BitConverter.GetBytes 方法以什么顺序返回字节数组

前言

我在“浅谈 ZipInteger”一文中的 ZipInteger 结构中使用了 BitConverter 类的 GetBytes 方法。当时我是假设 GetBytes 方法根据 IsLittleEndian 的值不同而按照不同的顺序返回字节数组。但是 MSDN 有关 BitConverter 类的文档中没有对此作出明确的说明。请参见我在 MSDN 论坛的一个贴子“请问 BitConverter.GetBytes 方法以什么顺序返回字节数组”:

在 MSDN 文档的“BigInteger 构造函数 (Byte[])”中提到:

value 数组中的各个字节应该为 little-endian 顺序,从最低序位字节到最高序位字节。 
将数值转换为字节数组的大多数方法,例如 BigInteger.ToByteArray 和 BitConverter.GetBytes,以 little-endian 顺序返回字节数组。

按照我的理解,该 MSDN 文档认为 BitConverter.GetBytes 方法总是以 little-endian 顺序返回字节数组,而不管 BitConverter.IsLittleEndian 的值如何。
而在 BitConverter 类的文档中并没有明确指出 BitConverter.GetBytes 方法应该以什么顺序返回字节数组。
不知我的理解是否正确。

该贴子并没有得到靠谱的答复。

测试程序

那么,我们写个程序来测试一下吧。下面就是 BitConverterTester.cs:

 

using System;

namespace Skyiv.Tester
{
  static class BitConverterTester
  {
    static void Main()
    {
      Console.WriteLine("      OS Version: " + Environment.OSVersion);
      Console.WriteLine("     CLR Version: " + Environment.Version);
      Console.WriteLine("  IsLittleEndian: " + BitConverter.IsLittleEndian);
      long n = 0x1234567890ABCDEF;
      double d = 1;
      Console.WriteLine(n.ToString("X") + ": " + BitConverter.ToString(BitConverter.GetBytes(n)));
      Console.WriteLine(d.ToString("F14") + ": " + BitConverter.ToString(BitConverter.GetBytes(d)));
    }
  }
}

 

这个程序在 Windows Server 2003 操作系统的 .NET Framework 4 环境下编译和运行:

C:\CS\BitConverterTester> csc BitConverterTester.cs
Microsoft(R) Visual C# 2010 编译器 4.0.30319.1 版
版权所有(C) Microsoft Corporation。保留所有权利。

C:\CS\BitConverterTester> BitConverterTester
      OS Version: Microsoft Windows NT 5.2.3790 Service Pack 2
     CLR Version: 4.0.30319.1
  IsLittleEndian: True
1234567890ABCDEF: EF-CD-AB-90-78-56-34-12
1.00000000000000: 00-00-00-00-00-00-F0-3F

C:\CS\BitConverterTester>

 

在 Ubuntu 10.10 操作系统的 Mono 2.8.2 环境下编译和运行:

ben@ben-m4000t:~/work/BitConverterTester$ dmcs BitConverterTester.cs
ben@ben-m4000t:~/work/BitConverterTester$ mono28 BitConverterTester.exe
      OS Version: Unix 2.6.35.24
     CLR Version: 4.0.30319.1
  IsLittleEndian: True
1234567890ABCDEF: EF-CD-AB-90-78-56-34-12
1.00000000000000: 00-00-00-00-00-00-F0-3F
ben@ben-m4000t:~/work/BitConverterTester$ 

 

这两次运行的结果都在预料之中,BitConverter 类的 GetBytes 方法以 Little-Endian 顺序返回字节数组。但是,在这两次运行中,IsLittleEndian 的值都为 True,所以还是没有解决我们的问题。

查看 Microsoft .NET Framework 4 中相关的源程序代码

祭出 Reflector 这个神器:

BitConverter.GetBytes 方法以什么顺序返回字节数组_第1张图片

在 mscorlib.dll 的 System 命名空间下找到 BitConverter 类:

