Re: [babel] Éric Vyncke's Discuss on draft-ietf-babel-yang-model-10: (with DISCUSS and COMMENT)

[Mahesh Jethanandani <[e-mail]>]

Hi Martin,

> On May 18, 2021, at 11:40 PM, Martin Björklund <[e-mail]> wrote:
> 
> Hi,
> 
>>> 
>>> The YANG module does not compile correctly with PYANG, it should be
>>> easy to fix
>>> though :-) See:
>>> https://yangcatalog.org/results/ietf-babel@2021-05-12_ietf.html
>>> Or is it a PYANG error ?
>> 
>> This is a tooling error (even though I have gone ahead and changed it
>> in the model). The ABNF grammar says the following for identity-stmt
>> in RFC 7950.
>> 
>>   identity-stmt       = identity-keyword sep identifier-arg-str optsep
>>                         (";" /
>>                          "{" stmtsep
>>                              ;; these stmts can appear in any order
>>                              *if-feature-stmt
>>                              *base-stmt
>>                              [status-stmt]
>>                              [description-stmt]
>>                              [reference-stmt]
>>                          "}") stmtsep
>> 
>> According to it, the ‘if-feature’ and ‘base-stmt’ can appear in any
>> order. But for some reason pyang version 2.4.0 insists on a particular
>> order. I will bring this up with tooling folks.
> 
> This is not a tooling error.  The error message is:
> 
>  [e-mail] <mailto:[e-mail]>:148: error: keyword "if-feature" not in
>    canonical order (see RFC 6020, Section 12)
> 
> [side note: the RFC ref in this error message should be updated to RFC
> 7950 Section 14]
> 
> The referenced section says:
> 
>   The ABNF grammar [RFC5234] [RFC7405] defines the canonical order.
> 
> When we publish YANG models in RFCs we use the canonical order in
> order to have a consistent format.  pyang checks this when the flag
> --ietf is used.

I am trying to understand, and this is more for my education, what rule within the grammar is determining the canonical order?

Also, should pyang report this as an error, or maybe a warning might be better? After all, this is being done for consistency reasons.

> 
> 
> /martin
> 
> 
> 
>> 
>>> 
>>> 
>>> 
>> ----------------------------------------------------------------------
>>> COMMENT:
>>> 
>> ----------------------------------------------------------------------
>>> 
>>> == COMMENTS ==
>>> 
>>> The related links on should be updated. E.g., the YANG catalog entry
>>> should be:
>>> https://www.yangcatalog.org/yang-search/module_details.php?module=ietf-babel@2021-05-12
>>> 
>>> -- Section 2.2 --
>>> I usually use the expanded tree view rather the YANG module itself to
>>> get a
>>> global view. Is there any reason why the full tree view is not
>>> included?
>> 
>> Added this to the Appendix.
>> 
>>> 
>>> -- Section 5 --
>>> Is there any reason why the doc shepherd is not acknowledged ?
>>> 
>>> == NITS ==
>>> 
>>> -- Section 2.3 --
>>> s/MAC based security/MAC-based security/ ?
>> 
>> Fixed.
>> 
>> Thanks.
>> 
>> Mahesh Jethanandani
>> [e-mail]

Mahesh Jethanandani
[e-mail]