On 11/17/2017 09:53 PM, Jeremy Harris wrote: > On 17/11/17 19:06, Ian Zimmerman wrote:
>> On 15/11/17 14:51, Gryzli Bugbear via Exim-users wrote:
>>>> I'm trying to find what is the encoding that exim uses, when logging
>>>> Topic(T=)/Subject to exim mainlog.
>>>>
>>>> Currently cyrillic subjects are logged to something like this :
>>>>
>>>> \321\201\321\200\320\276\320\272\320\276\320\262\320\265
>> Doesn't a message with such subject violate the spec? It ought to be
>> rfc-2047-encoded, right? Or does exim decode it and then log the
>> decoded version? That would be a mistake, IMHO; it should just log the
>> rfc-2047 form.
> However, it's what it does. What I described was being applied
> to the $h_subject contents, so it's had mime-decoding applied
> by that point.
