forked from D-Net/dnet-hadoop
This commit is contained in:
parent
769aa8178a
commit
44fab140de
File diff suppressed because it is too large
Load Diff
File diff suppressed because it is too large
Load Diff
|
|
@ -1,10 +1,14 @@
|
||||||
package eu.dnetlib.dhp.resulttocommunityfromsemrel;
|
package eu.dnetlib.dhp.resulttocommunityfromsemrel;
|
||||||
|
|
||||||
|
import static eu.dnetlib.dhp.PropagationConstant.*;
|
||||||
|
|
||||||
import com.fasterxml.jackson.databind.ObjectMapper;
|
import com.fasterxml.jackson.databind.ObjectMapper;
|
||||||
import eu.dnetlib.dhp.QueryInformationSystem;
|
import eu.dnetlib.dhp.QueryInformationSystem;
|
||||||
import eu.dnetlib.dhp.TypedRow;
|
import eu.dnetlib.dhp.TypedRow;
|
||||||
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
|
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
|
||||||
import eu.dnetlib.dhp.schema.oaf.*;
|
import eu.dnetlib.dhp.schema.oaf.*;
|
||||||
|
import java.util.*;
|
||||||
|
import java.util.stream.Collectors;
|
||||||
import org.apache.commons.io.IOUtils;
|
import org.apache.commons.io.IOUtils;
|
||||||
import org.apache.hadoop.fs.FileSystem;
|
import org.apache.hadoop.fs.FileSystem;
|
||||||
import org.apache.spark.SparkConf;
|
import org.apache.spark.SparkConf;
|
||||||
|
|
@ -16,28 +20,25 @@ import org.apache.spark.sql.Row;
|
||||||
import org.apache.spark.sql.SparkSession;
|
import org.apache.spark.sql.SparkSession;
|
||||||
import scala.Tuple2;
|
import scala.Tuple2;
|
||||||
|
|
||||||
import java.util.*;
|
|
||||||
import java.util.stream.Collectors;
|
|
||||||
|
|
||||||
import static eu.dnetlib.dhp.PropagationConstant.*;
|
|
||||||
|
|
||||||
public class SparkResultToCommunityThroughSemRelJob3 {
|
public class SparkResultToCommunityThroughSemRelJob3 {
|
||||||
public static void main(String[] args) throws Exception {
|
public static void main(String[] args) throws Exception {
|
||||||
|
|
||||||
final ArgumentApplicationParser parser = new ArgumentApplicationParser(IOUtils
|
final ArgumentApplicationParser parser =
|
||||||
.toString(SparkResultToCommunityThroughSemRelJob3.class
|
new ArgumentApplicationParser(
|
||||||
.getResourceAsStream("/eu/dnetlib/dhp/resulttocommunityfromsemrel/input_propagationresultcommunityfromsemrel_parameters.json")));
|
IOUtils.toString(
|
||||||
|
SparkResultToCommunityThroughSemRelJob3.class.getResourceAsStream(
|
||||||
|
"/eu/dnetlib/dhp/resulttocommunityfromsemrel/input_communitytoresult_parameters.json")));
|
||||||
parser.parseArgument(args);
|
parser.parseArgument(args);
|
||||||
|
|
||||||
SparkConf conf = new SparkConf();
|
SparkConf conf = new SparkConf();
|
||||||
conf.set("hive.metastore.uris", parser.get("hive_metastore_uris"));
|
conf.set("hive.metastore.uris", parser.get("hive_metastore_uris"));
|
||||||
final SparkSession spark = SparkSession
|
final SparkSession spark =
|
||||||
.builder()
|
SparkSession.builder()
|
||||||
.appName(SparkResultToCommunityThroughSemRelJob3.class.getSimpleName())
|
.appName(SparkResultToCommunityThroughSemRelJob3.class.getSimpleName())
|
||||||
.master(parser.get("master"))
|
.master(parser.get("master"))
|
||||||
.config(conf)
|
.config(conf)
|
||||||
.enableHiveSupport()
|
.enableHiveSupport()
|
||||||
.getOrCreate();
|
.getOrCreate();
|
||||||
|
|
||||||
final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext());
|
final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext());
|
||||||
final String inputPath = parser.get("sourcePath");
|
final String inputPath = parser.get("sourcePath");
|
||||||
|
|
@ -45,42 +46,48 @@ public class SparkResultToCommunityThroughSemRelJob3 {
|
||||||
|
|
||||||
final List<String> allowedsemrel = Arrays.asList(parser.get("allowedsemrels").split(";"));
|
final List<String> allowedsemrel = Arrays.asList(parser.get("allowedsemrels").split(";"));
|
||||||
|
|
||||||
final List<String> communityIdList = QueryInformationSystem.getCommunityList(parser.get("isLookupUrl"));
|
final List<String> communityIdList =
|
||||||
|
QueryInformationSystem.getCommunityList(parser.get("isLookupUrl"));
|
||||||
|
|
||||||
createOutputDirs(outputPath, FileSystem.get(spark.sparkContext().hadoopConfiguration()));
|
createOutputDirs(outputPath, FileSystem.get(spark.sparkContext().hadoopConfiguration()));
|
||||||
|
|
||||||
|
JavaRDD<Publication> publication_rdd =
|
||||||
|
sc.textFile(inputPath + "/publication")
|
||||||
|
.map(item -> new ObjectMapper().readValue(item, Publication.class));
|
||||||
|
|
||||||
JavaRDD<Publication> publication_rdd = sc.textFile(inputPath + "/publication")
|
JavaRDD<Dataset> dataset_rdd =
|
||||||
.map(item -> new ObjectMapper().readValue(item, Publication.class));
|
sc.textFile(inputPath + "/dataset")
|
||||||
|
.map(item -> new ObjectMapper().readValue(item, Dataset.class));
|
||||||
|
|
||||||
JavaRDD<Dataset> dataset_rdd = sc.textFile(inputPath + "/dataset")
|
JavaRDD<OtherResearchProduct> orp_rdd =
|
||||||
.map(item -> new ObjectMapper().readValue(item, Dataset.class));
|
sc.textFile(inputPath + "/otherresearchproduct")
|
||||||
|
.map(
|
||||||
|
item ->
|
||||||
|
new ObjectMapper()
|
||||||
|
.readValue(item, OtherResearchProduct.class));
|
||||||
|
|
||||||
JavaRDD<OtherResearchProduct> orp_rdd = sc.textFile(inputPath + "/otherresearchproduct")
|
JavaRDD<Software> software_rdd =
|
||||||
.map(item -> new ObjectMapper().readValue(item, OtherResearchProduct.class));
|
sc.textFile(inputPath + "/software")
|
||||||
|
.map(item -> new ObjectMapper().readValue(item, Software.class));
|
||||||
|
|
||||||
JavaRDD<Software> software_rdd = sc.textFile(inputPath + "/software")
|
