WIP: graph cleaner implementation

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

@@ -61,7 +61,7 @@ public class Vocabulary implements Serializable {
 	}
 
 	public VocabularyTerm getTermBySynonym(final String syn) {
-		return getTerm(synonyms.get(syn));
+		return getTerm(synonyms.get(syn.toLowerCase()));
 	}
 
 	public Qualifier getTermAsQualifier(final String termId) {

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

@@ -3,7 +3,6 @@ package eu.dnetlib.dhp.oa.graph.raw.common;
 
 import java.io.Serializable;
 import java.util.*;
-import java.util.function.Supplier;
 import java.util.stream.Collectors;
 
 import org.apache.commons.lang3.StringUtils;
@@ -14,7 +13,7 @@ import eu.dnetlib.enabling.is.lookup.rmi.ISLookUpService;
 
 public class VocabularyGroup implements Serializable {
 
-	public static final String VOCABULARIES_XQUERY = "for $x in collection(' /db/DRIVER/VocabularyDSResources/VocabularyDSResourceType') \n"
+	public static final String VOCABULARIES_XQUERY = "for $x in collection('/db/DRIVER/VocabularyDSResources/VocabularyDSResourceType') \n"
 		+
 		"let $vocid := $x//VOCABULARY_NAME/@code\n" +
 		"let $vocname := $x//VOCABULARY_NAME/text()\n" +
@@ -46,7 +45,7 @@ public class VocabularyGroup implements Serializable {
 				}
 
 				vocs.addTerm(vocId, termId, termName);
-				vocs.addSynonyms(vocId, termId, termId);
+				// vocs.addSynonyms(vocId, termId, termId);
 			}
 		}
 
@@ -58,7 +57,7 @@ public class VocabularyGroup implements Serializable {
 				final String syn = arr[2].trim();
 
 				vocs.addSynonyms(vocId, termId, syn);
-				vocs.addSynonyms(vocId, termId, termId);
+				// vocs.addSynonyms(vocId, termId, termId);
 			}
 		}
 
@@ -135,7 +134,7 @@ public class VocabularyGroup implements Serializable {
 		Optional
 			.ofNullable(vocs.get(id))
 			.orElseThrow(() -> new IllegalArgumentException("missing vocabulary id: " + vocId))
-			.addSynonym(syn, termId);
+			.addSynonym(syn.toLowerCase(), termId);
 	}
 
 }

dhp-workflows/dhp-graph-mapper/src/test/java/eu/dnetlib/dhp/oa/graph/clean/CleaningFunctionTest.java

@@ -60,8 +60,8 @@ public class CleaningFunctionTest {
 
 		assertNotNull(p_out);
 
-		assertEquals("eng", p_out.getLanguage().getClassid());
-		assertEquals("English", p_out.getLanguage().getClassname());
+		assertEquals("und", p_out.getLanguage().getClassid());
+		assertEquals("Undetermined", p_out.getLanguage().getClassname());
 
 		assertEquals("0018", p_out.getInstance().get(0).getInstancetype().getClassid());
 		assertEquals("Annotation", p_out.getInstance().get(0).getInstancetype().getClassname());

dhp-workflows/dhp-graph-mapper/src/test/resources/eu/dnetlib/dhp/oa/graph/clean/result.json

@@ -281,8 +281,8 @@
         "value": "VIRTA"
       },
       "instancetype": {
-        "classid": "Comentario",
-        "classname": "Comentario",
+        "classid": "Comment/debate",
+        "classname": "Comment/debate",
         "schemeid": "dnet:publication_resource",
         "schemename": "dnet:publication_resource"
       },
@@ -317,8 +317,8 @@
     "vol": ""
   },
   "language": {
-    "classid": "en",
-    "classname": "en",
+    "classid": "UNKNOWN",
+    "classname": "UNKNOWN",
     "schemeid": "dnet:languages",
     "schemename": "dnet:languages"
   },