BitConverter.GetBytes 方法以什么顺序返回字节数组_第2张图片

如上图如示,IsLittleEndian 是 BitConverter 类的静态只读字段。

BitConverter.GetBytes 方法以什么顺序返回字节数组_第3张图片

如上图所示,在 BitConverter 类的静态构造函数中,直接把 IsLittleEndian 这个静态只读字段的值赋值为 true。由于我没有 Microsoft 实现 BitConverter 类的 C# 源程序代码,不知道是 Microsoft 的 C# 源程序中就是直接这样写呢,还是实际上是有根据平台来判断的,但是 C# 编译器在具体平台上优化了这段代码。

BitConverter.GetBytes 方法以什么顺序返回字节数组_第4张图片

如上图所示,GetBytes(Int64) 方法也非常简单,直接通过不安全的指针转换就得到了相应的字节数组。学过 C 语言中的朋友想必非常熟悉这种做法。这下清楚了,BitConverter 类的 GetBytes 方法是根据 IsLittleEndian 的值的不同按照不同顺序来返回字节数组的。MSDN 文档在“BigInteger 构造函数 (Byte[]) ”中的相关说法:“例如 BigInteger.ToByteArray 和 BitConverter.GetBytes,以 little-endian 顺序返回字节数组”是不正确的。

继续看下去:

BitConverter.GetBytes 方法以什么顺序返回字节数组_第5张图片

上图中的 ToInt64 方法也明确地根据 IsLittleEndian 的值采取不同的动作。其实,既然在静态构造函数中明确地给 IsLittleEndian 赋值为 true,这里也可以省略对 IsLittleEndian 的判断,直接按照 IsLittleEndian 的值为 true 去做就行了,可以稍微节省点代码,提高点速度。

BitConverter.GetBytes 方法以什么顺序返回字节数组_第6张图片

我们看到,GetBytes(Double) 方法也非常简单。

查看 Mono 2.8.2 中相关的源代码

我们可以到 http://ftp.novell.com/pub/mono/sources/mono/ 下载 Mono 的源代码,然后按以下方法找到 BitConverter 类的源程序代码:

ben@ben-m4000t:~$ cd src/mono-2.8.2
ben@ben-m4000t:~/src/mono-2.8.2$ find . -name BitConverter.cs
./mcs/class/corlib/System/BitConverter.cs
ben@ben-m4000t:~/src/mono-2.8.2$

下面就是 BitConverter.cs:

 

//
// System.BitConverter.cs
//
// Author:
//   Matt Kimball ([email protected])
//
//
// Copyright (C) 2004 Novell, Inc (http://www.novell.com)
//
// Permission is hereby granted, free of charge, to any person obtaining
// a copy of this software and associated documentation files (the
// "Software"), to deal in the Software without restriction, including
// without limitation the rights to use, copy, modify, merge, publish,
// distribute, sublicense, and/or sell copies of the Software, and to
// permit persons to whom the Software is furnished to do so, subject to
// the following conditions:
// 
// The above copyright notice and this permission notice shall be
// included in all copies or substantial portions of the Software.
// 
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
// EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
// MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
// NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
// LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
// OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
// WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
//

using System.Text;

namespace System
{
	public
	static
	class BitConverter
	{
		static readonly bool SwappedWordsInDouble = DoubleWordsAreSwapped ();
		public static readonly bool IsLittleEndian = AmILittleEndian ();

		static unsafe bool AmILittleEndian ()
		{
			// binary representations of 1.0:
			// big endian: 3f f0 00 00 00 00 00 00
			// little endian: 00 00 00 00 00 00 f0 3f
			// arm fpa little endian: 00 00 f0 3f 00 00 00 00
			double d = 1.0;
			byte *b = (byte*)&d;
			return (b [0] == 0);
		}

