[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: RelaxNG question

• To: XML Dev <xml-dev@l...>
• Subject: Re: RelaxNG question
• From: rjelliffe <ricko@a...>
• Date: Thu, 21 Aug 2003 02:05:22 +1000
• In-reply-to: <830178CE7378FC40BC6F1DDADCFDD1D171A993@R...>
• References: <830178CE7378FC40BC6F1DDADCFDD1D171A993@R...>
• User-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.3.1) Gecko/20030428

Dare Obasanjo wrote:

>Because your favorite schema language does not support the feature does not make it a daft constraint. The ability to negate wildcards is an oft requested feature of users of W3C XML Schema and makes sense if one could write something akin to 
>
>	<xs:any namespace="not(##targetNamespace)" /> 
>
>although in this case I assume Tim would want something akin to  
>
>
>	<xs:any namespace="not(##targetNamespace ##local)" /> 
>
This kind of test is very easy in Schematron. Something like:--

<rule context="eg:container/*">
    <ssert test="namespace()!=' ' ">Any children of an eg:container must 
be in a namespace</assert>
</rule>

or
 
<rule context="eg:container//*">
    <assert test="namespace()!=' ' ">Any descendants of an eg:container 
must be in a namespace</assert>
</rule>

Cheers
Rick Jelliffe

(Apologies to anyone who has tried to contact me for the last week, while I
moved over to Linux after the virus.)

• References:
  • RE: RelaxNG question
    • From: "Dare Obasanjo" <dareo@m...>

• Prev by Date: Re: Postel's Law Has No Exceptions
• Next by Date: RE: Are we ready for the namespace ID registry, yet? (w as: RelaxNG question)
• Previous by thread: Re: RelaxNG question
• Next by thread: RE: RelaxNG question
• Index(es):
  • Date
  • Thread