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

[graph raw] fixed mapping of the original resource type from the Datacite format

This commit is contained in:
Claudio Atzori 2024-02-09 10:19:53 +01:00
parent ba1a0e7b4f
commit 1416f16b35
3 changed files with 106 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
dhp-workflows/dhp-graph-mapper/src/main/java/eu/dnetlib/dhp/oa/graph/raw/OdfToOafMapper.java
View File

@ -238,11 +238,23 @@ public class OdfToOafMapper extends AbstractMdRecordToOafMapper {
(Element) doc
.selectSingleNode(
"//*[local-name()='metadata']/*[local-name() = 'resource']/*[local-name() = 'resourceType']"))
.map(element -> {
final String resourceTypeURI = element.attributeValue("uri");
final String resourceTypeAnyURI = element.attributeValue("anyURI");
final String resourceTypeTxt = element.getText();
final String resourceTypeGeneral = element.attributeValue("resourceTypeGeneral");
.map(e -> {
final String resourceTypeURI = Optional
.ofNullable(e.attributeValue("uri"))
.filter(StringUtils::isNotBlank)
.orElse(null);
final String resourceTypeAnyURI = Optional
.ofNullable(e.attributeValue("anyURI"))
.filter(StringUtils::isNotBlank)
.orElse(null);
final String resourceTypeTxt = Optional
.ofNullable(e.getText())
.filter(StringUtils::isNotBlank)
.orElse(null);
final String resourceTypeGeneral = Optional
.ofNullable(e.attributeValue("resourceTypeGeneral"))
.filter(StringUtils::isNotBlank)
.orElse(null);
return ObjectUtils
.firstNonNull(resourceTypeURI, resourceTypeAnyURI, resourceTypeTxt, resourceTypeGeneral);

32
dhp-workflows/dhp-graph-mapper/src/test/java/eu/dnetlib/dhp/oa/graph/raw/MappersTest.java
View File

@ -1171,6 +1171,34 @@ class MappersTest {
}
@Test
void test_Zenodo2() throws IOException {
final String xml = IOUtils.toString(Objects.requireNonNull(getClass().getResourceAsStream("odf_zenodo2.xml")));
final List<Oaf> list = new OdfToOafMapper(vocs, false, true).processMdRecord(xml);
assertEquals(3, list.size());
Publication p = cleanup((Publication) list.get(0), vocs);
assertNotNull(p.getInstance());
assertEquals(1, p.getInstance().size());
final Instance instance = p.getInstance().get(0);
assertNotNull(instance.getInstanceTypeMapping());
assertEquals(1, instance.getInstanceTypeMapping().size());
Optional<InstanceTypeMapping> coarType = instance
.getInstanceTypeMapping()
.stream()
.filter(itm -> ModelConstants.OPENAIRE_COAR_RESOURCE_TYPES_3_1.equals(itm.getVocabularyName()))
.findFirst();
assertTrue(coarType.isPresent());
assertNotNull(coarType.get().getOriginalType());
assertNull(coarType.get().getTypeCode());
assertNull(coarType.get().getTypeLabel());
}
@Test
void testROHub2() throws IOException {
final String xml = IOUtils
@ -1229,7 +1257,7 @@ class MappersTest {
}
@Test
public void testD4ScienceTraining() throws IOException {
void testD4ScienceTraining() throws IOException {
final String xml = IOUtils
.toString(Objects.requireNonNull(getClass().getResourceAsStream("d4science-1-training.xml")));
final List<Oaf> list = new OdfToOafMapper(vocs, false, true).processMdRecord(xml);
@ -1240,7 +1268,7 @@ class MappersTest {
}
@Test
public void testD4ScienceDataset() throws IOException {
void testD4ScienceDataset() throws IOException {
final String xml = IOUtils
.toString(Objects.requireNonNull(getClass().getResourceAsStream("d4science-2-dataset.xml")));
final List<Oaf> list = new OdfToOafMapper(vocs, false, true).processMdRecord(xml);

59
dhp-workflows/dhp-graph-mapper/src/test/resources/eu/dnetlib/dhp/oa/graph/raw/odf_zenodo2.xml Normal file
View File

@ -0,0 +1,59 @@
<?xml version="1.0" encoding="UTF-8"?>
<record xmlns:oaf="http://namespace.openaire.eu/oaf"
xmlns:oai="http://www.openarchives.org/OAI/2.0/"
xmlns:datacite="http://datacite.org/schema/kernel-3"
xmlns:dr="http://www.driver-repository.eu/namespace/dr"
xmlns:dri="http://www.driver-repository.eu/namespace/dri">
<header xmlns="http://www.openarchives.org/OAI/2.0/">
<identifier>oai:zenodo.org:1596086</identifier>
<datestamp>2020-01-20T13:50:28Z</datestamp>
<setSpec>openaire</setSpec>
<dr:dateOfTransformation>2024-02-08T11:03:10.994Z</dr:dateOfTransformation>
<dri:objIdentifier>od______2659::036d5555a6688ed00c8d0da97bdece3b</dri:objIdentifier>
<dri:dateOfCollection>2024-02-08T11:03:10.994Z</dri:dateOfCollection>
<dri:dateOfTransformation>2024-02-08T11:03:10.994Z</dri:dateOfTransformation>
</header>
<metadata>
<resource xmlns="http://datacite.org/schema/kernel-4"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://datacite.org/schema/kernel-4 http://schema.datacite.org/meta/kernel-4.1/metadata.xsd">
<identifier identifierType="URL">https://zenodo.org/record/1596086</identifier>
<alternateIdentifiers xmlns="http://datacite.org/schema/kernel-3"/>
<creators>
<creator>
<creatorName>Bonney, T. G.</creatorName>
<givenName>T. G.</givenName>
<familyName>Bonney</familyName>
</creator>
</creators>
<titles>
<title>Ice Blocks on a Moraine</title>
</titles>
<publisher>Zenodo</publisher>
<publicationYear>1889</publicationYear>
<dates>
<date dateType="Issued">1889-08-22</date>
</dates>
<resourceType resourceTypeGeneral="JournalArticle"/>
<relatedIdentifiers>
<relatedIdentifier relatedIdentifierType="DOI" relationType="IsIdenticalTo"
>10.1038/040391a0</relatedIdentifier>
</relatedIdentifiers>
<rightsList>
<rights rightsURI="https://creativecommons.org/publicdomain/zero/1.0/legalcode"
>Creative Commons Zero v1.0 Universal</rights>
<rights rightsURI="info:eu-repo/semantics/openAccess">Open Access</rights>
</rightsList>
<descriptions>
<description descriptionType="Abstract">n/a</description>
</descriptions>
</resource>
<dr:CobjCategory type="publication">0001</dr:CobjCategory>
<oaf:dateAccepted>1889-08-22</oaf:dateAccepted>
<oaf:accessrights>OPEN</oaf:accessrights>
<oaf:license>http://creativecommons.org/publicdomain/zero/1.0/legalcode</oaf:license>
<oaf:language/>
<oaf:hostedBy name="ZENODO" id="opendoar____::2659"/>
<oaf:collectedFrom name="ZENODO" id="opendoar____::2659"/>
</metadata>
</record>