花了半天的时间写了个简单的rpc框架,是因为我最初看dubbo源码的时候发现dubbo虽然看起来很庞大,但是隐隐约约总感觉,其实其绝大多数功能,都是基于可扩张性和服务治理的需要而编写的。我看过dubbo和grpc的源码,这两个都是非常优秀的rpc框架,但是为了让初学rpc的同学能够快速抓住rpc的核心,所以写了这篇文章,希望看过的同学,再次去看dubbo的源码的时候,能够抓住这个核心去看。
一:rpc协议的接口
RpcProtocol.java

public interface RpcProtocol  {
 void export(int port);
 Object refer(Class inrerfaceClass,String host, int port);

}

这个接口类只提供两个接口,那是因为对于一个rpc框架来说,本质上就只需要两个接口,一个是consumer引用provider的服务,一个是provider接收到consumer的请求之后对外暴露服务。
下面是具体的实现。代码不复杂,可以直接复制到idea,慢慢调试

二:rpc协议的具体实现

RpcCore.java

public class RpcCore implements RpcProtocol{

private Socket socket;
private ObjectOutputStream objectOutputStream;
private ObjectInputStream objectInputStream;

private ServerSocket serverSocket;
private Map<String,List<Object>> services=new ConcurrentHashMap<String, List<Object>>();
private Map<String,Map<String,Object>> interfaceAtrributes=new ConcurrentHashMap<>();

@Override
public void export(int port){
    start(port);
}

@Override
public Object refer(final Class interfaceClass,String host, int port){
    connect(host,port);
    return Proxy.newProxyInstance(interfaceClass.getClassLoader(), new Class[]{interfaceClass},
            new InvocationHandler() {
                @Override
                public Object invoke(Object proxy, Method method, Object[] args) throws Throwable {
                    String interfaceName=interfaceClass.getName();
                    String fullName= (String) interfaceAtrributes.get(interfaceName).get("fullName");
                    return get(fullName,method,args);
                }
            });
}

public Object get(String interfaceFullName,Method method,Object[] parames){
    Object result=null;
    try {
        objectOutputStream.writeUTF(interfaceFullName);
        objectOutputStream.writeUTF(method.getName());
        objectOutputStream.writeObject(method.getParameterTypes());
        objectOutputStream.writeObject(parames);
        objectOutputStream.flush();
        result=objectInputStream.readObject();
    } catch (IOException e) {
        e.printStackTrace();
    } catch (ClassNotFoundException e) {
        e.printStackTrace();
    } finally {
        try {
            if (objectOutputStream!=null) {
                objectOutputStream.close();
            }
            if (objectInputStream!=null) {
                objectInputStream.close();
            }
        } catch (IOException e) {
            e.printStackTrace();
        }
    }
    return result;

}

private void start(int port) {
    try {
        serverSocket = new ServerSocket();
        serverSocket.bind(new InetSocketAddress("localhost", port));
        init();
    } catch (IOException e) {
        e.printStackTrace();
    }
    while (true) {
        System.out.println("server has started success port is --->"+port);
        Socket socket = null;
        try {
            socket = serverSocket.accept();
            new Thread(new Processsor(socket,services)).start();
        } catch (IOException e) {
            e.printStackTrace();
        }

    }

}

public  void init(){
    RpcDemo rpcDemo=new RpcDemoImplProvider();
    String group="rpcDemo";
    String version="1.0.0";
    String fullName=RpcDemo.class.getName()+"&"+group+"&"+version;
    List<Object> rpcDemoInstances=services.get(fullName);
    if (rpcDemoInstances==null){
        rpcDemoInstances=new ArrayList();
        rpcDemoInstances.add(rpcDemo);
    }
    services.put(fullName,rpcDemoInstances);
}

public void connect(String host, int port) {
    try {
        storeInterface();
        socket = new Socket();
        socket.connect(new InetSocketAddress(host, port));
        objectOutputStream=new ObjectOutputStream(socket.getOutputStream());
        objectInputStream=new ObjectInputStream(socket.getInputStream());
    } catch (IOException e) {
        e.printStackTrace();
    }

