forked from D-Net/dnet-hadoop
test classes for dump of results related to funders
This commit is contained in:
parent
b37b9352d7
commit
bde6d337dd
|
@ -1,4 +1,133 @@
|
|||
package eu.dnetlib.dhp.oa.graph.dump.funderresult;
|
||||
|
||||
import com.fasterxml.jackson.databind.ObjectMapper;
|
||||
import eu.dnetlib.dhp.oa.graph.dump.funderresults.SparkResultLinkedToProject;
|
||||
import eu.dnetlib.dhp.schema.oaf.Publication;
|
||||
import eu.dnetlib.dhp.schema.oaf.Result;
|
||||
import org.apache.commons.io.FileUtils;
|
||||
import org.apache.spark.SparkConf;
|
||||
import org.apache.spark.api.java.JavaRDD;
|
||||
import org.apache.spark.api.java.JavaSparkContext;
|
||||
import org.apache.spark.sql.Encoders;
|
||||
import org.apache.spark.sql.SparkSession;
|
||||
import org.junit.jupiter.api.AfterAll;
|
||||
import org.junit.jupiter.api.Assertions;
|
||||
import org.junit.jupiter.api.BeforeAll;
|
||||
import org.junit.jupiter.api.Test;
|
||||
import org.slf4j.Logger;
|
||||
import org.slf4j.LoggerFactory;
|
||||
|
||||
import java.io.IOException;
|
||||
import java.nio.file.Files;
|
||||
import java.nio.file.Path;
|
||||
import java.util.HashMap;
|
||||
|
||||
public class ResultLinkedToProjectTest {
|
||||
|
||||
private static final ObjectMapper OBJECT_MAPPER = new ObjectMapper();
|
||||
|
||||
private static SparkSession spark;
|
||||
|
||||
private static Path workingDir;
|
||||
|
||||
private static final Logger log = LoggerFactory
|
||||
.getLogger(eu.dnetlib.dhp.oa.graph.dump.funderresult.ResultLinkedToProjectTest.class);
|
||||
|
||||
private static HashMap<String, String> map = new HashMap<>();
|
||||
|
||||
@BeforeAll
|
||||
public static void beforeAll() throws IOException {
|
||||
workingDir = Files
|
||||
.createTempDirectory(eu.dnetlib.dhp.oa.graph.dump.funderresult.ResultLinkedToProjectTest.class.getSimpleName());
|
||||
log.info("using work dir {}", workingDir);
|
||||
|
||||
SparkConf conf = new SparkConf();
|
||||
conf.setAppName(eu.dnetlib.dhp.oa.graph.dump.funderresult.ResultLinkedToProjectTest.class.getSimpleName());
|
||||
|
||||
conf.setMaster("local[*]");
|
||||
conf.set("spark.driver.host", "localhost");
|
||||
conf.set("hive.metastore.local", "true");
|
||||
conf.set("spark.ui.enabled", "false");
|
||||
conf.set("spark.sql.warehouse.dir", workingDir.toString());
|
||||
conf.set("hive.metastore.warehouse.dir", workingDir.resolve("warehouse").toString());
|
||||
|
||||
spark = SparkSession
|
||||
.builder()
|
||||
.appName(eu.dnetlib.dhp.oa.graph.dump.funderresult.ResultLinkedToProjectTest.class.getSimpleName())
|
||||
.config(conf)
|
||||
.getOrCreate();
|
||||
}
|
||||
|
||||
@AfterAll
|
||||
public static void afterAll() throws IOException {
|
||||
FileUtils.deleteDirectory(workingDir.toFile());
|
||||
spark.stop();
|
||||
}
|
||||
|
||||
@Test
|
||||
public void testNoMatch() throws Exception {
|
||||
|
||||
final String sourcePath = getClass()
|
||||
.getResource("/eu/dnetlib/dhp/oa/graph/dump/funderresource/nomatch/papers.json")
|
||||
.getPath();
|
||||
|
||||
final String relationPath = getClass()
|
||||
.getResource("/eu/dnetlib/dhp/oa/graph/dump/funderresource/nomatch/relations.json")
|
||||
.getPath();
|
||||
|
||||
SparkResultLinkedToProject.main(new String[] {
|
||||
"-isSparkSessionManaged", Boolean.FALSE.toString(),
|
||||
"-outputPath", workingDir.toString() + "/preparedInfo",
|
||||
"-sourcePath", sourcePath,
|
||||
"-resultTableName", "eu.dnetlib.dhp.schema.oaf.Publication",
|
||||
"-relationPath", relationPath
|
||||
|
||||
});