		static unsafe bool DoubleWordsAreSwapped ()
		{
			// binary representations of 1.0:
			// big endian: 3f f0 00 00 00 00 00 00
			// little endian: 00 00 00 00 00 00 f0 3f
			// arm fpa little endian: 00 00 f0 3f 00 00 00 00
			double d = 1.0;
			byte *b = (byte*)&d;
			return b [2] == 0xf0;
		}

		public static long DoubleToInt64Bits (double value)
		{
			return ToInt64 (GetBytes (value), 0);
		}

		public static double Int64BitsToDouble (long value)
		{
			return ToDouble (GetBytes (value), 0);
		}

		internal static double InternalInt64BitsToDouble (long value)
		{
			return SwappableToDouble (GetBytes (value), 0);
		}
		
		unsafe static byte[] GetBytes (byte *ptr, int count)
		{
			byte [] ret = new byte [count];

			for (int i = 0; i < count; i++) {
				ret [i] = ptr [i];
			}

			return ret;
		}

		unsafe public static byte[] GetBytes (bool value)
		{
			return GetBytes ((byte *) &value, 1);
		}

		unsafe public static byte[] GetBytes (char value)
		{
			return GetBytes ((byte *) &value, 2);
		}

		unsafe public static byte[] GetBytes (short value)
		{
			return GetBytes ((byte *) &value, 2);
		}

		unsafe public static byte[] GetBytes (int value)
		{
			return GetBytes ((byte *) &value, 4);
		}

		unsafe public static byte[] GetBytes (long value)
		{
			return GetBytes ((byte *) &value, 8);
		}

		[CLSCompliant (false)]
		unsafe public static byte[] GetBytes (ushort value)
		{
			return GetBytes ((byte *) &value, 2);
		}

		[CLSCompliant (false)]
		unsafe public static byte[] GetBytes (uint value)
		{
			return GetBytes ((byte *) &value, 4);
		}

		[CLSCompliant (false)]
		unsafe public static byte[] GetBytes (ulong value)
		{
			return GetBytes ((byte *) &value, 8);
		}

		unsafe public static byte[] GetBytes (float value)
		{
			return GetBytes ((byte *) &value, 4);
		}

		unsafe public static byte[] GetBytes (double value)
		{
			if (SwappedWordsInDouble) {
				byte[] data = new byte [8];
				byte *p = (byte*)&value;
				data [0] = p [4];
				data [1] = p [5];
				data [2] = p [6];
				data [3] = p [7];
				data [4] = p [0];
				data [5] = p [1];
				data [6] = p [2];
				data [7] = p [3];
				return data;
			} else {
				return GetBytes ((byte *) &value, 8);
			}
		}

		unsafe static void PutBytes (byte *dst, byte[] src, int start_index, int count)
		{
			if (src == null)
				throw new ArgumentNullException ("value");

			if (start_index < 0 || (start_index > src.Length - 1))
				throw new ArgumentOutOfRangeException ("startIndex", "Index was"
					+ " out of range. Must be non-negative and less than the"
					+ " size of the collection.");

			// avoid integer overflow (with large pos/neg start_index values)
			if (src.Length - count < start_index)
				throw new ArgumentException ("Destination array is not long"
					+ " enough to copy all the items in the collection."
					+ " Check array index and length.");

			for (int i = 0; i < count; i++)
				dst[i] = src[i + start_index];
		}

		unsafe public static bool ToBoolean (byte[] value, int startIndex)
		{
			if (value == null) 
				throw new ArgumentNullException ("value");

			if (startIndex < 0 || (startIndex > value.Length - 1))
				throw new ArgumentOutOfRangeException ("startIndex", "Index was"
					+ " out of range. Must be non-negative and less than the"
					+ " size of the collection.");

			if (value [startIndex] != 0)
				return true;
			
			return false;
		}

		unsafe public static char ToChar (byte[] value, int startIndex)
		{
			char ret;

			PutBytes ((byte *) &ret, value, startIndex, 2);

			return ret;
		}

		unsafe public static short ToInt16 (byte[] value, int startIndex)
		{
			short ret;

