Anthony B. Coates wrote

> I would suggest that this won't give you any meaningful results.  The 
> complexity of an XSLT transformation is driven by the 
> structural/formatting differences between the input and the output 
> formats.  It is not directly related to the complexity of the schemata
> for  the input/output formats.

agreed....I would suggest even a simple diff and counting the nodes,
between input and result trees, would be a rough indicator of XSLT
complexity...though these indirect type measurements can be quite brittle.

-Jim Fuller