W3C home > Mailing lists > Public > xmlschema-dev@w3.org > April 2002

RE: [Gml30.rwg] RE: [RESEND] Derivation by restriction

From: <Simon.Cox@csiro.au>
Date: Wed, 3 Apr 2002 08:13:56 +0800
Message-ID: <116D27C8E12BD411B3AB00B0D022B0B8010E5505@yate.wa.csiro.au>
To: bobbitt@posc.org, Simon.Cox@csiro.au
Cc: xmlschema-dev@w3.org, gml30.rwg@opengis.org
John Bobbitt wrote:

Why not use a choice? 

<xsd:group name="geometryContainer"> 
  <xsd:choice> 
    <xsd:element ref="PolygonContainer"/> 
    <xsd:element ref="CirlceContainer"/> 
  </xsd:choice> 
</xsd:group>  

Yes, this is possible.  It requires that we specify in advance all the
members of the choice group.  

This might work for geometry, if we believe that we do really know all the
types we are ever going to use.  
I guess that modelling geometry.xsd on 19107 is our assurance of this ;-)

But in the more general case, we use substitution groups (which are actually
implicit choice groups) rather than explicit choice groups *precisely* to
take advantage of extensibility.  It is possible add members to a
substitution group at any stage - in application schemas, in GML4, whatever.


Don't forget that we are dealing with XML Schema. There are ways to do
things other than by restrictions and extensions and substitutionGroups. 
Received on Tuesday, 2 April 2002 19:23:37 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 11 January 2011 00:14:30 GMT