ETL 清洗

@Configuration
public class ETLPipelineConfig {
    @Bean
    public ETLPipeline etlPipeline(
            DocumentLoader documentLoader,
            EmbeddingModel embeddingModel,
            VectorStore vectorStore) {
        return ETLPipeline.builder()
                .documentLoader(documentLoader)
                .embeddingModel(embeddingModel)
                .vectorStore(vectorStore)
                .build();
    }
}

spring.ai.etl.pipeline.enabled=true
spring.ai.etl.pipeline.chunk-size=1000
spring.ai.etl.pipeline.chunk-overlap=200
spring.ai.etl.pipeline.batch-size=100

@Component
public class CustomTransformer implements DocumentTransformer {
    @Override
    public Document transform(Document document) {
        // 自定义转换逻辑
        return transformedDocument;
    }
}

management.endpoints.web.exposure.include=etl-pipeline
management.endpoint.etl-pipeline.enabled=true