I am not asking for a change, but ...

IMHO it would be better if YANG would not allow whitespace in enum names.
It is a rarely needed freedom that degrades usability and can lead to mistakes
and higher tool development costs. So it is bad for both people and tools.

  • It is misleading, as a human user can mistakenly think that "this is legal" is actually 3 separate values
  • Some tools tend to consider spaces as separators
  • When creating code from YANG this needs special handling

regards Balazs

On 2019. 02. 22. 12:07, Juergen Schoenwaelder wrote:
On Fri, Feb 22, 2019 at 10:55:31AM +0000, Balázs Lengyel wrote:
Do people really use enum names with whitespace within?
I always had the feeling that the name of the enum is like an identifier, so
it should follow maybe not the exact same but at least similar rules.

When such an enum name is translated into a programing language AFAIK the
translation needs to replace the spaces with something.
Not just spaces. Tool implementors will have to address this and what
needs special handling varies from target language to target language
(some target languages or implementations of target languages also
have interesting length contraints).

I think it is fine if BCPs like RFC 8407 provide guidelines that
people should avoid characters that are likely to be problematic but
the YANG specification is relatively clear that YANG itself puts
little constraints on such names (and hence tool makers in charge
to handle this).

/js

-- 
Balazs Lengyel                       Ericsson Hungary Ltd.
Senior Specialist
Mobile: +36-70-330-7909              email: [email protected] 

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature

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

Reply via email to