M5 komplett umgesetzt

pdf-umbenenner-bootstrap/src/main/java/de/gecheckt/pdf/umbenenner/bootstrap/BootstrapRunner.java

@@ -9,31 +9,43 @@ import org.apache.logging.log4j.LogManager;
 import org.apache.logging.log4j.Logger;
 
 import de.gecheckt.pdf.umbenenner.adapter.in.cli.SchedulerBatchCommand;
+import de.gecheckt.pdf.umbenenner.adapter.out.ai.OpenAiHttpAdapter;
 import de.gecheckt.pdf.umbenenner.adapter.out.bootstrap.validation.InvalidStartConfigurationException;
 import de.gecheckt.pdf.umbenenner.adapter.out.bootstrap.validation.StartConfigurationValidator;
+import de.gecheckt.pdf.umbenenner.adapter.out.clock.SystemClockAdapter;
 import de.gecheckt.pdf.umbenenner.adapter.out.configuration.ConfigurationLoadingException;
 import de.gecheckt.pdf.umbenenner.adapter.out.configuration.PropertiesConfigurationPortAdapter;
 import de.gecheckt.pdf.umbenenner.adapter.out.fingerprint.Sha256FingerprintAdapter;
 import de.gecheckt.pdf.umbenenner.adapter.out.lock.FilesystemRunLockPortAdapter;
 import de.gecheckt.pdf.umbenenner.adapter.out.pdfextraction.PdfTextExtractionPortAdapter;
+import de.gecheckt.pdf.umbenenner.adapter.out.prompt.FilesystemPromptPortAdapter;
 import de.gecheckt.pdf.umbenenner.adapter.out.sourcedocument.SourceDocumentCandidatesPortAdapter;
 import de.gecheckt.pdf.umbenenner.adapter.out.sqlite.SqliteDocumentRecordRepositoryAdapter;
 import de.gecheckt.pdf.umbenenner.adapter.out.sqlite.SqliteProcessingAttemptRepositoryAdapter;
 import de.gecheckt.pdf.umbenenner.adapter.out.sqlite.SqliteSchemaInitializationAdapter;
 import de.gecheckt.pdf.umbenenner.adapter.out.sqlite.SqliteUnitOfWorkAdapter;
+import de.gecheckt.pdf.umbenenner.adapter.out.targetcopy.FilesystemTargetFileCopyAdapter;
+import de.gecheckt.pdf.umbenenner.adapter.out.targetfolder.FilesystemTargetFolderAdapter;
 import de.gecheckt.pdf.umbenenner.application.config.RuntimeConfiguration;
 import de.gecheckt.pdf.umbenenner.application.config.startup.StartConfiguration;
 import de.gecheckt.pdf.umbenenner.application.port.in.BatchRunOutcome;
 import de.gecheckt.pdf.umbenenner.application.port.in.BatchRunProcessingUseCase;
+import de.gecheckt.pdf.umbenenner.application.port.out.AiInvocationPort;
+import de.gecheckt.pdf.umbenenner.application.port.out.ClockPort;
 import de.gecheckt.pdf.umbenenner.application.port.out.ConfigurationPort;
+import de.gecheckt.pdf.umbenenner.application.port.out.TargetFileCopyPort;
+import de.gecheckt.pdf.umbenenner.application.port.out.TargetFolderPort;
 import de.gecheckt.pdf.umbenenner.application.port.out.DocumentPersistenceException;
 import de.gecheckt.pdf.umbenenner.application.port.out.DocumentRecordRepository;
 import de.gecheckt.pdf.umbenenner.application.port.out.FingerprintPort;
 import de.gecheckt.pdf.umbenenner.application.port.out.PersistenceSchemaInitializationPort;
 import de.gecheckt.pdf.umbenenner.application.port.out.ProcessingAttemptRepository;
 import de.gecheckt.pdf.umbenenner.application.port.out.ProcessingLogger;
+import de.gecheckt.pdf.umbenenner.application.port.out.PromptPort;
 import de.gecheckt.pdf.umbenenner.application.port.out.RunLockPort;
 import de.gecheckt.pdf.umbenenner.application.port.out.UnitOfWorkPort;
+import de.gecheckt.pdf.umbenenner.application.service.AiNamingService;
+import de.gecheckt.pdf.umbenenner.application.service.AiResponseValidator;
 import de.gecheckt.pdf.umbenenner.application.service.DocumentProcessingCoordinator;
 import de.gecheckt.pdf.umbenenner.application.usecase.DefaultBatchRunProcessingUseCase;
 import de.gecheckt.pdf.umbenenner.bootstrap.adapter.Log4jProcessingLogger;
@@ -71,11 +83,15 @@ import de.gecheckt.pdf.umbenenner.domain.model.RunId;
  * <ul>
  *   <li>{@link PropertiesConfigurationPortAdapter} — loads configuration from properties and environment.</li>
  *   <li>{@link FilesystemRunLockPortAdapter} — ensures exclusive execution via a lock file.</li>
- *   <li>{@link SqliteSchemaInitializationAdapter} — initializes SQLite schema at startup.</li>
+ *   <li>{@link SqliteSchemaInitializationAdapter} — initializes SQLite schema (including target-copy
+ *       schema evolution) at startup.</li>
  *   <li>{@link Sha256FingerprintAdapter} — provides content-based document identification.</li>
  *   <li>{@link SqliteDocumentRecordRepositoryAdapter} — manages document master records.</li>
  *   <li>{@link SqliteProcessingAttemptRepositoryAdapter} — maintains attempt history.</li>
