partial implementation of entities from db

2020-01-20 16:04:19 +01:00 · 2020-01-20 16:04:19 +01:00 · b35c59eb42
parent 81f82b5d34
commit b35c59eb42
7 changed files with 702 additions and 538 deletions
--- a/dhp-build/dhp-build-properties-maven-plugin/pom.xml
+++ b/dhp-build/dhp-build-properties-maven-plugin/pom.xml
@ -76,6 +76,41 @@
                </configuration>
            </plugin>
        </plugins>
+        <pluginManagement>
+        	<plugins>
+        		<!--This plugin's configuration is used to store Eclipse m2e settings only. It has no influence on the Maven build itself.-->
+        		<plugin>
+        			<groupId>org.eclipse.m2e</groupId>
+        			<artifactId>lifecycle-mapping</artifactId>
+        			<version>1.0.0</version>
+        			<configuration>
+        				<lifecycleMappingMetadata>
+        					<pluginExecutions>
+        						<pluginExecution>
+        							<pluginExecutionFilter>
+        								<groupId>
+        									org.apache.maven.plugins
+        								</groupId>
+        								<artifactId>
+        									maven-plugin-plugin
+        								</artifactId>
+        								<versionRange>
+        									[3.2,)
+        								</versionRange>
+        								<goals>
+        									<goal>descriptor</goal>
+        								</goals>
+        							</pluginExecutionFilter>
+        							<action>
+        								<ignore></ignore>
+        							</action>
+        						</pluginExecution>
+        					</pluginExecutions>
+        				</lifecycleMappingMetadata>
+        			</configuration>
+        		</plugin>
+        	</plugins>
+        </pluginManagement>
    </build>

 </project>
--- a/dhp-schemas/src/main/java/eu/dnetlib/dhp/schema/dli/Relation.java
+++ b/dhp-schemas/src/main/java/eu/dnetlib/dhp/schema/dli/Relation.java
@ -5,6 +5,11 @@ import java.util.List;

 public class Relation implements Serializable {

+	/**
+	 *
+	 */
+	private static final long serialVersionUID = -9103706796710618813L;
+
 	private String source;

 	private String target;
@ -17,7 +22,7 @@ public class Relation implements Serializable {
 		return source;
 	}

-    public void setSource(String source) {
+	public void setSource(final String source) {
 		this.source = source;
 	}

@ -25,7 +30,7 @@ public class Relation implements Serializable {
 		return target;
 	}

-    public void setTarget(String target) {
+	public void setTarget(final String target) {
 		this.target = target;
 	}

@ -33,7 +38,7 @@ public class Relation implements Serializable {
 		return provenance;
 	}

-    public void setProvenance(List<Provenance> provenance) {
+	public void setProvenance(final List<Provenance> provenance) {
 		this.provenance = provenance;
 	}

@ -41,7 +46,7 @@ public class Relation implements Serializable {
 		return semantic;
 	}

-    public void setSemantic(RelationSemantic semantic) {
+	public void setSemantic(final RelationSemantic semantic) {
 		this.semantic = semantic;
 	}
 }
--- a/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/migration/MigrateDbEntitiesApplication.java
+++ b/dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/migration/MigrateDbEntitiesApplication.java
@ -2,11 +2,17 @@ package eu.dnetlib.dhp.migration;

 import java.io.Closeable;
 import java.io.IOException;
+import java.sql.Array;
 import java.sql.ResultSet;
+import java.sql.SQLException;
+import java.util.ArrayList;
 import java.util.Arrays;
+import java.util.Date;
+import java.util.List;
 import java.util.function.Consumer;

 import org.apache.commons.io.IOUtils;
+import org.apache.commons.lang3.StringUtils;
 import org.apache.commons.logging.Log;
 import org.apache.commons.logging.LogFactory;

