sabeel
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Pull Requests Activity

[cleaning] deactivating the cleaning of FOS subjects found in the metadata provided by repositories

This commit is contained in:
Claudio Atzori 2024-05-03 14:31:12 +02:00
parent a5d13d5d27
commit 0486227185
1 changed files with 28 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

38
dhp-workflows/dhp-graph-mapper/src/main/java/eu/dnetlib/dhp/oa/graph/clean/CleaningRuleMap.java
View File

@ -4,6 +4,7 @@ package eu.dnetlib.dhp.oa.graph.clean;
import java.io.Serializable; import java.io.Serializable;
import java.util.HashMap; import java.util.HashMap;
import java.util.Objects; import java.util.Objects;
import java.util.Optional;
import java.util.concurrent.atomic.AtomicReference; import java.util.concurrent.atomic.AtomicReference;
import org.apache.commons.lang3.SerializationUtils; import org.apache.commons.lang3.SerializationUtils;
@ -29,7 +30,10 @@ public class CleaningRuleMap extends HashMap<Class<?>, SerializableConsumer<Obje
mapping.put(AccessRight.class, o -> cleanQualifier(vocabularies, (AccessRight) o)); mapping.put(AccessRight.class, o -> cleanQualifier(vocabularies, (AccessRight) o));
mapping.put(Country.class, o -> cleanCountry(vocabularies, (Country) o)); mapping.put(Country.class, o -> cleanCountry(vocabularies, (Country) o));
mapping.put(Relation.class, o -> cleanRelation(vocabularies, (Relation) o)); mapping.put(Relation.class, o -> cleanRelation(vocabularies, (Relation) o));
mapping.put(Subject.class, o -> cleanSubject(vocabularies, (Subject) o));
// commenting out the subject cleaning until we decide if we want to it or not and the implementation will
// be completed. At the moment it is not capable of expanding the whole hierarchy.
// mapping.put(Subject.class, o -> cleanSubject(vocabularies, (Subject) o));
return mapping; return mapping;
} }
@ -38,8 +42,15 @@ public class CleaningRuleMap extends HashMap<Class<?>, SerializableConsumer<Obje
// TODO cleaning based on different subject vocabs can be added here // TODO cleaning based on different subject vocabs can be added here
} }
/**
* The procedure cleans out the subject values, using a vocabulary identified by the field subject.qualifier.classid.
*
* @param vocabularyId
* @param vocabularies
* @param subject
*/
private static void cleanSubjectForVocabulary(String vocabularyId, VocabularyGroup vocabularies, private static void cleanSubjectForVocabulary(String vocabularyId, VocabularyGroup vocabularies,
Subject subject) { Subject subject) {
vocabularies.find(vocabularyId).ifPresent(vocabulary -> { vocabularies.find(vocabularyId).ifPresent(vocabulary -> {
if (ModelConstants.DNET_SUBJECT_KEYWORD.equalsIgnoreCase(subject.getQualifier().getClassid())) { if (ModelConstants.DNET_SUBJECT_KEYWORD.equalsIgnoreCase(subject.getQualifier().getClassid())) {
@ -49,14 +60,21 @@ public class CleaningRuleMap extends HashMap<Class<?>, SerializableConsumer<Obje
subject.getQualifier().setClassid(vocabularyId); subject.getQualifier().setClassid(vocabularyId);
subject.getQualifier().setClassname(vocabulary.getName()); subject.getQualifier().setClassname(vocabulary.getName());
} }
} else if (vocabularyId.equals(subject.getQualifier().getClassid()) && } else {
Objects.nonNull(subject.getDataInfo()) && final String provenanceActionClassId = Optional.ofNullable(subject.getDataInfo())
!"subject:fos".equals(subject.getDataInfo().getProvenanceaction())) { .map(DataInfo::getProvenanceaction)
Qualifier syn = vocabulary.getSynonymAsQualifier(subject.getValue()); .map(Qualifier::getClassid)
VocabularyTerm term = vocabulary.getTerm(subject.getValue()); .orElse(null);
if (Objects.isNull(syn) && Objects.isNull(term)) {
subject.getQualifier().setClassid(ModelConstants.DNET_SUBJECT_KEYWORD); if (vocabularyId.equals(subject.getQualifier().getClassid()) &&
subject.getQualifier().setClassname(ModelConstants.DNET_SUBJECT_KEYWORD); !"subject:fos".equals(provenanceActionClassId)) {
Qualifier syn = vocabulary.getSynonymAsQualifier(subject.getValue());
VocabularyTerm term = vocabulary.getTerm(subject.getValue());
if (Objects.isNull(syn) && Objects.isNull(term)) {
subject.getQualifier().setClassid(ModelConstants.DNET_SUBJECT_KEYWORD);
subject.getQualifier().setClassname(ModelConstants.DNET_SUBJECT_KEYWORD);
}
} }
} }
}); });