merging duplicates using IdentifierComparator

This commit is contained in:
Claudio Atzori 2022-11-11 16:10:25 +01:00
parent 5af5a8ae42
commit ddff0e8999
8 changed files with 590 additions and 111 deletions

View File

@ -1,12 +1,12 @@
package eu.dnetlib.dhp.oa.dedup;
import java.util.Collection;
import java.util.Comparator;
import java.util.Iterator;
import java.util.List;
import java.lang.reflect.InvocationTargetException;
import java.util.*;
import java.util.stream.Collectors;
import org.apache.commons.beanutils.BeanUtils;
import org.apache.commons.lang3.StringUtils;
import org.apache.spark.api.java.function.MapFunction;
import org.apache.spark.api.java.function.MapGroupsFunction;
import org.apache.spark.sql.Dataset;
@ -15,14 +15,12 @@ import org.apache.spark.sql.SparkSession;
import com.fasterxml.jackson.databind.DeserializationFeature;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.google.common.collect.Iterators;
import com.google.common.collect.Lists;
import eu.dnetlib.dhp.oa.dedup.model.Identifier;
import eu.dnetlib.dhp.oa.merge.AuthorMerger;
import eu.dnetlib.dhp.schema.common.ModelSupport;
import eu.dnetlib.dhp.schema.oaf.*;
import eu.dnetlib.dhp.schema.oaf.utils.IdentifierFactory;
import scala.Tuple2;
public class DedupRecordFactory {
@ -79,23 +77,25 @@ public class DedupRecordFactory {
public static <T extends OafEntity> T entityMerger(
String id, Iterator<Tuple2<String, T>> entities, long ts, DataInfo dataInfo, Class<T> clazz)
throws IllegalAccessException, InstantiationException {
throws IllegalAccessException, InstantiationException, InvocationTargetException {
T entity = clazz.newInstance();
entity.setDataInfo(dataInfo);
final Comparator<Identifier<T>> idComparator = new IdentifierComparator<>();
final Collection<String> dates = Lists.newArrayList();
final List<List<Author>> authors = Lists.newArrayList();
final Comparator<Identifier<T>> idComparator = new IdentifierComparator<T>().reversed();
final List<T> entityList = Lists
final LinkedList<T> entityList = Lists
.newArrayList(entities)
.stream()
.map(t -> Identifier.newInstance(t._2()))
.sorted(idComparator)
.map(Identifier::getEntity)
.collect(Collectors.toList());
.collect(Collectors.toCollection(LinkedList::new));
final T entity = clazz.newInstance();
final T first = entityList.removeFirst();
BeanUtils.copyProperties(entity, first);
final Collection<String> dates = Lists.newArrayList();
final List<List<Author>> authors = Lists.newArrayList();
entityList
.forEach(
@ -103,12 +103,11 @@ public class DedupRecordFactory {
entity.mergeFrom(duplicate);
if (ModelSupport.isSubClass(duplicate, Result.class)) {
Result r1 = (Result) duplicate;
if (r1.getAuthor() != null && !r1.getAuthor().isEmpty())
if (r1.getAuthor() != null && StringUtils.isNotBlank(r1.getDateofacceptance().getValue()))
authors.add(r1.getAuthor());
if (r1.getDateofacceptance() != null)
dates.add(r1.getDateofacceptance().getValue());
}
});
// set authors and date

View File

@ -7,6 +7,7 @@ import java.io.BufferedReader;
import java.io.FileReader;
import java.io.IOException;
import java.io.Serializable;
import java.lang.reflect.InvocationTargetException;
import java.nio.file.Paths;
import java.util.*;
import java.util.stream.Collectors;
@ -54,7 +55,7 @@ class EntityMergerTest implements Serializable {
}
@Test
void softwareMergerTest() throws InstantiationException, IllegalAccessException {
void softwareMergerTest() throws InstantiationException, IllegalAccessException, InvocationTargetException {
List<Tuple2<String, Software>> softwares = readSample(
testEntityBasePath + "/software_merge.json", Software.class);
@ -69,7 +70,7 @@ class EntityMergerTest implements Serializable {
}
@Test
void publicationMergerTest() throws InstantiationException, IllegalAccessException {
void publicationMergerTest() throws InstantiationException, IllegalAccessException, InvocationTargetException {
Publication pub_merged = DedupRecordFactory
.entityMerger(dedupId, publications.iterator(), 0, dataInfo, Publication.class);
@ -134,7 +135,7 @@ class EntityMergerTest implements Serializable {
}
@Test
void publicationMergerTest2() throws InstantiationException, IllegalAccessException {
