Nomad changes

docs/snippets/dart/metadata/language_detection.md

@@ -0,0 +1,32 @@
+```dart title="Dart"
+import 'package:kreuzberg/kreuzberg.dart';
+
+Future<void> main() async {
+  final languageDetection = LanguageDetectionConfig(
+    enabled: true,
+    minConfidence: 0.5,
+    detectMultiple: false,
+  );
+
+  final config = ExtractionConfig(
+    useCache: true,
+    enableQualityProcessing: true,
+    forceOcr: false,
+    disableOcr: false,
+    resultFormat: ResultFormat.unified,
+    outputFormat: OutputFormat.plain(),
+    includeDocumentStructure: false,
+    maxArchiveDepth: 3,
+    useLayoutForMarkdown: false,    languageDetection: languageDetection,
+  );
+
+  final result = await KreuzbergBridge.extractFile('document.pdf', null, config);
+
+  final detected = result.detectedLanguages;
+  if (detected != null && detected.isNotEmpty) {
+    print('Primary language: ${detected.first}');
+  } else {
+    print('No language detected');
+  }
+}
+```

docs/snippets/dart/metadata/language_detection_multilingual.md

@@ -0,0 +1,36 @@
+```dart title="Dart"
+import 'package:kreuzberg/kreuzberg.dart';
+
+Future<void> main() async {
+  final languageDetection = LanguageDetectionConfig(
+    enabled: true,
+    minConfidence: 0.3,
+    detectMultiple: true,
+  );
+
+  final config = ExtractionConfig(
+    useCache: true,
+    enableQualityProcessing: true,
+    forceOcr: false,
+    disableOcr: false,
+    resultFormat: ResultFormat.unified,
+    outputFormat: OutputFormat.plain(),
+    includeDocumentStructure: false,
+    maxArchiveDepth: 3,
+    useLayoutForMarkdown: false,    languageDetection: languageDetection,
+  );
+
+  final result = await KreuzbergBridge.extractFile('multilingual.pdf', null, config);
+
+  final detected = result.detectedLanguages;
+  if (detected == null || detected.isEmpty) {
+    print('No languages detected');
+    return;
+  }
+
+  print('Detected ${detected.length} language(s):');
+  for (final language in detected) {
+    print('  - $language');
+  }
+}
+```

docs/snippets/dart/metadata/metadata.md

@@ -0,0 +1,38 @@
+```dart title="Dart"
+import 'package:kreuzberg/kreuzberg.dart';
+
+Future<void> main() async {
+  final result = await KreuzbergBridge.extractFile('document.pdf', null);
+
+  final metadata = result.metadata;
+
+  if (metadata.title != null) {
+    print('Title: ${metadata.title}');
+  }
+  if (metadata.subject != null) {
+    print('Subject: ${metadata.subject}');
+  }
+  if (metadata.authors != null) {
+    print('Authors: ${metadata.authors!.join(', ')}');
+  }
+  if (metadata.keywords != null) {
+    print('Keywords: ${metadata.keywords!.join(', ')}');
+  }
+  if (metadata.language != null) {
+    print('Language: ${metadata.language}');
+  }
+  if (metadata.createdAt != null) {
+    print('Created: ${metadata.createdAt}');
+  }
+  if (metadata.modifiedAt != null) {
+    print('Modified: ${metadata.modifiedAt}');
+  }
+  if (metadata.extractionDurationMs != null) {
+    print('Extraction took: ${metadata.extractionDurationMs} ms');
+  }
+
+  for (final entry in metadata.additional.entries) {
+    print('Additional[${entry.key}]: ${entry.value}');
+  }
+}
+```

docs/snippets/dart/metadata/page_boundaries.md

@@ -0,0 +1,31 @@
+```dart title="Dart"
+import 'package:kreuzberg/kreuzberg.dart';
+
+Future<void> main() async {
+  final result = await KreuzbergBridge.extractFile('document.pdf', null);
+
+  final pages = result.metadata.pages;
+  if (pages == null) {
+    print('No page structure available');
+    return;
+  }
+
+  final boundaries = pages.boundaries;
+  if (boundaries == null || boundaries.isEmpty) {
+    print('No page boundaries available');
+    return;
+  }
+
+  final content = result.content;
+  for (final boundary in boundaries.take(3)) {
+    final start = boundary.byteStart.toInt();
+    final end = boundary.byteEnd.toInt();
+    final pageText = content.substring(start, end);
+    final previewEnd = pageText.length < 100 ? pageText.length : 100;
+
+    print('Page ${boundary.pageNumber}:');
+    print('  Byte range: $start-$end');
+    print('  Preview: ${pageText.substring(0, previewEnd)}...');
+  }
+}
+```