@ -15,14 +21,21 @@ import eu.dnetlib.dhp.schema.oaf.DataInfo;
 import eu.dnetlib.dhp.schema.oaf.Datasource;
 import eu.dnetlib.dhp.schema.oaf.Organization;
 import eu.dnetlib.dhp.schema.oaf.Project;
+import eu.dnetlib.dhp.schema.oaf.Qualifier;
 import eu.dnetlib.dhp.schema.oaf.Relation;
+import eu.dnetlib.dhp.schema.oaf.StructuredProperty;

 public class MigrateDbEntitiesApplication extends AbstractMigrateApplication implements Closeable {

+	private static final Qualifier ENTITYREGISTRY_PROVENANCE_ACTION = MigrationUtils
+			.qualifier("sysimport:crosswalk:entityregistry", "sysimport:crosswalk:entityregistry", "dnet:provenance_actions", "dnet:provenance_actions");
+
 	private static final Log log = LogFactory.getLog(MigrateDbEntitiesApplication.class);

 	private final DbClient dbClient;

+	private final long lastUpdateTimestamp;
+
 	public static void main(final String[] args) throws Exception {
 		final ArgumentApplicationParser parser = new ArgumentApplicationParser(
 				IOUtils.toString(MigrateDbEntitiesApplication.class.getResourceAsStream("/eu/dnetlib/dhp/migration/migrate_db_entities_parameters.json")));
@ -51,6 +64,7 @@ public class MigrateDbEntitiesApplication extends AbstractMigrateApplication imp
 			final String dbPassword) throws Exception {
 		super(hdfsPath, hdfsNameNode, hdfsUser);
 		this.dbClient = new DbClient(dbUrl, dbUser, dbPassword);
+		this.lastUpdateTimestamp = new Date().getTime();
 	}

 	public void execute(final String sqlFile, final Consumer<ResultSet> consumer) throws Exception {
@ -61,7 +75,7 @@ public class MigrateDbEntitiesApplication extends AbstractMigrateApplication imp
 	public void processDatasource(final ResultSet rs) {
 		try {

-			final DataInfo info = MigrationUtils.dataInfo(null, null, null, null, null, null); // TODO
+			final DataInfo info = prepareDataInfo(rs);

 			final Datasource ds = new Datasource();

@ -74,8 +88,8 @@ public class MigrateDbEntitiesApplication extends AbstractMigrateApplication imp
 			ds.setExtraInfo(null); // TODO
 			ds.setOaiprovenance(null); // TODO

-			ds.setDatasourcetype(null); // Qualifier datasourcetype) {
-			ds.setOpenairecompatibility(null); // Qualifier openairecompatibility) {
+			ds.setDatasourcetype(prepareQualifierSplitting(rs.getString("datasourcetype")));
+			ds.setOpenairecompatibility(prepareQualifierSplitting(rs.getString("openairecompatibility")));
 			ds.setOfficialname(MigrationUtils.field(rs.getString("officialname"), info));
 			ds.setEnglishname(MigrationUtils.field(rs.getString("englishname"), info));
 			ds.setWebsiteurl(MigrationUtils.field(rs.getString("websiteurl"), info));
@ -86,7 +100,7 @@ public class MigrateDbEntitiesApplication extends AbstractMigrateApplication imp
 			ds.setLongitude(MigrationUtils.field(Double.toString(rs.getDouble("longitude")), info));
 			ds.setDateofvalidation(MigrationUtils.field(rs.getDate("dateofvalidation").toString(), info));
 			ds.setDescription(MigrationUtils.field(rs.getString("description"), info));
-			ds.setSubjects(null); // List<StructuredProperty> subjects) {
+			ds.setSubjects(prepareListOfStructProps(rs.getArray("subjects"), info));
 			ds.setOdnumberofitems(MigrationUtils.field(Double.toString(rs.getInt("odnumberofitems")), info));
 			ds.setOdnumberofitemsdate(MigrationUtils.field(rs.getDate("odnumberofitemsdate").toString(), info));
 			ds.setOdpolicies(MigrationUtils.field(rs.getString("odpolicies"), info));
@ -110,12 +124,15 @@ public class MigrateDbEntitiesApplication extends AbstractMigrateApplication imp
 			ds.setPolicies(null); // List<KeyValue> // TODO
 			ds.setJournal(null); // Journal // TODO

