[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]
Re: [xml-dev] Feasibility of "do all application coding in the XMLlanguages"?
- From: Rick Jelliffe <rjelliffe@allette.com.au>
- To: "xml-dev@lists.xml.org" <xml-dev@lists.xml.org>
- Date: Thu, 04 Dec 2008 19:59:51 +1100
Readers who are following this thread and are interested in some
guidance about "what kinds of XSD structures are commonly found in
data-binding tools" should look at the W3C document
Basic XML Schema Patterns for Databinding Version 1.0
http://www.w3.org/TR/xmlschema-patterns/
The gist is in Appendix F:
Supported XML Schema elements, attributes and simple types
Now of course, most people will in fact be asking the dual question:
"what kinds of XSD structures are unwise or unsafe?" For that, if your
structure uses an element marked N/A or does imatch one of the patterns
linked to, then there is the kind of structure that data-binding tools
may not implement (well, minimally or at all).
That does not mean you should necessarily avoid them, of course. In most
cases you could derive a simpler schema and just use that. For example,
if your schema was (a, b|c, d[2-32]) you would have to remodel it
(a, b?, c?, b?, d*)
The allowed paths are modeled using XPaths, so I guess it would be
possible to convert this document into a Schematron schema with each
XPath (with the initial "." removed) being a rule context with <assert
test="true()"/> so swallow the element and default rule at the end like
<rule context="*">
<report test="true()">Declaration found which did not match an
expected Databinding pattern</report>
</rule>
Validation would then give a list of elements that didn't match those
patterns. If these included any significant parts declarations, you
might have to prepare some workaround, I guess.
Cheers
Rick Jelliffe
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]