docs/snippets/dart/metadata/page_tracking_basic.md

@@ -0,0 +1,38 @@
+```dart title="Dart"
+import 'package:kreuzberg/kreuzberg.dart';
+
+Future<void> main() async {
+  final pageConfig = PageConfig(
+    extractPages: true,
+    insertPageMarkers: false,
+    markerFormat: '<!-- page {page} -->',
+  );
+
+  final config = ExtractionConfig(
+    useCache: true,
+    enableQualityProcessing: true,
+    forceOcr: false,
+    disableOcr: false,
+    resultFormat: ResultFormat.unified,
+    outputFormat: OutputFormat.plain(),
+    includeDocumentStructure: false,
+    maxArchiveDepth: 3,
+    useLayoutForMarkdown: false,    pages: pageConfig,
+  );
+
+  final result = await KreuzbergBridge.extractFile('document.pdf', null, config);
+
+  final pages = result.pages;
+  if (pages == null) {
+    print('No per-page content available');
+    return;
+  }
+
+  for (final page in pages) {
+    print('Page ${page.pageNumber}:');
+    print('  Content: ${page.content.length} chars');
+    print('  Tables: ${page.tables.length}');
+    print('  Images: ${page.images.length}');
+  }
+}
+```

docs/snippets/dart/metadata/tables.md

@@ -0,0 +1,20 @@
+```dart title="Dart"
+import 'package:kreuzberg/kreuzberg.dart';
+
+Future<void> main() async {
+  final result = await KreuzbergBridge.extractFile('document.pdf', null);
+
+  for (final table in result.tables) {
+    print('Table on page ${table.pageNumber} with ${table.cells.length} rows');
+    print(table.markdown);
+
+    for (final row in table.cells) {
+      print(row);
+    }
+
+    if (table.boundingBox != null) {
+      print('Bounding box: ${table.boundingBox}');
+    }
+  }
+}
+```

docs/snippets/dart/metadata/vector_database_integration.md

@@ -0,0 +1,92 @@
+```dart title="Dart"
+import 'package:kreuzberg/kreuzberg.dart';
+
+class VectorRecord {
+  final String id;
+  final List<double> embedding;
+  final String content;
+  final Map<String, Object?> metadata;
+
+  const VectorRecord({
+    required this.id,
+    required this.embedding,
+    required this.content,
+    required this.metadata,
+  });
+}
+
+void storeInVectorDatabase(List<VectorRecord> records) {
+  for (final record in records) {
+    if (record.embedding.isEmpty) {
+      continue;
+    }
+    print(
+      'Storing ${record.id}: ${record.content.length} chars, '
+      '${record.embedding.length} dims',
+    );
+  }
+}
+
+Future<List<VectorRecord>> extractAndVectorize(
+  String documentPath,
+  String documentId,
+) async {
+  final embedding = EmbeddingConfig(
+    model: EmbeddingModelType.preset(name: 'balanced'),
+    normalize: true,
+    batchSize: 32,
+    showDownloadProgress: false,
+  );
+
+  final chunking = ChunkingConfig(
+    maxCharacters: 512,
+    overlap: 50,
+    trim: true,
+    chunkerType: ChunkerType.text,
+    embedding: embedding,
+    sizing: ChunkSizing.characters(),
+    prependHeadingContext: false,
+  );
+
+  final config = ExtractionConfig(
+    useCache: true,
+    enableQualityProcessing: true,
+    forceOcr: false,
+    disableOcr: false,
+    resultFormat: ResultFormat.unified,
+    outputFormat: OutputFormat.plain(),
+    includeDocumentStructure: false,
+    maxArchiveDepth: 3,
+    useLayoutForMarkdown: false,    chunking: chunking,
+  );
+
+  final result = await KreuzbergBridge.extractFile(documentPath, null, config);
+  final chunks = result.chunks ?? const <Chunk>[];
+
+  final records = <VectorRecord>[];
+  for (var index = 0; index < chunks.length; index++) {
+    final chunk = chunks[index];
+    final embeddingValues = chunk.embedding?.toList() ?? const <double>[];
+
+    records.add(
+      VectorRecord(
+        id: '${documentId}_chunk_$index',
+        content: chunk.content,
+        embedding: embeddingValues,
+        metadata: {
+          'document_id': documentId,
+          'chunk_index': index,
+          'content_length': chunk.content.length,
+        },
+      ),
+    );
+  }
+
+  storeInVectorDatabase(records);
+  return records;
+}
+
+Future<void> main() async {
+  await extractAndVectorize('document.pdf', 'doc-1');
+}
+```