+			ds.setDataInfo(info);
+			ds.setLastupdatetimestamp(lastUpdateTimestamp);
+
 			// rs.getString("datasourceid");
 			rs.getArray("identities");
 			// rs.getString("officialname");
 			// rs.getString("englishname");
 			// rs.getString("contactemail");
-			rs.getString("openairecompatibility");  // COMPLEX ...@@@...
+			// rs.getString("openairecompatibility"); // COMPLEX ...@@@...
 			// rs.getString("websiteurl");
 			// rs.getString("logourl");
 			// rs.getArray("accessinfopackage");
@ -124,15 +141,15 @@ public class MigrateDbEntitiesApplication extends AbstractMigrateApplication imp
 			// rs.getString("namespaceprefix");
 			// rs.getInt("odnumberofitems"); // NULL
 			// rs.getDate("odnumberofitemsdate"); // NULL
-			rs.getArray("subjects");
+			// rs.getArray("subjects");
 			// rs.getString("description");
 			// rs.getString("odpolicies"); // NULL
 			// rs.getArray("odlanguages");
 			// rs.getArray("odcontenttypes");
-			rs.getBoolean("inferred");  // false
-			rs.getBoolean("deletedbyinference");// false
-			rs.getDouble("trust");  // 0.9
-			rs.getString("inferenceprovenance"); // NULL
+			// rs.getBoolean("inferred"); // false
+			// rs.getBoolean("deletedbyinference");// false
+			// rs.getDouble("trust"); // 0.9
+			// rs.getString("inferenceprovenance"); // NULL
 			// rs.getDate("dateofcollection");
 			// rs.getDate("dateofvalidation");
 			// rs.getDate("releasestartdate");
@ -152,21 +169,22 @@ public class MigrateDbEntitiesApplication extends AbstractMigrateApplication imp
 			rs.getArray("policies");
 			// rs.getString("collectedfromid");
 			// rs.getString("collectedfromname");
-			rs.getString("datasourcetype");  // COMPLEX XXX@@@@....
-			rs.getString("provenanceaction"); // 'sysimport:crosswalk:entityregistry@@@sysimport:crosswalk:entityregistry@@@dnet:provenance_actions@@@dnet:provenance_actions'
+			// rs.getString("datasourcetype"); // COMPLEX XXX@@@@....
+			// rs.getString("provenanceaction"); //
+			// 'sysimport:crosswalk:entityregistry@@@sysimport:crosswalk:entityregistry@@@dnet:provenance_actions@@@dnet:provenance_actions'
 			// AS provenanceaction,
 			rs.getString("journal");  // CONCAT(d.issn, '@@@', d.eissn, '@@@', d.lissn) AS journal

 			emitOaf(ds);
 		} catch (final Exception e) {
-			// TODO: handle exception
+			throw new RuntimeException(e);
 		}
 	}

 	public void processProject(final ResultSet rs) {
 		try {

-			final DataInfo info = MigrationUtils.dataInfo(null, null, null, null, null, null); // TODO
+			final DataInfo info = prepareDataInfo(rs);

 			final Project p = new Project();

