Netty框架编程之字节序大小端问题研究

1. 背景

netty提供的LengthFieldBasedFrameDecoder解码器中，有个获取报文头长度方法getUnadjustedFrameLength，该方法有个参数ByteOrder，用来指定字节序是BIG_ENDIAN还是LITTLE_ENDIAN。
BIG_ENDIAN：In this order, the bytes of a multibyte value are ordered from most significant to least significant
LITTLE_ENDIAN：In this order, the bytes of a multibyte value are ordered from least significant to most significant

2. 代码分析

• 报文说明
  代码中使用“Hello world!”字符为例，该字符占用12字节，报文头占2字节，共14字节，如下所示：

0	1	2	3	4	5	6	7	8	9	10	11	12	13
0X00	0X0C	H	e	l	l	o		w	o	r	l	d	!

• 代码测试

import java.nio.ByteOrder;

import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

import io.netty.buffer.ByteBuf;
import io.netty.buffer.Unpooled;
import io.netty.handler.codec.DecoderException;

public class EndianTest {

	public static final Logger log = LoggerFactory.getLogger(EndianTest.class);
	
	public static void main(String[] args) {
		byte[] src = "Hello World!".getBytes();
		byte[] dst = new byte[14];
		// [1]. BIG_ENDIAN（大端）模式，报文头字节序列
		dst[0] = 0X00;
		dst[1] = 0X0C;
		// [2]. LITTLE_ENDIAN（小端）模式，报文头字节序列
//		dst[0] = 0X0C;
//		dst[1] = 0X00;
		
		for (int t=0; t<src.length; t++) {
			dst[t+2] = src[t];
		}
		
		ByteBuf byteBuf = Unpooled.copiedBuffer(dst);
		// [3]. 大端测试
		long n = getUnadjustedFrameLength(byteBuf, 0, 2, ByteOrder.BIG_ENDIAN);
		// [4]. 小端测试
		//long n = getUnadjustedFrameLength(byteBuf, 0, 2, ByteOrder.LITTLE_ENDIAN);
		
		log.info("数据包报头长度：[{}]", n);
	}
	
	// 取自 LengthFieldBasedFrameDecoder类
	public static long getUnadjustedFrameLength(ByteBuf buf, int offset, int length, ByteOrder order) {
        buf = buf.order(order);
        long frameLength;
        switch (length) {
        case 1:
            frameLength = buf.getUnsignedByte(offset);
            break;
        case 2:
            frameLength = buf.getUnsignedShort(offset);
            break;
        case 3:
            frameLength = buf.getUnsignedMedium(offset);
            break;
        case 4:
            frameLength = buf.getUnsignedInt(offset);
            break;
        case 8:
            frameLength = buf.getLong(offset);
            break;
        default:
            throw new DecoderException(
                    "unsupported lengthFieldLength: " + 2 + " (expected: 1, 2, 3, 4, or 8)");
        }
        return frameLength;
    }
}

• 测试结果

[1]-[3]组合测试结果正常，日志如下

09:22:26.448 [main] INFO EndianTest - 数据包报头长度：[12]

[1]-[4]组合测试结果有误，日志如下

09:20:05.287 [main] INFO EndianTest - 数据包报头长度：[3072]

[2]-[4]组合测试结果正常，日志如下

09:20:52.396 [main] INFO EndianTest - 数据包报头长度：[12]

[2]-[3]组合测试结果有误，日志如下

09:21:23.824 [main] INFO EndianTest - 数据包报头长度：[3072]

3. 结论

BIG_ENDIAN（大端）模式，字节序：

0	1	2	3	4	5	6	7	8	9	10	11	12	13
0X00	0X0C	H	e	l	l	o		w	o	r	l	d	!

LITTLE_ENDIAN（小端）模式，字节序：

0	1	2	3	4	5	6	7	8	9	10	11	12	13
0X0C	0X00	H	e	l	l	o		w	o	r	l	d	!

综上所述，两者区别就是报文头的字节顺序，通俗的说，大端模式可以理解为前置补零，小端模式理解为后置补零。