Metacat

the current implementation of xpath used in metacat for the purpose of returning specific fields of an xml document within a returnset is not sufficient for highly structured documents. more or full xpath support needs to be implemented. currently, if you specify "/x/y" as a return field and the document is structured
<x>
<y>
<z>...</z>
<z>...</z>
</y>
</x>

metacat will return nothing. metacat seems to require that you specify the exact xpath to a node or it just returns null. This really doesn't work for any document that contains formatting (such as <p>, <i>, <b>, etc) within the xml.