How to fetch double (FLOAT64) from FSUIPC using C#

[jspringe]

I'm new to C#. Using the methods provided in the CSHARP SDK I have been able to fetch and use most datatypes, but I'm having trouble with the 8 byte FLOAT64 values.

The overloaded methods provided are:

public bool FSUIPC_Get(ref int Token, ref byte Result)

public bool FSUIPC_Get(ref int Token, ref short Result)

public bool FSUIPC_Get(ref int Token, ref int Result)

public bool FSUIPC_Get(ref int Token, ref long Result)

public bool FSUIPC_Get(ref int Token, int dwSize, ref byte[] Result)

I assume to fetch an 8 byte value that I will use a new byte[8] array. So the question: how do I convert the 8 separate bytes returned from the get into a C# double value.

Thanks for any help,

Jason

[Pete Dowson]

I'm new to C#.

That's more than me! :wink:

I assume to fetch an 8 byte value that I will use a new byte[8] array. So the question: how do I convert the 8 separate bytes returned from the get into a C# double value.

If C# supports the double type, why can't you use that directly, like the other numerical types? Can't you add a further 'overload' (whatever that is, forgive my ignorance)?

Regards,

Pete

[jspringe]

Pete,

Just got up some courage to look inside Scott McCrory's C# library. Looks simple enough to add a FSUIPC_Get double method. However, he is using a builtin class called "Marshal" that can read the datatypes he provided methods for. The Marshal class does not have a method to read type double. This is what I have so far and where I'm stuck:

public bool FSUIPC_Get(ref int Token, ref double Result)

{

int Size = 8; // 8 bytes in a double

if ((Token < 0) || (Token > IPC_BUFFER_SIZE - (4 + Size)) )

{ //Token out of range

Result = 0;

return false;

}

IntPtr heapbuf = Marshal.AllocHGlobal(Size);

Marshal.Copy(IPC, Token + 4, heapbuf, Size);

Result = Marshal.Read??????(heapbuf);

Marshal.FreeHGlobal(heapbuf);

if (IPCdr[Token] )

{

IPCdr[Token] = false; // reset data ready flag

return true;

}

else

{ // if (data ready flag not set)

return false;

}

}

Marshal supports the following reads:

Marshal.ReadByte

Marshal.ReadInt16

Marshal.ReadInt32

Marshal.ReadInt64

Marshal.ReadIntPtr

So, I either need an example of how to call one of the existing methods that Scott provided (my original question on how to convert a byte[] to double) or I need an answer on how to add a FSUIPC_Get that takes a "double".

Thanks,

Jason

[Pete Dowson]

This is what I have so far and where I'm stuck:

Ah, I'm afraid I would be stuck even before you!

But: "Marshal.ReadInt64" will presumably read an 8-byte (64-bit) integer. Can you use casts in C#? If so just copy the int64 you read into a double using an explicit cast.

Regards,

Pete

[jspringe]

Well, I finally came up with an answer. There is a conversion method from byte[] to type double (took forever to find it). It is the "BitConverter" class.

byte[] buffer = new byte[255];

// get loaded weight

bResult = m_fsuipc.FSUIPC_Read(0x30C0, 8, ref token, ref dwResult);

bResult = m_fsuipc.FSUIPC_Process(ref dwResult);

bResult = m_fsuipc.FSUIPC_Get(ref token, 8, ref buffer);

double dLoadedWeight = BitConverter.ToDouble(buffer, 0);

Jason

[Pete Dowson]

Well, I finally came up with an answer. There is a conversion method from byte[] to type double (took forever to find it). It is the "BitConverter" class.

Well done!

All this complication and obscure inefficient-looking code does make me wonder -- why choose C#? It seems to have the worst of VB and C++ with none of the flexibility and power of the latter or ease of the former.

Regards,

Pete

[jspringe]

Why use C#?

Well good or bad, it is the closest thing to Java and true OO (I still prefer Java). Really, it is extremely simple. The only time it can be difficult is when you need to interface to legacy modules (and that's just because it is my first time developing an app in C#). The syntax and structure is much like Java which makes it easier to switch between the two on different projects (I have completely forgotten my c programming skills from 20 years ago -- char* far * what?). A long time ago I vowed never to program in VB. Just not ready to eat crow on that one yet.

Jason

[Pete Dowson]

The only time it can be difficult is when you need to interface to legacy modules

"Legacy"? Is C and C++ relegated to a "legacy" category now? Most all of FS2004 is written in C++ (some parts are C and ASM leftovers, but very little now). And pretty much all of Windows is too. How can all these modern modules be classed as "legacy"?

Regards,

Pete