Modified parse_token() to allow for '-' characters in molecule names (derived from file names, etc.)
Instead of throwing a RelaxError if an invalid range is found, it is assumed that the '-' character
does not signify a range and is treated as a string character which is part of the token.
