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

[AffiliationIngestion]refactoring

This commit is contained in:
Miriam Baglioni 2024-06-29 18:36:47 +02:00
parent 9cbe966b4a
commit a2b708bb71
3 changed files with 44 additions and 35 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
dhp-workflows/dhp-graph-provision/src/main/java/eu/dnetlib/dhp/oa/provision/model/ProvisionModelSupport.java
View File

@ -5,7 +5,6 @@ import java.io.StringReader;
import java.util.*; import java.util.*;
import java.util.stream.Collectors; import java.util.stream.Collectors;
import eu.dnetlib.dhp.schema.solr.ExternalReference;
import org.apache.commons.lang3.StringUtils; import org.apache.commons.lang3.StringUtils;
import org.dom4j.Document; import org.dom4j.Document;
import org.dom4j.DocumentException; import org.dom4j.DocumentException;
@ -31,6 +30,7 @@ import eu.dnetlib.dhp.schema.solr.Context;
import eu.dnetlib.dhp.schema.solr.Country; import eu.dnetlib.dhp.schema.solr.Country;
import eu.dnetlib.dhp.schema.solr.Datasource; import eu.dnetlib.dhp.schema.solr.Datasource;
import eu.dnetlib.dhp.schema.solr.EoscIfGuidelines; import eu.dnetlib.dhp.schema.solr.EoscIfGuidelines;
import eu.dnetlib.dhp.schema.solr.ExternalReference;
import eu.dnetlib.dhp.schema.solr.Instance; import eu.dnetlib.dhp.schema.solr.Instance;
import eu.dnetlib.dhp.schema.solr.Journal; import eu.dnetlib.dhp.schema.solr.Journal;
import eu.dnetlib.dhp.schema.solr.Measure; import eu.dnetlib.dhp.schema.solr.Measure;
@ -562,10 +562,16 @@ public class ProvisionModelSupport {
.orElse(null); .orElse(null);
} }
private static List<ExternalReference> mapExternalReference(List<eu.dnetlib.dhp.schema.oaf.ExternalReference> externalReference) { private static List<ExternalReference> mapExternalReference(
return Optional.ofNullable(externalReference) List<eu.dnetlib.dhp.schema.oaf.ExternalReference> externalReference) {
.map(ext -> ext.stream() return Optional
.map(e -> ExternalReference.newInstance( .ofNullable(externalReference)
.map(
ext -> ext
.stream()
.map(
e -> ExternalReference
.newInstance(
e.getSitename(), e.getSitename(),
e.getLabel(), e.getLabel(),
e.getAlternateLabel(), e.getAlternateLabel(),
@ -573,8 +579,8 @@ public class ProvisionModelSupport {
mapCodeLabel(e.getQualifier()), mapCodeLabel(e.getQualifier()),
e.getRefidentifier(), e.getRefidentifier(),
e.getQuery())) e.getQuery()))
.collect(Collectors.toList())) .collect(Collectors.toList()))
.orElse(Lists.newArrayList()); .orElse(Lists.newArrayList());
} }
private static List<Context> asContext(List<eu.dnetlib.dhp.schema.oaf.Context> ctxList, private static List<Context> asContext(List<eu.dnetlib.dhp.schema.oaf.Context> ctxList,

56
dhp-workflows/dhp-graph-provision/src/test/java/eu/dnetlib/dhp/oa/provision/SolrConfigExploreTest.java
View File

