przemyslaw.jacewicz
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Merge remote-tracking branch 'origin/master' into doiboost

This commit is contained in:
Sandro La Bruzzo 2020-04-22 15:00:57 +02:00
parent 157915988c 9e4d58f505
commit bb6c9785b4
13 changed files with 1140 additions and 1211 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkCreateDedupRecord.java
View File

@ -11,8 +11,6 @@ import eu.dnetlib.pace.config.DedupConfig;
import java.io.IOException;
import org.apache.commons.io.IOUtils;
import org.apache.spark.SparkConf;
import org.apache.spark.api.java.function.MapFunction;
import org.apache.spark.sql.Encoders;
import org.apache.spark.sql.SaveMode;
import org.apache.spark.sql.SparkSession;
import org.dom4j.DocumentException;
@ -72,12 +70,9 @@ public class SparkCreateDedupRecord extends AbstractSparkAction {
Class<OafEntity> clazz = ModelSupport.entityTypes.get(EntityType.valueOf(subEntity));
DedupRecordFactory.createDedupRecord(spark, mergeRelPath, entityPath, clazz)
.map(
(MapFunction<OafEntity, String>)
value -> OBJECT_MAPPER.writeValueAsString(value),
Encoders.STRING())
.write()
.mode(SaveMode.Overwrite)
.option("compression", "gzip")
.json(outputPath);
}
}

19
dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/SparkUpdateEntity.java
View File

@ -33,7 +33,7 @@ public class SparkUpdateEntity extends AbstractSparkAction {
private static final Logger log = LoggerFactory.getLogger(SparkUpdateEntity.class);
final String IDJSONPATH = "$.id";
private static final String IDJSONPATH = "$.id";
public SparkUpdateEntity(ArgumentApplicationParser parser, SparkSession spark) {
super(parser, spark);
@ -65,27 +65,25 @@ public class SparkUpdateEntity extends AbstractSparkAction {
log.info("workingPath: '{}'", workingPath);
log.info("dedupGraphPath: '{}'", dedupGraphPath);
final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext());
final JavaSparkContext sc = JavaSparkContext.fromSparkContext(spark.sparkContext());
// for each entity
ModelSupport.entityTypes.forEach(
(entity, clazz) -> {
final String outputPath = dedupGraphPath + "/" + entity;
(type, clazz) -> {
final String outputPath = dedupGraphPath + "/" + type;
removeOutputDir(spark, outputPath);
JavaRDD<String> sourceEntity =
sc.textFile(
DedupUtility.createEntityPath(
graphBasePath, entity.toString()));
DedupUtility.createEntityPath(graphBasePath, type.toString()));
if (mergeRelExists(workingPath, entity.toString())) {
if (mergeRelExists(workingPath, type.toString())) {
final String mergeRelPath =
DedupUtility.createMergeRelPath(
workingPath, "*", entity.toString());
DedupUtility.createMergeRelPath(workingPath, "*", type.toString());
final String dedupRecordPath =
DedupUtility.createDedupRecordPath(
workingPath, "*", entity.toString());
workingPath, "*", type.toString());
final Dataset<Relation> rel =
spark.read().load(mergeRelPath).as(Encoders.bean(Relation.class));
@ -107,7 +105,6 @@ public class SparkUpdateEntity extends AbstractSparkAction {
MapDocumentUtil.getJPathString(
IDJSONPATH, s),
s));
JavaRDD<String> map =
entitiesWithId
.leftOuterJoin(mergedIds)

20
dhp-workflows/dhp-dedup-openaire/src/test/java/eu/dnetlib/dhp/oa/dedup/SparkDedupTest.java
View File

@ -72,7 +72,7 @@ public class SparkDedupTest implements Serializable {
spark =
SparkSession.builder()
.appName(SparkCreateSimRels.class.getSimpleName())
.appName(SparkDedupTest.class.getSimpleName())
.master("local[*]")
.config(new SparkConf())
.getOrCreate();
@ -272,7 +272,17 @@ public class SparkDedupTest implements Serializable {
.distinct()
.count();
assertEquals(831, publications);
long mergedSw =
spark.read()
.load(testOutputBasePath + "/" + testActionSetId + "/software_mergerel")
.as(Encoders.bean(Relation.class))
.where("relClass=='merges'")
.javaRDD()
.map(Relation::getTarget)
.distinct()
.count();
assertEquals(897, publications);
assertEquals(835, organizations);
assertEquals(100, projects);
assertEquals(100, datasource);
@ -288,8 +298,14 @@ public class SparkDedupTest implements Serializable {
.filter(this::isDeletedByInference)
.count();
long deletedSw =
jsc.textFile(testDedupGraphBasePath + "/software")
.filter(this::isDeletedByInference)
.count();
assertEquals(mergedOrgs, deletedOrgs);
assertEquals(mergedPubs, deletedPubs);
assertEquals(mergedSw, deletedSw);
}
@Test

200
dhp-workflows/dhp-dedup-openaire/src/test/resources/eu/dnetlib/dhp/dedup/entities/datasource
View File

File diff suppressed because one or more lines are too long

840
dhp-workflows/dhp-graph-mapper/src/main/java/eu/dnetlib/dhp/oa/graph/raw/AbstractMdRecordToOafMapper.java
View File

