Merge pull request 'SDG - FOS' (#181) from SDG into beta

Reviewed-on: D-Net/dnet-hadoop#181
This commit is contained in:
Miriam Baglioni 2022-01-07 11:31:19 +01:00
commit af8a33638d
10 changed files with 466 additions and 381 deletions

View File

@ -19,7 +19,7 @@ public class Constants {
public static final String DOI = "doi";
public static final char DEFAULT_DELIMITER = ',';
public static final String DEFAULT_DELIMITER = ",";
public static final String UPDATE_DATA_INFO_TYPE = "update";
public static final String UPDATE_SUBJECT_FOS_CLASS_ID = "subject:fos";
@ -55,7 +55,8 @@ public class Constants {
.map((MapFunction<String, R>) value -> OBJECT_MAPPER.readValue(value, clazz), Encoders.bean(clazz));
}
public static StructuredProperty getSubject(String sbj, String classid, String classname, String diqualifierclassid) {
public static StructuredProperty getSubject(String sbj, String classid, String classname,
String diqualifierclassid) {
if (sbj.equals(NULL))
return null;
StructuredProperty sp = new StructuredProperty();

View File

@ -0,0 +1,91 @@
package eu.dnetlib.dhp.actionmanager.createunresolvedentities;
import static eu.dnetlib.dhp.actionmanager.Constants.DEFAULT_DELIMITER;
import static eu.dnetlib.dhp.actionmanager.Constants.isSparkSessionManaged;
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
import java.io.Serializable;
import java.util.Optional;
import org.apache.commons.io.IOUtils;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
import org.apache.spark.SparkConf;
import org.apache.spark.api.java.function.MapFunction;
import org.apache.spark.sql.*;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import eu.dnetlib.dhp.actionmanager.createunresolvedentities.model.FOSDataModel;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
public class GetFOSSparkJob implements Serializable {
private static final Logger log = LoggerFactory.getLogger(GetFOSSparkJob.class);
public static void main(String[] args) throws Exception {
String jsonConfiguration = IOUtils
.toString(
GetFOSSparkJob.class
.getResourceAsStream(
"/eu/dnetlib/dhp/actionmanager/createunresolvedentities/get_input_parameters.json"));
final ArgumentApplicationParser parser = new ArgumentApplicationParser(jsonConfiguration);
parser.parseArgument(args);
Boolean isSparkSessionManaged = isSparkSessionManaged(parser);
log.info("isSparkSessionManaged: {}", isSparkSessionManaged);
// the path where the original fos csv file is stored
final String sourcePath = parser.get("sourcePath");
log.info("sourcePath {}", sourcePath);
// the path where to put the file as json
final String outputPath = parser.get("outputPath");
log.info("outputPath {}", outputPath);
final String delimiter = Optional
.ofNullable(parser.get("delimiter"))
.orElse(DEFAULT_DELIMITER);
SparkConf sconf = new SparkConf();
runWithSparkSession(
sconf,
isSparkSessionManaged,
spark -> {
getFOS(
spark,
sourcePath,
outputPath,
delimiter);
});
}
private static void getFOS(SparkSession spark, String sourcePath, String outputPath, String delimiter) {
Dataset<Row> fosData = spark
.read()
.format("csv")
.option("sep", delimiter)
.option("inferSchema", "true")
.option("header", "true")
.option("quotes", "\"")
.load(sourcePath);
fosData.map((MapFunction<Row, FOSDataModel>) r -> {
FOSDataModel fosDataModel = new FOSDataModel();
fosDataModel.setDoi(r.getString(0).toLowerCase());
fosDataModel.setLevel1(r.getString(1));
fosDataModel.setLevel2(r.getString(2));
fosDataModel.setLevel3(r.getString(3));
return fosDataModel;
}, Encoders.bean(FOSDataModel.class))
.write()
.mode(SaveMode.Overwrite)
.json(outputPath);
}
}

