[ACCEPTED]-how to convert a value type to byte[] in C#?-arrays

Accepted answer
Score: 21

To convert any value types (not just primitive 1 types) to byte arrays and vice versa:

    public T FromByteArray<T>(byte[] rawValue)
    {
        GCHandle handle = GCHandle.Alloc(rawValue, GCHandleType.Pinned);
        T structure = (T)Marshal.PtrToStructure(handle.AddrOfPinnedObject(), typeof(T));
        handle.Free();
        return structure;
    }

    public byte[] ToByteArray(object value, int maxLength)
    {
        int rawsize = Marshal.SizeOf(value);
        byte[] rawdata = new byte[rawsize];
        GCHandle handle =
            GCHandle.Alloc(rawdata,
            GCHandleType.Pinned);
        Marshal.StructureToPtr(value,
            handle.AddrOfPinnedObject(),
            false);
        handle.Free();
        if (maxLength < rawdata.Length) {
            byte[] temp = new byte[maxLength];
            Array.Copy(rawdata, temp, maxLength);
            return temp;
        } else {
            return rawdata;
        }
    }
Score: 15

The BitConverter class might be what you are looking 3 for. Example:

int input = 123;
byte[] output = BitConverter.GetBytes(input);

If your enum was known to be 2 an Int32 derived type, you could simply 1 cast its values first:

BitConverter.GetBytes((int)commands.one);
Score: 3

You can use the BitConverter.GetBytes method to do this.

0

Score: 3

For anyone who's interested in how it works 3 without using BitConverter you can do it like this:

// Convert double to byte[]
public unsafe byte[] pack(double d) {
    byte[] packed = new byte[8]; // There are 8 bytes in a double
    void* ptr = &d; // Get a reference to the memory containing the double
    for (int i = 0; i < 8; i++) { // Each one of the 8 bytes needs to be added to the byte array
        packed[i] = (byte)(*(UInt64 *)ptr >> (8 * i)); // Bit shift so that each chunk of 8 bits (1 byte) is cast as a byte and added to array 
    }
    return packed;
}

// Convert byte[] to double
public unsafe double unpackDouble(byte[] data) {
    double unpacked = 0.0; // Prepare a chunk of memory ready for the double
    void* ptr = &unpacked; // Reference the double memory
    for (int i = 0; i < data.Length; i++) {
        *(UInt64 *)ptr |= ((UInt64)data[i] << (8 * i)); // Get the bits into the right place and OR into the double
    }
    return unpacked;
}

In 2 reality it's much easier and safer to use 1 BitConverter but it's fun to know!

More Related questions