While testing test scripts I noticed that their output is defferent
when running directly or under Harness.
I have scripts like this: (called a.t)
use strict;
use warnings;
use Test::More tests => 2;
is(2, 2);
is(2, 3);
and I'd like to test if their output is what I expect.
(These are examples for my training)
A simple script to test the above is this: (called t.t)
use strict;
use warnings;
use Test::More tests => 1;
system "$^X a.t > out 2>&1";
open my $in, '<out' or die;
my @lines = <$in>;
my @expected = <DATA>;
is_deeply(\...@lines, \...@expected);
__DATA__
1..2
ok 1
not ok 2
# Failed test at a.t line 7.
# got: '2'
# expected: '3'
# Looks like you failed 1 test of 2.
If I run "perl t.t" I get OK
If I run perl -MTest::Harness -e'runtests("t.t")'
it fails with
# Failed test at t.t line 11.
# Structures begin differing at:
# $got->[3] = '
# '
# $expected->[3] = '# Failed test at a.t line 7.
# '
that is, under Harness there is an extra empty row
between the not ok line and the explanation line:
# Failed test at a.t line 7.
Is this intentional?
If not should this be fixed?
Gabor
