I've written a program which is distributed as a single script, but implemented as a series of modules. Distributing the script as one file avoids the need to "install", or mess with include-paths.

Unfortunately this means that my script looks like this:

# ... pod for module1
# ... code for module1

# ... POD for module2
# .... code for module2

# package main
# pod
# code
[download]

This screws up my preferred usage of POd::Usage, as it shows the pod for all the script. I've read the documentation for Pod::Usage, where it suggets using Pod::Find and the current package (__PACKAGE__) but I cannot make it work, as the following demonstrates.

Is there a simple solution, or am I out of luck?

Test program follows:

#!/usr/bin/perl -w

use strict;
use warnings;

use Getopt::Long;
use Pod::Find qw(pod_where);
use Pod::Usage;



=head1 NAME

Foo - Do something.

=cut

=head1 SYNOPSIS

  use strict;
  use Foo;

  my $f = Foo->new();
  $f->doStuff();

=cut

=head1 DESCRIPTION

This class encapsulates stuff.

=cut

package Foo;


=head2 new

Constructor

=cut

sub new()
{
    my ( $proto, %supplied ) = (@_);
    my $class = ref($proto) || $proto;

    my $self = {};

    bless( $self, $class );
    return $self;

}


=begin doc

Do stuff

=end doc

=cut

sub do_stuff()
{
}





=head1 NAME

My Script - Do something

=cut

=head1 SYNOPSIS

Usage:

  $0 [options]

  options

   --help     Show the help
   --man      Show the manual.

=cut

=head1 DESCRIPTION

This is the utility to do stuff ..  It uses the Foo module internally.

=cut

package main;




#
#  Configuration variables
#
my %CONFIG;




#
#  Parse the command line.
#
exit
  if (
    !Getopt::Long::GetOptions(

        # Help options
        "help",   \$CONFIG{ 'help' },
        "manual", \$CONFIG{ 'manual' },
    ) );


Pod::Usage::pod2usage() if ( $CONFIG{ 'help' } );
Pod::Usage::pod2usage(-verbose => 2 ) if ( $CONFIG{ 'manual' } );


print "I am alive\n";
exit( 0 );
[download]