Merge pull request #518 from jmartisk/easy-rag-ingestion-strategies

Ability to turn off ingestion with Easy RAG

easy-rag/runtime/src/main/java/io/quarkiverse/langchain4j/easyrag/runtime/EasyRagConfig.java

@@ -48,4 +48,12 @@ public interface EasyRagConfig {
     @WithDefault("5")
     Integer maxResults();
 
+    /**
+     * The strategy to decide whether document ingestion into the store should happen at startup or not.
+     * The default is ON. Changing to OFF generally only makes sense if running against a persistent embedding store
+     * that was already populated.
+     */
+    @WithDefault("ON")
+    IngestionStrategy ingestionStrategy();
+
 }

easy-rag/runtime/src/main/java/io/quarkiverse/langchain4j/easyrag/runtime/EasyRagRecorder.java

@@ -31,6 +31,10 @@ public class EasyRagRecorder {
     private static final Logger LOGGER = Logger.getLogger(EasyRagRecorder.class);
 
     public void ingest(EasyRagConfig config, BeanContainer beanContainer) {
+        if (config.ingestionStrategy() == IngestionStrategy.OFF) {
+            LOGGER.info("Skipping document ingestion as per configuration");
+            return;
+        }
         EmbeddingStore<TextSegment> embeddingStore = beanContainer.beanInstance(EmbeddingStore.class);
         EmbeddingModel embeddingModel = beanContainer.beanInstance(EmbeddingModel.class);
 

easy-rag/runtime/src/main/java/io/quarkiverse/langchain4j/easyrag/runtime/IngestionStrategy.java

@@ -0,0 +1,11 @@
+package io.quarkiverse.langchain4j.easyrag.runtime;
+
+public enum IngestionStrategy {
+
+    // TODO: we'd like to also have a "if-empty" and "overwrite" strategy
+    // but these require some enhancements to the EmbeddingStore API
+
+    ON,
+    OFF
+
+}