it is prefered and more smoother that the getChildren()
method in Node returns and empty NodeList instead of
null when no childs exists.
Also org.htmlparser.util.NodeList to implement
java.lang.Iterable so 1.5 foreach can target it instead
of getting toNodeArray()