[Dev-C++] typedef of pointers to functions compile error

[Ben M. \(finlaylabs\) <bm...@fi...>]

I have a prob with code that contains pointers to functions, that will =
not compile with Dev-C++. Other (knowlegeable)friends can make small =
mods and get it to compile on windows with VC++6.0, and on BSD systems =
etc, But Dev-C++ seems to choke on this one issue. I have used Dev-C++ =
for 6 months now and have been very impressed.
I am taking the liberty of pasting a post to Devshed below, that =
explains the issue. One friend (mitakeet) states that "Likely DevC wants =
some other way to declare the pointer. My research yesterday indicated =
that many compilers do not follow a standard, I just tried things until =
they worked."

He declared the pointer using this style:
#ifdef WIN32
    #define FNCTPTR=20
#else //*nix
    #define FNCTPTR &Data::
#endif

class Data{
private:

typedef int (Data::*FUNC_GET)(void);
and then
int GetData(FUNC_GET fn){
        return (this->*fn)();
};
but that does not work on Dev-C++. Neither does my method, basically =
straight from the books by Stroustrup. The orig question post is below, =
with nice red highlighting to ease your viewing. How should I be =
declaring the pointer to function for Dev-C++?? Thanks for the help and =
the EXCELLENT software,

Ben Mclean


-------------------------------------------------------------------------=
-------
A friend and I are stumped on why the following wont compile. We get =
many error messages all as follows, for any function that contains calls =
to ddxFunc():=20

Calculations.cpp: In member function `double =
calculations::ddt3_Stress_xx(int, int, int)':=20
Calculations.cpp:176: no matching function for call to =
`Calculations::ddxFunc(<unknown type>, int&, int&, int&)'=20
Calculations.h:61: candidates are: double Calculations::ddxFunc(double =
(Calculations::*)(int, int, int), int, int, int)=20

The function ddt3_Stress_xx is below, the only point being it makes use =
of ddxFunc:=20

code:--------------------------------------------------------------------=
----------double Calculations::ddt3_Stress_xx(int i,int j,int k){
    double ddt3stress;
    ddt3stress =3D calcStiffness(11,i,j,k)*ddxFunc(ddt2_Vel_x,i,j,k)
                + calcStiffness(12,i,j,k)*ddyFunc(ddt2_Vel_y,i,j,k)
                + calcStiffness(13,i,j,k)*ddzFunc(ddt2_Vel_z,i,j,k);
    if(myParams->symmetry>3){
        ddt3stress +=3D =
calcStiffness(14,i,j,k)*(ddzFunc(ddt2_Vel_y,i,j,k)
                    + ddyFunc(ddt2_Vel_z,i,j,k))
                    + calcStiffness(15,i,j,k)*(ddzFunc(ddt2_Vel_x,i,j,k)
                    + ddxFunc(ddt2_Vel_z,i,j,k))
                    + calcStiffness(16,i,j,k)*(ddyFunc(ddt2_Vel_x,i,j,k)
                    + ddxFunc(ddt2_Vel_y,i,j,k));
    }
    return ddt3stress;
}------------------------------------------------------------------------=
------
As far as we can determine we have correctly typedef'd etc etc such that =
it should be clear the FUNC argument is of type double. An extract of =
Calculations.h and .cpp showing how we have done this is below. What are =
we doing wrong, or what are we not doing that we need to be? I would =
greatly apprecate some help on this.=20

Thanks in anticipation,=20
finlaylabs.=20


code:--------------------------------------------------------------------=
----------//*************************************************************=
***************
//Calculations.h

class Calculations{
    // typedef for passing functions as args
    typedef double (Calculations::*FUNC)(int,int,int);
    // should be clear that FUNC arguments return type double
public:

private:

    double ddxFunc(FUNC,int,int,int);
    double ddyFunc(FUNC,int,int,int);
    double ddzFunc(FUNC,int,int,int);
//etc............


//***********************************************************************=
*****
//Calculations.cpp

#include <math.h>

#include "Calculations.h"

// derivatives of function-based values:
//dd*Func() can calc derivatives for ddt2_vel_x() etc function-derived =
values
double Calculations::ddxFunc(FUNC fn,int i,int j,int k){
    double xderv =3D 0.0;
    for(int K_index=3D1;K_index<=3DmyParams->K;K_index++){
        xderv +=3D weightcoeff(K_index,myParams->dx)
                *tapercoeff(K_index,myParams->dx)
                =
*((this->*fn)(i+K_index,j,k)-(this->*fn)(i+1-K_index,j,k));
    }
    return xderv;
}

-------------------------------------------------------------------------=
-----