Yes, thats how I've been doing it

$editted_copy = $string;
# Look for codepoints not in Basic Latin
while ( $string  =~ s/(\P{InBasic_Latin})// ) {   
  my $U_char = $1;
  my $U_codepoint = ord($U_char);  
  #and try to replace them
  if( 
     defined( $subs{$U_codepoint} )
  && exists( $subs{$U_codepoint} )){
    $editted_copy =~ s/$U_char/$subs{$U_codepoint}/;
  }
  else{
    #add the missing U_codepoint by hand to the %subs hash
    #and iterate till no more U_codepoints causing problems 
  }
[download]

(I was just hoping for something prettier)
RichardH