Re: Struggling with LWS (2616) vs LWSP (2831bis)

Julian Reschke wrote:
 
> For now, I'm just trying to make progress on the ABNF conversion;
> *not* on changing the allowed syntax (for which I think we don't
> have consensus yet).

Fine, for 2831bis I arrived at 
<http://article.gmane.org/gmane.ietf.sasl/2758>
amended by
<http://article.gmane.org/gmane.ietf.sasl/2759>
using the definition  LWSP = *(WSP / CRLF WSP)

That's the same LWSP as in STD 68 with "apparently empty lines".

To fix this *later* you can simply replace LWSP by FWS (RFC 2822):

FWS     = ([*WSP CRLF] 1*WSP) / obs-FWS

*Later* put the "multi-CRLF" crap into <obs-FWS>, and say that it
MUST NOT be generated, compare RFC 2822 chapter 4.2.  The syntax
for <obs-FWS> could be improved, I didn't see that in the 2822upd
discussions:

obs-FWS = 1*WSP *(CRLF 1*WSP)

I think that should be obs-FWS = 1*WSP 2*(CRLF 1*WSP) or similar.

 Frank

Received on Tuesday, 27 May 2008 11:42:16 UTC