void publicationMergerTest2() throws InstantiationException, IllegalAccessException, InvocationTargetException {
Publication pub_merged = DedupRecordFactory
.entityMerger(dedupId, publications2.iterator(), 0, dataInfo, Publication.class);
@ -146,7 +147,7 @@ class EntityMergerTest implements Serializable {
}
@Test
void publicationMergerTest3() throws InstantiationException, IllegalAccessException {
void publicationMergerTest3() throws InstantiationException, IllegalAccessException, InvocationTargetException {
Publication pub_merged = DedupRecordFactory
.entityMerger(dedupId, publications3.iterator(), 0, dataInfo, Publication.class);
@ -156,7 +157,8 @@ class EntityMergerTest implements Serializable {
}
@Test
void publicationMergerTest4() throws InstantiationException, IllegalStateException, IllegalAccessException {
void publicationMergerTest4()
throws InstantiationException, IllegalStateException, IllegalAccessException, InvocationTargetException {
Publication pub_merged = DedupRecordFactory
.entityMerger(dedupId, publications4.iterator(), 0, dataInfo, Publication.class);
@ -166,7 +168,8 @@ class EntityMergerTest implements Serializable {
}
@Test
void publicationMergerTest5() throws InstantiationException, IllegalStateException, IllegalAccessException {
void publicationMergerTest5()
throws InstantiationException, IllegalStateException, IllegalAccessException, InvocationTargetException {
System.out
.println(

View File

@ -0,0 +1,389 @@
package eu.dnetlib.dhp.oa.dedup;
import static java.nio.file.Files.createTempDirectory;
import static org.apache.spark.sql.functions.count;
import static org.junit.jupiter.api.Assertions.*;
import static org.mockito.Mockito.lenient;
import java.io.File;
import java.io.IOException;
import java.io.Serializable;
import java.net.URISyntaxException;
import java.nio.file.Paths;
import java.util.*;
import java.util.stream.Collectors;
import com.fasterxml.jackson.core.JsonProcessingException;
import org.apache.commons.io.FileUtils;
import org.apache.commons.io.IOUtils;
import org.apache.spark.SparkConf;
import org.apache.spark.api.java.JavaSparkContext;
import org.apache.spark.api.java.function.FilterFunction;
import org.apache.spark.api.java.function.FlatMapFunction;
import org.apache.spark.api.java.function.MapFunction;
import org.apache.spark.sql.Dataset;
import org.apache.spark.sql.Encoders;
import org.apache.spark.sql.SparkSession;
import org.junit.jupiter.api.*;
import org.junit.jupiter.api.extension.ExtendWith;
import org.mockito.Mock;
import org.mockito.Mockito;
import org.mockito.junit.jupiter.MockitoExtension;
import com.fasterxml.jackson.databind.DeserializationFeature;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.google.common.collect.Sets;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import eu.dnetlib.dhp.schema.common.ModelConstants;
import eu.dnetlib.dhp.schema.oaf.*;
import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpException;
import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpService;
@ExtendWith(MockitoExtension.class)
@TestMethodOrder(MethodOrderer.OrderAnnotation.class)
public class SparkDedupPublicationTest implements Serializable {
@Mock(serializable = true)
ISLookUpService isLookUpService;
private static SparkSession spark;
private static JavaSparkContext jsc;
private static String testGraphBasePath;
private static String testOutputBasePath;
private static String testDedupGraphBasePath;
private static final String testActionSetId = "test-orchestrator";
@BeforeAll
public static void cleanUp() throws IOException, URISyntaxException {
testGraphBasePath = Paths
.get(SparkDedupPublicationTest.class.getResource("/eu/dnetlib/dhp/dedup/entities2").toURI())
.toFile()
.getAbsolutePath();
testOutputBasePath = createTempDirectory(SparkDedupPublicationTest.class.getSimpleName() + "-")
.toAbsolutePath()
.toString();
testDedupGraphBasePath = createTempDirectory(SparkDedupPublicationTest.class.getSimpleName() + "-")
.toAbsolutePath()
.toString();
FileUtils.deleteDirectory(new File(testOutputBasePath));
FileUtils.deleteDirectory(new File(testDedupGraphBasePath));
final SparkConf conf = new SparkConf();
conf.set("spark.sql.shuffle.partitions", "10");
