通道(Channel):用于源节点与目标节点的链接。在 Java NIO 中负责缓冲区中数据的传输。Channel 自己不存储数据,所以须要配合缓冲区进行传输数组
|--FileChannel 本地io
|--SocketChannel TCP网络io
|--ServerSocketChannel TCP网络io
|--DatagramChannel UDP网络io网络
本地 IO:FileInputStream、FileOutputStream、 RandomAccessFileapp
网络IO:Socket、ServerSocket、DatagramSocketdom
//利用通道完成文件的复制(非直接缓冲区) @Test public void test1(){ long start = System.currentTimeMillis(); FileInputStream fis = null; FileOutputStream fos = null; //①获取通道 FileChannel inChannel = null; FileChannel outChannel = null; try { fis = new FileInputStream("/Users/tentsuuhou/Desktop/01.mp4"); fos = new FileOutputStream("/Users/tentsuuhou/Desktop/09.mp4"); inChannel = fis.getChannel(); outChannel = fos.getChannel(); //②分配指定大小的缓冲区 ByteBuffer buf = ByteBuffer.allocate(1024); //③将通道中的数据存入缓冲区中 while(inChannel.read(buf) != -1){ buf.flip(); //切换读取数据的模式 //④将缓冲区中的数据写入通道中 outChannel.write(buf); buf.clear(); //清空缓冲区 } } catch (IOException e) { e.printStackTrace(); } finally { if(outChannel != null){ try { outChannel.close(); } catch (IOException e) { e.printStackTrace(); } } if(inChannel != null){ try { inChannel.close(); } catch (IOException e) { e.printStackTrace(); } } if(fos != null){ try { fos.close(); } catch (IOException e) { e.printStackTrace(); } } if(fis != null){ try { fis.close(); } catch (IOException e) { e.printStackTrace(); } } } long end = System.currentTimeMillis(); System.out.println("耗费时间为:" + (end - start)); }
我用的是mac系统,01.mp4视频25M左右,第一次复制时间1024ms,而后覆盖从新执行,大概左右在683ms,若是能够,你们能够试试1G以上的视频,看看须要多久工具
//使用直接缓冲区完成文件的复制(内存映射文件) @Test public void test2(){ long start = System.currentTimeMillis(); FileChannel inChannel = null; FileChannel outChannel= null; try { inChannel = FileChannel.open(Paths.get("/Users/tentsuuhou/Desktop/01.mp4"), StandardOpenOption.READ); outChannel = FileChannel.open(Paths.get("/Users/tentsuuhou/Desktop/08.mp4"), StandardOpenOption.WRITE, StandardOpenOption.READ, StandardOpenOption.CREATE); //内存映射文件 MappedByteBuffer inMappedBuf = inChannel.map(MapMode.READ_ONLY, 0, inChannel.size()); MappedByteBuffer outMappedBuf = outChannel.map(MapMode.READ_WRITE, 0, inChannel.size()); //直接对缓冲区进行数据的读写操做 byte[] dst = new byte[inMappedBuf.limit()]; inMappedBuf.get(dst); outMappedBuf.put(dst); } catch (IOException e){ e.printStackTrace(); } finally { try { inChannel.close(); } catch (IOException e) { e.printStackTrace(); } try { outChannel.close(); } catch (IOException e) { e.printStackTrace(); } } long end = System.currentTimeMillis(); System.out.println("耗费时间为:" + (end - start)); }
第一次:371ms,第二次104ms,证实直接缓冲区比间接缓冲区效率搞了不少。编码
直接字节缓冲区还能够经过FileChannel的map()方法将文件区域直接映射到内存中来建立。该方法返回 MappedByteBuffer 。Java 平台的实现有助于经过 JNI 从本机代码建立直接字节缓冲区。若是以上这些缓冲区 中的某个缓冲区实例指的是不可访问的内存区域,则试图访问该区域不会更改该缓冲区的内容,而且将会在 访问期间或稍后的某个时间致使抛出不肯定的异常。spa