@ -10,6 +10,20 @@ import static eu.dnetlib.dhp.oa.graph.raw.common.OafMapperUtils.oaiIProvenance;
import static eu.dnetlib.dhp.oa.graph.raw.common.OafMapperUtils.qualifier;
import static eu.dnetlib.dhp.oa.graph.raw.common.OafMapperUtils.structuredProperty;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Date;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import org.apache.commons.lang3.StringUtils;
import org.dom4j.Document;
import org.dom4j.DocumentFactory;
import org.dom4j.DocumentHelper;
import org.dom4j.Node;
import eu.dnetlib.dhp.oa.graph.raw.common.MigrationConstants;
import eu.dnetlib.dhp.schema.oaf.Author;
import eu.dnetlib.dhp.schema.oaf.DataInfo;
import eu.dnetlib.dhp.schema.oaf.Dataset;
@ -27,452 +41,388 @@ import eu.dnetlib.dhp.schema.oaf.Relation;
import eu.dnetlib.dhp.schema.oaf.Result;
import eu.dnetlib.dhp.schema.oaf.Software;
import eu.dnetlib.dhp.schema.oaf.StructuredProperty;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Date;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import org.apache.commons.lang3.StringUtils;
import org.dom4j.Document;
import org.dom4j.DocumentFactory;
import org.dom4j.DocumentHelper;
import org.dom4j.Node;
public abstract class AbstractMdRecordToOafMapper {
protected final Map<String, String> code2name;
protected static final Qualifier MAIN_TITLE_QUALIFIER =
qualifier("main title", "main title", "dnet:dataCite_title", "dnet:dataCite_title");
protected static final Qualifier PUBLICATION_RESULTTYPE_QUALIFIER =
qualifier(
"publication",
"publication",
"dnet:result_typologies",
"dnet:result_typologies");
protected static final Qualifier DATASET_RESULTTYPE_QUALIFIER =
qualifier("dataset", "dataset", "dnet:result_typologies", "dnet:result_typologies");
protected static final Qualifier SOFTWARE_RESULTTYPE_QUALIFIER =
qualifier("software", "software", "dnet:result_typologies", "dnet:result_typologies");
protected static final Qualifier OTHER_RESULTTYPE_QUALIFIER =
qualifier("other", "other", "dnet:result_typologies", "dnet:result_typologies");
protected static final Qualifier REPOSITORY_QUALIFIER =
qualifier(
"sysimport:crosswalk:repository",
"sysimport:crosswalk:repository",
"dnet:provenanceActions",
"dnet:provenanceActions");
protected AbstractMdRecordToOafMapper(final Map<String, String> code2name) {
this.code2name = code2name;
}
public List<Oaf> processMdRecord(final String xml) {
try {
final Map<String, String> nsContext = new HashMap<>();
nsContext.put("dr", "http://www.driver-repository.eu/namespace/dr");
nsContext.put("dri", "http://www.driver-repository.eu/namespace/dri");
nsContext.put("oaf", "http://namespace.openaire.eu/oaf");
nsContext.put("oai", "http://www.openarchives.org/OAI/2.0/");
nsContext.put("prov", "http://www.openarchives.org/OAI/2.0/provenance");
nsContext.put("dc", "http://purl.org/dc/elements/1.1/");
nsContext.put("datacite", "http://datacite.org/schema/kernel-3");
DocumentFactory.getInstance().setXPathNamespaceURIs(nsContext);
final Document doc =
DocumentHelper.parseText(
xml.replaceAll(
"http://datacite.org/schema/kernel-4",
"http://datacite.org/schema/kernel-3"));
final String type = doc.valueOf("//dr:CobjCategory/@type");
final KeyValue collectedFrom =
keyValue(
createOpenaireId(10, doc.valueOf("//oaf:collectedFrom/@id"), true),
doc.valueOf("//oaf:collectedFrom/@name"));
final KeyValue hostedBy =
StringUtils.isBlank(doc.valueOf("//oaf:hostedBy/@id"))
? collectedFrom
: keyValue(
createOpenaireId(10, doc.valueOf("//oaf:hostedBy/@id"), true),
doc.valueOf("//oaf:hostedBy/@name"));
final DataInfo info = prepareDataInfo(doc);
final long lastUpdateTimestamp = new Date().getTime();
return createOafs(doc, type, collectedFrom, hostedBy, info, lastUpdateTimestamp);
} catch (final Exception e) {
throw new RuntimeException(e);
}
}
protected List<Oaf> createOafs(
final Document doc,
final String type,
final KeyValue collectedFrom,
final KeyValue hostedBy,
final DataInfo info,
final long lastUpdateTimestamp) {
final List<Oaf> oafs = new ArrayList<>();
switch (type.toLowerCase()) {
case "":
case "publication":
final Publication p = new Publication();
populateResultFields(p, doc, collectedFrom, hostedBy, info, lastUpdateTimestamp);
p.setResulttype(PUBLICATION_RESULTTYPE_QUALIFIER);
p.setJournal(prepareJournal(doc, info));
oafs.add(p);
break;
case "dataset":
final Dataset d = new Dataset();
populateResultFields(d, doc, collectedFrom, hostedBy, info, lastUpdateTimestamp);
d.setResulttype(DATASET_RESULTTYPE_QUALIFIER);
d.setStoragedate(prepareDatasetStorageDate(doc, info));
d.setDevice(prepareDatasetDevice(doc, info));
d.setSize(prepareDatasetSize(doc, info));
d.setVersion(prepareDatasetVersion(doc, info));
d.setLastmetadataupdate(prepareDatasetLastMetadataUpdate(doc, info));
d.setMetadataversionnumber(prepareDatasetMetadataVersionNumber(doc, info));
d.setGeolocation(prepareDatasetGeoLocations(doc, info));
oafs.add(d);
break;
case "software":
final Software s = new Software();
populateResultFields(s, doc, collectedFrom, hostedBy, info, lastUpdateTimestamp);
s.setResulttype(SOFTWARE_RESULTTYPE_QUALIFIER);
s.setDocumentationUrl(prepareSoftwareDocumentationUrls(doc, info));
s.setLicense(prepareSoftwareLicenses(doc, info));
s.setCodeRepositoryUrl(prepareSoftwareCodeRepositoryUrl(doc, info));
s.setProgrammingLanguage(prepareSoftwareProgrammingLanguage(doc, info));
oafs.add(s);
break;
case "otherresearchproducts":
default:
final OtherResearchProduct o = new OtherResearchProduct();
populateResultFields(o, doc, collectedFrom, hostedBy, info, lastUpdateTimestamp);
o.setResulttype(OTHER_RESULTTYPE_QUALIFIER);
o.setContactperson(prepareOtherResearchProductContactPersons(doc, info));
o.setContactgroup(prepareOtherResearchProductContactGroups(doc, info));
o.setTool(prepareOtherResearchProductTools(doc, info));
oafs.add(o);
break;
}
if (!oafs.isEmpty()) {
oafs.addAll(addProjectRels(doc, collectedFrom, info, lastUpdateTimestamp));
oafs.addAll(addOtherResultRels(doc, collectedFrom, info, lastUpdateTimestamp));
}
return oafs;
}
private List<Oaf> addProjectRels(
final Document doc,
final KeyValue collectedFrom,
final DataInfo info,
final long lastUpdateTimestamp) {
final List<Oaf> res = new ArrayList<>();
final String docId = createOpenaireId(50, doc.valueOf("//dri:objIdentifier"), false);
for (final Object o : doc.selectNodes("//oaf:projectid")) {
final String projectId = createOpenaireId(40, ((Node) o).getText(), true);
final Relation r1 = new Relation();
r1.setRelType("resultProject");
r1.setSubRelType("outcome");
r1.setRelClass("isProducedBy");
r1.setSource(docId);
r1.setTarget(projectId);
r1.setCollectedfrom(Arrays.asList(collectedFrom));
r1.setDataInfo(info);
r1.setLastupdatetimestamp(lastUpdateTimestamp);
res.add(r1);
final Relation r2 = new Relation();
r2.setRelType("resultProject");
r2.setSubRelType("outcome");
r2.setRelClass("produces");
r2.setSource(projectId);
r2.setTarget(docId);
r2.setCollectedfrom(Arrays.asList(collectedFrom));
r2.setDataInfo(info);
r2.setLastupdatetimestamp(lastUpdateTimestamp);
res.add(r2);
}
return res;
}
protected abstract List<Oaf> addOtherResultRels(
final Document doc,
final KeyValue collectedFrom,
final DataInfo info,
final long lastUpdateTimestamp);
private void populateResultFields(
final Result r,
final Document doc,
final KeyValue collectedFrom,
final KeyValue hostedBy,
final DataInfo info,
final long lastUpdateTimestamp) {
r.setDataInfo(info);
r.setLastupdatetimestamp(lastUpdateTimestamp);
r.setId(createOpenaireId(50, doc.valueOf("//dri:objIdentifier"), false));
