Jos, If your element is global, why do you need that additional document element? Why not make your element a document element?
Validation can also be done using data binding tools such as Castor. Alex Quoting Jos van den Oever <[EMAIL PROTECTED]>: > Hello Aleksander, > > Actually, what I'm doing is using elements in a DOM with a substitute > and meaningless Document and DocumentElement. I'm doing this because > WSDL specifies the datatypes on the level of element or Schema type. > One of the problems however is that the DocumentElement can be > validated too. If it's restriction says it needs an <a/> and a <b/> > element, then the document will not validate without both. For simple > situations, I could add a dummy element, but this is not possible with > more comples Schemas. > > You are right about the local definitions. I am talking about global > elements and global types. Very simple example: how can I validate an > attribute as a 'xs:float'? Beforehand I know not what Schema I will > get and what data. The data I get happens to be (or should be) type > 'xs:float' according to the dynamically loaded Schema. How do I > validate it? > > Cheers, Jos > > > On Fri, 19 Nov 2004 09:50:12 -0800, Aleksandar Milanovic > <[EMAIL PROTECTED]> wrote: > > I know of a work-around. If the element is global, you can create a new > copy > > of that element and its child elements by using importNode and then append > the > > new copy to a new document which you can validate. > > > > If the element is locally defined, then I think that validation is > meaningless > > because the local element has no meaning without its context. > > > > Take my words with caution though. I am not an XML Schema expert. > > > > Alex > > > > > > > > Quoting Jos van den Oever <[EMAIL PROTECTED]>: > > > > > Hello Ramin, > > > > > > I have to deal with XML Schema's that are defined elsewhere. > > > > > > Imagine a document with two required elements <a/> and <b/>. I don't > > > have <a/> but only <b/>. Now I want to now if <b/> is valid, even > > > though the complete document is not valid, because <a/> lacks. > > > > > > If XSModel wasn't readonly, I could make a ripped version of a given > > > Schema and use that for validation. It wouldn't be very > > > straightforward though. > > > > > > Cheers, Jos > > > > > > On Fri, 19 Nov 2004 12:03:16 -0500, [EMAIL PROTECTED] > > > <[EMAIL PROTECTED]> wrote: > > > > > > > > In your XML schema, just define the rule for that one element. > > > > > > > > Cheers, > > > > -ramin > > > > > > > > > > > > > > > > > > > > > > > > > > > > "Jos van den Oever" <[EMAIL PROTECTED]> > > > > > > > > 11/19/2004 11:59 AM > > > > > > > > Please respond to > > > > [EMAIL PROTECTED]; Please respond to > > > > "Jos van den Oever" <[EMAIL PROTECTED]> > > > > > > > > > > > > To [EMAIL PROTECTED] > > > > > > > > cc > > > > > > > > Subject validate a single element > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Hello all, > > > > > > > > Is it possible to validate only parts of an XML document with Xerces? > This > > > > > > > would be useful in Web Services were data typing is done by specifying > XML > > > > > > > Schema single types or elements. > > > > > > > > Cheers, Jos > > > > > > > > --------------------------------------------------------------------- > > > > To unsubscribe, e-mail: [EMAIL PROTECTED] > > > > For additional commands, e-mail: [EMAIL PROTECTED] > > > > > > > > > > > > > > > > > > > > > > --------------------------------------------------------------------- > > > > > > > To unsubscribe, e-mail: [EMAIL PROTECTED] > > > For additional commands, e-mail: [EMAIL PROTECTED] > > > > > > > > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
