D-Net
/
dnet-hadoop
14
5
Fork 11
Code Issues 14 Pull Requests 12 Projects 2 Releases 2 Wiki Activity

[enrichment] adding relations in one side only

This commit is contained in:
Miriam Baglioni 2023-02-15 16:41:28 +01:00
parent 624c62f62d
commit eca7ebab6d
3 changed files with 17 additions and 41 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

35
dhp-workflows/dhp-enrichment/src/main/java/eu/dnetlib/dhp/projecttoresult/SparkResultToProjectThroughSemRelJob.java
View File

@ -107,30 +107,17 @@ public class SparkResultToProjectThroughSemRelJob {
potentialUpdate
.getProjectSet()
.forEach(
projectId -> {
newRelations
.add(
getRelation(
resId,
projectId,
ModelConstants.IS_PRODUCED_BY,
ModelConstants.RESULT_PROJECT,
ModelConstants.OUTCOME,
PROPAGATION_DATA_INFO_TYPE,
PROPAGATION_RELATION_RESULT_PROJECT_SEM_REL_CLASS_ID,
PROPAGATION_RELATION_RESULT_PROJECT_SEM_REL_CLASS_NAME));
newRelations
.add(
getRelation(
projectId,
resId,
ModelConstants.PRODUCES,
ModelConstants.RESULT_PROJECT,
ModelConstants.OUTCOME,
PROPAGATION_DATA_INFO_TYPE,
PROPAGATION_RELATION_RESULT_PROJECT_SEM_REL_CLASS_ID,
PROPAGATION_RELATION_RESULT_PROJECT_SEM_REL_CLASS_NAME));
});
projectId -> newRelations
.add(
getRelation(
projectId,
resId,
ModelConstants.PRODUCES,
ModelConstants.RESULT_PROJECT,
ModelConstants.OUTCOME,
PROPAGATION_DATA_INFO_TYPE,
PROPAGATION_RELATION_RESULT_PROJECT_SEM_REL_CLASS_ID,
PROPAGATION_RELATION_RESULT_PROJECT_SEM_REL_CLASS_NAME)));
return newRelations.iterator();
};
}

10
dhp-workflows/dhp-enrichment/src/main/java/eu/dnetlib/dhp/resulttoorganizationfrominstrepo/SparkResultToOrganizationFromIstRepoJob.java
View File

@ -139,10 +139,12 @@ public class SparkResultToOrganizationFromIstRepoJob {
organizations
.forEach(
orgId -> newRelations
.addAll(
getOrganizationRelationPair(
orgId,
resultId,
.add(
getRelation(
resultId, orgId,
ModelConstants.HAS_AUTHOR_INSTITUTION,
ModelConstants.RESULT_ORGANIZATION,
ModelConstants.AFFILIATION, PROPAGATION_DATA_INFO_TYPE,
PROPAGATION_RELATION_RESULT_ORGANIZATION_INST_REPO_CLASS_ID,
PROPAGATION_RELATION_RESULT_ORGANIZATION_INST_REPO_CLASS_NAME))

13
dhp-workflows/dhp-enrichment/src/main/java/eu/dnetlib/dhp/resulttoorganizationfromsemrel/SparkResultToOrganizationFromSemRel.java
View File

@ -211,19 +211,6 @@ public class SparkResultToOrganizationFromSemRel implements Serializable {
.groupByKey((MapFunction<Relation, String>) r -> r.getSource() + r.getTarget(), Encoders.STRING())
.mapGroups(
(MapGroupsFunction<String, Relation, Relation>) (k, it) -> it.next(), Encoders.bean(Relation.class))
.flatMap(
(FlatMapFunction<Relation, Relation>) r -> Arrays
.asList(
r, getRelation(
r.getTarget(), r.getSource(), ModelConstants.IS_AUTHOR_INSTITUTION_OF,
ModelConstants.RESULT_ORGANIZATION,
ModelConstants.AFFILIATION,
PROPAGATION_DATA_INFO_TYPE,
PROPAGATION_RELATION_RESULT_ORGANIZATION_SEM_REL_CLASS_ID,
PROPAGATION_RELATION_RESULT_ORGANIZATION_SEM_REL_CLASS_NAME))
.iterator()
, Encoders.bean(Relation.class))
.write()
.mode(SaveMode.Append)