Character encoding question

I maintain an application that downloads messages from IMAP servers. I am
working
with a server now that is giving me a folder name encoded as

"Cabinet/test Ordner &AOQA,AD2-"

My customer is telling me this is supposed to be "test Ordner äüö"

There is another example where "t&APY-st" decodes to "töst". This one I
understand - I take the three character sequence between the & and -
and base64 decode it, and end up with a single character. All the examples
I have seen of this type of encoding use triplets for each character.
However
I have no idea how to decode the first example above. The first three
characters
do decode to " ä" as expected but everything I try to extract those last two
characters fails. I'm beginning to suspect that the IMAP server is not
correctly
encoding the folder name. Any ideas?

On Wed, 20 Aug 2008 22:24:16 -0700, Jeff McKay wrote:

> There is another example where "t&APY-st" decodes to "töst". This one I
> understand - I take the three character sequence between the & and - and
> base64 decode it, and end up with a single character. All the examples
> I have seen of this type of encoding use triplets for each character.
> However
> I have no idea how to decode the first example above.

IMAP specifications are a good place to start. Open RFC 3501 and look for
mailbox names / modified UTF-7.