[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Parsing the Injection-Info: header field


Hi Charles,


I understand that RFC 2045 now extends the syntax of RFC 5322 (instead
of RFC 822).


How can that possibly be so, since it was written umpteen years earlier?
Anything that "extended RFC 5322 would have to be written _after_ that.


A misunderstanding of mine.
I thought that if an article was to comply with RFC 5322, then the syntax
of extended fields (defined by MIME), was to be restricted to what RFC 5322
allows.  But that does not seem to be the case.

Thanks, Antti-Juhani and Charles, for having explained what should be
done for MIME.



RFC 2045, and its relations, provides syntax for many new header fields.
From reading RFC 2045 and RFC 822, you can deduce *exactly* where CFWS is
allowed. If RFC were to be re-written (a highly desirable thing BTW), then
if would redefine its syntax with explicite CFWS, presumably in exactly
the same places as now, unless it chose to declare some of them to be
obsolete as 5322 has done.

It is clear that, according to RFC 2045, CFWS is allowed on both sides of
the '=' in a <parameter> (and why shouldn't it?). And, just to make sure, I
checked my interpretation with Keith Moore before writing the original
version of the paragraph which you quoted.


Yeah, no problem.  Sorry for the disturbance of my thread.

For what it is worth, I also had an answer from the YAM group meanwhile:

 http://www.ietf.org/mail-archive/web/yam/current/msg00279.html

   [Ned Freed]
   "RFCs 2045 and 2046 were written before DRUMS and RFCs
   2821/2822/5321/5322 and therefore used the syntactic convensions of RFC 822,
   where LWSP is implicitly allowed between tokens. The fact that DRUMS opted to
   make the places where LWSP is allowed explicit rather than implicit doesn't
   change the syntax rules for 2045 and 2046 in any way. shape or form.

   I also note that had RFC 5322 had any effect on MIME syntax, the header of the
   document would show that it updates RFCs 2045 and 2046. The document header
   says no such thing.

   P.S. When revised MIME specifications come out they will probably switch to the
   explicit LWSP approach since it seems that's the preferred way to do it now."

--
Julien ÉLIE

« Cela n'a rien de remarquable. Il suffit d'appuyer
 sur la bonne touche au bon moment et l'instrument
joue tout seul. » (J.-S. Bach)