Using big numbers in Perl

Question

I have a scenario where I take 2 very big binary strings (having 100 characters) and I need to add them.

The issue is that I am getting the answer in the form 2.000xxxxxxxxxxe+2, whereas I need the precise answer, as another 100 character long string.

chomp($str1=<STDIN>);
chomp($str2=<STDIN>);
print "Str 1 is $str1\n";
print "Str 2 is $str2\n";
$t = $str1 + $str2;
print "Sum is $t\n";

Sample Input

1001101111101011011100101100100110111011111011000100111100111110111101011011011100111001100011111010
1001101111101011011100101100100110111011111011000100111100111110111101011011011100111001100011111010

Sample Output
Str1 is

1001101111101011011100101100100110111011111011000100111100111110111101011011011100111001100011111010

Str2 is

1001101111101011011100101100100110111011111011000100111100111110111101011011011100111001100011111010

Sum is

2.0022022220202e+099

My question is that if str1 and str2 can be printed successfully without using BigInt, then why its sum cannot be printed. — Saurabh Shrivastava, Nov 19 '14 at 11:13
@user3388005, because you are printing the strings. When you add them the values are converted to numbers (some kind of floating point), and you get the number result. — Qtax, Nov 19 '14 at 11:17
you might want to use pack/unpack: http://perldoc.perl.org/perlpacktut.html — bytepusher, Nov 19 '14 at 11:28
I have never used pack, tutorial seems difficult to understand. Can you tell me how to use pack for this particular problem. — Saurabh Shrivastava, Nov 19 '14 at 11:44
why don't you read the docs for BigInt as suggested? There is an as_bin output method — Vorsprung, Nov 19 '14 at 11:46
I am not allowed to use extra module like Big Int. I have to run my code on an online site and there is no way that i can add an extra module. — Saurabh Shrivastava, Nov 19 '14 at 11:57
Then you will have to reimplement addition for numeric strings of arbitrary length. Your ad-hoc implementation is almost certainly slower, buggier, and less secure than a standard library; perhaps you can do it really poorly in order to persuade management to reassess this particularly braindead policy. — tripleee, Nov 19 '14 at 13:35
But it's not an "extra module"; it's included in the Perl distribution. — tripleee, Nov 19 '14 at 13:37

score 4 · Answer 1 · answered Nov 19 '14 at 12:18

As already suggested, you can use Math::BigInt core module,

use Math::BigInt;

# chomp($str1=<STDIN>);
# chomp($str2=<STDIN>);
# print "Str 1 is $str1\n";
# print "Str 2 is $str2\n";

my $t = Math::BigInt->new("0b$str1") + Math::BigInt->new("0b$str2");

print $t->as_bin;

score 0 · Answer 2 · edited May 23 '17 at 12:12

In order to perform arithmetic on your strings, Perl converts them to floating-point numbers, which are inherently imprecise. If you want to avoid that, use Math::BigInt as already suggested ... or roll your own.

######## WARNING/GUARANTEE: This is practically certain to be
# slower, buggier, less portable, and less secure than Math::BigInt.
# In fact, I planted a security hole just to prove a point.  Enjoy.

use strict;
use warnings;

sub addition {
    my ($int1, $int2) = @_;
    my @int1 = reverse split //, $int1;
    my @int2 = reverse split //, $int2;
    my $len = scalar(@int1>@int2 ? @int1 : @int2);

    my @result;
    my $carry = 0;
    for (my $i=0; $i < $len; ++$i)
    {
        $int1[$i] |= 0;
        $int2[$i] |= 0;
        my $sum = $carry + $int1[$i] + $int2[$i];
        if ($sum >= 10)
        {
            $carry = int($sum / 10);
            $sum %= 10;
        }
        push @result, $sum;
    }
    push @result, $carry if $carry;
    return join ('', reverse @result);
}

Using big numbers in Perl

2 Answers2