}

private void storeInterface(){
    String group="rpcDemo";
    String version="1.0.0";
    String fullName=RpcDemo.class.getName()+"&"+group+"&"+version;
    Map<String,Object> attributes=interfaceAtrributes.get(fullName);
    if (attributes==null){
        attributes=new ConcurrentHashMap(100);
        attributes.put("group",group);
        attributes.put("version",version);
        attributes.put("fullName",fullName);
    }
    interfaceAtrributes.put(RpcDemo.class.getName(),attributes);

}

class Processsor implements Runnable {
    private Socket socket;
    private ObjectInputStream objectInputStream;
    private ObjectOutputStream objectOutputStream;
    private Map<String,List<Object>> services;
    private Processsor(Socket socket,Map<String,List<Object>> services) {
        this.socket = socket;
        this.services=services;
    }

    @Override
    public void run() {
        System.out.println((((InetSocketAddress) socket.getRemoteSocketAddress()).getPort()));

        try {
            objectInputStream = new ObjectInputStream(socket.getInputStream());
            objectOutputStream = new ObjectOutputStream(socket.getOutputStream());
            String interfaceFullName=objectInputStream.readUTF();
            String methodName=objectInputStream.readUTF();
            Class[] parameTypes= (Class[]) objectInputStream.readObject();
            Object[] objects= (Object[]) objectInputStream.readObject();
            String interfaceName=interfaceFullName.split("&")[0];
            Class service=Class.forName(interfaceName);
            Method method=service.getMethod(methodName,parameTypes);
            Object instances=services.get(interfaceFullName).get(0);
            Object result = method.invoke(instances, objects);
            objectOutputStream.writeObject(result);
            objectOutputStream.flush();
            objectOutputStream.close();
            objectInputStream.close();
        } catch (Exception e) {
            e.printStackTrace();
        }

    }
}

}

三:rpc测试的接口
所谓接口说白了就是协议,与http,mqtt等其他的协议本质上没什么区别,只不过rpc是强依赖,而后两个是弱依赖而已,另外之所以把实体作为内部类,是为了表达一种思想,rpc的实体类和异常都是协议的一部分,应该将他们放到一起。
Rpcdemo.java

public interface RpcDemo {
Student getStudent(Integer id,String name);
class Student implements Serializable{
    public long id;
    public String name;
    public int age;
    public boolean man;

    public Student(long id, String name, int age, boolean man) {
        this.id = id;
        this.name = name;
        this.age = age;
        this.man = man;
    }

    public long getId() {
        return id;
    }

    public void setId(long id) {
        this.id = id;
    }

    public String getName() {
        return name;
    }

    public void setName(String name) {
        this.name = name;
    }

    public int getAge() {
        return age;
    }

    public void setAge(int age) {
        this.age = age;
    }

    public boolean isMan() {
        return man;
    }

    public void setMan(boolean man) {
        this.man = man;
    }

    @Override
    public String toString() {
        return "Student{" +
                "id=" + id +
                ", name='" + name + '\'' +
                ", age=" + age +
                ", man=" + man +
                '}';
    }
}

}

四:接口的实现:

RpcDemoImplProvider.java

public class RpcDemoImplProvider implements RpcDemo{
public Student getStudent(Integer id,String name){
    return new Student(1234,"zhangsan",20,true);
}

public static void main(String[] args) {
    RpcCore rpcCore=new RpcCore();
    rpcCore.export(8087);
}

}

五:RpcDemoConsumer.java

消费端。

public class RpcDemoConsumer {
public static void main(String[] args) {
    RpcCore rpcCore=new RpcCore();
    RpcDemo rpcDemo = (RpcDemo)rpcCore.refer(RpcDemo.class, "127.0.0.1", 8087);

    System.out.println(" 远程调用成功");
    System.out.println("返回的结果是---->"+rpcDemo.getStudent(111,"zhangsan"));
}

}

六:先启动RpcDemoImplProvider
三百行代码完成一个简单的rpc框架

七:启动RpcDemoConsumer
三百行代码完成一个简单的rpc框架

总共也就250行代码左右。over