Merge pull request 'adding context information to projects and datasources' (#407) from taggingProjects into beta
Reviewed-on: #407
This commit is contained in:
commit
9e700a8b0d
|
@ -95,7 +95,7 @@ public class SparkAtomicActionScoreJob implements Serializable {
|
|||
|
||||
return projectScores.map((MapFunction<BipProjectModel, Project>) bipProjectScores -> {
|
||||
Project project = new Project();
|
||||
// project.setId(bipProjectScores.getProjectId());
|
||||
project.setId(bipProjectScores.getProjectId());
|
||||
project.setMeasures(bipProjectScores.toMeasures());
|
||||
return project;
|
||||
}, Encoders.bean(Project.class))
|
||||
|
|
|
@ -6,6 +6,7 @@ import java.io.IOException;
|
|||
import java.io.InputStreamReader;
|
||||
import java.net.HttpURLConnection;
|
||||
import java.net.URL;
|
||||
import java.util.List;
|
||||
|
||||
import org.jetbrains.annotations.NotNull;
|
||||
|
||||
|
|
|
@ -3,14 +3,10 @@ package eu.dnetlib.dhp.api;
|
|||
|
||||
import java.io.IOException;
|
||||
import java.io.Serializable;
|
||||
import java.util.ArrayList;
|
||||
import java.util.List;
|
||||
import java.util.Map;
|
||||
import java.util.Objects;
|
||||
import java.util.*;
|
||||
import java.util.stream.Collectors;
|
||||
|
||||
import javax.management.Query;
|
||||
|
||||
import org.jetbrains.annotations.NotNull;
|
||||
import org.slf4j.Logger;
|
||||
import org.slf4j.LoggerFactory;
|
||||
|
||||
|
@ -24,7 +20,10 @@ import eu.dnetlib.dhp.bulktag.community.CommunityConfiguration;
|
|||
import eu.dnetlib.dhp.bulktag.community.Provider;
|
||||
import eu.dnetlib.dhp.bulktag.criteria.VerbResolver;
|
||||
import eu.dnetlib.dhp.bulktag.criteria.VerbResolverFactory;
|
||||
import eu.dnetlib.dhp.resulttocommunityfromorganization.SparkResultToCommunityFromOrganizationJob;
|
||||
import eu.dnetlib.dhp.schema.common.ModelSupport;
|
||||
import eu.dnetlib.dhp.schema.oaf.Datasource;
|
||||
import eu.dnetlib.dhp.schema.oaf.Organization;
|
||||
import eu.dnetlib.dhp.schema.oaf.Project;
|
||||
|
||||
/**
|
||||
* @author miriam.baglioni
|
||||
|
@ -58,7 +57,7 @@ public class Utils implements Serializable {
|
|||
if (d.getEnabled() == null || Boolean.FALSE.equals(d.getEnabled()))
|
||||
return null;
|
||||
Provider p = new Provider();
|
||||
p.setOpenaireId("10|" + d.getOpenaireId());
|
||||
p.setOpenaireId(ModelSupport.getIdPrefix(Datasource.class) + "|" + d.getOpenaireId());
|
||||
p.setSelectionConstraints(d.getSelectioncriteria());
|
||||
if (p.getSelectionConstraints() != null)
|
||||
p.getSelectionConstraints().setSelection(resolver);
|
||||
|
@ -113,6 +112,7 @@ public class Utils implements Serializable {
|
|||
*/
|
||||
public static CommunityEntityMap getCommunityOrganization(String baseURL) throws IOException {
|
||||
CommunityEntityMap organizationMap = new CommunityEntityMap();
|
||||
String entityPrefix = ModelSupport.getIdPrefix(Organization.class);
|
||||
getValidCommunities(baseURL)
|
||||
.forEach(community -> {
|
||||
String id = community.getId();
|
||||
|
@ -124,9 +124,9 @@ public class Utils implements Serializable {
|
|||
if (!organizationMap
|
||||
.keySet()
|
||||
.contains(
|
||||
"20|" + o))
|
||||
organizationMap.put("20|" + o, new ArrayList<>());
|
||||
organizationMap.get("20|" + o).add(community.getId());
|
||||
entityPrefix + "|" + o))
|
||||
organizationMap.put(entityPrefix + "|" + o, new ArrayList<>());
|
||||
organizationMap.get(entityPrefix + "|" + o).add(community.getId());
|
||||
});
|
||||
} catch (IOException e) {
|
||||
throw new RuntimeException(e);
|
||||
|
@ -138,7 +138,7 @@ public class Utils implements Serializable {
|
|||
|
||||
public static CommunityEntityMap getCommunityProjects(String baseURL) throws IOException {
|
||||
CommunityEntityMap projectMap = new CommunityEntityMap();
|
||||
|
||||
String entityPrefix = ModelSupport.getIdPrefix(Project.class);
|
||||
getValidCommunities(baseURL)
