Bug #31650 Lindfalas
Submitted: 16 Oct 2007 19:36 Modified: 30 May 2013 6:37
Reporter: Sören Forsdahl Email Updates:
Status: Closed Impact on me:
None 
Category:Connector / ODBC Severity:S2 (Serious)
Version:5.1 alpha OS:Windows (Vista)
Assigned to: CPU Architecture:Any
Tags: national characters

[16 Oct 2007 19:36] Sören Forsdahl
Description:
When I try this version I get som strange characters where it should be my national characters.
It works with 3.51 driver.

Its varchar fields and with utf8 characters set.
All fields in the databas is right.
So it must be the 5.1 alpha connector that cant translate characters right.

How to repeat:
Its the same all the time
[17 Oct 2007 9:15] Susanne Ebrecht
Hi Sören,

thanks for writing a bug report. Unfortunately we need some more informations.
What do you exactly mean with: the character set won't work?
Can you give us a simple test/example, that we can try to reproduce it?

Regards,

Susanne
[17 Oct 2007 9:34] Sören Forsdahl
If I use my swedish language, I got 3 national charachers ( å ä ö)
If I use them with 5.1 alpha I get som wierd text insteed of this characters.
I have read that this was also a problem with 5.0.
I use UTF-8 charset and all work good with the 3.51 version.

Greetz
[17 Oct 2007 10:27] Susanne Ebrecht
Hi Sören,

we need an example from you. What do you do exactly? Do you use _latin1' or _utf8' or do you use set names?

Please, send us your test code.

Also it make sense, when you send us the trace file too.

Regards,

Susanne
[30 Oct 2007 8:19] Susanne Ebrecht
Hi Sören,

Can you tell us, which Software you used?
Where do you see the wrong output from the swedish characters?

Kind Regard,

Susanne
[18 Nov 2007 0:00] Bugs System
No feedback was provided for this bug for over a month, so it is
being suspended automatically. If you are able to provide the
information that was originally requested, please do so and change
the status of the bug back to "Open".
[19 Nov 2007 11:37] Susanne Ebrecht
Sören,

we still need the information, what exactly you are doing.

Kind Regards,

Susanne
[19 Dec 2007 13:35] Tonci Grgin
I presume that data was probably inserted with MyODBC 3.51 under Latin1 (or similar) c-set into UTF8 table/field. Afterwards data was read with MyODBC 5.x, which is UTF8 capable, resulting in double conversion and garbage in output. It also results in MyODBC 3.51 being able to read the data as it does not care for UTF8.

Sören, if your whole table is affected that way, you can simply mark your fields as binary, then again as utf8 like this (thus effectively refreshing c-set info):

ALTER TABLE t_name MODIFY f_name VARCHAR(xx) CHARACTER SET binary;
ALTER TABLE t_name MODIFY f_name VARCHAR(xx) CHARACTER SET utf8;

If you want to extract particular data in such field, you can:

SET NAMES latin1; (or check appropriate cset in MyODBC 5 configuration)
SELECT data FROM t_name;

Leaving this report to expire.
[20 Dec 2007 0:00] Bugs System
No feedback was provided for this bug for over a month, so it is
being suspended automatically. If you are able to provide the
information that was originally requested, please do so and change
the status of the bug back to "Open".
[30 May 2013 6:37] Bogdan Degtyariov
I'm closing this bug because I can not continue without feedback from the reporter. If you have new info, please reopen the report.