generic number in c++

Question

I want to make a BigInteger object (for practice). I want the overloaded operators to accept any numerical datatype. I can do this polymorphically, but since it would be impractical to overload the 20+ binary operators for each of the 20~ numerical types, I would really like to do something like this:

X & operator+(const anynum input)
{
  return this->value += input;
}

...
main()
{
  X a = 1;
  a = a + 1;
  a = a + 1L;
}

sorry, my question is: "is this possible"?

I researched this most of last night I read through the operator overloading entry on cpp.com, the list of overloadable operators on wikipedia, various posts on stack overflow.

Instead of writing binary operators for every possible pair, why not write one binary operator for bigint + bigint, and then have cast operators to convert from other numerical types to bigint? — triple_r, Jul 27 '15 at 19:58
facepalm. Yup. It has been way too long since I have done C++. @tripl_r post this as an answer and I will choose it. — chrisgotter, Jul 27 '15 at 20:25
chrisgotter, before you go with the cast solution, consider @LightnessRacesinOrbit 's suggestion of using a templated function. You will have one function, no casting, and the compiler will scream bloody blue murder if someone tries to stuff in a string or other non numeric. — user4581301, Jul 27 '15 at 21:03

user4581301 · Accepted Answer · 2015-07-27T21:26:34.557

Possible through slightly different means. See example:

#include <iostream>

class X
{
public:
    X(int val) :
            value(val)
    {

    }
// this is the important part
    template<class TYPE>
    X & operator+(TYPE input)
    {
        value += input; // just demonstrating with a simple int
                        // a real bignum will be significantly more complex
                        // and you may well find that one function does not fit all cases
                        // for your particular bignum implementation
        return *this;
    }
// end of important part
    void print()
    {
        std::cout << value << std::endl;
    }
private:
    int value; 
};

int main()
{
    short s = 1;
    unsigned short us = 1;
    int i = 1;
    long l = 1;
    long long ll = 1;
    float f = 1.0;
    double d = 1.0;
    X a(1);
    a.print();
    a + 1; // templated function is automatically implemented by compiler 
           // for the data type in the call. Use all the data types you want. 
           // The compiler will make one for each one used. 
    a.print();
    a + 1L;
    a.print();
    a + 1.0;
    a.print();
    a + s;
    a.print();
    a + us;
    a.print();
    a + i;
    a.print();
    a + l;
    a.print();
    a + ll;
    a.print();
    a + f;
    a.print();
    a + d;
    a.print();
//  a + "Hi!"; Ka BOOOOOM! Cant add a string to an integer
    a + 'A'; // it does, however, allow this through because math on characters
             // is a time-honoured tradition.
    a.print();
}

For completeness, here's the output:

interesting, for some reason I thought that wouldn't work, I will have to give it a try though — chrisgotter, Jul 27 '15 at 21:23

generic number in c++

1 Answers1