Re: [xml-dev] The awesome power of Schematron + XPath 2.0 ...Able to ex

[Rick Jelliffe <rjelliffe@allette.com.au>]

On Wed, 2007-10-24 at 19:49 -0400, Costello, Roger L. wrote:

> It was a very enlightening experience.  Schematron+XPath 2.0 was able
> to implement all of my data requirements (including all grammar
> constraints). Conversely, XML Schemas was only able to implement the
> grammar constraints (which are actually of lesser importance to me than
> my other data requirements).  

> Thanks Rick!

My pleasure.

We intend updating the ISO Schematron spec to include a standard binding
for XSLT2. One of the issues was whether to allow user-defined functions
or not. The solution seems to be to allow @queryBinding='xpath2' which
has no extensions and @queryBinding='xslt2' which allows xslt elements
for defining functions. This is analagous to the way that we have
@queryBinding='xpath' and @queryBinding='xslt' where the latter allows
<xsl:key> elements.  Ideas and comments are very welcome. 

I think there are two extremes for language development: the
mathematicists want the most power in the most terseness, while the
engineerists want the most organize-ability and integrate-ability. RELAX
NG is the former, XSD is the latter. I would put Schematron in the
latter actually: it may look simple but actually it pays a lot of
attention to organization: phases, abstract patterns, abstract rules,
the split between diagnostic text and assertion text, flags, roles,
etc. 

Its just that these are different abstractions than the XSD ones (type,
derivation, grammars, facets, substitution.) The XSD abstractions do
reify the various things that parameter entities did in DTDs, however
the question needs to be asked "Is what the parameter entities did
actually particularly useful, or just low-hanging fruit?"  (A
hypostatization problem resulting from bottom up design?) 

Cheers
Rick Jelliffe