ReplayingDecoder是ByteToMessageDecoder的一个特殊变体,ReplayingDecoder继承了ByteToMessageDecoder。ReplayingDecoder和ByteToMessageDecoder最大的不同在于,ReplayingDecoder允许让你实现decode()方法,就像已经接收到所有所需的字节,而不用去检查所需字节的可用性。
在自定义协议传递ByteBuf格式
head通常是4个字节长度,读取出来的int数字表示要在body读的字节长度。
下面看一下ReplayingDecoder和ByteToMessageDecoder具体的区别。
public class IntegerHeaderFrameDecoder extends ByteToMessageDecoder {
@Override
protected void decode( ChannelHandlerContext ctx,
ByteBuf buf, List<Object> out) throws Exception {
if (buf.readableBytes() < 4) {
return;
}
buf.markReaderIndex();
int length = buf.readInt();
if (buf.readableBytes() < length) {
buf.resetReaderIndex();
return;
}
out.add(buf.readBytes(length));
}
}
int length = buf.readInt();将readerindex移动到body的起始位置。
###ReplayingDecoder的实现
public class IntegerHeaderFrameDecoder
extends ReplayingDecoder<Void> {
protected void decode(ChannelHandlerContext ctx,
ByteBuf buf) throws Exception {
out.add(buf.readBytes(buf.readInt()));
}
}
ReplayingDecoder是怎么做到的呢?
ReplayingDecoder传递一个专门的ByteBuf实现,当缓冲区中没有足够的数据时,会抛出某种类型的一个Error。在上面的IntegerHeaderFrameDecoder,当调用buf.readInt()时,只能假定缓冲区有4个或更多的字节。如果,缓冲区中确实有4个字节,会返回你所期望的integer header。否则,会抛出这个Error。如果ReplayingDecoder捕获了这个Error,会重置缓冲区的readerindex到缓冲区的开始位置,并且当更多的数据到达缓冲区时,会再次调用decode()方法。
需要注意一些限制:如果网络缓慢而且消息格式复杂,性能会变差,在这种情况,你的解码器不得不解码同一部分的信息。解码一条信息,decode()方法可以被多次调用。如下代码不能正常工作:
public class MyDecoder extends ReplayingDecoder<Void> {
private final Queue<Integer> values = new LinkedList<Integer>();
@Override
public void decode(.., ByteBuf buf, List<Object> out) throws Exception {
// A message contains 2 integers.
values.offer(buf.readInt());
values.offer(buf.readInt());
// This assertion will fail intermittently since values.offer()
// can be called more than two times!
assert values.size() == 2;
out.add(values.poll() + values.poll());
}
}
正确的做法如下:
public class MyDecoder extends ReplayingDecoder<Void> {
private final Queue<Integer> values = new LinkedList<Integer>();
@Override
public void decode(.., ByteBuf buf, List<Object> out) throws Exception {
// Revert the state of the variable that might have been changed
// since the last partial decode.
values.clear();
// A message contains 2 integers.
values.offer(buf.readInt());
values.offer(buf.readInt());
// Now we know this assertion will never fail.
assert values.size() == 2;
out.add(values.poll() + values.poll());
}
}
复杂解码器的性能因为checkpoint()方法显著的提高。checkpoint()方法更新缓冲区的“初始”位置,ReplayingDecoder将回滚缓冲区的readerIndex到最后调用checkpoint()方法的位置。
管理解码器状态的最简单方法是创建表示解码器当前状态的Enum类型,并在状态改变时调用checkpoint(T)方法。你可以有许多你想要的状态取决于你想解码的信息的复杂性:
public enum MyDecoderState {
READ_LENGTH,
READ_CONTENT;
}
public class IntegerHeaderFrameDecoder
extends ReplayingDecoder<MyDecoderState> {
private int length;
public IntegerHeaderFrameDecoder() {
// Set the initial state.
super(MyDecoderState.READ_LENGTH);
}
@Override
protected void decode( ChannelHandlerContext ctx,
ByteBuf buf, List<Object> out) throws Exception {
switch (state()) {
case READ_LENGTH:
length = buf.readInt();
checkpoint(MyDecoderState.READ_CONTENT);
case READ_CONTENT:
ByteBuf frame = buf.readBytes(length);
checkpoint(MyDecoderState.READ_LENGTH);
out.add(frame);
break;
default:
throw new Error("Shouldn't reach here.");
}
}
}
调用无参的checkpoint()
public class IntegerHeaderFrameDecoder
extends ReplayingDecoder<Void> {
private boolean readLength
private int length;
Override
protected void decode( ChannelHandlerContext ctx,
ByteBuf buf, List<Object> out) throws Exception {
if (!readLength) {
length = buf.readInt();
readLength = true;
checkpoint();
}
if (readLength) {
ByteBuf frame = buf.readBytes(length);
readLength = false;
checkpoint();
out.add(frame);
}
}
}