@ -192,9 +210,9 @@ public class MigrateDbEntitiesApplication extends AbstractMigrateApplication imp
 			p.setEcsc39(MigrationUtils.field(Boolean.toString(rs.getBoolean("ecsc39")), info));
 			p.setOamandatepublications(MigrationUtils.field(Boolean.toString(rs.getBoolean("oamandatepublications")), info));
 			p.setEcarticle29_3(MigrationUtils.field(Boolean.toString(rs.getBoolean("ecarticle29_3")), info));
-			p.setSubjects(null); // List<StructuredProperty> //TODO
+			p.setSubjects(prepareListOfStructProps(rs.getArray("subjects"), info));
 			p.setFundingtree(null); // List<Field<String>> //TODO
-			p.setContracttype(null); // Qualifier //TODO
+			p.setContracttype(prepareQualifierSplitting(rs.getString("contracttype")));
 			p.setOptional1(MigrationUtils.field(rs.getString("optional1"), info));
 			p.setOptional2(MigrationUtils.field(rs.getString("optional2"), info));
 			p.setJsonextrainfo(MigrationUtils.field(rs.getString("jsonextrainfo"), info));
@ -207,6 +225,9 @@ public class MigrateDbEntitiesApplication extends AbstractMigrateApplication imp
 			p.setTotalcost(new Float(rs.getDouble("totalcost")));
 			p.setFundedamount(new Float(rs.getDouble("fundedamount")));

+			p.setDataInfo(info);
+			p.setLastupdatetimestamp(lastUpdateTimestamp);
+
 			// rs.getString("projectid");
 			// rs.getString("code");
 			// rs.getString("websiteurl");
@ -222,13 +243,13 @@ public class MigrateDbEntitiesApplication extends AbstractMigrateApplication imp
 			// rs.getBoolean("ecarticle29_3");
 			// rs.getDate("dateofcollection");
 			// rs.getDate("dateoftransformation");
-			rs.getBoolean("inferred");
-			rs.getBoolean("deletedbyinference");
-			rs.getDouble("trust");
-			rs.getString("inferenceprovenance");
+			// rs.getBoolean("inferred");
+			// rs.getBoolean("deletedbyinference");
+			// rs.getDouble("trust");
+			// rs.getString("inferenceprovenance");
 			// rs.getString("optional1");
 			// rs.getString("optional2");
-			rs.getString("jsonextrainfo");
+			// rs.getString("jsonextrainfo");
 			// rs.getString("contactfullname");
 			// rs.getString("contactfax");
 			// rs.getString("contactphone");
@ -248,14 +269,14 @@ public class MigrateDbEntitiesApplication extends AbstractMigrateApplication imp
 			emitOaf(p);

 		} catch (final Exception e) {
-			// TODO: handle exception
+			throw new RuntimeException(e);
 		}
 	}

 	public void processOrganization(final ResultSet rs) {
 		try {

-			final DataInfo info = MigrationUtils.dataInfo(null, null, null, null, null, null); // TODO
+			final DataInfo info = prepareDataInfo(rs);

 			final Organization o = new Organization();

@ -269,7 +290,7 @@ public class MigrateDbEntitiesApplication extends AbstractMigrateApplication imp
 			o.setOaiprovenance(null); // OAIProvenance // TODO
 			o.setLegalshortname(MigrationUtils.field("legalshortname", info));
 			o.setLegalname(MigrationUtils.field("legalname", info));
-			o.setAlternativeNames(null); // List<Field<String>> //TODO
+			o.setAlternativeNames(new ArrayList<>());
 			o.setWebsiteurl(MigrationUtils.field("websiteurl", info));
 			o.setLogourl(MigrationUtils.field("logourl", info));
 			o.setEclegalbody(MigrationUtils.field(Boolean.toString(rs.getBoolean("eclegalbody")), info));
@ -283,7 +304,10 @@ public class MigrateDbEntitiesApplication extends AbstractMigrateApplication imp
 			o.setEcenterprise(MigrationUtils.field(Boolean.toString(rs.getBoolean("ecenterprise")), info));
 			o.setEcsmevalidated(MigrationUtils.field(Boolean.toString(rs.getBoolean("ecsmevalidated")), info));
 			o.setEcnutscode(MigrationUtils.field(Boolean.toString(rs.getBoolean("ecnutscode")), info));
-			o.setCountry(null); // Qualifier country) {
+			o.setCountry(prepareQualifierSplitting(rs.getString("country")));
+
+			o.setDataInfo(info);
+			o.setLastupdatetimestamp(lastUpdateTimestamp);