|
||||
.forEach(community -> {
|
||||
int page = -1;
|
||||
|
@ -155,9 +155,9 @@ public class Utils implements Serializable {
|
|||
ContentModel.class);
|
||||
if (cm.getContent().size() > 0) {
|
||||
cm.getContent().forEach(p -> {
|
||||
if (!projectMap.keySet().contains("40|" + p.getOpenaireId()))
|
||||
projectMap.put("40|" + p.getOpenaireId(), new ArrayList<>());
|
||||
projectMap.get("40|" + p.getOpenaireId()).add(community.getId());
|
||||
if (!projectMap.keySet().contains(entityPrefix + "|" + p.getOpenaireId()))
|
||||
projectMap.put(entityPrefix + "|" + p.getOpenaireId(), new ArrayList<>());
|
||||
projectMap.get(entityPrefix + "|" + p.getOpenaireId()).add(community.getId());
|
||||
});
|
||||
}
|
||||
} catch (IOException e) {
|
||||
|
@ -174,4 +174,41 @@ public class Utils implements Serializable {
|
|||
.map(community -> community.getId())
|
||||
.collect(Collectors.toList());
|
||||
}
|
||||
|
||||
public static List<EntityCommunities> getDatasourceCommunities(String baseURL) throws IOException {
|
||||
List<CommunityModel> validCommunities = getValidCommunities(baseURL);
|
||||
HashMap<String, Set<String>> map = new HashMap<>();
|
||||
String entityPrefix = ModelSupport.getIdPrefix(Datasource.class) + "|";
|
||||
|
||||
validCommunities.forEach(c -> {
|
||||
try {
|
||||
new ObjectMapper()
|
||||
.readValue(QueryCommunityAPI.communityDatasource(c.getId(), baseURL), DatasourceList.class)
|
||||
.forEach(d -> {
|
||||
if (!map.keySet().contains(d.getOpenaireId()))
|
||||
map.put(d.getOpenaireId(), new HashSet<>());
|
||||
|
||||
map.get(d.getOpenaireId()).add(c.getId());
|
||||
});
|
||||
} catch (IOException e) {
|
||||
throw new RuntimeException(e);
|
||||
}
|
||||
});
|
||||
|
||||
List<EntityCommunities> temp = map
|
||||
.keySet()
|
||||
.stream()
|
||||
.map(k -> EntityCommunities.newInstance(entityPrefix + k, getCollect(k, map)))
|
||||
.collect(Collectors.toList());
|
||||
|
||||
return temp;
|
||||
|
||||
}
|
||||
|
||||
@NotNull
|
||||
private static List<String> getCollect(String k, HashMap<String, Set<String>> map) {
|
||||
List<String> temp = map.get(k).stream().collect(Collectors.toList());
|
||||
return temp;
|
||||
}
|
||||
|
||||
}
|
||||
|
|
|
@ -0,0 +1,40 @@
|
|||
|
||||
package eu.dnetlib.dhp.api.model;
|
||||
|
||||
import java.io.Serializable;
|
||||
import java.util.List;
|
||||
|
||||
import eu.dnetlib.dhp.schema.common.ModelSupport;
|
||||
import eu.dnetlib.dhp.schema.oaf.Datasource;
|
||||
|
||||
/**
|
||||
* @author miriam.baglioni
|
||||
* @Date 13/02/24
|
||||
*/
|
||||
public class EntityCommunities implements Serializable {
|
||||
private String entityId;
|
||||
private List<String> communitiesId;
|
||||
|
||||
public String getEntityId() {
|
||||
return entityId;
|
||||
}
|
||||
|
||||
public void setEntityId(String entityId) {
|
||||
this.entityId = entityId;
|
||||
}
|
||||
|
||||
public List<String> getCommunitiesId() {
|
||||
return communitiesId;
|
||||
}
|
||||
|
||||
public void setCommunitiesId(List<String> communitiesId) {
|
||||
this.communitiesId = communitiesId;
|
||||
}
|
||||
|
||||
public static EntityCommunities newInstance(String dsid, List<String> csid) {
|
||||
EntityCommunities dsc = new EntityCommunities();
|
||||
dsc.entityId = dsid;
|
||||
dsc.communitiesId = csid;
|
||||
return dsc;
|
||||
}
|
||||
}
|
|
@ -4,9 +4,23 @@ package eu.dnetlib.dhp.bulktag;
|
|||
import static eu.dnetlib.dhp.PropagationConstant.removeOutputDir;
|
||||
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
|
||||
|
||||
import java.io.BufferedOutputStream;
|
||||
import java.io.IOException;
|
||||
import java.nio.charset.StandardCharsets;
|
||||
import java.util.*;
|
||||
import java.util.stream.Collectors;
|
||||
import java.util.zip.GZIPOutputStream;
|
||||
|
||||
import org.apache.avro.TestAnnotation;
|
||||
import org.apache.commons.compress.archivers.tar.TarArchiveEntry;