View File

@ -1,80 +0,0 @@
package eu.dnetlib.dhp.actionmanager.createunresolvedentities;
import java.io.IOException;
import java.io.InputStreamReader;
import java.io.Serializable;
import java.util.Objects;
import java.util.Optional;
import java.util.zip.GZIPInputStream;
import org.apache.commons.io.IOUtils;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import static eu.dnetlib.dhp.actionmanager.Constants.DEFAULT_DELIMITER;
public class GetInputData implements Serializable {
private static final Logger log = LoggerFactory.getLogger(GetInputData.class);
public static void main(final String[] args) throws Exception {
final ArgumentApplicationParser parser = new ArgumentApplicationParser(
IOUtils
.toString(
Objects
.requireNonNull(
GetInputData.class
.getResourceAsStream(
"/eu/dnetlib/dhp/actionmanager/createunresolvedentities/get_input_parameters.json"))));
parser.parseArgument(args);
// the path where the original fos csv file is stored
final String sourcePath = parser.get("sourcePath");
log.info("sourcePath {}", sourcePath);
// the path where to put the file as json
final String outputPath = parser.get("outputPath");
log.info("outputPath {}", outputPath);
final String hdfsNameNode = parser.get("hdfsNameNode");
log.info("hdfsNameNode {}", hdfsNameNode);
final String classForName = parser.get("classForName");
log.info("classForName {}", classForName);
final char delimiter = Optional
.ofNullable(parser.get("delimiter"))
.map(s -> s.charAt(0))
.orElse(DEFAULT_DELIMITER);
log.info("delimiter {}", delimiter);
Configuration conf = new Configuration();
conf.set("fs.defaultFS", hdfsNameNode);
FileSystem fileSystem = FileSystem.get(conf);
new GetInputData().doRewrite(sourcePath, outputPath, classForName, delimiter, fileSystem);
}
public void doRewrite(String inputPath, String outputFile, String classForName, char delimiter, FileSystem fs)
throws IOException, ClassNotFoundException {
// reads the csv and writes it as its json equivalent
try (InputStreamReader reader = new InputStreamReader(new GZIPInputStream(fs.open(new Path(inputPath))))) {
eu.dnetlib.dhp.common.collection.GetCSV.getCsv(fs, reader, outputFile, classForName, delimiter);
}
}
}

View File

@ -0,0 +1,89 @@
package eu.dnetlib.dhp.actionmanager.createunresolvedentities;
import static eu.dnetlib.dhp.actionmanager.Constants.DEFAULT_DELIMITER;
import static eu.dnetlib.dhp.actionmanager.Constants.isSparkSessionManaged;
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
import java.io.Serializable;
import java.util.Optional;
import org.apache.commons.io.IOUtils;
import org.apache.spark.SparkConf;
import org.apache.spark.api.java.function.FilterFunction;
import org.apache.spark.api.java.function.MapFunction;
import org.apache.spark.sql.*;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import eu.dnetlib.dhp.actionmanager.createunresolvedentities.model.FOSDataModel;
import eu.dnetlib.dhp.actionmanager.createunresolvedentities.model.SDGDataModel;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
public class GetSDGSparkJob implements Serializable {
private static final Logger log = LoggerFactory.getLogger(GetSDGSparkJob.class);
public static void main(String[] args) throws Exception {
String jsonConfiguration = IOUtils
.toString(
GetSDGSparkJob.class
.getResourceAsStream(
"/eu/dnetlib/dhp/actionmanager/createunresolvedentities/get_input_parameters.json"));
final ArgumentApplicationParser parser = new ArgumentApplicationParser(jsonConfiguration);
parser.parseArgument(args);
Boolean isSparkSessionManaged = isSparkSessionManaged(parser);
log.info("isSparkSessionManaged: {}", isSparkSessionManaged);
// the path where the original fos csv file is stored
final String sourcePath = parser.get("sourcePath");
log.info("sourcePath {}", sourcePath);
// the path where to put the file as json
final String outputPath = parser.get("outputPath");
log.info("outputPath {}", outputPath);
final String delimiter = Optional
.ofNullable(parser.get("delimiter"))
.orElse(DEFAULT_DELIMITER);
SparkConf sconf = new SparkConf();
runWithSparkSession(
sconf,
isSparkSessionManaged,
spark -> {
getSDG(
spark,
sourcePath,
outputPath,
delimiter);
});
}
private static void getSDG(SparkSession spark, String sourcePath, String outputPath, String delimiter) {
Dataset<Row> sdgData = spark
.read()
.format("csv")
.option("sep", delimiter)
.option("quotes", "\"")
.load(sourcePath);
sdgData.map((MapFunction<Row, SDGDataModel>) r -> {
SDGDataModel sdgDataModel = new SDGDataModel();
sdgDataModel.setDoi(r.getString(0).toLowerCase());
sdgDataModel.setSbj(r.getString(1));
return sdgDataModel;
}, Encoders.bean(SDGDataModel.class))
.filter((FilterFunction<SDGDataModel>) sdg -> sdg.getSbj() != null)
.write()
.mode(SaveMode.Overwrite)
.json(outputPath);
}
}

