Hi!

I have Win10, I can see and type local (slovenian) characters in mail, subject and all other locations. So I assume I have correct settings for codepage and everything.

But when I receive mail, I can not see name of file, if it has local characters.

Any idea?

BR, Vito

This may depend on how the attachments are encoded by the sender. If you open the message in raw view there should be lines like this immediately preceding each attachment's (presumably also encoded) contents (only separarted by a single blank line from the contents):

Content-Type: image/jpeg; name="image001.jpg"

and/or:

Content-Disposition: inline; filename="image001.jpg"

These ones are obviouly for an image file, but the point here is: If the filenames of your attachments contain non-ASCII characters (which the above ones don't) they need to be converted and indicate what character set is being used for decoding them back to the originally, so what do the only of the message you quoted look like? The character set's name would be shown within the encoded line next to the filename like this:

Content-disposition: attachment; filename="=?UTF-8?Q?=C3=A4=C3=B6=C3=BC.txt?="

In this case the character set is UTF-8 which is the universal one including almost all languages on earth and should be chosen by whatever sender sends attachments and emails using non-ASCII characters.

PS: =C3=A4=C3=B6=C3=BC.txt should be decoded to äöü.txt doing a UTF-8 to West European character decoding ...

Thank you. RAW looks like this:

Funny, but in file info I can see correct characters :-)

ISO/IEC 8859-2 should be correct settings for our country (Slovenia).

Any idea what should I do?

Yep, that's correct.

How long are the Content-type: and the Content-disposition lines respectiveliy in total? I'm afraid they both exceed the maximum allowed header line length of 78 characters (excluding two for the line breaks), and if so there's nothing you can do about it. For checking out whether this is the reason you might try the following after moving or copying your test message back to Pegasus Mail's New Mail Folder: In Pegasus Mail's main menu open Help > About Pegasus Mail > Info Button for getting the New mailbox location, then open the respective Windows directory via Windows File Explorer. Next right click the test message entry in the Pegasus Mail's New Mail folder and go down to its Message properties menu item at the bottom. After opening the respective dialog there'll be an entry in the top section showing the message's filename.

After creating a backup copy of this file you can now open the message in this file in a plain text editor (no Office application!!!) such as the former Windows Notepad where you'll have the raw view available which you can edit now: So go down to the header section you've copied above and insert a line break just behind the Content-type: Application/Octet-stream; part leaving the blank behind the semicolon as the first character of the following line (so Pegasus Mail knows it's the continuation of the previous line). The same woul have to be applied to the Content-disposition line. For your convenience I'm providing the proper lines here to simply copy and paste (hope there's no typo since I had to retype it):

Content-type: Application/Octet-stream;
  name="=?ISO-8859-2?Q?=B9=E8=E6=F0=BE.txt?="; type=text
Content-disposition: Attachment;
  filename="=?ISO-8859-2?Q?=B9=E8=E6=F0=BE.txt?="

This way the line lenght shouldn't prevent the filenames from being truncated and being read and decoded properly. IOW: Either sender's email application or server script needs to be fixed according to Internet standards defined by RFC 5322 (ok, it's a should only, not a must anymore).

Thank you. But this did not help. Not it claims that type is unknown.

On the other hand, wrong characters are shown immediately when I add attachment to mail. No matter if I drag/drop or go with 'Add attachment':

After 'Open' I see this:

This is not sent yet.

BR, Vito

What are your settings on Tools > Options > Advanced settings > Default MIME character set? I would recommend UTF-8. But this would only affect your own messages, not incoming ones, if at all. The above is only for diagnostic pruposes anyway, it's not a solution of any kind, just trying to figure out what's going wrong.

It was 8859-2. I have tried also UTF-8, but the result is the same. In raw I can see the change, but in att. it is always the same

Can you please forward a test message with such an attachment to <beta-reports [at] pmail.gen.nz>? Please use this thread's title as the email's subject so I can tell it apart from all the spam beeing sent to this account.

TIA

Done. Thank you.

Thanks, here's the first view I get from preview's attachment pane:

What we have here are UTF-8 multibyte characters displayed as (most likely) Windows-1252 single byte characters instead of conversion to Unicode. Here's what 7-Zip displays on my machine: þƒåðº.txt, and that's what I see in Windows Explorer as well. Then I have this in the File information section:

   ---- File information -----------
     File:  šèæðž.txt
     ...
     ...
     Type:  Text

If using an online decoder for only decoding the quoted printable encoded filename "=C5=A1=C4=8D=C4=87=C4=91=C5=BE.txt" I'm getting this: "ščćđž.txt". Now which one is the right one after decoding?

Great! You have the third option, which I never had :-)

What I get is:

It should be

Ok, now I have something to check out and pass on to David Harris for figuring out a solution. It looks like there's more than just a display issue to be solved ...

Great! Thank you!

Best Regards, Vito

You're welcome.