Originally posted by: intp

Hi,

We're using WTX 8.4 and we need to read a database field with cyrillic characters.

We have tried a DBQUERY in a type tree only with an item field in Cyrillic, also in UTF-8 and UTF-16, but the result/file saved doesn't contain the correct characters.

We've tried with the field in an Oracle and in a MS SQL Server table, if you read the field in the sql client it returns it correctly, but the dbquery in the wtx map can not retrieve it.

Please, any help

Originally posted by: ibm_soa_gbl

Hi 

Some special characters we cannot see in Text editors like notepad.

Open the result of the wtx in Dos prompt and check you can see it or not.

Regards

Gopi

Originally posted by: intp

Hi,

We've checked it with other editors, and it's not a problem of visualization.

If we use the odbc, instead of native driver, we can see an error:

Error in SQLFetch

Message: [IBM WTX][ODBC Oracle Wire Protocol driver]String data code page conversion failed.

or  with the field in SDL Server:

Error in SQLFetch

Message: [IBM WTX][ODBC SQL Server Wire Protocol driver]String data code page conversion failed. Error in column 1.

It seems that the database adapter can not manage the code page.

Originally posted by: Sidhumohan

First Question in What is the encoding format of the input file? and second the type tree language setting.

One more issue is you need to set an environmental variable at system level for handling such kind of characters.

Add an environmental user variable at system level as below 

Variable Name - NLS_LANG

Value - AMERICAN_AMERICA.AL32UTF8

Rerun the map and check. This should work else will try something else.

Originally posted by: intp

Hi,

We've checked the database language:

SELECT * from NLS_DATABASE_PARAMETERS;

NLS_LANGUAGE = AMERICAN

NLS_CHARACTERSET = UTF8

And in the client (WTX server), NLS_LANG in the registry:

SPANISH_SPAIN.WE8MSWIN1252

If we do the query in the sql developer in the WTXserver, it returns correctly the characters:

Мытищи

So it's strange that the sql developer can read it and the adapter not, isn't it?

Can we read the data in WTX without conversion and then managing it in the mapping?

The type tree uses Western, UTF-8. We've tried UTF-16 and cyrillic (ibm-1256) too.

Thanks,

Originally posted by: intp

Hi,

We have added an environmental user variable

Variable Name - NLS_LANG

Value - SPANISH_SPAIN.AL32UTF8

And the type tree with language native and it can read it correctly.

Now, we''ll check how it affect to other existing mappings, or if there's an option to run them with different configuration...

Thanks,