Bignum will probably do what you want with minimum fuss.

use 5.012;
use bignum;
say 1/18690250981623985612839659816239856;
__END__
Output:
0.00000000000000000000000000000000005350382940192655191511204088952992
+471243
[download]

You will probably want to read up on bignum for options on how to control accuracy and precision (more info in the related Math::BigInt). These libraries support arbitrary precision, within the limits of available memory, that is.