r.setOriginalId(Arrays.asList(doc.valueOf("//dri:objIdentifier")));
r.setCollectedfrom(Arrays.asList(collectedFrom));
r.setPid(
prepareListStructProps(
doc,
"//oaf:identifier",
"@identifierType",
"dnet:pid_types",
"dnet:pid_types",
info));
r.setDateofcollection(doc.valueOf("//dr:dateOfCollection"));
r.setDateoftransformation(doc.valueOf("//dr:dateOfTransformation"));
r.setExtraInfo(new ArrayList<>()); // NOT PRESENT IN MDSTORES
r.setOaiprovenance(prepareOAIprovenance(doc));
r.setAuthor(prepareAuthors(doc, info));
r.setLanguage(prepareLanguages(doc));
r.setCountry(new ArrayList<>()); // NOT PRESENT IN MDSTORES
r.setSubject(prepareSubjects(doc, info));
r.setTitle(prepareTitles(doc, info));
r.setRelevantdate(prepareRelevantDates(doc, info));
r.setDescription(prepareDescriptions(doc, info));
r.setDateofacceptance(prepareField(doc, "//oaf:dateAccepted", info));
r.setPublisher(preparePublisher(doc, info));
r.setEmbargoenddate(prepareField(doc, "//oaf:embargoenddate", info));
r.setSource(prepareSources(doc, info));
r.setFulltext(new ArrayList<>()); // NOT PRESENT IN MDSTORES
r.setFormat(prepareFormats(doc, info));
r.setContributor(prepareContributors(doc, info));
r.setResourcetype(prepareResourceType(doc, info));
r.setCoverage(prepareCoverages(doc, info));
r.setContext(new ArrayList<>()); // NOT PRESENT IN MDSTORES
r.setExternalReference(new ArrayList<>()); // NOT PRESENT IN MDSTORES
r.setInstance(prepareInstances(doc, info, collectedFrom, hostedBy));
}
protected abstract Qualifier prepareResourceType(Document doc, DataInfo info);
protected abstract List<Instance> prepareInstances(
Document doc, DataInfo info, KeyValue collectedfrom, KeyValue hostedby);
protected abstract List<Field<String>> prepareSources(Document doc, DataInfo info);
protected abstract List<StructuredProperty> prepareRelevantDates(Document doc, DataInfo info);
protected abstract List<Field<String>> prepareCoverages(Document doc, DataInfo info);
protected abstract List<Field<String>> prepareContributors(Document doc, DataInfo info);
protected abstract List<Field<String>> prepareFormats(Document doc, DataInfo info);
protected abstract Field<String> preparePublisher(Document doc, DataInfo info);
protected abstract List<Field<String>> prepareDescriptions(Document doc, DataInfo info);
protected abstract List<StructuredProperty> prepareTitles(Document doc, DataInfo info);
protected abstract List<StructuredProperty> prepareSubjects(Document doc, DataInfo info);
protected abstract Qualifier prepareLanguages(Document doc);
protected abstract List<Author> prepareAuthors(Document doc, DataInfo info);
protected abstract List<Field<String>> prepareOtherResearchProductTools(
Document doc, DataInfo info);
protected abstract List<Field<String>> prepareOtherResearchProductContactGroups(
Document doc, DataInfo info);
protected abstract List<Field<String>> prepareOtherResearchProductContactPersons(
Document doc, DataInfo info);
protected abstract Qualifier prepareSoftwareProgrammingLanguage(Document doc, DataInfo info);
protected abstract Field<String> prepareSoftwareCodeRepositoryUrl(Document doc, DataInfo info);
protected abstract List<StructuredProperty> prepareSoftwareLicenses(
Document doc, DataInfo info);
protected abstract List<Field<String>> prepareSoftwareDocumentationUrls(
Document doc, DataInfo info);
protected abstract List<GeoLocation> prepareDatasetGeoLocations(Document doc, DataInfo info);
protected abstract Field<String> prepareDatasetMetadataVersionNumber(
Document doc, DataInfo info);
protected abstract Field<String> prepareDatasetLastMetadataUpdate(Document doc, DataInfo info);
protected abstract Field<String> prepareDatasetVersion(Document doc, DataInfo info);
protected abstract Field<String> prepareDatasetSize(Document doc, DataInfo info);
protected abstract Field<String> prepareDatasetDevice(Document doc, DataInfo info);
protected abstract Field<String> prepareDatasetStorageDate(Document doc, DataInfo info);
private Journal prepareJournal(final Document doc, final DataInfo info) {
final Node n = doc.selectSingleNode("//oaf:journal");
if (n != null) {
final String name = n.getText();
final String issnPrinted = n.valueOf("@issn");
final String issnOnline = n.valueOf("@eissn");
final String issnLinking = n.valueOf("@lissn");
final String ep = n.valueOf("@ep");
final String iss = n.valueOf("@iss");
final String sp = n.valueOf("@sp");
final String vol = n.valueOf("@vol");
final String edition = n.valueOf("@edition");
if (StringUtils.isNotBlank(name)) {
return journal(
name,
issnPrinted,
issnOnline,
issnLinking,
ep,
iss,
sp,
vol,
edition,
null,
null,
info);
}
}
return null;
}
protected Qualifier prepareQualifier(
final Node node, final String xpath, final String schemeId, final String schemeName) {
final String classId = node.valueOf(xpath);
final String className = code2name.get(classId);
return qualifier(classId, className, schemeId, schemeName);
}
protected List<StructuredProperty> prepareListStructProps(
final Node node,
final String xpath,
final String xpathClassId,
final String schemeId,
final String schemeName,
final DataInfo info) {
final List<StructuredProperty> res = new ArrayList<>();
for (final Object o : node.selectNodes(xpath)) {
final Node n = (Node) o;
final String classId = n.valueOf(xpathClassId);
final String className = code2name.get(classId);
res.add(
structuredProperty(
n.getText(), classId, className, schemeId, schemeName, info));
}
return res;
}
protected List<StructuredProperty> prepareListStructProps(
final Node node, final String xpath, final Qualifier qualifier, final DataInfo info) {
final List<StructuredProperty> res = new ArrayList<>();
for (final Object o : node.selectNodes(xpath)) {
final Node n = (Node) o;
res.add(structuredProperty(n.getText(), qualifier, info));
}
return res;
}
protected List<StructuredProperty> prepareListStructProps(
final Node node, final String xpath, final DataInfo info) {
final List<StructuredProperty> res = new ArrayList<>();
for (final Object o : node.selectNodes(xpath)) {
final Node n = (Node) o;
res.add(
structuredProperty(
n.getText(),
n.valueOf("@classid"),
n.valueOf("@classname"),
n.valueOf("@schemeid"),
n.valueOf("@schemename"),
info));
}
return res;
}
protected OAIProvenance prepareOAIprovenance(final Document doc) {
final Node n =
doc.selectSingleNode(
"//*[local-name()='provenance']/*[local-name()='originDescription']");
if (n == null) {
return null;
}
final String identifier = n.valueOf("./*[local-name()='identifier']");
final String baseURL = n.valueOf("./*[local-name()='baseURL']");
;
final String metadataNamespace = n.valueOf("./*[local-name()='metadataNamespace']");
;
final boolean altered = n.valueOf("@altered").equalsIgnoreCase("true");
final String datestamp = n.valueOf("./*[local-name()='datestamp']");
;
final String harvestDate = n.valueOf("@harvestDate");
;
return oaiIProvenance(
identifier, baseURL, metadataNamespace, altered, datestamp, harvestDate);
}
protected DataInfo prepareDataInfo(final Document doc) {
final Node n = doc.selectSingleNode("//oaf:datainfo");
if (n == null) {
return dataInfo(false, null, false, false, REPOSITORY_QUALIFIER, "0.9");
}
final String paClassId = n.valueOf("./oaf:provenanceaction/@classid");
final String paClassName = n.valueOf("./oaf:provenanceaction/@classname");
final String paSchemeId = n.valueOf("./oaf:provenanceaction/@schemeid");
final String paSchemeName = n.valueOf("./oaf:provenanceaction/@schemename");
final boolean deletedbyinference =
