BrBETA_dnet-hadoop/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/opencitations/CreateActionSetSparkJob.java

182 lines
5.6 KiB
Java

package eu.dnetlib.dhp.actionmanager.opencitations;
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
import java.io.IOException;
import java.io.Serializable;
import java.util.*;
import org.apache.commons.cli.ParseException;
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.function.FilterFunction;
import org.apache.spark.api.java.function.FlatMapFunction;
import org.apache.spark.sql.Encoders;
import org.apache.spark.sql.SparkSession;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import eu.dnetlib.dhp.schema.action.AtomicAction;
import eu.dnetlib.dhp.schema.common.ModelConstants;
import eu.dnetlib.dhp.schema.common.ModelSupport;
import eu.dnetlib.dhp.schema.oaf.*;
import eu.dnetlib.dhp.schema.oaf.utils.CleaningFunctions;
import eu.dnetlib.dhp.schema.oaf.utils.IdentifierFactory;
import scala.Tuple2;
public class CreateActionSetSparkJob implements Serializable {
public static final String OPENCITATIONS_CLASSID = "sysimport:crosswalk:opencitations";
public static final String OPENCITATIONS_CLASSNAME = "Imported from OpenCitations";
private static final String ID_PREFIX = "50|doi_________::";
private static final String TRUST = "0.91";
private static final Logger log = LoggerFactory.getLogger(CreateActionSetSparkJob.class);
private static final ObjectMapper OBJECT_MAPPER = new ObjectMapper();
public static void main(final String[] args) throws IOException, ParseException {
final ArgumentApplicationParser parser = new ArgumentApplicationParser(
IOUtils
.toString(
Objects
.requireNonNull(
CreateActionSetSparkJob.class
.getResourceAsStream(
"/eu/dnetlib/dhp/actionmanager/opencitations/as_parameters.json"))));
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.toString());
final String outputPath = parser.get("outputPath");
log.info("outputPath {}", outputPath);
final boolean shouldDuplicateRels = Optional
.ofNullable(parser.get("shouldDuplicateRels"))
.map(Boolean::valueOf)
.orElse(Boolean.FALSE);
SparkConf conf = new SparkConf();
runWithSparkSession(
conf,
isSparkSessionManaged,
spark -> {
extractContent(spark, inputPath, outputPath, shouldDuplicateRels);
});
}
private static void extractContent(SparkSession spark, String inputPath, String outputPath,
boolean shouldDuplicateRels) {
spark
.sqlContext()
.createDataset(spark.sparkContext().textFile(inputPath + "/*", 6000), Encoders.STRING())
.flatMap(
(FlatMapFunction<String, Relation>) value -> createRelation(value, shouldDuplicateRels).iterator(),
Encoders.bean(Relation.class))
.filter((FilterFunction<Relation>) value -> value != null)
.toJavaRDD()
.map(p -> new AtomicAction(p.getClass(), p))
.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<Relation> createRelation(String value, boolean duplicate) {
String[] line = value.split(",");
if (!line[1].startsWith("10.")) {
return new ArrayList<>();
}
List<Relation> relationList = new ArrayList<>();
String citing = ID_PREFIX + IdentifierFactory.md5(CleaningFunctions.normalizePidValue("doi", line[1]));
final String cited = ID_PREFIX + IdentifierFactory.md5(CleaningFunctions.normalizePidValue("doi", line[2]));
relationList
.addAll(
getRelations(
citing,
cited));
if (duplicate && line[1].endsWith(".refs")) {
citing = ID_PREFIX + IdentifierFactory
.md5(CleaningFunctions.normalizePidValue("doi", line[1].substring(0, line[1].indexOf(".refs"))));
relationList.addAll(getRelations(citing, cited));
}
return relationList;
}
private static Collection<Relation> getRelations(String citing, String cited) {
return Arrays
.asList(
getRelation(citing, cited, ModelConstants.CITES),
getRelation(cited, citing, ModelConstants.IS_CITED_BY));
}
public static Relation getRelation(
String source,
String target,
String relclass) {
Relation r = new Relation();
r.setCollectedfrom(getCollectedFrom());
r.setSource(source);
r.setTarget(target);
r.setRelClass(relclass);
r.setRelType(ModelConstants.RESULT_RESULT);
r.setSubRelType(ModelConstants.CITATION);
r
.setDataInfo(
getDataInfo());
return r;
}
public static List<KeyValue> getCollectedFrom() {
KeyValue kv = new KeyValue();
kv.setKey(ModelConstants.OPENOCITATIONS_ID);
kv.setValue(ModelConstants.OPENOCITATIONS_NAME);
return Arrays.asList(kv);
}
public static DataInfo getDataInfo() {
DataInfo di = new DataInfo();
di.setInferred(false);
di.setDeletedbyinference(false);
di.setTrust(TRUST);
di
.setProvenanceaction(
getQualifier(OPENCITATIONS_CLASSID, OPENCITATIONS_CLASSNAME, ModelConstants.DNET_PROVENANCE_ACTIONS));
return di;
}
public static Qualifier getQualifier(String class_id, String class_name,
String qualifierSchema) {
Qualifier pa = new Qualifier();
pa.setClassid(class_id);
pa.setClassname(class_name);
pa.setSchemeid(qualifierSchema);
pa.setSchemename(qualifierSchema);
return pa;
}
}