antonis.lempesis
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
1
Fork 1
Code Issues Pull Requests Projects Releases Wiki Activity

tests to generate records for the EOSC-Future demo with the EOSC Jupyter Notebbok subject

This commit is contained in:
Alessia Bardi 2021-09-24 17:11:56 +02:00
parent df15a4dc9f
commit b924276e18
4 changed files with 167 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
dhp-workflows/dhp-graph-provision/src/test/java/eu/dnetlib/dhp/oa/provision/IndexRecordTransformerTest.java
View File

@ -76,6 +76,18 @@ public class IndexRecordTransformerTest {
testRecordTransformation(record);
}
@Test
public void testForEOSCFutureDataTransferPilot() throws IOException, TransformerException {
final String record = IOUtils.toString(getClass().getResourceAsStream("eosc-future/data-transfer-pilot.xml"));
testRecordTransformation(record);
}
@Test
public void testForEOSCFutureTraining() throws IOException, TransformerException {
final String record = IOUtils.toString(getClass().getResourceAsStream("eosc-future/training-notebooks-seadatanet.xml"));
testRecordTransformation(record);
}
private void testRecordTransformation(final String record) throws IOException, TransformerException {
final String fields = IOUtils.toString(getClass().getResourceAsStream("fields.xml"));
final String xslt = IOUtils.toString(getClass().getResourceAsStream("layoutToRecordTransformer.xsl"));

72
dhp-workflows/dhp-graph-provision/src/test/resources/eu/dnetlib/dhp/oa/provision/eosc-future/data-transfer-pilot.xml Normal file
View File

@ -0,0 +1,72 @@
<record>
<result xmlns:dri="http://www.driver-repository.eu/namespace/dri" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
<header>
<dri:objIdentifier>r37b0ad08687::dec0d8520e726f2adda9a51280ac7299</dri:objIdentifier>
<dri:dateOfCollection>2021-09-22T08:53:16Z</dri:dateOfCollection>
<dri:status>under curation</dri:status>
<counters />
</header>
<metadata>
<oaf:entity xmlns:oaf="http://namespace.openaire.eu/oaf"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://namespace.openaire.eu/oaf http://namespace.openaire.eu/oaf http://www.openaire.eu/schema/0.2/oaf-0.2.xsd">
<oaf:result>
<title classid="main title" classname="main title" schemeid="dnet:dataCite_title" schemename="dnet:dataCite_title">EGI-Foundation/data-transfer-pilot: Include libraries in environment.yml</title>
<creator rank="1" name="" surname="">Giuseppe La Rocca</creator>
<creator rank="2" name="" surname="">Enol Fernández</creator>
<creator rank="3" name="" surname="">Andrea Manzi</creator>
<dateofacceptance />
