DeepSeek API 调用 - Spring Boot 实现

1. 项目依赖

在 pom.xml 中添加以下依赖：

<dependencies>
    <dependency>
        <groupId>org.springframework.boot</groupId>
        <artifactId>spring-boot-starter-webflux</artifactId>
    </dependency>
    <dependency>
        <groupId>org.projectlombok</groupId>
        <artifactId>lombok</artifactId>
    </dependency>
    <dependency>
        <groupId>com.fasterxml.jackson.core</groupId>
        <artifactId>jackson-databind</artifactId>
    </dependency>
</dependencies>

2. 项目结构

deepseek-project/
├── src/main/java/com/example/deepseek/
│   ├── DeepSeekApplication.java
│   ├── config/
│   │   └── DeepSeekConfig.java
│   ├── model/
│   │   ├── ChatRequest.java
│   │   ├── ChatResponse.java
│   │   └── Message.java
│   └── service/
│       └── DeepSeekService.java
└── conversation.txt

3. 完整代码实现

3.1 配置类 DeepSeekConfig.java

package com.example.deepseek.config;

import org.springframework.beans.factory.annotation.Value;
import org.springframework.context.annotation.Configuration;

@Configuration
@Getter
public class DeepSeekConfig {
    @Value("${deepseek.api.url}")
    private String apiUrl;

    @Value("${deepseek.api.key}")
    private String apiKey;
}

3.2 请求/响应模型

Message.java:

package com.example.deepseek.model;

import lombok.Data;

@Data
public class Message {
    private String role;
    private String content;
}

ChatRequest.java:

package com.example.deepseek.model;

import lombok.Data;
import java.util.List;

@Data
public class ChatRequest {
    private String model = "deepseek-ai/DeepSeek-V3";
    private List<Message> messages;
    private boolean stream = true;
    private int max_tokens = 2048;
    private double temperature = 0.7;
    private double top_p = 0.7;
    private int top_k = 50;
    private double frequency_penalty = 0.5;
    private int n = 1;
    private ResponseFormat response_format = new ResponseFormat("text");

    @Data
    public static class ResponseFormat {
        private String type;
        
        public ResponseFormat(String type) {
            this.type = type;
        }
    }
}

ChatResponse.java:

package com.example.deepseek.model;

import lombok.Data;
import java.util.List;

@Data
public class ChatResponse {
    private List<Choice> choices;

    @Data
    public static class Choice {
        private Delta delta;
    }

    @Data
    public static class Delta {
        private String content;
    }
}

3.3 服务类 DeepSeekService.java

package com.example.deepseek.service;

import com.example.deepseek.config.DeepSeekConfig;
import com.example.deepseek.model.ChatRequest;
import com.example.deepseek.model.ChatResponse;
import com.example.deepseek.model.Message;
import com.fasterxml.jackson.databind.ObjectMapper;
import lombok.RequiredArgsConstructor;
import org.springframework.stereotype.Service;
import org.springframework.web.reactive.function.client.WebClient;
import reactor.core.publisher.Flux;

import java.io.FileWriter;
import java.io.IOException;
import java.io.PrintWriter;
import java.time.LocalDateTime;
import java.time.format.DateTimeFormatter;
import java.util.Collections;
import java.util.Scanner;

@Service
@RequiredArgsConstructor
public class DeepSeekService {
    private final DeepSeekConfig config;
    private final WebClient.Builder webClientBuilder;
    private final ObjectMapper objectMapper = new ObjectMapper();

    public void startInteractiveChat() {
        try (Scanner scanner = new Scanner(System.in);
             PrintWriter fileWriter = new PrintWriter(new FileWriter("conversation.txt", true))) {

            while (true) {
                System.out.print("\n请输入您的问题 (输入 q 退出): ");
                String question = scanner.nextLine().trim();

                if ("q".equalsIgnoreCase(question)) {
                    System.out.println("程序已退出");
                    break;
                }

                // 保存问题
                saveToFile(fileWriter, question, true);

