Chinese in Transformer

[MellBI]

Hi,
I am trying to produce a global cube with customers and their names.
I have a issue with e.g. the Chinese customers having a Chinese sign in the name, they are displayed as a question mark (?). Anyone who knows how to show the correct sign?
The data is stored in a MySQL database and is displayed correct in the database, (UTF-8).

How do I solve this?

Thanks
MellBI

[Skippystrand]

I might have a couple options to your solution. I experienced the same issue this week:

I'm currently using Sybase.
Our Sybase system might already be UTF-8 aware, so that the input stream needs to be copied in from a UTF-8 character set. In other words, if we DO NOT interpret the raw files and just load them into the Sybase system we will get the right data.
If we load the content into another database that is UTF-8 as garbled ASCII characters, we will get garbled ascii, since it did the valid type conversion for us.

SELECT T.vendor_id, T.language FROM dbo.test_foreign T
UPDATE dbo.test_foreign SET language='Blumen für dein Grab' WHERE vendor_id='ABC_COMMANDERINCHIEF_F003_FR' AND language='Blumen für dein'

id            vendor_id           language                 
1              ABC_ALIAS_F015_DE     Première frappe               
2              ABC_COMMANDERINCHIEF_F003_FR    Blumen für dein Grab       

If you are looking at the Cube and you show ??? marks, try doing a test report using your Cognos Browser:    (Taken from my notes on same issue: It looks like we have a working system in Cognos with UTF-8 characters displaying without having to set UTF-8 in the web browser. It is auto detecting the UTF-8 characters in my browser and displaying the correct output).

One last trick - See if your SQL has a "adapter" to place on top. It's kinda like when you slide a second "window" or reading glass over the top.

[MellBI]

Hi,

As far as I understand we are not able to use all character sets in the same cube. We would have to have one server with each locale.
We solved it by building reports directly against the datawarehouse instead.

/MellBI