Skip to content

Commit

Permalink
Merged in CST-11738 (pull request DSpace#1129)
Browse files Browse the repository at this point in the history 
[CST-11738] TUHH: ORCID lookup with more data

Approved-by: Stefano Maffei
  • Loading branch information
@eskander17 @steph-ieffam
eskander17 authored and steph-ieffam committed Oct 25, 2023
2 parents 1fcd95e + eeadc46 commit c060a18
Show file tree
Hide file tree
Showing 13 changed files with 691 additions and 33 deletions.
7 changes: 5 additions & 2 deletions dspace-api/src/main/java/org/dspace/authority/filler/ExternalDataProviderImportFiller.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -7,7 +7,6 @@
*/
package org.dspace.authority.filler;

import static org.apache.commons.collections.CollectionUtils.isEmpty;
import static org.apache.commons.lang3.StringUtils.isBlank;
import static org.apache.commons.lang3.StringUtils.removeStart;
import static org.apache.commons.lang3.StringUtils.startsWith;
Expand Down Expand Up @@ -110,7 +109,11 @@ private void enrichItemWithExternalData(Context context, Item item, ExternalData
}

private boolean notAlreadyPresent(Item item, MetadataValueDTO value) {
return isEmpty(itemService.getMetadata(item, value.getSchema(), value.getElement(), value.getQualifier(), ANY));
List<MetadataValue> metadataValues = itemService.getMetadata(item, value.getSchema(),
value.getElement(), value.getQualifier(), ANY);

return metadataValues.stream().noneMatch(metadataValue ->
metadataValue.getValue().equals(value.getValue()));
}

private boolean isTitleNotSet(Item item) {
Expand Down
81 changes: 70 additions & 11 deletions dspace-api/src/main/java/org/dspace/external/provider/impl/OrcidV3AuthorDataProvider.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -16,6 +16,7 @@
import java.util.Collections;
import java.util.LinkedList;
import java.util.List;
import java.util.Map;
import java.util.Objects;
import java.util.Optional;
import java.util.stream.Collectors;
Expand All @@ -27,6 +28,7 @@
import org.apache.http.impl.client.HttpClientBuilder;
import org.apache.logging.log4j.LogManager;
import org.apache.logging.log4j.Logger;
import org.dspace.content.MetadataFieldName;
import org.dspace.content.dto.MetadataValueDTO;
import org.dspace.external.OrcidRestConnector;
import org.dspace.external.model.ExternalDataObject;
Expand All @@ -35,6 +37,7 @@
import org.json.JSONObject;
import org.orcid.jaxb.model.v3.release.common.OrcidIdentifier;
import org.orcid.jaxb.model.v3.release.record.Person;
import org.orcid.jaxb.model.v3.release.record.Record;
import org.orcid.jaxb.model.v3.release.search.Result;
import org.springframework.beans.factory.annotation.Autowired;

Expand All @@ -60,6 +63,8 @@ public class OrcidV3AuthorDataProvider extends AbstractExternalDataProvider {

private XMLtoBio converter;

private Map<String, String> externalIdentifiers;

public static final String ORCID_ID_SYNTAX = "\\d{4}-\\d{4}-\\d{4}-(\\d{3}X|\\d{4})";
private static final int MAX_INDEX = 10000;

Expand Down Expand Up @@ -113,12 +118,13 @@ public void init() throws IOException {

@Override
public Optional<ExternalDataObject> getExternalDataObject(String id) {
Person person = getBio(id);
ExternalDataObject externalDataObject = convertToExternalDataObject(person);
Record record = getBio(id);
ExternalDataObject externalDataObject = convertToExternalDataObject(record);
return Optional.of(externalDataObject);
}

protected ExternalDataObject convertToExternalDataObject(Person person) {
protected ExternalDataObject convertToExternalDataObject(Record record) {
Person person = record.getPerson();
ExternalDataObject externalDataObject = new ExternalDataObject(sourceIdentifier);
if (person.getName() != null) {
String lastName = "";
Expand All @@ -141,6 +147,12 @@ protected ExternalDataObject convertToExternalDataObject(Person person) {
externalDataObject
.addMetadata(new MetadataValueDTO("dc", "identifier", "uri", null,
orcidUrl + '/' + person.getName().getPath()));

appendOtherNames(externalDataObject, person);
appendResearcherUrls(externalDataObject, person);
appendExternalIdentifiers(externalDataObject, person);
appendAffiliations(externalDataObject, record);

if (!StringUtils.isBlank(lastName) && !StringUtils.isBlank(firstName)) {
externalDataObject.setDisplayValue(lastName + ", " + firstName);
externalDataObject.setValue(lastName + ", " + firstName);
Expand All @@ -157,24 +169,64 @@ protected ExternalDataObject convertToExternalDataObject(Person person) {
return externalDataObject;
}

private void appendOtherNames(ExternalDataObject externalDataObject, Person person) {
person.getOtherNames().getOtherNames().forEach(otherName ->
externalDataObject.addMetadata(new MetadataValueDTO("crisrp", "name", "variant", null,
otherName.getContent())));
}