 			// rs.getString("organizationid");
 			// rs.getString("legalshortname");
@ -300,87 +324,160 @@ public class MigrateDbEntitiesApplication extends AbstractMigrateApplication imp
 			// rs.getBoolean("ecenterprise");
 			// rs.getBoolean("ecsmevalidated");
 			// rs.getBoolean("ecnutscode");
-			rs.getDate("dateofcollection");
-			rs.getDate("dateoftransformation");
-			rs.getBoolean("inferred");
-			rs.getBoolean("deletedbyinference");
-			rs.getDouble("trust");
-			rs.getString("inferenceprovenance");
+			// rs.getDate("dateofcollection");
+			// rs.getDate("dateoftransformation");
+			// rs.getBoolean("inferred");
+			// rs.getBoolean("deletedbyinference");
+			// rs.getDouble("trust");
+			// rs.getString("inferenceprovenance");
 			// rs.getString("collectedfromid");
 			// rs.getString("collectedfromname");
-			rs.getString("country");
+			// rs.getString("country");
 			rs.getString("provenanceaction");
 			rs.getArray("pid");

 			emitOaf(o);
 		} catch (final Exception e) {
-			// TODO: handle exception
+			throw new RuntimeException(e);
 		}
 	}

 	public void processDatasourceOrganization(final ResultSet rs) {

 		try {
-			final Relation r = new Relation();
+			final DataInfo info = prepareDataInfo(rs);
+			final String orgId = MigrationUtils.createOpenaireId("20", rs.getString("organization"));
+			final String dsId = MigrationUtils.createOpenaireId("10", rs.getString("datasource"));

-			r.setRelType(null); // TODO
-			r.setSubRelType(null); // TODO
-			r.setRelClass(null); // TODO
-			r.setSource(null); // TODO
-			r.setTarget(null); // TODO
-			r.setCollectedFrom(MigrationUtils.listKeyValues("", ""));
+			final Relation r1 = new Relation();
+			r1.setRelType("datasourceOrganization");
+			r1.setSubRelType("provision");
+			r1.setRelClass("isProvidedBy");
+			r1.setSource(dsId);
+			r1.setTarget(orgId);
+			r1.setCollectedFrom(null);// TODO
+			r1.setDataInfo(info);
+			r1.setLastupdatetimestamp(lastUpdateTimestamp);
+			emitOaf(r1);

-			rs.getString("datasource");
-			rs.getString("organization");
-			rs.getDate("startdate");  // NULL
-			rs.getDate("enddate");    // NULL
-			rs.getBoolean("inferred"); // false
-			rs.getBoolean("deletedbyinference"); // false
-			rs.getDouble("trust");                // 0.9
-			rs.getString("inferenceprovenance"); // NULL
-			rs.getString("semantics");  // 'providedBy@@@provided
+			final Relation r2 = new Relation();
+			r2.setRelType("datasourceOrganization");
+			r2.setSubRelType("provision");
+			r2.setRelClass("provides");
+			r2.setSource(orgId);
+			r2.setTarget(dsId);
+			r2.setCollectedFrom(null);  // TODO
+			r2.setDataInfo(info);
+			r1.setLastupdatetimestamp(lastUpdateTimestamp);
+			emitOaf(r2);
+
+			// rs.getString("datasource");
+			// rs.getString("organization");
+			// rs.getDate("startdate"); // NULL
+			// rs.getDate("enddate"); // NULL
+			// rs.getBoolean("inferred"); // false
+			// rs.getBoolean("deletedbyinference"); // false
+			// rs.getDouble("trust"); // 0.9
+			// rs.getString("inferenceprovenance"); // NULL
+			// rs.getString("semantics"); // 'providedBy@@@provided
 			// by@@@dnet:datasources_organizations_typologies@@@dnet:datasources_organizations_typologies' AS
 			// semantics,
-			rs.getString("provenanceaction"); // d.provenanceaction || '@@@' || d.provenanceaction ||
+			// rs.getString("provenanceaction"); // d.provenanceaction || '@@@' || d.provenanceaction ||
 			// '@@@dnet:provenanceActions@@@dnet:provenanceActions' AS provenanceaction

