XHTML parser fails to recognize properly the type of links

<a href="something#other">

    /**
     * Recognize the passed reference and figure out what type of link it should be:
     * <ul>
     *   <li>UC1: the reference points to a valid URL, we return a reference of type "url",
     *       e.g. {@code http://server/path/reference#anchor}</li>
     *   <li>UC2: the reference is not a valid URL, we return a reference of type "path",
     *       e.g. {@code path/reference#anchor}</li>
     *   <li>UC3: there is an anchor but no reference, we return a reference of type "doc",
     *       e.g. {@code #anchor}. In this case we point to the current content. </li>
     *   <li>UC4: the reference is empty, we return a reference of type "doc". In this case we point to the current
     *       content. </li>
     * </ul>
     *
     * @param rawReference the full reference (e.g. "/some/path/something#other")
     * @return the properly typed {@link ResourceReference} matching the use cases
     */