  *   <li>{@link SqliteUnitOfWorkAdapter} — coordinates atomic persistence operations.</li>
+ *   <li>{@link FilesystemTargetFolderAdapter} — resolves unique filenames in the configured target folder.</li>
+ *   <li>{@link FilesystemTargetFileCopyAdapter} — copies source documents to the target folder via
+ *       a temporary file and final move/rename.</li>
  * </ul>
  * <p>
  * Schema initialization is performed exactly once in {@link #run()} before the batch processing loop
@@ -162,12 +178,22 @@ public class BootstrapRunner {
      *   <li>{@link SourceDocumentCandidatesPortAdapter} for PDF candidate discovery.</li>
      *   <li>{@link PdfTextExtractionPortAdapter} for PDFBox-based text and page count extraction.</li>
      *   <li>{@link Sha256FingerprintAdapter} for SHA-256 content fingerprinting.</li>
-     *   <li>{@link SqliteSchemaInitializationAdapter} for SQLite schema DDL at startup.</li>
+     *   <li>{@link SqliteSchemaInitializationAdapter} for SQLite schema DDL and target-copy schema
+     *       evolution at startup.</li>
      *   <li>{@link SqliteDocumentRecordRepositoryAdapter} for document master record CRUD.</li>
      *   <li>{@link SqliteProcessingAttemptRepositoryAdapter} for attempt history CRUD.</li>
      *   <li>{@link SqliteUnitOfWorkAdapter} for atomic persistence operations.</li>
+     *   <li>{@link FilesystemTargetFolderAdapter} for duplicate-safe filename resolution in the
+     *       configured {@code target.folder}.</li>
+     *   <li>{@link FilesystemTargetFileCopyAdapter} for copying source documents to the target folder
+     *       via a temporary file and final atomic move/rename.</li>
      * </ul>
      * <p>
+     * Target folder availability and write access are validated in
+     * {@link #loadAndValidateConfiguration()} via {@link StartConfigurationValidator} before
+     * schema initialisation and batch processing begin. If the target folder does not yet exist,
+     * the validator creates it; failure to do so is a hard startup error.
+     * <p>
      * Schema initialisation is performed explicitly in {@link #run()} before the batch loop
      * begins. Failure during initialisation aborts the run with exit code 1.
      */
@@ -189,8 +215,23 @@ public class BootstrapRunner {
             UnitOfWorkPort unitOfWorkPort =
                     new SqliteUnitOfWorkAdapter(jdbcUrl);
             ProcessingLogger coordinatorLogger = new Log4jProcessingLogger(DocumentProcessingCoordinator.class);
+            TargetFolderPort targetFolderPort = new FilesystemTargetFolderAdapter(startConfig.targetFolder());
+            TargetFileCopyPort targetFileCopyPort = new FilesystemTargetFileCopyAdapter(startConfig.targetFolder());
             DocumentProcessingCoordinator documentProcessingCoordinator =
-                    new DocumentProcessingCoordinator(documentRecordRepository, processingAttemptRepository, unitOfWorkPort, coordinatorLogger);
+                    new DocumentProcessingCoordinator(documentRecordRepository, processingAttemptRepository, unitOfWorkPort, targetFolderPort, targetFileCopyPort, coordinatorLogger);
+
+            // Wire AI naming pipeline
+            AiInvocationPort aiInvocationPort = new OpenAiHttpAdapter(startConfig);
+            PromptPort promptPort = new FilesystemPromptPortAdapter(startConfig.promptTemplateFile());
+            ClockPort clockPort = new SystemClockAdapter();
+            AiResponseValidator aiResponseValidator = new AiResponseValidator(clockPort);
+            AiNamingService aiNamingService = new AiNamingService(
+                    aiInvocationPort,
+                    promptPort,
+                    aiResponseValidator,
+                    startConfig.apiModel(),
+                    startConfig.maxTextCharacters());
+
             ProcessingLogger useCaseLogger = new Log4jProcessingLogger(DefaultBatchRunProcessingUseCase.class);
             return new DefaultBatchRunProcessingUseCase(
                     runtimeConfig,
@@ -199,6 +240,7 @@ public class BootstrapRunner {
                     new PdfTextExtractionPortAdapter(),
                     fingerprintPort,
                     documentProcessingCoordinator,
+                    aiNamingService,
                     useCaseLogger);
         };
         this.commandFactory = SchedulerBatchCommand::new;
@@ -272,8 +314,17 @@ public class BootstrapRunner {
 
