how to represent number bigger than long in c

Question

I want a variable to represent the number 600851475143 in C but this value is larger than the maximum value of long i.e 2147483647.What data type should I use?

`unsigned long`? `long long`? `unsigned long long`? Use a bigint library? — , Sep 07 '13 at 15:52
...strings? `double`? `long double`? Everything depends on your needs! — nneonneo, Sep 07 '13 at 15:55
Search for C data types and its sizes. Some even compilers support a 128 bits integer. — The Mask, Sep 07 '13 at 16:06
You should tell us what are you going to do with that number if you want a better advice. You gave us too little info on the problem you are trrying to solve. — LorenzoDonati4Ukraine-OnStrike, Sep 07 '13 at 16:24

chux - Reinstate Monica · Accepted Answer · 2013-09-07T17:22:48.877

4

Use integer type long long.
It is good for at least the range -9223372036854775807 <= x <= +9223372036854775807.

#include <stdio.h>
int main(int argc, char *argv[]) {
  long long big = 600851475143;
  printf("%lld\n", big);  // prints 600851475143
  return 0;
}

Alternatively, one could use int64_t or uint64_t - 64-bit types.

The range mentioned above is the minimal range specified in C11 5.2.4.2.1.

edited Sep 07 '13 at 17:22

answered Sep 07 '13 at 16:43

chux - Reinstate Monica

143,097
13
135
256

SevenBits · Answer 2 · 2013-09-07T20:17:22.847

3

Use an unsigned long. This will allow only numbers >= 0, but it will extend your range. I would make sure that your long type is 64-bit, as this will also extend your range of values.

You can also use gmplib, which is released under the LGPL: http://gmplib.org/

edited Sep 07 '13 at 20:17

answered Sep 07 '13 at 15:55

SevenBits

2,836
1
20
33

1

Probably a typo, but just to avoid leaving incorrect info in a post: "This will allow only numbers **>= 0**" – LorenzoDonati4Ukraine-OnStrike Sep 07 '13 at 16:20
@SevenBits To insure the integer type is 64-bit, simple use `int64_t` or `uint64_t`. – chux - Reinstate Monica Sep 07 '13 at 17:20
@LorenzoDonati Thanks, yes, it was a typo. I'll correct it now. – SevenBits Sep 07 '13 at 20:17

score 2 · Answer 3 · answered Sep 07 '13 at 16:04

Besides using an arbitrary precision integer math library, as suggested by @H2CO3, which is probably you best bet if you plan to work with such big numbers without worrying for overflows, you should check your compiler/platform documentation for the maximum representable number.

If you don't need signed arithmetic you could use unsigned long or unsigned long long, otherwise signed long long. Keep in mind that the standard doesn't mandate their size exactly, only a minimum is specified (from C99 draft standard - Annex E):

#define LONG_MAX +2147483647
#define LONG_MIN -2147483647
#define LLONG_MAX +9223372036854775807
#define LLONG_MIN -9223372036854775807
// [...]
#define ULONG_MAX 4294967295
#define ULLONG_MAX 18446744073709551615

but the size of thoose types it self is implementation defined again, isn't it? Or why on all plattforms i know long's size differs in 32 and 64 bit arch's? — dhein, Sep 07 '13 at 17:03
@Zaibis Lorenzo Donati posted the _minimum_. The values are implementation defined, but will _at least_ cover the above range, thus a `long` could be 4 or more 8-bit bytes. — chux - Reinstate Monica, Sep 07 '13 at 17:15

how to represent number bigger than long in c

3 Answers3

Linked