Re: tests & logging

[adrianh]

Is there module or a hack to log the failed test/s? How in the world do you track these type of issues down?

Are all 40k tests in one file? If not just run the test file that's failing.

You can also use runtests from TAP::Harness with the -q or -Q option to only show test failure.

[andreas1234567]

Are all 40k tests in one file?

In that case I would seriously consider splitting the tests into smaller sets (i.e. files). Current modules on CPAN could serve as examples, e.g. a simple and easily understood one such as Number::Bytes::Human. Some test files contain just a handful tests/lines each:

[Number-Bytes-Human-0.07]# wc -l t/*
   7 t/01use.t
  42 t/02basic.t
  27 t/03large.t
  31 t/04base1000.t
  14 t/05zero.t
  15 t/06si.t
   9 t/07unit.t
  29 t/08base1M.t
 137 t/10parse_args.t
  17 t/20oo.t
   8 t/90pod.t
   8 t/98pod-coverage.t
 344 total

Run all tests:

[Number-Bytes-Human-0.07]# make test
PERL_DL_NONLAZY=1 /usr/bin/perl "-MExtUtils::Command::MM" "-e" "test_h
+arness(0, 'blib/lib', 'blib/arch')" t/*.t
t/01use.............ok 1/1# Testing Number::Bytes::Human 0.07, Perl 5.
+008005, /usr/bin/perl
t/01use.............ok
t/02basic...........ok
t/03large...........ok 1/4number too large (>= 1024**10) at t/03large.
+t line 26
t/03large...........ok
t/04base1000........ok
t/05zero............ok
t/06si..............ok
t/07unit............ok
t/08base1M..........ok
t/10parse_args......ok
t/20oo..............ok
t/90pod.............ok
t/98pod-coverage....ok
All tests successful.
Files=12, Tests=128,  1 wallclock secs ( 0.89 cusr +  0.10 csys =  0.9
+9 CPU)

Run a single test:

[Number-Bytes-Human-0.07]# prove t/05zero.t
t/05zero....ok
All tests successful.
Files=1, Tests=6,  0 wallclock secs ( 0.05 cusr +  0.00 csys =  0.05 C
+PU)