Fix #41: Historischen KI-Dateinamen für übersprungene Dokumente in Ergebnistabelle anzeigen

Neue Komponenten:
- ResolveHistoricalFileNameUseCase (port/in) und DefaultResolveHistoricalFileNameUseCase (usecase)
- GuiHistoricalFileNamePort (GUI-interner Port, folgt dem Muster von GuiManualFileRenamePort)

GuiBatchRunCoordinator ruft in toRow() für SKIPPED-Zeilen ohne finalName den
historicalFileNamePort auf und trägt den Rückgabewert als neuen Dateinamen ein.

Bootstrap verdrahtet resolveHistoricalFileNameForGui als GuiHistoricalFileNamePort
und übergibt ihn über GuiStartupContext an den GUI-Adapter.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

pdf-umbenenner-application/src/main/java/de/gecheckt/pdf/umbenenner/application/port/in/ResolveHistoricalFileNameUseCase.java

@@ -0,0 +1,37 @@
+package de.gecheckt.pdf.umbenenner.application.port.in;
+
+import java.util.Optional;
+
+import de.gecheckt.pdf.umbenenner.domain.model.DocumentFingerprint;
+
+/**
+ * Inbound port for resolving the historical target filename for a previously processed document.
+ * <p>
+ * Used to populate the "Neuer Dateiname" column in the GUI result table for documents that
+ * were skipped in the current run because they were already in a terminal state from a
+ * previous run. For documents that previously reached
+ * {@link de.gecheckt.pdf.umbenenner.domain.model.ProcessingStatus#SUCCESS}, the last
+ * successfully written target filename is returned. For all other terminal states
+ * (e.g. {@code FAILED_FINAL}) that were never copied to the target folder, an empty
+ * {@link Optional} is returned.
+ * <p>
+ * <strong>Architecture boundary:</strong> Implementations of this port must not expose
+ * JDBC, SQLite, filesystem or HTTP types through this interface. All infrastructure details
+ * remain in the adapter layer.
+ */
+public interface ResolveHistoricalFileNameUseCase {
+
+    /**
+     * Returns the last successfully written target filename for the document identified
+     * by the given fingerprint, or an empty {@link Optional} if no such filename exists.
+     * <p>
+     * The method never throws: technical failures during the repository lookup are caught
+     * and result in an empty return value.
+     *
+     * @param fingerprint content-based document identity; must not be {@code null}
+     * @return the last target filename written for this document, or empty if the document
+     *         never reached a successful terminal state or the lookup failed
+     * @throws NullPointerException if {@code fingerprint} is {@code null}
+     */
+    Optional<String> resolveHistoricalFileName(DocumentFingerprint fingerprint);
+}

pdf-umbenenner-application/src/main/java/de/gecheckt/pdf/umbenenner/application/port/in/package-info.java

@@ -28,6 +28,13 @@
  *       — Event and summary value types carried to the observer</li>
  * </ul>
  * <p>
+ * <p>
+ * Query use cases (for GUI adapters):
+ * <ul>
+ *   <li>{@link de.gecheckt.pdf.umbenenner.application.port.in.ResolveHistoricalFileNameUseCase}
+ *       — Resolves the last known target filename for a document identified by its fingerprint</li>
+ * </ul>
+ * <p>
  * Architecture Rule: Inbound ports are independent of implementation and contain no business logic.
  * They define "what can be done to the application". All dependencies point inward;
  * adapters depend on ports, not vice versa.

pdf-umbenenner-application/src/main/java/de/gecheckt/pdf/umbenenner/application/usecase/DefaultResolveHistoricalFileNameUseCase.java

@@ -0,0 +1,69 @@
+package de.gecheckt.pdf.umbenenner.application.usecase;
+
+import java.util.Objects;
+import java.util.Optional;
+
+import de.gecheckt.pdf.umbenenner.application.port.in.ResolveHistoricalFileNameUseCase;
+import de.gecheckt.pdf.umbenenner.application.port.out.DocumentRecordLookupResult;
+import de.gecheckt.pdf.umbenenner.application.port.out.DocumentRecordRepository;
+import de.gecheckt.pdf.umbenenner.application.port.out.DocumentTerminalSuccess;
+import de.gecheckt.pdf.umbenenner.domain.model.DocumentFingerprint;
+
+/**
+ * Default implementation of {@link ResolveHistoricalFileNameUseCase}.
+ * <p>
+ * Queries the {@link DocumentRecordRepository} for the master record of the given fingerprint.
+ * If the record represents a document that previously reached a successful terminal state,
+ * the last known target filename ({@code lastTargetFileName}) is returned.
+ * <p>
+ * For all other terminal states (e.g. documents that finally failed without ever producing
+ * a target copy) or when no master record exists, an empty {@link Optional} is returned.
+ * Technical failures during the repository lookup are caught silently and treated as
+ * an absent result so that the calling GUI layer is never forced to handle exceptions
+ * from this query path.
+ */
+public class DefaultResolveHistoricalFileNameUseCase implements ResolveHistoricalFileNameUseCase {
+
+    private final DocumentRecordRepository documentRecordRepository;
+
+    /**
+     * Creates the use case with the required document master record repository.
+     *
+     * @param documentRecordRepository repository for reading document master records;
+     *                                 must not be {@code null}
+     * @throws NullPointerException if {@code documentRecordRepository} is {@code null}
+     */
+    public DefaultResolveHistoricalFileNameUseCase(
+            DocumentRecordRepository documentRecordRepository) {
+        this.documentRecordRepository = Objects.requireNonNull(
+                documentRecordRepository, "documentRecordRepository must not be null");
+    }
+
+    /**
+     * Returns the last successfully written target filename for the given fingerprint,
+     * or an empty {@link Optional} if no such filename exists.
+     * <p>
+     * The method queries the document master record. Only documents with a prior
+     * successful terminal state carry a non-null {@code lastTargetFileName}. For all
+     * other states (unknown, processable, finally failed) and on any technical lookup
+     * failure, an empty {@link Optional} is returned.
+     *
+     * @param fingerprint content-based document identity; must not be {@code null}
+     * @return the historical target filename, or empty if not available
+     * @throws NullPointerException if {@code fingerprint} is {@code null}
+     */
+    @Override
+    public Optional<String> resolveHistoricalFileName(DocumentFingerprint fingerprint) {
+        Objects.requireNonNull(fingerprint, "fingerprint must not be null");
+        try {
+            DocumentRecordLookupResult result =
+                    documentRecordRepository.findByFingerprint(fingerprint);
+            if (result instanceof DocumentTerminalSuccess success) {
+                return Optional.ofNullable(success.record().lastTargetFileName());
+            }
+            return Optional.empty();
+        } catch (Exception e) {
+            return Optional.empty();
+        }
+    }
+}