On Jan 4, 2008 2:48 AM, R J Scheuerle Jr <[EMAIL PROTECTED]> wrote:

> Rich Scheuerle
> IBM Web Services
> Apache Axis2 ([EMAIL PROTECTED])
> 512-838-5115 (IBM TL 678-5115)
>
> Glen Daniels <[EMAIL PROTECTED]> wrote on 12/28/2007 12:49:21 AM:
>
> > Hey Amila:
> >
> > >     won't need them.  But for SOAP 1.1, the situation is different.
>  The
> > >     encoding spec says you MUST encode all complex object types
> astop-level
>
> > >     members of the serialization.  Therefore ALL conforming SOAP 1.1
> > >     encoding implementations will be putting out stuff that looks like
> this:
> > >
> > >     <soap:body>
> > >       <operation>
> > >         <arg1 href="#obj1"/>
> > >       </operation>
> > >       <multiref id="obj1">
> > >         <name>the real argument</name>
> > >         <color>blue</color>
> > >       </multiref>
> > >     </soap:body>
> > >
> > > is this means it is allowed to have more than one xml element in the
> > > soap:body ?
> >
> > Yes.  But - I was incorrect about the MUST above!  My hands were typing
> > a little ahead of my brain there. :)  Multirefs are not in fact
> > *required* by SOAP 1.1 section 5, but some implementations (Axis 1.X
> > included) will by default serialize all complex objects as multirefs
> > since it speeds writing (if you don't do it that way you have to walk
> > the entire data graph to see what objects are referred to multiple times
>
> > before serializing).
> >
> > > I have to think about bit. We can resolve the problem with parsing as
> I
> > > have mentioned earlier but can not think about a way to serialize with
>
> > > multirefs.
> >
> > Again, my bad - we're not forced to do so, so it's ok for us not to, as
> > long as we accept that we're not going to be able to do real graphs.  If
>
> > no one wants this particular functionality, I'm fine with punting on it.
> >
> > Your idea about parsing the XML to remove the hrefs and generate a
> > larger "virtual" expanded document on the reading side should work fine,
>
> > although it will potentially cause repeated data if a multiref is used
> > many times.  Probably not a big deal.  I'll respond to the rest of your
> > other note in another message.
>
>
> This will only work for graphs which are strictly trees.
> It will not work for graphs which contain loops.  (Even worse, it will
> result in
> an infinite loop).
>
> So this highlights an additional concern.  Does ADB, in document/literal
> mode, contain code that
> detects loops and gracefully throws an exception ?
>
No. It can handle loops in schema level. but for object structure there can
not be any loops. Any way with ADB multiref serialization is not also
possible.

Amila.

> I know from prior support work that MS does
> trace loops.
>
>
>
>
> >
> > Thanks,
> > --Glen
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: [EMAIL PROTECTED]
> > For additional commands, e-mail: [EMAIL PROTECTED]
> >
>



-- 
Amila Suriarachchi,
WSO2 Inc.

Reply via email to