JavaRDD<Relation> relation_rdd =
|
||||||
.map(item -> new ObjectMapper().readValue(item, Software.class));
|
sc.textFile(inputPath + "/relation")
|
||||||
|
.map(item -> new ObjectMapper().readValue(item, Relation.class));
|
||||||
|
|
||||||
JavaRDD<Relation> relation_rdd = sc.textFile(inputPath + "/relation")
|
org.apache.spark.sql.Dataset<Publication> publication =
|
||||||
.map(item -> new ObjectMapper().readValue(item, Relation.class));
|
spark.createDataset(publication_rdd.rdd(), Encoders.bean(Publication.class));
|
||||||
|
|
||||||
|
org.apache.spark.sql.Dataset<Relation> relation =
|
||||||
|
spark.createDataset(relation_rdd.rdd(), Encoders.bean(Relation.class));
|
||||||
|
|
||||||
org.apache.spark.sql.Dataset<Publication> publication = spark.createDataset(publication_rdd.rdd(),
|
org.apache.spark.sql.Dataset<Dataset> dataset =
|
||||||
Encoders.bean(Publication.class));
|
spark.createDataset(dataset_rdd.rdd(), Encoders.bean(Dataset.class));
|
||||||
|
|
||||||
org.apache.spark.sql.Dataset<Relation> relation = spark.createDataset(relation_rdd.rdd(),
|
org.apache.spark.sql.Dataset<OtherResearchProduct> other =
|
||||||
Encoders.bean(Relation.class));
|
spark.createDataset(orp_rdd.rdd(), Encoders.bean(OtherResearchProduct.class));
|
||||||
|
|
||||||
org.apache.spark.sql.Dataset<Dataset> dataset = spark.createDataset(dataset_rdd.rdd(),
|
|
||||||
Encoders.bean(Dataset.class));
|
|
||||||
|
|
||||||
org.apache.spark.sql.Dataset<OtherResearchProduct> other = spark.createDataset(orp_rdd.rdd(),
|
|
||||||
Encoders.bean(OtherResearchProduct.class));
|
|
||||||
|
|
||||||
org.apache.spark.sql.Dataset<Software> software = spark.createDataset(software_rdd.rdd(),
|
|
||||||
Encoders.bean(Software.class));
|
|
||||||
|
|
||||||
|
org.apache.spark.sql.Dataset<Software> software =
|
||||||
|
spark.createDataset(software_rdd.rdd(), Encoders.bean(Software.class));
|
||||||
|
|
||||||
publication.createOrReplaceTempView("publication");
|
publication.createOrReplaceTempView("publication");
|
||||||
relation.createOrReplaceTempView("relation");
|
relation.createOrReplaceTempView("relation");
|
||||||
|
|
@ -90,327 +97,490 @@ public class SparkResultToCommunityThroughSemRelJob3 {
|
||||||
|
|
||||||
String communitylist = getConstraintList(" co.id = '", communityIdList);
|
String communitylist = getConstraintList(" co.id = '", communityIdList);
|
||||||
|
|
||||||
String semrellist = getConstraintList(" relClass = '", allowedsemrel );
|
String semrellist = getConstraintList(" relClass = '", allowedsemrel);
|
||||||
|
|
||||||
|
String query =
|
||||||
|
"Select source, community_context, target "
|
||||||
|
+ "from (select id, collect_set(co.id) community_context "
|
||||||
|
+ "from publication "
|
||||||
|
+ "lateral view explode (context) c as co "
|
||||||
|
+ "where datainfo.deletedbyinference = false "
|
||||||
|
+ communitylist
|
||||||
|
+ " group by id) p "
|
||||||
|
+ "JOIN "
|
||||||
|
+ "(select * "
|
||||||
|
+ "from relation "
|
||||||
|
+ "where datainfo.deletedbyinference = false "
|
||||||
|
+ semrellist
|
||||||
|
+ ") r "
|
||||||
|
+ "ON p.id = r.source";
|
||||||
|
|
||||||
String query = "Select source, community_context, target " +
|
org.apache.spark.sql.Dataset<Row> publication_context = spark.sql(query);
|
||||||
"from (select id, collect_set(co.id) community_context " +
|
|
||||||
"from publication " +
|
|
||||||
"lateral view explode (context) c as co " +
|
|
||||||
"where datainfo.deletedbyinference = false "+ communitylist +
|
|
||||||
" group by id) p " +
|
|
||||||
"JOIN " +
|
|
||||||
"(select * " +
|
|
||||||
"from relation " +
|
|
||||||
"where datainfo.deletedbyinference = false " + semrellist + ") r " +
|
|
||||||
"ON p.id = r.source";
|
|
||||||
|
|
||||||
|
|
||||||
org.apache.spark.sql.Dataset<Row> publication_context = spark.sql( query);
|
|
||||||
publication_context.createOrReplaceTempView("publication_context");
|
publication_context.createOrReplaceTempView("publication_context");
|
||||||
|
|
||||||
//( source, (mes, dh-ch-, ni), target )
|
// ( source, (mes, dh-ch-, ni), target )
|
||||||
query = "select target , collect_set(co) " +
|
query =
|
||||||
"from (select target, community_context " +
|
"select target , collect_set(co) "
|
||||||
"from publication_context pc join publication p on " +
|
+ "from (select target, community_context "
|
||||||
"p.id = pc.source) tmp " +
|
+ "from publication_context pc join publication p on "
|
||||||
"lateral view explode (community_context) c as co " +
|
+ "p.id = pc.source) tmp "
|
||||||
"group by target";
|
+ "lateral view explode (community_context) c as co "
|
||||||
|
+ "group by target";
|
||||||
|
|
||||||
|
|
||||||
org.apache.spark.sql.Dataset<Row> toupdatepublicationreresult = spark.sql(query);
|
org.apache.spark.sql.Dataset<Row> toupdatepublicationreresult = spark.sql(query);
|
||||||
org.apache.spark.sql.Dataset<Row> toupdatesoftwareresult = getUpdateCommunitiesForTable(spark, "software");
|
org.apache.spark.sql.Dataset<Row> toupdatesoftwareresult =
|
||||||
org.apache.spark.sql.Dataset<Row> toupdatedatasetresult = getUpdateCommunitiesForTable(spark, "dataset");
|
getUpdateCommunitiesForTable(spark, "software");
|
||||||
org.apache.spark.sql.Dataset<Row> toupdateotherresult = getUpdateCommunitiesForTable(spark, "other");
|
org.apache.spark.sql.Dataset<Row> toupdatedatasetresult =
|
||||||
|
getUpdateCommunitiesForTable(spark, "dataset");
|
||||||
|
org.apache.spark.sql.Dataset<Row> toupdateotherresult =
|
||||||
|
getUpdateCommunitiesForTable(spark, "other");
|
||||||
|
|
||||||
