#!/usr/bin/perl
use strict ;
use Getopt::Long ;
use TAP::Harness ;
use Test::Builder ;
use vars qw($VERSION) ;
$VERSION = "0.01" ;
sub TAP::Parser::Aggregator::ok { $_[0]->passed() - ($_[0]->skipped() + $_[0]->todo()) }
my %METRICS = (
planned => 'Expected number of planned tests',
passed => 'Expected number of passed tests (including skip and toto tests)',
ok => 'Expected number of passed tests (excluding skip and toto tests)',
failed => 'Expected number of failed tests',
skipped => 'Expected number of skipped tests',
todo => 'Expected number of todo tests',
) ;
my @METRICS = keys %METRICS ;
my %opts = () ;
GetOptions(\%opts, "runtests", map { "$_:1" } @METRICS) or usage() ;
usage() if -t \*STDIN ;
my $harness = new TAP::Harness({verbosity => -3}) ;
my $agg = $harness->runtests([\*STDIN, 'stdin']) ;
my $tb = new Test::Builder() ;
my $tap = '' ;
$tb->output(\$tap) if $opts{runtests} ;
$tb->plan(tests => scalar(@METRICS)) ;
foreach my $m (@METRICS){
my $result = scalar($agg->$m()) || 0 ;
my $expected = defined($opts{$m}) ? $opts{$m} : -1 ;
if ($expected >= 0){
$tb->is_eq($result, $expected, "$METRICS{$m} == $expected") ;
}
else {
$tb->skip("$METRICS{$m} not specified") ;
}
}
if (! $opts{runtests}){
# Just print out the TAP output
print $tap ;
}
else {
# Send it through the harness!
open(MEM, '<', \$tap) ;
TAP::Harness->new()->runtests([\*MEM, 'TAP']) ;
close(MEM) ;
}
#################################################
sub usage {
my $more = shift ;
print STDERR <<USAGE;
Usage: $0 OPTION...
metatap is a simple testing tool for TAP producers. It reads TAP from standard
input, along with the expected results on the command line. It then produces
(and optionally runs) more TAP describing how the input TAP was interpreted by
TAP::Harness.
USAGE
print STDERR
(map { sprintf(" --%-10s $METRICS{$_}\n", "$_:") } @METRICS),
" --runtests: Run the produced TAP though TAP::Harness\n" ;
exit 1 unless $more ;
}
__DATA__
=head1 NAME
metatap - Create TAP about TAP
=head1 DESCRIPTION
metatap is a simple testing tool for TAP producers. It reads TAP from standard
input, along with the expected results on the command line. It then produces
(and optionally runs) TAP describing how the input TAP was interpreted by
L<TAP::Harness>.
For example, if you send to following TAP:
1..5
ok 1 - 1 equals 1
ok 2 - 1 equals 1
ok 3 # skip because
not ok 4 - 1 equals 2
ok 5 - 1 == 2 # TODO because
into
$ metatap --planned=5 --ok=2 --failed=1 --skip=1 --todo=0
it will produce the following (TAP) output:
1..6
ok 1 - Expected number of passed tests (excluding skip and toto tests) == 2
ok 2 - Expected number of planned tests == 5
ok 3 # skip Expected number of passed tests (including skip and toto tests) not specified
ok 4 - Expected number of failed tests == 1
not ok 5 - Expected number of todo tests == 0
ok 6 - Expected number of skipped tests == 1
adding the --runtests option, the produced TAP will be run in a TAP::Harness,
yielding:
TAP .. Failed 1/6 subtests
(less 1 skipped subtest: 4 okay)
Test Summary Report
-------------------
TAP (Wstat: 0 Tests: 6 Failed: 1)
Failed test: 5
Files=1, Tests=6, 0 wallclock secs ( 0.00 usr + 0.00 sys = 0.00 CPU)
Result: FAIL
=head1 SEE ALSO
L<TAP::Harness>, L<Test::Builder>.
=head1 AUTHOR
Patrick LeBoutillier <patl@cpan.org> is the author of metatap.
=head1 COPYRIGHT
Copyright (c) 2010-2011, Patrick LeBoutillier.
All Rights Reserved. This module is free software. It may be used,
redistributed and/or modified under the terms of the Perl Artistic
License. See http://www.perl.com/perl/misc/Artistic.html for more
details.
=cut