-			emitOaf(r);
 		} catch (final Exception e) {
-			// TODO: handle exception
+			throw new RuntimeException(e);
 		}
 	}

 	public void processProjectOrganization(final ResultSet rs) {
 		try {
-			final Relation r = new Relation();
+			final DataInfo info = prepareDataInfo(rs);
+			final String orgId = MigrationUtils.createOpenaireId("20", rs.getString("resporganization"));
+			final String projectId = MigrationUtils.createOpenaireId("40", rs.getString("project"));

-			r.setRelType(null); // TODO
-			r.setSubRelType(null); // TODO
-			r.setRelClass(null); // TODO
-			r.setSource(null); // TODO
-			r.setTarget(null); // TODO
-			r.setCollectedFrom(null);
+			final Relation r1 = new Relation();
+			r1.setRelType("projectOrganization");
+			r1.setSubRelType("participation");
+			r1.setRelClass("isParticipant");
+			r1.setSource(projectId);
+			r1.setTarget(orgId);
+			r1.setCollectedFrom(null);// TODO
+			r1.setDataInfo(info);
+			r1.setLastupdatetimestamp(lastUpdateTimestamp);
+			emitOaf(r1);

-			rs.getString("project");
-			rs.getString("resporganization");
-			rs.getInt("participantnumber");
-			rs.getDouble("contribution");
-			rs.getDate("startdate");// null
-			rs.getDate("enddate");// null
-			rs.getBoolean("inferred");// false
-			rs.getBoolean("deletedbyinference"); // false
-			rs.getDouble("trust");
-			rs.getString("inferenceprovenance"); // NULL
-			rs.getString("semantics"); // po.semanticclass || '@@@' || po.semanticclass ||
+			final Relation r2 = new Relation();
+			r2.setRelType("projectOrganization");
+			r2.setSubRelType("participation");
+			r2.setRelClass("hasParticipant");
+			r2.setSource(orgId);
+			r2.setTarget(projectId);
+			r2.setCollectedFrom(null);  // TODO
+			r2.setDataInfo(info);
+			r1.setLastupdatetimestamp(lastUpdateTimestamp);
+			emitOaf(r2);
+
+			// rs.getString("project");
+			// rs.getString("resporganization");
+			// rs.getInt("participantnumber");
+			// rs.getDouble("contribution");
+			// rs.getDate("startdate");// null
+			// rs.getDate("enddate");// null
+			// rs.getBoolean("inferred");// false
+			// rs.getBoolean("deletedbyinference"); // false
+			// rs.getDouble("trust");
+			// rs.getString("inferenceprovenance"); // NULL
+			// rs.getString("semantics"); // po.semanticclass || '@@@' || po.semanticclass ||
 			// '@@@dnet:project_organization_relations@@@dnet:project_organization_relations' AS semantics,
-			rs.getString("provenanceaction"); // 'sysimport:crosswalk:entityregistry@@@sysimport:crosswalk:entityregistry@@@dnet:provenance_actions@@@dnet:provenance_actions'
+			// rs.getString("provenanceaction"); //
+			// 'sysimport:crosswalk:entityregistry@@@sysimport:crosswalk:entityregistry@@@dnet:provenance_actions@@@dnet:provenance_actions'
 			// AS provenanceaction
-			emitOaf(r);
+
 		} catch (final Exception e) {
-			// TODO: handle exception
+			throw new RuntimeException(e);
 		}
 	}

