Re: [netmod] XML and prefix

tom petch <ietfc@btconnect.com> wrote:
> From: Martin Björklund <mbj+ietf@4668.se>
> Sent: 14 January 2022 11:23
> 
> Hi,
> 
> Ok, I think I understand what he means.  With this XML:
> 
>   <alarm-category xmlns:nsfmi="urn:ietf:params:xml:ns:yang\
>                               :ietf-i2nsf-nsf-monitoring">
>       nsfmi:memory-alarm
>   </alarm-category>
> 
> the prefix "nsfmi" is present in the element data, which means that
> in order to implement this properly, the XML parser must pass the
> namespace mappings to the user code.
> 
> So he proposes to add to the draft:
> (from https://mailarchive.ietf.org/arch/msg/i2nsf/06kJ7vS6X-0hUmGHrCWN-jVzU7M)
> 
> 11.  XML Examples for I2NSF NSF Monitoring
> 
>    This section shows the XML examples of I2NSF NSF Monitoring data
>    delivered via Monitoring Interface from an NSF.  In order for the XML
>    to work, the prefix in the element that uses "identityref" type
>    (e.g., /i2nsf-event/i2nsf-system-detection-alarm/alarm-category/) in
>    the YANG module described in this document MUST be the same as the
>    namespace prefix for urn:ietf:params:xml:ns:yang:ietf-i2nsf-nsf-
>    monitoring.  The XML software MUST be chosen that makes the namespace
>    prefix information available.
> 
> 
> I think this is a bit odd.  Who is supposed to act on the first MUST?
> This text is about an example, which is what it is, and it happens to
> be correct.
> 
> Also, the text about XML software seems unnecessary to me.  It follows
> from the definition of an identityref in RFC 7950 that the namespace
> mapping is needed to parse this correctly.
> 
> 
> <tp2>
> Martin,
> 
> Again, thank you.  My initial take was that defining the prefix in alarm-category makes that the default so that the prefix is not needed on 'memory-alarm' but my XML is getting a bit rusty!

Almost ;-)  RFC 7950 says:

   If the prefix is not
   present, the namespace of the identityref is the default namespace
   in effect on the element that contains the identityref value.

So in this particular case, the example could instead be written as:

  <alarm-category>memory-alarm</alarm-category>

b/c the default namespace in effect is

  urn:ietf:params:xml:ns:yang:ietf-i2nsf-nsf-monitoring

/martin

> 
> Tom Petch
> /martin
> 
> 
> 
> 
> 
> Ladislav Lhotka <ladislav.lhotka@nic.cz> wrote:
> > On 14. 01. 22 11:39, Martin Björklund wrote:
> > > Hi,
> > > I don't understand the problem either.  He writes:
> > >
> > >> Sorry, but this has the same problem in figure 11.1 that we've just
> > >> been
> > >> discussing with Ian.
> > > Can you send a pointer to that discussion?  Perhaps there's more
> > > context there.
> >
> > Right. I also suspect that the last sentence should have been
> >
> > "I don't think it's OK for the draft to say those things."
> >
> > Lada
> >
> > > /martin
> > > tom petch <ietfc@btconnect.com> wrote:
> > >> I see that IANA have taken to asking XML Registry experts about the
> > >> registration of YANG namespaces at Last Call, or perhaps they have
> > >> always done this but have only recently put the e-mail on a public
> > >> list.  Anyhow, the experts have taken it upon themselves to comment on
> > >> the XML examples and I do not understand this comment.  This comes
> > >> from
> > >> [IANA #1217705] Expert Review for
> > >> draft-ietf-i2nsf-nsf-monitoring-data-model-12 (xml-registry)
> > >> by Tim Bray 17 dec 2021 03:03
> > >>
> > >> ===============================
> > >> Sorry, but this has the same problem in figure 11.1 that we've just
> > >> been
> > >> discussing with Ian.
> > >>
> > >> For it to work, (a) the prefix in the alarm-category element MUST be
> > >> the
> > >> same as the namespace prefix for
> > >> urn:ietf:params:xml:ns:yang:ietf-i2nsf-nsf-monitoring, which means
> > >> that XML
> > >> software MUST be chosen that makes the namespace prefix information
> > >> available.  I don't think it's OK for the draft not to say those
> > >> thigns.
> > >>
> > >> <alarm-category
> > >>             xmlns:nsfmi="urn:ietf:params:xml:ns:yang:\
> > >>                        ietf-i2nsf-nsf-monitoring">
> > >>             nsfmi:memory-alarm
> > >>           </alarm-category>
> > >> =================================================
> > >> a) I am unclear what the problem is - I thought that XML allowed great
> > >> freedom with prefix even if the IETF would rather not
> > >> b) this suggestion seems to be that all I-D with XML examples, which
> > >> is pretty much every I-D with a YANG module in it, needs to carry a
> > >> warning about what XML software to choose, which seems rather a
> > >> burden.  Thoughts?
> > >>
> > >> Tom Petch
> > >>
> > >>
> > >>
> > >>
> > >> _______________________________________________
> > >> netmod mailing list
> > >> netmod@ietf.org
> > >> https://www.ietf.org/mailman/listinfo/netmod
> > > _______________________________________________
> > > netmod mailing list
> > > netmod@ietf.org
> > > https://www.ietf.org/mailman/listinfo/netmod
> >
> > --
> > Ladislav Lhotka
> > Head, CZ.NIC Labs
> > PGP Key ID: 0xB8F92B08A9F76C67