@Test public void test7(){ long start = System.currentTimeMillis(); FileChannel in = null; FileChannel out = null; try { in = (FileChannel) Files.newByteChannel(Paths.get("/Users/tentsuuhou/Desktop/01.mp4"),StandardOpenOption.READ); out = (FileChannel) Files.newByteChannel(Paths.get("/Users/tentsuuhou/Desktop/07.mp4"),StandardOpenOption.READ,StandardOpenOption.WRITE,StandardOpenOption.CREATE); MappedByteBuffer inMap = in.map(MapMode.READ_ONLY,0,in.size()); MappedByteBuffer outMap = out.map(MapMode.READ_WRITE,0,in.size()); byte[] bytes = new byte[inMap.limit()]; //直接对缓冲区进行数据的读写操做 inMap.get(bytes); outMap.put(bytes); } catch (IOException i){ i.printStackTrace(); } finally { try { in.close(); } catch (IOException e) { e.printStackTrace(); } try { out.close(); } catch (IOException e) { e.printStackTrace(); } } long end = System.currentTimeMillis(); System.out.println(end-start); }
一样使用的是直接缓冲区,速度和上面基本一致,只不过这个是用newByteChannel() 方法建立通道。code
4.1 transferFrom()视频
4.2 transferTo()blog
//通道之间的数据传输(直接缓冲区) @Test public void test3(){ FileChannel inChannel = null; FileChannel outChannel = null; try { inChannel = FileChannel.open(Paths.get("/Users/tentsuuhou/Desktop/01.mp4"), StandardOpenOption.READ); outChannel = FileChannel.open(Paths.get("/Users/tentsuuhou/Desktop/06.mp4"), StandardOpenOption.WRITE, StandardOpenOption.READ, StandardOpenOption.CREATE); //使用下面transferTo 或者 transferFrom 均可以达到目的 //inChannel.transferTo(0, inChannel.size(), outChannel); outChannel.transferFrom(inChannel, 0, inChannel.size()); } catch (IOException e){ e.printStackTrace(); } finally { try { inChannel.close(); } catch (IOException e) { e.printStackTrace(); } try { outChannel.close(); } catch (IOException e) { e.printStackTrace(); } } }
5.1 分散读取(Scattering Reads):将通道中的数据分散到多个缓冲区中
5.2 汇集写入(Gathering Writes):将多个缓冲区中的数据汇集到通道中
//分散和汇集 @Test public void test4() throws IOException{ RandomAccessFile raf1 = new RandomAccessFile("1.txt", "rw"); //1. 获取通道 FileChannel channel1 = raf1.getChannel(); //2. 分配指定大小的缓冲区 ByteBuffer buf1 = ByteBuffer.allocate(100); ByteBuffer buf2 = ByteBuffer.allocate(1024); //3. 分散读取 ByteBuffer[] bufs = {buf1, buf2}; channel1.read(bufs); for (ByteBuffer byteBuffer : bufs) { byteBuffer.flip(); } System.out.println(new String(bufs[0].array(), 0, bufs[0].limit())); System.out.println("-----------------"); System.out.println(new String(bufs[1].array(), 0, bufs[1].limit())); //4. 汇集写入 RandomAccessFile raf2 = new RandomAccessFile("2.txt", "rw"); FileChannel channel2 = raf2.getChannel(); channel2.write(bufs); }
没有什么效率的问题,只是记住有这么一个功能就行了
6.1 编码:字符串 -> 字节数组
6.2 解码:字节数组 -> 字符串
//字符集 @Test public void test6() throws IOException{ Charset cs1 = Charset.forName("utf-8"); //获取编码器 CharsetEncoder ce = cs1.newEncoder(); //获取解码器 CharsetDecoder cd = cs1.newDecoder(); CharBuffer cBuf = CharBuffer.allocate(1024); cBuf.put("我爱开源中国,做者MuJiuTian!"); cBuf.flip(); //编码 ByteBuffer bBuf = ce.encode(cBuf); for (int i = 0; i < 36; i++) { System.out.println(bBuf.get()); } //解码 bBuf.flip(); CharBuffer cBuf2 = cd.decode(bBuf); System.out.println(cBuf2.toString()); System.out.println("------------------------------------------------------"); Charset cs2 = Charset.forName("utf-8"); bBuf.flip(); CharBuffer cBuf3 = cs2.decode(bBuf); System.out.println(cBuf3.toString()); }