Message-ID: <ElDoheS00WBwQvhEU4@andrew.cmu.edu>
Date: Thu,  7 Mar 1996 15:47:38 -0500 (EST)
From: John Gardiner Myers <jgm+@CMU.EDU>
To: drums@cs.utk.edu
Subject: Re: Message format document outline
In-Reply-To: <MailManager.826226747.10347.mrc@Ikkoku-Kan.Panda.COM>
References: <MailManager.826226747.10347.mrc@Ikkoku-Kan.Panda.COM>

Mark Crispin <MRC@Panda.COM> writes:
> > These are all illegal per RFC 822 and should continue to be illegal.
> 
> Where does 822 say this?  I'm pretty sure that I remember an emphatic
> statement otherwise.

822 has no rule or sequence of rules to permit it.  3.1.2 states, in
part:

        Once a field has been unfolded, it may be viewed as being com-
        posed of a field-name followed by a colon (":"), followed by a
        field-body, and  terminated  by  a  carriage-return/line-feed.

So the field-body starts after the colon.  It also states:

               Field-names, unstructured field bodies  and  structured
               field bodies each are scanned by their own, independent
               "lexical" analyzers.

The section that permits free insertion of linear-white-space is
3.1.4, which only defines the lexer for strucured field bodies.  The
free insertion rule cannot be applied to the unstructured field body
or field-name lexical analyzer.

>   Certainly that was the case in the working group; that's
> why the BNF has things like "From" ":" and not "From:".

"From" and ":" correspond to different nonterminals in the rule for
field:

     field       =  field-name ":" [ field-body ] CRLF

They are different tokens, so should appear in the grammar as such.
In other documents you have such things as:

   unsubscribe_obs ::= "UNSUBSCRIBE" SPACE "MAILBOX" SPACE mailbox

instead of:

   unsubscribe_obs ::= "UNSUBSCRIBE MAILBOX " mailbox

-- 
_.John G. Myers		Internet: jgm+@CMU.EDU
			LoseNet:  ...!seismo!ihnp4!wiscvm.wisc.edu!give!up