antonis.lempesis
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
1
Fork 1
Code Issues Pull Requests Projects Releases Wiki Activity

refactoring

This commit is contained in:
Miriam Baglioni 2020-04-23 12:35:49 +02:00
parent 8c5dac5cc3
commit d8dc31d4af
3 changed files with 197 additions and 161 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

91
dhp-workflows/dhp-propagation/src/main/java/eu/dnetlib/dhp/orcidtoresultfromsemrel/PrepareResultOrcidAssociationStep1.java
View File

@ -1,38 +1,38 @@
package eu.dnetlib.dhp.orcidtoresultfromsemrel;
import static eu.dnetlib.dhp.PropagationConstant.*;
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkHiveSession;
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 java.util.Arrays;
import java.util.List;
import org.apache.commons.io.IOUtils;
import org.apache.hadoop.io.compress.GzipCodec;
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 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"));
String jsonConfiguration =
IOUtils.toString(
SparkOrcidToResultFromSemRelJob3.class.getResourceAsStream(
"/eu/dnetlib/dhp/orcidtoresultfromsemrel/input_prepareorcidtoresult_parameters.json"));
final ArgumentApplicationParser parser = new ArgumentApplicationParser(
jsonConfiguration);
final ArgumentApplicationParser parser = new ArgumentApplicationParser(jsonConfiguration);
parser.parseArgument(args);
@ -51,33 +51,44 @@ public class PrepareResultOrcidAssociationStep1 {
final List<String> allowedsemrel = Arrays.asList(parser.get("allowedsemrels").split(";"));
log.info("allowedSemRel: {}", new Gson().toJson(allowedsemrel));
final String resultType = resultClassName.substring(resultClassName.lastIndexOf(".") + 1).toLowerCase();
final String resultType =
resultClassName.substring(resultClassName.lastIndexOf(".") + 1).toLowerCase();
log.info("resultType: {}", resultType);
Class<? extends Result> resultClazz = (Class<? extends Result>) Class.forName(resultClassName);
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,
runWithSparkHiveSession(
conf,
isSparkSessionManaged,
spark -> {
if (isTest(parser)) {
removeOutputDir(spark, outputPath);
}
prepareInfo(spark, inputPath, outputPath, resultClazz, resultType, allowedsemrel);
prepareInfo(
spark, inputPath, outputPath, resultClazz, resultType, allowedsemrel);
});
}
private static <R extends Result> void prepareInfo(SparkSession spark, String inputPath,
String outputPath, Class<R> resultClazz,
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 -> OBJECT_MAPPER.readValue(item, Relation.class)).rdd(), Encoders.bean(Relation.class));
org.apache.spark.sql.Dataset<Relation> relation =
spark.createDataset(
sc.textFile(inputPath + "/relation")
.map(item -> OBJECT_MAPPER.readValue(item, Relation.class))
.rdd(),
Encoders.bean(Relation.class));
relation.createOrReplaceTempView("relation");
log.info("Reading Graph table from: {}", inputPath + "/" + resultType);
@ -86,25 +97,27 @@ public class PrepareResultOrcidAssociationStep1 {
result.createOrReplaceTempView("result");
getPossibleResultOrcidAssociation(spark, allowedsemrel, outputPath + "/" + resultType);
}
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";
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))
@ -118,6 +131,4 @@ public class PrepareResultOrcidAssociationStep1 {
// .text(outputPath)
// ;
}
}

49
dhp-workflows/dhp-propagation/src/main/java/eu/dnetlib/dhp/orcidtoresultfromsemrel/PrepareResultOrcidAssociationStep2.java
View File

