Bignum divison and value assignment in c++

Question

I'm writing my own bignum class for operating on large numbers. So far I've overloaded the operator= and operator+. How do I perform long division?

Also, right now I can only assign values within the range of integers to the bignum object. How do I assign values that fall outside the int range? Is it possible to do this without strings?

#include <iostream>

using namespace std;
class bignum
{
public:
    int number[20];
    bignum operator + (bignum);
    bignum operator = (int);
    void output()
};

bignum bignum::operator= (int j)
{
    int f;
    f=j;
    for(int k=0; k<=19; k++)
    {
       number[k]=0;
    }
    for(int l=19; l>=0,f>0; l--)
    {
       number[l]=(f%10);
       f/=10;
    }
}

bignum bignum::operator+ (bignum b) 
{
    bignum a;
    int carry=0;
    for(int k=0; k<=19; k++)
    {
        a.number[k]=0;
    }
    for(int i=19; i>=0; i--)
    {
        a.number[i]= number[i]+b.number[i]+a.number[i];
        if(a.number[i]>9)
        {
        carry=(a.number[i]/10);
        a.number[i-1]+=carry;
        a.number[i]=(a.number[i]%10);
        }
    }

    return a;
}

int main()
{
    bignum a,b,c;
    a=9999;
    b=a+a;
    //for(int k=1; k<=9; k++)
    //b.number[k]=0;
    //b=a+a;
    for(int k=0; k<=19; k++)
    cout<<b.number[k];
    cin.get();
}

if you need to work with very big numbers then aproximation algorithms are the way. Also for assigning is better to use hexadecimal strings instead of decimal (its far less CPU eating to encode decode hex if your bignums use binary form of representation) — Spektre, Sep 05 '13 at 23:16

score 1 · Accepted Answer · answered Jun 10 '11 at 14:06

For assigning big numbers the usual way is strings. This usually only happens at the interface to your calculations and so is not a problem. Another way would be to have a template constructor that takes a range of chars that you interpret as a Two's Compliment number (I prefer this method, but I havent seen it used very often).

If you are serious about implementing a BigNum class then I would suggest that you read up about Expression Templates as a way to reduce the cost of temporaries in natural looking code.

Bignum divison and value assignment in c++

1 Answers1