refactoring

This commit is contained in:
Miriam Baglioni 2020-04-16 15:53:34 +02:00
parent 08227cfcbd
commit fd5d792e35
13 changed files with 841 additions and 80 deletions

View File

@ -121,9 +121,9 @@ public class SparkCountryPropagationJob2 {
.map(r -> new Tuple2<>(r.getId(), r), .map(r -> new Tuple2<>(r.getId(), r),
Encoders.tuple(Encoders.STRING(), Encoders.bean(resultClazz))); Encoders.tuple(Encoders.STRING(), Encoders.bean(resultClazz)));
Dataset<Tuple2<String, ResultCountrySet>> potential_update_pair = potentialUpdates.map(pu -> new Tuple2<>(pu.getResultId(), // Dataset<Tuple2<String, ResultCountrySet>> potential_update_pair = potentialUpdates.map(pu -> new Tuple2<>(pu.getResultId(),
pu), // pu),
Encoders.tuple(Encoders.STRING(), Encoders.bean(ResultCountrySet.class))); // Encoders.tuple(Encoders.STRING(), Encoders.bean(ResultCountrySet.class)));
Dataset<R> new_table = result_pair Dataset<R> new_table = result_pair
.joinWith(potentialUpdates, result_pair.col("_1").equalTo(potentialUpdates.col("resultId")), "left_outer") .joinWith(potentialUpdates, result_pair.col("_1").equalTo(potentialUpdates.col("resultId")), "left_outer")

View File

@ -0,0 +1,88 @@
package eu.dnetlib.dhp.orcidtoresultfromsemrel;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.google.gson.Gson;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import eu.dnetlib.dhp.schema.oaf.Relation;
import eu.dnetlib.dhp.schema.oaf.Result;
import org.apache.commons.io.IOUtils;
import org.apache.spark.SparkConf;
import org.apache.spark.api.java.JavaSparkContext;
import org.apache.spark.sql.Dataset;
import org.apache.spark.sql.Encoders;
import org.apache.spark.sql.SparkSession;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import java.util.Arrays;
import java.util.List;
import java.util.Optional;
import static eu.dnetlib.dhp.PropagationConstant.*;
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkHiveSession;
public class PrepareResultOrcidAssociation {
private static final Logger log = LoggerFactory.getLogger(PrepareResultOrcidAssociation.class);
private static final ObjectMapper OBJECT_MAPPER = new ObjectMapper();
public static void main(String[] args) throws Exception {
String jsonConfiguration = IOUtils.toString(SparkOrcidToResultFromSemRelJob3.class
.getResourceAsStream("/eu/dnetlib/dhp/orcidtoresultfromsemrel/input_prepareorcidtoresult_parameters.json"));
final ArgumentApplicationParser parser = new ArgumentApplicationParser(
jsonConfiguration);
parser.parseArgument(args);
Boolean isSparkSessionManaged = isSparkSessionManaged(parser);
log.info("isSparkSessionManaged: {}", isSparkSessionManaged);
String inputPath = parser.get("sourcePath");
log.info("inputPath: {}", inputPath);
final String outputPath = parser.get("outputPath");
log.info("outputPath: {}", outputPath);
final String resultClassName = parser.get("resultTableName");
log.info("resultTableName: {}", resultClassName);
final List<String> allowedsemrel = Arrays.asList(parser.get("allowedsemrel").split(";"));
log.info("allowedSemRel: {}", new Gson().toJson(allowedsemrel));
final String resultType = resultClassName.substring(resultClassName.lastIndexOf(".") + 1).toLowerCase();
log.info("resultType: {}", resultType);
Class<? extends Result> resultClazz = (Class<? extends Result>) Class.forName(resultClassName);
SparkConf conf = new SparkConf();
conf.set("hive.metastore.uris", parser.get("hive_metastore_uris"));
runWithSparkHiveSession(conf, isSparkSessionManaged,
spark -> {
if (isTest(parser)) {
removeOutputDir(spark, outputPath);
}
prepareInfo(spark, inputPath, outputPath, resultClazz, resultType);
});
}
private static <R extends Result> void prepareInfo(SparkSession spark, String inputPath,
String outputPath, Class<R> resultClazz,
String resultType) {
//read the relation table and the table related to the result it is using
final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext());
org.apache.spark.sql.Dataset<Relation> relation = spark.createDataset(sc.textFile(inputPath + "/relation")
.map(item -> new ObjectMapper().readValue(item, Relation.class)).rdd(), Encoders.bean(Relation.class));
relation.createOrReplaceTempView("relation");
log.info("Reading Graph table from: {}", inputPath + "/" + resultType);
Dataset<R> result = readPathEntity(spark, inputPath + "/" + resultType, resultClazz);
}
}

