n00b C++ Question

[Master_Nabo]

I am using VSC++.net2k3, and I was wondering how to use the DWORD type in C++.

I am porting some code over from C# in order to make a Dynamic Link Library in C++, and I discovered that my complier does not reconsize the DWORD as an identifier, and it didn't reconize ZeroMemory or CopyMemory for some random reason.

Code Select Expand


#pragma once

using namespace System;

namespace CBnetHashing
{
...
}

static void CBnetHashing::XSha1::HashData(System::Byte *inbuff, System::Int32 nLength, System::Byte *outbuff)
{
DWORD dhashbuff[128];
ZeroMemory(dhashbuff, 128*4);
if (nLength > 64)
throw new System::ArgumentException("Data cannot be more than 64 bytes");
CopyMemory(dhashbuff, inbuff, nLength);
int i;
DWORD a, b, c, d, e, f;...
}


(the dots represents code that I have yet to finish.)

Basically, what I am trying to do is declair an array of the DWORD type (was used in C# code) into C++, ZeroMemory all of the addresses, and CopyMemory from inbuff to dhasbuff to a certain length.

[UserLoser.]

Code Select Expand


typedef unsigned long DWORD;


or try including header files such as windows.h.

[Master_Nabo]

Thanks, I can't believe I forgot the header file.

In the C# code, it refers to static functions, yet when I compile them, I get errors.

Code Select Expand


namespace CBnetHashing
{
public __gc class XSha1
{
public:
static void HashData(System::Byte *inbuff, System::Int32 nLength, System::Byte *outbuff);
private:
static System::UInt32 Rol(System::UInt32 n, int shift);
};


How can I fix this (I don't remember hearing about static function)?

[UserLoser.]

What errors?

[Master_Nabo]

error C2724: 'CBnetHashing::XSha1::Rol' : 'static' should not be used on member functions defined at file scope

[EpicOfTimeWasted]

DWORD should actually be:

Code Select Expand

typedef unsigned int DWORD;

This makes sure that your double word doesn't turn into a quad word when it's compiled for a 64 bit CPU.

[Kp]

DWORD should actually be:

Code Select Expand

typedef unsigned int DWORD;

This makes sure that your double word doesn't turn into a quad word when it's compiled for a 64 bit CPU.

Technically, on a 64bit CPU, a doubleword should be 128 bits.  Of course, it's rare you ever see a Windows programmer using the right definition of a machine word anyway.

[Master_Nabo]

Thanks for the info but...

Where do I put that?  When I include it before the name, my compiler says that:

error C2146: syntax error : missing ';' before identifier 'dhashbuff'

So could you post an example of making an array of DWORD?

From what I can make sence, it tries to make DWORD an interger, and wants to throw the rest of the declaration away. 

[Kp]

Code Select Expand

DWORD dw[5] /* big enough for a SHA output */;
DWORD *pdw = new DWORD[10] /* this line must be in a function, not at file scope */;

[MyndFyre]

A couple thoughts....

You definitely didn't see a "DWORD" type in a C# library.  The only "DWORD" type in a C# library would be System.UInt32 (System::UInt32), which is accessible through MC++, which brings me around to my next question, which is -- if you're using OTHER MC++ types like System::Byte, why aren't you using System::UInt32 instead of DWORD?

Also, the C# "static" keyword means that you don't need to have an instance of an object to access that particular member.  As I recall, in C++, you don't have to have an instance of an object, but you'll get an exception for trying to access a NULL pointer if the function you call uses the this pointer at all.  I could be wrong on that, though.

[Master_Nabo]

Nah, my project partner thought that DWORD was the C++ equivelent of the C# uint, which it isn't (maybe I should have thought more into this before posting, oh well).

Sure, I'll try and use System::UInt32 instead (this is what I originally planed to use, but Shout tried to convince me otherwise).

[R.a.B.B.i.T]

or try including header files such as windows.h.

[shout]

He is porting my BnetHashing.dll to C++, which is available in C# here.

[LordVader]

This page, talks about general issues when trying to compile C# like code under C++ and general idea's on conversions & fixes.

Most of the time it will just require (TYPE) cast infront of an instruction|function..
Where -> TYPE <- would be DWORD|INT etc.. or other 'type' of unit or structure known to the compiler.

Example from the page I mentioned.

In C this will work:

Code Select Expand


HBITMAP hbmOldBuffer = SelectObject(hdcBuffer, hbmBuffer);

But in C++ requires a cast:

Code Select Expand


HBITMAP hbmOldBuffer = (HBITMAP)SelectObject(hdcBuffer, hbmBuffer);