Making Programs Faster

`parse_text`

Old tokenizer code:

     split /([A-Z]  <          # Escape code and open bracket
             (?: <+ \s ) ?     # Possible extended delimiter
            )/x;

New tokenizer:

     split /([A-Z]  <          # Escape code and open bracket
             (?: [^<>]* >      # ...and the rest of the escape sequence
               | (?: <+ \s )?  # OR a possible extended delimiter
             ))/x;