Merge branch '24-robusteres-hinzufugen-der-lizenzinformationen' into 'main'

Resolve "Robusteres Hinzufügen der Lizenzinformationen"

Closes #24

See merge request !21

src/main/java/de/landsh/opendata/csw2dcat/MDMetadata2Dataset.java

@@ -25,8 +25,10 @@ import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
 import java.io.IOException;
+import java.io.UnsupportedEncodingException;
 import java.net.MalformedURLException;
 import java.net.URL;
+import java.net.URLEncoder;
 import java.time.LocalDate;
 import java.util.*;
 import java.util.stream.Collectors;
@@ -47,6 +49,7 @@ public class MDMetadata2Dataset {
     boolean ignoreInvalidMapping = true;
     private CswInterface cswInterface;
     private BridgeSettings settings;
+    private final Map<String, JSONObject> url2license = new HashMap<>();
 
     public MDMetadata2Dataset(Model model, CswInterface cswInterface, BridgeSettings bridgeSettings) {
         this(model);
@@ -77,6 +80,12 @@ public class MDMetadata2Dataset {
             final String value = entry.getString(key);
             keywordMapping.put(key, value);
         }
+
+        JSONArray licenses = new JSONArray(new JSONTokener(Objects.requireNonNull(getClass().getResourceAsStream("/licenses.json"))));
+        for (Object l : licenses) {
+            JSONObject license = (JSONObject) l;
+            url2license.put(license.getString("uri"), license);
+        }
     }
 
     static String getTextOrNull(Node node) {
@@ -533,18 +542,43 @@ public class MDMetadata2Dataset {
     }
 
     private void addLicenseInformation(Resource resource, List<Node> otherLegalConstraints) {
-        final JSONObject licenceInformation = findLicenseInformation(otherLegalConstraints);
-        if (licenceInformation != null) {
-            resource.addLiteral(DCATAPde.licenseAttributionByText, licenceInformation.getString("quelle"));
-            if (licenceInformation.has("url") && licenceInformation.getString("url").startsWith("http://dcat-ap.de/def/licenses/")) {
-                // Some publishers specify the licence as the URL and not the id.
-                resource.addProperty(DCTerms.license, model.createResource(licenceInformation.getString("url")));
+        final JSONObject licenseInformation = findLicenseInformation(otherLegalConstraints);
+    
+        if (licenseInformation != null) {
+            resource.addLiteral(DCATAPde.licenseAttributionByText, licenseInformation.getString("quelle"));
+    
+            if (licenseInformation.has("url")) {
+                String licenseURL = licenseInformation.getString("url");
+    
+                if (licenseURL.startsWith("http://dcat-ap.de/def/licenses/")) {
+                    resource.addProperty(DCTerms.license, model.createResource(licenseURL));
+                } else {
+                    final JSONObject license = url2license.get(licenseURL);
+    
+                    if (license != null && license.has("uri")) {
+                        resource.addProperty(DCTerms.license, model.createResource(license.getString("uri")));
+                    } else {
+                        log.info("Unknown license: {}", licenseURL);
+                        handleUnknownLicense(resource, licenseInformation.getString("id"));
+                    }
+                }
             } else {
-                resource.addProperty(DCTerms.license, model.createResource("http://dcat-ap.de/def/licenses/" + licenceInformation.getString("id")));
+                handleUnknownLicense(resource, licenseInformation.getString("id"));
             }
         }
     }
-
+  
+    private void handleUnknownLicense(Resource resource, String licenseId) {
+        try {
+            licenseId = URLEncoder.encode(licenseId, "UTF-8")
+                .replaceAll("%2F", "/"); // Replace the encoded slash with the original slash
+            resource.addProperty(DCTerms.license, model.createResource("http://dcat-ap.de/def/licenses/" + licenseId));
+            log.info("Falling back to: {}", licenseId);
+        } catch (UnsupportedEncodingException e) {
+            log.info("Unable to encode licenseId: {}", e);
+        }
+    }
+  
     private Resource guessFormat(String url) {
         final String lowerCaseURL = url.toLowerCase();
         if (lowerCaseURL.contains("service=wms")) {