While converting a couple of Test::Deep tests to Test2::V0 at $work I noticed that T2's float isn't quite a drop in replacement for Deep's num when tolerance is passed to both.
Test::Deep's check is inclusive:
abs($got - $self->{val}) <= $tolerancehttps://metacpan.org/release/RJBS/Test-Deep-1.205/source/lib/Test/Deep/Number.pm#L31
Whereas Test2's is exclusive:
$equal = 1 if
$got > $input - $tolerance &&
$got < $input + $tolerance;
https://metacpan.org/release/EXODIST/Test-Simple-1.302214/source/lib/Test2/Compare/Float.pm#L93
My current workaround is to apply a tiny fudge factor to the tolerance in the test file with float( $exp, tolerance => $tol + 1e-8 ) but this is less than ideal as it shows up in the diagnostics and is generally harder to read.
I wonder if it would make sense to have either an inclusive flag or inclusive_tolerance arg to float?
While converting a couple of Test::Deep tests to Test2::V0 at $work I noticed that T2's
floatisn't quite a drop in replacement for Deep'snumwhen tolerance is passed to both.Test::Deep's check is inclusive:
https://metacpan.org/release/RJBS/Test-Deep-1.205/source/lib/Test/Deep/Number.pm#L31
Whereas Test2's is exclusive:
https://metacpan.org/release/EXODIST/Test-Simple-1.302214/source/lib/Test2/Compare/Float.pm#L93
My current workaround is to apply a tiny fudge factor to the tolerance in the test file with
float( $exp, tolerance => $tol + 1e-8 )but this is less than ideal as it shows up in the diagnostics and is generally harder to read.I wonder if it would make sense to have either an inclusive flag or inclusive_tolerance arg to float?