[graph raw] let the mapping exceptions propagate

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

@@ -28,10 +28,7 @@ import java.util.Optional;
 import java.util.Set;
 
 import org.apache.commons.lang3.StringUtils;
-import org.dom4j.Document;
-import org.dom4j.DocumentFactory;
-import org.dom4j.DocumentHelper;
-import org.dom4j.Node;
+import org.dom4j.*;
 
 import com.google.common.collect.Lists;
 import com.google.common.collect.Sets;
@@ -112,43 +109,40 @@ public abstract class AbstractMdRecordToOafMapper {
 		this.forceOriginalId = false;
 	}
 
-	public List<Oaf> processMdRecord(final String xml) {
-		try {
-			DocumentFactory.getInstance().setXPathNamespaceURIs(nsContext);
+	public List<Oaf> processMdRecord(final String xml) throws DocumentException {
 
-			final Document doc = DocumentHelper
-				.parseText(
-					xml
-						.replaceAll(DATACITE_SCHEMA_KERNEL_4, DATACITE_SCHEMA_KERNEL_3)
-						.replaceAll(DATACITE_SCHEMA_KERNEL_4_SLASH, DATACITE_SCHEMA_KERNEL_3)
-						.replaceAll(DATACITE_SCHEMA_KERNEL_3_SLASH, DATACITE_SCHEMA_KERNEL_3));
+		DocumentFactory.getInstance().setXPathNamespaceURIs(nsContext);
 
-			final KeyValue collectedFrom = getProvenanceDatasource(
-				doc, "//oaf:collectedFrom/@id", "//oaf:collectedFrom/@name");
+		final Document doc = DocumentHelper
+			.parseText(
+				xml
+					.replaceAll(DATACITE_SCHEMA_KERNEL_4, DATACITE_SCHEMA_KERNEL_3)
+					.replaceAll(DATACITE_SCHEMA_KERNEL_4_SLASH, DATACITE_SCHEMA_KERNEL_3)
+					.replaceAll(DATACITE_SCHEMA_KERNEL_3_SLASH, DATACITE_SCHEMA_KERNEL_3));
 
-			if (collectedFrom == null) {
-				return Lists.newArrayList();
-			}
+		final KeyValue collectedFrom = getProvenanceDatasource(
+			doc, "//oaf:collectedFrom/@id", "//oaf:collectedFrom/@name");
 
-			final KeyValue hostedBy = StringUtils.isBlank(doc.valueOf("//oaf:hostedBy/@id"))
-				? collectedFrom
-				: getProvenanceDatasource(doc, "//oaf:hostedBy/@id", "//oaf:hostedBy/@name");
-
-			if (hostedBy == null) {
-				return Lists.newArrayList();
-			}
-
-			final DataInfo info = prepareDataInfo(doc, invisible);
-			final long lastUpdateTimestamp = new Date().getTime();
-
-			final List<Instance> instances = prepareInstances(doc, info, collectedFrom, hostedBy);
-
-			final String type = getResultType(doc, instances);
-
-			return createOafs(doc, type, instances, collectedFrom, info, lastUpdateTimestamp);
-		} catch (final Exception e) {
-			throw new RuntimeException(e);
+		if (collectedFrom == null) {
+			return Lists.newArrayList();
 		}
+
+		final KeyValue hostedBy = StringUtils.isBlank(doc.valueOf("//oaf:hostedBy/@id"))
+			? collectedFrom
+			: getProvenanceDatasource(doc, "//oaf:hostedBy/@id", "//oaf:hostedBy/@name");
+
+		if (hostedBy == null) {
+			return Lists.newArrayList();
+		}
+
+		final DataInfo info = prepareDataInfo(doc, invisible);
+		final long lastUpdateTimestamp = new Date().getTime();
+
+		final List<Instance> instances = prepareInstances(doc, info, collectedFrom, hostedBy);
+
+		final String type = getResultType(doc, instances);
+
+		return createOafs(doc, type, instances, collectedFrom, info, lastUpdateTimestamp);
 	}
 
 	protected String getResultType(final Document doc, final List<Instance> instances) {
@@ -499,22 +493,6 @@ public abstract class AbstractMdRecordToOafMapper {
 		return vocs.getTermAsQualifier(schemeId, classId);
 	}
 
-	protected List<StructuredProperty> prepareListStructProps(
-		final Node node,
-		final String xpath,
-		final String xpathClassId,
-		final String schemeId,
-		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).trim();
-			res.add(structuredProperty(n.getText(), prepareQualifier(classId, schemeId), info));
-		}
-		return res;
-	}
-
 	protected List<StructuredProperty> prepareListStructPropsWithValidQualifier(
 		final Node node,
 		final String xpath,

dhp-workflows/dhp-graph-mapper/src/main/java/eu/dnetlib/dhp/oa/graph/raw/GenerateEntitiesApplication.java

@@ -17,6 +17,7 @@ import org.apache.spark.SparkConf;
 import org.apache.spark.api.java.JavaRDD;
 import org.apache.spark.api.java.JavaSparkContext;
 import org.apache.spark.sql.SparkSession;
+import org.dom4j.DocumentException;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
@@ -158,7 +159,7 @@ public class GenerateEntitiesApplication {
 		final String id,
 		final String s,
 		final boolean shouldHashId,
-		final VocabularyGroup vocs) {
+		final VocabularyGroup vocs) throws DocumentException {
 		final String type = StringUtils.substringAfter(id, ":");
 
 		switch (type.toLowerCase()) {

dhp-workflows/dhp-graph-mapper/src/test/java/eu/dnetlib/dhp/oa/graph/raw/GenerateEntitiesApplicationTest.java

@@ -9,6 +9,7 @@ import java.io.IOException;
 import java.util.List;
 
 import org.apache.commons.io.IOUtils;
+import org.dom4j.DocumentException;
 import org.junit.jupiter.api.BeforeEach;
 import org.junit.jupiter.api.Test;
 import org.junit.jupiter.api.extension.ExtendWith;
@@ -44,7 +45,7 @@ class GenerateEntitiesApplicationTest {
 	}
 
 	@Test
-	void testMergeResult() throws IOException {
+	void testMergeResult() throws IOException, DocumentException {
 		Result publication = getResult("oaf_record.xml", Publication.class);
 		Result dataset = getResult("odf_dataset.xml", Dataset.class);
 		Result software = getResult("odf_software.xml", Software.class);
@@ -76,7 +77,8 @@ class GenerateEntitiesApplicationTest {
 		assertEquals(resultType, merge.getResulttype().getClassid());
 	}
 
-	protected <T extends Result> Result getResult(String xmlFileName, Class<T> clazz) throws IOException {
+	protected <T extends Result> Result getResult(String xmlFileName, Class<T> clazz)
+		throws IOException, DocumentException {
 		final String xml = IOUtils.toString(getClass().getResourceAsStream(xmlFileName));
 		return new OdfToOafMapper(vocs, false, true)
 			.processMdRecord(xml)