|
||||
import org.apache.commons.compress.archivers.tar.TarArchiveInputStream;
|
||||
import org.apache.commons.compress.compressors.gzip.GzipCompressorInputStream;
|
||||
import org.apache.commons.io.IOUtils;
|
||||
import org.apache.hadoop.conf.Configuration;
|
||||
import org.apache.hadoop.fs.FSDataInputStream;
|
||||
import org.apache.hadoop.fs.FSDataOutputStream;
|
||||
import org.apache.hadoop.fs.FileSystem;
|
||||
import org.apache.hadoop.fs.Path;
|
||||
import org.apache.spark.SparkConf;
|
||||
import org.apache.spark.api.java.function.FilterFunction;
|
||||
import org.apache.spark.api.java.function.MapFunction;
|
||||
|
@ -17,17 +31,24 @@ import org.apache.spark.sql.SparkSession;
|
|||
import org.slf4j.Logger;
|
||||
import org.slf4j.LoggerFactory;
|
||||
|
||||
import com.fasterxml.jackson.core.JsonProcessingException;
|
||||
import com.fasterxml.jackson.databind.ObjectMapper;
|
||||
import com.google.gson.Gson;
|
||||
import com.sun.media.sound.ModelInstrumentComparator;
|
||||
|
||||
import eu.dnetlib.dhp.api.Utils;
|
||||
import eu.dnetlib.dhp.api.model.CommunityEntityMap;
|
||||
import eu.dnetlib.dhp.api.model.EntityCommunities;
|
||||
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
|
||||
import eu.dnetlib.dhp.bulktag.community.*;
|
||||
import eu.dnetlib.dhp.schema.common.EntityType;
|
||||
import eu.dnetlib.dhp.schema.common.ModelConstants;
|
||||
import eu.dnetlib.dhp.schema.common.ModelSupport;
|
||||
import eu.dnetlib.dhp.schema.oaf.Context;
|
||||
import eu.dnetlib.dhp.schema.oaf.Datasource;
|
||||
import eu.dnetlib.dhp.schema.oaf.Project;
|
||||
import eu.dnetlib.dhp.schema.oaf.Result;
|
||||
import eu.dnetlib.dhp.schema.oaf.utils.OafMapperUtils;
|
||||
import scala.Tuple2;
|
||||
|
||||
public class SparkBulkTagJob {
|
||||
|
||||
|
@ -47,6 +68,7 @@ public class SparkBulkTagJob {
|
|||
.getResourceAsStream(
|
||||
"/eu/dnetlib/dhp/wf/subworkflows/bulktag/input_bulkTag_parameters.json"));
|
||||
|
||||
log.info(args.toString());
|
||||
final ArgumentApplicationParser parser = new ArgumentApplicationParser(jsonConfiguration);
|
||||
parser.parseArgument(args);
|
||||
|
||||
|
@ -65,8 +87,20 @@ public class SparkBulkTagJob {
|
|||
final String baseURL = parser.get("baseURL");
|
||||
log.info("baseURL: {}", baseURL);
|
||||
|
||||
ProtoMap protoMappingParams = new Gson().fromJson(parser.get("pathMap"), ProtoMap.class);
|
||||
log.info("pathMap: {}", new Gson().toJson(protoMappingParams));
|
||||
log.info("pathMap: {}", parser.get("pathMap"));
|
||||
String protoMappingPath = parser.get("pathMap");
|
||||
|
||||
final String hdfsNameNode = parser.get("nameNode");
|
||||
log.info("nameNode: {}", hdfsNameNode);
|
||||
|
||||
Configuration configuration = new Configuration();
|
||||
configuration.set("fs.defaultFS", hdfsNameNode);
|
||||
FileSystem fs = FileSystem.get(configuration);
|
||||
|
||||
String temp = IOUtils.toString(fs.open(new Path(protoMappingPath)), StandardCharsets.UTF_8);
|
||||
log.info("protoMap: {}", temp);
|
||||
ProtoMap protoMap = new Gson().fromJson(temp, ProtoMap.class);
|
||||
log.info("pathMap: {}", new Gson().toJson(protoMap));
|
||||
|
||||
SparkConf conf = new SparkConf();
|
||||
CommunityConfiguration cc;
|
||||
|
@ -88,10 +122,130 @@ public class SparkBulkTagJob {
|
|||
isSparkSessionManaged,
|
||||
spark -> {
|
||||
extendCommunityConfigurationForEOSC(spark, inputPath, cc);
|
||||
execBulkTag(spark, inputPath, outputPath, protoMappingParams, cc);
|
||||
execBulkTag(
|
||||
spark, inputPath, outputPath, protoMap, cc);
|
||||
execDatasourceTag(spark, inputPath, outputPath, Utils.getDatasourceCommunities(baseURL));
|
||||
execProjectTag(spark, inputPath, outputPath, Utils.getCommunityProjects(baseURL));
|
||||
});
|
||||
}
|
||||
|
||||
private static void execProjectTag(SparkSession spark, String inputPath, String outputPath,
|
||||