spark = SparkSession
.builder()
.appName(SparkDedupPublicationTest.class.getSimpleName())
.master("local[*]")
.config(conf)
.getOrCreate();
jsc = JavaSparkContext.fromSparkContext(spark.sparkContext());
}
@BeforeEach
public void setUp() throws IOException, ISLookUpException {
lenient()
.when(isLookUpService.getResourceProfileByQuery(Mockito.contains(testActionSetId)))
.thenReturn(classPathResourceAsString("/eu/dnetlib/dhp/dedup/profiles/mock_orchestrator_publication.xml"));
lenient()
.when(isLookUpService.getResourceProfileByQuery(Mockito.contains("publication")))
.thenReturn(classPathResourceAsString("/eu/dnetlib/dhp/dedup/conf/pub.curr.conf.json"));
}
@Test
@Order(1)
void createSimRelsTest() throws Exception {
ArgumentApplicationParser parser = new ArgumentApplicationParser(
classPathResourceAsString("/eu/dnetlib/dhp/oa/dedup/createSimRels_parameters.json"));
parser
.parseArgument(
new String[] {
"--graphBasePath", testGraphBasePath,
"--actionSetId", testActionSetId,
"--isLookUpUrl", "lookupurl",
"--workingPath", testOutputBasePath,
"--numPartitions", "5"
});
new SparkCreateSimRels(parser, spark).run(isLookUpService);
long pubs_simrel = spark
.read()
.load(DedupUtility.createSimRelPath(testOutputBasePath, testActionSetId, "publication"))
.count();
assertEquals(62, pubs_simrel);
}
@Test
@Order(2)
void cutMergeRelsTest() throws Exception {
ArgumentApplicationParser parser = new ArgumentApplicationParser(
classPathResourceAsString("/eu/dnetlib/dhp/oa/dedup/createCC_parameters.json"));
parser
.parseArgument(
new String[] {
"--graphBasePath", testGraphBasePath,
"--actionSetId", testActionSetId,
"--isLookUpUrl", "lookupurl",
"--workingPath", testOutputBasePath,
"--cutConnectedComponent", "3"
});
new SparkCreateMergeRels(parser, spark).run(isLookUpService);
long pubs_mergerel = spark
.read()
.load(testOutputBasePath + "/" + testActionSetId + "/publication_mergerel")
.as(Encoders.bean(Relation.class))
.filter((FilterFunction<Relation>) r -> r.getRelClass().equalsIgnoreCase("merges"))
.groupBy("source")
.agg(count("target").alias("cnt"))
.select("source", "cnt")
.where("cnt > 3")
.count();
assertEquals(0, pubs_mergerel);
FileUtils.deleteDirectory(new File(testOutputBasePath + "/" + testActionSetId + "/publication_mergerel"));
}
@Test
@Order(3)
void createMergeRelsTest() throws Exception {
ArgumentApplicationParser parser = new ArgumentApplicationParser(
classPathResourceAsString("/eu/dnetlib/dhp/oa/dedup/createCC_parameters.json"));
parser
.parseArgument(
new String[] {
"--graphBasePath", testGraphBasePath,
"--actionSetId", testActionSetId,
"--isLookUpUrl", "lookupurl",
"--workingPath", testOutputBasePath
});
new SparkCreateMergeRels(parser, spark).run(isLookUpService);
final Dataset<Relation> pubs = spark
.read()
.load(testOutputBasePath + "/" + testActionSetId + "/publication_mergerel")
.as(Encoders.bean(Relation.class));
final List<Relation> merges = pubs
.filter("source == '50|doi_dedup___::d5021b53204e4fdeab6ff5d5bc468032'")
.collectAsList();
assertEquals(3, merges.size());
Set<String> dups = Sets
.newHashSet(
"50|doi_________::3b1d0d8e8f930826665df9d6b82fbb73",
"50|doi_________::d5021b53204e4fdeab6ff5d5bc468032",
"50|arXiv_______::c93aeb433eb90ed7a86e29be00791b7c");
merges.forEach(r -> {
assertEquals(ModelConstants.RESULT_RESULT, r.getRelType());
assertEquals(ModelConstants.DEDUP, r.getSubRelType());
assertEquals(ModelConstants.MERGES, r.getRelClass());
assertTrue(dups.contains(r.getTarget()));
});
final List<Relation> mergedIn = pubs
.filter("target == '50|doi_dedup___::d5021b53204e4fdeab6ff5d5bc468032'")
.collectAsList();
assertEquals(3, mergedIn.size());
mergedIn.forEach(r -> {
assertEquals(ModelConstants.RESULT_RESULT, r.getRelType());
assertEquals(ModelConstants.DEDUP, r.getSubRelType());
assertEquals(ModelConstants.IS_MERGED_IN, r.getRelClass());
assertTrue(dups.contains(r.getSource()));
});
assertEquals(24, pubs.count());
}
@Test
@Order(4)
void createDedupRecordTest() throws Exception {
ArgumentApplicationParser parser = new ArgumentApplicationParser(
classPathResourceAsString("/eu/dnetlib/dhp/oa/dedup/createDedupRecord_parameters.json"));
parser
.parseArgument(
new String[] {
