SpringBoot調(diào)用DeepSeek?API的完整操作指南

更新時間：2025年02月10日 08:17:48 作者：老大白菜

這篇文章主要為大家詳細介紹了SpringBoot調(diào)用DeepSeek?API的完整操作指南,文中的示例代碼講解詳細,感興趣的小伙伴可以跟隨小編一起學(xué)習(xí)一下

1. 項目依賴

在 pom.xml 中添加以下依賴：

<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. 項目結(jié)構(gòu)

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. 完整代碼實現(xiàn)

3.1 配置類 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 請求/響應(yīng)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 服務(wù)類 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請輸入您的問題 (輸入 q 退出): ");
                String question = scanner.nextLine().trim();

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

                // 保存問題
                saveToFile(fileWriter, question, true);

                // 發(fā)起對話請求
                Flux<String> responseFlux = sendChatRequest(question);

                StringBuilder fullResponse = new StringBuilder();
                responseFlux
                    .doOnNext(chunk -> {
                        System.out.print(chunk);
                        fullResponse.append(chunk);
                    })
                    .doOnComplete(() -> {
                        // 保存完整回復(fù)
                        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 主應(yīng)用類 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. 代碼詳解

4.1 關(guān)鍵特性

使用 Spring WebFlux 的響應(yīng)式編程模型

流式處理 API 響應(yīng)

文件記錄對話

錯誤處理和異常管理

4.2 主要組件

DeepSeekConfig: 管理 API 配置

DeepSeekService: 處理對話邏輯和 API 交互

模型類: 定義請求和響應(yīng)結(jié)構(gòu)

5. 使用方法

替換 application.properties 中的 YOUR_API_KEY

運行 DeepSeekApplication

在控制臺輸入問題

輸入 ‘q’ 退出程序

查看 conversation.txt 獲取對話記錄

6. 性能和可擴展性

使用響應(yīng)式編程提高并發(fā)性能

靈活的配置管理

易于擴展和定制

7. 注意事項

確保正確配置 API Key

處理網(wǎng)絡(luò)異常

注意內(nèi)存使用

總結(jié)

Spring Boot 實現(xiàn)提供了一個健壯、可擴展的 DeepSeek API 調(diào)用方案，利用響應(yīng)式編程提供高效的流式對話體驗。

到此這篇關(guān)于SpringBoot調(diào)用DeepSeek API的完整操作指南的文章就介紹到這了,更多相關(guān)SpringBoot調(diào)用DeepSeek API內(nèi)容請搜索腳本之家以前的文章或繼續(xù)瀏覽下面的相關(guān)文章希望大家以后多多支持腳本之家！

您可能感興趣的文章: