The interesting thing with W3C's schema language is that since they've taken
a more OOP approach, there are many different ways you can write schemas to
validate a specific document, it's all up to your style. What I read from
the original poster was a schema that would validate the following XML:
<ParentNode>
<A/>
<B/>
<A/>
<A/>
<B/>
<A/>
<A/>
<B/>
<B/>
<!-- and many more -->
</ParentNode>
The schema I whipped up would validate that instance. My schema is
definately not all inclusive though, as you described. I tend to avoid
xsd:choice nodes, as is my style, which was why my first instinct was to
have a sequence with optional children. I could have easily done a schema
with two optional sequences that had one of each node. Or choice too. It's
all up to the schema author to decide which practices he/she wishes to
follow.
Bryce K. Nielsen
SysOnyx, Inc. (www.sysonyx.com)
Makers of xmlArchitect, the Smart XML Schema Editor
(http://www.sysonyx.com/products/xmlarchitect)