<resulttype classid="software" classname="software" schemeid="dnet:result_typologies" schemename="dnet:result_typologies" />
<language classid="" classname="" schemeid="dnet:languages" schemename="dnet:languages" />
<description>This notebook is used to demonstrate how a scientist from one of the PaNOSC RIs can use the resources provided by EGI to perform analysis on the data sets obtained during an expirement.</description>
<country classid="" classname="" schemeid="" schemename="" />
<subject classid="keyword" classname="keyword" schemeid="dnet:subject_classification_typologies" schemename="dnet:subject_classification_typologies">EOSC Jupyter Notebook</subject>
<relevantdate classid="" classname="" schemeid="" schemename="" />
<publisher>Zenodo</publisher>
<embargoenddate />
<journal issn="" eissn="" lissn="" ep="" iss="" sp="" vol="" />
<source />
<fulltext />
<format />
<storagedate />
<resourcetype classid="" classname="" schemeid="" schemename="" />
<device />
<size />
<version />
<lastmetadataupdate />
<metadataversionnumber />
<documentationUrl />
<codeRepositoryUrl />
<programmingLanguage classid="" classname="" schemeid="" schemename="" />
<contactperson />
<contactgroup />
<tool />
<originalId>oai:zenodo.org:4218562</originalId>
<collectedfrom name="Zenodo" id="re3data_____::7b0ad08687b2c960d5aeef06f811d5e6" />
<pid classid="oai" classname="Open Archives Initiative" schemeid="dnet:pid_types" schemename="dnet:pid_types">oai:zenodo.org:4218562</pid>
<pid classid="doi" classname="Digital Object Identifier" schemeid="dnet:pid_types" schemename="dnet:pid_types">10.5281/zenodo.4218562</pid>
<bestaccessright classid="OPEN" classname="Open Access" schemeid="dnet:access_modes" schemename="dnet:access_modes" />
<datainfo>
<inferred>false</inferred>
<deletedbyinference>false</deletedbyinference>
<trust>0.9</trust>
<inferenceprovenance />
<provenanceaction classid="user:insert" classname="user:insert" schemeid="dnet:provenanceActions" schemename="dnet:provenanceActions" />
</datainfo>
<rels></rels>
<children>
<instance id="r37b0ad08687::dec0d8520e726f2adda9a51280ac7299">
<instancetype classid="0029" classname="Software" schemeid="dnet:publication_resource" schemename="dnet:publication_resource" />
<collectedfrom name="Zenodo" id="re3data_____::7b0ad08687b2c960d5aeef06f811d5e6" />
<hostedby name="Zenodo" id="re3data_____::7b0ad08687b2c960d5aeef06f811d5e6" />
<accessright classid="OPEN" classname="Open Access" schemeid="dnet:access_modes" schemename="dnet:access_modes" />
<dateofacceptance />
<webresource>
<url>https://zenodo.org/record/4218562</url>
</webresource>
</instance>
</children>
</oaf:result>
</oaf:entity>
</metadata>
</result>
</record>

