D-Net
/
dnet-hadoop
14
5
Fork 11
Code Issues 14 Pull Requests 13 Projects 2 Releases 2 Wiki Activity

fixed problem with timestamp

This commit is contained in:
Sandro La Bruzzo 2020-04-15 15:23:54 +02:00
parent c36239e693
commit 74a7fac774
3 changed files with 7 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
dhp-workflows/dhp-graph-provision-scholexplorer/src/main/java/eu/dnetlib/dhp/provision/update/DataciteClientIterator.java
View File

@ -27,10 +27,12 @@ public class DataciteClientIterator implements Iterator<String> {
final String esIndex;
final ObjectMapper mapper = new ObjectMapper();
public DataciteClientIterator(final String esHost, final String esIndex, final long timestamp) throws IOException {
public DataciteClientIterator(final String esHost, final String esIndex, long timestamp) throws IOException {
this.esHost = esHost;
this.esIndex = esIndex;
// THIS FIX IS NECESSARY to avoid different timezone
timestamp -= (60 *60 *2);
final String body =getResponse(String.format("http://%s:9200/%s/_search?scroll=1m", esHost, esIndex), String.format("{\"size\":1000, \"query\":{\"range\":{\"timestamp\":{\"gte\":%d}}}}", timestamp));
scrollId= getJPathString(scrollIdPath, body);
buffer = getBlobs(body);

4
dhp-workflows/dhp-graph-provision-scholexplorer/src/main/resources/eu/dnetlib/dhp/sx/synch/oozie_app/workflow.xml
View File

@ -50,7 +50,7 @@
<arg>-ih</arg><arg>ip-90-147-167-25.ct1.garrservices.it</arg>
<arg>-in</arg><arg>datacite</arg>
</java>
<ok to="End"/>
<ok to="resolveScholix"/>
<error to="Kill"/>
</action>
@ -60,7 +60,7 @@
<name-node>${nameNode}</name-node>
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>Resolve and generate Scholix</name>
<name>resolve and generate Scholix</name>
<class>eu.dnetlib.dhp.provision.update.SparkResolveScholixTarget</class>
<jar>dhp-graph-provision-scholexplorer-${projectVersion}.jar</jar>
<spark-opts>--executor-memory ${sparkExecutorMemory} --driver-memory=${sparkDriverMemory} ${sparkExtraOPT} --conf spark.dynamicAllocation.maxExecutors="32" </spark-opts>

5
dhp-workflows/dhp-graph-provision-scholexplorer/src/test/java/eu/dnetlib/dhp/provision/DataciteClientTest.java
View File

@ -6,11 +6,13 @@ import eu.dnetlib.dhp.provision.scholix.ScholixResource;
import eu.dnetlib.dhp.provision.update.CrossrefClient;
import eu.dnetlib.dhp.provision.update.Datacite2Scholix;
import eu.dnetlib.dhp.provision.update.DataciteClient;
import eu.dnetlib.dhp.provision.update.DataciteClientIterator;
import eu.dnetlib.scholexplorer.relation.RelationMapper;
import org.apache.commons.io.IOUtils;
import org.junit.jupiter.api.Assertions;
import org.junit.jupiter.api.Test;
import java.io.IOException;
import java.util.List;
@ -25,7 +27,6 @@ public class DataciteClientTest {
System.out.println(new ObjectMapper().writeValueAsString(s));
}
public void testResolveDataset() throws Exception {
DataciteClient dc = new DataciteClient("ip-90-147-167-25.ct1.garrservices.it");
ScholixResource datasetByDOI = dc.getDatasetByDOI("10.17182/hepdata.15392.v1/t5");
@ -37,7 +38,5 @@ public class DataciteClientTest {
ScholixResource crossrefByDOI = cr.getResourceByDOI("10.26850/1678-4618eqj.v35.1.2010.p41-46");
Assertions.assertNotNull(crossrefByDOI);
System.out.println(new ObjectMapper().writeValueAsString(crossrefByDOI));
}
}