Jürgen Schönwälder <j.schoenwael...@jacobs-university.de> wrote:
> On Tue, Mar 08, 2022 at 11:20:57AM +0100, Martin Björklund wrote:
> > Hi,
> >
> > You didn't answer my first question about what we actually mean - do
> > we mean the "URI"
>
> "The uri type represents a Uniform Resource Identifier
> (URI) as defined by STD 66. [...]"
>
> STD 66 resolves to RFC 3986. Maybe I do not understand your question.
RFC 3986 defines the syntax for "URI" and "URI-reference", see
specifically section 4 and 4.1 of RFC 3986. To avoid confusion, I
suggest:
"The uri type represents a Uniform Resource Identifier
(URI) as defined by the rule 'URI' in RFC 3986.
or "the ABNF rule 'URI'..."
(side note, is it better to refer to STD 66 rather than the RFC?)
/martin
>
> > I agree with you re the problem with pattern in this case.
> >
> > The simplest pattern would be:
> >
> > pattern "[a-z][a-z0-9+.-]*:.*"; // matches the mandatory scheme
> >
> > but I think you will say that either we have a complete correct
> > pattern, or no pattern at all ;-)
>
> I personally do not believe that a pattern needs to catch every
> nuance. The only requirement is that a pattern must not reject any
> valid values. For me, the description ultimately defines the type and
> a pattern merely helps to catch some of the syntactic details in an
> automated fashion.
>
> My experience is that people have different ideas about how many
> details a pattern should capture and hence pattern have a tendency to
> grow increasingly complex over time.
>
> I am happy to add the proposed pattern since covering the scheme part
> is apparently better than having nothing (but I am sure this will be
> the beginning of a journey and not its end).
>
> /js
>
> --
> Jürgen Schönwälder Jacobs University Bremen gGmbH
> Phone: +49 421 200 3587 Campus Ring 1 | 28759 Bremen | Germany
> Fax: +49 421 200 3103 <https://www.jacobs-university.de/>
_______________________________________________
netmod mailing list
netmod@ietf.org
https://www.ietf.org/mailman/listinfo/netmod
