seconf phase of data preparation. Groups all the possible updates by id

This commit is contained in:
Miriam Baglioni 2020-04-16 16:08:51 +02:00
parent 548ba915ac
commit 0e5399bf74
1 changed files with 43 additions and 71 deletions

View File

@ -1,34 +1,27 @@
package eu.dnetlib.dhp.orcidtoresultfromsemrel; package eu.dnetlib.dhp.orcidtoresultfromsemrel;
import com.fasterxml.jackson.databind.ObjectMapper; import com.fasterxml.jackson.databind.ObjectMapper;
import com.google.gson.Gson;
import eu.dnetlib.dhp.application.ArgumentApplicationParser; import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import eu.dnetlib.dhp.schema.oaf.Relation;
import eu.dnetlib.dhp.schema.oaf.Result;
import org.apache.commons.io.IOUtils; import org.apache.commons.io.IOUtils;
import org.apache.hadoop.io.compress.GzipCodec;
import org.apache.spark.SparkConf; import org.apache.spark.SparkConf;
import org.apache.spark.api.java.JavaSparkContext; import org.apache.spark.sql.*;
import org.apache.spark.sql.Dataset;
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.Logger;
import org.slf4j.LoggerFactory; import org.slf4j.LoggerFactory;
import scala.Tuple2;
import java.util.Arrays; import java.util.HashSet;
import java.util.List; import java.util.Set;
import static eu.dnetlib.dhp.PropagationConstant.*; import static eu.dnetlib.dhp.PropagationConstant.*;
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkHiveSession; import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
public class PrepareResultOrcidAssociationStep1 { public class PrepareResultOrcidAssociationStep2 {
private static final Logger log = LoggerFactory.getLogger(PrepareResultOrcidAssociationStep1.class); private static final Logger log = LoggerFactory.getLogger(PrepareResultOrcidAssociationStep2.class);
private static final ObjectMapper OBJECT_MAPPER = new ObjectMapper(); private static final ObjectMapper OBJECT_MAPPER = new ObjectMapper();
public static void main(String[] args) throws Exception { public static void main(String[] args) throws Exception {
String jsonConfiguration = IOUtils.toString(SparkOrcidToResultFromSemRelJob3.class String jsonConfiguration = IOUtils.toString(SparkOrcidToResultFromSemRelJob3.class
.getResourceAsStream("/eu/dnetlib/dhp/orcidtoresultfromsemrel/input_prepareorcidtoresult_parameters.json")); .getResourceAsStream("/eu/dnetlib/dhp/orcidtoresultfromsemrel/input_prepareorcidtoresult2_parameters.json"));
final ArgumentApplicationParser parser = new ArgumentApplicationParser( final ArgumentApplicationParser parser = new ArgumentApplicationParser(
jsonConfiguration); jsonConfiguration);
@ -44,76 +37,55 @@ public class PrepareResultOrcidAssociationStep1 {
final String outputPath = parser.get("outputPath"); final String outputPath = parser.get("outputPath");
log.info("outputPath: {}", outputPath); log.info("outputPath: {}", outputPath);
final String resultClassName = parser.get("resultTableName");
log.info("resultTableName: {}", resultClassName);
final List<String> allowedsemrel = Arrays.asList(parser.get("allowedsemrel").split(";"));
log.info("allowedSemRel: {}", new Gson().toJson(allowedsemrel));
final String resultType = resultClassName.substring(resultClassName.lastIndexOf(".") + 1).toLowerCase();
log.info("resultType: {}", resultType);
Class<? extends Result> resultClazz = (Class<? extends Result>) Class.forName(resultClassName);
SparkConf conf = new SparkConf(); SparkConf conf = new SparkConf();
conf.set("hive.metastore.uris", parser.get("hive_metastore_uris"));
runWithSparkHiveSession(conf, isSparkSessionManaged, runWithSparkSession(conf, isSparkSessionManaged,
spark -> { spark -> {
if (isTest(parser)) { if (isTest(parser)) {
removeOutputDir(spark, outputPath); removeOutputDir(spark, outputPath);
} }
prepareInfo(spark, inputPath, outputPath, resultClazz, resultType, allowedsemrel); mergeInfo(spark, inputPath, outputPath);
}); });
} }
private static <R extends Result> void prepareInfo(SparkSession spark, String inputPath, private static void mergeInfo(SparkSession spark, String inputPath, String outputPath) {
String outputPath, Class<R> resultClazz,
String resultType,
List<String> allowedsemrel) {
//read the relation table and the table related to the result it is using Dataset<ResultOrcidList> resultOrcidAssoc = readAssocResultOrcidList(spark, inputPath);
final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext());
org.apache.spark.sql.Dataset<Relation> relation = spark.createDataset(sc.textFile(inputPath + "/relation")
.map(item -> new ObjectMapper().readValue(item, Relation.class)).rdd(), Encoders.bean(Relation.class));
relation.createOrReplaceTempView("relation");
log.info("Reading Graph table from: {}", inputPath + "/" + resultType); resultOrcidAssoc
Dataset<R> result = readPathEntity(spark, inputPath + "/" + resultType, resultClazz); .toJavaRDD()
.mapToPair(r -> new Tuple2<>(r.getResultId(), r))
result.createOrReplaceTempView("result"); .reduceByKey((a, b) -> {
if (a == null) {
getPossibleResultOrcidAssociation(spark, allowedsemrel, outputPath); return b;
}
if (b == null) {
return a;
}
Set<String> orcid_set = new HashSet<>();
a.getAuthorList().stream().forEach(aa -> orcid_set.add(aa.getOrcid()));
b.getAuthorList().stream().forEach(aa -> {
if (!orcid_set.contains(aa.getOrcid())) {
a.getAuthorList().add(aa);
orcid_set.add(aa.getOrcid());
}
});
return a;
})
.map(c -> c._2())
.map(r -> OBJECT_MAPPER.writeValueAsString(r))
.saveAsTextFile(outputPath, GzipCodec.class);
} }
private static void getPossibleResultOrcidAssociation(SparkSession spark, List<String> allowedsemrel, String outputPath){ private static Dataset<ResultOrcidList> readAssocResultOrcidList(SparkSession spark, String relationPath) {
String query = " select target resultId, author authorList" + return spark
" from (select id, collect_set(named_struct('name', name, 'surname', surname, 'fullname', fullname, 'orcid', orcid)) author " + .read()
" from ( " + .textFile(relationPath)
" select id, MyT.fullname, MyT.name, MyT.surname, MyP.value orcid " + .map(value -> OBJECT_MAPPER.readValue(value, ResultOrcidList.class), Encoders.bean(ResultOrcidList.class));
" from result " +
" lateral view explode (author) a as MyT " +
" lateral view explode (MyT.pid) p as MyP " +
" where MyP.qualifier.classid = 'ORCID') tmp " +
" group by id) r_t " +
" join (" +
" select source, target " +
" from relation " +
" where datainfo.deletedbyinference = false " +
getConstraintList(" relclass = '" ,allowedsemrel) + ") rel_rel " +
" on source = id";
spark.sql(query)
.as(Encoders.bean(ResultOrcidList.class))
.toJSON()
.write()
.mode(SaveMode.Append)
.option("compression","gzip")
.text(outputPath)
;
} }
} }