"--graphBasePath", testGraphBasePath,
"--actionSetId", testActionSetId,
"--isLookUpUrl", "lookupurl",
"--workingPath", testOutputBasePath
});
new SparkCreateDedupRecord(parser, spark).run(isLookUpService);
final ObjectMapper mapper = new ObjectMapper()
.configure(DeserializationFeature.FAIL_ON_UNKNOWN_PROPERTIES, false);
final Dataset<Publication> roots = spark
.read()
.textFile(testOutputBasePath + "/" + testActionSetId + "/publication_deduprecord")
.map(
(MapFunction<String, Publication>) value -> mapper.readValue(value, Publication.class),
Encoders.bean(Publication.class));
assertEquals(2, roots.count());
final Dataset<Publication> pubs = spark
.read()
.textFile(DedupUtility.createEntityPath(testGraphBasePath, "publication"))
.map(
(MapFunction<String, Publication>) value -> mapper.readValue(value, Publication.class),
Encoders.bean(Publication.class));
verifyRoot_case_1(roots, pubs);
verifyRoot_case_2(roots, pubs);
}
private static void verifyRoot_case_1(Dataset<Publication> roots, Dataset<Publication> pubs) {
Publication root = roots
.filter("id = '50|doi_dedup___::d5021b53204e4fdeab6ff5d5bc468032'")
.first();
assertNotNull(root);
Publication crossref_duplicate = pubs
.filter("id = '50|doi_________::d5021b53204e4fdeab6ff5d5bc468032'")
.collectAsList()
.get(0);
assertEquals(crossref_duplicate.getJournal().getName(), root.getJournal().getName());
assertEquals(crossref_duplicate.getJournal().getIssnPrinted(), root.getJournal().getIssnPrinted());
assertEquals(crossref_duplicate.getPublisher().getValue(), root.getPublisher().getValue());
Set<String> rootPids = root
.getPid()
.stream()
.map(StructuredProperty::getValue)
.collect(Collectors.toCollection(HashSet::new));
Set<String> dupPids = crossref_duplicate
.getPid()
.stream()
.map(StructuredProperty::getValue)
.collect(Collectors.toCollection(HashSet::new));
assertFalse(Sets.intersection(rootPids, dupPids).isEmpty());
assertTrue(rootPids.contains("10.1109/jstqe.2022.3205716"));
Optional<Instance> instance_cr = root
.getInstance()
.stream()
.filter(i -> i.getCollectedfrom().getValue().equals("Crossref"))
.findFirst();
assertTrue(instance_cr.isPresent());
assertEquals("OPEN", instance_cr.get().getAccessright().getClassid());
assertEquals("Open Access", instance_cr.get().getAccessright().getClassname());
assertEquals(OpenAccessRoute.hybrid, instance_cr.get().getAccessright().getOpenAccessRoute());
assertEquals(
"IEEE Journal of Selected Topics in Quantum Electronics", instance_cr.get().getHostedby().getValue());
assertEquals("0001", instance_cr.get().getInstancetype().getClassid());
assertEquals("Article", instance_cr.get().getInstancetype().getClassname());
}
private void verifyRoot_case_2(Dataset<Publication> roots, Dataset<Publication> pubs) throws JsonProcessingException {
Publication root = roots
.filter("id = '50|doi_dedup___::18aff3b55fb6876466a5d4bd82434885'")
.first();
assertNotNull(root);
Publication crossref_duplicate = pubs
.filter("id = '50|doi_________::18aff3b55fb6876466a5d4bd82434885'")
.first();
//System.err.println(new ObjectMapper().writeValueAsString(root));
assertEquals(crossref_duplicate.getJournal().getName(), root.getJournal().getName());
assertEquals(crossref_duplicate.getJournal().getIssnOnline(), root.getJournal().getIssnOnline());
assertEquals(crossref_duplicate.getJournal().getVol(), root.getJournal().getVol());
assertEquals(crossref_duplicate.getPublisher().getValue(), root.getPublisher().getValue());
Set<String> dups_cf = pubs
.collectAsList()
.stream()
.flatMap(p -> p.getCollectedfrom().stream())
.map(KeyValue::getValue)
.collect(Collectors.toCollection(HashSet::new));
Set<String> root_cf = root
.getCollectedfrom()
.stream()
.map(KeyValue::getValue)
.collect(Collectors.toCollection(HashSet::new));
assertTrue(Sets.difference(root_cf, dups_cf).isEmpty());
}
@Test
@Order(6)
void updateEntityTest() throws Exception {
ArgumentApplicationParser parser = new ArgumentApplicationParser(
classPathResourceAsString("/eu/dnetlib/dhp/oa/dedup/updateEntity_parameters.json"));
parser
.parseArgument(
new String[] {
"-i", testGraphBasePath, "-w", testOutputBasePath, "-o", testDedupGraphBasePath
});
new SparkUpdateEntity(parser, spark).run(isLookUpService);