View File

@ -1,11 +1,13 @@
package eu.dnetlib.dhp.actionmanager.createunresolvedentities;
import eu.dnetlib.dhp.actionmanager.createunresolvedentities.model.SDGDataModel;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import eu.dnetlib.dhp.schema.oaf.Result;
import eu.dnetlib.dhp.schema.oaf.StructuredProperty;
import eu.dnetlib.dhp.utils.DHPUtils;
import static eu.dnetlib.dhp.actionmanager.Constants.*;
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
import java.io.Serializable;
import java.util.ArrayList;
import java.util.List;
import org.apache.commons.io.IOUtils;
import org.apache.spark.SparkConf;
import org.apache.spark.api.java.function.MapFunction;
@ -16,11 +18,12 @@ import org.apache.spark.sql.SaveMode;
import org.apache.spark.sql.SparkSession;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import java.io.Serializable;
import java.util.ArrayList;
import java.util.List;
import static eu.dnetlib.dhp.actionmanager.Constants.*;
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
import eu.dnetlib.dhp.actionmanager.createunresolvedentities.model.SDGDataModel;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import eu.dnetlib.dhp.schema.oaf.Result;
import eu.dnetlib.dhp.schema.oaf.StructuredProperty;
import eu.dnetlib.dhp.utils.DHPUtils;
public class PrepareSDGSparkJob implements Serializable {
private static final Logger log = LoggerFactory.getLogger(PrepareSDGSparkJob.class);
@ -62,15 +65,18 @@ public class PrepareSDGSparkJob implements Serializable {
private static void doPrepare(SparkSession spark, String sourcePath, String outputPath) {
Dataset<SDGDataModel> sdgDataset = readPath(spark, sourcePath, SDGDataModel.class);
sdgDataset.groupByKey((MapFunction<SDGDataModel,String>)r -> r.getDoi().toLowerCase(),Encoders.STRING())
sdgDataset
.groupByKey((MapFunction<SDGDataModel, String>) r -> r.getDoi().toLowerCase(), Encoders.STRING())
.mapGroups((MapGroupsFunction<String, SDGDataModel, Result>) (k, it) -> {
Result r = new Result();
r.setId(DHPUtils.generateUnresolvedIdentifier(k, DOI));
SDGDataModel first = it.next();
List<StructuredProperty> sbjs = new ArrayList<>();
sbjs.add(getSubject(first.getSbj(), SDG_CLASS_ID, SDG_CLASS_NAME, UPDATE_SUBJECT_SDG_CLASS_ID));
it.forEachRemaining(s -> sbjs.add(getSubject(s.getSbj(),SDG_CLASS_ID, SDG_CLASS_NAME, UPDATE_SUBJECT_SDG_CLASS_ID)));
it
.forEachRemaining(
s -> sbjs
.add(getSubject(s.getSbj(), SDG_CLASS_ID, SDG_CLASS_NAME, UPDATE_SUBJECT_SDG_CLASS_ID)));
r.setSubject(sbjs);
return r;
}, Encoders.bean(Result.class))
@ -80,7 +86,4 @@ public class PrepareSDGSparkJob implements Serializable {
.json(outputPath + "/sdg");
}
}

View File

@ -1,9 +1,10 @@
package eu.dnetlib.dhp.actionmanager.createunresolvedentities.model;
import com.opencsv.bean.CsvBindByPosition;
import java.io.Serializable;
import com.opencsv.bean.CsvBindByPosition;
public class SDGDataModel implements Serializable {
@CsvBindByPosition(position = 0)
@ -14,7 +15,6 @@ public class SDGDataModel implements Serializable{
// @CsvBindByName(column = "sdg")
private String sbj;
public SDGDataModel() {
}
@ -37,7 +37,6 @@ public class SDGDataModel implements Serializable{
this.doi = doi;
}
public String getSbj() {
return sbj;
}

View File

@ -19,15 +19,9 @@
"paramRequired": false
},
{
"paramName": "hnn",
"paramLongName": "hdfsNameNode",
"paramDescription": "the path used to store the HostedByMap",
"paramRequired": true
},
{
"paramName": "cfn",
"paramLongName": "classForName",
"paramDescription": "the path used to store the HostedByMap",
"paramRequired": true
"paramName": "d",
"paramLongName": "delimiter",
"paramDescription": "the delimiter if different from the default one (,)",
"paramRequired": false
}
]