@ -1,12 +1,13 @@
package eu.dnetlib.dhp.oa.provision; package eu.dnetlib.dhp.oa.provision;
import static org.junit.jupiter.api.Assertions.assertEquals;
import java.io.File; import java.io.File;
import java.io.IOException; import java.io.IOException;
import java.net.URI; import java.net.URI;
import java.nio.file.Path; import java.nio.file.Path;
import eu.dnetlib.dhp.oa.provision.model.SerializableSolrInputDocument;
import org.apache.commons.io.FileUtils; import org.apache.commons.io.FileUtils;
import org.apache.commons.io.IOUtils; import org.apache.commons.io.IOUtils;
import org.apache.solr.client.solrj.SolrQuery; import org.apache.solr.client.solrj.SolrQuery;
@ -32,14 +33,13 @@ import org.junit.jupiter.api.io.TempDir;
import org.mockito.Mock; import org.mockito.Mock;
import org.mockito.Mockito; import org.mockito.Mockito;
import org.mockito.junit.jupiter.MockitoExtension; import org.mockito.junit.jupiter.MockitoExtension;
import eu.dnetlib.dhp.oa.provision.utils.ISLookupClient;
import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpException;
import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpService;
import org.slf4j.Logger; import org.slf4j.Logger;
import org.slf4j.LoggerFactory; import org.slf4j.LoggerFactory;
import static org.junit.jupiter.api.Assertions.assertEquals; import eu.dnetlib.dhp.oa.provision.model.SerializableSolrInputDocument;
import eu.dnetlib.dhp.oa.provision.utils.ISLookupClient;
import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpException;
import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpService;
@ExtendWith(MockitoExtension.class) @ExtendWith(MockitoExtension.class)
public class SolrConfigExploreTest { public class SolrConfigExploreTest {
@ -91,7 +91,7 @@ public class SolrConfigExploreTest {
SparkConf conf = new SparkConf(); SparkConf conf = new SparkConf();
conf.setAppName(XmlIndexingJobTest.class.getSimpleName()); conf.setAppName(XmlIndexingJobTest.class.getSimpleName());
conf.registerKryoClasses(new Class[] { conf.registerKryoClasses(new Class[] {
SerializableSolrInputDocument.class SerializableSolrInputDocument.class
}); });
conf.setMaster("local[1]"); conf.setMaster("local[1]");
@ -101,10 +101,10 @@ public class SolrConfigExploreTest {
conf.set("spark.sql.warehouse.dir", workingDir.resolve("spark").toString()); conf.set("spark.sql.warehouse.dir", workingDir.resolve("spark").toString());
spark = SparkSession spark = SparkSession
.builder() .builder()
.appName(SolrConfigExploreTest.class.getSimpleName()) .appName(SolrConfigExploreTest.class.getSimpleName())
.config(conf) .config(conf)
.getOrCreate(); .getOrCreate();
// random unassigned HTTP port // random unassigned HTTP port
final int jettyPort = 0; final int jettyPort = 0;
@ -134,35 +134,35 @@ public class SolrConfigExploreTest {
log.info(new ConfigSetAdminRequest.List().process(miniCluster.getSolrClient()).toString()); log.info(new ConfigSetAdminRequest.List().process(miniCluster.getSolrClient()).toString());
log log
.info( .info(
CollectionAdminRequest.ClusterStatus CollectionAdminRequest.ClusterStatus
.getClusterStatus() .getClusterStatus()
.process(miniCluster.getSolrClient()) .process(miniCluster.getSolrClient())
.toString()); .toString());
NamedList<Object> res = createCollection( NamedList<Object> res = createCollection(
miniCluster.getSolrClient(), SHADOW_COLLECTION, 4, 2, 20, CONFIG_NAME); miniCluster.getSolrClient(), SHADOW_COLLECTION, 4, 2, 20, CONFIG_NAME);
res.forEach(o -> log.info(o.toString())); res.forEach(o -> log.info(o.toString()));
// miniCluster.getSolrClient().setDefaultCollection(SHADOW_COLLECTION); // miniCluster.getSolrClient().setDefaultCollection(SHADOW_COLLECTION);
res = createCollection( res = createCollection(
