| Subject: | AUTOMATED_TESTING and failure t/98_pod_coverage.t |
Hello,
We use AUTOMATED_TESTING=1 to build a large number of modules
in an automated manner, but the logic in t/98_pod_coverage.t
ironically broke this recently. We don't have RELEASE_TESTING set.
"unless" combined with and/or error prone, so I'm not sure what
the intention is below:
t/98_pod_coverage.t
# Don't run tests during end-user installs
use Test::More;
unless ( $ENV{AUTOMATED_TESTING} or $ENV{RELEASE_TESTING} ) {
plan( skip_all => "Author tests not required for installation" );
}
# Failed test 'Test IPC::Run::Win32Helper that all modules are documented.'
# at t/98_pod_coverage.t line 38.
# IPC::Run::Win32Helper: requiring 'IPC::Run::Win32Helper' failed
# Failed test 'Test IPC::Run::Win32IO that all modules are documented.'
# at t/98_pod_coverage.t line 39.
# IPC::Run::Win32IO: requiring 'IPC::Run::Win32IO' failed
# Failed test 'Test IPC::Run::Win32Pump that all modules are documented.'
# at t/98_pod_coverage.t line 40.
# IPC::Run::Win32Pump: requiring 'IPC::Run::Win32Pump' failed
# Looks like you failed 3 tests of 7.
End-users also use AUTOMATED_TESTING in practice, usually to
make chatty installations like Term::* to just do it.
There's TEST_AUTHOR=1 as a recommended alternative:
http://search.cpan.org/~thaljef/Test-Perl-Critic-1.02/lib/Test/Perl/Critic.pm
Thanks,
Brad