@ -1,7 +1,12 @@
package eu.dnetlib.dhp.orcidtoresultfromsemrel;
import static eu.dnetlib.dhp.PropagationConstant.*;
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import java.util.HashSet;
import java.util.Set;
import org.apache.commons.io.IOUtils;
import org.apache.hadoop.io.compress.GzipCodec;
import org.apache.spark.SparkConf;
@ -9,22 +14,20 @@ import org.apache.spark.sql.*;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import scala.Tuple2;
import java.util.HashSet;
import java.util.Set;
import static eu.dnetlib.dhp.PropagationConstant.*;
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
public class PrepareResultOrcidAssociationStep2 {
private static final Logger log = LoggerFactory.getLogger(PrepareResultOrcidAssociationStep2.class);
private static final Logger log =
LoggerFactory.getLogger(PrepareResultOrcidAssociationStep2.class);
private static final ObjectMapper OBJECT_MAPPER = new ObjectMapper();
public static void main(String[] args) throws Exception {
String jsonConfiguration = IOUtils.toString(PrepareResultOrcidAssociationStep2.class
.getResourceAsStream("/eu/dnetlib/dhp/orcidtoresultfromsemrel/input_prepareorcidtoresult_parameters2.json"));
String jsonConfiguration =
IOUtils.toString(
PrepareResultOrcidAssociationStep2.class.getResourceAsStream(
"/eu/dnetlib/dhp/orcidtoresultfromsemrel/input_prepareorcidtoresult_parameters2.json"));
final ArgumentApplicationParser parser = new ArgumentApplicationParser(
jsonConfiguration);
final ArgumentApplicationParser parser = new ArgumentApplicationParser(jsonConfiguration);
parser.parseArgument(args);
@ -39,7 +42,9 @@ public class PrepareResultOrcidAssociationStep2 {
SparkConf conf = new SparkConf();
runWithSparkSession(conf, isSparkSessionManaged,
runWithSparkSession(
conf,
isSparkSessionManaged,
spark -> {
if (isTest(parser)) {
removeOutputDir(spark, outputPath);
@ -50,7 +55,8 @@ public class PrepareResultOrcidAssociationStep2 {
private static void mergeInfo(SparkSession spark, String inputPath, String outputPath) {
Dataset<ResultOrcidList> resultOrcidAssoc = readAssocResultOrcidList(spark, inputPath + "/publication")
Dataset<ResultOrcidList> resultOrcidAssoc =
readAssocResultOrcidList(spark, inputPath + "/publication")
.union(readAssocResultOrcidList(spark, inputPath + "/dataset"))
.union(readAssocResultOrcidList(spark, inputPath + "/otherresearchproduct"))
.union(readAssocResultOrcidList(spark, inputPath + "/software"));
@ -58,7 +64,8 @@ public class PrepareResultOrcidAssociationStep2 {
resultOrcidAssoc
.toJavaRDD()
.mapToPair(r -> new Tuple2<>(r.getResultId(), r))
.reduceByKey((a, b) -> {
.reduceByKey(
(a, b) -> {
if (a == null) {
return b;
}
@ -68,7 +75,9 @@ public class PrepareResultOrcidAssociationStep2 {
Set<String> orcid_set = new HashSet<>();
a.getAuthorList().stream().forEach(aa -> orcid_set.add(aa.getOrcid()));
b.getAuthorList().stream().forEach(aa -> {
b.getAuthorList().stream()
.forEach(
aa -> {
if (!orcid_set.contains(aa.getOrcid())) {
a.getAuthorList().add(aa);
orcid_set.add(aa.getOrcid());
@ -81,14 +90,12 @@ public class PrepareResultOrcidAssociationStep2 {
.saveAsTextFile(outputPath, GzipCodec.class);
}
private static Dataset<ResultOrcidList> readAssocResultOrcidList(SparkSession spark, String relationPath) {
return spark
.read()
private static Dataset<ResultOrcidList> readAssocResultOrcidList(
SparkSession spark, String relationPath) {
return spark.read()
.textFile(relationPath)
.map(value -> OBJECT_MAPPER.readValue(value, ResultOrcidList.class), Encoders.bean(ResultOrcidList.class));
.map(
value -> OBJECT_MAPPER.readValue(value, ResultOrcidList.class),
Encoders.bean(ResultOrcidList.class));
}
}

106
dhp-workflows/dhp-propagation/src/main/java/eu/dnetlib/dhp/orcidtoresultfromsemrel/SparkOrcidToResultFromSemRelJob3.java
View File