miniCluster.getSolrClient(), PUBLIC_COLLECTION, 4, 2, 20, CONFIG_NAME); miniCluster.getSolrClient(), PUBLIC_COLLECTION, 4, 2, 20, CONFIG_NAME);
res.forEach(o -> log.info(o.toString())); res.forEach(o -> log.info(o.toString()));
admin = new SolrAdminApplication(miniCluster.getZkClient().getZkServerAddress()); admin = new SolrAdminApplication(miniCluster.getZkClient().getZkServerAddress());
CollectionAdminResponse rsp = (CollectionAdminResponse) admin CollectionAdminResponse rsp = (CollectionAdminResponse) admin
.createAlias(ProvisionConstants.PUBLIC_ALIAS_NAME, PUBLIC_COLLECTION); .createAlias(ProvisionConstants.PUBLIC_ALIAS_NAME, PUBLIC_COLLECTION);
assertEquals(0, rsp.getStatus()); assertEquals(0, rsp.getStatus());
rsp = (CollectionAdminResponse) admin.createAlias(ProvisionConstants.SHADOW_ALIAS_NAME, SHADOW_COLLECTION); rsp = (CollectionAdminResponse) admin.createAlias(ProvisionConstants.SHADOW_ALIAS_NAME, SHADOW_COLLECTION);
assertEquals(0, rsp.getStatus()); assertEquals(0, rsp.getStatus());
log log
.info( .info(
CollectionAdminRequest.ClusterStatus CollectionAdminRequest.ClusterStatus
.getClusterStatus() .getClusterStatus()
.process(miniCluster.getSolrClient()) .process(miniCluster.getSolrClient())
.toString()); .toString());
} }
@ -180,7 +180,8 @@ public class SolrConfigExploreTest {
new XmlIndexingJob(spark, inputPath, SHADOW_FORMAT, ProvisionConstants.SHADOW_ALIAS_NAME, batchSize) new XmlIndexingJob(spark, inputPath, SHADOW_FORMAT, ProvisionConstants.SHADOW_ALIAS_NAME, batchSize)
.run(isLookupClient); .run(isLookupClient);
Assertions.assertEquals(0, miniCluster.getSolrClient().commit(ProvisionConstants.SHADOW_ALIAS_NAME).getStatus()); Assertions
.assertEquals(0, miniCluster.getSolrClient().commit(ProvisionConstants.SHADOW_ALIAS_NAME).getStatus());
String[] queryStrings = { String[] queryStrings = {
"cancer", "cancer",
@ -200,7 +201,8 @@ public class SolrConfigExploreTest {
// System.out.println(rsp.getExplainMap()); // System.out.println(rsp.getExplainMap());
for (SolrDocument doc : rsp.getResults()) { for (SolrDocument doc : rsp.getResults()) {
log.info( log
.info(
doc.get("score") + "\t" + doc.get("score") + "\t" +
doc.get("__indexrecordidentifier") + "\t" + doc.get("__indexrecordidentifier") + "\t" +
doc.get("resultidentifier") + "\t" + doc.get("resultidentifier") + "\t" +
@ -216,7 +218,7 @@ public class SolrConfigExploreTest {
} }
protected static NamedList<Object> createCollection(CloudSolrClient client, String name, int numShards, protected static NamedList<Object> createCollection(CloudSolrClient client, String name, int numShards,
int replicationFactor, int maxShardsPerNode, String configName) throws Exception { int replicationFactor, int maxShardsPerNode, String configName) throws Exception {
ModifiableSolrParams modParams = new ModifiableSolrParams(); ModifiableSolrParams modParams = new ModifiableSolrParams();
modParams.set(CoreAdminParams.ACTION, CollectionParams.CollectionAction.CREATE.name()); modParams.set(CoreAdminParams.ACTION, CollectionParams.CollectionAction.CREATE.name());
modParams.set("name", name); modParams.set("name", name);

3
dhp-workflows/dhp-graph-provision/src/test/java/eu/dnetlib/dhp/oa/provision/SolrConfigTest.java
View File

@ -85,7 +85,8 @@ public class SolrConfigTest extends SolrTest {
new XmlIndexingJob(spark, inputPath, SHADOW_FORMAT, ProvisionConstants.SHADOW_ALIAS_NAME, batchSize) new XmlIndexingJob(spark, inputPath, SHADOW_FORMAT, ProvisionConstants.SHADOW_ALIAS_NAME, batchSize)
.run(isLookupClient); .run(isLookupClient);
Assertions.assertEquals(0, miniCluster.getSolrClient().commit(ProvisionConstants.SHADOW_ALIAS_NAME).getStatus()); Assertions
.assertEquals(0, miniCluster.getSolrClient().commit(ProvisionConstants.SHADOW_ALIAS_NAME).getStatus());
String[] queryStrings = { String[] queryStrings = {
"cancer", "cancer",