At 12:27 PM +0000 2/2/04, Robin LaFontaine wrote:
Elliotte,

I'm wondering how you are going to test these tools. The 
canonicalization process should not change the data, though it will 
re-order attributes and change prefixes.

After I've canonicalized, a straight text diff or binary compare can 
find any differences. No fancy infoset comparisons are necessary. My 
plan, if I can find a reliable tool, is just to generate some 
canonicalized output from various input, then compare that output to 
the output of my own canonicalizer to see if it's working correctly 
or not. Given the current state of these tools and the lack of a 
normative test suite I wouldn't be surprised if some of the 
differences I turn up are bugs in the independent canonicalizer, so 
I'll have to check each difference manually against the spec. 
However, I should eventually be able to generate a reasonable test 
suite for canonical XML by comparing enough independent 
implementations and using the spec to figure out whose right when 
they disagree.
-- 

   Elliotte Rusty Harold
   elharo@m...
   Effective XML (Addison-Wesley, 2003)
   http://www.cafeconleche.org/books/effectivexml            
   http://www.amazon.com/exec/obidos/ISBN%3D0321150406/ref%3Dnosim/cafeaulaitA