- From: Dave Raggett <dsr@w3.org>
- Date: Fri, 7 Apr 2000 09:58:39 +0100 (GMT Daylight Time)
- To: Curt Arnold <carnold@houston.rr.com>
- cc: www-forms@w3.org
On Fri, 7 Apr 2000, Curt Arnold wrote: > I think it would be very beneficial if the forms datatypes were > consistent and/or layered upon schema datatypes. Definitely the > requirements are different. In forms, the desire is to allow a > variety of potential representation and to allow ambiguity. In > schema, the desire is to have an efficient, unambiguous > representation. For example, nobody would want a user interface > that made you type in 2000-04-07 in a date field. The > underlying value space is the same, so all that XML Schema has > done with value-space constraints is still applicable. All that > is really necessary is to define the translation from "form > lexical" to "schema lexical" representation and then let the > underlying schema type validation systems to take over. The XForms datatypes are layered upon XML schema dataypes and the intention (mentioned in the abstract of the draft XForms data model) is to provide an XSLT filter to map XForms data models into XML Schemas that describe the XML encoding of form data for transfer to the server. Note that a validator for XML Schemas will be weaker that one based on XForms, since XML Schemas as currently specified lacks the means to describe the constraints between data items that XForms represents with its expression language. > For example, this form translation layer might convert "June" in > form lexical representation into 2000-06-01 in schema lexical > representation by using the current year to eliminate the > ambiguity and by implying the first of the month. The XForm's representation of dates is that used by XML Schema, i.e. ISO 8601. However, the user won't be expected to enter dates in this format. The model-view-controller paradigm allows for a mapping between user friendly (and locale dependent) date formats and ISO 8601. We are working on another draft that covers the user interface which will make this clearer. > A lot of features you have described in your data hierarchy, > open enumerations, conjunction types, etc, I have lobbied for in > XML Schema datatypes. I hope that some more will make it in the > next public draft of XML Datatypes. Thanks for the feedback. Regards, -- Dave Raggett <dsr@w3.org> http://www.w3.org/People/Raggett tel/fax: +44 122 578 3011 (or 2521) +44 385 320 444 (mobile) World Wide Web Consortium (on assignment from HP Labs)
Received on Friday, 7 April 2000 04:58:47 UTC