On Tue, Sep 26, 2006 at 10:52:12AM -0400, David Golden wrote:
>    On 9/26/06, Tim Bunce <[EMAIL PROTECTED]> wrote:
> 
>      Err, thanks. But there's almost no information about the problem.
> 
>      I'm CC'ing this to [EMAIL PROTECTED] to ask if anyone knows
>      why it said "Detailed output could not be captured with test.pl and 
> make."
> 
>    This is a limitation in the current version of CPAN::Reporter.
> 
>    My understanding is that ExtUtils::MakeMaker creates a Makefile
>    that executes test.pl directly and uses the result code to
>    determine success or failure.  It does not pass test.pl to
>    Test::Harness, so CPAN::Reporter can't reliably parse for success.
>    (And it gets run right after all tests from t/*.t).
> 
>    As a result, I can't tee the output of test.pl as that obscures the
>    exit code.  I can't just capture without teeing or there would be
>    no visible output, no way to respond to prompts and no way to know
>    if the test has hung.  So when test.pl exists and "make test" is
>    the command from CPAN.pm, CPAN::Reporter punts. [...]

So the simple existence of a test.pl file prevents CPAN::Reporter
capturing the output of the t/*.t tests?

>    I'll grant it needs a better diagnostic.

True :)

>    Andreas has suggested running "make test" again on failure, the
>    second time with output captured.  I haven't gotten around to
>    implementing it -- and I'm not sure if I want to do so without
>    adding a prompt and having an option to turn it off.  In part, I
>    was waiting to see if it would be even be a problem -- I don't
>    think many distros still use test.pl.
> 
>    Well, DBI is important enough for it to be a problem.
> 
>    test.pl is flagged as being "mostly an empty shell".  Can it be
>    removed entirely or moved to the examples directory?

I'll rename it for the next release. It's not very useful these days.

>    Or do I need to prioritize getting some sort of output capture even
>    in the case of test.pl?  In this case, since t/*.t is run as well,
>    the Test::Harness "All tests successful" line would get picked up
>    and the the results of test.pl would be ignored.

That seems like the best approach. Make won't run test.pl is any of the
t/*.t tests failed. So if you know that none failed but make failed
then you'll know it was test.pl that failed.

Thanks for your work on this David.

Tim.

>    Regards,
>    David Golden
> 
> References
> 
>    Visible links
>    1. mailto:[EMAIL PROTECTED]
>    2. mailto:[email protected]

Reply via email to