DocumentReader RAG Data Source Integration

Basic Usage

The Spring AI Alibaba official community provides many DocumentReader plugin extensions. In RAG scenarios, when integrating private domain data from different sources and formats, these plugins are very useful as they help developers quickly read data, avoiding the trouble of repetitive development.

Taking Lark/Feishu document library as an example, here’s the basic usage of the official community DocumentReader implementation for data integration:

1. Add Maven dependency

<dependency>
  <groupId>com.alibaba.cloud.ai</groupId>
  <artifactId>feishu-document-reader</artifactId>
  <version>${spring.ai.alibaba.version}</version>
</dependency>

2. Write code to read documents and write to vector database

FeiShuResource feiShuResource = FeiShuResource.builder()
      .appId("xxxxx")
      .appSecret("xxxxxxx")
      .build();
FeiShuDocumentReader reader = new FeiShuDocumentReader(feishuResourcde);

List<Document> documentList = reader.get();

TokenTextSplitter splitter = new TokenTextSplitter();
List<Document> chunks = splitter.apply(documentList);

vectorStore.add(chunks);

Community Implementation List