Making C++ operators

[Arkimedes]

I am thinking about making some operators in C++, but I have no clue how to approach it. Something for "to the power of". Alternatively, I could simply make a void method; but I would prefer an operator.

[sec_ware]

Hi

Operator overloading[1] is what you want I think.
The idea is to redefine an operator in the following
formal way (here: "to the power of")

double operator^(double a,  double b){
	return(exp(b*log(a)));
}

However, the above example won't work if I recall correctly,
since operator overloading requires a class/structure for some
reason. I do not know whether this is true in general, or whether
it can be bypassed.

One working, but inconsequent example would be

struct mydouble{
	double value;
};

double operator^(mydouble a,  double b){
	return(exp(b*log(a.value)));
}

Cheers

P.s. Certainly, the above construction for the power of
is superior to a naive for-loop for two reasons:
(i) it is more general (double in the exponent is possible),
thus the square can be taken
(ii) for an integer "b" larger than 20 or so, it is
faster. The reason: while the for-loop approach grows linearly
in "b" the other does not.


[1] http://msdn.microsoft.com/library/de...tm/overl_9.asp

[csl]

sec_ware has hit the nail on the head, to overload an operator, atleast one argument must be a user defined type as below

#include <iostream>
#include <cmath> // include for pow()
using namespace std;
struct PowerFloat { // define PowerFloat type
       float f;
       float operator^(float b) { // change the meaning of the ^ operator
             return pow(f,b); // compute the actual pow() and return the value
       } 
};
int main() {
       float a = 2.0;
       PowerFloat b; 
       b.f = 5;
       cout << (b^a); // output the value of pow using ^ within () due to operator presedence
       return 0;
}

[whizkid2300]

Just to jump in and add my two cents... They are both correct, when overloading an operator you have to do it within a class... If I am not mistaken it only works within that class also.

I wanna say there are some operators that are forbidden also.

But, I do what all great men do when they don't know.... Turn to a great women and tell her to get my coffee... (Please women don't take offense in that... it was a joke.)

I use google....

http://www.cplusplus.com/doc/tutorial/classes2.html

should give you a nice little read on it

[Juridian]

Erm, you don't necessarily have to overload an operator in a class. You can make non member operators to do things like the insertion and extraction operators to handle your user defined classes.

As an example:

std::istream& operator>>(std::istream& is, <user defined class>& v )
{
// code input data for your user defined class
return is;
}

This operator can then be used in the standard way to use "cin >> <your class>;" to get data for your class.

You might also look into using templates to make the non member operator more generic.