Boolean.parseBoolean(n.valueOf("./oaf:deletedbyinference"));
final String inferenceprovenance = n.valueOf("./oaf:inferenceprovenance");
final Boolean inferred = Boolean.parseBoolean(n.valueOf("./oaf:inferred"));
final String trust = n.valueOf("./oaf:trust");
return dataInfo(
deletedbyinference,
inferenceprovenance,
inferred,
false,
qualifier(paClassId, paClassName, paSchemeId, paSchemeName),
trust);
}
protected Field<String> prepareField(final Node node, final String xpath, final DataInfo info) {
return field(node.valueOf(xpath), info);
}
protected List<Field<String>> prepareListFields(
final Node node, final String xpath, final DataInfo info) {
return listFields(info, prepareListString(node, xpath));
}
protected List<String> prepareListString(final Node node, final String xpath) {
final List<String> res = new ArrayList<>();
for (final Object o : node.selectNodes(xpath)) {
final String s = ((Node) o).getText().trim();
if (StringUtils.isNotBlank(s)) {
res.add(s);
}
}
return res;
}
protected final Map<String, String> code2name;
protected static final Qualifier MAIN_TITLE_QUALIFIER =
qualifier("main title", "main title", "dnet:dataCite_title", "dnet:dataCite_title");
protected AbstractMdRecordToOafMapper(final Map<String, String> code2name) {
this.code2name = code2name;
}
public List<Oaf> processMdRecord(final String xml) {
try {
final Map<String, String> nsContext = new HashMap<>();
nsContext.put("dr", "http://www.driver-repository.eu/namespace/dr");
nsContext.put("dri", "http://www.driver-repository.eu/namespace/dri");
nsContext.put("oaf", "http://namespace.openaire.eu/oaf");
nsContext.put("oai", "http://www.openarchives.org/OAI/2.0/");
nsContext.put("prov", "http://www.openarchives.org/OAI/2.0/provenance");
nsContext.put("dc", "http://purl.org/dc/elements/1.1/");
nsContext.put("datacite", "http://datacite.org/schema/kernel-3");
DocumentFactory.getInstance().setXPathNamespaceURIs(nsContext);
final Document doc =
DocumentHelper.parseText(xml.replaceAll("http://datacite.org/schema/kernel-4", "http://datacite.org/schema/kernel-3"));
final String type = doc.valueOf("//dr:CobjCategory/@type");
final KeyValue collectedFrom =
keyValue(createOpenaireId(10, doc.valueOf("//oaf:collectedFrom/@id"), true), doc.valueOf("//oaf:collectedFrom/@name"));
final KeyValue hostedBy =
StringUtils.isBlank(doc.valueOf("//oaf:hostedBy/@id"))
? collectedFrom
: keyValue(createOpenaireId(10, doc.valueOf("//oaf:hostedBy/@id"), true), doc.valueOf("//oaf:hostedBy/@name"));
final DataInfo info = prepareDataInfo(doc);
final long lastUpdateTimestamp = new Date().getTime();
return createOafs(doc, type, collectedFrom, hostedBy, info, lastUpdateTimestamp);
} catch (final Exception e) {
throw new RuntimeException(e);
}
}
protected List<Oaf> createOafs(
final Document doc,
final String type,
final KeyValue collectedFrom,
final KeyValue hostedBy,
final DataInfo info,
final long lastUpdateTimestamp) {
final List<Oaf> oafs = new ArrayList<>();
switch (type.toLowerCase()) {
case "":
case "publication":
final Publication p = new Publication();
populateResultFields(p, doc, collectedFrom, hostedBy, info, lastUpdateTimestamp);
p.setResulttype(MigrationConstants.PUBLICATION_RESULTTYPE_QUALIFIER);
p.setJournal(prepareJournal(doc, info));
oafs.add(p);
break;
case "dataset":
final Dataset d = new Dataset();
populateResultFields(d, doc, collectedFrom, hostedBy, info, lastUpdateTimestamp);
d.setResulttype(MigrationConstants.DATASET_RESULTTYPE_QUALIFIER);
d.setStoragedate(prepareDatasetStorageDate(doc, info));
d.setDevice(prepareDatasetDevice(doc, info));
d.setSize(prepareDatasetSize(doc, info));
d.setVersion(prepareDatasetVersion(doc, info));
d.setLastmetadataupdate(prepareDatasetLastMetadataUpdate(doc, info));
d.setMetadataversionnumber(prepareDatasetMetadataVersionNumber(doc, info));
d.setGeolocation(prepareDatasetGeoLocations(doc, info));
oafs.add(d);
break;
case "software":
final Software s = new Software();
populateResultFields(s, doc, collectedFrom, hostedBy, info, lastUpdateTimestamp);
s.setResulttype(MigrationConstants.SOFTWARE_RESULTTYPE_QUALIFIER);
s.setDocumentationUrl(prepareSoftwareDocumentationUrls(doc, info));
s.setLicense(prepareSoftwareLicenses(doc, info));
s.setCodeRepositoryUrl(prepareSoftwareCodeRepositoryUrl(doc, info));
s.setProgrammingLanguage(prepareSoftwareProgrammingLanguage(doc, info));
oafs.add(s);
break;
case "otherresearchproducts":
default:
final OtherResearchProduct o = new OtherResearchProduct();
populateResultFields(o, doc, collectedFrom, hostedBy, info, lastUpdateTimestamp);
o.setResulttype(MigrationConstants.OTHER_RESULTTYPE_QUALIFIER);
o.setContactperson(prepareOtherResearchProductContactPersons(doc, info));
o.setContactgroup(prepareOtherResearchProductContactGroups(doc, info));
o.setTool(prepareOtherResearchProductTools(doc, info));
oafs.add(o);
break;
}
if (!oafs.isEmpty()) {
oafs.addAll(addProjectRels(doc, collectedFrom, info, lastUpdateTimestamp));
oafs.addAll(addOtherResultRels(doc, collectedFrom, info, lastUpdateTimestamp));
}
return oafs;
}
private List<Oaf> addProjectRels(
final Document doc,
final KeyValue collectedFrom,
final DataInfo info,
final long lastUpdateTimestamp) {
final List<Oaf> res = new ArrayList<>();
final String docId = createOpenaireId(50, doc.valueOf("//dri:objIdentifier"), false);
for (final Object o : doc.selectNodes("//oaf:projectid")) {
final String projectId = createOpenaireId(40, ((Node) o).getText(), true);
final Relation r1 = new Relation();
r1.setRelType("resultProject");
r1.setSubRelType("outcome");
r1.setRelClass("isProducedBy");
r1.setSource(docId);
r1.setTarget(projectId);
r1.setCollectedfrom(Arrays.asList(collectedFrom));
r1.setDataInfo(info);
r1.setLastupdatetimestamp(lastUpdateTimestamp);
res.add(r1);
final Relation r2 = new Relation();
r2.setRelType("resultProject");
r2.setSubRelType("outcome");
r2.setRelClass("produces");
r2.setSource(projectId);
r2.setTarget(docId);
r2.setCollectedfrom(Arrays.asList(collectedFrom));
r2.setDataInfo(info);
r2.setLastupdatetimestamp(lastUpdateTimestamp);
res.add(r2);
}
return res;
}
protected abstract List<Oaf> addOtherResultRels(
final Document doc,
final KeyValue collectedFrom,
final DataInfo info,
final long lastUpdateTimestamp);
private void populateResultFields(
final Result r,
final Document doc,
final KeyValue collectedFrom,
final KeyValue hostedBy,
final DataInfo info,
final long lastUpdateTimestamp) {
r.setDataInfo(info);
r.setLastupdatetimestamp(lastUpdateTimestamp);
r.setId(createOpenaireId(50, doc.valueOf("//dri:objIdentifier"), false));
r.setOriginalId(Arrays.asList(doc.valueOf("//dri:objIdentifier")));
r.setCollectedfrom(Arrays.asList(collectedFrom));
r.setPid(prepareListStructProps(doc, "//oaf:identifier", "@identifierType", "dnet:pid_types", "dnet:pid_types", info));
r.setDateofcollection(doc.valueOf("//dr:dateOfCollection"));
r.setDateoftransformation(doc.valueOf("//dr:dateOfTransformation"));
r.setExtraInfo(new ArrayList<>()); // NOT PRESENT IN MDSTORES
r.setOaiprovenance(prepareOAIprovenance(doc));
r.setAuthor(prepareAuthors(doc, info));
r.setLanguage(prepareLanguages(doc));
r.setCountry(new ArrayList<>()); // NOT PRESENT IN MDSTORES
r.setSubject(prepareSubjects(doc, info));
r.setTitle(prepareTitles(doc, info));
r.setRelevantdate(prepareRelevantDates(doc, info));
r.setDescription(prepareDescriptions(doc, info));
r.setDateofacceptance(prepareField(doc, "//oaf:dateAccepted", info));
r.setPublisher(preparePublisher(doc, info));
r.setEmbargoenddate(prepareField(doc, "//oaf:embargoenddate", info));
