From 853333bdde97db3d7ddc4c82fbe6ab5fa486ce83 Mon Sep 17 00:00:00 2001 From: miconis Date: Mon, 20 Sep 2021 16:21:47 +0200 Subject: [PATCH] implementation of the whitelist for similarity relations --- .../dhp/oa/dedup/SparkWhitelistSimRels.java | 151 +++ .../dhp/oa/dedup/scan/oozie_app/workflow.xml | 32 + .../oa/dedup/whitelistSimRels_parameters.json | 38 + .../dnetlib/dhp/oa/dedup/SparkDedupTest.java | 1181 +++++++++-------- .../dnetlib/dhp/dedup/whitelist.simrels.txt | 2 + 5 files changed, 856 insertions(+), 548 deletions(-) create mode 100644 dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkWhitelistSimRels.java create mode 100644 dhp-workflows/dhp-dedup-openaire/src/main/resources/eu/dnetlib/dhp/oa/dedup/whitelistSimRels_parameters.json create mode 100644 dhp-workflows/dhp-dedup-openaire/src/test/resources/eu/dnetlib/dhp/dedup/whitelist.simrels.txt diff --git a/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkWhitelistSimRels.java b/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkWhitelistSimRels.java new file mode 100644 index 000000000..fa7d33570 --- /dev/null +++ b/dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkWhitelistSimRels.java @@ -0,0 +1,151 @@ +package eu.dnetlib.dhp.oa.dedup; + +import java.io.IOException; +import java.util.Optional; + +import org.apache.commons.io.IOUtils; +import org.apache.spark.SparkConf; +import org.apache.spark.api.java.JavaPairRDD; +import org.apache.spark.api.java.JavaRDD; +import org.apache.spark.api.java.JavaSparkContext; +import org.apache.spark.api.java.function.FilterFunction; +import org.apache.spark.api.java.function.MapFunction; +import org.apache.spark.api.java.function.PairFunction; +import org.apache.spark.sql.Dataset; +import org.apache.spark.sql.Encoders; +import org.apache.spark.sql.SaveMode; +import org.apache.spark.sql.SparkSession; +import org.dom4j.DocumentException; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; +import org.xml.sax.SAXException; + +import eu.dnetlib.dhp.application.ArgumentApplicationParser; +import eu.dnetlib.dhp.oa.dedup.model.Block; +import eu.dnetlib.dhp.schema.oaf.DataInfo; +import eu.dnetlib.dhp.schema.oaf.Relation; +import eu.dnetlib.dhp.utils.ISLookupClientFactory; +import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpException; +import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpService; +import eu.dnetlib.pace.config.DedupConfig; +import eu.dnetlib.pace.model.MapDocument; +import eu.dnetlib.pace.util.MapDocumentUtil; +import scala.Tuple2; +import scala.Tuple3; + +public class SparkWhitelistSimRels extends AbstractSparkAction { + + private static final Logger log = LoggerFactory.getLogger(SparkCreateSimRels.class); + + private static final String WHITELIST_SEPARATOR = "####"; + + public SparkWhitelistSimRels(ArgumentApplicationParser parser, SparkSession spark) { + super(parser, spark); + } + + public static void main(String[] args) throws Exception { + ArgumentApplicationParser parser = new ArgumentApplicationParser( + IOUtils + .toString( + SparkCreateSimRels.class + .getResourceAsStream( + "/eu/dnetlib/dhp/oa/dedup/whitelistSimRels_parameters.json"))); + parser.parseArgument(args); + + SparkConf conf = new SparkConf(); + new SparkWhitelistSimRels(parser, getSparkSession(conf)) + .run(ISLookupClientFactory.getLookUpService(parser.get("isLookUpUrl"))); + } + + @Override + public void run(ISLookUpService isLookUpService) + throws DocumentException, IOException, ISLookUpException, SAXException { + + // read oozie parameters + final String graphBasePath = parser.get("graphBasePath"); + final String isLookUpUrl = parser.get("isLookUpUrl"); + final String actionSetId = parser.get("actionSetId"); + final String workingPath = parser.get("workingPath"); + final int numPartitions = Optional + .ofNullable(parser.get("numPartitions")) + .map(Integer::valueOf) + .orElse(NUM_PARTITIONS); + final String whiteListPath = parser.get("whiteListPath"); + + log.info("numPartitions: '{}'", numPartitions); + log.info("graphBasePath: '{}'", graphBasePath); + log.info("isLookUpUrl: '{}'", isLookUpUrl); + log.info("actionSetId: '{}'", actionSetId); + log.info("workingPath: '{}'", workingPath); + log.info("whiteListPath: '{}'", whiteListPath); + + JavaSparkContext sc = JavaSparkContext.fromSparkContext(spark.sparkContext()); + + //file format: source####target + Dataset> whiteListRels = spark.createDataset(sc + .textFile(whiteListPath) + //check if the line is in the correct format: id1####id2 + .filter(s -> s.contains(WHITELIST_SEPARATOR) && s.split(WHITELIST_SEPARATOR).length == 2) + .map(s -> new Tuple2<>(s.split(WHITELIST_SEPARATOR)[0], s.split(WHITELIST_SEPARATOR)[1])) + .rdd(), + Encoders.tuple(Encoders.STRING(), Encoders.STRING())); + + // for each dedup configuration + for (DedupConfig dedupConf : getConfigurations(isLookUpService, actionSetId)) { + + final String entity = dedupConf.getWf().getEntityType(); + final String subEntity = dedupConf.getWf().getSubEntityValue(); + log.info("Adding whitelist simrels for: '{}'", subEntity); + + final String outputPath = DedupUtility.createSimRelPath(workingPath, actionSetId, subEntity); + + Dataset> entities = spark.createDataset(sc + .textFile(DedupUtility.createEntityPath(graphBasePath, subEntity)) + .repartition(numPartitions) + .mapToPair( + (PairFunction) s -> { + MapDocument d = MapDocumentUtil.asMapDocumentWithJPath(dedupConf, s); + return new Tuple2<>(d.getIdentifier(), "present"); + }) + .rdd(), + Encoders.tuple(Encoders.STRING(), Encoders.STRING())); + + Dataset> whiteListRels1 = whiteListRels + .joinWith(entities, whiteListRels.col("_1").equalTo(entities.col("_1")), "inner") + .map((MapFunction, Tuple2>, Tuple2>) Tuple2::_1, Encoders.tuple(Encoders.STRING(), Encoders.STRING())); + + Dataset> whiteListRels2 = whiteListRels1 + .joinWith(entities, whiteListRels1.col("_2").equalTo(entities.col("_1")), "inner") + .map((MapFunction, Tuple2>, Tuple2>) Tuple2::_1, Encoders.tuple(Encoders.STRING(), Encoders.STRING())); + + Dataset whiteListSimRels = whiteListRels2 + .map((MapFunction, Relation>) + r -> createSimRel(r._1(), r._2(), entity), + Encoders.bean(Relation.class) + ); + + saveParquet(whiteListSimRels, outputPath, SaveMode.Append); + } + } + + private Relation createSimRel(String source, String target, String entity) { + final Relation r = new Relation(); + r.setSource(source); + r.setTarget(target); + r.setSubRelType("dedupSimilarity"); + r.setRelClass("isSimilarTo"); + r.setDataInfo(new DataInfo()); + + switch (entity) { + case "result": + r.setRelType("resultResult"); + break; + case "organization": + r.setRelType("organizationOrganization"); + break; + default: + throw new IllegalArgumentException("unmanaged entity type: " + entity); + } + return r; + } +} diff --git a/dhp-workflows/dhp-dedup-openaire/src/main/resources/eu/dnetlib/dhp/oa/dedup/scan/oozie_app/workflow.xml b/dhp-workflows/dhp-dedup-openaire/src/main/resources/eu/dnetlib/dhp/oa/dedup/scan/oozie_app/workflow.xml index 342d83e8e..02fdd8431 100644 --- a/dhp-workflows/dhp-dedup-openaire/src/main/resources/eu/dnetlib/dhp/oa/dedup/scan/oozie_app/workflow.xml +++ b/dhp-workflows/dhp-dedup-openaire/src/main/resources/eu/dnetlib/dhp/oa/dedup/scan/oozie_app/workflow.xml @@ -20,6 +20,10 @@ workingPath path for the working directory + + whiteListPath + path for the whitelist of similarity relations + dedupGraphPath path for the output graph @@ -130,6 +134,34 @@ --workingPath${workingPath} --numPartitions8000 + + + + + + + yarn + cluster + Add Whitelist Similarity Relations + eu.dnetlib.dhp.oa.dedup.SparkWhitelistSimRels + dhp-dedup-openaire-${projectVersion}.jar + + --executor-memory=${sparkExecutorMemory} + --executor-cores=${sparkExecutorCores} + --driver-memory=${sparkDriverMemory} + --conf spark.extraListeners=${spark2ExtraListeners} + --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners} + --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress} + --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir} + --conf spark.sql.shuffle.partitions=3840 + + --graphBasePath${graphBasePath} + --isLookUpUrl${isLookUpUrl} + --actionSetId${actionSetId} + --workingPath${workingPath} + --whiteListPath${whiteListPath} + --numPartitions8000 + diff --git a/dhp-workflows/dhp-dedup-openaire/src/main/resources/eu/dnetlib/dhp/oa/dedup/whitelistSimRels_parameters.json b/dhp-workflows/dhp-dedup-openaire/src/main/resources/eu/dnetlib/dhp/oa/dedup/whitelistSimRels_parameters.json new file mode 100644 index 000000000..0a5cad7c4 --- /dev/null +++ b/dhp-workflows/dhp-dedup-openaire/src/main/resources/eu/dnetlib/dhp/oa/dedup/whitelistSimRels_parameters.json @@ -0,0 +1,38 @@ +[ + { + "paramName": "la", + "paramLongName": "isLookUpUrl", + "paramDescription": "address for the LookUp", + "paramRequired": true + }, + { + "paramName": "asi", + "paramLongName": "actionSetId", + "paramDescription": "action set identifier (name of the orchestrator)", + "paramRequired": true + }, + { + "paramName": "i", + "paramLongName": "graphBasePath", + "paramDescription": "the base path of the raw graph", + "paramRequired": true + }, + { + "paramName": "w", + "paramLongName": "workingPath", + "paramDescription": "path of the working directory", + "paramRequired": true + }, + { + "paramName": "np", + "paramLongName": "numPartitions", + "paramDescription": "number of partitions for the similarity relations intermediate phases", + "paramRequired": false + }, + { + "paramName": "wl", + "paramLongName": "whiteListPath", + "paramDescription": "whitelist file path for the addition of custom simrels", + "paramRequired": true + } +] \ No newline at end of file diff --git a/dhp-workflows/dhp-dedup-openaire/src/test/java/eu/dnetlib/dhp/oa/dedup/SparkDedupTest.java b/dhp-workflows/dhp-dedup-openaire/src/test/java/eu/dnetlib/dhp/oa/dedup/SparkDedupTest.java index 2f992bd78..fa03f93a6 100644 --- a/dhp-workflows/dhp-dedup-openaire/src/test/java/eu/dnetlib/dhp/oa/dedup/SparkDedupTest.java +++ b/dhp-workflows/dhp-dedup-openaire/src/test/java/eu/dnetlib/dhp/oa/dedup/SparkDedupTest.java @@ -5,13 +5,16 @@ 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.mockito.Mockito.lenient; import java.io.File; +import java.io.FileReader; import java.io.IOException; import java.io.Serializable; import java.net.URISyntaxException; import java.nio.file.Paths; +import java.util.List; import org.apache.commons.io.FileUtils; import org.apache.commons.io.IOUtils; @@ -45,552 +48,634 @@ import scala.Tuple2; @TestMethodOrder(MethodOrderer.OrderAnnotation.class) public class SparkDedupTest 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(SparkDedupTest.class.getResource("/eu/dnetlib/dhp/dedup/entities").toURI()) - .toFile() - .getAbsolutePath(); - testOutputBasePath = createTempDirectory(SparkDedupTest.class.getSimpleName() + "-") - .toAbsolutePath() - .toString(); - - testDedupGraphBasePath = createTempDirectory(SparkDedupTest.