Add bootstrapped GND labels to ID fields during conversion

See #24

app/apps/Convert.java

@@ -10,6 +10,8 @@
 import java.io.Reader;
 import java.io.StringReader;
 import java.io.StringWriter;
+import java.net.InetAddress;
+import java.net.UnknownHostException;
 import java.nio.charset.StandardCharsets;
 import java.nio.file.Files;
 import java.nio.file.Paths;
@@ -35,6 +37,11 @@
 import org.culturegraph.mf.framework.ObjectReceiver;
 import org.culturegraph.mf.framework.helpers.DefaultObjectPipe;
 import org.culturegraph.mf.framework.helpers.DefaultStreamPipe;
+import org.elasticsearch.action.get.GetResponse;
+import org.elasticsearch.client.transport.TransportClient;
+import org.elasticsearch.common.settings.Settings;
+import org.elasticsearch.common.transport.InetSocketTransportAddress;
+import org.elasticsearch.transport.client.PreBuiltTransportClient;
 import org.xml.sax.InputSource;
 import org.xml.sax.SAXException;
 
@@ -49,6 +56,8 @@
 import com.typesafe.config.ConfigObject;
 
 import ORG.oclc.oai.harvester2.app.RawWrite;
+import controllers.HomeController;
+import models.AuthorityResource;
 import models.GndOntology;
 import play.Logger;
 import play.libs.Json;
@@ -57,12 +66,29 @@ public class Convert {
 
 	private static final Config CONFIG = ConfigFactory.parseFile(new File("conf/application.conf"));
 
+	private static final Settings SETTINGS = Settings.builder()
+			.put("cluster.name", HomeController.config("index.cluster")).build();
+
+	private static final TransportClient CLIENT = new PreBuiltTransportClient(SETTINGS);
+
+	static {
+		CONFIG.getStringList("index.hosts").forEach((host) -> {
+			try {
+				CLIENT.addTransportAddress(new InetSocketTransportAddress(InetAddress.getByName(host), 9300));
+			} catch (UnknownHostException e) {
+				e.printStackTrace();
+			}
+		});
+	}
+
 	static String config(String id) {
 		return CONFIG.getString(id);
 	}
 
 	static final Map<String, Object> context = load();
 
+	private static final Map<String, String> labelCache = new HashMap<>();
+
 	static class OpenOaiPmh extends DefaultObjectPipe<String, ObjectReceiver<Reader>> {
 
 		ByteArrayOutputStream stream = new ByteArrayOutputStream();
@@ -173,7 +199,8 @@ private static Model preprocess(Model model, String id) {
 				// See https://github.com/hbz/lobid-gnd/issues/85
 				// See https://github.com/hbz/lobid-gnd/issues/24
 				String localVocab = "http://d-nb.info/standards/";
-				String object = o.toString().startsWith(localVocab) ? GndOntology.label(o.toString()) : o.toString();
+				String object = o.toString().startsWith(localVocab) ? GndOntology.label(o.toString())
+						: boostrapLabel(o.toString());
 				Statement labelStatement = model.createLiteralStatement(model.createResource(o.toString()),
 						model.createProperty(label), object);
 				toAdd.add(labelStatement);
@@ -217,6 +244,22 @@ private static Model preprocess(Model model, String id) {
 		return model;
 	}
 
+	private static String boostrapLabel(String fullId) {
+		if (!fullId.startsWith(AuthorityResource.DNB_PREFIX)) {
+			return fullId;
+		}
+		String id = fullId.substring(AuthorityResource.DNB_PREFIX.length());
+		return labelCache.containsKey(id) ? labelCache.get(id) : getLabelFromIndex(id, "preferredName");
+	}
+
+	private static String getLabelFromIndex(String id, String field) {
+		GetResponse response = CLIENT.prepareGet(config("index.name"), config("index.type"), id).execute().actionGet();
+		if (response.isExists()) {
+			return response.getSourceAsMap().get(field).toString();
+		}
+		return id;
+	}
+
 	private static String secondLevelTypeFor(String gnd, String type) {
 		String key = type.substring(gnd.length());
 		ConfigObject object = CONFIG.getObject("types");

test/apps/ConvertTest.java

@@ -133,6 +133,24 @@ public void testIriFieldStructure() throws FileNotFoundException {
 		assertTrue(!label.toString().isEmpty());
 	}
 
+	@Test
+	public void testOntologyLabelEnrichment() throws FileNotFoundException {
+		String jsonLd = jsonLdFor("118624822");
+		JsonNode area = Json.parse(jsonLd).get("geographicAreaCode").elements().next();
+		assertIsObjectWithIdAndLabel(area);
+		assertEquals("http://d-nb.info/standards/vocab/gnd/geographic-area-code#XD-US", area.get("id").textValue());
+		assertEquals("USA", area.get("label").textValue());
+	}
+
+	@Test
+	public void testGndLabelEnrichment() throws FileNotFoundException {
+		String jsonLd = jsonLdFor("1081942517");
+		JsonNode author = Json.parse(jsonLd).get("firstAuthor").elements().next();
+		assertIsObjectWithIdAndLabel(author);
+		assertEquals("http://d-nb.info/gnd/118624822", author.get("id").textValue());
+		assertEquals("Twain, Mark", author.get("label").textValue());
+	}
+
 	@Test
 	public void testTriplesToFramedJsonLd() throws FileNotFoundException {
 		Model model = ModelFactory.createDefaultModel();