Re: [xacml] Namespaces and xpaths

[Erik Rissanen <mirty@sics.se>]

Good. I have added this as issue 78 in the issues list with you Daniel
as the champion.

Regards,
Erik

Daniel Engovatov wrote:
> It seems that our XPath binding is underspecified - we need to define
> the in scope namespaces resolution (things that XQuery does in its
> prologue) and define the content node (".")
> Minimal list that we need to define is in appendix D of the XPath 2.0
> document: http://www.w3.org/TR/xpath20/#id-impl-defined-items, Section
> F.1 and F.2 in XDM document:
> http://www.w3.org/TR/xpath-datamodel/#implementation-defined and
> Appendix F in F&O spec: http://www.w3.org/TR/xquery-operators/#impl-def
>
> That is a fairly long list that may require its own mini-spec.   I have
> filed a work item on this, so I guess I would own the list.
>
> Daniel;
>
>
> -----Original Message-----
> From: Erik Rissanen [mailto:mirty@sics.se] 
> Sent: Thursday, March 29, 2007 5:31 AM
> To: xacml@lists.oasis-open.org
> Subject: [xacml] Namespaces and xpaths
>
> All,
>
> I have a question about namespaces and xpath expressions. There are a
> few standard xpath functions in XACML 2.0 in section A.3.15. These
> functions take strings as parameters, where the strings are to be
> interpreted as xpath expressions.
>
> How is resolving of namespace prefixes in these strings/expressions
> done? The specification doesn't say anything about this. (Or I didn't
> find it.)
>
> Is the xpath function going to trace back its input to the xml document
> from where the string came from? This doesn't seem possible in general.
> Or does it find the <Apply> element from where the invocation came, and
> use only those declarations? But this is confusing since the string may
> originate from somewhere else.
>
> Maybe we need an xpath expression data type to embed the namespace
> associations with the expression? I think such a type would be valuable
> also if users create new xpath related functions.
>
> If we define such a data type, we could also define a cast from a
> string, in which case no namespaces will be defined.
>
> Similarly, there is no text about namespace resolving in the text about
> <AttributeSelector>, but here it seems obvious that the xpath matching
> must use the namespaces in effect at the <AttributeSelector> element.
> But maybe some text would be good anyway?
>
> Regards,
> Erik
>
>
> Notice:  This email message, together with any attachments, may contain information  of  BEA Systems,  Inc.,  its subsidiaries  and  affiliated entities,  that may be confidential,  proprietary,  copyrighted  and/or legally privileged, and is intended solely for the use of the individual or entity named in this message. If you are not the intended recipient, and have received this message in error, please immediately return this by email and then delete it.
>