createUpdateForResultDatasetWrite(toupdatesoftwareresult.toJavaRDD(), outputPath, "software_update",
|
createUpdateForResultDatasetWrite(
|
||||||
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME, communityIdList);
|
toupdatesoftwareresult.toJavaRDD(),
|
||||||
|
outputPath,
|
||||||
|
"software_update",
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID,
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME,
|
||||||
|
communityIdList);
|
||||||
|
|
||||||
createUpdateForResultDatasetWrite(toupdatedatasetresult.toJavaRDD(), outputPath, "dataset_update",
|
createUpdateForResultDatasetWrite(
|
||||||
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME, communityIdList);
|
toupdatedatasetresult.toJavaRDD(),
|
||||||
|
outputPath,
|
||||||
|
"dataset_update",
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID,
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME,
|
||||||
|
communityIdList);
|
||||||
|
|
||||||
createUpdateForResultDatasetWrite(toupdatepublicationreresult.toJavaRDD(), outputPath, "publication_update",
|
createUpdateForResultDatasetWrite(
|
||||||
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME, communityIdList);
|
toupdatepublicationreresult.toJavaRDD(),
|
||||||
|
outputPath,
|
||||||
|
"publication_update",
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID,
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME,
|
||||||
|
communityIdList);
|
||||||
|
|
||||||
createUpdateForResultDatasetWrite(toupdateotherresult.toJavaRDD(), outputPath, "other_update",
|
createUpdateForResultDatasetWrite(
|
||||||
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME, communityIdList);
|
toupdateotherresult.toJavaRDD(),
|
||||||
|
outputPath,
|
||||||
|
"other_update",
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID,
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME,
|
||||||
|
communityIdList);
|
||||||
|
|
||||||
updateForDatasetDataset(toupdatedatasetresult.toJavaRDD(), dataset.toJavaRDD(), outputPath, "dataset",
|
updateForDatasetDataset(
|
||||||
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME, communityIdList);
|
toupdatedatasetresult.toJavaRDD(),
|
||||||
|
dataset.toJavaRDD(),
|
||||||
|
outputPath,
|
||||||
|
"dataset",
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID,
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME,
|
||||||
|
communityIdList);
|
||||||
|
|
||||||
updateForOtherDataset(toupdateotherresult.toJavaRDD(), other.toJavaRDD(), outputPath, "otherresearchproduct",
|
updateForOtherDataset(
|
||||||
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME, communityIdList);
|
toupdateotherresult.toJavaRDD(),
|
||||||
|
other.toJavaRDD(),
|
||||||
|
outputPath,
|
||||||
|
"otherresearchproduct",
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID,
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME,
|
||||||
|
communityIdList);
|
||||||
|
|
||||||
updateForSoftwareDataset(toupdatesoftwareresult.toJavaRDD(), software.toJavaRDD(), outputPath, "software",
|
updateForSoftwareDataset(
|
||||||
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME, communityIdList);
|
toupdatesoftwareresult.toJavaRDD(),
|
||||||
|
software.toJavaRDD(),
|
||||||
|
outputPath,
|
||||||
|
"software",
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID,
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME,
|
||||||
|
communityIdList);
|
||||||
|
|
||||||
updateForPublicationDataset(toupdatepublicationreresult.toJavaRDD(), publication.toJavaRDD(), outputPath, "publication",
|
updateForPublicationDataset(
|
||||||
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME, communityIdList);
|
toupdatepublicationreresult.toJavaRDD(),
|
||||||
//
|
publication.toJavaRDD(),
|
||||||
|
outputPath,
|
||||||
|
"publication",
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID,
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME,
|
||||||
|
communityIdList);
|
||||||
|
//
|
||||||
|
|
||||||
/*
|
/*
|
||||||
JavaPairRDD<String, TypedRow> resultLinkedToCommunities = publication
|
JavaPairRDD<String, TypedRow> resultLinkedToCommunities = publication
|
||||||
.map(p -> getTypedRow(communityIdList, p.getContext(), p.getId(),"publication"))
|
.map(p -> getTypedRow(communityIdList, p.getContext(), p.getId(),"publication"))
|
||||||
.filter(p -> !(p == null))
|
|
||||||
.mapToPair(toPair())
|
|
||||||
.union(datasets
|
|
||||||
.map(p -> getTypedRow(communityIdList, p.getContext(), p.getId(),"dataset"))
|
|
||||||
.filter(p -> !(p == null))
|
.filter(p -> !(p == null))
|
||||||
.mapToPair(toPair())
|
.mapToPair(toPair())
|
||||||
)
|
.union(datasets
|
||||||
.union(software
|
.map(p -> getTypedRow(communityIdList, p.getContext(), p.getId(),"dataset"))
|
||||||
.map(p -> getTypedRow(communityIdList, p.getContext(), p.getId(),"software"))
|
.filter(p -> !(p == null))
|
||||||
.filter(p -> !(p == null))
|
.mapToPair(toPair())
|
||||||
.mapToPair(toPair())
|
)
|
||||||
)
|
.union(software
|
||||||
.union(other
|
.map(p -> getTypedRow(communityIdList, p.getContext(), p.getId(),"software"))
|
||||||
.map(p -> getTypedRow(communityIdList, p.getContext(), p.getId(),"otherresearchproduct"))
|
.filter(p -> !(p == null))
|
||||||
.filter(p -> !(p == null))
|
.mapToPair(toPair())
|
||||||
.mapToPair(toPair())
|
)
|
||||||
);
|
.union(other
|
||||||
|
.map(p -> getTypedRow(communityIdList, p.getContext(), p.getId(),"otherresearchproduct"))
|
||||||
|
.filter(p -> !(p == null))
|
||||||
|
.mapToPair(toPair())
|
||||||
|
);
|
||||||
|
|
||||||
JavaPairRDD<String, TypedRow> to_add_result_communities = resultLinkedToCommunities.join(result_result).map(r -> r._2()._1().setSourceId(r._2()._2().getTargetId()))
