Re: [apps-discuss] Review of: draft-yevstifeyev-abnf-separated-lists-02

[Julian Reschke <[e-mail]>]

On 11.12.2010 17:10, Dave CROCKER wrote:
> ...
> At this point I am wondering whether it would not be cleaner to have the
> notation be:
>
> <n>*<m>[^(<a>)]element
>
> In other words, augment the existing repetition construct with a clean
> separator construct. The up-arrow only specifies a separator, it does
> not also specify repetition.
>
> Hence:
>
> 1*10^":"tag-val
>
> would be 1 to 10 tag-vals separated by colons, and
>
> *^" "words
>
> would be zero or more words separated by a space.
> ...

Looks better to me.

I believe it would be useful to have a look at existing specs that use 
list productions, and see how they could use this format.

That might discover missing pieces, or just things that are done 
differently across specs, which may make it harder to find a common syntax.

For instance, in HTTP we have:

"Wherever this construct is used, null elements are allowed, but do not 
contribute to the count of elements present. That is, "(element), , 
(element) " is permitted, but counts as only two elements. Therefore, 
where at least one element is required, at least one non-null element 
MUST be present. Default values are 0 and infinity so that "#element" 
allows any number, including zero; "1#element" requires at least one; 
and "1#2element" allows one or two." -- 
<http://greenbytes.de/tech/webdav/rfc2616.html#rfc.section.2.1>

I don't think the proposal addresses this use case, so it appears it 
would be unusable for expressing the HTTP ABNF.

Best regards, Julian