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", "200"); - spark = SparkSession - .builder() - .appName(SparkDedupTest.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( - IOUtils - .toString( - SparkDedupTest.class - .getResourceAsStream( - "/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"))); - - lenient() - .when(isLookUpService.getResourceProfileByQuery(Mockito.contains("publication"))) - .thenReturn( - IOUtils - .toString( - SparkDedupTest.class - .getResourceAsStream( - "/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"))); - - lenient() - .when(isLookUpService.getResourceProfileByQuery(Mockito.contains("dataset"))) - .thenReturn( - IOUtils - .toString( - SparkDedupTest.class - .getResourceAsStream( - "/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"))); - } - - @Test - @Order(1) - void createSimRelsTest() throws Exception { - - ArgumentApplicationParser parser = new ArgumentApplicationParser( - IOUtils - .toString( - SparkCreateSimRels.class - .getResourceAsStream( - "/eu/dnetlib/dhp/oa/dedup/createSimRels_parameters.json"))); - - parser - .parseArgument( - new String[] { - "-i", testGraphBasePath, - "-asi", testActionSetId, - "-la", "lookupurl", - "-w", testOutputBasePath, - "-np", "50" - }); - - new SparkCreateSimRels(parser, spark).run(isLookUpService); - - long orgs_simrel = spark - .read() - .load(DedupUtility.createSimRelPath(testOutputBasePath, testActionSetId, "organization")) - .count(); - - long pubs_simrel = spark - .read() - .load(DedupUtility.createSimRelPath(testOutputBasePath, testActionSetId, "publication")) - .count(); - - long sw_simrel = spark - .read() - .load(DedupUtility.createSimRelPath(testOutputBasePath, testActionSetId, "software")) - .count(); - - long ds_simrel = spark - .read() - .load(DedupUtility.createSimRelPath(testOutputBasePath, testActionSetId, "dataset")) - .count(); - - long orp_simrel = spark - .read() - .load(DedupUtility.createSimRelPath(testOutputBasePath, testActionSetId, "otherresearchproduct")) - .count(); - - assertEquals(3082, orgs_simrel); - assertEquals(7036, pubs_simrel); - assertEquals(336, sw_simrel); - assertEquals(442, ds_simrel); - assertEquals(6750, orp_simrel); - } - - @Test - @Order(2) - void cutMergeRelsTest() throws Exception { - - ArgumentApplicationParser parser = new ArgumentApplicationParser( - IOUtils - .toString( - SparkCreateMergeRels.class - .getResourceAsStream( - "/eu/dnetlib/dhp/oa/dedup/createCC_parameters.json"))); - - parser - .parseArgument( - new String[] { - "-i", - testGraphBasePath, - "-asi", - testActionSetId, - "-la", - "lookupurl", - "-w", - testOutputBasePath, - "-cc", - "3" - }); - - new SparkCreateMergeRels(parser, spark).run(isLookUpService); - - long orgs_mergerel = spark - .read() - .load(testOutputBasePath + "/" + testActionSetId + "/organization_mergerel") - .as(Encoders.bean(Relation.class)) - .filter((FilterFunction) r -> r.getRelClass().equalsIgnoreCase("merges")) - .groupBy("source") - .agg(count("target").alias("cnt")) - .select("source", "cnt") - .where("cnt > 3") - .count(); - - long pubs_mergerel = spark - .read() - .load(testOutputBasePath + "/" + testActionSetId + "/publication_mergerel") - .as(Encoders.bean(Relation.class)) - .filter((FilterFunction) r -> r.getRelClass().equalsIgnoreCase("merges")) - .groupBy("source") - .agg(count("target").alias("cnt")) - .select("source", "cnt") - .where("cnt > 3") - .count(); - long sw_mergerel = spark - .read() - .load(testOutputBasePath + "/" + testActionSetId + "/software_mergerel") - .as(Encoders.bean(Relation.class)) - .filter((FilterFunction) r -> r.getRelClass().equalsIgnoreCase("merges")) - .groupBy("source") - .agg(count("target").alias("cnt")) - .select("source", "cnt") - .where("cnt > 3") - .count(); - - long ds_mergerel = spark - .read() - .load(testOutputBasePath + "/" + testActionSetId + "/dataset_mergerel") - .as(Encoders.bean(Relation.class)) - .filter((FilterFunction) r -> r.getRelClass().equalsIgnoreCase("merges")) - .groupBy("source") - .agg(count("target").alias("cnt")) - .select("source", "cnt") - .where("cnt > 3") - .count(); - - long orp_mergerel = spark - .read() - .load(testOutputBasePath + "/" + testActionSetId + "/otherresearchproduct_mergerel") - .as(Encoders.bean(Relation.class)) - .filter((FilterFunction) r -> r.getRelClass().equalsIgnoreCase("merges")) - .groupBy("source") - .agg(count("target").alias("cnt")) - .select("source", "cnt") - .where("cnt > 3") - .count(); - - assertEquals(0, orgs_mergerel); - assertEquals(0, pubs_mergerel); - assertEquals(0, sw_mergerel); - assertEquals(0, ds_mergerel); - assertEquals(0, orp_mergerel); - - FileUtils.deleteDirectory(new File(testOutputBasePath + "/" + testActionSetId + "/organization_mergerel")); - FileUtils.deleteDirectory(new File(testOutputBasePath + "/" + testActionSetId + "/publication_mergerel")); - FileUtils.deleteDirectory(new File(testOutputBasePath + "/" + testActionSetId + "/software_mergerel")); - FileUtils.deleteDirectory(new File(testOutputBasePath + "/" + testActionSetId + "/dataset_mergerel")); - FileUtils - .deleteDirectory(new File(testOutputBasePath + "/" + testActionSetId + "/otherresearchproduct_mergerel")); - } - - @Test - @Order(3) - void createMergeRelsTest() throws Exception { - - ArgumentApplicationParser parser = new ArgumentApplicationParser( - IOUtils - .toString( - SparkCreateMergeRels.class - .getResourceAsStream( - "/eu/dnetlib/dhp/oa/dedup/createCC_parameters.json"))); - - parser - .parseArgument( - new String[] { - "-i", - testGraphBasePath, - "-asi", - testActionSetId, - "-la", - "lookupurl", - "-w", - testOutputBasePath - }); - - new SparkCreateMergeRels(parser, spark).run(isLookUpService); - - long orgs_mergerel = spark - .read() - .load(testOutputBasePath + "/" + testActionSetId + "/organization_mergerel") - .count(); - long pubs_mergerel = spark - .read() - .load(testOutputBasePath + "/" + testActionSetId + "/publication_mergerel") - .count(); - long sw_mergerel = spark - .read() - .load(testOutputBasePath + "/" + testActionSetId + "/software_mergerel") - .count(); - long ds_mergerel = spark - .read() - .load(testOutputBasePath + "/" + testActionSetId + "/dataset_mergerel") - .count(); - - long orp_mergerel = spark - .read() - .load(testOutputBasePath + "/" + testActionSetId + "/otherresearchproduct_mergerel") - .count(); - - assertEquals(1272, orgs_mergerel); - assertEquals(1438, pubs_mergerel); - assertEquals(286, sw_mergerel); - assertEquals(472, ds_mergerel); - assertEquals(718, orp_mergerel); - - } - - @Test - @Order(4) - void createDedupRecordTest() throws Exception { - - ArgumentApplicationParser parser = new ArgumentApplicationParser( - IOUtils - .toString( - SparkCreateDedupRecord.class - .getResourceAsStream( - "/eu/dnetlib/dhp/oa/dedup/createDedupRecord_parameters.json"))); - parser - .parseArgument( - new String[] { - "-i", - testGraphBasePath, - "-asi", - testActionSetId, - "-la", - "lookupurl", - "-w", - testOutputBasePath - }); - - new SparkCreateDedupRecord(parser, spark).run(isLookUpService); - - 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(); - long ds_deduprecord = jsc.textFile(testOutputBasePath + "/" + testActionSetId + "/dataset_deduprecord").count(); - long orp_deduprecord = jsc - .textFile( - testOutputBasePath + "/" + testActionSetId + "/otherresearchproduct_deduprecord") - .count(); - - assertEquals(85, orgs_deduprecord); - assertEquals(65, pubs_deduprecord); - assertEquals(51, sw_deduprecord); - assertEquals(97, ds_deduprecord); - assertEquals(89, orp_deduprecord); - } - - @Test - @Order(5) - void updateEntityTest() throws Exception { - - ArgumentApplicationParser parser = new ArgumentApplicationParser( - IOUtils - .toString( - SparkUpdateEntity.class - .getResourceAsStream( - "/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 organizations = jsc.textFile(testDedupGraphBasePath + "/organization").count(); - long publications = jsc.textFile(testDedupGraphBasePath + "/publication").count(); - long projects = jsc.textFile(testDedupGraphBasePath + "/project").count(); - long datasource = jsc.textFile(testDedupGraphBasePath + "/datasource").count(); - long softwares = jsc.textFile(testDedupGraphBasePath + "/software").count(); - long dataset = jsc.textFile(testDedupGraphBasePath + "/dataset").count(); - long otherresearchproduct = jsc.textFile(testDedupGraphBasePath + "/otherresearchproduct").count(); - - long mergedOrgs = spark - .read() - .load(testOutputBasePath + "/" + testActionSetId + "/organization_mergerel") - .as(Encoders.bean(Relation.class)) - .where("relClass=='merges'") - .javaRDD() - .map(Relation::getTarget) - .distinct() - .count(); - - long mergedPubs = spark - .read() - .load(testOutputBasePath + "/" + testActionSetId + "/publication_mergerel") - .as(Encoders.bean(Relation.class)) - .where("relClass=='merges'") - .javaRDD() - .map(Relation::getTarget) - .distinct() - .count(); - - long mergedSw = spark - .read() - .load(testOutputBasePath + "/" + testActionSetId + "/software_mergerel") - .as(Encoders.bean(Relation.class)) - .where("relClass=='merges'") - .javaRDD() - .map(Relation::getTarget) - .distinct() - .count(); - - long mergedDs = spark - .read() - .load(testOutputBasePath + "/" + testActionSetId + "/dataset_mergerel") - .as(Encoders.bean(Relation.class)) - .where("relClass=='merges'") - .javaRDD() - .map(Relation::getTarget) - .distinct() - .count(); - - long mergedOrp = spark - .read() - .load(testOutputBasePath + "/" + testActionSetId + "/otherresearchproduct_mergerel") - .as(Encoders.bean(Relation.class)) - .where("relClass=='merges'") - .javaRDD() - .map(Relation::getTarget) - .distinct() - .count(); - - assertEquals(896, publications); - assertEquals(838, organizations); - assertEquals(100, projects); - assertEquals(100, datasource); - assertEquals(200, softwares); - assertEquals(389, dataset); - assertEquals(517, otherresearchproduct); - - long deletedOrgs = jsc - .textFile(testDedupGraphBasePath + "/organization") - .filter(this::isDeletedByInference) - .count(); - - long deletedPubs = jsc - .textFile(testDedupGraphBasePath + "/publication") - .filter(this::isDeletedByInference) - .count(); - - long deletedSw = jsc - .textFile(testDedupGraphBasePath + "/software") - .filter(this::isDeletedByInference) - .count(); - - long deletedDs = jsc - .textFile(testDedupGraphBasePath + "/dataset") - .filter(this::isDeletedByInference) - .count(); - - long deletedOrp = jsc - .textFile(testDedupGraphBasePath + "/otherresearchproduct") - .filter(this::isDeletedByInference) - .count(); - - assertEquals(mergedOrgs, deletedOrgs); - assertEquals(mergedPubs, deletedPubs); - assertEquals(mergedSw, deletedSw); - assertEquals(mergedDs, deletedDs); - assertEquals(mergedOrp, deletedOrp); - } - - @Test - @Order(6) - void propagateRelationTest() throws Exception { - - ArgumentApplicationParser parser = new ArgumentApplicationParser( - IOUtils - .toString( - SparkPropagateRelation.class - .getResourceAsStream( - "/eu/dnetlib/dhp/oa/dedup/propagateRelation_parameters.json"))); - parser - .parseArgument( - new String[] { - "-i", testGraphBasePath, "-w", testOutputBasePath, "-o", testDedupGraphBasePath - }); - - new SparkPropagateRelation(parser, spark).run(isLookUpService); - - long relations = jsc.textFile(testDedupGraphBasePath + "/relation").count(); - - assertEquals(4860, relations); - - // check deletedbyinference - final Dataset mergeRels = spark - .read() - .load(DedupUtility.createMergeRelPath(testOutputBasePath, "*", "*")) - .as(Encoders.bean(Relation.class)); - final JavaPairRDD mergedIds = mergeRels - .where("relClass == 'merges'") - .select(mergeRels.col("target")) - .distinct() - .toJavaRDD() - .mapToPair( - (PairFunction) r -> new Tuple2(r.getString(0), "d")); - - JavaRDD toCheck = jsc - .textFile(testDedupGraphBasePath + "/relation") - .mapToPair(json -> new Tuple2<>(MapDocumentUtil.getJPathString("$.source", json), json)) - .join(mergedIds) - .map(t -> t._2()._1()) - .mapToPair(json -> new Tuple2<>(MapDocumentUtil.getJPathString("$.target", json), json)) - .join(mergedIds) - .map(t -> t._2()._1()); - - long deletedbyinference = toCheck.filter(this::isDeletedByInference).count(); - long updated = toCheck.count(); - - assertEquals(updated, deletedbyinference); - } - - @Test - @Order(7) - void testRelations() throws Exception { - testUniqueness("/eu/dnetlib/dhp/dedup/test/relation_1.json", 12, 10); - testUniqueness("/eu/dnetlib/dhp/dedup/test/relation_2.json", 10, 2); - } - - private void testUniqueness(String path, int expected_total, int expected_unique) { - Dataset rel = spark - .read() - .textFile(getClass().getResource(path).getPath()) - .map( - (MapFunction) s -> new ObjectMapper().readValue(s, Relation.class), - Encoders.bean(Relation.class)); - - assertEquals(expected_total, rel.count()); - assertEquals(expected_unique, rel.distinct().count()); - } - - @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"); - } + @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"; + private static String whitelistPath; + private static List whiteList; + + private static String WHITELIST_SEPARATOR = "####"; + + @BeforeAll + public static void cleanUp() throws IOException, URISyntaxException { + + testGraphBasePath = Paths + .get(SparkDedupTest.class.getResource("/eu/dnetlib/dhp/dedup/entities").toURI()) + .toFile() + .getAbsolutePath(); + testOutputBasePath = createTempDirectory(SparkDedupTest.class.getSimpleName() + "-") + .toAbsolutePath() + .toString(); + + testDedupGraphBasePath = createTempDirectory(SparkDedupTest.class.getSimpleName() + "-") + .toAbsolutePath() + .toString(); + + whitelistPath = Paths + .get(SparkDedupTest.class.getResource("/eu/dnetlib/dhp/dedup/whitelist.simrels.txt").toURI()) + .toFile() + .getAbsolutePath(); + whiteList = IOUtils.readLines(new FileReader(whitelistPath)); + + FileUtils.deleteDirectory(new File(testOutputBasePath)); + FileUtils.deleteDirectory(new File(testDedupGraphBasePath)); + + final SparkConf conf = new SparkConf(); + conf.set("spark.sql.shuffle.partitions", "200"); + spark = SparkSession + .builder() + .appName(SparkDedupTest.