|
JavaPairRDD<String, TypedRow> to_add_result_communities = resultLinkedToCommunities.join(result_result).map(r -> r._2()._1().setSourceId(r._2()._2().getTargetId()))
|
||||||
.mapToPair(toPair());
|
.mapToPair(toPair());
|
||||||
|
|
||||||
JavaPairRDD<String, Result> pubs = publications.mapToPair(p -> new Tuple2<>(p.getId(),p));
|
JavaPairRDD<String, Result> pubs = publications.mapToPair(p -> new Tuple2<>(p.getId(),p));
|
||||||
JavaPairRDD<String, Result> dss = datasets.mapToPair(p -> new Tuple2<>(p.getId(),p));
|
JavaPairRDD<String, Result> dss = datasets.mapToPair(p -> new Tuple2<>(p.getId(),p));
|
||||||
JavaPairRDD<String, Result> sfw = software.mapToPair(p -> new Tuple2<>(p.getId(),p));
|
JavaPairRDD<String, Result> sfw = software.mapToPair(p -> new Tuple2<>(p.getId(),p));
|
||||||
JavaPairRDD<String, Result> orp = other.mapToPair(p -> new Tuple2<>(p.getId(),p));
|
JavaPairRDD<String, Result> orp = other.mapToPair(p -> new Tuple2<>(p.getId(),p));
|
||||||
|
|
||||||
updateResultForCommunity(pubs, to_add_result_communities, outputPath, "publication", PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME);
|
updateResultForCommunity(pubs, to_add_result_communities, outputPath, "publication", PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME);
|
||||||
updateResultForCommunity(dss, to_add_result_communities, outputPath, "dataset", PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME);
|
updateResultForCommunity(dss, to_add_result_communities, outputPath, "dataset", PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME);
|
||||||
updateResultForCommunity(sfw, to_add_result_communities, outputPath, "software", PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME);
|
updateResultForCommunity(sfw, to_add_result_communities, outputPath, "software", PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME);
|
||||||
updateResultForCommunity(orp, to_add_result_communities, outputPath, "otherresearchproduct", PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME);
|
updateResultForCommunity(orp, to_add_result_communities, outputPath, "otherresearchproduct", PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME);
|
||||||
//leftouterjoin result.to_add_result_communities (result = java pair rdd result) [left outer join perche' li voglio tutti anche quelli che non ho aggiornato]
|
//leftouterjoin result.to_add_result_communities (result = java pair rdd result) [left outer join perche' li voglio tutti anche quelli che non ho aggiornato]
|
||||||
//per quelli che matchano cercare nel risultato se i context da aggiungere sono gia' presenti. Se non ci sono aggiungerli, altrimenti nulla
|
//per quelli che matchano cercare nel risultato se i context da aggiungere sono gia' presenti. Se non ci sono aggiungerli, altrimenti nulla
|
||||||
*/
|
*/
|
||||||
}
|
}
|
||||||
|
|
||||||
private static org.apache.spark.sql.Dataset<Row> getUpdateCommunitiesForTable(SparkSession spark, String table){
|
private static org.apache.spark.sql.Dataset<Row> getUpdateCommunitiesForTable(
|
||||||
String query = "SELECT target_id, collect_set(co.id) context_id " +
|
SparkSession spark, String table) {
|
||||||
" FROM (SELECT t.id target_id, s.context source_context " +
|
String query =
|
||||||
" FROM context_software s " +
|
"SELECT target_id, collect_set(co.id) context_id "
|
||||||
" JOIN " + table + " t " +
|
+ " FROM (SELECT t.id target_id, s.context source_context "
|
||||||
" ON s.target = t.id " +
|
+ " FROM context_software s "
|
||||||
" UNION ALL " +
|
+ " JOIN "
|
||||||
" SELECT t.id target_id, d.context source_context " +
|
+ table
|
||||||
" FROM dataset_context d " +
|
+ " t "
|
||||||
" JOIN " + table + " t" +
|
+ " ON s.target = t.id "
|
||||||
" ON s.target = t.id " +
|
+ " UNION ALL "
|
||||||
" UNION ALL " +
|
+ " SELECT t.id target_id, d.context source_context "
|
||||||
" SELECT t.id target_id, p.context source_context " +
|
+ " FROM dataset_context d "
|
||||||
" FROM publication_context p" +
|
+ " JOIN "
|
||||||
" JOIN " + table +" t " +
|
+ table
|
||||||
" on p.target = t.id " +
|
+ " t"
|
||||||
" UNION ALL " +
|
+ " ON s.target = t.id "
|
||||||
" SELECT t.id target_id, o.context source_context " +
|
+ " UNION ALL "
|
||||||
" FROM other_context o " +
|
+ " SELECT t.id target_id, p.context source_context "
|
||||||
" JOIN " + table + " t " +
|
+ " FROM publication_context p"
|
||||||
" ON o.target = t.id) TMP " +
|
+ " JOIN "
|
||||||
" LATERAL VIEW EXPLODE(source_context) MyT as co " +
|
+ table
|
||||||
" GROUP BY target_id" ;
|
+ " t "
|
||||||
|
+ " on p.target = t.id "
|
||||||
|
+ " UNION ALL "
|
||||||
|
+ " SELECT t.id target_id, o.context source_context "
|
||||||
|
+ " FROM other_context o "
|
||||||
|
+ " JOIN "
|
||||||
|
+ table
|
||||||
|
+ " t "
|
||||||
|
+ " ON o.target = t.id) TMP "
|
||||||
|
+ " LATERAL VIEW EXPLODE(source_context) MyT as co "
|
||||||
|
+ " GROUP BY target_id";
|
||||||
|
|
||||||
return spark.sql(query);
|
return spark.sql(query);
|
||||||
}
|
}
|
||||||
|
|
||||||
private static JavaRDD<Result> createUpdateForResultDatasetWrite(JavaRDD<Row> toupdateresult, String outputPath, String type, String class_id, String class_name, List<String> communityIdList){
|
private static JavaRDD<Result> createUpdateForResultDatasetWrite(
|
||||||
return toupdateresult.map(r -> {
|
JavaRDD<Row> toupdateresult,
|
||||||
List<Context> contextList = new ArrayList();
|
String outputPath,
|
||||||
List<String> toAddContext = r.getList(1);
|
String type,
|
||||||
for (String cId : toAddContext) {
|
