charset error when sending an e-mail
Categories
(Thunderbird :: Untriaged, defect)
Tracking
(Not tracked)
People
(Reporter: geckozone, Unassigned)
Details
Steps to reproduce:
Send a simple e-mail containing the message "déjà"
Actual results:
Message sent has this source (via CTRL+U) (personal info replaced) :
Message-ID: <some_id>
Date: Sun, 28 May 2023 17:33:05 +0200
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
Thunderbird/102.11.0
Content-Language: fr
To: SOMEONE <someone@example.org>
From: ME <me@example.org>
Subject: =?UTF-8?B?ZMOpasOg?=
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
déjÃ
Expected results:
Two solutions :
- Message should be "utf-8" compliant (here it has been converted via "latin-1")
- Message should contain the correct charset "latin-1" (or "ISO-8859-1")
For information, on the geckozone forum someone told me to add
"intl.charset.detector" with value "universal_charset_detector"
line did not exist, I had to create it.
And then :
"mailnews.force_charset_override" with value "false"
I still have the same problem.
Comment 2•2 years ago
|
||
Eh, I think that's just the view source editor wanting to display it as latin1?
Nope, I use a python lib to read e-mails through IMAP (Imbox) and content is latin-1 encoded and headers say :
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Comment 4•2 years ago
|
||
(In reply to geckozone from comment #0)
Message sent has this source (via CTRL+U) (personal info replaced) :
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bitdéjÃ
Yes, this is the RAW UTF-8 code represented with Windows-1252(? I say).
I don't recall TB ever doing it any other way.
And I seem to remember that there is already a bug for that. And I further think that its result was WONTFIX.
But I can not find it at the moment.
Expected results:
Two solutions :
- Message should be "utf-8" compliant (here it has been converted via "latin-1")
- Message should contain the correct charset "latin-1" (or "ISO-8859-1")
I don't understand what you are trying to say.
You can get the Unicode representation via Menu: View -> Repair Text Encoding
Comment 5•2 years ago
|
||
(In reply to Alfred Peters from comment #4)
And I seem to remember that there is already a bug for that.
Found it.
And I further think that its result was WONTFIX.
I was wrong about that.
But I can not find it at the moment.
Comment 6•2 years ago
|
||
(In reply to Alfred Peters from comment #5)
And I further think that its result was WONTFIX.
I was wrong about that.
I must have remembered Bug 1448373. ;-)
Maybe I did not tell my problem clearly, it is not a View source" problem.
If I just send the email and try to read it via IMAP, its content is latin-1 (or ISO-8859-1) encoded while header says it's utf-8 encoded.
So, when I use an external lib to receive the e-mail, it uses utf-8 to decode resulting in wrong content.
Therefore, header should say "latin-1" or content should be utf-8 encoded.
Comment 8•2 years ago
|
||
Thunderbird never sends anything other than UTF-8.
Ok, maybe a python problem so. Apparently e-mail is utf-8 encoded and the "View source" bug has pushed me on TB. Thx
Description
•