r.setSource(prepareSources(doc, info));
r.setFulltext(new ArrayList<>()); // NOT PRESENT IN MDSTORES
r.setFormat(prepareFormats(doc, info));
r.setContributor(prepareContributors(doc, info));
r.setResourcetype(prepareResourceType(doc, info));
r.setCoverage(prepareCoverages(doc, info));
r.setContext(new ArrayList<>()); // NOT PRESENT IN MDSTORES
r.setExternalReference(new ArrayList<>()); // NOT PRESENT IN MDSTORES
r.setInstance(prepareInstances(doc, info, collectedFrom, hostedBy));
}
protected abstract Qualifier prepareResourceType(Document doc, DataInfo info);
protected abstract List<Instance> prepareInstances(
Document doc,
DataInfo info,
KeyValue collectedfrom,
KeyValue hostedby);
protected abstract List<Field<String>> prepareSources(Document doc, DataInfo info);
protected abstract List<StructuredProperty> prepareRelevantDates(Document doc, DataInfo info);
protected abstract List<Field<String>> prepareCoverages(Document doc, DataInfo info);
protected abstract List<Field<String>> prepareContributors(Document doc, DataInfo info);
protected abstract List<Field<String>> prepareFormats(Document doc, DataInfo info);
protected abstract Field<String> preparePublisher(Document doc, DataInfo info);
protected abstract List<Field<String>> prepareDescriptions(Document doc, DataInfo info);
protected abstract List<StructuredProperty> prepareTitles(Document doc, DataInfo info);
protected abstract List<StructuredProperty> prepareSubjects(Document doc, DataInfo info);
protected abstract Qualifier prepareLanguages(Document doc);
protected abstract List<Author> prepareAuthors(Document doc, DataInfo info);
protected abstract List<Field<String>> prepareOtherResearchProductTools(
Document doc,
DataInfo info);
protected abstract List<Field<String>> prepareOtherResearchProductContactGroups(
Document doc,
DataInfo info);
protected abstract List<Field<String>> prepareOtherResearchProductContactPersons(
Document doc,
DataInfo info);
protected abstract Qualifier prepareSoftwareProgrammingLanguage(Document doc, DataInfo info);
protected abstract Field<String> prepareSoftwareCodeRepositoryUrl(Document doc, DataInfo info);
protected abstract List<StructuredProperty> prepareSoftwareLicenses(
Document doc,
DataInfo info);
protected abstract List<Field<String>> prepareSoftwareDocumentationUrls(
Document doc,
DataInfo info);
protected abstract List<GeoLocation> prepareDatasetGeoLocations(Document doc, DataInfo info);
protected abstract Field<String> prepareDatasetMetadataVersionNumber(
Document doc,
DataInfo info);
protected abstract Field<String> prepareDatasetLastMetadataUpdate(Document doc, DataInfo info);
protected abstract Field<String> prepareDatasetVersion(Document doc, DataInfo info);
protected abstract Field<String> prepareDatasetSize(Document doc, DataInfo info);
protected abstract Field<String> prepareDatasetDevice(Document doc, DataInfo info);
protected abstract Field<String> prepareDatasetStorageDate(Document doc, DataInfo info);
private Journal prepareJournal(final Document doc, final DataInfo info) {
final Node n = doc.selectSingleNode("//oaf:journal");
if (n != null) {
final String name = n.getText();
final String issnPrinted = n.valueOf("@issn");
final String issnOnline = n.valueOf("@eissn");
final String issnLinking = n.valueOf("@lissn");
final String ep = n.valueOf("@ep");
final String iss = n.valueOf("@iss");
final String sp = n.valueOf("@sp");
final String vol = n.valueOf("@vol");
final String edition = n.valueOf("@edition");
if (StringUtils.isNotBlank(name)) { return journal(name, issnPrinted, issnOnline, issnLinking, ep, iss, sp, vol, edition, null, null, info); }
}
return null;
}
protected Qualifier prepareQualifier(
final Node node,
final String xpath,
final String schemeId,
final String schemeName) {
final String classId = node.valueOf(xpath);
final String className = code2name.get(classId);
return qualifier(classId, className, schemeId, schemeName);
}
protected List<StructuredProperty> prepareListStructProps(
final Node node,
final String xpath,
final String xpathClassId,
final String schemeId,
final String schemeName,
final DataInfo info) {
final List<StructuredProperty> res = new ArrayList<>();
for (final Object o : node.selectNodes(xpath)) {
final Node n = (Node) o;
final String classId = n.valueOf(xpathClassId);
final String className = code2name.get(classId);
res.add(structuredProperty(n.getText(), classId, className, schemeId, schemeName, info));
}
return res;
}
protected List<StructuredProperty> prepareListStructProps(
final Node node,
final String xpath,
final Qualifier qualifier,
final DataInfo info) {
final List<StructuredProperty> res = new ArrayList<>();
for (final Object o : node.selectNodes(xpath)) {
final Node n = (Node) o;
res.add(structuredProperty(n.getText(), qualifier, info));
}
return res;
}
protected List<StructuredProperty> prepareListStructProps(
final Node node,
final String xpath,
final DataInfo info) {
final List<StructuredProperty> res = new ArrayList<>();
for (final Object o : node.selectNodes(xpath)) {
final Node n = (Node) o;
res.add(structuredProperty(n.getText(), n.valueOf("@classid"), n.valueOf("@classname"), n.valueOf("@schemeid"), n.valueOf("@schemename"), info));
}
return res;
}
protected OAIProvenance prepareOAIprovenance(final Document doc) {
final Node n =
doc.selectSingleNode("//*[local-name()='provenance']/*[local-name()='originDescription']");
if (n == null) { return null; }
final String identifier = n.valueOf("./*[local-name()='identifier']");
final String baseURL = n.valueOf("./*[local-name()='baseURL']");;
final String metadataNamespace = n.valueOf("./*[local-name()='metadataNamespace']");;
final boolean altered = n.valueOf("@altered").equalsIgnoreCase("true");
final String datestamp = n.valueOf("./*[local-name()='datestamp']");;
final String harvestDate = n.valueOf("@harvestDate");;
return oaiIProvenance(identifier, baseURL, metadataNamespace, altered, datestamp, harvestDate);
}
protected DataInfo prepareDataInfo(final Document doc) {
final Node n = doc.selectSingleNode("//oaf:datainfo");
if (n == null) { return dataInfo(false, null, false, false, MigrationConstants.REPOSITORY_PROVENANCE_ACTIONS, "0.9"); }
final String paClassId = n.valueOf("./oaf:provenanceaction/@classid");
final String paClassName = n.valueOf("./oaf:provenanceaction/@classname");
final String paSchemeId = n.valueOf("./oaf:provenanceaction/@schemeid");
final String paSchemeName = n.valueOf("./oaf:provenanceaction/@schemename");
final boolean deletedbyinference =
Boolean.parseBoolean(n.valueOf("./oaf:deletedbyinference"));
final String inferenceprovenance = n.valueOf("./oaf:inferenceprovenance");
final Boolean inferred = Boolean.parseBoolean(n.valueOf("./oaf:inferred"));
final String trust = n.valueOf("./oaf:trust");
return dataInfo(deletedbyinference, inferenceprovenance, inferred, false, qualifier(paClassId, paClassName, paSchemeId, paSchemeName), trust);
}
protected Field<String> prepareField(final Node node, final String xpath, final DataInfo info) {
return field(node.valueOf(xpath), info);
}
protected List<Field<String>> prepareListFields(
final Node node,
final String xpath,
final DataInfo info) {
return listFields(info, prepareListString(node, xpath));
}
protected List<String> prepareListString(final Node node, final String xpath) {
final List<String> res = new ArrayList<>();
for (final Object o : node.selectNodes(xpath)) {
final String s = ((Node) o).getText().trim();
if (StringUtils.isNotBlank(s)) {
res.add(s);
}
}
return res;
}
}