                // 发起对话请求
                Flux<String> responseFlux = sendChatRequest(question);

                StringBuilder fullResponse = new StringBuilder();
                responseFlux
                    .doOnNext(chunk -> {
                        System.out.print(chunk);
                        fullResponse.append(chunk);
                    })
                    .doOnComplete(() -> {
                        // 保存完整回复
                        saveToFile(fileWriter, fullResponse.toString(), false);
                        System.out.println("\n----------------------------------------");
                        fileWriter.println("\n----------------------------------------");
                        fileWriter.flush();
                    })
                    .blockLast();
            }
        } catch (IOException e) {
            e.printStackTrace();
        }
    }

    private Flux<String> sendChatRequest(String question) {
        ChatRequest request = new ChatRequest();
        Message userMessage = new Message();
        userMessage.setRole("user");
        userMessage.setContent(question);
        request.setMessages(Collections.singletonList(userMessage));

        return webClientBuilder.build()
            .post()
            .uri(config.getApiUrl())
            .header("Authorization", "Bearer " + config.getApiKey())
            .header("Content-Type", "application/json")
            .bodyValue(request)
            .retrieve()
            .bodyToFlux(String.class)
            .filter(line -> line.startsWith("data: ") && !line.equals("data: [DONE]"))
            .map(line -> {
                try {
                    String jsonStr = line.substring(6);
                    ChatResponse response = objectMapper.readValue(jsonStr, ChatResponse.class);
                    return response.getChoices().get(0).getDelta().getContent();
                } catch (Exception e) {
                    return "";
                }
            })
            .filter(content -> !content.isEmpty());
    }

    private void saveToFile(PrintWriter fileWriter, String content, boolean isQuestion) {
        String timestamp = LocalDateTime.now().format(DateTimeFormatter.ofPattern("yyyy-MM-dd HH:mm:ss"));
        if (isQuestion) {
            fileWriter.printf("\n[%s] Question:\n%s\n\n[%s] Answer:\n", timestamp, content, timestamp);
        } else {
            fileWriter.print(content);
        }
        fileWriter.flush();
    }
}

3.4 主应用类 DeepSeekApplication.java

package com.example.deepseek;

import com.example.deepseek.service.DeepSeekService;
import org.springframework.boot.SpringApplication;
import org.springframework.boot.autoconfigure.SpringBootApplication;
import org.springframework.context.ConfigurableApplicationContext;

@SpringBootApplication
public class DeepSeekApplication {
    public static void main(String[] args) {
        ConfigurableApplicationContext context = SpringApplication.run(DeepSeekApplication.class, args);
        DeepSeekService deepSeekService = context.getBean(DeepSeekService.class);
        deepSeekService.startInteractiveChat();
    }
}

3.5 配置文件 application.properties

deepseek.api.url=https://api.siliconflow.cn/v1/chat/completions
deepseek.api.key=YOUR_API_KEY

4. 代码详解

4.1 关键特性

1. 使用 Spring WebFlux 的响应式编程模型
2. 流式处理 API 响应
3. 文件记录对话
4. 错误处理和异常管理

4.2 主要组件

• DeepSeekConfig: 管理 API 配置
• DeepSeekService: 处理对话逻辑和 API 交互
• 模型类: 定义请求和响应结构

5. 使用方法

1. 替换 application.properties 中的 YOUR_API_KEY
2. 运行 DeepSeekApplication
3. 在控制台输入问题
4. 输入 ‘q’ 退出程序
5. 查看 conversation.txt 获取对话记录

6. 性能和可扩展性

• 使用响应式编程提高并发性能
• 灵活的配置管理
• 易于扩展和定制

7. 注意事项

• 确保正确配置 API Key
• 处理网络异常
• 注意内存使用

总结

Spring Boot 实现提供了一个健壮、可扩展的 DeepSeek API 调用方案，利用响应式编程提供高效的流式对话体验。

立即体验

快来体验 DeepSeek：https://cloud.siliconflow.cn/i/vnCCfVaQ