一、前言
上一节我们在nl2sql技术实现自动sql生成介绍了自然语言转sql的基本思路,今天我们通过实操来解决这一技术实现,通过Spring AI Alibaba开源的nl2sql框架进行这块技术的讲解。
二、环境要求
jdk:17+
spring boot:3.4.0以上
spring ai:1.0.0
spring ai alibaba:1.0.0.3
mysql:5.7+
三、技术实现
1、maven依赖
<dependency> <groupId>com.alibaba.cloud.ai</groupId> <artifactId>spring-ai-alibaba-starter-nl2sql</artifactId> <exclusions> <exclusion> <groupId>com.alibaba.cloud.ai</groupId> <artifactId>spring-ai-alibaba-starter-dashscope</artifactId> </exclusion> </exclusions> </dependency> <dependency> <groupId>org.springframework.boot</groupId> <artifactId>spring-boot-starter-webflux</artifactId> </dependency> <dependency> <groupId>org.springframework.ai</groupId> <artifactId>spring-ai-advisors-vector-store</artifactId> </dependency>
上面依赖引入了alibaba的nl2sql、spring boot webflux和向量数据库的组件
2、application配置
配置Mysql数据源、chatbi数据源和openai使用模型和向量库,LLM模型、文本嵌入模型以及向量库使用的都是阿里提供的模型,可根据实际情况进行调整,比如使用其他开源的模型或者向量库。
spring: datasource: url: jdbc:mysql://127.0.0.1:3306/orderdb?useUnicode=true&characterEncoding=utf8&zeroDateTimeBehavior=convertToNull&useSSL=false&serverTimezone=GMT%2B8 driverClassName: com.mysql.cj.jdbc.Driver username: root password: 123456 ai: openai: base-url: https://dashscope.aliyuncs.com/compatible-mode api-key: your-key model: qwen3-max embedding: model: text-embedding-v4 vectorstore: analytic: enabled: false collectName: chatBi regionId: cn-hangzhou dbInstanceId: '' managerAccount: '' managerAccountPassword: '' namespace: chat namespacePassword: chat defaultTopK: 6 defaultSimilarityThreshold: 0.75 accessKeyId: '' accessKeySecret: '' chatBi: dbConfig: url: jdbc:mysql://127.0.0.1:3306/orderdb?useUnicode=true&characterEncoding=utf8&zeroDateTimeBehavior=convertToNull&useSSL=false&serverTimezone=GMT%2B8 username: root password: 123456 schema: orderdb connection-type: jdbc dialect-type: mysql
3、测试验证
写一个Nl2sql的接口类,设置chatbi的数据源、表信息,将以上信息通过向量库服务通过文本嵌入模型存储到向量库中,便于生成sql时进行检索。代码如下:
java
import com.alibaba.cloud.ai.connector.config.DbConfig;
import com.alibaba.cloud.ai.request.SchemaInitRequest;
import com.alibaba.cloud.ai.service.simple.SimpleNl2SqlService;
import com.alibaba.cloud.ai.service.simple.SimpleVectorStoreService;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.web.bind.annotation.PostMapping;
import org.springframework.web.bind.annotation.RequestBody;
import org.springframework.web.bind.annotation.RestController;
import java.util.Arrays;
@RestController
public class SimpleChatController {
@Autowired
private SimpleNl2SqlService simpleNl2SqlService;
@Autowired
private SimpleVectorStoreService simpleVectorStoreService;
@Autowired
private DbConfig dbConfig;
@PostMapping("/simpleChat")
public String simpleNl2Sql(@RequestBody String input) throws Exception {
SchemaInitRequest schemaInitRequest = new SchemaInitRequest();
schemaInitRequest.setDbConfig(dbConfig);
schemaInitRequest.setTables(Arrays.asList("orders"));
simpleVectorStoreService.schema(schemaInitRequest);
return simpleNl2SqlService.nl2sql(input);
}
}
启动类配置下alibaba注入的package:
java
import org.springframework.boot.autoconfigure.SpringBootApplication;
import org.springframework.boot.builder.SpringApplicationBuilder;
@SpringBootApplication(scanBasePackages = {"com.alibaba.cloud.ai", "com.llm.ai.spring.demo.nl2sql"})
public class Nl2SqlApplication {
public static void main(String[] args) {
new SpringApplicationBuilder(Nl2SqlApplication.class).run(args);
}
}
在postman上发起请求,请求如下:

通过如上发起的请求,可以看到nl2sql根据用户的userMessage自动生成了统计的sql,具体的工作原理如上一篇的介绍的一样,下一篇我将通过spring ai框架实现自定义开发的nl2sql组件,敬请关注,今天就介绍到这里。