long publications = jsc.textFile(testDedupGraphBasePath + "/publication").count();
long mergedPubs = spark
.read()
.load(testOutputBasePath + "/" + testActionSetId + "/publication_mergerel")
.as(Encoders.bean(Relation.class))
.where("relClass=='merges'")
.javaRDD()
.map(Relation::getTarget)
.distinct()
.count();
assertEquals(14, publications);
long deletedPubs = jsc
.textFile(testDedupGraphBasePath + "/publication")
.filter(this::isDeletedByInference)
.count();
assertEquals(mergedPubs, deletedPubs);
}
@AfterAll
public static void finalCleanUp() throws IOException {
FileUtils.deleteDirectory(new File(testOutputBasePath));
FileUtils.deleteDirectory(new File(testDedupGraphBasePath));
}
public boolean isDeletedByInference(String s) {
return s.contains("\"deletedbyinference\":true");
}
private static String classPathResourceAsString(String path) throws IOException {
return IOUtils
.toString(
SparkDedupPublicationTest.class
.getResourceAsStream(path));
}
}

View File

@ -4,8 +4,7 @@ package eu.dnetlib.dhp.oa.dedup;
import static java.nio.file.Files.createTempDirectory;
import static org.apache.spark.sql.functions.count;
import static org.junit.jupiter.api.Assertions.assertEquals;
import static org.junit.jupiter.api.Assertions.assertTrue;
import static org.junit.jupiter.api.Assertions.*;
import static org.mockito.Mockito.lenient;
import java.io.File;
@ -14,7 +13,11 @@ import java.io.IOException;
import java.io.Serializable;
import java.net.URISyntaxException;
import java.nio.file.Paths;
import java.util.HashSet;
import java.util.List;
import java.util.Optional;
import java.util.Set;
import java.util.stream.Collectors;
import org.apache.commons.io.FileUtils;
import org.apache.commons.io.IOUtils;
@ -35,10 +38,13 @@ import org.mockito.Mock;
import org.mockito.Mockito;
import org.mockito.junit.jupiter.MockitoExtension;
import com.fasterxml.jackson.databind.DeserializationFeature;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.google.common.collect.Sets;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import eu.dnetlib.dhp.schema.oaf.Relation;
import eu.dnetlib.dhp.schema.common.ModelConstants;
import eu.dnetlib.dhp.schema.oaf.*;
import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpException;
import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpService;
import eu.dnetlib.pace.util.MapDocumentUtil;
@ -105,57 +111,27 @@ public class SparkDedupTest implements Serializable {
lenient()
.when(isLookUpService.getResourceProfileByQuery(Mockito.contains(testActionSetId)))
.thenReturn(
IOUtils
.toString(
SparkDedupTest.class
.getResourceAsStream(
"/eu/dnetlib/dhp/dedup/profiles/mock_orchestrator.xml")));
.thenReturn(classPathResourceAsString("/eu/dnetlib/dhp/dedup/profiles/mock_orchestrator.xml"));
lenient()
.when(isLookUpService.getResourceProfileByQuery(Mockito.contains("organization")))
.thenReturn(
IOUtils
.toString(
SparkDedupTest.class
.getResourceAsStream(
"/eu/dnetlib/dhp/dedup/conf/org.curr.conf.json")));
.thenReturn(classPathResourceAsString("/eu/dnetlib/dhp/dedup/conf/org.curr.conf.json"));
lenient()
.when(isLookUpService.getResourceProfileByQuery(Mockito.contains("publication")))
.thenReturn(
IOUtils
.toString(
SparkDedupTest.class
.getResourceAsStream(
"/eu/dnetlib/dhp/dedup/conf/pub.curr.conf.json")));
.thenReturn(classPathResourceAsString("/eu/dnetlib/dhp/dedup/conf/pub.curr.conf.json"));
lenient()
.when(isLookUpService.getResourceProfileByQuery(Mockito.contains("software")))
.thenReturn(
IOUtils
.toString(
SparkDedupTest.class
.getResourceAsStream(
"/eu/dnetlib/dhp/dedup/conf/sw.curr.conf.json")));
.thenReturn(classPathResourceAsString("/eu/dnetlib/dhp/dedup/conf/sw.curr.conf.json"));
lenient()
.when(isLookUpService.getResourceProfileByQuery(Mockito.contains("dataset")))
.thenReturn(
IOUtils
.toString(
SparkDedupTest.class
.getResourceAsStream(
"/eu/dnetlib/dhp/dedup/conf/ds.curr.conf.json")));
.thenReturn(classPathResourceAsString("/eu/dnetlib/dhp/dedup/conf/ds.curr.conf.json"));
lenient()
.when(isLookUpService.getResourceProfileByQuery(Mockito.contains("otherresearchproduct")))
.thenReturn(
IOUtils
.toString(
SparkDedupTest.class
.getResourceAsStream(
