[Users] Corrupted subjects

[Slavko]

Dňa 26. septembra 2022 19:31:30 UTC používateľ mahlamytsike at gmail.com napísal:

>Message body content tags are following:
>Subject: O_ teavitus (copied directly from Ctrl-U message source)

Nobody will help, until you provide raw value, eg with hexdump, but
my guess is broken encoding. Anyway, header with non US-ASCII
chars have to be MIME encoded, as the SMTPUTF8 extension is not
widelly supported yet.

>Content-Type: text/plain; charset=iso-8859-1; format=flowed
>
>When I choose Default, the question mark is show everywhere. When I choost ISO-8859-1, the message is rendered correctly.

The message cleanly defines ISO-8859-1 encoding. While it is surprising, 
that it fails in UTF-8 case, as the ISO-8859-1 is UTF-8 subset, that seems
as message which contain some broken chars (bytes). I will not be surprised,
when it will contain multibyte chars breaked in between (eg. with end of
line chars). That means, that charsed value in Content-Type header doesn't
reflects real state. But without raw content, i can only guess.

>Message tags are following:
>Subject: =?utf-8?B?0KHQtdC60YDQtdGC0Ysg0L/Rg9GC0LXRiNC10YHRgtCy0LXQ?=
> =?utf-8?B?vdC90LjQutCwLiDQntGC0LLQtdGH0LDQtdC8INC90LAg0LLQvtC/0YDQ?=
> =?utf-8?B?vtGB0Ysg0Log0LDQstGC0L7RgtGD0YDQuNGB0YLQsNC8?=

That encoded Subject header is broken, it breaks UTF-8 chars in between
multibyte values in particular chunks. No one can expect right result on
decode, as every chunk have to be valid UTF-8 (in this case) string. Their
MUA/MSA is bad, not able to work properly with multibyte charsets.

Joining separate chunks together manully provides result in Cyrilic.

>Content-Type: text/html; charset=utf-8
>Content-Transfer-Encoding: base64

That is about body, not affects headers...

regards



-- 
Slavko
https://www.slavino.sk/