View File

@ -107,17 +107,30 @@
</action>
<action name="getFOS">
<java>
<main-class>eu.dnetlib.dhp.actionmanager.createunresolvedentities.GetInputData</main-class>
<arg>--hdfsNameNode</arg><arg>${nameNode}</arg>
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Gets Data from FOS csv file</name>
<class>eu.dnetlib.dhp.actionmanager.createunresolvedentities.GetFOSSparkJob</class>
<jar>dhp-aggregation-${projectVersion}.jar</jar>
<spark-opts>
--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.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${fosPath}</arg>
<arg>--outputPath</arg><arg>${workingDir}/input/fos</arg>
<arg>--classForName</arg><arg>eu.dnetlib.dhp.actionmanager.createunresolvedentities.model.FOSDataModel</arg>
</java>
</spark>
<ok to="prepareFos"/>
<error to="Kill"/>
</action>
<action name="prepareFos">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
@ -144,17 +157,30 @@
</action>
<action name="getSDG">
<java>
<main-class>eu.dnetlib.dhp.actionmanager.createunresolvedentities.GetInputData</main-class>
<arg>--hdfsNameNode</arg><arg>${nameNode}</arg>
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Gets Data from SDG csv file</name>
<class>eu.dnetlib.dhp.actionmanager.createunresolvedentities.GetSDGSparkJob</class>
<jar>dhp-aggregation-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory=${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}${spark2EventLogDir}
--conf spark.sql.warehouse.dir=${sparkSqlWarehouseDir}
</spark-opts>
<arg>--sourcePath</arg><arg>${sdgPath}</arg>
<arg>--outputPath</arg><arg>${workingDir}/input/sdg</arg>
<arg>--classForName</arg><arg>eu.dnetlib.dhp.actionmanager.createunresolvedentities.model.SDGDataModel</arg>
</java>
</spark>
<ok to="prepareSDG"/>
<error to="Kill"/>
</action>
<action name="prepareSDG">
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>

