If it really is an XML ID (declared as such in the DTD or schema), you 
may be able to speed that up by replacing
	//excel_database[@id=current()/text()]/@name
with
	id(current())/@name
(Take the value of the current node, do a table lookup to find the 
element which carries that id value, take its name attribute.) The 
advantage is that this avoids having to search the whole document for 
the desired entry; the disadvantage is that unless it's properly 
declared it doesn't work.

If you have to use the search approach, you may want to consider 
modifying it:
	//excel_database[@id=current()][1]/@name
... assuming that the ID is unique, this generates the same result but 
avoids wasting time searching for another instance.

BTW, testing current()/text() checks whether a single text=-node child 
has that value -- which may or may not be what you want if there's 
markup inside the current() node. If you want the string value of the 
whole node, you can just test it directly or use the string() function 
(which comes out to the same thing).

-- 
Joe Kesselman / Beware the fury of a patient man. -- John Dryden