W3C home > Mailing lists > Public > www-forms@w3.org > December 2004

RE: Question re bind constraint

From: John Boyer <JBoyer@PureEdge.com>
Date: Wed, 15 Dec 2004 09:03:08 -0800
Message-ID: <7874BFCCD289A645B5CE3935769F0B52D8842D@tigger.pureedge.com>
To: "Chris Picton" <xforms-list@tangent.co.za>, <www-forms@w3.org>

Hi Chris,

The bind constraint appearing in your response does not
quite match the one I sent you.  Nice.  Clearly, having
current() on the brain caused me to have a lapse of 
current on the brain. :-)

Since XPath's nodeset to nodeset comparison already does a 
simple sequential search, you don't need current and can
go forward without any extension to XForms 1.0.

FYI the use case we want to solve with current() seems to
involve only slightly more complicated cases of table lookup
in which the found node is actually used.  The 1.1 working
draft (http://www.w3.org/TR/2004/WD-xforms11-20041115/) contains 
an example.


-----Original Message-----
From: Chris Picton [mailto:xforms-list@tangent.co.za]
Sent: Wednesday, December 15, 2004 1:49 AM
To: www-forms@w3.org
Subject: RE: Question re bind constraint

On Tue, 2004-12-14 at 12:57 -0500, Kugelman, John wrote:
> I believe you can do the following:
>     <model>
>       <instance id="data">
>         <field/>
>       </instance>
>       <instance id="numbers">
>         <numbers>
>           <number>1</number>
>           <number>10</number>
>           <number>23</number>
>         </numbers>
>       </instance>
>       <bind nodeset="instance('data')/field"
>             constraint="not(. = instance('numbers')/number)" />
>     </model>

Thanks.  That works and clarifies the spec for me.
Received on Wednesday, 15 December 2004 17:04:32 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:36:14 UTC