1000
dhp-workflows/dhp-graph-mapper/src/main/java/eu/dnetlib/dhp/oa/graph/raw/MigrateDbEntitiesApplication.java
View File

File diff suppressed because it is too large Load Diff

22
dhp-workflows/dhp-graph-mapper/src/main/java/eu/dnetlib/dhp/oa/graph/raw/common/MigrationConstants.java Normal file
View File

@ -0,0 +1,22 @@
package eu.dnetlib.dhp.oa.graph.raw.common;
import static eu.dnetlib.dhp.oa.graph.raw.common.OafMapperUtils.qualifier;
import eu.dnetlib.dhp.schema.oaf.Qualifier;
public class MigrationConstants {
public static final Qualifier PUBLICATION_RESULTTYPE_QUALIFIER =
qualifier("publication", "publication", "dnet:result_typologies", "dnet:result_typologies");
public static final Qualifier DATASET_RESULTTYPE_QUALIFIER =
qualifier("dataset", "dataset", "dnet:result_typologies", "dnet:result_typologies");
public static final Qualifier SOFTWARE_RESULTTYPE_QUALIFIER =
qualifier("software", "software", "dnet:result_typologies", "dnet:result_typologies");
public static final Qualifier OTHER_RESULTTYPE_QUALIFIER =
qualifier("other", "other", "dnet:result_typologies", "dnet:result_typologies");
public static final Qualifier REPOSITORY_PROVENANCE_ACTIONS =
qualifier("sysimport:crosswalk:repository", "sysimport:crosswalk:repository", "dnet:provenanceActions", "dnet:provenanceActions");
public static final Qualifier ENTITYREGISTRY_PROVENANCE_ACTION =
qualifier("sysimport:crosswalk:entityregistry", "sysimport:crosswalk:entityregistry", "dnet:provenanceActions", "dnet:provenanceActions");
}