View File

@ -0,0 +1,119 @@
package eu.dnetlib.dhp.orcidtoresultfromsemrel;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.google.gson.Gson;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import eu.dnetlib.dhp.schema.oaf.Relation;
import eu.dnetlib.dhp.schema.oaf.Result;
import org.apache.commons.io.IOUtils;
import org.apache.spark.SparkConf;
import org.apache.spark.api.java.JavaSparkContext;
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.slf4j.Logger;
import org.slf4j.LoggerFactory;
import java.util.Arrays;
import java.util.List;
import static eu.dnetlib.dhp.PropagationConstant.*;
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkHiveSession;
public class PrepareResultOrcidAssociationStep1 {
private static final Logger log = LoggerFactory.getLogger(PrepareResultOrcidAssociationStep1.class);
private static final ObjectMapper OBJECT_MAPPER = new ObjectMapper();
public static void main(String[] args) throws Exception {
String jsonConfiguration = IOUtils.toString(SparkOrcidToResultFromSemRelJob3.class
.getResourceAsStream("/eu/dnetlib/dhp/orcidtoresultfromsemrel/input_prepareorcidtoresult_parameters.json"));
final ArgumentApplicationParser parser = new ArgumentApplicationParser(
jsonConfiguration);
parser.parseArgument(args);
Boolean isSparkSessionManaged = isSparkSessionManaged(parser);
log.info("isSparkSessionManaged: {}", isSparkSessionManaged);
String inputPath = parser.get("sourcePath");
log.info("inputPath: {}", inputPath);
final String outputPath = parser.get("outputPath");
log.info("outputPath: {}", outputPath);
final String resultClassName = parser.get("resultTableName");
log.info("resultTableName: {}", resultClassName);
final List<String> allowedsemrel = Arrays.asList(parser.get("allowedsemrel").split(";"));
log.info("allowedSemRel: {}", new Gson().toJson(allowedsemrel));
final String resultType = resultClassName.substring(resultClassName.lastIndexOf(".") + 1).toLowerCase();
log.info("resultType: {}", resultType);
Class<? extends Result> resultClazz = (Class<? extends Result>) Class.forName(resultClassName);
SparkConf conf = new SparkConf();
conf.set("hive.metastore.uris", parser.get("hive_metastore_uris"));
runWithSparkHiveSession(conf, isSparkSessionManaged,
spark -> {
if (isTest(parser)) {
removeOutputDir(spark, outputPath);
}
prepareInfo(spark, inputPath, outputPath, resultClazz, resultType, allowedsemrel);
});
}
private static <R extends Result> void prepareInfo(SparkSession spark, String inputPath,
String outputPath, Class<R> resultClazz,
String resultType,
List<String> allowedsemrel) {
//read the relation table and the table related to the result it is using
final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext());
org.apache.spark.sql.Dataset<Relation> relation = spark.createDataset(sc.textFile(inputPath + "/relation")
.map(item -> new ObjectMapper().readValue(item, Relation.class)).rdd(), Encoders.bean(Relation.class));
relation.createOrReplaceTempView("relation");
log.info("Reading Graph table from: {}", inputPath + "/" + resultType);
Dataset<R> result = readPathEntity(spark, inputPath + "/" + resultType, resultClazz);
result.createOrReplaceTempView("result");
getPossibleResultOrcidAssociation(spark, allowedsemrel, outputPath);
}
private static void getPossibleResultOrcidAssociation(SparkSession spark, List<String> allowedsemrel, String outputPath){
String query = " select target resultId, author authorList" +
" from (select id, collect_set(named_struct('name', name, 'surname', surname, 'fullname', fullname, 'orcid', orcid)) author " +
" from ( " +
" select id, MyT.fullname, MyT.name, MyT.surname, MyP.value orcid " +
" from result " +
" lateral view explode (author) a as MyT " +
" lateral view explode (MyT.pid) p as MyP " +
" where MyP.qualifier.classid = 'ORCID') tmp " +
" group by id) r_t " +
" join (" +
" select source, target " +
" from relation " +
" where datainfo.deletedbyinference = false " +
getConstraintList(" relclass = '" ,allowedsemrel) + ") rel_rel " +
" on source = id";
spark.sql(query)
.as(Encoders.bean(ResultOrcidList.class))
.toJSON()
.write()
.mode(SaveMode.Append)
.option("compression","gzip")
.text(outputPath)
;
}
}

