I have to disagree with the index array not being useful. In fact, I see it as being absolutely necessary in the example I used, due to multiple <random> tags being used. Otherwise, how would you be able to distinguish which LI values belong to which RANDOM tag? It’s not really possible to do that without the index array.

The more I look at using xml_parse_into_struct though, the more I see that it’s use will probably cause more problems than it solves. For now I’ll concentrate more on streamlining the current code and getting it to parse templates more efficiently. At least I know that what I’ve written so far works (if slowly).

TRUE.  The function, xml_parse_into_struct is powerful, though parsing XML is one of the earliest steps in terms of going back to the drawing board.  I wonder how that struct was designed to be used?

  [tag] => NOTE
  [value] =>
  [type] => cdata
  [level] => 1

What do you think is with that cdata type?

Since the ‘value’ element is empty, I would guess (all I can do, since there’s no context supplied) that the NOTE tag has only child elements (or even a single element, or perhaps it’s an atomic tag with neither contained text nor children? Again, without context it’s hard to say, since I’m not completely familiar with the structure of the arrays that are created by the function.