使用SpringBoot快速集成LangChain4j, 实现AI的丝滑调用

1. 简介

前几章简单测试了一下LangChain4J的特性, 本章使用SpringBoot快速集成LangChain4J, 实现丝滑调用大模型, 往期内容传送门

LangChain4J官方提供了SpringBoot Starter, 本章就使用Starter进行快速集成.

2. 环境信息

使用SDK版本信息如下：

Java: 21

SpringBoot: 3.4.5

LangChain4j: 1.0.1

LLM: (使用在线的百炼(阿里)平台)

embedding模型: text-embedding-v3

chat模型: qwen-plus

PGVector(postgresql版本的向量数据库, 文章最后有相关的docker-compose): 0.8.0-pg17

3. Maven

100

101

102

103

<?xml version="1.0" encoding="UTF-8"?>

<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"

xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 https://maven.apache.org/xsd/maven-4.0.0.xsd">

<groupId>org.springframework.boot</groupId>

<artifactId>spring-boot-starter-parent</artifactId>

</parent>

<artifactId>langchain-test</artifactId>

<version>0.0.1-SNAPSHOT</version>

<name>langchain-test</name>

<description>langchain-test</description>

<java.version>21</java.version>

<guava.version>33.0.0-jre</guava.version>

</properties>

<groupId>dev.langchain4j</groupId>

<artifactId>langchain4j-bom</artifactId>

<scope>import</scope>

</dependency>

</dependencies>

</dependencyManagement>

<groupId>org.springframework.boot</groupId>

<artifactId>spring-boot-starter-web</artifactId>

</dependency>

<groupId>dev.langchain4j</groupId>

<artifactId>langchain4j-open-ai-spring-boot-starter</artifactId>

</dependency>

<groupId>dev.langchain4j</groupId>

<artifactId>langchain4j-spring-boot-starter</artifactId>

</dependency>

<groupId>dev.langchain4j</groupId>

<artifactId>langchain4j-open-ai</artifactId>

</dependency>

<groupId>com.google.guava</groupId>

<artifactId>guava</artifactId>

<version>${guava.version}</version>

</dependency>

<groupId>dev.langchain4j</groupId>

<artifactId>langchain4j-reactor</artifactId>

</dependency>

<groupId>dev.langchain4j</groupId>

<artifactId>langchain4j-pgvector</artifactId>

</dependency>

<groupId>org.projectlombok</groupId>

<artifactId>lombok</artifactId>

</dependency>

<groupId>org.springframework.boot</groupId>

<artifactId>spring-boot-starter-test</artifactId>

</dependency>

</dependencies>

<build>

<groupId>org.apache.maven.plugins</groupId>

<artifactId>maven-compiler-plugin</artifactId>

</plugin>

<groupId>org.springframework.boot</groupId>

<artifactId>spring-boot-maven-plugin</artifactId>

<groupId>org.projectlombok</groupId>

<artifactId>lombok</artifactId>

</exclude>

</excludes>

</configuration>

</plugin>

</plugins>

</build>

</project>

4. 配置信息

4.1 application.yml

通过在配置文件中声明模型信息即可实现对应模型的自动注入

langchain4j:

open-ai:

# 普通聊天模型

chat-model:

api-key: ${LLM_API_KEY}

model-name: qwen-plus

base-url: https://dashscope.aliyuncs.com/compatible-mode/v1

# 流式相应模型

streaming-chat-model:

api-key: ${LLM_API_KEY}

model-name: qwen-plus

base-url: https://dashscope.aliyuncs.com/compatible-mode/v1

# 向量模型

embedding-model:

api-key: ${LLM_API_KEY}

model-name: text-embedding-v3

base-url: https://dashscope.aliyuncs.com/compatible-mode/v1

4.2 配置类

主要声明了

聊天记忆提供类, 关联了记忆存储对象

向量存储对象,这里使用的是pgvector

内容检索器(RAG-检索实现)

package com.ldx.langchaintest.config;

import com.ldx.langchaintest.service.PersistentChatMemoryStore;

import dev.langchain4j.data.segment.TextSegment;

import dev.langchain4j.memory.chat.ChatMemoryProvider;

import dev.langchain4j.memory.chat.MessageWindowChatMemory;

import dev.langchain4j.model.embedding.EmbeddingModel;

import dev.langchain4j.rag.content.retriever.ContentRetriever;

import dev.langchain4j.rag.content.retriever.EmbeddingStoreContentRetriever;

import dev.langchain4j.store.embedding.EmbeddingStore;

import dev.langchain4j.store.embedding.pgvector.DefaultMetadataStorageConfig;

import dev.langchain4j.store.embedding.pgvector.PgVectorEmbeddingStore;

import dev.langchain4j.store.memory.chat.InMemoryChatMemoryStore;

import org.springframework.context.annotation.Bean;

import org.springframework.context.annotation.Configuration;

/**

* config

* @author ludangxin

* @date 2025/5/15

@Configuration

public class AiConfiguration {

/**

* 聊天记忆提供者

* @param persistentChatMemoryStore 对话内容持久化对象

* @return 对话记忆 provider

@Bean

public ChatMemoryProvider jdbcChatMemoryProvider(PersistentChatMemoryStore persistentChatMemoryStore) {

return memoryId -> MessageWindowChatMemory

.builder()

