C++ Unions in C#, how do they work underneath?

Question

i have implemented c++ unions in c#, just to verify that i´m understandig it. But it seems that i understand nothing. I expected some times a completly different output. My Code:

using System;
using System.Runtime.InteropServices;

namespace ConsoleApplication2
{
[StructLayout(LayoutKind.Explicit)]
class union
{
    [FieldOffset(0)]
    public double d;
    [FieldOffset(0)]
    public float f0;
    [FieldOffset(4)]
    public float f1;
    [FieldOffset(0)]
    public int i0;
    [FieldOffset(4)]
    public int i1;
    [FieldOffset(0)]
    public short s0;
    [FieldOffset(2)]
    public short s1;
    [FieldOffset(4)]
    public short s2;
    [FieldOffset(6)]
    public short s3;

}

class Program
{
    static void Main(string[] args)
    {
        union su = new union();
        su.f0 = 19.012012F;
        su.f1 = 3.14159265F;
        Console.WriteLine(su.d);
        Console.WriteLine(su.i0);
        Console.WriteLine(su.i1);
        Console.WriteLine(su.s0);
        Console.WriteLine(su.s1);
        Console.WriteLine(su.s2);
        Console.WriteLine(su.s3);
        Console.ReadLine();
    }
}
}

And my output is:

50,1238786690385
1100486810
1078530011
6298
16792
4059
16457

For example, i thought s0 would be 30209 and not 6298. Can someone explain how it works?

im not sure how to resolve the int and short types. s0 is located in first two bytes of f0, so how do i have to resolve the bits, that i get 6298? — Shintos, Jun 17 '13 at 19:16
Stupid me, has known the answer, but was unable to do basic math. It is really simple. Translate your float in IEEE 754. And then check your offset, that is your starting point. Make sure you know the length of your type in bit or better in byte. Then resolve your number like you have learnt before. For Example: The last 16 bit(fraction) of f0 are: 0001 1000 (Byte 1) 1001 1010(Byte 0) s0 is a short integer so it size is 16-Bit or 2 Bytes. Now starting to resolve s0 by starting at the LSB to: 2 + 8 + 16 + 128 + 2048 + 4096 = 6298 I´m truly sorry for the inconvenience. — Shintos, Jun 17 '13 at 19:54

score 1 · Accepted Answer · answered Jun 18 '13 at 06:22

Stupid me, has known the answer, but was unable to do basic math. It is really simple. Translate your float in IEEE 754. And then check your offset(in bytes), that is your starting point. Make sure you know the length of your type in bit or better in byte. Then resolve your number like you have learnt before.
For Example:
The last 16 bit(fraction) of f0 are: 0001 1000 (Byte 1) 1001 1010(Byte 0)
s0 is a short integer so it size is 16-Bit or 2 Bytes. Now starting to resolve s0 by starting at the LSB to:
2 + 8 + 16 + 128 + 2048 + 4096 = 6298

So nothing special here.
I´m truly sorry for the inconvenience.

score 0 · Answer 2 · edited May 23 '17 at 10:32

0

I think you may need to use struct other than class. Here is something more: Discriminated union in C#

edited May 23 '17 at 10:32

Community

1
1

answered Jun 17 '13 at 19:09

NonStatic

951
1
8
27

C++ Unions in C#, how do they work underneath?

2 Answers2