- From: Fennell, Philip <philip.fennell@hp.com>
- Date: Wed, 7 Feb 2007 12:51:07 -0000
- To: <www-forms@w3.org>
Hello, I have a scenario where there is a source document (model) and a rendering (view). I generate an XForm that allows me to both edit the values of attributes in the source document and see values of attributes generated in the view. For example, I may define an object in the source but it has properties that are computed by the viewer. Notes: a) Source and view properties are mutually exclusive. b) View properties are regarded as read-only. In these types of situation I see two possible options: 1) Multiple instances. ---------------------- Create a 'model' xforms:instance and a 'view' xforms:instance in the same xforms:model. The xforms:bind elements reference the two instances using the instance(xs:IDREF) function. I believe I'm right in assuming that I would use the instance() function on the xforms:submission element's ref attribute to indicate which instance to use (unless the 'model' instance is the first within the xforms:model and therefore the default instance for submission). 2) relevant = "false()" ----------------------- Merge the view properties with those of the source element (see note 'a' above) in a single xforms:instance and then upon their respective bindings set the 'view' properties to relevant="false()". In this case, when the form is submitted, I believe the irrelevant nodes will be removed from the instance that is submitted. All bindings will point to the same xforms:instance. Please could I have some 'best practice' guidance as to which option would be considered more practical/preferable? Regards Philip Fennell
Received on Wednesday, 7 February 2007 22:39:20 UTC