#!/usr/bin/perl

use strict;
use warnings;

while(<DATA>) {
  /^#/ and next;
  chomp;

  if ( m{ <([^>]*?)> [^<]*? </\1> }x )
  {
    while ( m{ <([^>]*?)> [^<]*? </\1> }gx )
    {
      my $token = $1;

      # Some verbose info. Note first line doesn't
      # get printed because it doesn't match regex
      print $_, "\n";
      print "Found <$token> and </$token>, removing...\n";

      s{<$token>}{}g;
      s{</$token>}{}g;
    }
  }

  else
  {
    # <A HREF="http://google.com">search engine</A>
    while ( m{ </([^>]*?)> }x )
    {
      my $close = $1;
      if ( m{ <($close[^>]*?)> [^<]*? </$close> }x )
      {
        my $open  = $1;
        print $_,"\n";
        print "Found <$open> and </$close>, removing\n";
        s{<$open>}{}g;
        s{</$close>}{}g;
        print $_,"\n";
      }
    }
  }

  # Of course, this will be hard to do if you
  # "don't know how many dashes, if any, will
  # be there."
  print "\t$_\n" for ( split (/-- /,$_) );
  print "\n";

}

__DATA__
This is a -- string of -- words
<b>This is a -- string of -- words</b>
This <b>is a -- string</b> of -- words
This <i>is</i> a -- <b>string</b> of -- words
This <i>is a -- <b>nested set</b> of</i> -- tokens
This is -- an awesome -- <A HREF="http://google.com">search engine</A>
Truly an -- ugly -- <A HREF="http://perl.com"><FONT COLOR="RED">nested</FONT> st
ring</A>
#This string -- <b>causes -- <A HREF="http://perl.com"><FONT COLOR="RED">my box</
b></FONT> to hang</A>