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( + IOUtils + .toString( + SparkDedupTest.class + .getResourceAsStream( + "/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"))); + + lenient() + .when(isLookUpService.getResourceProfileByQuery(Mockito.contains("publication"))) + .thenReturn( + IOUtils + .toString( + SparkDedupTest.class + .getResourceAsStream( + "/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"))); + + lenient() + .when(isLookUpService.getResourceProfileByQuery(Mockito.contains("dataset"))) + .thenReturn( + IOUtils + .toString( + SparkDedupTest.class + .getResourceAsStream( + "/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"))); + } + + @Test + @Order(1) + void createSimRelsTest() throws Exception { + + ArgumentApplicationParser parser = new ArgumentApplicationParser( + IOUtils + .toString( + SparkCreateSimRels.class + .getResourceAsStream( + "/eu/dnetlib/dhp/oa/dedup/createSimRels_parameters.json"))); + + parser + .parseArgument( + new String[]{ + "-i", testGraphBasePath, + "-asi", testActionSetId, + "-la", "lookupurl", + "-w", testOutputBasePath, + "-np", "50" + }); + + new SparkCreateSimRels(parser, spark).run(isLookUpService); + + long orgs_simrel = spark + .read() + .load(DedupUtility.createSimRelPath(testOutputBasePath, testActionSetId, "organization")) + .count(); + + long pubs_simrel = spark + .read() + .load(DedupUtility.createSimRelPath(testOutputBasePath, testActionSetId, "publication")) + .count(); + + long sw_simrel = spark + .read() + .load(DedupUtility.createSimRelPath(testOutputBasePath, testActionSetId, "software")) + .count(); + + long ds_simrel = spark + .read() + .load(DedupUtility.createSimRelPath(testOutputBasePath, testActionSetId, "dataset")) + .count(); + + long orp_simrel = spark + .read() + .load(DedupUtility.createSimRelPath(testOutputBasePath, testActionSetId, "otherresearchproduct")) + .count(); + + assertEquals(3082, orgs_simrel); + assertEquals(7036, pubs_simrel); + assertEquals(336, sw_simrel); + assertEquals(442, ds_simrel); + assertEquals(6750, orp_simrel); + } + + @Test + @Order(2) + void whitelistSimRelsTest() throws Exception { + + ArgumentApplicationParser parser = new ArgumentApplicationParser( + IOUtils + .toString( + SparkWhitelistSimRels.class + .getResourceAsStream( + "/eu/dnetlib/dhp/oa/dedup/whitelistSimRels_parameters.json"))); + + parser + .parseArgument( + new String[]{ + "-i", testGraphBasePath, + "-asi", testActionSetId, + "-la", "lookupurl", + "-w", testOutputBasePath, + "-np", "50", + "-wl", whitelistPath + }); + + new SparkWhitelistSimRels(parser, spark).run(isLookUpService); + + long orgs_simrel = spark + .read() + .load(DedupUtility.createSimRelPath(testOutputBasePath, testActionSetId, "organization")) + .count(); + + long pubs_simrel = spark + .read() + .load(DedupUtility.createSimRelPath(testOutputBasePath, testActionSetId, "publication")) + .count(); + + long ds_simrel = spark + .read() + .load(DedupUtility.createSimRelPath(testOutputBasePath, testActionSetId, "dataset")) + .count(); + + long orp_simrel = spark + .read() + .load(DedupUtility.createSimRelPath(testOutputBasePath, testActionSetId, "otherresearchproduct")) + .count(); + + //entities simrels supposed to be equal to the number of previous step (no rels in whitelist) + assertEquals(3082, orgs_simrel); + assertEquals(7036, pubs_simrel); + assertEquals(442, ds_simrel); + assertEquals(6750, orp_simrel); + + //entities simrels to be different from the number of previous step (new simrels in the whitelist) + Dataset sw_simrel = spark + .read() + .load(DedupUtility.createSimRelPath(testOutputBasePath, testActionSetId, "software")); + + //check if the first relation in the whitelist exists + assertTrue(sw_simrel + .as(Encoders.bean(Relation.class)) + .toJavaRDD() + .filter(rel -> + rel.getSource().equalsIgnoreCase(whiteList.get(0).split(WHITELIST_SEPARATOR)[0]) && rel.getTarget().equalsIgnoreCase(whiteList.get(0).split(WHITELIST_SEPARATOR)[1])).count() > 0); + //check if the second relation in the whitelist exists + assertTrue(sw_simrel + .as(Encoders.bean(Relation.class)) + .toJavaRDD() + .filter(rel -> + rel.getSource().equalsIgnoreCase(whiteList.get(1).split(WHITELIST_SEPARATOR)[0]) && rel.getTarget().equalsIgnoreCase(whiteList.get(1).split(WHITELIST_SEPARATOR)[1])).count() > 0); + + assertEquals(338, sw_simrel.count()); + + } + + @Test + @Order(3) + void cutMergeRelsTest() throws Exception { + + ArgumentApplicationParser parser = new ArgumentApplicationParser( + IOUtils + .toString( + SparkCreateMergeRels.class + .getResourceAsStream( + "/eu/dnetlib/dhp/oa/dedup/createCC_parameters.json"))); + + parser + .parseArgument( + new String[]{ + "-i", + testGraphBasePath, + "-asi", + testActionSetId, + "-la", + "lookupurl", + "-w", + testOutputBasePath, + "-cc", + "3" + }); + + new SparkCreateMergeRels(parser, spark).run(isLookUpService); + + long orgs_mergerel = spark + .read() + .load(testOutputBasePath + "/" + testActionSetId + "/organization_mergerel") + .as(Encoders.bean(Relation.class)) + .filter((FilterFunction) r -> r.getRelClass().equalsIgnoreCase("merges")) + .groupBy("source") + .agg(count("target").alias("cnt")) + .select("source", "cnt") + .where("cnt > 3") + .count(); + + long pubs_mergerel = spark + .read() + .load(testOutputBasePath + "/" + testActionSetId + "/publication_mergerel") + .as(Encoders.bean(Relation.class)) + .filter((FilterFunction) r -> r.getRelClass().equalsIgnoreCase("merges")) + .groupBy("source") + .agg(count("target").alias("cnt")) + .select("source", "cnt") + .where("cnt > 3") + .count(); + long sw_mergerel = spark + .read() + .load(testOutputBasePath + "/" + testActionSetId + "/software_mergerel") + .as(Encoders.bean(Relation.class)) + .filter((FilterFunction) r -> r.getRelClass().equalsIgnoreCase("merges")) + .groupBy("source") + .agg(count("target").alias("cnt")) + .select("source", "cnt") + .where("cnt > 3") + .count(); + + long ds_mergerel = spark + .read() + .load(testOutputBasePath + "/" + testActionSetId + "/dataset_mergerel") + .as(Encoders.bean(Relation.class)) + .filter((FilterFunction) r -> r.getRelClass().equalsIgnoreCase("merges")) + .groupBy("source") + .agg(count("target").alias("cnt")) + .select("source", "cnt") + .where("cnt > 3") + .count(); + + long orp_mergerel = spark + .read() + .load(testOutputBasePath + "/" + testActionSetId + "/otherresearchproduct_mergerel") + .as(Encoders.bean(Relation.class)) + .filter((FilterFunction) r -> r.getRelClass().equalsIgnoreCase("merges")) + .groupBy("source") + .agg(count("target").alias("cnt")) + .select("source", "cnt") + .where("cnt > 3") + .count(); + + assertEquals(0, orgs_mergerel); + assertEquals(0, pubs_mergerel); + assertEquals(0, sw_mergerel); + assertEquals(0, ds_mergerel); + assertEquals(0, orp_mergerel); + + FileUtils.deleteDirectory(new File(testOutputBasePath + "/" + testActionSetId + "/organization_mergerel")); + FileUtils.deleteDirectory(new File(testOutputBasePath + "/" + testActionSetId + "/publication_mergerel")); + FileUtils.deleteDirectory(new File(testOutputBasePath + "/" + testActionSetId + "/software_mergerel")); + FileUtils.deleteDirectory(new File(testOutputBasePath + "/" + testActionSetId + "/dataset_mergerel")); + FileUtils + .deleteDirectory(new File(testOutputBasePath + "/" + testActionSetId + "/otherresearchproduct_mergerel")); + } + + @Test + @Order(4) + void createMergeRelsTest() throws Exception { + + ArgumentApplicationParser parser = new ArgumentApplicationParser( + IOUtils + .toString( + SparkCreateMergeRels.class + .getResourceAsStream( + "/eu/dnetlib/dhp/oa/dedup/createCC_parameters.json"))); + + parser + .parseArgument( + new String[]{ + "-i", + testGraphBasePath, + "-asi", + testActionSetId, + "-la", + "lookupurl", + "-w", + testOutputBasePath + }); + + new SparkCreateMergeRels(parser, spark).run(isLookUpService); + + long orgs_mergerel = spark + .read() + .load(testOutputBasePath + "/" + testActionSetId + "/organization_mergerel") + .count(); + long pubs_mergerel = spark + .read() + .load(testOutputBasePath + "/" + testActionSetId + "/publication_mergerel") + .count(); + long sw_mergerel = spark + .read() + .load(testOutputBasePath + "/" + testActionSetId + "/software_mergerel") + .count(); + long ds_mergerel = spark + .read() + .load(testOutputBasePath + "/" + testActionSetId + "/dataset_mergerel") + .count(); + + long orp_mergerel = spark + .read() + .load(testOutputBasePath + "/" + testActionSetId + "/otherresearchproduct_mergerel") + .count(); + + assertEquals(1272, orgs_mergerel); + assertEquals(1438, pubs_mergerel); + assertEquals(286, sw_mergerel); + assertEquals(472, ds_mergerel); + assertEquals(718, orp_mergerel); + + } + + @Test + @Order(5) + void createDedupRecordTest() throws Exception { + + ArgumentApplicationParser parser = new ArgumentApplicationParser( + IOUtils + .toString( + SparkCreateDedupRecord.class + .getResourceAsStream( + "/eu/dnetlib/dhp/oa/dedup/createDedupRecord_parameters.json"))); + parser + .parseArgument( + new String[]{ + "-i", + testGraphBasePath, + "-asi", + testActionSetId, + "-la", + "lookupurl", + "-w", + testOutputBasePath + }); + + new SparkCreateDedupRecord(parser, spark).run(isLookUpService); + + 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(); + long ds_deduprecord = jsc.textFile(testOutputBasePath + "/" + testActionSetId + "/dataset_deduprecord").count(); + long orp_deduprecord = jsc + .textFile( + testOutputBasePath + "/" + testActionSetId + "/otherresearchproduct_deduprecord") + .count(); + + assertEquals(85, orgs_deduprecord); + assertEquals(65, pubs_deduprecord); + assertEquals(49, sw_deduprecord); + assertEquals(97, ds_deduprecord); + assertEquals(89, orp_deduprecord); + } + + @Test + @Order(6) + void updateEntityTest() throws Exception { + + ArgumentApplicationParser parser = new ArgumentApplicationParser( + IOUtils + .toString( + SparkUpdateEntity.class + .getResourceAsStream( + "/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 organizations = jsc.textFile(testDedupGraphBasePath + "/organization").count(); + long publications = jsc.textFile(testDedupGraphBasePath + "/publication").count(); + long projects = jsc.textFile(testDedupGraphBasePath + "/project").count(); + long datasource = jsc.textFile(testDedupGraphBasePath + "/datasource").count(); + long softwares = jsc.textFile(testDedupGraphBasePath + "/software").count(); + long dataset = jsc.textFile(testDedupGraphBasePath + "/dataset").count(); + long otherresearchproduct = jsc.textFile(testDedupGraphBasePath + "/otherresearchproduct").count(); + + long mergedOrgs = spark + .read() + .load(testOutputBasePath + "/" + testActionSetId + "/organization_mergerel") + .as(Encoders.bean(Relation.class)) + .where("relClass=='merges'") + .javaRDD() + .map(Relation::getTarget) + .distinct() + .count(); + + long mergedPubs = spark + .read() + .load(testOutputBasePath + "/" + testActionSetId + "/publication_mergerel") + .as(Encoders.bean(Relation.class)) + .where("relClass=='merges'") + .javaRDD() + .map(Relation::getTarget) + .distinct() + .count(); + + long mergedSw = spark + .read() + .load(testOutputBasePath + "/" + testActionSetId + "/software_mergerel") + .as(Encoders.bean(Relation.class)) + .where("relClass=='merges'") + .javaRDD() + .map(Relation::getTarget) + .distinct() + .count(); + + long mergedDs = spark + .read() + .load(testOutputBasePath + "/" + testActionSetId + "/dataset_mergerel") + .as(Encoders.bean(Relation.class)) + .where("relClass=='merges'") + .javaRDD() + .map(Relation::getTarget) + .distinct() + .count(); + + long mergedOrp = spark + .read() + .load(testOutputBasePath + "/" + testActionSetId + "/otherresearchproduct_mergerel") + .as(Encoders.bean(Relation.class)) + .where("relClass=='merges'") + .javaRDD() + .map(Relation::getTarget) + .distinct() + .count(); + + assertEquals(896, publications); + assertEquals(838, organizations); + assertEquals(100, projects); + assertEquals(100, datasource); + assertEquals(198, softwares); + assertEquals(389, dataset); + assertEquals(517, otherresearchproduct); + + long deletedOrgs = jsc + .textFile(testDedupGraphBasePath + "/organization") + .filter(this::isDeletedByInference) + .count(); + + long deletedPubs = jsc + .textFile(testDedupGraphBasePath + "/publication") + .filter(this::isDeletedByInference) + .count(); + + long deletedSw = jsc + .textFile(testDedupGraphBasePath + "/software") + .filter(this::isDeletedByInference) + .count(); + + long deletedDs = jsc + .textFile(testDedupGraphBasePath + "/dataset") + .filter(this::isDeletedByInference) + .count(); + + long deletedOrp = jsc + .textFile(testDedupGraphBasePath + "/otherresearchproduct") + .filter(this::isDeletedByInference) + .count(); + + assertEquals(mergedOrgs, deletedOrgs); + assertEquals(mergedPubs, deletedPubs); + assertEquals(mergedSw, deletedSw); + assertEquals(mergedDs, deletedDs); + assertEquals(mergedOrp, deletedOrp); + } + + @Test + @Order(7) + void propagateRelationTest() throws Exception { + + ArgumentApplicationParser parser = new ArgumentApplicationParser( + IOUtils + .toString( + SparkPropagateRelation.class + .getResourceAsStream( + "/eu/dnetlib/dhp/oa/dedup/propagateRelation_parameters.json"))); + parser + .parseArgument( + new String[]{ + "-i", testGraphBasePath, "-w", testOutputBasePath, "-o", testDedupGraphBasePath + }); + + new SparkPropagateRelation(parser, spark).run(isLookUpService); + + long relations = jsc.textFile(testDedupGraphBasePath + "/relation").count(); + + assertEquals(4860, relations); + + // check deletedbyinference + final Dataset mergeRels = spark + .read() + .load(DedupUtility.createMergeRelPath(testOutputBasePath, "*", "*")) + .as(Encoders.bean(Relation.class)); + final JavaPairRDD mergedIds = mergeRels + .where("relClass == 'merges'") + .select(mergeRels.col("target")) + .distinct() + .toJavaRDD() + .mapToPair( + (PairFunction) r -> new Tuple2(r.getString(0), "d")); + + JavaRDD toCheck = jsc + .textFile(testDedupGraphBasePath + "/relation") + .mapToPair(json -> new Tuple2<>(MapDocumentUtil.getJPathString("$.source", json), json)) + .join(mergedIds) + .map(t -> t._2()._1()) + .mapToPair(json -> new Tuple2<>(MapDocumentUtil.getJPathString("$.target", json), json)) + .join(mergedIds) + .map(t -> t._2()._1()); + + long deletedbyinference = toCheck.filter(this::isDeletedByInference).count(); + long updated = toCheck.count(); + + assertEquals(updated, deletedbyinference); + } + + @Test + @Order(8) + void testRelations() throws Exception { + testUniqueness("/eu/dnetlib/dhp/dedup/test/relation_1.json", 12, 10); + testUniqueness("/eu/dnetlib/dhp/dedup/test/relation_2.json", 10, 2); + } + + private void testUniqueness(String path, int expected_total, int expected_unique) { + Dataset rel = spark + .read() + .textFile(getClass().getResource(path).getPath()) + .map( + (MapFunction) s -> new ObjectMapper().readValue(s, Relation.class), + Encoders.bean(Relation.class)); + + assertEquals(expected_total, rel.count()); + assertEquals(expected_unique, rel.distinct().count()); + } + + @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"); + } } diff --git a/dhp-workflows/dhp-dedup-openaire/src/test/resources/eu/dnetlib/dhp/dedup/whitelist.simrels.txt b/dhp-workflows/dhp-dedup-openaire/src/test/resources/eu/dnetlib/dhp/dedup/whitelist.simrels.txt new file mode 100644 index 000000000..862ca466d --- /dev/null +++ b/dhp-workflows/dhp-dedup-openaire/src/test/resources/eu/dnetlib/dhp/dedup/whitelist.simrels.txt @@ -0,0 +1,2 @@ +50|r37b0ad08687::f645b9729d1e1025a72c57883f0f2cac####50|r37b0ad08687::4c55b436743b5c49fa32cd582fd9e1aa +50|datacite____::a90f49f9fde5393c00633bea6e4e374a####50|datacite____::5f55cdee77303ba8a2bf9996c32a330c \ No newline at end of file