39
dhp-workflows/dhp-graph-mapper/src/main/resources/eu/dnetlib/dhp/oa/graph/hive/oozie_app/workflow.xml
View File

@ -21,6 +21,28 @@
<name>sparkExecutorCores</name>
<description>number of cores used by single executor</description>
</property>
<property>
<name>oozieActionShareLibForSpark2</name>
<description>oozie action sharelib for spark 2.*</description>
</property>
<property>
<name>spark2ExtraListeners</name>
<value>com.cloudera.spark.lineage.NavigatorAppListener</value>
<description>spark 2.* extra listeners classname</description>
</property>
<property>
<name>spark2SqlQueryExecutionListeners</name>
<value>com.cloudera.spark.lineage.NavigatorQueryListener</value>
<description>spark 2.* sql query execution listeners classname</description>
</property>
<property>
<name>spark2YarnHistoryServerAddress</name>
<description>spark 2.* yarn history server address</description>
</property>
<property>
<name>spark2EventLogDir</name>
<description>spark 2.* event log dir location</description>
</property>
</parameters>
<global>
@ -35,6 +57,10 @@
<name>oozie.launcher.mapred.job.queue.name</name>
<value>${oozieLauncherQueueName}</value>
</property>
<property>
<name>oozie.action.sharelib.for.spark</name>
<value>${oozieActionShareLibForSpark2}</value>
</property>
</configuration>
</global>
@ -52,14 +78,15 @@
<class>eu.dnetlib.dhp.oa.graph.GraphHiveImporterJob</class>
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
<spark-opts>
--executor-memory ${sparkExecutorMemory}
--executor-cores ${sparkExecutorCores}
--executor-memory=${sparkExecutorMemory}
--executor-cores=${sparkExecutorCores}
--driver-memory=${sparkDriverMemory}
--conf spark.extraListeners="com.cloudera.spark.lineage.NavigatorAppListener"
--conf spark.sql.queryExecutionListeners="com.cloudera.spark.lineage.NavigatorQueryListener"
--conf spark.sql.warehouse.dir="/user/hive/warehouse"
--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>-mt</arg> <arg>yarn-cluster</arg>
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
<arg>--hive_db_name</arg><arg>${hive_db_name}</arg>
<arg>--hive_metastore_uris</arg><arg>${hive_metastore_uris}</arg>

96
dhp-workflows/dhp-graph-provision/src/main/java/eu/dnetlib/dhp/oa/provision/CreateRelatedEntitiesJob_phase1.java
View File

@ -11,11 +11,13 @@ import eu.dnetlib.dhp.oa.provision.model.RelatedEntity;
import eu.dnetlib.dhp.oa.provision.model.SortableRelation;
import eu.dnetlib.dhp.schema.common.EntityType;
import eu.dnetlib.dhp.schema.common.ModelSupport;
import eu.dnetlib.dhp.schema.oaf.OafEntity;
import eu.dnetlib.dhp.schema.oaf.*;
import java.util.List;
import java.util.Objects;
import java.util.Optional;
import java.util.stream.Collectors;
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.Dataset;
import org.apache.spark.sql.Encoders;
@ -104,16 +106,12 @@ public class CreateRelatedEntitiesJob_phase1 {
SparkSession spark,
String inputRelationsPath,
String inputEntityPath,
Class<E> entityClazz,
Class<E> clazz,
String outputPath) {
Dataset<Tuple2<String, SortableRelation>> relsByTarget =
readPathRelation(spark, inputRelationsPath)
.filter(
(FilterFunction<SortableRelation>)
value ->
value.getDataInfo().getDeletedbyinference()
== false)
.filter("dataInfo.deletedbyinference == false")
.map(
(MapFunction<SortableRelation, Tuple2<String, SortableRelation>>)
r -> new Tuple2<>(r.getTarget(), r),
@ -122,10 +120,11 @@ public class CreateRelatedEntitiesJob_phase1 {
.cache();
Dataset<Tuple2<String, RelatedEntity>> entities =
readPathEntity(spark, inputEntityPath, entityClazz)
readPathEntity(spark, inputEntityPath, clazz)
.filter("dataInfo.invisible == false")
.map(
(MapFunction<E, RelatedEntity>)
value -> asRelatedEntity(value, entityClazz),
value -> asRelatedEntity(value, clazz),
Encoders.bean(RelatedEntity.class))
.map(
(MapFunction<RelatedEntity, Tuple2<String, RelatedEntity>>)
@ -146,7 +145,7 @@ public class CreateRelatedEntitiesJob_phase1 {
Encoders.bean(EntityRelEntity.class))
.write()
.mode(SaveMode.Overwrite)
.parquet(outputPath + "/" + EntityType.fromClass(entityClazz));
.parquet(outputPath + "/" + EntityType.fromClass(clazz));
}
private static <E extends OafEntity> Dataset<E> readPathEntity(
@ -161,6 +160,81 @@ public class CreateRelatedEntitiesJob_phase1 {
Encoders.bean(entityClazz));
}
public static <E extends OafEntity> RelatedEntity asRelatedEntity(E entity, Class<E> clazz) {
final RelatedEntity re = new RelatedEntity();
re.setId(entity.getId());
re.setType(EntityType.fromClass(clazz).name());
re.setPid(entity.getPid());
re.setCollectedfrom(entity.getCollectedfrom());
switch (EntityType.fromClass(clazz)) {
case publication:
case dataset:
case otherresearchproduct:
case software:
Result result = (Result) entity;
if (result.getTitle() != null && !result.getTitle().isEmpty()) {
re.setTitle(result.getTitle().stream().findFirst().get());
}
re.setDateofacceptance(getValue(result.getDateofacceptance()));
re.setPublisher(getValue(result.getPublisher()));
re.setResulttype(result.getResulttype());
re.setInstances(result.getInstance());
// TODO still to be mapped
// re.setCodeRepositoryUrl(j.read("$.coderepositoryurl"));
break;
case datasource:
Datasource d = (Datasource) entity;
re.setOfficialname(getValue(d.getOfficialname()));
re.setWebsiteurl(getValue(d.getWebsiteurl()));
re.setDatasourcetype(d.getDatasourcetype());
re.setOpenairecompatibility(d.getOpenairecompatibility());
break;
case organization:
Organization o = (Organization) entity;
re.setLegalname(getValue(o.getLegalname()));
re.setLegalshortname(getValue(o.getLegalshortname()));
re.setCountry(o.getCountry());
re.setWebsiteurl(getValue(o.getWebsiteurl()));
break;
case project:
Project p = (Project) entity;
re.setProjectTitle(getValue(p.getTitle()));
re.setCode(getValue(p.getCode()));
re.setAcronym(getValue(p.getAcronym()));
re.setContracttype(p.getContracttype());
List<Field<String>> f = p.getFundingtree();
if (!f.isEmpty()) {
re.setFundingtree(
f.stream().map(s -> s.getValue()).collect(Collectors.toList()));
}
break;
}
return re;
}
private static String getValue(Field<String> field) {
return getFieldValueWithDefault(field, "");
}
private static <T> T getFieldValueWithDefault(Field<T> f, T defaultValue) {
return Optional.ofNullable(f)
.filter(Objects::nonNull)
.map(x -> x.getValue())
.orElse(defaultValue);
}
/**
* Reads a Dataset of eu.dnetlib.dhp.oa.provision.model.SortableRelation objects from a newline
* delimited json text file,