.id(memoryId)

// 这里使用了自定义的会话存储对象, 可以通过其实现对话过程内容的持久化

// 本地测试的话可以使用 InMemoryChatMemoryStore对象实现内存存储

.chatMemoryStore(persistentChatMemoryStore)

.maxMessages(5)

.build();

}

/**

* 向量存储对象

* @param embeddingModel 向量模型

* @return 向量存储对象

public EmbeddingStore<TextSegment> embeddingStore(EmbeddingModel embeddingModel) {

return PgVectorEmbeddingStore

.builder()

.host("localhost") // 必需：PostgresSQL 实例的主机

.port(5431) // 必需：PostgresSQL 实例的端口

.database("postgres") // 必需：数据库名称

.user("root") // 必需：数据库用户

.password("123456") // 必需：数据库密码

.table("my_embeddings") // 必需：存储嵌入的表名

.dimension(embeddingModel.dimension()) // 必需：嵌入的维度

.metadataStorageConfig(DefaultMetadataStorageConfig.defaultConfig()) // 元数据存储配置

.build();

}

/**

* 内容检索器

* @param embeddingModel 向量模型

* @return 内容检索器

@Bean

public ContentRetriever contentRetriever(EmbeddingModel embeddingModel) {

return EmbeddingStoreContentRetriever

.builder()

.embeddingStore(this.embeddingStore(embeddingModel))

.embeddingModel(embeddingModel)

.maxResults(10)

.minScore(0.65)

.build();

}

4.3 聊天记忆持久化

实现了ChatMemoryStore接口, 这里测试使用的是map存储的, 生产环境中可以持久化到数据库中

chat过程中消息会通过ChatMemory调用ChatMemoryStore对聊天内容进行持久化/获取

package com.ldx.langchaintest.service;

import com.google.common.collect.ArrayListMultimap;

import dev.langchain4j.data.message.ChatMessage;

import dev.langchain4j.store.memory.chat.ChatMemoryStore;

import org.springframework.stereotype.Service;

import java.util.List;

@Service

public class PersistentChatMemoryStore implements ChatMemoryStore {

final ArrayListMultimap<Object, ChatMessage> messagesStore = ArrayListMultimap.create();

@Override

public List<ChatMessage> getMessages(Object memoryId) {

return messagesStore.get(memoryId);

}

@Override

public void updateMessages(Object memoryId, List<ChatMessage> messages) {

messagesStore.put(memoryId, messages.getLast());

}

@Override

public void deleteMessages(Object memoryId) {

messagesStore.removeAll(memoryId);

}

4.4 tools

package com.ldx.langchaintest.tools;

import dev.langchain4j.agent.tool.P;

import dev.langchain4j.agent.tool.Tool;

import lombok.extern.slf4j.Slf4j;

import org.springframework.stereotype.Component;

/**

* tools

* @author ludangxin

* @date 2025/5/15

@Slf4j

@Component

public class SysTools {

@Tool("根据用户的名称获取对应的code")

public String getUserCodeByUsername(@P("用户名称") String username) {

log.info("get user code by username:{}", username);

if ("张铁牛".equals(username)) {

return "003";

}

return "000";

}

5. 核心源码

5.1 Ai Services

@AiService 将实现AiService的自动注入

wiringMode = EXPLICIT: 用户自己指定相关的bean

缺省:wiringMode = AUTOMATIC: 项目启动时自动在环境中找对应的对象实现注入,如果有多个(比如:chatModel),启动报错

这里举了几种典型的场景如

普通聊天 chat()

聊天记忆&流式输出 chatWithStream()

提取指定内容并将结果结构化 extractPerson()

提示词占位替换 mockUsername()

rag text-sql chatWithSql()

package com.ldx.langchaintest.service;

import com.ldx.langchaintest.domain.Person;

import dev.langchain4j.service.MemoryId;

import dev.langchain4j.service.SystemMessage;

import dev.langchain4j.service.UserMessage;

import dev.langchain4j.service.V;

import dev.langchain4j.service.spring.AiService;

import reactor.core.publisher.Flux;

import java.util.List;

import static dev.langchain4j.service.spring.AiServiceWiringMode.EXPLICIT;

/**

* ai svc

* @author ludangxin

* @date 2025/5/16

@AiService(wiringMode = EXPLICIT,

chatModel = "openAiChatModel",

streamingChatModel = "openAiStreamingChatModel",

chatMemoryProvider = "chatMemoryProvider",

contentRetriever = "contentRetriever",

tools = {"sysTools"})

public interface AiSqlAssistantService {

String chat(String message);

@SystemMessage("👉 将文本改写成类似小红书的 Emoji 风格")

Flux<String> chatWithStream(@MemoryId String memoryId, @UserMessage String message);

@SystemMessage("请在用户提供的文本中提取出人员信息")

Person extractPerson(@UserMessage String message);

@UserMessage("需要你帮我mock人员姓名, 帮我生成{{total}}个")

List<String> mockUsername(@V("total") Integer total);

