xmlXPath

doc = xmlRead("http://www.w3.org/TR/2009/REC-xml-names-20091208/xml-names-10-3e.xml");

// Retrieve the nodes with name equals to "note"
xp = xmlXPath(doc, "//note");
s = size(xp);
for i=1:s(2)
    xmlDump(xp(i))
end

// Count the nodes with name equals to "note"
xp = xmlXPath(doc, "count(//note)")

// Retrieve the node with id="Philosophy"
xp = xmlXPath(doc, "//*[@id=""Philosophy""]");
s = size(xp);
if (s(2) <> 0) then
    xmlDump(xp(1))
end

// Retrieve the nodes with an attribute num equals to the number 5
xp = xmlXPath(doc, "//*[number(@num)=5]");
s = size(xp);
if (s(2) <> 0) then
    xmlDump(xp(1))
end

// Get the name and the content of all the attributes of nodes named 'emph'
xp = xmlXPath(doc, "//emph/@*");
xp.name
xp.content

xmlDelete(doc);

// Query using namespaces
doc = xmlReadStr("<root xmlns:scilab=""http://www.scilab.org""><scilab:a att=""foo"" rib=""bar""><b>Hello</b></scilab:a></root>");

// We search for element named a
xmlXPath(doc, "//a") // => nothing
xmlXPath(doc, "//scilab:a", ["scilab" "http://www.scilab.org"]) // => OK

// This code will fail:
// xmlXPath(doc, "//scilab:a") // => error

xmlDelete(doc);

// Query starting on an element
doc = xmlReadStr("<root att=""attribute""><a a1=""A1"" a2=""A2"" a3=""A3""><b>Hello</b><c>Scilab</c><b>World</b></a><b>Nothing</b></root>");
e = doc.root.children(1);

// Get the attributes of e
xp = xmlXPath(e, "@*");
xmlAsText(xp)

// Get the 'b' from e
xp = xmlXPath(e, "b");
xmlAsText(xp)

xmlDelete(doc);