RE: Question about XML Schema validation

Hi Eric,

Yes, the validity of a node comes from the following channels:

1) defined schema related to the node (which also implies the type MIP)
2) required
3) constraint MIP

(assuming the node is relevant, of course).

See Section 4.3.5, the xforms-revalidate event.

Cheers,
John Boyer

-----Original Message-----
From: www-forms-request@w3.org [mailto:www-forms-request@w3.org]On
Behalf Of Erik Bruchez
Sent: Wednesday, June 08, 2005 3:09 AM
To: www-forms@w3.org
Subject: Re: Question about XML Schema validation



John Boyer wrote:
 > Hi Erik,
 >
 > The erratum is erroneous, and the working group is actively working
 > on it.
 >
 > The issue has been thoroughly discussed on our last telecon, and its
 > resolution is scheduled for our face to face meeting next week.
 >
 > I expect there will be some kind of change in what E29 says.

Thanks.

Another question I have regards validity and how it relates to the
"required" model item property, if at all.

It seems that "required" mainly impacts submission (according to
6.1.3: "indicates that a non-empty instance data node is required
before a submission of instance data can occur"). Plus, a control
bound to required node may have some visual feedback.

So far, so good, except that the section on submission is a little
fuzzy: "Any selected instance data found to be *invalid* stops submit
processing...". Definition of "invalid instance data" is not
clear. Does this include checking for "required"? 6.1.3 appears to say
so, but then it's not validity, it's being required! However, this
suggests that in somebody's mind, "valid" is tied to "required".

An example:

If you say that an "instance data node" (BTW "instance data node"
doesn't have a definition in the spec as far as I can tell) is bound
to, say, a type "xs:date", and if it is not required, does the node
become valid, or invalid? It seems to me that according to the spec
the properties on the node will be:

o valid: false()
o required: false() (default)

Is this the intended result? Should then the UI and controls be in
charge of marking the node as "invalid" for the user by combining
those two model item properties?

Intuitively, I would have imagined that "required" would influence
"valid": if a node would be otherwise invalid as per a bound type, but
is actually empty and not required, then it is no longer invalid.

Am I the only one confused?

-Erik

Received on Wednesday, 8 June 2005 16:16:05 UTC