Nomad changes

docs/snippets/kotlin/utils/chunking.md

@@ -0,0 +1,19 @@
+```kotlin title="Kotlin"
+import dev.kreuzberg.*
+import java.nio.file.Paths
+import java.util.Optional
+
+fun main() {
+    val chunking = ChunkingConfig.builder()
+        .withMaxCharacters(1500L)
+        .withOverlap(200L)
+        .build()
+
+    val config = ExtractionConfig.builder()
+        .withChunking(Optional.of(chunking))
+        .build()
+
+    val result = Kreuzberg.extractFileSync(Paths.get("document.pdf"), null, config)
+    println("Chunks: ${result.chunks()?.size ?: 0}")
+}
+```

docs/snippets/kotlin/utils/chunking_rag.md

@@ -0,0 +1,35 @@
+```kotlin title="Kotlin"
+import dev.kreuzberg.*
+import java.nio.file.Paths
+import java.util.Optional
+
+fun main() {
+    val embedding = EmbeddingConfig.builder()
+        .withModel(EmbeddingModelType.Preset("balanced"))
+        .withNormalize(true)
+        .build()
+
+    val chunking = ChunkingConfig.builder()
+        .withMaxCharacters(500L)
+        .withOverlap(50L)
+        .withEmbedding(Optional.of(embedding))
+        .build()
+
+    val config = ExtractionConfig.builder()
+        .withChunking(Optional.of(chunking))
+        .build()
+
+    val result = Kreuzberg.extractFileSync(Paths.get("research_paper.pdf"), null, config)
+    for (chunk in result.chunks().orEmpty()) {
+        val metadata = chunk.metadata()
+        println("Chunk ${metadata.chunkIndex() + 1}/${metadata.totalChunks()}")
+        println("Position: ${metadata.byteStart()}-${metadata.byteEnd()}")
+        val text = chunk.content()
+        val preview = text.take(100)
+        println("Content: $preview...")
+        chunk.embedding()?.let { vector ->
+            println("Embedding: ${vector.size} dimensions")
+        }
+    }
+}
+```

docs/snippets/kotlin/utils/embedding_with_chunking.md

@@ -0,0 +1,27 @@
+```kotlin title="Kotlin"
+import dev.kreuzberg.*
+import java.nio.file.Paths
+import java.util.Optional
+
+fun main() {
+    val embedding = EmbeddingConfig.builder()
+        .withModel(EmbeddingModelType.Preset("balanced"))
+        .withNormalize(true)
+        .withBatchSize(32L)
+        .withShowDownloadProgress(false)
+        .build()
+
+    val chunking = ChunkingConfig.builder()
+        .withMaxCharacters(1024L)
+        .withOverlap(100L)
+        .withEmbedding(Optional.of(embedding))
+        .build()
+
+    val config = ExtractionConfig.builder()
+        .withChunking(Optional.of(chunking))
+        .build()
+
+    val result = Kreuzberg.extractFileSync(Paths.get("document.pdf"), null, config)
+    println("Chunks with embeddings: ${result.chunks()?.size ?: 0}")
+}
+```

docs/snippets/kotlin/utils/keyword_extraction_example.md

@@ -0,0 +1,22 @@
+```kotlin title="Kotlin"
+import dev.kreuzberg.*
+import java.nio.file.Paths
+import java.util.Optional
+
+fun main() {
+    val keywords = KeywordConfig.builder()
+        .withAlgorithm(KeywordAlgorithm.Yake)
+        .withMaxKeywords(10L)
+        .withMinScore(0.3f)
+        .build()
+
+    val config = ExtractionConfig.builder()
+        .withKeywords(Optional.of(keywords))
+        .build()
+
+    val result = Kreuzberg.extractFileSync(Paths.get("research_paper.pdf"), null, config)
+    result.extractedKeywords()?.let { extracted ->
+        println("Keywords: $extracted")
+    }
+}
+```