View File

@ -0,0 +1,26 @@
package eu.dnetlib.dhp.orcidtoresultfromsemrel;
import java.io.Serializable;
import java.util.ArrayList;
import java.util.List;
public class ResultWithOrcid implements Serializable {
String id;
List<AutoritativeAuthor> authorList = new ArrayList<>();
public String getId() {
return id;
}
public void setId(String id) {
this.id = id;
}
public List<AutoritativeAuthor> getAuthorList() {
return authorList;
}
public void setAuthorList(List<AutoritativeAuthor> authorList) {
this.authorList = authorList;
}
}

View File

@ -1,4 +0,0 @@
package eu.dnetlib.dhp.orcidtoresultfromsemrel;
public class ResultWithOrcid {
}

View File

@ -0,0 +1,398 @@
package eu.dnetlib.dhp.orcidtoresultfromsemrel;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import eu.dnetlib.dhp.schema.oaf.*;
import org.apache.commons.io.IOUtils;
import org.apache.commons.lang3.StringUtils;
import org.apache.hadoop.fs.FileSystem;
import org.apache.spark.SparkConf;
import org.apache.spark.api.java.JavaPairRDD;
import org.apache.spark.api.java.JavaSparkContext;
import org.apache.spark.sql.Encoders;
import org.apache.spark.sql.Row;
import org.apache.spark.sql.SparkSession;
import scala.Tuple2;
import java.util.*;
import static eu.dnetlib.dhp.PropagationConstant.*;
public class SparkOrcidToResultFromSemRelJob2 {
public static void main(String[] args) throws Exception {
final ArgumentApplicationParser parser = new ArgumentApplicationParser(IOUtils.toString(SparkOrcidToResultFromSemRelJob2.class.getResourceAsStream("/eu/dnetlib/dhp/orcidtoresultfromremrel/input_orcidtoresult_parameters.json")));
parser.parseArgument(args);
SparkConf conf = new SparkConf();
conf.set("hive.metastore.uris", parser.get("hive_metastore_uris"));
final SparkSession spark = SparkSession
.builder()
.appName(SparkOrcidToResultFromSemRelJob2.class.getSimpleName())
.master(parser.get("master"))
.config(conf)
.enableHiveSupport()
.getOrCreate();
final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext());
final String inputPath = parser.get("sourcePath");
final String outputPath = "/tmp/provision/propagation/orcidtoresult";
final List<String> allowedsemrel = Arrays.asList(parser.get("allowedsemrels").split(";"));
boolean writeUpdate = TRUE.equals(parser.get("writeUpdate"));
boolean saveGraph = TRUE.equals(parser.get("saveGraph"));
createOutputDirs(outputPath, FileSystem.get(spark.sparkContext().hadoopConfiguration()));
org.apache.spark.sql.Dataset<Relation> relation = spark.createDataset(sc.textFile(inputPath + "/relation")
.map(item -> new ObjectMapper().readValue(item, Relation.class)).rdd(), Encoders.bean(Relation.class));
org.apache.spark.sql.Dataset<Dataset> dataset = spark.createDataset(sc.textFile(inputPath + "/dataset")
.map(item -> new ObjectMapper().readValue(item, eu.dnetlib.dhp.schema.oaf.Dataset.class)).rdd(),
Encoders.bean(eu.dnetlib.dhp.schema.oaf.Dataset.class));
org.apache.spark.sql.Dataset<OtherResearchProduct> other = spark.createDataset(sc.textFile(inputPath + "/otherresearchproduct")
.map(item -> new ObjectMapper().readValue(item, eu.dnetlib.dhp.schema.oaf.OtherResearchProduct.class)).rdd(),
Encoders.bean(eu.dnetlib.dhp.schema.oaf.OtherResearchProduct.class));
org.apache.spark.sql.Dataset<Software> software = spark.createDataset(sc.textFile(inputPath + "/software")
.map(item -> new ObjectMapper().readValue(item, eu.dnetlib.dhp.schema.oaf.Software.class)).rdd(),
Encoders.bean(eu.dnetlib.dhp.schema.oaf.Software.class));
org.apache.spark.sql.Dataset<Publication> publication = spark.createDataset(sc.textFile(inputPath + "/publication")
.map(item -> new ObjectMapper().readValue(item, eu.dnetlib.dhp.schema.oaf.Publication.class)).rdd(),
