Re: [Cbor] cddl 0.8.17: Add .abnf (draft-ietf-cbor-cddl-control)

[Paul Kyzivat <[e-mail]>]

On 2/25/21 5:58 PM, Carsten Bormann wrote:

> The only surprise for me (beyond the fact that this stuff is so easy that I needed less than six hours to implement this PoC, after all) was how thoroughly I had forgotten about the leading whitespace limitation in RFC 5234 (before Julian Reschke reminded me in a completely unrelated thread a couple of days ago).  Of course, the bap tool has some adaptive handling here that makes this “feature” invisible to spec writers.
...
> We could add text that removes the common leading whitespace before processing the ABNF proper; this still requires the catted pieces to agree on the leading whitespace (which is bad if you are catting together pieces from different sources).  We could simply stipulate, say, two spaces.

Perhaps off topic here, but ...

IMO it would be good to make a revision to 5234 to resolve the leading 
whitespace issue. As it stands, there must be multitude of hacks to work 
around it, and AFAIK the assumptions made within those hacks are poorly 
documented and probably inconsistent.

The first thing to resolve would be to decide how to resolve the 
ambiguity when leading whitespace is allowed. Unfortunately I don't 
think there is any backward compatible extension that could itself be 
defined in abnf.

	Thanks,
	Paul