XPATH evaluation with libxml

[Bart]

Hello,

I am real new to using LIBXML and have a question about XPATH
evaluations. The question may show a real ignorance of the LIBXML
structure, don't so assume I know what I am talking about.

I am trying to use the XPATH functions to pull the data for individual
data elements. The order of the elements whose value I must find are
not necessarily contiguous in the XML document. So walking the tree
will not work.

I have figured out how to make the call to xmlXPathEvalExpression
function with a full XPATH starting at the root of the document. But
the documents I am working with are so large that this can be too slow
for my purpose.

I can identify a specific node of the whole document that I know that
several values I need to find are contained within. Can I use the
xmlXPathEvalExpression to find the branch I need to work with, then by
manipulating the context or the XPATH object of the
xmlXPathEvalExpression call to then do further calls to
xmlXPathEvalExpression to find specific elements in the now segregated
branch? I am thinking that I would be working with a much smaller tree
and the XPATH evaluations would work much faster.

Is this doable?

Thanks

Bart Torbert
(e-mail address removed)