There should be no need to do so, unless it's to make downstream maintenance slightly easier. Write a Perl routine to replace the standard Perl variables, like $., by their English names, and remember to follow the performance-related advice in English.

#!/usr/bin/perl
use strict;
use warnings;

use PPI;

my $script;
{
    local $/ = undef;
    $script = <DATA>;
}

my $doc = PPI::Document->new(\$script);
my $vars = $doc->find('PPI::Statement::Variable'); 
my @var_names = map { $_->variables } @$vars;

print "$_\n" for @var_names;
__END__
#!/usr/bin/perl
use strict;
use warnings;

my $foo = 'bar';
our $blah = 'asdf';

{
    local $_ = 'hello world';
    print $_, $/;
}
[download]

I was hoping there was a tutorial but I couldn't find one. It seems like PPI::Transform would be the right tool if you can figure out how to use it. I remember reading this article a few years ago - perhaps it will help/inspire you.

You want to look for PPI::Token::Magic probably.