CommunityEntityMap communityProjects) {
|
||||
Dataset<Project> projects = readPath(spark, inputPath + "project", Project.class);
|
||||
Dataset<EntityCommunities> pc = spark
|
||||
.createDataset(
|
||||
communityProjects
|
||||
.keySet()
|
||||
.stream()
|
||||
.map(k -> EntityCommunities.newInstance(k, communityProjects.get(k)))
|
||||
.collect(Collectors.toList()),
|
||||
Encoders.bean(EntityCommunities.class));
|
||||
|
||||
projects
|
||||
.joinWith(pc, projects.col("id").equalTo(pc.col("entityId")), "left")
|
||||
.map((MapFunction<Tuple2<Project, EntityCommunities>, Project>) t2 -> {
|
||||
Project ds = t2._1();
|
||||
if (t2._2() != null) {
|
||||
List<String> context = Optional
|
||||
.ofNullable(ds.getContext())
|
||||
.map(v -> v.stream().map(c -> c.getId()).collect(Collectors.toList()))
|
||||
.orElse(new ArrayList<>());
|
||||
|
||||
if (!Optional.ofNullable(ds.getContext()).isPresent())
|
||||
ds.setContext(new ArrayList<>());
|
||||
t2._2().getCommunitiesId().forEach(c -> {
|
||||
if (!context.contains(c)) {
|
||||
Context con = new Context();
|
||||
con.setId(c);
|
||||
con
|
||||
.setDataInfo(
|
||||
Arrays
|
||||
.asList(
|
||||
OafMapperUtils
|
||||
.dataInfo(
|
||||
false, TaggingConstants.BULKTAG_DATA_INFO_TYPE, true, false,
|
||||
OafMapperUtils
|
||||
.qualifier(
|
||||
TaggingConstants.CLASS_ID_DATASOURCE,
|
||||
TaggingConstants.CLASS_NAME_BULKTAG_DATASOURCE,
|
||||
ModelConstants.DNET_PROVENANCE_ACTIONS,
|
||||
ModelConstants.DNET_PROVENANCE_ACTIONS),
|
||||
"1")));
|
||||
ds.getContext().add(con);
|
||||
}
|
||||
});
|
||||
}
|
||||
return ds;
|
||||
}, Encoders.bean(Project.class))
|
||||
.write()
|
||||
.mode(SaveMode.Overwrite)
|
||||
.option("compression", "gzip")
|
||||
.json(outputPath + "project");
|
||||
|
||||
readPath(spark, outputPath + "project", Datasource.class)
|
||||
.write()
|
||||
.mode(SaveMode.Overwrite)
|
||||
.option("compression", "gzip")
|
||||
.json(inputPath + "project");
|
||||
}
|
||||
|
||||
private static void execDatasourceTag(SparkSession spark, String inputPath, String outputPath,
|
||||
List<EntityCommunities> datasourceCommunities) {
|
||||
Dataset<Datasource> datasource = readPath(spark, inputPath + "datasource", Datasource.class);
|
||||
|
||||
Dataset<EntityCommunities> dc = spark
|
||||
.createDataset(datasourceCommunities, Encoders.bean(EntityCommunities.class));
|
||||
|
||||
datasource
|
||||
.joinWith(dc, datasource.col("id").equalTo(dc.col("entityId")), "left")
|
||||
.map((MapFunction<Tuple2<Datasource, EntityCommunities>, Datasource>) t2 -> {
|
||||
Datasource ds = t2._1();
|
||||
if (t2._2() != null) {
|
||||
|
||||
List<String> context = Optional
|
||||
.ofNullable(ds.getContext())
|
||||
.map(v -> v.stream().map(c -> c.getId()).collect(Collectors.toList()))
|
||||
.orElse(new ArrayList<>());
|
||||
|
||||
if (!Optional.ofNullable(ds.getContext()).isPresent())
|
||||
ds.setContext(new ArrayList<>());
|
||||
|
||||
t2._2().getCommunitiesId().forEach(c -> {
|
||||
if (!context.contains(c)) {
|
||||
Context con = new Context();
|
||||
con.setId(c);
|
||||
con
|
||||
.setDataInfo(
|
||||
Arrays
|
||||
.asList(
|
||||
OafMapperUtils
|
||||
.dataInfo(
|
||||
false, TaggingConstants.BULKTAG_DATA_INFO_TYPE, true, false,
|
||||
OafMapperUtils
|
||||
.qualifier(
|
||||
TaggingConstants.CLASS_ID_DATASOURCE,
|
||||
TaggingConstants.CLASS_NAME_BULKTAG_DATASOURCE,
|
||||
ModelConstants.DNET_PROVENANCE_ACTIONS,
|
||||
ModelConstants.DNET_PROVENANCE_ACTIONS),
|
||||
"1")));
|
||||
ds.getContext().add(con);
|
||||
}
|
||||
});
|
||||
}
|
||||
return ds;
|
||||
}, Encoders.bean(Datasource.class))
|
||||
.write()
|
||||
.mode(SaveMode.Overwrite)
|
||||
.option("compression", "gzip")
|
||||
.json(outputPath + "datasource");
|
||||
|
||||
readPath(spark, outputPath + "datasource", Datasource.class)
|
||||
.write()
|
||||
.mode(SaveMode.Overwrite)
|
||||
.option("compression", "gzip")
