传统的IO传输,服务端会一直等待客户端传来数据,不能做其他事,便处于阻塞状态,即阻塞IO。
通过把一个文件从客户端传输到服务端的案例来理解一下BIO:
首先是客户端的代码:
@Test
public void client(){
try {
//1.获取通道
SocketChannel clientChannel = SocketChannel.open(
new InetSocketAddress("127.0.0.1", 7979));
System.out.println("client is start connecting...");
FileChannel fileChannel = FileChannel.open(Paths.get("1.jpg"), StandardOpenOption.READ);
//2.分配指定大小的缓冲区
ByteBuffer buffer = ByteBuffer.allocate(1024);
//3.读取本地文件,并向服务端发送文件
while(fileChannel.read(buffer)!=-1){
buffer.flip();
clientChannel.write(buffer);
buffer.clear();
}
clientChannel.shutdownOutput();//告诉服务端数据已经发送完毕,否则客户端和服务端都会处于阻塞状态
//4.接收服务端的反馈
int len = 0;
while((len = clientChannel.read(buffer))!=-1){
buffer.flip();
System.out.println(new String(buffer.array(),0,len));
buffer.clear();
}
//5.关闭通道
clientChannel.close();
fileChannel.close();
} catch (IOException e) {
e.printStackTrace();
}
}
服务端代码如下:
@Test
public void server(){
try {
//1.获取通道
ServerSocketChannel serverChannel = ServerSocketChannel.open();
FileChannel outChannel = FileChannel.open(
Paths.get("bio.jpg"), StandardOpenOption.WRITE,StandardOpenOption.CREATE);
//2.绑定端口
serverChannel.bind(new InetSocketAddress(7979));
System.out.println("Server start in port:7979...");
//3.获取客户端连接
SocketChannel socketChannel = serverChannel.accept();
//4分配指定到小的缓冲区
ByteBuffer byteBuffer = ByteBuffer.allocate(1024);
//5.读取数据
while (socketChannel.read(byteBuffer)!=-1){
byteBuffer.flip();
outChannel.write(byteBuffer);
byteBuffer.clear();
}
//6.发送反馈给客户端
byteBuffer.put("服务端接收数据成功!".getBytes());
byteBuffer.flip();
socketChannel.write(byteBuffer);
//7.关闭通道
socketChannel.close();
outChannel.close();
serverChannel.close();
} catch (IOException e) {
e.printStackTrace();
}
}
总结一下使用NIO完成网络通信的三个核心:
1.通道(Channel):负责连接;
2.缓冲区(Buffer):负责数据的存取;
3.选择器(Selector):是SelectableChannel的多路复用器,用于监控SelectableChannel的IO状况
本文还未用到选择器,在nio编程中会使用到。