+	private DataInfo prepareDataInfo(final ResultSet rs) throws SQLException {
+		final Boolean deletedbyinference = rs.getBoolean("deletedbyinference");
+		final String inferenceprovenance = rs.getString("inferenceprovenance");
+		final Boolean inferred = rs.getBoolean("inferred");
+		final String trust = rs.getString("trust");
+		return MigrationUtils.dataInfo(deletedbyinference, inferenceprovenance, inferred, false, ENTITYREGISTRY_PROVENANCE_ACTION, trust);
+	}
+
+	private Qualifier prepareQualifierSplitting(final String s) {
+		if (StringUtils.isBlank(s)) { return null; }
+		final String[] arr = s.split("@@@");
+		return arr.length == 4 ? MigrationUtils.qualifier(arr[0], arr[1], arr[2], arr[3]) : null;
+	}
+
+	private StructuredProperty prepareStructProp(final String s, final DataInfo dataInfo) {
+		if (StringUtils.isBlank(s)) { return null; }
+		final String[] parts = s.split("###");
+		if (parts.length == 2) {
+			final String value = parts[0];
+			final String[] arr = parts[1].split("@@@");
+			if (arr.length == 4) { return MigrationUtils.structuredProperty(value, arr[0], arr[1], arr[2], arr[3], dataInfo); }
+		}
+		return null;
+	}
+
+	private List<StructuredProperty> prepareListOfStructProps(final Array array, final DataInfo dataInfo) throws SQLException {
+		final List<StructuredProperty> res = new ArrayList<>();
+		if (array != null) {
+			for (final String s : (String[]) array.getArray()) {
+				final StructuredProperty sp = prepareStructProp(s, dataInfo);
+				if (sp != null) {
+					res.add(sp);
+				}
+			}
+		}
+
+		return res;
+	}
+
 	@Override
 	public void close() throws IOException {
 		super.close();
--- a/dhp-workflows/dhp-dedup/pom.xml
+++ b/dhp-workflows/dhp-dedup/pom.xml
@ -10,6 +10,37 @@

    <artifactId>dhp-dedup</artifactId>
    
+    <build>
+    	<plugins>
+    	     <plugin>
+                <groupId>net.alchim31.maven</groupId>
+                <artifactId>scala-maven-plugin</artifactId>
+                <version>4.0.1</version>
+                <executions>
+                    <execution>
+                        <id>scala-compile-first</id>
+                        <phase>initialize</phase>
+                        <goals>
+                            <goal>add-source</goal>
+                            <goal>compile</goal>
+                        </goals>
+                    </execution>
+                    <execution>
+                        <id>scala-test-compile</id>
+                        <phase>process-test-resources</phase>
+                        <goals>
+                            <goal>testCompile</goal>
+                        </goals>
+                    </execution>
+                </executions>
+                <configuration>
+                    <scalaVersion>${scala.version}</scalaVersion>
+                </configuration>
+            </plugin>
+    	</plugins>
+    
+    </build>
+
    <dependencies>

        <dependency>
--- a/dhp-workflows/dhp-graph-mapper/src/main/java/eu/dnetlib/dhp/graph/GraphMappingUtils.java
+++ b/dhp-workflows/dhp-graph-mapper/src/main/java/eu/dnetlib/dhp/graph/GraphMappingUtils.java
@ -1,10 +1,18 @@
 package eu.dnetlib.dhp.graph;

-import com.google.common.collect.Maps;
-import eu.dnetlib.dhp.schema.oaf.*;
-
 import java.util.Map;

+import com.google.common.collect.Maps;
+
+import eu.dnetlib.dhp.schema.oaf.Dataset;
+import eu.dnetlib.dhp.schema.oaf.Datasource;
+import eu.dnetlib.dhp.schema.oaf.Organization;
+import eu.dnetlib.dhp.schema.oaf.OtherResearchProduct;
+import eu.dnetlib.dhp.schema.oaf.Project;
+import eu.dnetlib.dhp.schema.oaf.Publication;
+import eu.dnetlib.dhp.schema.oaf.Relation;
+import eu.dnetlib.dhp.schema.oaf.Software;
+
 public class GraphMappingUtils {

 	public final static Map<String, Class> types = Maps.newHashMap();
--- a/dhp-workflows/dhp-graph-mapper/src/main/java/eu/dnetlib/dhp/graph/SparkGraphImporterJob.java
+++ b/dhp-workflows/dhp-graph-mapper/src/main/java/eu/dnetlib/dhp/graph/SparkGraphImporterJob.java
@ -1,7 +1,5 @@
 package eu.dnetlib.dhp.graph;

