Front page | perl.perl5.porters |
Postings from February 2020
On Sat, Feb 1, 2020 at 9:18 AM James E Keenan <jkeenan@pobox.com> wrote: > > I'm having trouble figuring out how to use Porting/bisect.pl to identify > the first point at which, with given configuration options, a file in > the test suite began to pass. It looks like --expect-fail differentiates between the error and failure states, not between good and bad exit values. I would use a wrapper script (or one-liner if that's easier) that called the .t file and inverts its $? to give the results that git bisect expects -- zero for the test fails (i.e. a "good" result for bisect), and non-zero for the test passing (a "bad" result for bisect) -- and pass that wrapper to bisect.pl instead of the .t file itself.Thread Previous