Problems with ODBC MSAccess and UTF8

[ThorkilG12]

Hi Please look at the attached pictures. All software is running on the Windows2012r2 server. Database2.zip Database2.zip

My guess is that something is wrong inside DBeaver. It's så strange that only row-id 15 (Lodz) but spelled with Polish letters is wrong.. Can anyone perhaps explain ? Thanks in advance.

[serge-rider]

I see that city name in 14th row is in ASCII. 15th row is in unicode. Why do you think it is something inside DBeaver?

[ThorkilG12]

Hey, and thanks for the swift answer :-)

All other two-byte special characters is ok. Even the € which is a three byte unicode. I use this table over and over to ensure that I have controle over the european countries, and ONLY DBeaver has problems with this when it comes to MS-Access...

PostGreSQL has no problems with this. Try to see the two files I have attached.

In my sample-data there is a lot of unicode... Why only the polish ones went wrong ? And here you can see how well it works in DBeaver/PostgreSQL

[image: image.png]

So I will still argue that DBeaver/MSAccess has "issues"

Venlig hilsen

Thorkil Johansen Tel: +45 4070 6547

Den søn. 17. feb. 2019 kl. 18.11 skrev Serge Rider <notifications@github.com

:

I see that city name in 14th row is in ASCII. 15th row is in unicode. Why do you think it is something inside DBeaver?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/dbeaver/dbeaver/issues/5299#issuecomment-464482674, or mute the thread https://github.com/notifications/unsubscribe-auth/ALjw1Zgyp20hyGu47qHlLMI24qGMjDkKks5vOY1UgaJpZM4a_o61 .

$allery €urocity Aberdeen cities Lodz Łódź Preußisch Oldendorf Überlingen Zagreb Ælligelyng Østerby Aalborg Ålestrup

[ThorkilG12]

The names of the cities has a lot of UTF-8 characters in them. But ONLY when I use DBeaver and the JDBC/ODBC connection I got this problem. As you can see, the Powershell uses the exact same ODBC driver and PS has no problems. Therefore I say (or guess) it is inside DBeaver that the solution has to be found.

[ThorkilG12]

Hi. Do we agree that there is an issue here ?

[serge-rider]

I'm not sure how this happened but even this is a bug then most likely it is originated in UCanAccess driver. If you are on windows then you could try ODBC MS Access driver to validate this: http://ucanaccess.sourceforge.net/site.html

[ThorkilG12]

As I documented in the first post, I can read through the Access driver without problems from PowerShell and other software. There is only one ODBC driver in use on my server. Since most of the characters (outside basic ANSI) if from codepage "Western" and since the two Polish characters is outside "Western" I'll suggest that you look into DBeaver to find the spot where you miss to use UTF correctly. Again: All kind of software on my server can do this correct, Only DBeaver fails.

[serge-rider]

This is what I see (using UCanAccess 4.0.2 driver):

So I can't even reproduce this issue. The problem is in your driver configuration (what exactly driver do you use?) or in your driver version (e.g. old UCanAccess or some old ODBC bridge driver).

[ThorkilG12]

I'm using 64bit 'Microsoft Access Driver (.mdb, .accdb)' Version 14.00.7180.5000 From Microsoft Cooperation

[ThorkilG12]

[serge-rider]

Could you try to UCanAccess driver (not ODBC) in DBeaver?

[uslss]

there is no update on ticket for a long time. Ticket closed