Java 接入 AI 接口

背景

如果你的项目使用 JDK 17+，推荐学习 Spring AI 框架。本篇以 JDK 8 项目为例，手搓 AI 接口的接入。

依赖引入

<!-- Gson for JSON processing -->
<dependency>
    <groupId>com.google.code.gson</groupId>
    <artifactId>gson</artifactId>
    <version>2.8.9</version>
</dependency>
<!-- HttpClient for making HTTP requests -->
<dependency>
    <groupId>org.apache.httpcomponents.client5</groupId>
    <artifactId>httpclient5</artifactId>
    <version>5.1.3</version>
</dependency>

AI 配置类

建议单独开一个配置类，后续可通过 Nacos 等方式动态更新：

public class AIConfig {
    private String apiKey;
    private String model;
    private String URL;
    private Double temperature;

    public AIConfig() {
        URL = "https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions";
        apiKey = "sk-xxx";
        model = "qwen-plus";
        temperature = 0.8;
    }
}

实体类

public class AIMessage {
    private String role;    // system / user
    private String content;
}

public class AIRequst {
    private String model;
    private List<AIMessage> messages;
    private double temperature;
}

核心调用实现

public ResultBody AIImport(String userId, String issue) {
    String key = "AI_" + userId;
    Object cacheObject = redisService.getCacheObject(key);
    String contextInfo = (cacheObject != null) ? (String) cacheObject : "";

    AIRequst openaiReq = new AIRequst();
    openaiReq.setModel(aiConfig.getModel());

    List<AIMessage> messages = new ArrayList<>();
    AIMessage systemMessage = new AIMessage();
    systemMessage.setRole("system");
    systemMessage.setContent("你是一个语音聊天助手，会讲冷笑话、鬼故事等。");
    messages.add(systemMessage);

    AIMessage userMessage = new AIMessage();
    userMessage.setRole("user");
    userMessage.setContent(String.format("上下文信息：\n%s\n\n用户问题：%s", contextInfo, issue));
    messages.add(userMessage);

    openaiReq.setMessages(messages);
    openaiReq.setTemperature(aiConfig.getTemperature());

    Gson gson = new Gson();
    String jsonRequest = gson.toJson(openaiReq);

    try {
        URL url = new URL(aiConfig.getURL());
        HttpURLConnection connection = (HttpURLConnection) url.openConnection();
        connection.setRequestMethod("POST");
        connection.setRequestProperty("Content-Type", "application/json");
        connection.setRequestProperty("Authorization", "Bearer " + aiConfig.getApiKey());
        connection.setDoOutput(true);

        try (OutputStream os = connection.getOutputStream()) {
            byte[] input = jsonRequest.getBytes(StandardCharsets.UTF_8);
            os.write(input, 0, input.length);
        }

        int responseCode = connection.getResponseCode();
        if (responseCode == 200) {
            try (BufferedReader br = new BufferedReader(
                    new InputStreamReader(connection.getInputStream(), StandardCharsets.UTF_8))) {
                StringBuilder response = new StringBuilder();
                String responseLine;
                while ((responseLine = br.readLine()) != null) {
                    response.append(responseLine.trim());
                }

                JsonObject jsonResponse = gson.fromJson(response.toString(), JsonObject.class);
                if (jsonResponse.getAsJsonArray("choices").isEmpty()) {
                    return ResultBody.error().message("AI 响应超时...");
                }

                JsonObject choice = jsonResponse.getAsJsonArray("choices").get(0).getAsJsonObject();
                JsonObject message = choice.getAsJsonObject("message");
                String answer = message.get("content").getAsString().trim();

                if (answer.isEmpty()) {
                    return ResultBody.error().message("AI 响应超时");
                }

                redisService.setCacheObject(key, issue, 300L, TimeUnit.MINUTES);
                return ResultBody.ok().message(answer);
            }
        } else {
            try (BufferedReader br = new BufferedReader(
                    new InputStreamReader(connection.getErrorStream(), StandardCharsets.UTF_8))) {
                StringBuilder errorResponse = new StringBuilder();
                String errorLine;
                while ((errorLine = br.readLine()) != null) {
                    errorResponse.append(errorLine.trim());
                }
                System.err.println("错误响应: " + errorResponse);
            }
            return ResultBody.error().message("AI 响应超时...");
        }
    } catch (IOException e) {
        e.printStackTrace();
        return ResultBody.error().message("AI 响应超时");
    }
}