[Evolution] Strange UTF8 problem in 1.4.3

[Erik Bågfors <erik bagfors nu>]

Hello!

I'm having some strange UTF8-problems in evo 1.4.3 that didn't exist in 1.4.0 and earlier.

Most mails I receive are in iso-8859-1 but sometimes I do get UTF-8 characters in evolution. Only sometimes.

If I look at the mail locally on the mailserver the mail looks correct.  
---------
Content-Transfer-Encoding: quoted-printable
....
....
Content-Type: text/plain; charset=iso-8859-1
....
....
Tak for hj=E6lpen.
---------

If I look at it in evo over imap:
------
Content-Transfer-Encoding: quoted-printable
....
....
Content-Type: text/plain; charset=iso-8859-1
....
....
Tak for hj=C3=A6lpen.
------
Somehow that latin1-character is now in utf-8.  If I go to "character encoding"->"UTF-8" the mail displays correctly.

I looks like this only happens over imap.  I moved a mbox file with this mail by scp and imported it into evolution and then it looks ok.

I also tried to snoop the network with ethereal and 
--------
A00184 SELECT INBOX
* FLAGS (\Answered \Flagged \Deleted \Seen \Recent)
* OK [PERMANENTFLAGS (\Answered \Flagged \Deleted \Seen)] Limited
* 2186 EXISTS
* 1 RECENT
* OK [UIDVALIDITY 997209422] Ok
A00184 OK [READ-WRITE] Ok
A00185 FETCH 2185 UID
* 2185 FETCH (UID 20412)
A00185 OK FETCH completed.
A00186 UID FETCH 20413:* (FLAGS RFC822.SIZE INTERNALDATE BODY.PEEK[HEADER.FIELDS.NOT (RECEIVED)])
* 2186 FETCH (UID 20413 FLAGS (\Recent) RFC822.SIZE 8287 INTERNALDATE "25-Jul-2003 11:18:14 +0200" BODY[HEADER.FIELDS.NOT ("RECEIVED")] {1375}
.....
.....
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
....
....


A00186 OK FETCH completed.
A00187 UID FETCH 20413 BODY
* 2186 FETCH (UID 20413 BODY ("text" "plain" ("charset" "iso-8859-1") NIL NIL "quoted-printable" 5331 157))
A00187 OK FETCH completed.
A00188 UID FETCH 20413 BODY.PEEK[HEADER]
* 2186 FETCH (UID 20413 BODY[HEADER] {2956}
....
....
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
....
....
Tak for hj=E6lpen.
--------------

So that looks correct!  How strange is that? :)

Has anyone else seen this?  I haven't found a bug about it but I might be blind! :)

/Erik

-- 
erik bagfors nu
fingerprint: 6666 A85B 95D3 D26B 296B 6C60 4F32 2C0B 693D 6E32