Programming Perl, 4th Edition

bigint

use bigint;

This pragma bypasses the architecture-dependent treatment of integer operations to work with very large numbers, as well as handling the special value NaN (for not a number):

use bigint;
say 2 ** 512;

This pragma works by overloading the numeric operators to use Math::BigInt to compute values. As such, it can be considerably slower than built-in operations. Still, slow right answers are always better than fast wrong ones.

You can load different implementing libraries, which may vary in performance. By default, bigint uses a pure Perl implementation, but you can load a faster library if you have it:

use bigint lib => 'GMP';

You can set the accuracy, which determines the number of significant digits in the answer:

use bigint a => 2;

Or, you can set the precision, which specifies the magnitude of the answer. A precision less than 0 is ignored:

use bigint p => –2;   # to the hundredths place (ignored)
use bigint p =>  1;   # rounded to 10