			PutBytes ((byte *) &ret, value, startIndex, 2);

			return ret;
		}

		unsafe public static int ToInt32 (byte[] value, int startIndex)
		{
			int ret;

			PutBytes ((byte *) &ret, value, startIndex, 4);

			return ret;
		}

		unsafe public static long ToInt64 (byte[] value, int startIndex)
		{
			long ret;

			PutBytes ((byte *) &ret, value, startIndex, 8);

			return ret;
		}

		[CLSCompliant (false)]
		unsafe public static ushort ToUInt16 (byte[] value, int startIndex)
		{
			ushort ret;

			PutBytes ((byte *) &ret, value, startIndex, 2);

			return ret;
		}

		[CLSCompliant (false)]
		unsafe public static uint ToUInt32 (byte[] value, int startIndex)
		{
			uint ret;

			PutBytes ((byte *) &ret, value, startIndex, 4);

			return ret;
		}

		[CLSCompliant (false)]
		unsafe public static ulong ToUInt64 (byte[] value, int startIndex)
		{
			ulong ret;

			PutBytes ((byte *) &ret, value, startIndex, 8);

			return ret;
		}

		unsafe public static float ToSingle (byte[] value, int startIndex)
		{
			float ret;

			PutBytes ((byte *) &ret, value, startIndex, 4);

			return ret;
		}

		unsafe public static double ToDouble (byte[] value, int startIndex)
		{
			double ret;

			if (SwappedWordsInDouble) {
				byte* p = (byte*)&ret;
				if (value == null)
					throw new ArgumentNullException ("value");

				if (startIndex < 0 || (startIndex > value.Length - 1))
					throw new ArgumentOutOfRangeException ("startIndex", "Index was"
						+ " out of range. Must be non-negative and less than the"
						+ " size of the collection.");

				// avoid integer overflow (with large pos/neg start_index values)
				if (value.Length - 8 < startIndex)
					throw new ArgumentException ("Destination array is not long"
						+ " enough to copy all the items in the collection."
						+ " Check array index and length.");

				p [0] = value [startIndex + 4];
				p [1] = value [startIndex + 5];
				p [2] = value [startIndex + 6];
				p [3] = value [startIndex + 7];
				p [4] = value [startIndex + 0];
				p [5] = value [startIndex + 1];
				p [6] = value [startIndex + 2];
				p [7] = value [startIndex + 3];

				return ret;
			}

			PutBytes ((byte *) &ret, value, startIndex, 8);

			return ret;
		}

		unsafe internal static double SwappableToDouble (byte[] value, int startIndex)
		{
			double ret;

			if (SwappedWordsInDouble) {
				byte* p = (byte*)&ret;
				if (value == null)
					throw new ArgumentNullException ("value");

				if (startIndex < 0 || (startIndex > value.Length - 1))
					throw new ArgumentOutOfRangeException ("startIndex", "Index was"
						+ " out of range. Must be non-negative and less than the"
						+ " size of the collection.");

				// avoid integer overflow (with large pos/neg start_index values)
				if (value.Length - 8 < startIndex)
					throw new ArgumentException ("Destination array is not long"
						+ " enough to copy all the items in the collection."
						+ " Check array index and length.");

				p [0] = value [startIndex + 4];
				p [1] = value [startIndex + 5];
				p [2] = value [startIndex + 6];
				p [3] = value [startIndex + 7];
				p [4] = value [startIndex + 0];
				p [5] = value [startIndex + 1];
				p [6] = value [startIndex + 2];
				p [7] = value [startIndex + 3];

				return ret;
			} else if (!IsLittleEndian) {
				byte* p = (byte*)&ret;
				if (value == null)
					throw new ArgumentNullException ("value");

				if (startIndex < 0 || (startIndex > value.Length - 1))
					throw new ArgumentOutOfRangeException ("startIndex", "Index was"
						+ " out of range. Must be non-negative and less than the"
						+ " size of the collection.");