"/eu/dnetlib/dhp/dedup/conf/orp.curr.conf.json")));
.thenReturn(classPathResourceAsString("/eu/dnetlib/dhp/dedup/conf/orp.curr.conf.json"));
}
@Test
@ -163,11 +139,7 @@ public class SparkDedupTest implements Serializable {
void createSimRelsTest() throws Exception {
ArgumentApplicationParser parser = new ArgumentApplicationParser(
IOUtils
.toString(
SparkCreateSimRels.class
.getResourceAsStream(
"/eu/dnetlib/dhp/oa/dedup/createSimRels_parameters.json")));
classPathResourceAsString("/eu/dnetlib/dhp/oa/dedup/createSimRels_parameters.json"));
parser
.parseArgument(
@ -207,7 +179,7 @@ public class SparkDedupTest implements Serializable {
.count();
assertEquals(3076, orgs_simrel);
assertEquals(7040, pubs_simrel);
assertEquals(7046, pubs_simrel);
assertEquals(336, sw_simrel);
assertEquals(442, ds_simrel);
assertEquals(6784, orp_simrel);
@ -223,11 +195,7 @@ public class SparkDedupTest implements Serializable {
void whitelistSimRelsTest() throws Exception {
ArgumentApplicationParser parser = new ArgumentApplicationParser(
IOUtils
.toString(
SparkWhitelistSimRels.class
.getResourceAsStream(
"/eu/dnetlib/dhp/oa/dedup/whitelistSimRels_parameters.json")));
classPathResourceAsString("/eu/dnetlib/dhp/oa/dedup/whitelistSimRels_parameters.json"));
parser
.parseArgument(
@ -264,7 +232,7 @@ public class SparkDedupTest implements Serializable {
// entities simrels supposed to be equal to the number of previous step (no rels in whitelist)
assertEquals(3076, orgs_simrel);
assertEquals(7040, pubs_simrel);
assertEquals(7046, pubs_simrel);
assertEquals(442, ds_simrel);
assertEquals(6784, orp_simrel);
// System.out.println("orgs_simrel = " + orgs_simrel);
@ -306,11 +274,7 @@ public class SparkDedupTest implements Serializable {
void cutMergeRelsTest() throws Exception {
ArgumentApplicationParser parser = new ArgumentApplicationParser(
IOUtils
.toString(
SparkCreateMergeRels.class
.getResourceAsStream(
"/eu/dnetlib/dhp/oa/dedup/createCC_parameters.json")));
classPathResourceAsString("/eu/dnetlib/dhp/oa/dedup/createCC_parameters.json"));
parser
.parseArgument(
@ -402,11 +366,7 @@ public class SparkDedupTest implements Serializable {
void createMergeRelsTest() throws Exception {
ArgumentApplicationParser parser = new ArgumentApplicationParser(
IOUtils
.toString(
SparkCreateMergeRels.class
.getResourceAsStream(
"/eu/dnetlib/dhp/oa/dedup/createCC_parameters.json")));
classPathResourceAsString("/eu/dnetlib/dhp/oa/dedup/createCC_parameters.json"));
parser
.parseArgument(
@ -427,10 +387,10 @@ public class SparkDedupTest implements Serializable {
.read()
.load(testOutputBasePath + "/" + testActionSetId + "/organization_mergerel")
.count();
long pubs_mergerel = spark
final Dataset<Relation> pubs = spark
.read()
.load(testOutputBasePath + "/" + testActionSetId + "/publication_mergerel")
.count();
.as(Encoders.bean(Relation.class));
long sw_mergerel = spark
.read()
.load(testOutputBasePath + "/" + testActionSetId + "/software_mergerel")
@ -445,8 +405,35 @@ public class SparkDedupTest implements Serializable {
.load(testOutputBasePath + "/" + testActionSetId + "/otherresearchproduct_mergerel")
.count();
final List<Relation> merges = pubs
.filter("source == '50|doi_dedup___::d5021b53204e4fdeab6ff5d5bc468032'")
.collectAsList();
assertEquals(3, merges.size());
Set<String> dups = Sets
.newHashSet(
"50|doi_________::3b1d0d8e8f930826665df9d6b82fbb73",
"50|doi_________::d5021b53204e4fdeab6ff5d5bc468032",
"50|arXiv_______::c93aeb433eb90ed7a86e29be00791b7c");
merges.forEach(r -> {
assertEquals(ModelConstants.RESULT_RESULT, r.getRelType());
assertEquals(ModelConstants.DEDUP, r.getSubRelType());
assertEquals(ModelConstants.MERGES, r.getRelClass());
assertTrue(dups.contains(r.getTarget()));
});
final List<Relation> mergedIn = pubs
.filter("target == '50|doi_dedup___::d5021b53204e4fdeab6ff5d5bc468032'")
.collectAsList();
assertEquals(3, mergedIn.size());
mergedIn.forEach(r -> {
assertEquals(ModelConstants.RESULT_RESULT, r.getRelType());
assertEquals(ModelConstants.DEDUP, r.getSubRelType());
assertEquals(ModelConstants.IS_MERGED_IN, r.getRelClass());
assertTrue(dups.contains(r.getSource()));
});
assertEquals(1268, orgs_mergerel);
assertEquals(1444, pubs_mergerel);
