[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: #1047 Path-identity: A proposal (perhaps)
Bruce Lilly wrote:
> That still permits foo!!tail-entry, foo!MISMATCH!tail-entry
Yes, and I wasn't consequent with the [FWS], next attempt:
path-list = [FWS] path-identity [FWS]
*( path-delimiter [FWS] path-identity [FWS] )
( "!" [ "POSTED!" ] ) [FWS] tail-entry [FWS]
path-delimiter = "!" [[ "MISMATCH" / "POSTED" ] "!" ]
That's a bit convoluted, maybe a simple list of delimiters is
better, see above:
( "!" / "!POSTED!" ) [FWS] tail-entry [FWS]
path-delimiter = "!" / "!!" / "!MISMATCH!" / "!POSTED!"
> doesn't address (no pun intended) where IP address literals
> are or are not allowed.
It was based on Charles' revised syntax for a <path-identity>:
| path-identity = ( ALPHA / DIGIT )
| *( ALPHA / DIGIT / "-" / "." / "_" ) /
| IPv4address / IPv6address ; see [RFC 3986]
> I made some quick notes this morning:
> path-identity "!POSTED" needs to be a group
Okay, path-identity "!POSTED!" tail-entry catches the normal
case, path-identity "!POSTED!" path-list is for reinjection.
> path-identity "!!" path-identity needs to be a group
Also okay, I have path-identity "!!" path-list, and no way to
combine it with a tail-entry.
> "MISMATCH!" path-identity needs to be a group
Dito as path-identity "!MISMATCH!" path-list, not with a tail.
> That text (and the draft ABNF) implies that
> Path: MISMATCH!example.net!tail-entry
> is a valid complete field, which might not be what is
> intended; the text doesn't say that there needs to be
> anything to the left of MISMATCH.
Then it was wrong, "!MISMATCH!" is the same idea as "!!", only
for a mismatch. In my hardcore nitpicking mode it should be:
path-identity "!" path-identity "!MISMATCH!" path-list
a!b!MISMATHCH!c!...!not-for-mail as in "I got this starting
with c!... but I think it should be b, anyway I am a, enjoy".
Or should this be a!!b!MISMATCH!c!...!not-for-mail (?) Bye.