[yang-tooling] Re: SID file issues / .sid file differences in pyang and core-wg/pyang

[Andy Bierman <[e-mail]>]

On Sat, Feb 7, 2026 at 5:05 AM Carsten Bormann <[e-mail]> wrote:

> On 2026-02-07, at 11:37, Michael Richardson <[e-mail]> wrote:
> >
> >> I don't really call it YANG if something as basic as choice/case is
> >> ignored.
> >
> > I think we can agree that it's an omission that needs fixing.
>
> That was a deliberate choice, and what needs fixing apparently is mainly
> the “description” in the YANG module for .sid Files.
>
>
I agree.  Not hard to believe you understood the examples better than the
text in RFC 7950.

The primary use case is representing instance data.
The existing use cases for choice/case identifiers use string names, and
they can continue to do so.

The SID item can be used at the encode/decode layer to deal with individual
nodes.
If a 'structure' is used, no extra schema nodes (e.g., rpc) are allowed.
They just convert to/from the internal format, so only the nodes expected
to be present in the data are specified.
The YANG validation can be done by a higher layer (or not at all).

(Note that, while we are fixing the .sid File YANG module, that we may
> already be extending .sid Files to support SIDs specifically for metadata
> names [1], similar to how identities get their own namespace.)
>
> [1]:
> https://mailarchive.ietf.org/arch/msg/core/totiCGYFFkh9mAZqL3otUnn0_iw/
>
>
We are trying to let a client use XML, JSON, or CBOR for any NETCONF or
RESTCONF message.
YANG Push does not use metadata, but several RPC operations use it.


>> Maybe a new SID file format for constrained devices should be developed
> >> if the file does not actually identify schema nodes.  A SID file that
> >> identified only data nodes might be suitable for 'structure' usage.
> >
> > I haven't been down the CORECONF path (yet), but I would not include .sid
> > files literally... I would process them through some build-time tooling
> that
> > would generate code for my constrained device.
>
> (I don’t usually look at “.sid” files before having processed them to CSV
> using the sid-csv tool.)
>
> Grüße, Carsten
>


Andy


>
> _______________________________________________
> yang-tooling mailing list -- [e-mail]
> To unsubscribe send an email to [e-mail]
>