assertEquals(1450, pubs.count());
assertEquals(286, sw_mergerel);
assertEquals(472, ds_mergerel);
assertEquals(738, orp_mergerel);
@ -463,11 +450,7 @@ public class SparkDedupTest implements Serializable {
void createDedupRecordTest() throws Exception {
ArgumentApplicationParser parser = new ArgumentApplicationParser(
IOUtils
.toString(
SparkCreateDedupRecord.class
.getResourceAsStream(
"/eu/dnetlib/dhp/oa/dedup/createDedupRecord_parameters.json")));
classPathResourceAsString("/eu/dnetlib/dhp/oa/dedup/createDedupRecord_parameters.json"));
parser
.parseArgument(
new String[] {
@ -483,12 +466,18 @@ public class SparkDedupTest implements Serializable {
new SparkCreateDedupRecord(parser, spark).run(isLookUpService);
final ObjectMapper mapper = new ObjectMapper()
.configure(DeserializationFeature.FAIL_ON_UNKNOWN_PROPERTIES, false);
final Dataset<Publication> pubs = spark
.read()
.textFile(testOutputBasePath + "/" + testActionSetId + "/publication_deduprecord")
.map(
(MapFunction<String, Publication>) value -> mapper.readValue(value, Publication.class),
Encoders.bean(Publication.class));
long orgs_deduprecord = jsc
.textFile(testOutputBasePath + "/" + testActionSetId + "/organization_deduprecord")
.count();
long pubs_deduprecord = jsc
.textFile(testOutputBasePath + "/" + testActionSetId + "/publication_deduprecord")
.count();
long sw_deduprecord = jsc
.textFile(testOutputBasePath + "/" + testActionSetId + "/software_deduprecord")
.count();
@ -499,11 +488,13 @@ public class SparkDedupTest implements Serializable {
.count();
assertEquals(86, orgs_deduprecord);
assertEquals(67, pubs_deduprecord);
assertEquals(68, pubs.count());
assertEquals(49, sw_deduprecord);
assertEquals(97, ds_deduprecord);
assertEquals(92, orp_deduprecord);
verifyRoot_1(mapper, pubs);
// System.out.println("orgs_deduprecord = " + orgs_deduprecord);
// System.out.println("pubs_deduprecord = " + pubs_deduprecord);
// System.out.println("sw_deduprecord = " + sw_deduprecord);
@ -511,16 +502,63 @@ public class SparkDedupTest implements Serializable {
// System.out.println("orp_deduprecord = " + orp_deduprecord);
}
private static void verifyRoot_1(ObjectMapper mapper, Dataset<Publication> pubs) {
Publication root = pubs
.filter("id = '50|doi_dedup___::d5021b53204e4fdeab6ff5d5bc468032'")
.first();
assertNotNull(root);
final Dataset<String> publication = spark
.read()
.textFile(DedupUtility.createEntityPath(testGraphBasePath, "publication"));
Publication crossref_duplicate = publication
.map(
(MapFunction<String, Publication>) value -> mapper.readValue(value, Publication.class),
Encoders.bean(Publication.class))
.filter("id = '50|doi_________::d5021b53204e4fdeab6ff5d5bc468032'")
.collectAsList()
.get(0);
assertEquals(crossref_duplicate.getJournal().getName(), root.getJournal().getName());
assertEquals(crossref_duplicate.getJournal().getIssnPrinted(), root.getJournal().getIssnPrinted());
assertEquals(crossref_duplicate.getPublisher().getValue(), root.getPublisher().getValue());
Set<String> rootPids = root
.getPid()
.stream()
.map(StructuredProperty::getValue)
.collect(Collectors.toCollection(HashSet::new));
Set<String> dupPids = crossref_duplicate
.getPid()
.stream()
.map(StructuredProperty::getValue)
.collect(Collectors.toCollection(HashSet::new));
assertFalse(Sets.intersection(rootPids, dupPids).isEmpty());
assertTrue(rootPids.contains("10.1109/jstqe.2022.3205716"));
Optional<Instance> instance_cr = root
.getInstance()
.stream()
.filter(i -> i.getCollectedfrom().getValue().equals("Crossref"))
.findFirst();
assertTrue(instance_cr.isPresent());
assertEquals("OPEN", instance_cr.get().getAccessright().getClassid());
assertEquals("Open Access", instance_cr.get().getAccessright().getClassname());
assertEquals(OpenAccessRoute.hybrid, instance_cr.get().getAccessright().getOpenAccessRoute());
assertEquals(
"IEEE Journal of Selected Topics in Quantum Electronics", instance_cr.get().getHostedby().getValue());
assertEquals("0001", instance_cr.get().getInstancetype().getClassid());
assertEquals("Article", instance_cr.get().getInstancetype().getClassname());
}
@Test
@Order(6)
void updateEntityTest() throws Exception {
ArgumentApplicationParser parser = new ArgumentApplicationParser(
