On Wed Jan 19 01:29:06 2011, payload wrote:
> On Mo, 2011-01-17 at 13:34 -0800, Moritz Lenz via RT wrote:
> > On 01/16/2011 07:47 PM, Gilbert R. Roehrbein (via RT) wrote:
> > > fixes the problem which you encounter when you try to evaluate
> > > +(23..2300000000)
> >
> > ... and creates others. Consider
> >
> > (0..^3.3).Numeric
> >
> > Where you patch makes it return 3, but 4 is the correct answer.
>
> shame :( but fixed it ^^
>
> > Maybe checking for ~~ Int instead of ~~ Numeric helps, but we also need
> > to more tests.
>
> does now work for following ranges
>
> for $(^3.3), $(0..3), $(-2..0), $(-2.9..3.1), $(1.9..3.1) {
> my $a := +$_;
> my $b := .elems;
> say "{.perl}\t\t$a != $b" if $a != $b;
> }
Sorry about the delay:
Added tests to S02-types/range.t, and applied a heavily modified version of the
patch (there
were some edge cases that were still failing with the patch as is.)
$ time ./perl6 -e 'say +(123..123121231231231212321)'
123121231231231212199
real 0m0.475s
user 0m0.369s
sys 0m0.101s
--
Will "Coke" Coleda