23
dhp-workflows/dhp-graph-provision/src/main/java/eu/dnetlib/dhp/oa/provision/PrepareRelationsJob.java
View File

@ -76,9 +76,6 @@ public class PrepareRelationsJob {
String outputPath = parser.get("outputPath");
log.info("outputPath: {}", outputPath);
int numPartitions = Integer.parseInt(parser.get("relPartitions"));
log.info("relPartitions: {}", numPartitions);
SparkConf conf = new SparkConf();
runWithSparkSession(
@ -86,27 +83,14 @@ public class PrepareRelationsJob {
isSparkSessionManaged,
spark -> {
removeOutputDir(spark, outputPath);
prepareRelationsFromPaths(spark, inputRelationsPath, outputPath, numPartitions);
prepareRelationsFromPaths(spark, inputRelationsPath, outputPath);
});
}
private static void prepareRelationsFromPaths(
SparkSession spark, String inputRelationsPath, String outputPath, int numPartitions) {
SparkSession spark, String inputRelationsPath, String outputPath) {
readPathRelation(spark, inputRelationsPath)
.filter(
(FilterFunction<SortableRelation>)
r -> {
try {
return r != null
&& r.getDataInfo() != null
&& !r.getDataInfo().getDeletedbyinference();
} catch (NullPointerException e) {
log.info(
"invalid NPE '{}'",
OBJECT_MAPPER.writeValueAsString(r));
throw e;
}
})
.filter("dataInfo.deletedbyinference == false")
.groupByKey(
(MapFunction<SortableRelation, String>) value -> value.getSource(),
Encoders.STRING())
@ -114,7 +98,6 @@ public class PrepareRelationsJob {
(FlatMapGroupsFunction<String, SortableRelation, SortableRelation>)
(key, values) -> Iterators.limit(values, MAX_RELS),
Encoders.bean(SortableRelation.class))
.repartition(numPartitions)
.write()
.mode(SaveMode.Overwrite)
.parquet(outputPath);

81
dhp-workflows/dhp-graph-provision/src/main/java/eu/dnetlib/dhp/oa/provision/utils/GraphMappingUtils.java
View File

@ -3,14 +3,8 @@ package eu.dnetlib.dhp.oa.provision.utils;
import static org.apache.commons.lang3.StringUtils.substringAfter;
import com.google.common.collect.Sets;
import eu.dnetlib.dhp.oa.provision.model.RelatedEntity;
import eu.dnetlib.dhp.schema.common.EntityType;
import eu.dnetlib.dhp.schema.oaf.*;
import java.util.List;
import java.util.Objects;
import java.util.Optional;
import java.util.Set;
import java.util.stream.Collectors;
public class GraphMappingUtils {
@ -18,81 +12,6 @@ public class GraphMappingUtils {
public static Set<String> authorPidTypes = Sets.newHashSet("orcid", "magidentifier");
public static <E extends OafEntity> RelatedEntity asRelatedEntity(E entity, Class<E> clazz) {
final RelatedEntity re = new RelatedEntity();
re.setId(entity.getId());
re.setType(EntityType.fromClass(clazz).name());
re.setPid(entity.getPid());
re.setCollectedfrom(entity.getCollectedfrom());
switch (EntityType.fromClass(clazz)) {
case publication:
case dataset:
case otherresearchproduct:
case software:
Result result = (Result) entity;
if (result.getTitle() == null && !result.getTitle().isEmpty()) {
re.setTitle(result.getTitle().stream().findFirst().get());
}
re.setDateofacceptance(getValue(result.getDateofacceptance()));
re.setPublisher(getValue(result.getPublisher()));
re.setResulttype(result.getResulttype());
re.setInstances(result.getInstance());
// TODO still to be mapped
// re.setCodeRepositoryUrl(j.read("$.coderepositoryurl"));
break;
case datasource:
Datasource d = (Datasource) entity;
re.setOfficialname(getValue(d.getOfficialname()));
re.setWebsiteurl(getValue(d.getWebsiteurl()));
re.setDatasourcetype(d.getDatasourcetype());
re.setOpenairecompatibility(d.getOpenairecompatibility());
break;
case organization:
Organization o = (Organization) entity;
re.setLegalname(getValue(o.getLegalname()));
re.setLegalshortname(getValue(o.getLegalshortname()));
re.setCountry(o.getCountry());
re.setWebsiteurl(getValue(o.getWebsiteurl()));
break;
case project:
Project p = (Project) entity;
re.setProjectTitle(getValue(p.getTitle()));
re.setCode(getValue(p.getCode()));
re.setAcronym(getValue(p.getAcronym()));
re.setContracttype(p.getContracttype());
List<Field<String>> f = p.getFundingtree();
if (!f.isEmpty()) {
re.setFundingtree(
f.stream().map(s -> s.getValue()).collect(Collectors.toList()));
}
break;
}
return re;
}
private static String getValue(Field<String> field) {
return getFieldValueWithDefault(field, "");
}
private static <T> T getFieldValueWithDefault(Field<T> f, T defaultValue) {
return Optional.ofNullable(f)
.filter(Objects::nonNull)
.map(x -> x.getValue())
.orElse(defaultValue);
}
public static String removePrefix(final String s) {
if (s.contains("|")) return substringAfter(s, "|");
return s;

1
dhp-workflows/dhp-graph-provision/src/main/resources/eu/dnetlib/dhp/oa/provision/oozie_app/workflow.xml
View File

@ -98,6 +98,7 @@
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
--conf spark.sql.shuffle.partitions=3840
</spark-opts>
<arg>--inputRelationsPath</arg><arg>${inputGraphRootPath}/relation</arg>
<arg>--outputPath</arg><arg>${workingDir}/relation</arg>

3
dhp-workflows/pom.xml
View File

@ -261,7 +261,8 @@
sparkDriverMemory,sparkExecutorMemory,sparkExecutorCores,
oozie.wf.application.path,projectVersion,oozie.use.system.libpath,
oozieActionShareLibForSpark1,spark1YarnHistoryServerAddress,spark1EventLogDir,
oozieActionShareLibForSpark2,spark2YarnHistoryServerAddress,spark2EventLogDir
oozieActionShareLibForSpark2,spark2YarnHistoryServerAddress,spark2EventLogDir,
sparkSqlWarehouseDir
</include>
<includeSystemProperties>true</includeSystemProperties>
<includePropertyKeysFromFiles>