IOUtils
.toString(
SparkUpdateEntity.class
.getResourceAsStream(
"/eu/dnetlib/dhp/oa/dedup/updateEntity_parameters.json")));
classPathResourceAsString("/eu/dnetlib/dhp/oa/dedup/updateEntity_parameters.json"));
parser
.parseArgument(
new String[] {
@ -587,7 +625,7 @@ public class SparkDedupTest implements Serializable {
.distinct()
.count();
assertEquals(898, publications);
assertEquals(902, publications);
assertEquals(839, organizations);
assertEquals(100, projects);
assertEquals(100, datasource);
@ -640,11 +678,7 @@ public class SparkDedupTest implements Serializable {
void propagateRelationTest() throws Exception {
ArgumentApplicationParser parser = new ArgumentApplicationParser(
IOUtils
.toString(
SparkPropagateRelation.class
.getResourceAsStream(
"/eu/dnetlib/dhp/oa/dedup/propagateRelation_parameters.json")));
classPathResourceAsString("/eu/dnetlib/dhp/oa/dedup/propagateRelation_parameters.json"));
parser
.parseArgument(
new String[] {
@ -714,4 +748,12 @@ public class SparkDedupTest implements Serializable {
public boolean isDeletedByInference(String s) {
return s.contains("\"deletedbyinference\":true");
}
private static String classPathResourceAsString(String path) throws IOException {
return IOUtils
.toString(
SparkDedupTest.class
.getResourceAsStream(path));
}
}

View File

@ -0,0 +1,24 @@
<RESOURCE_PROFILE>
<HEADER>
<RESOURCE_IDENTIFIER value=""/>
<RESOURCE_TYPE value="DedupOrchestrationDSResourceType"/>
<RESOURCE_KIND value="DedupOrchestrationDSResources"/>
<RESOURCE_URI value=""/>
<DATE_OF_CREATION value="2001-12-31T12:00:00"/>
</HEADER>
<BODY>
<CONFIGURATION enabled="true">
<DEDUPLICATION>
<ENTITY code="20" label="Organization" name="organization"/>
<ACTION_SET id="test-orchestrator"/>
<SCAN_SEQUENCE>
<SCAN id="publication"/>
</SCAN_SEQUENCE>
</DEDUPLICATION>
</CONFIGURATION>
<STATUS>
<LAST_UPDATE value="2001-12-31T12:00:00"/>
</STATUS>
<SECURITY_PARAMETERS>SECURITY_PARAMETERS</SECURITY_PARAMETERS>
</BODY>
</RESOURCE_PROFILE>

View File

@ -19,7 +19,29 @@ log4j.logger.org.eclipse.jetty=WARN
log4j.logger.org.apache.hadoop.mapreduce.lib.output.FileOutputCommitterFactory=WARN
log4j.logger.org.apache.hadoop.mapreduce.lib.output.FileOutputCommitter=WARN
log4j.logger.org.apache.parquet.hadoop.ParquetOutputFormat=WARN
log4j.logger.org.apache.parquet.hadoop.InternalParquetRecordWriter=WARN
#log4j.logger.org.apache.parquet.hadoop.ParquetOutputFormat=WARN
#log4j.logger.org.apache.parquet.hadoop.InternalParquetRecordWriter=WARN
log4j.logger.org.apache.hadoop.io.compress.CodecPool=WARN
log4j.logger.org.apache.parquet.hadoop.codec.CodecConfig=WARN
#log4j.logger.org.apache.hadoop.io.compress=WARN
#log4j.logger.org.apache.parquet.hadoop.codec.CodecConfig=WARN
log4j.logger.parquet.hadoop.ColumnChunkPageWriteStore=ERROR
log4j.logger.com.jayway.jsonpath.internal.path.CompiledPath=WARN
log4j.logger.org.apache.parquet.hadoop.ParquetRecordReader=ERROR
log4j.logger.parquet.hadoop=WARN
log4j.logger.org.eclipse.jetty.server.handler.ContextHandlerCollection=WARN
log4j.logger.org.spark_project.jetty.util.component.ContainerLifeCycle=WARN
log4j.logger.org.apache.hadoop.mapred.FileInputFormat=WARN
log4j.logger.org.spark_project.jetty.servlet.ServletHandler=WARN
log4j.logger.org.apache.commons.beanutils.converters.BooleanConverter=WARN
log4j.logger.org.apache.commons.beanutils.converters.StringConverter=WARN
log4j.logger.org.apache.commons.beanutils.converters.LongConverter=WARN
log4j.logger.org.apache.commons.beanutils.converters.ArrayConverter=WARN
log4j.logger.org.apache.commons.beanutils.converters.FloatConverter=WARN
log4j.logger.org.apache.commons.beanutils.converters.IntegerConverter=WARN
log4j.logger.org.apache.commons.beanutils.converters.DoubleConverter=WARN
log4j.logger.org.apache.commons.beanutils.converters.CharacterConverter=WARN
log4j.logger.org.apache.commons.beanutils.converters.ByteConverter=WARN
log4j.logger.org.apache.commons.beanutils.converters.BigIntegerConverter=WARN
log4j.logger.org.apache.commons.beanutils.converters.BigDecimalConverter=WARN
log4j.logger.org.apache.commons.beanutils.converters.ShortConverter=WARN
log4j.logger.org.apache.commons.beanutils.BeanUtils=WARN