I have a question about why setting flexible="true" on a <structure> element
is limited to situations in which ordered="false".

First, I don't understand why this restriction is necessary. Why is it any
harder to ignore unknown elements in the ordered case than in the unordered
case? Can't you just ignore them?

Secondly, this makes JiBX useless on some very common and important use
cases: XML schemas that allow arbitrary nested elements from other
namespaces.

For example, any schema that has one or more of these in it:

  <xsd:any namespace="##other" processContents="lax" minOccurs="0"
maxOccurs="unbounded"/>

If you want to write a parser that accepts any valid document adhering to
such a schema, you need to be able to have your parser automatically ignore
the nested stuff from other namespaces that you don't recognize. This is a
common pattern in XML documents to allow independent parties to
independently define extensions that are automatically (a)
backward-compatible and (b) don't conflict. But there doesn't seem to be a
way to accomplish this with JiBX when the namespace="##other" element(s)
appear among an ordered set of elements.

What am I missing?

Thanks,
-Archie

-- 
Archie L. Cobbs
------------------------------------------------------------------------------
This SF.net email is sponsored by 

Make an app they can't live without
Enter the BlackBerry Developer Challenge
http://p.sf.net/sfu/RIM-dev2dev 
_______________________________________________
jibx-users mailing list
jibx-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jibx-users

Reply via email to