W3C home > Mailing lists > Public > xmlschema-dev@w3.org > April 2009

Re: [XML Schema 1.1] Using doc() in xs:assert ... the referenced document needs a schema?

From: C. M. Sperberg-McQueen <cmsmcq@blackmesatech.com>
Date: Mon, 27 Apr 2009 09:47:04 -0600
Cc: "C. M. Sperberg-McQueen" <cmsmcq@blackmesatech.com>, "'xmlschema-dev@w3.org'" <xmlschema-dev@w3.org>
Message-Id: <534735D7-C9DC-4387-AD00-A85868F6057B@blackmesatech.com>
To: "Costello, Roger L." <costello@mitre.org>
On 27 Apr 2009, at 08:18 , Costello, Roger L. wrote:

>
> Here's a simple schema that uses xs:assert to check that the value  
> of the <country> element matches one listed in an instance document  
> (countries.xml):
>
> ---------------------------------------------------------------
> <?xml version="1.0"?>
> <schema xmlns="http://www.w3.org/2001/XMLSchema"
>        elementFormDefault="qualified">
>
>    <element name="Example">
>        <complexType>
>            <sequence>
>                <element name="country">
>                    <simpleType>
>                        <restriction base="string">
>                            <assert test="$value =  
> doc('countries.xml')//country" />
>                        </restriction>
>                    </simpleType>
>                </element>
>            </sequence>
>        </complexType>
>    </element>
>
> </schema>
> ---------------------------------------------------------------
>
>
> While running this example, I discovered these requirements:
>
>   - the instance document referenced by xs:assert (countries.xml)
>     must have a schema, and
>
>   - I have to import that schema.
>
> Why?

Can you provide a little more information?  Where did you discover
these requirements?  They don't look at first glance as if they
were requirements imposed by the XSD 1.1 spec.

Unless I am mistaken about the rules of XPath expression evaluation,
the issue of whether documents retrieved using the doc() function
must have a schema or not does not in fact arise in XSD 1.1.  The
Last Call draft of XSD 1.1 says that the 'available documents' property
of the dynamic context for XPath evaluation on assertions
is the empty set.  That should mean, if I understand correctly,
that the subexpression doc('countries.xml') returns the empty
node sequence.

Making the 'available documents' property be the empty set helps
ensure (a) better interoperability between validators, and (b) the
context-independence of schema-validity assessment of an element
or attribute against a declaration or type definition.  The
context-independence of validation against a type is important to
many users of XSD (although not, I suspect, to all).

If you want to validate cross-document references in this way, you
should certainly look at SML, which is designed precisely to
allow validation across document boundaries; in SML, your
constraint would be a straightforward keyref constraint requiring
the value of the 'country' element to match one of the keys in
'country.xml'.

>
> Would someone please explain the rationale for these requirements?


I'm happy to explain the rationale for various aspects of the
design of XSD, as I understand the rationale.  But as I said
the requirements you identify don't seem to stem from XSD; if
you can say a bit more about where they come from, perhaps someone
with knowledge of the source can speak up.

-- 
****************************************************************
* C. M. Sperberg-McQueen, Black Mesa Technologies LLC
* http://www.blackmesatech.com
* http://cmsmcq.com/mib
* http://balisage.net
****************************************************************
Received on Monday, 27 April 2009 15:47:42 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 11 January 2011 00:15:11 GMT