Re: [xml-dev] Is it valid in XSD to have choice where all particlesare o

[Michael Kay <mike@saxonica.com>]

On 15/12/2011 14:56, Jirka Kosek wrote:
> Hi,
>
> I have came across interesting problem in XSD. Assume the following
> snippet of XML schema:
>
> <xs:complexType name="foo">
>    <xs:sequence>
>      ...
>      <xs:choice>
>        <xs:element name="a" minOccurs="0"/>
>        <xs:element name="b" minOccurs="0"/>
>      </xs:choice>
>      ...
>    </xs:sequence>
> </xs:complexType>
>
> I have tried several validator and none of them reports error witch such
> schema. However in section "3.8.4 Model Group Validation Rules" in XML
> Schema Part 1 there is written:
>
> "Nothing in the above should be understood as ruling out groups whose
> {particles} is empty: *although no sequence can be ·valid· with respect
> to such a group whose {compositor} is choice*, the empty sequence is
> ·valid· with respect to empty groups whose {compositor} is sequence or all."
That's saying you can write

<xs:choice minOccurs="1"/>

meaning there's a menu of options to choose from, and you've got to 
choose something, but the list of options to choose from is empty; so 
nothing can be valid against the choice.

Or to put it another way, it's like having a condition nextElement = 
oneOf(a,b,c) where the list a,b,c is empty and the condition is 
therefore always false.
>
> Should this be understood as that above fragment of XSD is incorrect? In your fragment, the list of particles inside the choice is not empty.
No.

Michael Kay