Encoders.bean(eu.dnetlib.dhp.schema.oaf.Publication.class));
relation.createOrReplaceTempView("relation");
String query = "Select source, target " +
"from relation " +
"where datainfo.deletedbyinference = false " + getConstraintList(" relclass = '" , allowedsemrel);
org.apache.spark.sql.Dataset<Row> result_result = spark.sql(query);
publication.createOrReplaceTempView("publication");
org.apache.spark.sql.Dataset<ResultWithOrcid> pubs_with_orcid = getResultWithOrcid("publication", spark)
.as(Encoders.bean(ResultWithOrcid.class));
dataset.createOrReplaceTempView("dataset");
org.apache.spark.sql.Dataset<ResultWithOrcid> dats_with_orcid = getResultWithOrcid("dataset", spark)
.as(Encoders.bean(ResultWithOrcid.class));
other.createOrReplaceTempView("orp");
org.apache.spark.sql.Dataset<ResultWithOrcid> orp_with_orcid = getResultWithOrcid("orp", spark)
.as(Encoders.bean(ResultWithOrcid.class));
dataset.createOrReplaceTempView("software");
org.apache.spark.sql.Dataset<ResultWithOrcid> software_with_orcid = getResultWithOrcid("software", spark)
.as(Encoders.bean(ResultWithOrcid.class));
//get the results having at least one author pid we are interested in
//target of the relation from at least one source with orcid.
//the set of authors contains all those that have orcid and are related to target
//from any source with allowed semantic relationship
JavaPairRDD<String, List<AutoritativeAuthor>> target_authorlist_from_pubs = getTargetAutoritativeAuthorList(pubs_with_orcid);
JavaPairRDD<String, List<AutoritativeAuthor>> target_authorlist_from_dats = getTargetAutoritativeAuthorList(dats_with_orcid);
JavaPairRDD<String, List<AutoritativeAuthor>> target_authorlist_from_orp = getTargetAutoritativeAuthorList(orp_with_orcid);
JavaPairRDD<String, List<AutoritativeAuthor>> target_authorlist_from_sw = getTargetAutoritativeAuthorList(software_with_orcid);
if(writeUpdate){
target_authorlist_from_dats.map(r -> new ObjectMapper().writeValueAsString(r))
.saveAsTextFile(outputPath + "/" + "update_dats");
target_authorlist_from_pubs.map(r -> new ObjectMapper().writeValueAsString(r))
.saveAsTextFile(outputPath + "/" + "update_pubs");
target_authorlist_from_orp.map(r -> new ObjectMapper().writeValueAsString(r))
.saveAsTextFile(outputPath + "/" + "update_orp");
target_authorlist_from_sw.map(r -> new ObjectMapper().writeValueAsString(r))
.saveAsTextFile(outputPath + "/" + "update_sw");
}
if(saveGraph){
sc.textFile(inputPath + "/publication")
.map(item -> new ObjectMapper().readValue(item, eu.dnetlib.dhp.schema.oaf.Publication.class))
.mapToPair(p -> new Tuple2<>(p.getId(),p))
.leftOuterJoin(target_authorlist_from_pubs)
.map(c -> {
Result r = c._2()._1();
if(!c._2()._2().isPresent()){
return r;
}
List<eu.dnetlib.dhp.schema.oaf.Author> toenrich_author = r.getAuthor();
List<AutoritativeAuthor> autoritativeAuthors = c._2()._2().get();
for(eu.dnetlib.dhp.schema.oaf.Author author: toenrich_author){
if (!containsAllowedPid(author)){
enrichAuthor(author, autoritativeAuthors);
}
}
return r;
});
}
}
private static void enrichAuthor(eu.dnetlib.dhp.schema.oaf.Author a, List<AutoritativeAuthor> au){
for (AutoritativeAuthor aa: au){
if(enrichAuthor(aa, a)){
return;
}
}
}
// private static JavaPairRDD<String, List<AutoritativeAuthor>> getTargetAutoritativeAuthorList(org.apache.spark.sql.Dataset<Row> result_result, org.apache.spark.sql.Dataset<ResultWithOrcid> pubs_with_orcid) {
// return pubs_with_orcid
// .toJavaRDD()
// .mapToPair(p -> new Tuple2<>(p.getId(), p.getAuthorList()))
// .join(result_result.toJavaRDD().mapToPair(rel -> new Tuple2<>(rel.getString(0), rel.getString(1))))
