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

Re: Envelope address in procmail?



On 23 February 2000, Gjermund Sørseth <gjermund@nextra.com> wrote:
> 
>    Liviu Daia writes:
> 
>    >
>    >   PS: The "From:" line in your message is invalid.
> 
> 
> I don't think so. Can you explain in what way it is invalid?
> 
> The From header in question was:
> 
>    From: "=?ISO-8859-1?Q?Gjermund S=8F8rseth?=" <gjermund@nextra.com>

    The character "8F" is not printable in the charset iso-8859-1 your
header advertises.  As pointed below, I suppose what you mean is

	From: "=?ISO-8859-1?Q?Gjermund S=F8rseth?=" <gjermund@nextra.com>

(without the first "8"), which decodes to

	From "Gjermund Sørseth" <gjermund@nextra.com>

On 23 February 2000, Geir Johannessen <joge@stud.ntnu.no> wrote:
> Besides the topic of procmail, but anyway, according to an RFC (don't
> remember which one just now) it is not legal to encode headers with
> mime.
[...]

    RFC 822 forbids sending messages with non-ASCII characters in
headers.  That's why those characters have to be encoded as either
Base64 or Quoted-Printable, as documented in RFC 2047.

On 23 February 2000, Holger Wahlen <H.Wahlen@gmx.de> wrote:
> RFC 2047 definitely allows (and defines) it. If Gjermund had
> "S=F8rseth" (no "8" between "=" and "F") in his header, that would
> be okay and would have to be read as "S", character F8 hex (in
> ISO-8859-1) = o with slash, "r", "s", "e", "t", "h", but this way it's
> "S", character 8F hex, "8", "r" etc.; if I remember correctly, 8F is
> undefined in ISO-8859-1.

    That's my point.

On 23 February 2000, Stan Ryckman <stanr@sunspot.tiac.net> wrote:
[...]
> > But I wouldn't worry too much as all widespread mailers (MUAs) seem
> > to understand that the From: line is mime encoded and decodes it
> > successfully.
> 
> elm 2.5 PL2 (or earlier) does not:
>   O  8   Feb 23 * =?ISO-8859-1?Q?Gje (64)   Re: Envelope address in procmail?
> Neither does my favorite mail pager ("less").  :-)
> 
> Last time I checked, elm was pretty "widespread".

    Oh, Elm can decode those headers just fine, provided they are
syntactically valid, and you compiled it with the right options.  I
can't be bothered to check right now, but IIRC version 2.5 can even
re-encode characters from one charset to another, allowing you to see
f.i. =?ISO-8859-2?Q?=BA= (latin-2 "s cedilla") as "s" if your charset
has encoding iso-8859-1.

    Regards,

    Liviu Daia

-- 
Dr. Liviu Daia               e-mail:   Liviu.Daia@imar.ro
Institute of Mathematics     web page: http://www.imar.ro/~daia
of the Romanian Academy      PGP key:  http://www.imar.ro/~daia/daia.asc