71
dhp-workflows/dhp-graph-provision/src/test/resources/eu/dnetlib/dhp/oa/provision/eosc-future/training-notebooks-seadatanet.xml Normal file
View File

@ -0,0 +1,71 @@
<record>
<result xmlns:dri="http://www.driver-repository.eu/namespace/dri" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
<header>
<dri:objIdentifier>r37b0ad08687::eb430fb7438e1533ba95d6aa50a477eb</dri:objIdentifier>
<dri:dateOfCollection>2021-09-22T08:53:13Z</dri:dateOfCollection>
<dri:status>under curation</dri:status>
<counters />
</header>
<metadata>
<oaf:entity xmlns:oaf="http://namespace.openaire.eu/oaf"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://namespace.openaire.eu/oaf http://namespace.openaire.eu/oaf http://www.openaire.eu/schema/0.2/oaf-0.2.xsd">
<oaf:result>
<title classid="main title" classname="main title" schemeid="dnet:dataCite_title" schemename="dnet:dataCite_title">EGI-Foundation/training-notebooks-seadatanet: Version 0.4</title>
<creator rank="1" name="" surname="">Enol Fernández</creator>
<dateofacceptance />
<resulttype classid="software" classname="software" schemeid="dnet:result_typologies" schemename="dnet:result_typologies" />
<language classid="" classname="" schemeid="dnet:languages" schemename="dnet:languages" />
<description>A sample notebook using SeaDataNet data to plot a map that shows surface temperature of Black Sea, Arctic Sea and Baltic Sea. The data is available at EGI DataHub with PID http://hdl.handle.net/21.T15999/qVk6JWQ (run at EGI Notebooks service for easy access to data).This release updates the PID for the data.</description>
<country classid="" classname="" schemeid="" schemename="" />
<subject classid="keyword" classname="keyword" schemeid="dnet:subject_classification_typologies" schemename="dnet:subject_classification_typologies">EOSC Jupyter Notebook</subject>
<relevantdate classid="" classname="" schemeid="" schemename="" />
<publisher>Zenodo</publisher>
<embargoenddate />
<journal issn="" eissn="" lissn="" ep="" iss="" sp="" vol="" />
<source />
<fulltext />
<format />
<storagedate />
<resourcetype classid="" classname="" schemeid="" schemename="" />
<device />
<size />
<version />
<lastmetadataupdate />
<metadataversionnumber />
<documentationUrl />
<codeRepositoryUrl />
<programmingLanguage classid="" classname="" schemeid="" schemename="" />
<contactperson />
<contactgroup />
<tool />
<originalId>oai:zenodo.org:3561323</originalId>
<collectedfrom name="Zenodo" id="re3data_____::7b0ad08687b2c960d5aeef06f811d5e6" />
<pid classid="oai" classname="Open Archives Initiative" schemeid="dnet:pid_types" schemename="dnet:pid_types">oai:zenodo.org:3561323</pid>
<pid classid="doi" classname="Digital Object Identifier" schemeid="dnet:pid_types" schemename="dnet:pid_types">10.5281/zenodo.3561323</pid>
<bestaccessright classid="OPEN" classname="Open Access" schemeid="dnet:access_modes" schemename="dnet:access_modes" />
<datainfo>
<inferred>false</inferred>
<deletedbyinference>false</deletedbyinference>
<trust>0.9</trust>
<inferenceprovenance />
<provenanceaction classid="user:insert" classname="user:insert" schemeid="dnet:provenanceActions" schemename="dnet:provenanceActions" />
</datainfo>
<rels></rels>
<children>
<instance id="r37b0ad08687::eb430fb7438e1533ba95d6aa50a477eb">
<instancetype classid="0029" classname="Software" schemeid="dnet:publication_resource" schemename="dnet:publication_resource" />
<collectedfrom name="Zenodo" id="re3data_____::7b0ad08687b2c960d5aeef06f811d5e6" />
<hostedby name="Zenodo" id="re3data_____::7b0ad08687b2c960d5aeef06f811d5e6" />
<accessright classid="OPEN" classname="Open Access" schemeid="dnet:access_modes" schemename="dnet:access_modes" />
<dateofacceptance />
<webresource>
<url>https://zenodo.org/record/3561323</url>
</webresource>
</instance>
</children>
</oaf:result>
</oaf:entity>
</metadata>
</result>
</record>

16
dhp-workflows/dhp-graph-provision/src/test/resources/eu/dnetlib/dhp/oa/provision/fields.xml
View File

@ -15,7 +15,13 @@
<FIELD indexable="true" multivalued="false" name="datasourcecompatibilityid" result="false" stat="false" tokenizable="false" xpath="//*[local-name()='entity']/*[local-name()='datasource']/openairecompatibility/@classid"/>
<FIELD indexable="true" multivalued="false" name="datasourcecompatibilityname" result="false" stat="false" tokenizable="false" xpath="//*[local-name()='entity']/*[local-name()='datasource']/openairecompatibility/@classname"/>
<FIELD indexable="true" multivalued="true" name="datasourcesubject" result="false" stat="false" type="ngramtext" xpath="//*[local-name()='entity']/*[local-name()='datasource']/subjects"/>
<FIELD indexable="true" name="versioning" result="false" stat="false" tokenizable="false" xpath="//*[local-name()='entity']/*[local-name()='datasource']/versioning"/><!-- ORGANIZATION FIELDS -->
<FIELD indexable="true" name="versioning" result="false" stat="false" tokenizable="false" xpath="//*[local-name()='entity']/*[local-name()='datasource']/versioning"/>
<!-- datasource fields for EOSC -->
<FIELD indexable="true" name="datasourcejurisdiction" result="false" stat="false" tokenizable="false" xpath="//*[local-name()='entity']/*[local-name()='datasource']/jurisdiction"/>
<FIELD indexable="true" name="datasourcethematic" result="false" stat="false" tokenizable="false" xpath="//*[local-name()='entity']/*[local-name()='datasource']/thematic"/>
<FIELD indexable="true" name="datasourceknowledge_graph" result="false" stat="false" tokenizable="false" xpath="//*[local-name()='entity']/*[local-name()='datasource']/knowledgegraph"/>
<FIELD indexable="true" name="datasourcecontentpolicy" result="false" stat="false" tokenizable="false" xpath="//*[local-name()='entity']/*[local-name()='datasource']/contentpolicy"/>
<!-- ORGANIZATION FIELDS -->
<FIELD indexable="true" name="organizationlegalshortname" result="false" stat="false" type="ngramtext" xpath="distinct-values(//*[local-name()='entity']/*[local-name()='organization']//legalshortname)"/>
<FIELD indexable="true" name="organizationlegalname" result="false" stat="false" type="ngramtext" xpath="distinct-values(//*[local-name()='entity']/*[local-name()='organization']//legalname)"/>
<FIELD indexable="true" name="organizationalternativenames" result="false" stat="false" type="ngramtext" xpath="distinct-values(//*[local-name()='entity']/*[local-name()='organization']//alternativeNames)"/>
@ -28,7 +34,8 @@
<FIELD indexable="true" name="organizationecenterprise" result="false" stat="false" xpath="//*[local-name()='entity']/*[local-name()='organization']/ecenterprise"/>
<FIELD indexable="true" name="organizationecsmevalidated" result="false" stat="false" xpath="//*[local-name()='entity']/*[local-name()='organization']/ecsmevalidated"/>
<FIELD indexable="true" name="organizationecnutscode" result="false" stat="false" xpath="//*[local-name()='entity']/*[local-name()='organization']/ecnutscode"/>
<FIELD indexable="true" multivalued="false" name="organizationcountryname" result="false" stat="false" tokenizable="false" xpath="//*[local-name()='entity']/*[local-name()='organization']/country/@classname"/><!-- PROJECT FIELDS -->
<FIELD indexable="true" multivalued="false" name="organizationcountryname" result="false" stat="false" tokenizable="false" xpath="//*[local-name()='entity']/*[local-name()='organization']/country/@classname"/>
<!-- PROJECT FIELDS -->
<FIELD indexable="true" name="projectcode" result="false" stat="false" type="ngramtext" xpath="//*[local-name()='entity']/*[local-name()='project']/code"/>
<FIELD indexable="true" name="projectcode_nt" result="false" stat="false" tokenizable="false" xpath="//*[local-name()='entity']/*[local-name()='project']/code"/>
<FIELD indexable="true" name="projectacronym" result="false" stat="false" type="ngramtext" xpath="//*[local-name()='entity']/*[local-name()='project']/acronym"/>
@ -79,6 +86,7 @@
<FIELD indexable="true" multivalued="true" name="resultauthor_nt" result="false" stat="false" type="string_ci" xpath="//*[local-name()='entity']/*[local-name()='result']/creator"/>
<FIELD indexable="true" multivalued="true" name="authorid" result="false" stat="false" type="string_ci" xpath="//*[local-name()='entity']/*[local-name()='result']/creator/@*[local-name() != 'rank' and local-name() != 'name' and local-name() != 'surname']"/>