|
||||
|
||||
final JavaSparkContext sc = JavaSparkContext.fromSparkContext(spark.sparkContext());
|
||||
|
||||
JavaRDD<Result> tmp = sc
|
||||
.textFile(workingDir.toString() + "/preparedInfo")
|
||||
.map(item -> OBJECT_MAPPER.readValue(item, Result.class));
|
||||
|
||||
org.apache.spark.sql.Dataset<Result> verificationDataset = spark
|
||||
.createDataset(tmp.rdd(), Encoders.bean(Result.class));
|
||||
|
||||
Assertions.assertEquals(0, verificationDataset.count());
|
||||
|
||||
}
|
||||
|
||||
@Test
|
||||
public void testMatchOne() throws Exception {
|
||||
|
||||
final String sourcePath = getClass()
|
||||
.getResource("/eu/dnetlib/dhp/oa/graph/dump/funderresource/match/papers.json")
|
||||
.getPath();
|
||||
|
||||
final String relationPath = getClass()
|
||||
.getResource("/eu/dnetlib/dhp/oa/graph/dump/funderresource/match/relations.json")
|
||||
.getPath();
|
||||
|
||||
SparkResultLinkedToProject.main(new String[] {
|
||||
"-isSparkSessionManaged", Boolean.FALSE.toString(),
|
||||
"-outputPath", workingDir.toString() + "/preparedInfo",
|
||||
"-sourcePath", sourcePath,
|
||||
"-resultTableName", "eu.dnetlib.dhp.schema.oaf.Publication",
|
||||
"-relationPath", relationPath
|
||||
|
||||
});
|
||||
|
||||
final JavaSparkContext sc = JavaSparkContext.fromSparkContext(spark.sparkContext());
|
||||
|
||||
JavaRDD<Publication> tmp = sc
|
||||
.textFile(workingDir.toString() + "/preparedInfo")
|
||||
.map(item -> OBJECT_MAPPER.readValue(item, Publication.class));
|
||||
|
||||
org.apache.spark.sql.Dataset<Publication> verificationDataset = spark
|
||||
.createDataset(tmp.rdd(), Encoders.bean(Publication.class));
|
||||
|
||||
Assertions.assertEquals(1, verificationDataset.count());
|
||||
|
||||
}
|
||||
|
||||
}
|
||||
|
|
|
@ -1,4 +1,142 @@
|
|||
package eu.dnetlib.dhp.oa.graph.dump.funderresult;
|
||||
|
||||
public class DumpResultPerFunderTest {
|
||||
import com.fasterxml.jackson.databind.ObjectMapper;
|
||||
import eu.dnetlib.dhp.oa.graph.dump.DumpJobTest;
|
||||
import eu.dnetlib.dhp.oa.graph.dump.SplitForCommunityTest;
|
||||
import eu.dnetlib.dhp.oa.graph.dump.community.CommunitySplit;
|
||||
import eu.dnetlib.dhp.oa.graph.dump.funderresults.SparkDumpFunderResults;
|
||||
import eu.dnetlib.dhp.schema.dump.oaf.community.CommunityResult;
|
||||
import org.apache.commons.io.FileUtils;
|
||||
import org.apache.spark.SparkConf;
|
||||
import org.apache.spark.api.java.JavaRDD;
|
||||
import org.apache.spark.api.java.JavaSparkContext;
|
||||
import org.apache.spark.sql.Encoders;
|
||||
import org.apache.spark.sql.SparkSession;
|
||||
import org.junit.jupiter.api.AfterAll;
|
||||
import org.junit.jupiter.api.Assertions;
|
||||
import org.junit.jupiter.api.BeforeAll;
|
||||
import org.junit.jupiter.api.Test;
|
||||
import org.slf4j.Logger;
|
||||
import org.slf4j.LoggerFactory;
|
||||
|
||||
import java.io.IOException;
|
||||
import java.nio.file.Files;
|
||||
import java.nio.file.Path;
|
||||
|
||||
public class SplitPerFunderTest {
|
||||
|
||||
private static final ObjectMapper OBJECT_MAPPER = new ObjectMapper();
|
||||
|
||||
private static SparkSession spark;
|
||||
|
||||
private static Path workingDir;
|
||||
|
||||
private static final Logger log = LoggerFactory.getLogger(SplitPerFunderTest.class);
|
||||
|
||||
@BeforeAll
|
||||
public static void beforeAll() throws IOException {
|
||||
workingDir = Files.createTempDirectory(SplitPerFunderTest.class.getSimpleName());
|
||||
log.info("using work dir {}", workingDir);
|
||||
|
||||
SparkConf conf = new SparkConf();