|
||||
.json(inputPath + "datasource");
|
||||
}
|
||||
|
||||
private static void extendCommunityConfigurationForEOSC(SparkSession spark, String inputPath,
|
||||
CommunityConfiguration cc) {
|
||||
|
||||
|
@ -129,6 +283,11 @@ public class SparkBulkTagJob {
|
|||
ProtoMap protoMappingParams,
|
||||
CommunityConfiguration communityConfiguration) {
|
||||
|
||||
try {
|
||||
System.out.println(new ObjectMapper().writeValueAsString(protoMappingParams));
|
||||
} catch (JsonProcessingException e) {
|
||||
throw new RuntimeException(e);
|
||||
}
|
||||
ModelSupport.entityTypes
|
||||
.keySet()
|
||||
.parallelStream()
|
||||
|
|
|
@ -38,7 +38,6 @@ public class ExecSubstringAction implements Serializable {
|
|||
}
|
||||
|
||||
public String execSubstring() {
|
||||
|
||||
return this.value.substring(Integer.valueOf(this.from), Integer.valueOf(this.to));
|
||||
|
||||
}
|
||||
|
|
|
@ -1,4 +1,4 @@
|
|||
sourcePath=/tmp/beta_provision/graph/10_graph_orcid_enriched
|
||||
sourcePath=/tmp/beta_provision/graph/09_graph_orcid_enriched
|
||||
resumeFrom=ResultProject
|
||||
allowedsemrelsorcidprop=isSupplementedBy;isSupplementTo
|
||||
allowedsemrelsresultproject=isSupplementedBy;isSupplementTo
|
||||
|
@ -7,20 +7,23 @@ datasourceWhitelistForCountryPropagation=10|opendoar____::16e6a3326dd7d868cbc926
|
|||
#allowedtypes=pubsrepository::institutional
|
||||
allowedtypes=Institutional
|
||||
outputPath=/tmp/miriam/graph/11_graph_orcid
|
||||
pathMap ={"author":"$['author'][*]['fullname']", \
|
||||
"title":"$['title'][*]['value']",\
|
||||
"orcid":"$['author'][*]['pid'][*][?(@['qualifier']['classid']=='orcid')]['value']" ,\
|
||||
"orcid_pending":"$['author'][*]['pid'][*][?(@['qualifier']['classid']=='orcid_pending')]['value']" ,\
|
||||
"contributor" : "$['contributor'][*]['value']",\
|
||||
"description" : "$['description'][*]['value']",\
|
||||
"subject" :"$['subject'][*]['value']" , \
|
||||
"fos" : "$['subject'][?(@['qualifier']['classid']=='FOS')].value" ,\
|
||||
"sdg" : "$['subject'][?(@['qualifier']['classid']=='SDG')].value",\
|
||||
"journal":"$['journal'].name",\
|
||||
"hostedby":"$['instance'][*]['hostedby']['key']",\
|
||||
"collectedfrom":"$['instance'][*]['collectedfrom']['key']",\
|
||||
"publisher":"$['publisher'].value",\
|
||||
"publicationyear":"$['dateofacceptance'].value"}
|
||||
pathMap ={"author":{"path":"$['author'][*]['fullname']"}, \
|
||||
"title":{"path":"$['title'][*]['value']"},\
|
||||
"orcid":{"path":"$['author'][*]['pid'][*][?(@['qualifier']['classid']=='orcid')]['value']"} ,\
|
||||
"orcid_pending":{"path":"$['author'][*]['pid'][*][?(@['qualifier']['classid']=='orcid_pending')]['value']"} ,\
|
||||
"contributor" : {"path":"$['contributor'][*]['value']"},\
|
||||
"description" : {"path":"$['description'][*]['value']"},\
|
||||
"subject" :{"path":"$['subject'][*]['value']"}, \
|
||||
"fos" : {"path":"$['subject'][?(@['qualifier']['classid']=='FOS')].value"} ,\
|
||||
"sdg" : {"path":"$['subject'][?(@['qualifier']['classid']=='SDG')].value"},\
|
||||
"journal":{"path":"$['journal'].name"},\
|
||||
"hostedby":{"path":"$['instance'][*]['hostedby']['key']"},\
|
||||
"collectedfrom":{"path":"$['instance'][*]['collectedfrom']['key']"},\
|
||||
"publisher":{"path":"$['publisher'].value"},\
|
||||
"publicationyear":{"path":"$['dateofacceptance'].value", "action":{"class":"eu.dnetlib.dhp.bulktag.actions.ExecSubstringAction",\
|
||||
"method":"execSubstring",\
|
||||
"params":[{"param_name":"From","param_value":0},\
|
||||
{"param_name":"To","param_value":4}]}}}
|
||||
blacklist=empty
|
||||
allowedpids=orcid;orcid_pending
|
||||
baseURL = https://services.openaire.eu/openaire/community/
|
||||
|
|
|
@ -33,6 +33,11 @@
|
|||
"paramName": "bu",
|
||||
"paramLongName": "baseURL",
|
||||
"paramDescription": "this parameter is to specify the api to be queried (beta or production)",
|
||||