-import com.fasterxml.jackson.databind.ObjectMapper;
-import eu.dnetlib.dhp.application.ArgumentApplicationParser;
 import org.apache.commons.io.IOUtils;
 import org.apache.hadoop.io.Text;
 import org.apache.spark.api.java.JavaRDD;
@ -9,13 +7,18 @@ import org.apache.spark.api.java.JavaSparkContext;
 import org.apache.spark.sql.Encoders;
 import org.apache.spark.sql.SaveMode;
 import org.apache.spark.sql.SparkSession;
+
+import com.fasterxml.jackson.databind.ObjectMapper;
+
+import eu.dnetlib.dhp.application.ArgumentApplicationParser;
 import scala.Tuple2;

 public class SparkGraphImporterJob {

-    public static void main(String[] args) throws Exception {
+	public static void main(final String[] args) throws Exception {

-        final ArgumentApplicationParser parser = new ArgumentApplicationParser(IOUtils.toString(SparkGraphImporterJob.class.getResourceAsStream("/eu/dnetlib/dhp/graph/input_graph_parameters.json")));
+		final ArgumentApplicationParser parser = new ArgumentApplicationParser(
+				IOUtils.toString(SparkGraphImporterJob.class.getResourceAsStream("/eu/dnetlib/dhp/graph/input_graph_parameters.json")));
 		parser.parseArgument(args);
 		final SparkSession spark = SparkSession
 				.builder()
--- a/pom.xml
+++ b/pom.xml
@ -1,4 +1,6 @@
-<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/maven-v4_0_0.xsd">
+<project xmlns="http://maven.apache.org/POM/4.0.0"
+	xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+	xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/maven-v4_0_0.xsd">

 	<modelVersion>4.0.0</modelVersion>
 	<groupId>eu.dnetlib.dhp</groupId>
@ -232,6 +234,12 @@
 				<version>1.0.0</version>
 			</dependency>

+			<dependency>
+				<groupId>org.mongodb</groupId>
+				<artifactId>mongo-java-driver</artifactId>
+				<version>${mongodb.driver.version}</version>
+			</dependency>
+
 			<dependency>
 				<groupId>org.apache.oozie</groupId>
 				<artifactId>oozie-client</artifactId>
@ -349,31 +357,7 @@
 					</execution>
 				</executions>
 			</plugin>
-            <plugin>
-                <groupId>net.alchim31.maven</groupId>
-                <artifactId>scala-maven-plugin</artifactId>
-                <version>4.0.1</version>
-                <executions>
-                    <execution>
-                        <id>scala-compile-first</id>
-                        <phase>initialize</phase>
-                        <goals>
-                            <goal>add-source</goal>
-                            <goal>compile</goal>
-                        </goals>
-                    </execution>
-                    <execution>
-                        <id>scala-test-compile</id>
-                        <phase>process-test-resources</phase>
-                        <goals>
-                            <goal>testCompile</goal>
-                        </goals>
-                    </execution>
-                </executions>
-                <configuration>
-                    <scalaVersion>${scala.version}</scalaVersion>
-                </configuration>
-            </plugin>
+
 		</plugins>

 		<extensions>
@ -421,6 +405,7 @@
 		<dhp.jackson.version>2.9.6</dhp.jackson.version>
 		<dhp.commons.lang.version>3.5</dhp.commons.lang.version>
 		<scala.version>2.11.12</scala.version>
+		<mongodb.driver.version>3.4.2</mongodb.driver.version>
 	</properties>
 </project>