@SystemMessage("你是一名sql分析专家我会将sql相关的ddl给你, 需要你根据ddl生成合理且可执行的sql语句并返回")

String chatWithSql(@MemoryId String memoryId, @UserMessage String message);

}

5.2 Controller

package com.ldx.langchaintest.controller;

import com.ldx.langchaintest.domain.Person;

import com.ldx.langchaintest.service.AiSqlAssistantService;

import dev.langchain4j.data.document.Document;

import dev.langchain4j.data.document.DocumentSplitter;

import dev.langchain4j.data.document.loader.ClassPathDocumentLoader;

import dev.langchain4j.data.document.splitter.DocumentByRegexSplitter;

import dev.langchain4j.data.embedding.Embedding;

import dev.langchain4j.data.segment.TextSegment;

import dev.langchain4j.model.embedding.EmbeddingModel;

import dev.langchain4j.model.output.Response;

import dev.langchain4j.store.embedding.EmbeddingStore;

import lombok.RequiredArgsConstructor;

import lombok.extern.slf4j.Slf4j;

import org.springframework.web.bind.annotation.GetMapping;

import org.springframework.web.bind.annotation.PathVariable;

import org.springframework.web.bind.annotation.RequestMapping;

import org.springframework.web.bind.annotation.RequestParam;

import org.springframework.web.bind.annotation.RestController;

import reactor.core.publisher.Flux;

import java.io.IOException;

import java.util.List;

/**

* ai controller

* @author ludangxin

* @date 2025/5/16

@Slf4j

@RestController

@RequestMapping("/ai/chat")

@RequiredArgsConstructor

public class AiServiceController {

private final EmbeddingModel embeddingModel;

private final EmbeddingStore<TextSegment> embeddingStore;

private final AiSqlAssistantService aiSqlAssistantService;

@GetMapping("/test")

public String test() {

return aiSqlAssistantService.chat("你是谁");

}

@GetMapping

public String chat(@RequestParam String userMessage) {

return aiSqlAssistantService.chat(userMessage);

}

@GetMapping(value = "/{id}/stream/memory", produces = "text/stream;charset=utf-8")

public Flux<String> streamMemory(@PathVariable String id, @RequestParam String userMessage) {

final Flux<String> chatResponse = aiSqlAssistantService.chatWithStream(id, userMessage);

return chatResponse

.doOnNext(partial -> log.info("chat stream partial data:{}", partial))

.doOnError(e -> log.error("stream output error", e))

.doOnComplete(() -> log.info("chat stream complete"));

}

@GetMapping("/extract/person")

public Person extractPerson(@RequestParam String userMessage) {

return aiSqlAssistantService.extractPerson(userMessage);

}

@GetMapping("/mock/username")

public List<String> mockUsername(@RequestParam(defaultValue = "0") Integer total) {

return aiSqlAssistantService.mockUsername(total);

}

@GetMapping(value = "/embedding")

public String aiEmbedding() throws IOException {

final Document document = ClassPathDocumentLoader.loadDocument("student_ddl.sql");

// 创建 SQL 感知的文档分割器

DocumentSplitter splitter = new DocumentByRegexSplitter(";",";",

2000, // 最大片段长度

100 // 重叠长度

);

final List<TextSegment> textSegments = splitter.split(document);

final Response<List<Embedding>> embedResult = embeddingModel.embedAll(textSegments);

final List<Embedding> content = embedResult.content();

embeddingStore.addAll(content, textSegments);

return "success";

}

@GetMapping(value = "/{id}/sql/generate")

public String aiEmbedding(@PathVariable String id, @RequestParam String userMessage) {

return aiSqlAssistantService.chatWithSql(id, userMessage);

}

6. 测试

6.1 测试chat

6.2 测试tool

6.3 测试流式&聊天记忆

第一次会话:

后端日志如下

第二次会话:

6.4 测试抽取用户信息

6.5 测试mock

6.6 测试embedding

数据库中信息如下:

6.7 测试text2sql

7. 小结

本章通过使用SpringBoot实现快速集成LangChain4J, 通过简单的配置实现了AI的调用, 总体使用感受还不错, 虽然是刚发布的正式版但是整体的集成、方法调用都挺丝滑的, 到这里这关于LangChain4J的全部内容已经完结了, 后续会出个SpringAI正式版的体验对比,感兴趣的可以关注下.

8. 源码

测试过程中的代码已全部上传至github, 欢迎点赞收藏仓库地址: https://github.com/ludangxin/langchain4j-test

PGVector

version: '3'

services:

pgvector:

container_name: pgvector

restart: "no"

image: pgvector/pgvector:0.8.0-pg17

privileged: true

ports:

- 5431:5432

environment:

POSTGRES_USER: root

POSTGRES_PASSWORD: 123456

PGDATA: /var/lib/postgresql/data/

volumes:

- ./data:/var/lib/postgresql/data/

from:https://www.cnblogs.com/ludangxin/p/18937440

M	T	W	T	F	S	S
« Mar
		1	2	3	4	5
6	7	8	9	10	11	12
13	14	15	16	17	18	19
20	21	22	23	24	25	26
27	28	29	30