| Subject: | BigFloat overloaded int() changing its input |
| Date: | Tue, 26 Apr 2016 16:37:06 +1000 |
| To: | bug-Math-BigInt [...] rt.cpan.org |
| From: | Kevin Ryde <user42_kevin [...] yahoo.com.au> |
With Math-BigInt 1.999719 on recent debian i386 perl 5.22.1, a program
use strict;
use Math::BigFloat;
my $x = Math::BigFloat->new('0.5');
print "x is $x\n";
my $i = int($x);
print "i is $i and x is now $x\n";
prints
x is 0.5
i is 0 and x is now 0
I hoped int($x) would return a rounded value and leave its input $x
unchanged. That being of course is how int() behaves on ordinary
scalars, and I believe how previous BigFloat behaved too.
This is per a report from Slaven Rezic on tests of some of my code failing
https://rt.cpan.org/Ticket/Display.html?id=113960
and noting there the problem seems tickled by the immediately previous
BigFloat 1.999718 in addition to 1.999719.