RE: Canonical XML & Thinking output comparing output

[Lynda VanVleet <lvanvleet@classiq.com>]

Relating to Carmelo's and Dave's posts I have published my BETA version of
the XML Tutorial that will be packaged with our XML Testing Tool Evaluation
CD

http://www.classiq.com/XMLTutorial_files/frame.htm

Slides 1-10 show the test harness view and test case view

Slide 12 shows the returned data content in a test case
Slides 15 & 16 show the saved expected results in a test case
Slide 18 shows the difference report  (we hope to use XSLT to show a better
browser view of this soon)
Slide 19 shows the data content comparison in the test case

We basically use our own version of Canonical XML for comparison.  We ignore
comments, white space, PIs, etc.  We do not order attributes but that will
be done before release.  We do comparison on data content at the byte level
but we only report at the element or attribute level.

Since we plan to exhibit at XML conferences soon we are converting this code
into a command line tool we can give away. It will take two XML files and
display a browser type canonical XML difference report.  In my naiveté, I
see this same technique working for comparison of the committees XSLT test
case results.

Lynda VanVleet
lvanvleet@classiq.com

-----Original Message-----
From: David_Marston@lotus.com [mailto:David_Marston@lotus.com]
Sent: Tuesday, September 26, 2000 2:07 PM
To: XSLT-Conformance@lists.oasis-open.org
Subject: Canonical XML


This may be worthy of an agenda item. Actually, the larger question of how
a test lab detects that "correct" output occurred is probably the agenda
item.

The W3C is moving ahead with the Canonical XML standard. If no major
stumbling blocks arise, it will probably reach Recommendation status by the
time we issue our first report and test suite. If they settle on a
minimalist approach to emiting namespace declarations, then everything we
need will probably be covered. (There will be some items that they will
leave as unable to canonicalize, but I think there are none that we'll
depend upon.)
.................David Marston