String class_id,
|
||||||
if (communityIdList.contains(cId)) {
|
String class_name,
|
||||||
Context newContext = new Context();
|
List<String> communityIdList) {
|
||||||
newContext.setId(cId);
|
return toupdateresult
|
||||||
newContext.setDataInfo(Arrays.asList(getDataInfo(PROPAGATION_DATA_INFO_TYPE, class_id, class_name)));
|
.map(
|
||||||
contextList.add(newContext);
|
r -> {
|
||||||
}
|
List<Context> contextList = new ArrayList();
|
||||||
|
List<String> toAddContext = r.getList(1);
|
||||||
|
for (String cId : toAddContext) {
|
||||||
|
if (communityIdList.contains(cId)) {
|
||||||
|
Context newContext = new Context();
|
||||||
|
newContext.setId(cId);
|
||||||
|
newContext.setDataInfo(
|
||||||
|
Arrays.asList(
|
||||||
|
getDataInfo(
|
||||||
|
PROPAGATION_DATA_INFO_TYPE,
|
||||||
|
class_id,
|
||||||
|
class_name)));
|
||||||
|
contextList.add(newContext);
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
}
|
if (contextList.size() > 0) {
|
||||||
|
Result ret = new Result();
|
||||||
if (contextList.size() > 0) {
|
ret.setId(r.getString(0));
|
||||||
Result ret = new Result();
|
ret.setContext(contextList);
|
||||||
ret.setId(r.getString(0));
|
return ret;
|
||||||
ret.setContext(contextList);
|
}
|
||||||
return ret;
|
return null;
|
||||||
}
|
})
|
||||||
return null;
|
.filter(r -> r != null);
|
||||||
}).filter(r -> r != null);
|
|
||||||
}
|
}
|
||||||
|
|
||||||
private static void updateForSoftwareDataset(JavaRDD<Row> toupdateresult, JavaRDD<Software> result, String outputPath, String type, String class_id, String class_name, List<String> communityIdList){
|
private static void updateForSoftwareDataset(
|
||||||
JavaPairRDD<String, Result> tmp = result.mapToPair(r -> new Tuple2(r.getId(), r));
|
JavaRDD<Row> toupdateresult,
|
||||||
getUpdateForResultDataset(toupdateresult, tmp, outputPath, type, class_id, class_name, communityIdList)
|
JavaRDD<Software> result,
|
||||||
.map(r -> (Software) r)
|
String outputPath,
|
||||||
.map(s -> new ObjectMapper().writeValueAsString(s))
|
String type,
|
||||||
.saveAsTextFile(outputPath + "/" + type);
|
String class_id,
|
||||||
|
String class_name,
|
||||||
|
List<String> communityIdList) {
|
||||||
|
JavaPairRDD<String, Result> tmp = result.mapToPair(r -> new Tuple2(r.getId(), r));
|
||||||
|
getUpdateForResultDataset(
|
||||||
|
toupdateresult,
|
||||||
|
tmp,
|
||||||
|
outputPath,
|
||||||
|
type,
|
||||||
|
class_id,
|
||||||
|
class_name,
|
||||||
|
communityIdList)
|
||||||
|
.map(r -> (Software) r)
|
||||||
|
.map(s -> new ObjectMapper().writeValueAsString(s))
|
||||||
|
.saveAsTextFile(outputPath + "/" + type);
|
||||||
}
|
}
|
||||||
|
|
||||||
private static void updateForDatasetDataset(JavaRDD<Row> toupdateresult, JavaRDD<Dataset> result, String outputPath, String type, String class_id, String class_name, List<String> communityIdList){
|
private static void updateForDatasetDataset(
|
||||||
JavaPairRDD<String, Result> tmp = result.mapToPair(r -> new Tuple2(r.getId(), r));
|
JavaRDD<Row> toupdateresult,
|
||||||
getUpdateForResultDataset(toupdateresult, tmp, outputPath, type, class_id, class_name, communityIdList)
|
JavaRDD<Dataset> result,
|
||||||
.map( r-> (Dataset)r)
|
String outputPath,
|
||||||
|
String type,
|
||||||
|
String class_id,
|
||||||
|
String class_name,
|
||||||
|
List<String> communityIdList) {
|
||||||
|
JavaPairRDD<String, Result> tmp = result.mapToPair(r -> new Tuple2(r.getId(), r));
|
||||||
|
getUpdateForResultDataset(
|
||||||
|
toupdateresult,
|
||||||
|
tmp,
|
||||||
|
outputPath,
|
||||||
|
type,
|
||||||
|
class_id,
|
||||||
|
class_name,
|
||||||
|
communityIdList)
|
||||||
|
.map(r -> (Dataset) r)
|
||||||
.map(d -> new ObjectMapper().writeValueAsString(d))
|
.map(d -> new ObjectMapper().writeValueAsString(d))
|
||||||
.saveAsTextFile(outputPath + "/" + type);
|
.saveAsTextFile(outputPath + "/" + type);
|
||||||
}
|
}
|
||||||
|
|
||||||
private static void updateForPublicationDataset(JavaRDD<Row> toupdateresult, JavaRDD<Publication> result, String outputPath, String type, String class_id, String class_name, List<String> communityIdList){
|
private static void updateForPublicationDataset(
|
||||||
JavaPairRDD<String, Result> tmp = result.mapToPair(r -> new Tuple2(r.getId(), r));
|
JavaRDD<Row> toupdateresult,
|
||||||
getUpdateForResultDataset(toupdateresult, tmp, outputPath, type, class_id, class_name, communityIdList)
|
JavaRDD<Publication> result,
|
||||||
.map(r -> (Publication)r)
|
String outputPath,
|
||||||
.map(p -> new ObjectMapper().writeValueAsString(p))
|
String type,
|
||||||
.saveAsTextFile(outputPath + "/" + type);
|
String class_id,
|
||||||
|
String class_name,
|
||||||
|
List<String> communityIdList) {
|
||||||
|
JavaPairRDD<String, Result> tmp = result.mapToPair(r -> new Tuple2(r.getId(), r));
|
||||||
|
getUpdateForResultDataset(
|
||||||
|
toupdateresult,
|
||||||
|
tmp,
|
||||||
|
outputPath,
|
||||||
|
type,
|
||||||
|
class_id,
|
||||||
|
class_name,
|
||||||
|
communityIdList)
|
||||||
|
.map(r -> (Publication) r)
|
||||||
|
.map(p -> new ObjectMapper().writeValueAsString(p))
|
||||||
|
.saveAsTextFile(outputPath + "/" + type);
|
||||||
}
|
}
|
||||||
|
|
||||||
private static void updateForOtherDataset(JavaRDD<Row> toupdateresult, JavaRDD<OtherResearchProduct> result, String outputPath, String type, String class_id, String class_name, List<String> communityIdList){
|
private static void updateForOtherDataset(
|
||||||
JavaPairRDD<String, Result> tmp = result.mapToPair(r -> new Tuple2(r.getId(), r));
|
JavaRDD<Row> toupdateresult,
|
||||||
