The only thing I have come up with so far is to add some special comments before and after the code that I don't want to execute under test, then make the test script read the script under test into a scalar or array, strip out the code that should not be executed, then eval what is left over to create the package and subroutines/methods, then run the tests against those subroutines. Is there a way of achieving the same effect without jumping through the extra hoops? Is there a way for a perl script to tell if it is being run from a shell versus being 'require'-ed? If it makes a difference, the script will be run on Win32.

#!/usr/local/bin/perl -w
#<ignore_during_test>
my $obj = SupportingPackage->new;
$obj->do_something;
$obj->do_something_else;
#</ignore_during_test>
package SupportingPackage;
sub new { bless {}, $_[0]; }
sub do_something { 1;}
sub do_something_else { 2;}
[download]

use strict; use warnings;
use Test::More tests => 2;
#... pull in package SupportingPackage somehow 
#without running code that is in main ...
# then:
ok(my $obj = SupportingPackage->new);
can_ok($obj, 'do_something');
is($obj->do_something, 1);
can_ok($obj, 'do_something_else');
is($obj->do_something_else, 2);
# etc.
[download]