How should text be normalized for tests?

Expat returns separate #text fragments for newlines and space. This
results in extra text nodes in my average DOM tree (unless normalize()
is called of-course). I believe this is acceptable but the
nodegetfirstchildnull test assumes that the second node is employeeId:

67       doc = load("staff");
68       elementList = doc.getElementsByTagName("employee");
69       employeeNode = elementList.item(0);
70       employeeList = employeeNode.getChildNodes();
71 --->  secondChildNode = employeeList.item(1);

This is a #text node for me.

-- 
May The Source be with you.

Received on Monday, 25 March 2002 00:56:01 UTC