"paramRequired": false
|
||||
}
|
||||
"paramRequired": true
|
||||
},{
|
||||
"paramName": "nn",
|
||||
"paramLongName": "nameNode",
|
||||
"paramDescription": "this parameter is to specify the api to be queried (beta or production)",
|
||||
"paramRequired": true
|
||||
}
|
||||
]
|
|
@ -53,10 +53,10 @@
|
|||
</property>
|
||||
<property>
|
||||
<name>memoryOverhead</name>
|
||||
<value>3G</value>
|
||||
<value>4G</value>
|
||||
</property>
|
||||
<property>
|
||||
<name>partitions</name>
|
||||
<value>3284</value>
|
||||
<value>15000</value>
|
||||
</property>
|
||||
</configuration>
|
|
@ -76,6 +76,7 @@
|
|||
<arg>--outputPath</arg><arg>${workingDir}/bulktag/</arg>
|
||||
<arg>--pathMap</arg><arg>${pathMap}</arg>
|
||||
<arg>--baseURL</arg><arg>${baseURL}</arg>
|
||||
<arg>--nameNode</arg><arg>${nameNode}</arg>
|
||||
</spark>
|
||||
<ok to="End"/>
|
||||
<error to="Kill"/>
|
||||
|
|
|
@ -6,14 +6,19 @@ import static eu.dnetlib.dhp.bulktag.community.TaggingConstants.ZENODO_COMMUNITY
|
|||
import java.io.IOException;
|
||||
import java.nio.file.Files;
|
||||
import java.nio.file.Path;
|
||||
import java.util.Arrays;
|
||||
import java.util.List;
|
||||
|
||||
import org.apache.commons.io.FileUtils;
|
||||
import org.apache.commons.io.IOUtils;
|
||||
import org.apache.hadoop.conf.Configuration;
|
||||
import org.apache.hadoop.fs.FileSystem;
|
||||
import org.apache.hadoop.fs.LocalFileSystem;
|
||||
import org.apache.spark.SparkConf;
|
||||
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.sql.Encoders;
|
||||
import org.apache.spark.sql.Row;
|
||||
import org.apache.spark.sql.SparkSession;
|
||||
|
@ -25,14 +30,16 @@ import org.slf4j.Logger;
|
|||
import org.slf4j.LoggerFactory;
|
||||
|
||||
import com.fasterxml.jackson.databind.ObjectMapper;
|
||||
import com.google.gson.Gson;
|
||||
|
||||
import eu.dnetlib.dhp.bulktag.community.ProtoMap;
|
||||
import eu.dnetlib.dhp.schema.oaf.*;
|
||||
|
||||
public class BulkTagJobTest {
|
||||
|
||||
private static final ObjectMapper OBJECT_MAPPER = new ObjectMapper();
|
||||
|
||||
public static final String pathMap = "{\"author\":{\"path\":\"$['author'][*]['fullname']\"}," +
|
||||
public static final String pathMap = "{\"protoMap\":{\"author\":{\"path\":\"$['author'][*]['fullname']\"}," +
|
||||
" \"title\":{\"path\":\"$['title'][*]['value']\"}, " +
|
||||
" \"orcid\":{\"path\":\"$['author'][*]['pid'][*][?(@['qualifier']['classid']=='orcid')]['value']\"} , " +
|
||||
" \"orcid_pending\":{\"path\":\"$['author'][*]['pid'][*][?(@['qualifier']['classid']=='orcid_pending')]['value']\"} ,"
|
||||
|
@ -51,7 +58,7 @@ public class BulkTagJobTest {
|
|||
"\"method\":\"execSubstring\"," +
|
||||
"\"params\":[" +
|
||||
"{\"paramName\":\"From\", \"paramValue\":0}, " +
|
||||
"{\"paramName\":\"To\",\"paramValue\":4}]}}}";
|
||||
"{\"paramName\":\"To\",\"paramValue\":4}]}}}}";
|
||||
|
||||
private static SparkSession spark;
|
||||
|
||||
|
@ -231,6 +238,14 @@ public class BulkTagJobTest {
|
|||
|
||||
@Test
|
||||
void bulktagBySubjectPreviousContextNoProvenanceTest() throws Exception {
|
||||
LocalFileSystem fs = FileSystem.getLocal(new Configuration());
|
||||
fs
|
||||
.copyFromLocalFile(
|
||||
false, new org.apache.hadoop.fs.Path(getClass()
|
||||
.getResource("/eu/dnetlib/dhp/bulktag/pathMap/")
|
||||
.getPath()),
|
||||
new org.apache.hadoop.fs.Path(workingDir.toString() + "/data/bulktagging/protoMap"));
|
||||
|
||||
final String sourcePath = getClass()
|
||||
.getResource(
|
||||
"/eu/dnetlib/dhp/bulktag/sample/dataset/update_subject/contextnoprovenance/")
|
||||
|
@ -246,7 +261,8 @@ public class BulkTagJobTest {
|
|||
|
||||
"-outputPath", workingDir.toString() + "/",
|
||||
|
||||
"-pathMap", pathMap
|
||||
"-pathMap", workingDir.toString() + "/data/bulktagging/protoMap",
|
||||
"-nameNode", "local"
|
||||
});
|
||||
|
||||
final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext());
