home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.mail.mh
- Path: sparky!uunet!spool.mu.edu!sol.ctr.columbia.edu!sirius.ctr.columbia.edu!seth
- From: seth@sirius.ctr.columbia.edu (Seth Robertson)
- Subject: Re: Problem with MH address parser
- References: <1992Dec16.071847.3949@sol.ctr.columbia.edu>
- Sender: nobody@ctr.columbia.edu
- Organization: Columbia University Center for Telecommunications Research
- Date: Wed, 16 Dec 1992 09:54:22 GMT
- Message-ID: <1992Dec16.095422.6331@sol.ctr.columbia.edu>
- X-Posted-From: sirius.ctr.columbia.edu
- NNTP-Posting-Host: sol.ctr.columbia.edu
- Lines: 26
-
- The short form is that it is not a valid address. The long form,
- which follows, was sent to me by Owen Rees.
-
- Thanks!
- -Seth
-
- ======================================================================
- From: Owen Rees <rtor@ansa.co.uk>
-
- Seth Robertson <seth@startide.ctr.columbia.edu> writes:
- > repl: bad addresses:
- > N. R. "Norm" Lunde <norm> -- no at-sign after local-part (Lunde)
- > testformail
- >
- > As far as I can tell, this is a valid address per 822
-
- It isn't a valid RFC822 address -
- see 3.3 Lexical Tokens - N. is the 'word' N followed by 'special' .
- see 6.1 [Address] Syntax - the dots make N.R."Norm" a 'local-part'
- which must be followed by an at sign. The error message is
- telling you what if found instead of an at.
- see 3.1.4 Structured Field Bodies - linear white space is permitted
- between lexical tokens, it (including parenthesized comments)
- just gets stripped out when converting to canonical form.
-
- RFC1123 modifies the rules slightly but not so as to affect the above.
-