PHP and Perl and Python/Ruby polyglot

polyglot.pl.php.py.rb

#<?php eval('echo "PHP Code\n";'); __halt_compiler(); ?>

print ((("b" + "0" == 0) and eval('"Perl Code\n"')) or (0 and "Ruby Code\n" or "Python Code"));

__DATA__ = 1
"""""
__END__

===== This is comment of all. =====

outputs:

  $ perl polyglot.pl.php.py.rb
  Perl Code
  $ php polyglot.pl.php.py.rb
  #PHP Code
  $ python polyglot.pl.php.py.rb
  Python Code
  $ ruby polyglot.pl.php.py.rb
  Ruby Code
  $

"""

If the eval(...) in the PHP section doesn't do anything, it can be compressed and the hash can be removed.
#<?php echo "\x08PHP Code\n";__halt_compiler();?>

you can run php with the output_buffering parameter enabled php --define output_buffering=On $file.
then in the first line fo your polyglot you can use ob_end_clean() to drop everything that was printed before and __halt_compiler() to not even read the rest, thus yielding a "real" PHP on stdout, without magic/invis characters. e.g.:

#define X[<?php ob_end_clean();echo"PHP\n";__halt_compiler();#print(("ALGOL68",newline))COMMENT]