				// avoid integer overflow (with large pos/neg start_index values)
				if (value.Length - 8 < startIndex)
					throw new ArgumentException ("Destination array is not long"
						+ " enough to copy all the items in the collection."
						+ " Check array index and length.");

				p [0] = value [startIndex + 7];
				p [1] = value [startIndex + 6];
				p [2] = value [startIndex + 5];
				p [3] = value [startIndex + 4];
				p [4] = value [startIndex + 3];
				p [5] = value [startIndex + 2];
				p [6] = value [startIndex + 1];
				p [7] = value [startIndex + 0];

				return ret;
			}

			PutBytes ((byte *) &ret, value, startIndex, 8);

			return ret;
		}
		
		public static string ToString (byte[] value)
		{
			if (value == null)
				throw new ArgumentNullException ("value");

			return ToString (value, 0, value.Length);
		}

		public static string ToString (byte[] value, int startIndex)
		{
			if (value == null)
				throw new ArgumentNullException ("value");

			return ToString (value, startIndex, value.Length - startIndex);
		}

		public static string ToString (byte[] value, int startIndex, int length)
		{
			if (value == null)
				throw new ArgumentNullException ("byteArray");

			// The 4th and last clause (start_index >= value.Length)
			// was added as a small fix to a very obscure bug.
			// It makes a small difference when start_index is
			// outside the range and length==0. 
			if (startIndex < 0 || startIndex >= value.Length) {
				// special (but valid) case (e.g. new byte [0])
				if ((startIndex == 0) && (value.Length == 0))
					return String.Empty;
				throw new ArgumentOutOfRangeException ("startIndex", "Index was"
					+ " out of range. Must be non-negative and less than the"
					+ " size of the collection.");
			}

			if (length < 0)
				throw new ArgumentOutOfRangeException ("length",
					"Value must be positive.");

			// note: re-ordered to avoid possible integer overflow
			if (startIndex > value.Length - length)
				throw new ArgumentException ("startIndex + length > value.Length");

			if (length == 0)
				return string.Empty;

			StringBuilder builder = new StringBuilder(length * 3 - 1);
			int end = startIndex + length;

			for (int i = startIndex; i < end; i++) {
				if (i > startIndex)
					builder.Append('-');
				
				char high = (char)((value[i] >> 4) & 0x0f);
				char low = (char)(value[i] & 0x0f);

				if (high < 10) 
					high += '0';
				else {
					high -= (char) 10;
					high += 'A';
				}

				if (low < 10)
					low += '0';
				else {
					low -= (char) 10;
					low += 'A';
				}
				builder.Append(high);
				builder.Append(low);
			}

			return builder.ToString ();
		}
	}
}

 

 

可以看出,在第 39 行通过第 41 到 50 行的 AmILittleEndian 方法给 IsLittleEndian 这个静态只读字段赋值。这个 AmILittleEndian 方法中的注释写得很明白,其实有三种不台的平台,除了 Big-Endian 和 Little-Endian 外,还有一种“arm fpa little endian”,在这里也归于 Little-Endian。但是在后面的方法中是要以不同的手段进行处理的。

另外,不象 Microsoft .NET Framework 4,Mono 中 BitConverter 类的各种重载的公有静态GetBytes 方法是统一调用第 78 到 87 行的私有静态 GetBytes 方法来实现其功能的。

从 Mono 中相关的源程序代码中也可以看出,BitConverter 类的 GetBytes 方法是根据 IsLittleEndian 的值的不同以不同的顺序返回字节数组的。

结论

总结一下,MSDN 文档在“BigInteger 构造函数 (Byte[]) ”中的相关说法:“例如 BigInteger.ToByteArray 和 BitConverter.GetBytes,以 little-endian 顺序返回字节数组”是不正确的。BitConverter 类的 GetBytes 方法是根据 IsLittleEndian 的值的不同以不同的顺序返回字节数组的。

你可能感兴趣的:(Converter)