getUpdateForResultDataset(toupdateresult, tmp, outputPath, type, class_id, class_name, communityIdList)
|
JavaRDD<OtherResearchProduct> result,
|
||||||
.map( r -> (OtherResearchProduct)r)
|
String outputPath,
|
||||||
.map( o -> new ObjectMapper().writeValueAsString(o))
|
String type,
|
||||||
.saveAsTextFile(outputPath + "/" + type);
|
String class_id,
|
||||||
|
String class_name,
|
||||||
|
List<String> communityIdList) {
|
||||||
|
JavaPairRDD<String, Result> tmp = result.mapToPair(r -> new Tuple2(r.getId(), r));
|
||||||
|
getUpdateForResultDataset(
|
||||||
|
toupdateresult,
|
||||||
|
tmp,
|
||||||
|
outputPath,
|
||||||
|
type,
|
||||||
|
class_id,
|
||||||
|
class_name,
|
||||||
|
communityIdList)
|
||||||
|
.map(r -> (OtherResearchProduct) r)
|
||||||
|
.map(o -> new ObjectMapper().writeValueAsString(o))
|
||||||
|
.saveAsTextFile(outputPath + "/" + type);
|
||||||
}
|
}
|
||||||
|
|
||||||
|
private static JavaRDD<Result> getUpdateForResultDataset(
|
||||||
|
JavaRDD<Row> toupdateresult,
|
||||||
|
JavaPairRDD<String, Result> result,
|
||||||
|
String outputPath,
|
||||||
|
String type,
|
||||||
|
String class_id,
|
||||||
|
String class_name,
|
||||||
|
List<String> communityIdList) {
|
||||||
|
return result.leftOuterJoin(
|
||||||
|
toupdateresult.mapToPair(r -> new Tuple2<>(r.getString(0), r.getList(1))))
|
||||||
|
.map(
|
||||||
|
c -> {
|
||||||
|
if (!c._2()._2().isPresent()) {
|
||||||
|
return c._2()._1();
|
||||||
|
}
|
||||||
|
|
||||||
|
List<Object> toAddContext = c._2()._2().get();
|
||||||
private static JavaRDD<Result> getUpdateForResultDataset(JavaRDD<Row> toupdateresult, JavaPairRDD<String, Result> result, String outputPath, String type, String class_id, String class_name, List<String> communityIdList){
|
Set<String> context_set = new HashSet<>();
|
||||||
return result.leftOuterJoin(toupdateresult.mapToPair(r -> new Tuple2<>(r.getString(0), r.getList(1))))
|
for (Object cId : toAddContext) {
|
||||||
.map(c -> {
|
String id = (String) cId;
|
||||||
if(! c._2()._2().isPresent()){
|
if (communityIdList.contains(id)) {
|
||||||
return c._2()._1();
|
context_set.add(id);
|
||||||
}
|
}
|
||||||
|
}
|
||||||
List<Object> toAddContext = c._2()._2().get();
|
for (Context context : c._2()._1().getContext()) {
|
||||||
Set<String> context_set = new HashSet<>();
|
if (context_set.contains(context)) {
|
||||||
for(Object cId: toAddContext){
|
context_set.remove(context);
|
||||||
String id = (String)cId;
|
|
||||||
if (communityIdList.contains(id)){
|
|
||||||
context_set.add(id);
|
|
||||||
}
|
|
||||||
}
|
|
||||||
for (Context context: c._2()._1().getContext()){
|
|
||||||
if(context_set.contains(context)){
|
|
||||||
context_set.remove(context);
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
List<Context> contextList = context_set.stream().map(co -> {
|
|
||||||
Context newContext = new Context();
|
|
||||||
newContext.setId(co);
|
|
||||||
newContext.setDataInfo(Arrays.asList(getDataInfo(PROPAGATION_DATA_INFO_TYPE, class_id, class_name)));
|
|
||||||
return newContext;
|
|
||||||
|
|
||||||
}).collect(Collectors.toList());
|
|
||||||
|
|
||||||
if(contextList.size() > 0 ){
|
|
||||||
Result r = new Result();
|
|
||||||
r.setId(c._1());
|
|
||||||
r.setContext(contextList);
|
|
||||||
return r;
|
|
||||||
}
|
|
||||||
return null;
|
|
||||||
}).filter(r -> r != null);
|
|
||||||
|
|
||||||
|
|
||||||
// return toupdateresult.mapToPair(r -> new Tuple2<>(r.getString(0), r.getList(1)))
|
|
||||||
// .join(result)
|
|
||||||
// .map(c -> {
|
|
||||||
// List<Object> toAddContext = c._2()._1();
|
|
||||||
// Set<String> context_set = new HashSet<>();
|
|
||||||
// for(Object cId: toAddContext){
|
|
||||||
// String id = (String)cId;
|
|
||||||
// if (communityIdList.contains(id)){
|
|
||||||
// context_set.add(id);
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
// for (Context context: c._2()._2().getContext()){
|
|
||||||
// if(context_set.contains(context)){
|
|
||||||
// context_set.remove(context);
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
//
|
|
||||||
// List<Context> contextList = context_set.stream().map(co -> {
|
|
||||||
// Context newContext = new Context();
|
|
||||||
// newContext.setId(co);
|
|
||||||
// newContext.setDataInfo(Arrays.asList(getDataInfo(PROPAGATION_DATA_INFO_TYPE, class_id, class_name)));
|
|
||||||
// return newContext;
|
|
||||||
//
|
|
||||||
// }).collect(Collectors.toList());
|
|
||||||
//
|
|
||||||
// if(contextList.size() > 0 ){
|
|
||||||
// Result r = new Result();
|
|
||||||
// r.setId(c._1());
|
|
||||||
// r.setContext(contextList);
|
|
||||||
// return r;
|
|
||||||
// }
|
|
||||||
// return null;
|
|
||||||
// })
|
|
||||||
// .filter(r -> r != null);
|
|
||||||
}
|
|
||||||
|
|
||||||
private static JavaRDD<Software> createUpdateForSoftwareDataset(JavaRDD<Row> toupdateresult, List<String> communityList,
|
|
||||||
JavaRDD<Software> result, String class_id, String class_name) {
|
|
||||||
return result
|
|
||||||
.mapToPair(s -> new Tuple2<>(s.getId(), s)).leftOuterJoin(getStringResultJavaPairRDD(toupdateresult, communityList))
|
|
||||||
.map(c -> {
|
|
||||||
Software oaf = c._2()._1();
|
|
||||||
if (c._2()._2().isPresent()) {
|
|
||||||
|
|
||||||
HashSet<String> contexts = new HashSet<>(c._2()._2().get());
|
|
||||||
|
|
||||||
for (Context context : oaf.getContext()) {
|
|
||||||
if (contexts.contains(context.getId())){
|
|
||||||
if (!context.getDataInfo().stream().map(di -> di.getInferenceprovenance())
|
|
||||||
.collect(Collectors.toSet()).contains(PROPAGATION_DATA_INFO_TYPE)){
|
|
||||||
