# keveri-cloud

**Repository Path**: S_Yang/keveri-cloud

## Basic Information

- **Project Name**: keveri-cloud
- **Description**: 校招项目准备
- **Primary Language**: Unknown
- **License**: Apache-2.0
- **Default Branch**: master
- **Homepage**: https://book.keveri.cloud
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 1
- **Created**: 2024-05-22
- **Last Updated**: 2024-05-22

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

<h1 align="center" style="margin: 30px 0 30px; font-weight: bold;">Keveri Cloud</h1>
<h4 align="center" style="margin: 30px 0 30px; font-weight: bold;">云开发基底（自己业余瞎玩的～～）</h4>
<p align="center">
<a href="https://gitee.com/tian-haoran/ai-chat/blob/master/LICENSE"><img src="https://img.shields.io/badge/license-Apache--2.0-green"></a>
<a href="hhttps://gitee.com/tian-haoran/keveri-cloud"><img src="https://img.shields.io/badge/version-v1.0-blue"></a>
</p>

# Keveri Cloud
24校招准备的小项目，v1.0 发布，其中组件均已上传 Maven 大家可凭坐标检索，后续教程会更新到<a href= "https://book.keveri.cloud">Keveri 的春招之路</a>。全部免费!!!!!! <br>
这里面都是一些日常开发的一些基础组件，当然也可能是作者有病（造轮子的病哈哈，还有仿照轮子，主打一个学习～～～（骗面试官））<br>
**项目中网关部分来自对 Shenyu 网关的拆解**


## Chat4j

### 概述

1. 可复用多厂商 API （未来会逐渐支持）
2. 支持设置代理地址
3. 支持 SSE
4. 多厂商静态加载

### 接入 API

多厂商抽象不容易、慢慢来哈

- 对话问答
- 图片生成

### 如何使用？

#### 导入依赖

```xml

<dependency>
    <groupId>cloud.keveri.dev</groupId>
    <artifactId>keveri-chat4j-spring-boot-starter</artifactId>
    <version>latest-version</version>
</dependency>
```

#### 使用案例
##### 配置
```yaml
server:
  port: 8080

chat:
  read-timeout: 4300
  write-timeout: 4300
  connect-timeout: 4300
  retry-count: 0

  provider:
    chatgpt:
      key: sk-LIoETxxxxxxxxxxxxxxxxxxxxxxx
      # 如果使用代理则需要配置如下
      isProxy: true
      proxyUrl: https://proxy.openai.com/
```
#### API 使用
```java
package cloud.keveri.example.chat4j.interfaces;

import cloud.keveri.chat4j.api.ChatBlender;
import cloud.keveri.chat4j.api.Response;
import cloud.keveri.chat4j.api.entity.chat.CompletionRequest;
import cloud.keveri.chat4j.api.entity.chat.Message;
import cloud.keveri.chat4j.api.entity.image.ImageRequest;
import cloud.keveri.chat4j.core.factory.ChatFactory;
import cloud.keveri.chat4j.provider.SupplierType;
import cloud.keveri.chat4j.provider.chatgpt.listener.DefaultChatgptEventSourceListener;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.PostMapping;
import org.springframework.web.bind.annotation.RestController;
import org.springframework.web.servlet.mvc.method.annotation.ResponseBodyEmitter;

import javax.servlet.http.HttpServletResponse;
import java.util.ArrayList;
import java.util.List;

/**
 *  测试处理器
 * <p>
 * 作者：沈自在
 **/
@RestController
public class TestController {

    @PostMapping("/api/chatgpt/completion/not")
    public Response chatgpt_completion_not_stream() {

        ChatBlender chatBlender = ChatFactory.acquireChatBlender(SupplierType.CHATGPT);

        Message message = new Message("user", "讲个笑话吧");
        List<Message> messages = new ArrayList<>();
        messages.add(message);

        CompletionRequest request = CompletionRequest.builder()
                .messages(messages)
                .build();

        Response response = chatBlender.completions(request);
        System.out.println("response = " + response);
        return response;
    }

    @GetMapping("/api/chatgpt/completion/sse")
    public ResponseBodyEmitter sseEmitter(String question, HttpServletResponse response) {
        // 1. 准备请求

        response.setContentType("text/event-stream");
        response.setCharacterEncoding("UTF-8");
        response.setHeader("Cache-Control", "no-cache");


        Message message = Message.of(question);
        List<Message> messages = new ArrayList<>();
        messages.add(message);

        CompletionRequest request = CompletionRequest.builder()
                .messages(messages)
                .stream(true)
                .build();

        // 2. sse
        ResponseBodyEmitter emitter = new ResponseBodyEmitter(3 * 60 * 1000L);
        DefaultChatgptEventSourceListener listener = new DefaultChatgptEventSourceListener(emitter);
        //ConsoleChatgptEventSourceListener listener = new ConsoleChatgptEventSourceListener();

        // 3. sdk
        ChatBlender chatBlender = ChatFactory.acquireChatBlender(SupplierType.CHATGPT);
        chatBlender.completions(request, listener);

        return emitter;
    }

    @GetMapping("/api/chatgpt/image")
    public Response image(String prompt) {
        ChatBlender chatBlender = ChatFactory.acquireChatBlender(SupplierType.CHATGPT);
        ImageRequest request = ImageRequest.builder()
                .prompt(prompt)
                .build();
        return chatBlender.generateImage(request);
    }
}
```

## Response4j

### 概述
提供简化的响应包装 starter，现支持：
1. 响应包装
2. 跳过注解，并支持跳过逻辑注入
3. 自定义响应内容

### 如何使用
#### 导入依赖
```xml
        <dependency>
            <groupId>cloud.keveri.dev</groupId>
            <artifactId>keveri-response4j-spring-boot-starter</artifactId>
            <version>latest version</version>
        </dependency>
```
#### 使用案例
##### 配置

##### 代码实现
###### 跳过逻辑
```java
@Component
public class CustomSkip implements SkipWrapCheckFunction {
    @Override
    public boolean isSkip(@NonNull MethodParameter methodParameter,
                          @NonNull Class<? extends HttpMessageConverter<?>> converterType) {
        Method method = methodParameter.getMethod();
        assert method != null;
        Class<?> returnType = method.getReturnType();
        return returnType == TestController.User.class;
    }
}
```
###### 测试接口
```java
@Controller
@RequestMapping("/response4j/test")
public class TestController {

    /**
     *
     */
    @Data
    @AllArgsConstructor
    @NoArgsConstructor
    public static class User {
        String name = "沈自在";
        Integer age = 18;
    }

    /**
     *
     */
    @GetMapping
    public String hello() {
        return "你好";
    }

    /**
     *
     */
    @GetMapping("/user01")
    @ResponseBody
    public User user01() {
        return new User();
    }

    @GetMapping("/user02")
    @ResponseBody
    @SkipWrap
    public User user02() {
        return new User("田浩然", 23);
    }
}

```

## Keveri-Gateway

这是一个学习项目，后续教程更新到  <a href= "https://book.keveri.cloud">Keveri 的春招之路</a>