|
||||
|
@ -316,6 +332,7 @@ public class BulkTagJobTest {
|
|||
final String sourcePath = getClass()
|
||||
.getResource("/eu/dnetlib/dhp/bulktag/sample/publication/update_datasource/")
|
||||
.getPath();
|
||||
|
||||
SparkBulkTagJob
|
||||
.main(
|
||||
new String[] {
|
||||
|
@ -325,7 +342,7 @@ public class BulkTagJobTest {
|
|||
"-taggingConf", taggingConf,
|
||||
|
||||
"-outputPath", workingDir.toString() + "/",
|
||||
|
||||
"-baseURL", "https://services.openaire.eu/openaire/community/",
|
||||
"-pathMap", pathMap
|
||||
});
|
||||
|
||||
|
@ -383,6 +400,71 @@ public class BulkTagJobTest {
|
|||
.count());
|
||||
}
|
||||
|
||||
@Test
|
||||
void datasourceTag() throws Exception {
|
||||
final String sourcePath = getClass()
|
||||
.getResource("/eu/dnetlib/dhp/bulktag/sample/publication/update_datasource/")
|
||||
.getPath();
|
||||
LocalFileSystem fs = FileSystem.getLocal(new Configuration());
|
||||
fs
|
||||
.copyFromLocalFile(
|
||||
false, new org.apache.hadoop.fs.Path(getClass()
|
||||
.getResource("/eu/dnetlib/dhp/bulktag/pathMap/")
|
||||
.getPath()),
|
||||
new org.apache.hadoop.fs.Path(workingDir.toString() + "/data/bulktagging/protoMap"));
|
||||
SparkBulkTagJob
|
||||
.main(
|
||||
new String[] {
|
||||
|
||||
"-isSparkSessionManaged", Boolean.FALSE.toString(),
|
||||
"-sourcePath", sourcePath,
|
||||
"-taggingConf", taggingConf,
|
||||
|
||||
"-outputPath", workingDir.toString() + "/",
|
||||
"-baseURL", "https://services.openaire.eu/openaire/community/",
|
||||
|
||||
"-pathMap", workingDir.toString() + "/data/bulktagging/protoMap/pathMap",
|
||||
"-nameNode", "local"
|
||||
});
|
||||
|
||||
final JavaSparkContext sc = JavaSparkContext.fromSparkContext(spark.sparkContext());
|
||||
|
||||
JavaRDD<Datasource> tmp = sc
|
||||
.textFile(workingDir.toString() + "/datasource")
|
||||
.map(item -> OBJECT_MAPPER.readValue(item, Datasource.class));
|
||||
|
||||
Assertions.assertEquals(3, tmp.count());
|
||||
org.apache.spark.sql.Dataset<Datasource> verificationDataset = spark
|
||||
.createDataset(tmp.rdd(), Encoders.bean(Datasource.class));
|
||||
|
||||
verificationDataset.createOrReplaceTempView("datasource");
|
||||
|
||||
String query = "select id, MyT.id community, MyD.provenanceaction.classid provenance, MyD.provenanceaction.classname name "
|
||||
+ "from datasource "
|
||||
+ "lateral view explode(context) c as MyT "
|
||||
+ "lateral view explode(MyT.datainfo) d as MyD "
|
||||
+ "where MyD.inferenceprovenance = 'bulktagging'";
|
||||
|
||||
org.apache.spark.sql.Dataset<Row> idExplodeCommunity = spark.sql(query);
|
||||
|
||||
idExplodeCommunity.show(false);
|
||||
|
||||
Assertions.assertEquals(3, idExplodeCommunity.count());
|
||||
Assertions
|
||||
.assertEquals(
|
||||
3, idExplodeCommunity.filter("provenance = 'community:datasource'").count());
|
||||
Assertions
|
||||
.assertEquals(
|
||||
3,
|
||||
idExplodeCommunity
|
||||
.filter("name = 'Bulktagging for Community - Datasource'")
|
||||
.count());
|
||||
|
||||
Assertions.assertEquals(2, idExplodeCommunity.filter("community = 'dh-ch'").count());
|
||||
Assertions.assertEquals(1, idExplodeCommunity.filter("community = 'clarin'").count());
|
||||
|
||||
}
|
||||
|
||||
@Test
|
||||
void bulktagByZenodoCommunityTest() throws Exception {
|
||||
final String sourcePath = getClass()
|
||||
|
@ -1699,4 +1781,40 @@ public class BulkTagJobTest {
|
|||
|
||||
}
|
||||
|
||||
@Test
|
||||
public void prova() throws Exception {
|
||||
LocalFileSystem fs = FileSystem.getLocal(new Configuration());
|
||||
fs
|
||||
.copyFromLocalFile(
|
||||
false, new org.apache.hadoop.fs.Path(getClass()
|
||||
.getResource("/eu/dnetlib/dhp/bulktag/pathMap/")
|
||||
.getPath()),
|
||||
new org.apache.hadoop.fs.Path(workingDir.toString() + "/data/bulktagging/protoMap"));
|
||||
|
||||
final String sourcePath = getClass()
|
||||
.getResource(
|
||||
"/eu/dnetlib/dhp/bulktag/sample/dataset/update_subject/contextnoprovenance/")
|
||||
.getPath();
|
||||
|
||||
ProtoMap prova = new Gson()
|
||||
.fromJson(
|
||||
"{\"author\":{\"path\":\"$['author'][]['fullname']\"},\"title\":{\"path\":\"$['title'][]['value']\"},\"orcid\":{\"path\":\"$['author'][]['pid'][][?(@['qualifier']['classid']=='orcid')]['value']\"},\"orcid_pending\":{\"path\":\"$['author'][]['pid'][][?(@['qualifier']['classid']=='orcid_pending')]['value']\"},\"contributor\":{\"path\":\"$['contributor'][]['value']\"},\"description\":{\"path\":\"$['description'][]['value']\"},\"subject\":{\"path\":\"$['subject'][]['value']\"},\"fos\":{\"path\":\"$['subject'][?(@['qualifier']['classid']=='FOS')].value\"},\"sdg\":{\"path\":\"$['subject'][?(@['qualifier']['classid']=='SDG')].value\"},\"journal\":{\"path\":\"$['journal'].name\"},\"hostedby\":{\"path\":\"$['instance'][]['hostedby']['key']\"},\"collectedfrom\":{\"path\":\"$['instance'][*]['collectedfrom']['key']\"},\"publisher\":{\"path\":\"$['publisher'].value\"},\"publicationyear\":{\"path\":\"$['dateofacceptance'].value\",\"action\":{\"clazz\":\"eu.dnetlib.dhp.bulktag.actions.ExecSubstringAction\",\"method\":\"execSubstring\",\"params\":[{\"paramName\":\"From\",\"paramValue\":0},{\"paramName\":\"To\",\"paramValue\":4}]}}}",
|
||||
ProtoMap.class);
|
||||
SparkBulkTagJob
|
||||
.main(
|
||||
new String[] {
|
||||
|
||||
"-isSparkSessionManaged", Boolean.FALSE.toString(),
|
||||
"-sourcePath", sourcePath,
|
||||
"-taggingConf", taggingConf,
|
||||
|
||||
"-outputPath", workingDir.toString() + "/",
|
||||
|
||||
"-pathMap", workingDir.toString() + "/data/bulktagging/protoMap/pathMap",
|
||||
"-baseURL", "none",
|
||||
"-nameNode", "local"
|
||||
});
|
||||
|
||||
}
|
||||
|
||||
}
|
||||
|
|
|
@ -0,0 +1,58 @@
|
|||
{
|
||||
"author":{
|
||||
"path":"$['author'][*]['fullname']"
|
||||
},
|
||||
"title":{
|
||||
"path":"$['title'][*]['value']"
|
||||
},
|
||||
"orcid":{
|
||||
"path":"$['author'][*]['pid'][*][?(@['qualifier']['classid']=='orcid')]['value']"
|
||||
},
|
||||
"orcid_pending":{
|
||||
"path":"$['author'][*]['pid'][*][?(@['qualifier']['classid']=='orcid_pending')]['value']"
|
||||
},
|
||||
"contributor":{
|
||||
"path":"$['contributor'][*]['value']"
|
||||
},
|
||||
"description":{
|
||||
"path":"$['description'][*]['value']"
|
||||
},
|
||||
"subject":{
|
||||
"path":"$['subject'][*]['value']"
|
||||
},
|
||||
"fos":{
|
||||
"path":"$['subject'][?(@['qualifier']['classid']=='FOS')].value"
|
||||
},
|
||||
"sdg":{
|
||||
"path":"$['subject'][?(@['qualifier']['classid']=='SDG')].value"
|
||||
},
|
||||
"journal":{
|
||||
"path":"$['journal'].name"
|
||||
},
|
||||
"hostedby":{
|
||||
"path":"$['instance'][*]['hostedby']['key']"
|
||||
},
|
||||
"collectedfrom":{
|
||||
"path":"$['instance'][*]['collectedfrom']['key']"
|
||||
},
|
||||
"publisher":{
|
||||
"path":"$['publisher'].value"
|
||||
},
|
||||
"publicationyear":{
|
||||
"path":"$['dateofacceptance'].value",
|
||||
"action":{
|
||||
"clazz":"eu.dnetlib.dhp.bulktag.actions.ExecSubstringAction",
|
||||
"method":"execSubstring",
|
||||
"params":[
|
||||
{
|
||||
"paramName":"From",
|
||||
"paramValue":0
|
||||
},
|
||||
{
|
||||
"paramName":"To",
|
||||
"paramValue":4
|
||||
}
|
||||
]
|
||||
}
|
||||
}
|
||||
}
|
File diff suppressed because one or more lines are too long
1
pom.xml
1
pom.xml
|
@ -888,6 +888,7 @@
|
|||
<mockito-core.version>3.3.3</mockito-core.version>
|
||||
<mongodb.driver.version>3.4.2</mongodb.driver.version>
|
||||
<vtd.version>[2.12,3.0)</vtd.version>
|
||||
<dhp-schemas.version>[5.17.3]</dhp-schemas.version>
|
||||
<dhp-schemas.version>[6.1.0]</dhp-schemas.version>
|
||||
<dnet-actionmanager-api.version>[4.0.3]</dnet-actionmanager-api.version>
|
||||
<dnet-actionmanager-common.version>[6.0.5]</dnet-actionmanager-common.version>
|
||||
|
|
Loading…
Reference in New Issue