private void appendResearcherUrls(ExternalDataObject externalDataObject, Person person) {
person.getResearcherUrls().getResearcherUrls().forEach(researcherUrl ->
externalDataObject.addMetadata(new MetadataValueDTO("oairecerif", "identifier", "url", null,
researcherUrl.getUrl().getValue())));
}

private void appendExternalIdentifiers(ExternalDataObject externalDataObject, Person person) {
if (getExternalIdentifiers() != null) {
person.getExternalIdentifiers()
.getExternalIdentifiers()
.forEach(externalIdentifier -> {
String metadataField = externalIdentifiers.get(externalIdentifier.getType());
if (StringUtils.isNotEmpty(metadataField)) {
MetadataFieldName field = new MetadataFieldName(metadataField);
externalDataObject.addMetadata(
new MetadataValueDTO(field.schema, field.element, field.qualifier, null,
externalIdentifier.getValue()));
}
});
}
}

private void appendAffiliations(ExternalDataObject externalDataObject, Record record) {
record.getActivitiesSummary()
.getEmployments()
.getEmploymentGroups()
.stream()
.flatMap(affiliationGroup ->
affiliationGroup.getActivities().stream())
.forEach(employmentSummary ->
externalDataObject.addMetadata(new MetadataValueDTO("person", "affiliation", "name",
null, employmentSummary.getOrganization().getName())));
}

/**
* Retrieve a Person object based on a given orcid identifier.
* Retrieve a Record object based on a given orcid identifier.
* @param id orcid identifier
* @return Person
* @return Record
*/
public Person getBio(String id) {
public Record getBio(String id) {
log.debug("getBio called with ID=" + id);
if (!isValid(id)) {
return null;
}
InputStream bioDocument = orcidRestConnector.get(id + ((id.endsWith("/person")) ? "" : "/person"), accessToken);
Person person = converter.convertSinglePerson(bioDocument);
InputStream bioDocument = orcidRestConnector.get(id, accessToken);
Record record = converter.convertToRecord(bioDocument);
try {
bioDocument.close();
} catch (IOException e) {
log.error(e.getMessage(), e);
}
return person;
return record;
}

/**
Expand All @@ -201,13 +253,13 @@ public List<ExternalDataObject> searchExternalDataObjects(String query, int star
log.debug("queryBio searchPath=" + searchPath + " accessToken=" + accessToken);
InputStream bioDocument = orcidRestConnector.get(searchPath, accessToken);
List<Result> results = converter.convert(bioDocument);
List<Person> bios = new LinkedList<>();
List<Record> bios = new LinkedList<>();
for (Result result : results) {
OrcidIdentifier orcidIdentifier = result.getOrcidIdentifier();
if (orcidIdentifier != null) {
log.debug("Found OrcidId=" + orcidIdentifier.toString());
String orcid = orcidIdentifier.getPath();
Person bio = getBio(orcid);
Record bio = getBio(orcid);
if (bio != null) {
bios.add(bio);
}
Expand Down Expand Up @@ -298,4 +350,11 @@ public void setOrcidRestConnector(OrcidRestConnector orcidRestConnector) {
this.orcidRestConnector = orcidRestConnector;
}

public Map<String, String> getExternalIdentifiers() {
return externalIdentifiers;
}

public void setExternalIdentifiers(Map<String, String> externalIdentifiers) {
this.externalIdentifiers = externalIdentifiers;
}
}
12 changes: 12 additions & 0 deletions dspace-api/src/main/java/org/dspace/external/provider/orcid/xml/XMLtoBio.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -14,6 +14,7 @@

import org.apache.logging.log4j.Logger;
import org.orcid.jaxb.model.v3.release.record.Person;
import org.orcid.jaxb.model.v3.release.record.Record;
import org.orcid.jaxb.model.v3.release.search.Result;
import org.orcid.jaxb.model.v3.release.search.Search;
import org.xml.sax.SAXException;
Expand Down Expand Up @@ -64,4 +65,15 @@ public Person convertSinglePerson(InputStream xml) {
}
return null;
}

public Record convertToRecord(InputStream xml) {
Record record = null;
try {
record = (Record) unmarshall(xml, Record.class);
return record;
} catch (SAXException | URISyntaxException e) {
log.error(e);
}
return record;
}
}
8 changes: 8 additions & 0 deletions dspace-api/src/test/java/org/dspace/authority/orcid/MockOrcid.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -51,6 +51,14 @@ public InputStream answer(InvocationOnMock invocation) {
}
});

when(orcidRestConnector.get(ArgumentMatchers.matches("^\\d{4}-\\d{4}-\\d{4}-\\d{4}$"), ArgumentMatchers.any()))
.thenAnswer(new Answer<InputStream>() {
@Override
public InputStream answer(InvocationOnMock invocation) {
return this.getClass().getResourceAsStream("orcid-record.xml");
}
});

setOrcidRestConnector(orcidRestConnector);
}

Expand Down
Loading

0 comments on commit c060a18

Please sign in to comment.