +01:00
|
|
|
prepareResults(spark, inputPath, outputPath);
|
2020-12-01 14:31:34 +01:00
|
|
|
});
|
|
|
|
}
|
|
|
|
|
2021-12-14 14:51:14 +01:00
|
|
|
private static <I extends Result> void prepareResults(SparkSession spark, String bipScorePath, String outputPath) {
|
2020-12-01 14:31:34 +01:00
|
|
|
|
2021-08-11 12:13:22 +02:00
|
|
|
final JavaSparkContext sc = JavaSparkContext.fromSparkContext(spark.sparkContext());
|
2020-12-01 14:31:34 +01:00
|
|
|
|
|
|
|
JavaRDD<BipDeserialize> bipDeserializeJavaRDD = sc
|
|
|
|
.textFile(bipScorePath)
|
|
|
|
.map(item -> OBJECT_MAPPER.readValue(item, BipDeserialize.class));
|
|
|
|
|
|
|
|
Dataset<BipScore> bipScores = spark
|
|
|
|
.createDataset(bipDeserializeJavaRDD.flatMap(entry -> entry.keySet().stream().map(key -> {
|
|
|
|
BipScore bs = new BipScore();
|
|
|
|
bs.setId(key);
|
|
|
|
bs.setScoreList(entry.get(key));
|
|
|
|
return bs;
|
|
|
|
}).collect(Collectors.toList()).iterator()).rdd(), Encoders.bean(BipScore.class));
|
|
|
|
|
2020-12-01 15:02:31 +01:00
|
|
|
bipScores
|
2021-12-14 11:12:17 +01:00
|
|
|
|
|
|
|
.map((MapFunction<BipScore, Result>) bs -> {
|
2021-01-07 10:07:30 +01:00
|
|
|
Result ret = new Result();
|
2020-12-01 14:31:34 +01:00
|
|
|
|
2021-12-14 11:12:17 +01:00
|
|
|
ret.setId(bs.getId());
|
|
|
|
|
|
|
|
ret.setMeasures(getMeasure(bs));
|
2020-12-01 14:31:34 +01:00
|
|
|
|
2021-01-08 16:29:17 +01:00
|
|
|
return ret;
|
|
|
|
}, Encoders.bean(Result.class))
|
2020-12-01 14:31:34 +01:00
|
|
|
.toJavaRDD()
|
2021-12-14 14:51:14 +01:00
|
|
|
.map(p -> new AtomicAction(Result.class, p))
|
2020-12-01 14:31:34 +01:00
|
|
|
.mapToPair(
|
|
|
|
aa -> new Tuple2<>(new Text(aa.getClazz().getCanonicalName()),
|
|
|
|
new Text(OBJECT_MAPPER.writeValueAsString(aa))))
|
|
|
|
.saveAsHadoopFile(outputPath, Text.class, Text.class, SequenceFileOutputFormat.class);
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
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());
|
2021-12-14 11:12:17 +01:00
|
|
|
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),
|
|
|
|
""));
|
2020-12-01 14:31:34 +01:00
|
|
|
return kv;
|
|
|
|
})
|
|
|
|
.collect(Collectors.toList()));
|
|
|
|
return m;
|
|
|
|
})
|
|
|
|
.collect(Collectors.toList());
|
|
|
|
}
|
|
|
|
|
|
|
|
private static void removeOutputDir(SparkSession spark, String path) {
|
|
|
|
HdfsSupport.remove(path, spark.sparkContext().hadoopConfiguration());
|
|
|
|
}
|
|
|
|
|
|
|
|
public static <R> Dataset<R> readPath(
|
|
|
|
SparkSession spark, String inputPath, Class<R> clazz) {
|
|
|
|
return spark
|
|
|
|
.read()
|
|
|
|
.textFile(inputPath)
|
|
|
|
.map((MapFunction<String, R>) value -> OBJECT_MAPPER.readValue(value, clazz), Encoders.bean(clazz));
|
|
|
|
}
|
2020-11-30 20:14:23 +01:00
|
|
|
|
|
|
|
}
|