Closed Bug 1176662 Opened 9 years ago Closed 3 years ago

Mojibake in subject line

Categories

(Thunderbird :: Message Reader UI, defect)

38 Branch
defect
Not set
normal

Tracking

(Not tracked)

RESOLVED INVALID

People

(Reporter: shini_ma, Unassigned)

Details

(Keywords: regression)

Attachments

(4 files)

Attached image screencapture.png
User Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:38.0) Gecko/20100101 Firefox/38.0
Build ID: 20150525141253

Steps to reproduce:

Nothing special, just browse the list of emails


Actual results:

The characters in the subject line of certain emails are wrongly decoded, and hence they are not readable. In the previous version of thunderbird, I haven't experienced this problem.


Expected results:

In this new version, the detection of the text encode is inappropriately modified.
Please provide a sample: save as .eml

(If the charset is GB2312, dupe of bug 1174580)
Keywords: regression
Attached file Mojibake email.eml
This is eml file of Mojibake email.
Looks to me that the Subject header isn't encoded, but contains the raw value

Subject: $B6HL3O"Mm(B
I don't see a valid bug here but just an invalid email. :)
Status: UNCONFIRMED → RESOLVED
Closed: 9 years ago
Resolution: --- → INVALID
Attached image screencapture.gif
This is the problem of character transformation.
There is no problem in *.eml file itself.
I guess the decoding the eml file is bad 
and display unreadable information.
Here is evidence. The other Japanese mailer correctly decode the character in the subject line.
Thanks for your help.
Status: RESOLVED → UNCONFIRMED
Resolution: INVALID → ---
I still don't see a valid bug here but just an invalid email "encoding" (at least the attached example mail is invalid, but not sure if that was already scrambled when transferring via mail servers or such).

> There is no problem in *.eml file itself.

There is, at least in the one attached here in this bug report.
The correct decoded subject line by another Japanese mailer shown in attached example is the result of the import of the eml file. So the eml file itself is not wrong. This mean this is not the matter of transfer via mail servers. As you can see the time stamp, the email itself is not new. This is the old email I received before. I could see this email correctly by the older version of Thunderbird. Thanks!
The attached mail doesn't show any differently in tb31 for me. (But maybe there's some pref/fallback charset I'd have to set)
(In reply to shini_ma from comment #7)
> The correct decoded subject line by another Japanese mailer shown in
> attached example is the result of the import of the eml file. So the eml
> file itself is not wrong. This mean this is not the matter of transfer via
> mail servers. As you can see the time stamp, the email itself is not new.
> This is the old email I received before. I could see this email correctly by
> the older version of Thunderbird. Thanks!

The eml file is obtained by the current version Thunderbird.
The subject line of the obtained eml file can be viewed by the another mailer as well as the previous Thunderbird, but not by the current version Thunderbird.
As you can see in screencapture.gif, the subject line is different between two mailers while the other information (sender, and date) are same.
Is there any difference in the .eml if you export it from 31 or 38?
(In reply to Magnus Melin from comment #10)
> Is there any difference in the .eml if you export it from 31 or 38?

Sorry, I could not catch what you mean.
What do you mean "export from 31"?

By the way, I found two things.
1. Even if I changed the text encoding setting, the subject line in the summary display was not affected and the characters in the mail body display were affected. 
2. The subject line in the mail body display is correctly decoded as shown in the attached image file.

I greatly appreciate your help.
I mean in
 A) thunderbird 31: save as .eml
 B) thunderbird 38: save as .eml

Is there a difference in the saved file for case A vs case B?
(In reply to Magnus Melin from comment #12)

I got it. 31 and 38 mean the version of Thunderbird. 
There is no difference.

I uninstall the current Thunderbird v38 and install the older thunderbird v31.
Then the Mojibake email is correctly shown including subject line which is wrongly decoded by v38.
I saved the email as eml file, and compare it with the one saved with thunderbird v38.
There is no difference between them.
As I wrote in previous comment posted on 2015-06-23 04:10:05 PDT, 
it can correctly show the characters in the mail body display even when I used v38.
So, it could not be the problem of the transfer from the mail server.

What to do with this now?

Flags: needinfo?(mkmelin+mozilla)

I think we can close this. The subject should be encoded to be valid.

Status: UNCONFIRMED → RESOLVED
Closed: 9 years ago3 years ago
Flags: needinfo?(mkmelin+mozilla)
Resolution: --- → INVALID
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: