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

Merge branch 'beta' into 9117_pubmed_affiliations

This commit is contained in:
Claudio Atzori 2023-11-22 13:52:53 +01:00
parent 2090003ea9 008fdf9d8a
commit 745039ad5b
5 changed files with 113 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
dhp-workflows/dhp-graph-mapper/src/main/java/eu/dnetlib/dhp/oa/graph/raw/AbstractMdRecordToOafMapper.java
View File

@ -29,7 +29,7 @@ public abstract class AbstractMdRecordToOafMapper {
protected final VocabularyGroup vocs;
protected static final UrlValidator URL_VALIDATOR = UrlValidator.getInstance();
protected static final UrlValidator URL_VALIDATOR = new UrlValidator(UrlValidator.ALLOW_2_SLASHES);
private final boolean invisible;

32
dhp-workflows/dhp-graph-mapper/src/main/resources/eu/dnetlib/dhp/oa/graph/resolution/oozie_app/config-default.xml
View File

@ -1,4 +1,12 @@
<configuration>
<property>
<name>jobTracker</name>
<value>yarnRM</value>
</property>
<property>
<name>nameNode</name>
<value>hdfs://nameservice1</value>
</property>
<property>
<name>oozie.use.system.libpath</name>
<value>true</value>
@ -7,4 +15,28 @@
<name>oozie.action.sharelib.for.spark</name>
<value>spark2</value>
</property>
<property>
<name>hive_metastore_uris</name>
<value>thrift://iis-cdh5-test-m3.ocean.icm.edu.pl:9083</value>
</property>
<property>
<name>spark2YarnHistoryServerAddress</name>
<value>http://iis-cdh5-test-gw.ocean.icm.edu.pl:18089</value>
</property>
<property>
<name>spark2ExtraListeners</name>
<value>com.cloudera.spark.lineage.NavigatorAppListener</value>
</property>
<property>
<name>spark2SqlQueryExecutionListeners</name>
<value>com.cloudera.spark.lineage.NavigatorQueryListener</value>
</property>
<property>
<name>sparkExecutorNumber</name>
<value>4</value>
</property>
<property>
<name>spark2EventLogDir</name>
<value>/user/spark/spark2ApplicationHistory</value>
</property>
</configuration>

14
dhp-workflows/dhp-graph-mapper/src/test/java/eu/dnetlib/dhp/oa/graph/raw/MappersTest.java
View File

@ -797,6 +797,20 @@ class MappersTest {
assertFalse(p_cleaned.getTitle().isEmpty());
}
@Test
void test_instance_url_validation() throws IOException {
final String xml = IOUtils.toString(Objects.requireNonNull(getClass().getResourceAsStream("idus_sevilla.xml")));
final List<Oaf> list = new OafToOafMapper(vocs, false, true).processMdRecord(xml);
final Publication p = (Publication) list.get(0);
assertNotNull(p.getInstance());
assertFalse(p.getInstance().isEmpty());
assertNotNull(p.getInstance().get(0).getUrl());
assertFalse(p.getInstance().get(0).getUrl().isEmpty());
assertEquals("https://idus.us.es/handle//11441/118940", p.getInstance().get(0).getUrl().get(0));
}
@Test
void testZenodo() throws IOException, DocumentException {
final String xml = IOUtils.toString(Objects.requireNonNull(getClass().getResourceAsStream("odf_zenodo.xml")));

65
dhp-workflows/dhp-graph-mapper/src/test/resources/eu/dnetlib/dhp/oa/graph/raw/idus_sevilla.xml Normal file
View File

@ -0,0 +1,65 @@
<?xml version="1.0" encoding="UTF-8"?>
<record xmlns:dc="http://purl.org/dc/elements/1.1/"
xmlns:dr="http://www.driver-repository.eu/namespace/dr"
xmlns:dri="http://www.driver-repository.eu/namespace/dri"
xmlns:oaf="http://namespace.openaire.eu/oaf"
xmlns:oai="http://www.openarchives.org/OAI/2.0/"
xmlns:prov="http://www.openarchives.org/OAI/2.0/provenance"
xmlns:xs="http://www.w3.org/2001/XMLSchema" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
<header xmlns="http://namespace.openaire.eu/">
<dri:objIdentifier>od______3272::6a4d00217a024a46ce9697ce98b13c2a</dri:objIdentifier>
<dri:recordIdentifier>oai:idus.us.es:11441/118940</dri:recordIdentifier>
<dri:dateOfCollection/>
<dri:mdFormat/>
<dri:mdFormatInterpretation/>
<dri:repositoryId/>
<dr:objectIdentifier/>
<dr:dateOfCollection>2021-08-20T12:32:32.826Z</dr:dateOfCollection>
<dr:dateOfTransformation>2023-07-04T15:47:55.397Z</dr:dateOfTransformation>
<oaf:datasourceprefix>od______3272</oaf:datasourceprefix>
</header>
<metadata xmlns="http://namespace.openaire.eu/">
<dc:title>El museo pictorico y escala optica : tomo I : theorica de la pintura en que se describe su origen ... y se aprueban con demonstraciomes mathematicas y filosoficas, sus mas radicales fundamentos</dc:title>
<dc:creator>Palomino de Castro y Velasco, Antonio, 1653-1726</dc:creator>
<dc:contributor>Rovira y Brocandel, Hipólito, 1693-1765</dc:contributor>
<dc:contributor>Palomino de Castro y Velasco, Antonio, 1653-1726</dc:contributor>
<dc:date>2021-08-12T08:59:53Z</dc:date>
<dc:date>1715</dc:date>
<dc:description>A 042(a)/063</dc:description>
<dc:format>application/pdf</dc:format>
<dc:identifier>https://idus.us.es/handle//11441/118940</dc:identifier>
<dc:language>spa</dc:language>
<dc:publisher>En Madrid : por Lucas Antonio de Bedmar ... : vendese en casa de Don Joseph de Villar y Villanueva, 1715</dc:publisher>
<dc:type>info:eu-repo/semantics/book</dc:type>
<dc:type>info:eu-repo/semantics/publishedVersion</dc:type>
<dr:CobjCategory type="publication">0002</dr:CobjCategory>
<oaf:dateAccepted>1715-01-01</oaf:dateAccepted>
<oaf:embargoenddate/>
<oaf:collectedDatasourceid>opendoar____::3272</oaf:collectedDatasourceid>
<oaf:accessrights>OPEN</oaf:accessrights>
<oaf:hostedBy id="opendoar____::3272" name="idUS. Depósito de Investigación de la Universidad de Sevilla."/>
<oaf:collectedFrom id="opendoar____::3272" name="idUS. Depósito de Investigación de la Universidad de Sevilla."/>
<oaf:identifier identifierType="landingPage">https://idus.us.es/handle//11441/118940</oaf:identifier>
<oaf:journal eissn="" ep="" iss="" issn="" sp="" vol=""/>
<oaf:license>http://creativecommons.org/licenses/by-nc-nd/4.0/</oaf:license>
</metadata>
<about>
<provenance xmlns="http://www.openarchives.org/OAI/2.0/provenance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/provenance http://www.openarchives.org/OAI/2.0/provenance.xsd">
<originDescription altered="true" harvestDate="2021-08-20T12:32:32.826Z">
<baseURL>http%3A%2F%2Fidus.us.es%2Foai%2Fdriver</baseURL>
<identifier>oai:idus.us.es:11441/118940</identifier>
<datestamp>2021-08-12T08:59:54Z</datestamp>
<metadataNamespace>http://www.openarchives.org/OAI/2.0/oai_dc/</metadataNamespace>
</originDescription>
</provenance>
<oaf:datainfo>
<oaf:inferred>false</oaf:inferred>
<oaf:deletedbyinference>false</oaf:deletedbyinference>
<oaf:trust>0.9</oaf:trust>
<oaf:inferenceprovenance/>
<oaf:provenanceaction classid="sysimport:crosswalk:repository"
classname="sysimport:crosswalk:repository"
schemeid="dnet:provenanceActions" schemename="dnet:provenanceActions"/>
</oaf:datainfo>
</about>
</record>

2
pom.xml
View File

@ -888,7 +888,7 @@
<mockito-core.version>3.3.3</mockito-core.version>
<mongodb.driver.version>3.4.2</mongodb.driver.version>
<vtd.version>[2.12,3.0)</vtd.version>
<dhp-schemas.version>[3.17.1]</dhp-schemas.version>
<dhp-schemas.version>[3.17.2]</dhp-schemas.version>
<dnet-actionmanager-api.version>[4.0.3]</dnet-actionmanager-api.version>
<dnet-actionmanager-common.version>[6.0.5]</dnet-actionmanager-common.version>
<dnet-openaire-broker-common.version>[3.1.6]</dnet-openaire-broker-common.version>