@ -1,10 +1,15 @@
package eu.dnetlib.dhp.orcidtoresultfromsemrel;
import static eu.dnetlib.dhp.PropagationConstant.*;
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkHiveSession;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import eu.dnetlib.dhp.schema.oaf.Author;
import eu.dnetlib.dhp.schema.oaf.Result;
import eu.dnetlib.dhp.schema.oaf.StructuredProperty;
import java.util.List;
import java.util.Optional;
import org.apache.commons.io.IOUtils;
import org.apache.commons.lang3.StringUtils;
import org.apache.spark.SparkConf;
@ -15,30 +20,25 @@ import org.apache.spark.sql.SparkSession;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import java.util.List;
import java.util.Optional;
import static eu.dnetlib.dhp.PropagationConstant.*;
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkHiveSession;
public class SparkOrcidToResultFromSemRelJob3 {
private static final Logger log = LoggerFactory.getLogger(SparkOrcidToResultFromSemRelJob3.class);
private static final Logger log =
LoggerFactory.getLogger(SparkOrcidToResultFromSemRelJob3.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_orcidtoresult_parameters.json"));
String jsonConfiguration =
IOUtils.toString(
SparkOrcidToResultFromSemRelJob3.class.getResourceAsStream(
"/eu/dnetlib/dhp/orcidtoresultfromsemrel/input_orcidtoresult_parameters.json"));
final ArgumentApplicationParser parser = new ArgumentApplicationParser(
jsonConfiguration);
final ArgumentApplicationParser parser = new ArgumentApplicationParser(jsonConfiguration);
parser.parseArgument(args);
Boolean isSparkSessionManaged = isSparkSessionManaged(parser);
log.info("isSparkSessionManaged: {}", isSparkSessionManaged);
final String inputPath = parser.get("sourcePath");
log.info("inputPath: {}", inputPath);
@ -51,18 +51,21 @@ public class SparkOrcidToResultFromSemRelJob3 {
final String resultClassName = parser.get("resultTableName");
log.info("resultTableName: {}", resultClassName);
final Boolean saveGraph = Optional
.ofNullable(parser.get("saveGraph"))
final Boolean saveGraph =
Optional.ofNullable(parser.get("saveGraph"))
.map(Boolean::valueOf)
.orElse(Boolean.TRUE);
log.info("saveGraph: {}", saveGraph);
Class<? extends Result> resultClazz = (Class<? extends Result>) Class.forName(resultClassName);
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,
runWithSparkHiveSession(
conf,
isSparkSessionManaged,
spark -> {
if (isTest(parser)) {
removeOutputDir(spark, outputPath);
@ -70,26 +73,34 @@ public class SparkOrcidToResultFromSemRelJob3 {
if (saveGraph)
execPropagation(spark, possibleUpdates, inputPath, outputPath, resultClazz);
});
}
private static <R extends Result> void execPropagation(SparkSession spark, String possibleUpdatesPath, String inputPath,
String outputPath, Class<R> resultClazz ) {
private static <R extends Result> void execPropagation(
SparkSession spark,
String possibleUpdatesPath,
String inputPath,
String outputPath,
Class<R> resultClazz) {
// read possible updates (resultId and list of possible orcid to add
Dataset<ResultOrcidList> possible_updates = readAssocResultOrcidList(spark, possibleUpdatesPath);
Dataset<ResultOrcidList> possible_updates =
readAssocResultOrcidList(spark, possibleUpdatesPath);
// read the result we have been considering
Dataset<R> result = readPathEntity(spark, inputPath, resultClazz);
// make join result left_outer with possible updates
result.joinWith(possible_updates, result.col("id").equalTo(possible_updates.col("resultId")),
result.joinWith(
possible_updates,
result.col("id").equalTo(possible_updates.col("resultId")),
"left_outer")
.map(value -> {
.map(
value -> {
R ret = value._1();
Optional<ResultOrcidList> rol = Optional.ofNullable(value._2());
if (rol.isPresent()) {
List<Author> toenrich_author = ret.getAuthor();
List<AutoritativeAuthor> autoritativeAuthors = rol.get().getAuthorList();
List<AutoritativeAuthor> autoritativeAuthors =
rol.get().getAuthorList();
for (Author author : toenrich_author) {
if (!containsAllowedPid(author)) {
enrichAuthor(author, autoritativeAuthors);
@ -98,21 +109,22 @@ public class SparkOrcidToResultFromSemRelJob3 {
}
return ret;
}, Encoders.bean(resultClazz))
},
Encoders.bean(resultClazz))
.toJSON()
.write()
.mode(SaveMode.Overwrite)
.option("compression", "gzip")
.text(outputPath);
}
private static Dataset<ResultOrcidList> readAssocResultOrcidList(SparkSession spark, String relationPath) {
return spark
.read()
private static Dataset<ResultOrcidList> readAssocResultOrcidList(
SparkSession spark, String relationPath) {
return spark.read()
.textFile(relationPath)
.map(value -> OBJECT_MAPPER.readValue(value, ResultOrcidList.class), Encoders.bean(ResultOrcidList.class));
.map(
value -> OBJECT_MAPPER.readValue(value, ResultOrcidList.class),
Encoders.bean(ResultOrcidList.class));
}
private static void enrichAuthor(Author a, List<AutoritativeAuthor> au) {
@ -121,26 +133,34 @@ public class SparkOrcidToResultFromSemRelJob3 {
return;
}
}
}
private static boolean enrichAuthor(AutoritativeAuthor autoritative_author, 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())) {
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())) {
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))) {
// 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;
}
}
@ -152,17 +172,16 @@ public class SparkOrcidToResultFromSemRelJob3 {
StructuredProperty p = new StructuredProperty();
p.setValue(autoritative_author.getOrcid());
p.setQualifier(getQualifier(PROPAGATION_AUTHOR_PID, PROPAGATION_AUTHOR_PID));
p.setDataInfo(getDataInfo(PROPAGATION_DATA_INFO_TYPE, PROPAGATION_ORCID_TO_RESULT_FROM_SEM_REL_CLASS_ID, PROPAGATION_ORCID_TO_RESULT_FROM_SEM_REL_CLASS_NAME));
p.setDataInfo(
getDataInfo(
PROPAGATION_DATA_INFO_TYPE,
PROPAGATION_ORCID_TO_RESULT_FROM_SEM_REL_CLASS_ID,
PROPAGATION_ORCID_TO_RESULT_FROM_SEM_REL_CLASS_NAME));
author.addPid(p);
}
return toaddpid;
}
private static boolean containsAllowedPid(Author a) {
for (StructuredProperty pid : a.getPid()) {
if (PROPAGATION_AUTHOR_PID.equals(pid.getQualifier().getClassid())) {
@ -171,5 +190,4 @@ public class SparkOrcidToResultFromSemRelJob3 {
}
return false;
}
}