// .mapToPair(c -> new Tuple2<>(c._2._2(), c._2()._1()))
// .reduceByKey((a, b) -> {
// if(a == null){
// return b;
// }
// if(b==null){
// return a;
// }
//
// Set<String> authSet = new HashSet<>();
// a.stream().forEach(au -> authSet.add(au.getOrcid()));
//
// b.stream().forEach(au -> {
// if (!authSet.contains(au.getOrcid())) {
// a.add(au);
// }
// }
// );
// return a;
// });
// }
private static JavaPairRDD<String, List<AutoritativeAuthor>> getTargetAutoritativeAuthorList( org.apache.spark.sql.Dataset<ResultWithOrcid> pubs_with_orcid) {
return pubs_with_orcid
.toJavaRDD()
.mapToPair(p -> new Tuple2<>(p.getId(), p.getAuthorList()))
.reduceByKey((a, b) -> {
if(a == null){
return b;
}
if(b==null){
return a;
}
Set<String> authSet = new HashSet<>();
a.stream().forEach(au -> authSet.add(au.getOrcid()));
b.stream().forEach(au -> {
if (!authSet.contains(au.getOrcid())) {
a.add(au);
}
}
);
return a;
});
}
private static org.apache.spark.sql.Dataset<Row> getResultWithOrcid(String table, SparkSession spark){
String query = " select target, author " +
" from (select id, collect_set(named_struct('name', name, 'surname', surname, 'fullname', fullname, 'orcid', orcid)) author " +
" from ( " +
" select id, MyT.fullname, MyT.name, MyT.surname, MyP.value orcid " +
" from " + table +
" lateral view explode (author) a as MyT " +
" lateral view explode (MyT.pid) p as MyP " +
" where MyP.qualifier.classid = 'ORCID') tmp " +
" group by id) r_t " +
" join (" +
" select source, target " +
" from relation " +
" where datainfo.deletedbyinference = false and (relclass = 'isSupplementedBy' or relclass = 'isSupplementTo') rel_rel " +
" on source = id";
return spark.sql(query);
}
private static boolean enrichAuthor(AutoritativeAuthor autoritative_author, eu.dnetlib.dhp.schema.oaf.Author author) {
boolean toaddpid = false;
if (StringUtils.isNoneEmpty(autoritative_author.getSurname())) {
if (StringUtils.isNoneEmpty(author.getSurname())) {
if (autoritative_author.getSurname().trim().equalsIgnoreCase(author.getSurname().trim())) {
//have the same surname. Check the name
if (StringUtils.isNoneEmpty(autoritative_author.getName())) {
if (StringUtils.isNoneEmpty(author.getName())) {
if (autoritative_author.getName().trim().equalsIgnoreCase(author.getName().trim())) {
toaddpid = true;
}
//they could be differently written (i.e. only the initials of the name in one of the two
if (autoritative_author.getName().trim().substring(0, 0).equalsIgnoreCase(author.getName().trim().substring(0, 0))) {
toaddpid = true;
}
}
}
}
}
}
if (toaddpid){
StructuredProperty pid = new StructuredProperty();
String aa_pid = autoritative_author.getOrcid();
pid.setValue(aa_pid);
pid.setQualifier(getQualifier(PROPAGATION_AUTHOR_PID, PROPAGATION_AUTHOR_PID ));
pid.setDataInfo(getDataInfo(PROPAGATION_DATA_INFO_TYPE, PROPAGATION_ORCID_TO_RESULT_FROM_SEM_REL_CLASS_ID, PROPAGATION_ORCID_TO_RESULT_FROM_SEM_REL_CLASS_NAME));
if(author.getPid() == null){
author.setPid(Arrays.asList(pid));
}else{
author.getPid().add(pid);
}
}
return toaddpid;
}
// private static List<Author> enrichAuthors(List<Author> autoritative_authors, List<Author> to_enrich_authors, boolean filter){
//// List<Author> autoritative_authors = p._2()._2().get().getAuthors();
//// List<Author> to_enrich_authors = r.getAuthor();
//
// return to_enrich_authors
// .stream()
// .map(a -> {
// if (filter) {
// if (containsAllowedPid(a)) {
// return a;
// }
// }
//
// List<Author> lst = autoritative_authors.stream()
// .map(aa -> enrichAuthor(aa, a)).filter(au -> !(au == null)).collect(Collectors.toList());
// if (lst.size() == 0) {
// return a;
// }
// return lst.get(0);//Each author can be enriched at most once. It cannot be the same as many different people
