On Wed, Nov 25, 2015 at 5:15 AM, Martin Bjorklund <[email protected]> wrote:

> Juergen Schoenwaelder <[email protected]> wrote:
> > On Wed, Nov 25, 2015 at 01:31:01PM +0100, Martin Bjorklund wrote:
> > > Juergen Schoenwaelder <[email protected]> wrote:
> > > > On Wed, Nov 25, 2015 at 09:48:55AM +0100, Ladislav Lhotka wrote:
> > > > > Hi,
> > > > >
> > > > > I'd like to resolve this issue. Here is a complete example (valid,
> I believe):
> > > > >
> > > >
> > > > Sorry for asking a stupid question but what is the use case of
> > > > referencing state data from an xpath expression in RPC input or
> > > > output parameters?
> > >
> > > It all started with leafrefs in notifications.  The algorithm in RFC
> > > 6643 produces leafrefs from notifications to state data.  See Y04.
> > >
> > > For the same reason, you may want to reference let's say an interface
> > > name in the /interfaces-state/interface list from an RPC input
> > > parameter.
> > >
> >
> > Yes, but where does xpath come into play?
>
> The leafref's path is XPath.  See the example in Y04.  It might be
> surprising that this is illegal (if we do Y04-04):
>
>    rpc foo {
>      input {
>        leaf ifref {
>          type leafref {
>            path "/if:interfaces-state/if:interface/if:name";
>          }
>          must "/if:interfaces-state/if:interface[if:name =
> current()/ifref]"
>             + "/if:enabled";
>        }
>      }
>    }
>
> The leafref is legal, but the must expression would be illegal.
>
>
>
I agree with Juergen about questioning the need for this new feature.
The path expression is limited.  XPath is not limited at all.
The code to find the "pointed-at" leaf is much simpler than full XPath.



> /martin
>
>
Andy


>
>
> _______________________________________________
> netmod mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/netmod
>
_______________________________________________
netmod mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/netmod

Reply via email to