I do my best thinking while driving home. Last night I thought that a possible
solution that I could try is exactly what you suggest below. So, I had another
look at the code today and tried the following code changes to
TXFMXPath::evaluateExpr(...) which seems to work (on the limited test data that
I am using).
The following code fragments come from TXFMXPath::evaluateExpr(...). My changes
are commented with my initials JDM. I have left in just enough context for you
to see where in TXFMXPath::evaluateExpr(...) I have added the code.
If you would care to look it over, I would appreciate.
As to your other comment "why dont I used a more complex XPath expression" - I
probably will, but I am an "old" programer who tries to modularise things into
small, easy-to-understand bits. I find long XPath expressions too taxing on the
brain.
ta John
switch (inputType) {
case DOM_NODE_DOCUMENT :
case DOM_NODE_XPATH_NODESET : // Added by JDM.
// do XPath over the whole document and, if the input was an XPath
Nodeset, then
// later intersect the result with the input nodelist
cd = XalanDOMString("/"); // Root node
cexpr = cd.c_str();
.....
.....
....
....
...
...
..
.
for (int i = 0; i < size; ++ i) {
if (lst.item(i) == xd)
m_XPathMap.addNode(document);
else {
item = xwn.mapNode(lst.item(i));
m_XPathMap.addNode(item);
}
}
// Start Insert by JDM
if (inputType == DOM_NODE_XPATH_NODESET)
{
//the input list was a XPATH nodeset, so we must intersect the results
of the XPath
//processing done above with the input nodeset
m_XPathMap.intersect(input->getXPathNodeList());
}
// End Insert by JDM
xpesd.uninstallExternalFunctionGlobal(XalanDOMString(URI_ID_DSIG),
XalanDOMString("here"));