//
// }).collect(Collectors.toList());
// }
//
// private static void writeResult(JavaPairRDD<String, Result> results, JavaPairRDD<String, TypedRow> toupdateresult,
// String outputPath, String type) {
//
// results.join(toupdateresult)
// .map(p -> {
// Result r = p._2()._1();
//
// List<Author> autoritative_authors = p._2()._2().getAuthors();
// List<eu.dnetlib.dhp.schema.oaf.Author> to_enrich_authors = r.getAuthor();
//
// r.setAuthor(enrichAutors(autoritative_authors, to_enrich_authors, false));
//// .stream()
//// .map(a -> {
//// if(filter) {
//// if (containsAllowedPid(a)) {
//// return a;
//// }
//// }
////
//// List<Author> lst = autoritative_authors.stream()
//// .map(aa -> enrichAuthor(aa, a)).filter(au -> !(au == null)).collect(Collectors.toList());
//// if(lst.size() == 0){
//// return a;
//// }
//// return lst.get(0);//Each author can be enriched at most once. It cannot be the same as many different people
////
//// }).collect(Collectors.toList()));
//
// return r;
// })
// .map(p -> new ObjectMapper().writeValueAsString(p))
// .saveAsTextFile(outputPath + "/" + type + "_update");
// }
// private static void updateResult(JavaPairRDD<String, Result> results, JavaPairRDD<String, TypedRow> toupdateresult,
// String outputPath, String type) {
// results.leftOuterJoin(toupdateresult)
// .map(p -> {
// Result r = p._2()._1();
// if (p._2()._2().isPresent()){
// List<AutoritativeAuthor> autoritative_authors = p._2()._2().get().getAuthors();
// List<eu.dnetlib.dhp.schema.oaf.Author> to_enrich_authors = r.getAuthor();
//
// r.setAuthor(enrichAutors(autoritative_authors, to_enrich_authors, true));
//// .stream()
//// .map(a -> {
//// if(filter) {
//// if (containsAllowedPid(a)) {
//// return a;
//// }
//// }
////
//// List<Author> lst = autoritative_authors.stream()
//// .map(aa -> enrichAuthor(aa, a)).filter(au -> !(au == null)).collect(Collectors.toList());
//// if(lst.size() == 0){
//// return a;
//// }
//// return lst.get(0);//Each author can be enriched at most once. It cannot be the same as many different people
////
//// }).collect(Collectors.toList()));
// }
// return r;
// })
// .map(p -> new ObjectMapper().writeValueAsString(p))
// .saveAsTextFile(outputPath+"/"+type);
// }
private static boolean containsAllowedPid(eu.dnetlib.dhp.schema.oaf.Author a) {
for (StructuredProperty pid : a.getPid()) {
if (PROPAGATION_AUTHOR_PID.equals(pid.getQualifier().getClassid())) {
return true;
}
}
return false;
}
}
/*private ResultProtos.Result.Metadata.Builder searchMatch(List<FieldTypeProtos.Author> author_list){
ResultProtos.Result.Metadata.Builder metadataBuilder = ResultProtos.Result.Metadata.newBuilder();
boolean updated = false;
for (FieldTypeProtos.Author a: author_list){
FieldTypeProtos.Author.Builder author = searchAuthor(a, autoritative_authors);
if(author != null){
updated = true;
metadataBuilder.addAuthor(author);
}else{
metadataBuilder.addAuthor(FieldTypeProtos.Author.newBuilder(a));
}
}
if(updated)
return metadataBuilder;
return null;
}
private FieldTypeProtos.Author.Builder searchAuthor(FieldTypeProtos.Author a, List<FieldTypeProtos.Author> author_list){
if(containsOrcid(a.getPidList()))
return null;
for(FieldTypeProtos.Author autoritative_author : author_list) {
if (equals(autoritative_author, a)) {
if(!containsOrcid(a.getPidList()))
return update(a, autoritative_author);
}
}
return null;
}
private boolean containsOrcid(List<FieldTypeProtos.KeyValue> pidList){
if(pidList == null)
return false;
return pidList
.stream()
.filter(kv -> kv.getKey().equals(PropagationConstants.AUTHOR_PID))
.collect(Collectors.toList()).size() > 0;
}
*/

