W3C home > Mailing lists > Public > xmlschema-dev@w3.org > August 2004

Re: XML Schema Question

From: Jeni Tennison <jeni@jenitennison.com>
Date: Mon, 23 Aug 2004 17:15:10 +0100
Message-ID: <1002855531.20040823171510@jenitennison.com>
To: "Michael Kay" <mhk@mhk.me.uk>
CC: "'Asleson, Ryan'" <asleson@BIWORLDWIDE.com>, xmlschema-dev@w3.org

Hi Mike,

>> This is the most frequent of FAQs. It can't be done in XML Schema.
> Actually, I think someone did discover a way of doing it. You can
> use a choice as in (2) below, and then use a uniqueness constraint
> to ensure there is only one father and only one mother. Under the
> element declaration for family, you add:
> <xs:unique name="one-father">
>   <xs:selector xpath="."/>
>   <xs:field xpath="father"/>
> </xs:unique>
> This will cause a validation error if a the xs:field expression
> selects more than one node, that is, if a family has more than one
> father.

Yes, Eric van der Vlist came up with this in his XML Schema book. I
was one of his technical editors, and I remember he was worried about
whether it was legal. I thought that it was at the time, but (after
the book came out) found something in the spec that made me question
that. I think it was Clause 2.2 of Schema Component Constraint:
Selector Value OK, which states:

  [The {selector}] must be an XPath expression involving the child
  axis whose abbreviated form is as given above.

I think this means that the child axis must be used in the expression.
In the above case, the selector "." does not involve the child axis,
so I don't think it's legal.

But it's extremely fuzzy, and I'm not sure whether the wording
actually reflects the intention.



Jeni Tennison
Received on Monday, 23 August 2004 16:15:26 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 23:15:23 UTC