<FIELD indexable="true" multivalued="true" name="authoridtype" result="false" stat="false" type="string_ci" xpath="//*[local-name()='entity']/*[local-name()='result']/creator/@*[local-name() != 'rank' and local-name() != 'name' and local-name() != 'surname']/local-name()"/>
<FIELD indexable="true" multivalued="true" name="orcidtypevalue" result="false" stat="false" type="string_ci" value="string-join((./@*[local-name() = 'orcid' or local-name() = 'orcid_pending'], ./@*[local-name() = 'orcid' or local-name() = 'orcid_pending']/local-name()), '||' )" xpath="//*[local-name()='entity']/*[local-name()='result']/creator"/>
<FIELD indexable="true" name="resulthostingdatasource" result="false" stat="false" tokenizable="false" value="distinct-values(concat(./@id, '||', ./@name))" xpath="//*[local-name()='entity']/*[local-name()='result']/children/instance/*[local-name()='hostedby']"/>
<FIELD indexable="true" name="resulthostingdatasourceid" result="false" stat="false" tokenizable="false" xpath="distinct-values(//*[local-name()='entity']/*[local-name()='result']/children/instance/*[local-name()='hostedby']/@id)"/>
<FIELD indexable="true" name="resulthostingdatasourcename" result="false" stat="false" tokenizable="false" xpath="distinct-values(//*[local-name()='entity']/*[local-name()='result']/children/instance/*[local-name()='hostedby']/@name)"/>
@ -105,7 +113,7 @@
<FIELD indexable="true" name="relorganizationname" result="false" stat="false" xpath="distinct-values(//*[local-name()='entity']/*//rel[./to/@type='organization']/legalname)"/>
<FIELD indexable="true" name="relorganizationshortname" result="false" stat="false" xpath="distinct-values(//*[local-name()='entity']/*//rel[./to/@type='organization']/legalshortname)"/>
<FIELD indexable="true" name="relresultid" result="false" stat="false" tokenizable="false" xpath="distinct-values(//*[local-name()='entity']/*//rel/to[@type='result'])"/>
<FIELD indexable="true" name="relresulttype" result="false" stat="false" tokenizable="false" xpath="distinct-values(//*[local-name()='entity']/*//rel/to/@type)"/>
<FIELD indexable="true" name="relresulttype" result="false" stat="false" tokenizable="false" xpath="distinct-values(//*[local-name()='entity']/*//rel/resulttype/@classid)"/>
<FIELD indexable="true" name="relclass" result="false" stat="false" tokenizable="false" xpath="distinct-values(//*[local-name()='entity']/*//rel/to/@class)"/>
<FIELD indexable="true" name="relfundinglevel0_id" result="false" stat="false" tokenizable="false" xpath="//*[local-name()='entity']//rel/funding/funding_level_0"/>
<FIELD indexable="true" name="relfundinglevel0_name" result="false" stat="false" tokenizable="false" xpath="//*[local-name()='entity']//rel/funding/funding_level_0/@name/string()"/>
@ -130,7 +138,7 @@
<FIELD indexable="true" name="collectedfromdatasourceid" result="false" stat="false" tokenizable="false" xpath="distinct-values(//*[local-name()='entity']/*/*[local-name()='collectedfrom']/@id | //*[local-name()='entity']/*//*[local-name() = 'instance']/*[local-name()='collectedfrom']/@id)"/>
<FIELD indexable="true" name="collectedfromname" result="false" stat="false" tokenizable="false" xpath="distinct-values(//*[local-name()='entity']/*/*[local-name()='collectedfrom']/@name | //*[local-name()='entity']/*//*[local-name() = 'instance']/*[local-name()='collectedfrom']/@name)"/>
<FIELD indexable="true" name="originalid" result="false" stat="false" tokenizable="false" type="string_ci" xpath="//*[local-name()='entity']/*/*[local-name()='originalId']"/>
<FIELD indexable="true" name="pid" result="false" stat="false" tokenizable="false" type="string_ci" xpath="//*[local-name()='entity']/*/pid/text()"/>
<FIELD indexable="true" name="pid" result="false" stat="false" tokenizable="false" type="string_ci" xpath="distinct-values(//*[local-name()='entity']/*/pid/text()|//*[local-name()='instance']/*[local-name()='alternateidentifier']/text())"/>
<FIELD indexable="true" name="pidclassid" result="false" stat="false" tokenizable="false" xpath="distinct-values(//*[local-name()='entity']/*/pid/@classid)"/>
<FIELD indexable="true" name="pidclassname" result="false" stat="false" tokenizable="false" xpath="distinct-values(//*[local-name()='entity']/*/pid/@classname)"/>
<FIELD indexable="true" name="inferred" result="false" stat="false" tokenizable="false" xpath="//*[local-name()='entity']//datainfo/inferred"/>