Inappropriate converting of umlauts in subject

[sofl88]

Hello.

We got a strange behaviour reported when someone is trying to copy the subject of a conversation/email (Schadensmeldung FÜ230414439), paste it somewhere else inside the browser (atleast in latest chrome) and try to delete (backspace) the character Ü, which gets not deleted but instead, turns into an U.

The header information seems to be correct

Subject: =?UTF-8?Q?Schadensmeldung_F=C3=9C230414439?=

The website https://unicodelookup.com/ reveals that the Ü is some sort of a combined character (?)

combined Ü

manual typed Ü

This is a problem for us since we sometimes use the subject (or parts of it) to compare or search things in our database where things usually correct encoded/converted.

Thanks in advance.

[freescout-helpdesk]

Are you using modern Microsoft OAuth IMAP authentication or APP_NEW_FETCHING_LIBRARY env variable?

[sofl88]

No to both.

[freescout-helpdesk]

Try to add the following to .env file and clear cache:

APP_NEW_FETCHING_LIBRARY=true

[sofl88]

I tried but the issue is still there

[sofl88]

Reproducible in the freescout demo if that helps

[sofl88]

It seems the falsy conversion takes place here

https://github.com/freescout-helpdesk/freescout/blob/2fe0f518be3bf29a3677257b19fa482ea7d17d8c/overrides/webklex/laravel-imap/src/IMAP/Message.php#L330

The issue is gone when I change this line to

$this->subject = iconv_mime_decode($header->subject, ICONV_MIME_DECODE_CONTINUE_ON_ERROR, 'UTF-8');

But I don't know if this is a suitable solution in general.

[freescout-helpdesk]

Now iconv_mime_decode() function will be used for UTF-8 encoded subjects.