On Thu, Sep 11, 2003 at 10:38:31AM +0200, Andreas Metzler wrote:
> On Thu, Sep 11, 2003 at 11:26:15AM +0000, Michael V. Sokolov wrote:
> > %Subj% with exim?
>
> You could use ${rfc2047:<string>} and headers_add/remove. But even if
> you did it would not help you, because there is no reliable way[1] to
> identify the used charset and ${rfc2047 only supports ISO-8859-1 so
> you would change e.g. "This ??? is a Euro-symbol" to
> "=?ISO-8859-1?Q?This_=A4_is_a_Euro-symbol?=" which is not an
> Euro-symbol at all but "currency sign".
>
> I know that is a lame example but if you took
> russian/chinese/UTF-8/whatever input and declared it was ISO-8859-1
> you'd really make things worse.
> cu andreas
> [1] statistical analysis with databases of all natural langauages to
> check against might work ;-)
Looks like you may also try to use charset from Content-Type to
determine this. It's far from a clean solution, but you may get rather
good results, especially if you won't change anything but russian
encodings :)
Tell me in case of success, or send report to russian exim-users
mailing list :)
Regards,
--
Igor A. Karpov phone: +380(44)238-0624 JID:jc@???
Unix System Administrator
E = mc^2 +/- 3d6
