NOVELL FORUMS

mikebell90 · #2 05-Nov-2009, 11:27 AM

I think you may be right. There is certainly no way to fix this short of
opening an SR.

In their defense I recall several weird corner cases not covered by the
RFC. But it's been a long time since I've seen em.

Anders Westerberg wrote:
> Hi.
>
> I'm not sure if its the client that decodes the message or not so if I'm
> in the wrong forum please correct me.
>
> I'm not a groupwise expert but I do know a lot about smtp/email and I'v
> analyzed how our groupwise system decodes MIME encoded headers and it
> does not seem to follow rfc2047, it ignores linear-white-space at the
> wrong places.
>
> As an example we have an incoming mail with a subject header like this:
>
> subject: =?ISO-8859-1?Q?=E4rendenr?= 12345
>
> This is a proper encoding of a subject header that decodes to "Ã„rende
> 12345", without quotation marks and the first char is a swedish a with
> two dots over it if u cant read it ;)!
>
> Groupwise decodes this as "Ã„rende12345", it ignores the whitespace
> between an encoded word and a non-encoded word which is wrong, according
> to rfc2047 whitespaces should only be ignored inside an encoded string
> or between two encoded strings. Strangely enough it seems that if it
> happens the other way around with a header like this:
>
> subject: 12345 =?ISO-8859-1?Q?=E4rendenr?=
>
> Groupwise decodes this as "12345 Ã„rendenr" which is correct. Of course
> you could encode the first example like this instead:
>
> subject: =?ISO-8859-1?Q?=E4rendenr_12345?=
>
> Groupwise would then decode this correctly as "Ã„rendenr 12345" but both
> ways to encode this is correct and Groupwise should decode both
> correctly. This also happens on other headers like a from: "firstname
> lastname" <emailaddress> and say just the firstname is
> encoded, Groupwise then decodes that as
> "firstnamelastname emailaddress". Is this something that is a general
> bug in Groupwise or is it possible to adjust this and something is wrong
> with our Groupwise installation?
>
> /anders
>