context.getDataInfo().add(getDataInfo(PROPAGATION_DATA_INFO_TYPE, class_id, class_name));
|
|
||||||
//community id already in the context of the result. Remove it from the set that has to be added
|
|
||||||
contexts.remove(context.getId());
|
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
}
|
|
||||||
List<Context> cc = oaf.getContext();
|
|
||||||
for(String cId: contexts){
|
|
||||||
Context context = new Context();
|
|
||||||
context.setId(cId);
|
|
||||||
context.setDataInfo(Arrays.asList(getDataInfo(PROPAGATION_DATA_INFO_TYPE, class_id, class_name)));
|
|
||||||
cc.add(context);
|
|
||||||
}
|
|
||||||
oaf.setContext(cc);
|
|
||||||
|
|
||||||
|
List<Context> contextList =
|
||||||
|
context_set.stream()
|
||||||
|
.map(
|
||||||
|
co -> {
|
||||||
|
Context newContext = new Context();
|
||||||
|
newContext.setId(co);
|
||||||
|
newContext.setDataInfo(
|
||||||
|
Arrays.asList(
|
||||||
|
getDataInfo(
|
||||||
|
PROPAGATION_DATA_INFO_TYPE,
|
||||||
|
class_id,
|
||||||
|
class_name)));
|
||||||
|
return newContext;
|
||||||
|
})
|
||||||
|
.collect(Collectors.toList());
|
||||||
|
|
||||||
|
if (contextList.size() > 0) {
|
||||||
|
Result r = new Result();
|
||||||
|
r.setId(c._1());
|
||||||
|
r.setContext(contextList);
|
||||||
|
return r;
|
||||||
|
}
|
||||||
|
return null;
|
||||||
|
})
|
||||||
|
.filter(r -> r != null);
|
||||||
|
|
||||||
|
// return toupdateresult.mapToPair(r -> new Tuple2<>(r.getString(0), r.getList(1)))
|
||||||
|
// .join(result)
|
||||||
|
// .map(c -> {
|
||||||
|
// List<Object> toAddContext = c._2()._1();
|
||||||
|
// Set<String> context_set = new HashSet<>();
|
||||||
|
// for(Object cId: toAddContext){
|
||||||
|
// String id = (String)cId;
|
||||||
|
// if (communityIdList.contains(id)){
|
||||||
|
// context_set.add(id);
|
||||||
|
// }
|
||||||
|
// }
|
||||||
|
// for (Context context: c._2()._2().getContext()){
|
||||||
|
// if(context_set.contains(context)){
|
||||||
|
// context_set.remove(context);
|
||||||
|
// }
|
||||||
|
// }
|
||||||
|
//
|
||||||
|
// List<Context> contextList = context_set.stream().map(co -> {
|
||||||
|
// Context newContext = new Context();
|
||||||
|
// newContext.setId(co);
|
||||||
|
//
|
||||||
|
// newContext.setDataInfo(Arrays.asList(getDataInfo(PROPAGATION_DATA_INFO_TYPE, class_id,
|
||||||
|
// class_name)));
|
||||||
|
// return newContext;
|
||||||
|
//
|
||||||
|
// }).collect(Collectors.toList());
|
||||||
|
//
|
||||||
|
// if(contextList.size() > 0 ){
|
||||||
|
// Result r = new Result();
|
||||||
|
// r.setId(c._1());
|
||||||
|
// r.setContext(contextList);
|
||||||
|
// return r;
|
||||||
|
// }
|
||||||
|
// return null;
|
||||||
|
// })
|
||||||
|
// .filter(r -> r != null);
|
||||||
|
}
|
||||||
|
|
||||||
|
private static JavaRDD<Software> createUpdateForSoftwareDataset(
|
||||||
|
JavaRDD<Row> toupdateresult,
|
||||||
|
List<String> communityList,
|
||||||
|
JavaRDD<Software> result,
|
||||||
|
String class_id,
|
||||||
|
String class_name) {
|
||||||
|
return result.mapToPair(s -> new Tuple2<>(s.getId(), s))
|
||||||
|
.leftOuterJoin(getStringResultJavaPairRDD(toupdateresult, communityList))
|
||||||
|
.map(
|
||||||
|
c -> {
|
||||||
|
Software oaf = c._2()._1();
|
||||||
|
if (c._2()._2().isPresent()) {
|
||||||
|
|
||||||
|
HashSet<String> contexts = new HashSet<>(c._2()._2().get());
|
||||||
|
|
||||||
|
for (Context context : oaf.getContext()) {
|
||||||
|
if (contexts.contains(context.getId())) {
|
||||||
|
if (!context.getDataInfo().stream()
|
||||||
|
.map(di -> di.getInferenceprovenance())
|
||||||
|
.collect(Collectors.toSet())
|
||||||
|
.contains(PROPAGATION_DATA_INFO_TYPE)) {
|
||||||
|
context.getDataInfo()
|
||||||
|
.add(
|
||||||
|
getDataInfo(
|
||||||
|
PROPAGATION_DATA_INFO_TYPE,
|
||||||
|
class_id,
|
||||||
|
class_name));
|
||||||
|
// community id already in the context of the result.
|
||||||
|
// Remove it from the set that has to be added
|
||||||
|
contexts.remove(context.getId());
|
||||||
|
}
|
||||||
|
}
|
||||||
|
}
|
||||||
|
List<Context> cc = oaf.getContext();
|
||||||
|
for (String cId : contexts) {
|
||||||
|
Context context = new Context();
|
||||||
|
context.setId(cId);
|
||||||
|
context.setDataInfo(
|
||||||
|
Arrays.asList(
|
||||||
|
getDataInfo(
|
||||||
|
PROPAGATION_DATA_INFO_TYPE,
|
||||||
|
class_id,
|
||||||
|
class_name)));
|
||||||
|
cc.add(context);
|
||||||
|
}
|
||||||
|
oaf.setContext(cc);
|
||||||
|
}
|
||||||
|
return oaf;
|
||||||
|
});
|
||||||
|
}
|
||||||
|
|
||||||
|
private static JavaPairRDD<String, List<String>> getStringResultJavaPairRDD(
|
||||||
|
JavaRDD<Row> toupdateresult, List<String> communityList) {
|
||||||
|
return toupdateresult.mapToPair(
|
||||||
|
c -> {
|
||||||
|
List<String> contextList = new ArrayList<>();
|
||||||
|
List<String> contexts = c.getList(1);
|
||||||
|
for (String context : contexts) {
|
||||||
|
if (communityList.contains(context)) {
|
||||||
|
contextList.add(context);
|
||||||
|
}
|
||||||
}
|
}
|
||||||
return oaf;
|
|
||||||
|
return new Tuple2<>(c.getString(0), contextList);
|
||||||
});
|
});
|
||||||
}
|
}
|
||||||
|
|
||||||
private static JavaPairRDD<String, List<String>> getStringResultJavaPairRDD(JavaRDD<Row> toupdateresult, List<String> communityList) {
|
private static org.apache.spark.sql.Dataset<Row> getContext(SparkSession spark, String table) {
|
||||||
return toupdateresult.mapToPair(c -> {
|
String query =
|
||||||
|
"SELECT relation.source, "
|
||||||
List<String> contextList = new ArrayList<>();
|
+ table
|
||||||
List<String> contexts = c.getList(1);
|
+ ".context , relation.target "
|
||||||