docs/snippets/kotlin/utils/quality_processing_example.md

@@ -0,0 +1,22 @@
+```kotlin title="Kotlin"
+import dev.kreuzberg.*
+import java.nio.file.Paths
+import java.util.Optional
+
+fun main() {
+    val config = ExtractionConfig.builder()
+        .withEnableQualityProcessing(true)
+        .build()
+
+    val result = Kreuzberg.extractFileSync(Paths.get("scanned_document.pdf"), null, config)
+
+    val score = result.qualityScore()
+    if (score != null) {
+        if (score < 0.5) {
+            println("Warning: Low quality extraction (%.2f)".format(score))
+        } else {
+            println("Quality score: %.2f".format(score))
+        }
+    }
+}
+```

docs/snippets/kotlin/utils/standalone_embed.md

@@ -0,0 +1,17 @@
+```kotlin title="Kotlin"
+import dev.kreuzberg.*
+import java.util.Optional
+
+fun main() {
+    val config = EmbeddingConfig.builder()
+        .withModel(EmbeddingModelType.Preset("balanced"))
+        .withNormalize(true)
+        .build()
+
+    val texts = listOf("Hello, world!", "Kreuzberg is fast")
+    val embeddings = Kreuzberg.embedTexts(texts, config)
+
+    println("Texts embedded: ${embeddings.size}")
+    println("Dimensions: ${embeddings[0].size}")
+}
+```

docs/snippets/kotlin/utils/token_reduction.md

@@ -0,0 +1,19 @@
+```kotlin title="Kotlin"
+import dev.kreuzberg.*
+import java.nio.file.Paths
+import java.util.Optional
+
+fun main() {
+    val tokenReduction = TokenReductionOptions.builder()
+        .withMode("moderate")
+        .withPreserveImportantWords(true)
+        .build()
+
+    val config = ExtractionConfig.builder()
+        .withTokenReduction(Optional.of(tokenReduction))
+        .build()
+
+    val result = Kreuzberg.extractFileSync(Paths.get("document.pdf"), null, config)
+    println(result.content())
+}
+```

docs/snippets/kotlin/utils/token_reduction_example.md

@@ -0,0 +1,19 @@
+```kotlin title="Kotlin"
+import dev.kreuzberg.*
+import java.nio.file.Paths
+import java.util.Optional
+
+fun main() {
+    val tokenReduction = TokenReductionOptions.builder()
+        .withMode("moderate")
+        .withPreserveImportantWords(true)
+        .build()
+
+    val config = ExtractionConfig.builder()
+        .withTokenReduction(Optional.of(tokenReduction))
+        .build()
+
+    val result = Kreuzberg.extractFileSync(Paths.get("verbose_document.pdf"), null, config)
+    println("Reduced content length: ${result.content().length}")
+}
+```

docs/snippets/kotlin/utils/vector_database_integration.md

@@ -0,0 +1,52 @@
+```kotlin title="Kotlin"
+import dev.kreuzberg.*
+import java.nio.file.Paths
+import java.util.Optional
+
+data class VectorRecord(
+    val id: String,
+    val content: String,
+    val embedding: List<Float>,
+    val metadata: Map<String, String>
+)
+
+fun extractAndVectorize(documentPath: String, documentId: String): List<VectorRecord> {
+    val embedding = EmbeddingConfig.builder()
+        .withModel(EmbeddingModelType.Preset("balanced"))
+        .withNormalize(true)
+        .withBatchSize(32L)
+        .build()
+
+    val chunking = ChunkingConfig.builder()
+        .withMaxCharacters(512L)
+        .withOverlap(50L)
+        .withEmbedding(Optional.of(embedding))
+        .build()
+
+    val config = ExtractionConfig.builder()
+        .withChunking(Optional.of(chunking))
+        .build()
+
+    val result = Kreuzberg.extractFileSync(Paths.get(documentPath), null, config)
+
+    val records = mutableListOf<VectorRecord>()
+    val chunks = result.chunks().orEmpty()
+    for ((index, chunk) in chunks.withIndex()) {
+        val vector = chunk.embedding()
+        if (vector != null) {
+            val metadata = mapOf(
+                "document_id" to documentId,
+                "chunk_index" to index.toString(),
+                "content_length" to chunk.content().length.toString()
+            )
+            records += VectorRecord(
+                id = "${documentId}_chunk_$index",
+                content = chunk.content(),
+                embedding = vector,
+                metadata = metadata
+            )
+        }
+    }
+    return records
+}
+```