View File

@ -10,7 +10,6 @@ import java.nio.file.Files;
import java.nio.file.Path;
import java.util.stream.Collectors;
import eu.dnetlib.dhp.actionmanager.createunresolvedentities.model.SDGDataModel;
import org.apache.commons.io.FileUtils;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
@ -26,6 +25,7 @@ import org.slf4j.LoggerFactory;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.dhp.actionmanager.createunresolvedentities.model.FOSDataModel;
import eu.dnetlib.dhp.actionmanager.createunresolvedentities.model.SDGDataModel;
import eu.dnetlib.dhp.schema.oaf.Result;
public class PrepareTest {
@ -148,37 +148,6 @@ public class PrepareTest {
}
@Test
void getFOSFileTest() throws IOException, ClassNotFoundException {
final String sourcePath = getClass()
.getResource("/eu/dnetlib/dhp/actionmanager/createunresolvedentities/fos/fos_sbs.csv")
.getPath();
final String outputPath = workingDir.toString() + "/fos.json";
new GetInputData()
.doRewrite(
sourcePath, outputPath, "eu.dnetlib.dhp.actionmanager.createunresolvedentities.model.FOSDataModel",
',', fs);
BufferedReader in = new BufferedReader(
new InputStreamReader(fs.open(new org.apache.hadoop.fs.Path(outputPath))));
String line;
int count = 0;
while ((line = in.readLine()) != null) {
FOSDataModel fos = new ObjectMapper().readValue(line, FOSDataModel.class);
System.out.println(new ObjectMapper().writeValueAsString(fos));
count += 1;
}
assertEquals(39, count);
}
@Test
void fosPrepareTest() throws Exception {
final String sourcePath = getClass()
@ -206,7 +175,6 @@ public class PrepareTest {
assertEquals(20, tmp.count());
assertEquals(1, tmp.filter(row -> row.getId().equals(doi1)).count());
assertTrue(
tmp
.filter(r -> r.getId().equals(doi1))
@ -249,35 +217,6 @@ public class PrepareTest {
}
@Test
void getSDGFileTest() throws IOException, ClassNotFoundException {
final String sourcePath = getClass()
.getResource("/eu/dnetlib/dhp/actionmanager/createunresolvedentities/sdg/sdg_sbs.csv")
.getPath();
final String outputPath = workingDir.toString() + "/sdg.json";
new GetInputData()
.doRewrite(
sourcePath, outputPath, "eu.dnetlib.dhp.actionmanager.createunresolvedentities.model.SDGDataModel",
',', fs);
BufferedReader in = new BufferedReader(
new InputStreamReader(fs.open(new org.apache.hadoop.fs.Path(outputPath))));
String line;
int count = 0;
while ((line = in.readLine()) != null) {
SDGDataModel sdg = new ObjectMapper().readValue(line, SDGDataModel.class);
System.out.println(new ObjectMapper().writeValueAsString(sdg));
count += 1;
}
assertEquals(37, count);
}
@Test
void sdgPrepareTest() throws Exception {
final String sourcePath = getClass()
@ -305,7 +244,6 @@ public class PrepareTest {
assertEquals(32, tmp.count());
assertEquals(1, tmp.filter(row -> row.getId().equals(doi1)).count());
assertTrue(
tmp
.filter(r -> r.getId().equals(doi1))
@ -321,33 +259,59 @@ public class PrepareTest {
.collect()
.contains("8. Economic growth"));
}
@Test
void test3() throws Exception {
final String sourcePath = "/Users/miriam.baglioni/Downloads/doi_fos_results_20_12_2021.csv.gz";
final String outputPath = workingDir.toString() + "/fos.json";
GetFOSSparkJob
.main(
new String[] {
"--isSparkSessionManaged", Boolean.FALSE.toString(),
"--sourcePath", sourcePath,
"-outputPath", outputPath
});
final JavaSparkContext sc = JavaSparkContext.fromSparkContext(spark.sparkContext());
JavaRDD<FOSDataModel> tmp = sc
.textFile(outputPath)
.map(item -> OBJECT_MAPPER.readValue(item, FOSDataModel.class));
tmp.foreach(t -> Assertions.assertTrue(t.getDoi() != null));
tmp.foreach(t -> Assertions.assertTrue(t.getLevel1() != null));
tmp.foreach(t -> Assertions.assertTrue(t.getLevel2() != null));
tmp.foreach(t -> Assertions.assertTrue(t.getLevel3() != null));
}
@Disabled
@Test
void test2() throws Exception {
void test4() throws Exception {
final String sourcePath = "/Users/miriam.baglioni/Downloads/doi_sdg_results_20_12_21.csv.gz";
final String outputPath = workingDir.toString() + "/sdg.json";
GetSDGSparkJob
.main(
new String[] {
"--isSparkSessionManaged", Boolean.FALSE.toString(),
"--sourcePath", sourcePath,
new GetInputData()
.doRewrite(
sourcePath, outputPath, "eu.dnetlib.dhp.actionmanager.createunresolvedentities.model.SDGDataModel",
',', fs);
"-outputPath", outputPath
BufferedReader in = new BufferedReader(
new InputStreamReader(fs.open(new org.apache.hadoop.fs.Path(outputPath))));
});
String line;
int count = 0;
while ((line = in.readLine()) != null) {
SDGDataModel sdg = new ObjectMapper().readValue(line, SDGDataModel.class);
final JavaSparkContext sc = JavaSparkContext.fromSparkContext(spark.sparkContext());
System.out.println(new ObjectMapper().writeValueAsString(sdg));
count += 1;
}
JavaRDD<SDGDataModel> tmp = sc
.textFile(outputPath)
.map(item -> OBJECT_MAPPER.readValue(item, SDGDataModel.class));
tmp.foreach(t -> Assertions.assertTrue(t.getDoi() != null));
tmp.foreach(t -> Assertions.assertTrue(t.getSbj() != null));
}
}

View File

@ -7,7 +7,6 @@ import java.nio.file.Path;
import java.util.List;
import java.util.stream.Collectors;
import eu.dnetlib.dhp.actionmanager.Constants;
import org.apache.commons.io.FileUtils;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
@ -25,6 +24,7 @@ import org.slf4j.LoggerFactory;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.dhp.actionmanager.Constants;
import eu.dnetlib.dhp.schema.common.ModelConstants;
import eu.dnetlib.dhp.schema.oaf.*;
@ -349,7 +349,6 @@ public class ProduceTest {
}
private JavaRDD<Result> getResultJavaRDDPlusSDG() throws Exception {
final String bipPath = getClass()
.getResource("/eu/dnetlib/dhp/actionmanager/createunresolvedentities/bip/bip.json")
@ -402,7 +401,6 @@ public class ProduceTest {
.map(item -> OBJECT_MAPPER.readValue(item, Result.class));
}
@Test
void produceTestSomeNumbersWithSDG() throws Exception {