|
||||
conf.setAppName(SplitPerFunderTest.class.getSimpleName());
|
||||
|
||||
conf.setMaster("local[*]");
|
||||
conf.set("spark.driver.host", "localhost");
|
||||
conf.set("hive.metastore.local", "true");
|
||||
conf.set("spark.ui.enabled", "false");
|
||||
conf.set("spark.sql.warehouse.dir", workingDir.toString());
|
||||
conf.set("hive.metastore.warehouse.dir", workingDir.resolve("warehouse").toString());
|
||||
|
||||
spark = SparkSession
|
||||
.builder()
|
||||
.appName(SplitPerFunderTest.class.getSimpleName())
|
||||
.config(conf)
|
||||
.getOrCreate();
|
||||
}
|
||||
|
||||
@AfterAll
|
||||
public static void afterAll() throws IOException {
|
||||
FileUtils.deleteDirectory(workingDir.toFile());
|
||||
spark.stop();
|
||||
}
|
||||
|
||||
@Test
|
||||
public void test1() throws Exception {
|
||||
|
||||
|
||||
final String sourcePath = getClass()
|
||||
.getResource("/eu/dnetlib/dhp/oa/graph/dump/funderresource/extendeddump")
|
||||
.getPath();
|
||||
|
||||
|
||||
SparkDumpFunderResults.main(new String[] {
|
||||
"-isSparkSessionManaged", Boolean.FALSE.toString(),
|
||||
"-outputPath", workingDir.toString() + "/split",
|
||||
"-sourcePath", sourcePath,
|
||||
"-relationPath", sourcePath
|
||||
|
||||
});
|
||||
|
||||
|
||||
final JavaSparkContext sc = JavaSparkContext.fromSparkContext(spark.sparkContext());
|
||||
|
||||
// FP7 3
|
||||
JavaRDD<CommunityResult> tmp = sc
|
||||
.textFile(workingDir.toString() + "/split/EC_FP7")
|
||||
.map(item -> OBJECT_MAPPER.readValue(item, CommunityResult.class));
|
||||
|
||||
org.apache.spark.sql.Dataset<CommunityResult> verificationDataset = spark
|
||||
.createDataset(tmp.rdd(), Encoders.bean(CommunityResult.class));
|
||||
|
||||
Assertions.assertEquals(3, verificationDataset.count());
|
||||
|
||||
Assertions
|
||||
.assertEquals(
|
||||
1, verificationDataset.filter("id = '50|dedup_wf_001::0d16b1714ab3077df73893a8ea57d776'").count());
|
||||
|
||||
|
||||
// CIHR 2
|
||||
tmp = sc
|
||||
.textFile(workingDir.toString() + "/split/CIHR")
|
||||
.map(item -> OBJECT_MAPPER.readValue(item, CommunityResult.class));
|
||||
Assertions.assertEquals(2, tmp.count());
|
||||
|
||||
//NWO 1
|
||||
tmp = sc
|
||||
.textFile(workingDir.toString() + "/split/NWO")
|
||||
.map(item -> OBJECT_MAPPER.readValue(item, CommunityResult.class));
|
||||
Assertions.assertEquals(1, tmp.count());
|
||||
|
||||
//NIH 3
|
||||
tmp = sc
|
||||
.textFile(workingDir.toString() + "/split/NIH")
|
||||
.map(item -> OBJECT_MAPPER.readValue(item, CommunityResult.class));
|
||||
Assertions.assertEquals(2, tmp.count());
|
||||
|
||||
//NSF 1
|
||||
tmp = sc
|
||||
.textFile(workingDir.toString() + "/split/NSF")
|
||||
.map(item -> OBJECT_MAPPER.readValue(item, CommunityResult.class));
|
||||
Assertions.assertEquals(1, tmp.count());
|
||||
|
||||
//SNSF 1
|
||||
tmp = sc
|
||||
.textFile(workingDir.toString() + "/split/SNSF")
|
||||
.map(item -> OBJECT_MAPPER.readValue(item, CommunityResult.class));
|
||||
Assertions.assertEquals(1, tmp.count());
|
||||
|
||||
//NHMRC 1
|
||||
tmp = sc
|
||||
.textFile(workingDir.toString() + "/split/NHMRC")
|
||||
.map(item -> OBJECT_MAPPER.readValue(item, CommunityResult.class));
|
||||
Assertions.assertEquals(1, tmp.count());
|
||||
|
||||
//H2020 3
|
||||
tmp = sc
|
||||
.textFile(workingDir.toString() + "/split/EC_H2020")
|
||||
.map(item -> OBJECT_MAPPER.readValue(item, CommunityResult.class));
|
||||
Assertions.assertEquals(3, tmp.count());
|
||||
|
||||
}
|
||||
}
|
||||
|
|
Loading…
Reference in New Issue