     /**
      * Loads configuration via {@link ConfigurationPort} and validates it via
-     * {@link StartConfigurationValidator}. Validation includes checking that the
-     * {@code sqlite.file} parent directory exists or is technically creatable.
+     * {@link StartConfigurationValidator}.
+     * <p>
+     * Validation includes:
+     * <ul>
+     *   <li>{@code source.folder}: must exist, be a directory, and be readable.</li>
+     *   <li>{@code target.folder}: must exist as a writable directory, or be technically
+     *       creatable (validator attempts {@code Files.createDirectories} if absent;
+     *       failure here is a hard startup error).</li>
+     *   <li>{@code sqlite.file}: parent directory must exist.</li>
+     *   <li>All numeric and URI constraints.</li>
+     * </ul>
      */
     private StartConfiguration loadAndValidateConfiguration() {
         ConfigurationPort configPort = configPortFactory.create();

pdf-umbenenner-bootstrap/src/main/java/de/gecheckt/pdf/umbenenner/bootstrap/package-info.java

@@ -28,11 +28,15 @@
  * Startup sequence:
  * <ul>
  *   <li>Load and validate complete startup configuration from properties file and environment variables</li>
- *   <li>Initialize SQLite persistence schema via {@link de.gecheckt.pdf.umbenenner.application.port.out.PersistenceSchemaInitializationPort},
+ *   <li>Validate target folder availability and write access; create target folder if absent
+ *       (failure is a hard startup error)</li>
+ *   <li>Initialize SQLite persistence schema (including target-copy schema evolution) via
+ *       {@link de.gecheckt.pdf.umbenenner.application.port.out.PersistenceSchemaInitializationPort},
  *       ensuring the database is ready before any batch processing</li>
  *   <li>Schema initialization failure is treated as a hard bootstrap error and causes exit code 1</li>
  *   <li>Create run lock adapter and acquire exclusive lock</li>
- *   <li>Wire all outbound adapters (document candidates, PDF extraction, fingerprint, persistence, logging)</li>
+ *   <li>Wire all outbound adapters (document candidates, PDF extraction, fingerprint, persistence,
+ *       target folder duplicate resolution, target file copy, logging)</li>
  *   <li>Wire and invoke the batch processing CLI adapter</li>
  *   <li>Map batch outcome to process exit code</li>
  * </ul>