All document parsing is accomplished through objects -- %tidy_node% and %tidy_attr% representing a node and attribute respectively. ]]> ]]>