Nice polyglot. It's extra-defensive. In the way of explanation, since not all Windows command scripts are so complicated,

@rem = '--*-Perl-*--
@echo off
perl -x -S "%~dpnx0" %1 %2 %3 %4 %5 %6 %7 %8 %9
goto endofperl
@rem ';
#!perl
#line 8

   YOUR PERL CODE HERE;

__END__
:endofperl
[download]

This can be run as a Windows command file itself, or as a Perl script directly, and the syntax hacks are designed to fool both interpreters into ignoring the parts intended for the other.

In perl, lines 1..5 are @rem = '...'; and the resulting array is otherwise ignored. In Windows script, that's five separate commands: (1) an unechoed remark @rem, (2) an unechoed command to turn off further automatic echoes @echo off, (3) an invocation of perl on the script itself with its first nine arguments, (4) a command to skip any other lines until one matching /^:endofperl/, and another unechoed remark @rem. The first line, as a bonus, suggests the file should be edited in perl-mode for some smart editors.

The perl -x tells perl to skip anything ahead of the shebang line matching /^\#\!/. This is only seen when the command script invokes Perl. It shouldn't be necessary since the @rem=''; trick should armor these lines, but it's defensive. The odd "%~dpnx0" reconstructs the %0 variable explicitly with all possible components: drive, path, name, extension.

The #line 8 trick aligns perl's error-reporting features with the actual polyglot script's lines. Otherwise, any errors would be reported, thanks to the perl -x as numbered relative to the shebang line.

Of course, perl quits reading at the __END__ token, so anything following is ignored by perl, but the last line is the aforementioned target for the command script's goto endofperl command.

--
[ e d @ h a l l e y . c c ]

---

---

• Are you posting in the right place? Check out Where do I post X? to know for sure.
• Posts may use any of the Perl Monks Approved HTML tags. Currently these include the following:

  <code> <a> <b> <big> <blockquote> <br /> <dd> <dl> <dt> <em> <font> <h1> <h2> <h3> <h4> <h5> <h6> <hr /> <i> <li> <nbsp> <ol> <p> <small> <strike> <strong> <sub> <sup> <table> <td> <th> <tr> <tt> <u> <ul>

• Snippets of code should be wrapped in <code> tags not <pre> tags. In fact, <pre> tags should generally be avoided. If they must be used, extreme care should be taken to ensure that their contents do not have long lines (<70 chars), in order to prevent horizontal scrolling (and possible janitor intervention).
• Want more info? How to link or How to display code and escape characters are good places to start.