View File

@ -1,18 +0,0 @@
<configuration>
<property>
<name>jobTracker</name>
<value>yarnRM</value>
</property>
<property>
<name>nameNode</name>
<value>hdfs://nameservice1</value>
</property>
<property>
<name>oozie.use.system.libpath</name>
<value>true</value>
</property>
<property>
<name>oozie.action.sharelib.for.spark</name>
<value>spark2</value>
</property>
</configuration>

View File

@ -1,55 +0,0 @@
<workflow-app name="orcid_to_result_from_semrel_propagation" xmlns="uri:oozie:workflow:0.5">
<parameters>
<property>
<name>sourcePath</name>
<description>the source path</description>
</property>
<property>
<name>allowedsemrels</name>
<description>the semantic relationships allowed for propagation</description>
</property>
<property>
<name>sparkDriverMemory</name>
<description>memory for driver process</description>
</property>
<property>
<name>sparkExecutorMemory</name>
<description>memory for individual executor</description>
</property>
<property>
<name>sparkExecutorCores</name>
<description>number of cores used by single executor</description>
</property>
</parameters>
<start to="OrcidToResultFromSemRelPropagation"/>
<kill name="Kill">
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
</kill>
<action name="OrcidToResultFromSemRelPropagation">
<spark xmlns="uri:oozie:spark-action:0.2">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>AffiliatioPropagation</name>
<class>eu.dnetlib.dhp.resulttoorganizationfrominstrepo.SparkResultToOrganizationFromIstRepoJob</class>
<jar>dhp-propagation-${projectVersion}.jar</jar>
<spark-opts>--executor-memory ${sparkExecutorMemory}
--executor-cores ${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.extraListeners="com.cloudera.spark.lineage.NavigatorAppListener"
--conf spark.sql.queryExecutionListeners="com.cloudera.spark.lineage.NavigatorQueryListener"
</spark-opts>
<arg>-mt</arg> <arg>yarn-cluster</arg>
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
<arg>--allowedsemrels</arg><arg>${allowedsemrels}</arg>
</spark>
<ok to="End"/>
<error to="Kill"/>
</action>
<end name="End"/>
</workflow-app>

View File

@ -16,5 +16,17 @@
"paramLongName":"allowedsemrels", "paramLongName":"allowedsemrels",
"paramDescription": "the allowed sematinc relations for propagation", "paramDescription": "the allowed sematinc relations for propagation",
"paramRequired": true "paramRequired": true
},
{
"paramName":"wu",
"paramLongName":"writeUpdate",
"paramDescription": "true if the update must be writte. No double check if information is already present",
"paramRequired": true
},
{
"paramName":"sg",
"paramLongName":"saveGraph",
"paramDescription": "true if the new version of the graph must be saved",
"paramRequired": true
} }
] ]

View File

@ -0,0 +1,32 @@
[
{
"paramName":"mt",
"paramLongName":"master",
"paramDescription": "should be local or yarn",
"paramRequired": true
},
{
"paramName":"s",
"paramLongName":"sourcePath",
"paramDescription": "the path of the sequencial file to read",
"paramRequired": true
},
{
"paramName":"as",
"paramLongName":"allowedsemrels",
"paramDescription": "the allowed sematinc relations for propagation",
"paramRequired": true
},
{
"paramName":"wu",
"paramLongName":"writeUpdate",
"paramDescription": "true if the update must be writte. No double check if information is already present",
"paramRequired": true
},
{
"paramName":"sg",
"paramLongName":"saveGraph",
"paramDescription": "true if the new version of the graph must be saved",
"paramRequired": true
}
]

