Simple calculator not working with big numbers in C

Question

I'm trying to build a simple calculator that would perform multiplication and addition.

I have this code

#include <stdio.h>
#include <string.h>

long result = 0;
long *resultPointer = &result;

long plus(long *current, long num) {
    return (*current + num);
}

long krat(long *current, long num) {
    return (*current * num);
}

int main() {
    char operator[4];
    long num;

  printf("%d\n", 0);
    while(scanf("%s %li", &operator[0], &num) != EOF){
        if (num > 0) {
            if (strcmp(operator, "krat") == 0) {
                *resultPointer = krat(&result, num);
            }

            if (strcmp(operator, "plus") == 0) {
                *resultPointer = plus(&result, num);
            }

            printf("%li\n", result);
        }
    }

    return 0;
}

this is the input for the program

plus 123456789
krat 123456789
plus 0
krat 2
krat 3
krat 4
krat 5
krat 6

and this is the output

0
123456789
15241578750190521
30483157500381042
91449472501143126
365797890004572504
1828989450022862520
-7472807373572376496

The problem is that when the numbers get bigger and bigger they turn to negative. Is this the problem with memory allocation for the variables and how to address this?

Answer 1

You're overflowing the variables.

You have two options here. Either find a solution for arbitrarily large numbers (there are libraries you can use) or accept that you cannot use numbers that are to big.

The largest number a signed 64 bit integer can hold (it's obvious from your output that long is 64 bit on your system) is 9223372036854775807 and the largest for an unsigned 64 bit is 18446744073709551615.

It could be worth mentioning that overflow only have a defined behavior for unsigned types.

Other comments

char[4] is not enough to hold "plus", since you need place for the '\0'. Use char[5] instead.

There is no reason to use pointers in this code. Unless you have a very good reason, I would suggest changing to:

long plus(long current, long num) {
    return current + num;
}

Which in turn means that you can completely skip the function. You don't need a function to perform addition of two integers since you have the + operator to do that very thing.

Also, your usage of scanf is unsafe. You may write past the end of the array. Do this instead: scanf("%4s %li" Note the 4. It gives a maxlength for the string. On top of that, you should not check scanf for EOF. Read the documentation about what it actually does return.

Answer 2

Variables in programming languages have different size. In c++, long size is at least 32 bits (according to Fundamental Types (C++) (Microsoft website) and C++ Data Types (tutorialspoint website)). So, when you assign a number bigger which needs more bits to store number, it turns to negative (one's complement and 2's complement system in storing numbers).

read C++ Data Types (tutorialspoint website) about the size of each variable in c++ language and also wikipedia has good articles about one's complement and two's complement.