Can't understand what is wrong

[Franky4]

I have got a task to create a program which would take two whole numbers up to 100 digits each and print out their difference in full digits without shortening. I wrote this little program which seems to work only for small numbers so I guess it's some C principal I don't understand. Looking for an answer.

  #include <stdio.h>
#include <conio.h>
#include <iostream>

const int size = 200;

double ivestis();
int laips(int, int);

int main(){
	
	char * mass;
	double C;
	mass = new char[size];
	int flag = 0;
	
	C = ivestis();
	
	for (int i=0; i<=size; i++) mass[i] = '*';
	for (int j=0; j<=size; j++){
	mass[j] = C / laips(10,size-j);
	C -= mass[j] * laips(10,size-j);
	}
	for (int k=0; k<=size; k++) if (mass[k] != 0 && mass[k] != '*') {
	flag = k;
	break;}
	for (int l=flag; l<=size; l++) printf("%d", mass[l]);

	getch();
	return 0;		
}

double ivestis(){
	double A,B;
	printf("Number A: ");
	std::cin >> A;
	printf("Number B: ");
	std::cin >> B;
	return (A-B);	
}

int laips(int sk, int n){
	int back = 1;
	for(int i=1; i<=n; i++)	back *= sk;
	return back;
}

[keskiverto]

Arithmetic types have rather limited number of digits that they actually can hold.
http://en.cppreference.com/w/cpp/language/types

That says ~15 for doubles. Check Boost for arbitrary precision types.
http://www.boost.org/

[mobotus]

http://www.cplusplus.com/forum/lounge/32041/