OpenSAML user discussion

["Clifford Johnson" <>]

Ok, I can buy that.  So is AttributeValue going to be extended to have
an accessor method to something that reflects whether or not is it
"nil"?  AttributeValue's schema definition includes "nillable=true" (I
believe the only element in SAML that does) and
org.opensaml.saml2.core.AttributeValue has no provision for reflecting
that.

Clifford

-----Original Message-----
From: Scott Cantor 
[mailto:]
 
Sent: Thursday, March 13, 2008 11:00 AM
To: 

Subject: RE: Detecting "xsi:nil" in AttributeValue

> (Based on my reading of XML Schema Part 1, the xsi:nil attribute is
defined
> in *every* schema with global scope suggesting that an "isNil()"
method
> should be in org.opensaml.xml.XMLObject along with getSchemaType()
[for
> xsi:type], getSchemaLocation() [for xsi:schemaLocation], and
> getNoSchemaLocation() [for xsi:noSchemaLocation].)

No, that wouldn't make sense. Not every type is nillable (in fact very
few
are). The attribute is global but it's legality isn't, same as with
xml:lang.

What makes sense is for nillable types to carry it as a flag, unless
they
extend other classes that would give you access to the arbitrary
attributes.

-- Scott