using iconv()

Evgeny Kotsuba evgen at shatura.laser.ru
Sat Jan 8 11:07:49 CET 2005


Hi,
David Relson wrote:

>Evgeny,
>
>I'm working on code to use iconv() rather than bogofilter's charsets and
>maps.  It's still experimental and I don't expect any time tomorrow to
>work on it.  If you're interested, I may be able to send you a copy Sunday.
>
I am not enthusiastic about iconv  and not enthusiastic about to store 
data in unicode.  If you decide to use it then please make it with 
defining macro and #ifdef / #else .
There is no sence in unicode if we don't can detect propely codepage and 
we can't now do it in all cases.

For example, if we have 8-bit characters in heades like:
--------------------------
From: "ф═╔╘╥╔╩" <radiodetali at fromru.com>
Subject: Ё╥╧─┴═
Date: Wed, 15 Oct 2003 15:59:21 +0400
MIME-Version: 1.0
Content-Type: text/html;
        charset="koi8-r"
X-Priority: 3
---------------------------
So when we find charset we need go back  to From: and Subject.

Heh... I see with my Mozilla mail and you probably see what will be with 
iconv... this message is in utf8 and I see a number of question marks in 
From: and Subject:  text  while it is
20 | 22 E4 CD C9 | D4 D2 C9 CA | 22
and
F0 D2 CF C4 | C1 CD

SY,
EK






More information about the bogofilter-dev mailing list