View File

@ -0,0 +1,38 @@
[
{
"paramName":"s",
"paramLongName":"sourcePath",
"paramDescription": "the path of the sequencial file to read",
"paramRequired": true
},
{
"paramName":"as",
"paramLongName":"allowedsemrels",
"paramDescription": "the allowed sematinc relations for propagation",
"paramRequired": true
},
{
"paramName":"h",
"paramLongName":"hive_metastore_uris",
"paramDescription": "the hive metastore uris",
"paramRequired": true
},
{
"paramName": "out",
"paramLongName": "outputPath",
"paramDescription": "the path used to store temporary output files",
"paramRequired": true
},
{
"paramName": "ssm",
"paramLongName": "isSparkSessionManaged",
"paramDescription": "true if the spark session is managed, false otherwise",
"paramRequired": false
},
{
"paramName":"tn",
"paramLongName":"resultTableName",
"paramDescription": "the name of the result table we are currently working on",
"paramRequired": true
}
]

View File

@ -0,0 +1,54 @@
<configuration>
<property>
<name>jobTracker</name>
<value>yarnRM</value>
</property>
<property>
<name>nameNode</name>
<value>hdfs://nameservice1</value>
</property>
<property>
<name>oozie.use.system.libpath</name>
<value>true</value>
</property>
<property>
<name>oozie.action.sharelib.for.spark</name>
<value>spark2</value>
</property>
<property>
<name>spark2YarnHistoryServerAddress</name>
<value>http://iis-cdh5-test-gw.ocean.icm.edu.pl:18089</value>
</property>
<property>
<name>spark2EventLogDir</name>
<value>/user/spark/spark2ApplicationHistory</value>
</property>
<property>
<name>spark2ExtraListeners</name>
<value>com.cloudera.spark.lineage.NavigatorAppListener</value>
</property>
<property>
<name>spark2SqlQueryExecutionListeners</name>
<value>com.cloudera.spark.lineage.NavigatorQueryListener</value>
</property>
<property>
<name>sparkExecutorNumber</name>
<value>4</value>
</property>
<property>
<name>sparkDriverMemory</name>
<value>15G</value>
</property>
<property>
<name>sparkExecutorMemory</name>
<value>6G</value>
</property>
<property>
<name>sparkExecutorCores</name>
<value>1</value>
</property>
<property>
<name>spark2MaxExecutors</name>
<value>50</value>
</property>
</configuration>

View File

@ -0,0 +1,71 @@
<workflow-app name="orcid_to_result_from_semrel_propagation" xmlns="uri:oozie:workflow:0.5">
<parameters>
<property>
<name>sourcePath</name>
<description>the source path</description>
</property>
<property>
<name>allowedsemrels</name>
<description>the semantic relationships allowed for propagation</description>
</property>
<!-- <property>-->
<!-- <name>sparkDriverMemory</name>-->
<!-- <description>memory for driver process</description>-->
<!-- </property>-->
<!-- <property>-->
<!-- <name>sparkExecutorMemory</name>-->
<!-- <description>memory for individual executor</description>-->
<!-- </property>-->
<!-- <property>-->
<!-- <name>sparkExecutorCores</name>-->
<!-- <description>number of cores used by single executor</description>-->
<!-- </property>-->
<property>
<name>writeUpdate</name>
<description>writes the information found for the update. No double check done if the information is already present</description>
</property>
<property>
<name>saveGraph</name>
<description>writes new version of the graph after the propagation step</description>
</property>
</parameters>
<start to="OrcidToResultFromSemRelPropagation"/>
<kill name="Kill">
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
</kill>
<action name="OrcidToResultFromSemRelPropagation">
<spark xmlns="uri:oozie:spark-action:0.2">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<master>yarn-cluster</master>
<mode>cluster</mode>
<name>OrcidToResultFromSemRelPropagation</name>
<class>eu.dnetlib.dhp.orcidtoresultfromsemrel.SparkOrcidToResultFromSemRelJob</class>
<jar>dhp-propagation-${projectVersion}.jar</jar>
<spark-opts>
--num-executors=${sparkExecutorNumber}
--executor-memory=${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}${spark2EventLogDir}
--conf spark.dynamicAllocation.enabled=true
--conf spark.dynamicAllocation.maxExecutors=${spark2MaxExecutors}
</spark-opts>
<arg>-mt</arg> <arg>yarn-cluster</arg>
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
<arg>--allowedsemrels</arg><arg>${allowedsemrels}</arg>
<arg>--writeUpdate</arg><arg>${writeUpdate}</arg>
<arg>--saveGraph</arg><arg>${saveGraph}</arg>
</spark>
<ok to="End"/>
<error to="Kill"/>
</action>
<end name="End"/>
</workflow-app>