htmlparse currently provides the uninterpreted tag attributes to the callback function. It would be useful if the package provided a standard way to parse these attributes instead of every user needing to do it themselves.