for (String context : contexts) {
|
+ "FROM "
|
||||||
if (communityList.contains(context)) {
|
+ table
|
||||||
contextList.add(context);
|
+ " JOIN relation "
|
||||||
}
|
+ "ON id = source";
|
||||||
}
|
|
||||||
|
|
||||||
return new Tuple2<>(c.getString(0) ,contextList);
|
|
||||||
});
|
|
||||||
}
|
|
||||||
|
|
||||||
|
|
||||||
private static org.apache.spark.sql.Dataset<Row> getContext(SparkSession spark, String table){
|
|
||||||
String query = "SELECT relation.source, " + table +".context , relation.target " +
|
|
||||||
"FROM " + table +
|
|
||||||
" JOIN relation " +
|
|
||||||
"ON id = source" ;
|
|
||||||
|
|
||||||
return spark.sql(query);
|
return spark.sql(query);
|
||||||
}
|
}
|
||||||
|
|
||||||
private static Boolean relatedToCommunities(Result r, List<String> communityIdList) {
|
private static Boolean relatedToCommunities(Result r, List<String> communityIdList) {
|
||||||
Set<String> result_communities = r.getContext()
|
Set<String> result_communities =
|
||||||
.stream()
|
r.getContext().stream().map(c -> c.getId()).collect(Collectors.toSet());
|
||||||
.map(c -> c.getId())
|
|
||||||
.collect(Collectors.toSet());
|
|
||||||
for (String communityId : result_communities) {
|
for (String communityId : result_communities) {
|
||||||
if (communityIdList.contains(communityId)) {
|
if (communityIdList.contains(communityId)) {
|
||||||
return true;
|
return true;
|
||||||
|
|
@ -419,44 +589,61 @@ public class SparkResultToCommunityThroughSemRelJob3 {
|
||||||
return false;
|
return false;
|
||||||
}
|
}
|
||||||
|
|
||||||
private static void updateResult(JavaPairRDD<String, Result> results, JavaPairRDD<String, TypedRow> toupdateresult, String outputPath, String type) {
|
private static void updateResult(
|
||||||
|
JavaPairRDD<String, Result> results,
|
||||||
|
JavaPairRDD<String, TypedRow> toupdateresult,
|
||||||
|
String outputPath,
|
||||||
|
String type) {
|
||||||
results.leftOuterJoin(toupdateresult)
|
results.leftOuterJoin(toupdateresult)
|
||||||
.map(p -> {
|
.map(
|
||||||
Result r = p._2()._1();
|
p -> {
|
||||||
if (p._2()._2().isPresent()){
|
Result r = p._2()._1();
|
||||||
Set<String> communityList = p._2()._2().get().getAccumulator();
|
if (p._2()._2().isPresent()) {
|
||||||
for(Context c: r.getContext()){
|
Set<String> communityList = p._2()._2().get().getAccumulator();
|
||||||
if (communityList.contains(c.getId())){
|
for (Context c : r.getContext()) {
|
||||||
//verify if the datainfo for this context contains propagation
|
if (communityList.contains(c.getId())) {
|
||||||
if (!c.getDataInfo().stream().map(di -> di.getInferenceprovenance()).collect(Collectors.toSet()).contains(PROPAGATION_DATA_INFO_TYPE)){
|
// verify if the datainfo for this context contains
|
||||||
c.getDataInfo().add(getDataInfo(PROPAGATION_DATA_INFO_TYPE, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME));
|
// propagation
|
||||||
//community id already in the context of the result. Remove it from the set that has to be added
|
if (!c.getDataInfo().stream()
|
||||||
communityList.remove(c.getId());
|
.map(di -> di.getInferenceprovenance())
|
||||||
|
.collect(Collectors.toSet())
|
||||||
|
.contains(PROPAGATION_DATA_INFO_TYPE)) {
|
||||||
|
c.getDataInfo()
|
||||||
|
.add(
|
||||||
|
getDataInfo(
|
||||||
|
PROPAGATION_DATA_INFO_TYPE,
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID,
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME));
|
||||||
|
// community id already in the context of the result.
|
||||||
|
// Remove it from the set that has to be added
|
||||||
|
communityList.remove(c.getId());
|
||||||
|
}
|
||||||
|
}
|
||||||
}
|
}
|
||||||
|
List<Context> cc = r.getContext();
|
||||||
|
for (String cId : communityList) {
|
||||||
|
Context context = new Context();
|
||||||
|
context.setId(cId);
|
||||||
|
context.setDataInfo(
|
||||||
|
Arrays.asList(
|
||||||
|
getDataInfo(
|
||||||
|
PROPAGATION_DATA_INFO_TYPE,
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID,
|
||||||
|
PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME)));
|
||||||
|
cc.add(context);
|
||||||
|
}
|
||||||
|
r.setContext(cc);
|
||||||
}
|
}
|
||||||
}
|
return r;
|
||||||
List<Context> cc = r.getContext();
|
})
|
||||||
for(String cId: communityList){
|
|
||||||
Context context = new Context();
|
|
||||||
context.setId(cId);
|
|
||||||
context.setDataInfo(Arrays.asList(getDataInfo(PROPAGATION_DATA_INFO_TYPE, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_ID, PROPAGATION_RESULT_COMMUNITY_SEMREL_CLASS_NAME)));
|
|
||||||
cc.add(context);
|
|
||||||
}
|
|
||||||
r.setContext(cc);
|
|
||||||
}
|
|
||||||
return r;
|
|
||||||
})
|
|
||||||
.map(p -> new ObjectMapper().writeValueAsString(p))
|
.map(p -> new ObjectMapper().writeValueAsString(p))
|
||||||
.saveAsTextFile(outputPath+"/"+type);
|
.saveAsTextFile(outputPath + "/" + type);
|
||||||
}
|
}
|
||||||
|
|
||||||
|
private static TypedRow getTypedRow(
|
||||||
|
List<String> communityIdList, List<Context> context, String id, String type) {
|
||||||
private static TypedRow getTypedRow(List<String> communityIdList, List<Context> context, String id, String type) {
|
Set<String> result_communities =
|
||||||
Set<String> result_communities = context
|
context.stream().map(c -> c.getId()).collect(Collectors.toSet());
|
||||||
.stream()
|
|
||||||
.map(c -> c.getId())
|
|
||||||
.collect(Collectors.toSet());
|
|
||||||
TypedRow tp = new TypedRow();
|
TypedRow tp = new TypedRow();
|
||||||
tp.setSourceId(id);
|
tp.setSourceId(id);
|
||||||
tp.setType(type);
|
tp.setType(type);
|
||||||
|
|
|
||||||
Loading…
Reference in New Issue