December 1, 2006 at 10:28 am
The problem: in Enterprise Manager, when entering Polish accented characters into a varchar field, they are saved as unaccented 'equivalent' letters. (EM is running directly on the SQL Server itself.)
The infrastructure: Windows 2003 Server (Polish version, Polish locale), SQL Server 2000 (installed with Polish_CI_AS as default collation), database USERDB (created with SQL_Latin1_General_CP1_CI_AS as default collation), table USERTABLE (created with database defaults), column USERCOLUMN (collation changed to SQL_Polish_CP1250_CI_AS).
Strange fact: USERTABLE contains other rows with Polish accented characters present which display on screen OK!! How they got there, I don't know, perhaps imported somehow... Changes/new entries lose their accents when saved.
I have inherited this pre-production database and have been tasked with sorting it out - help!!
December 1, 2006 at 12:38 pm
Here are the steps Polish Collation for the database and columns and Nvarchar, NChar and NText for the data types. If the application is .NET do Polish encoding in the application layer, in VS2005 the advanced let you save the file as Unicode and the code page you choose. Hope this helps.
Kind regards,
Gift Peddie
December 2, 2006 at 4:49 am
Thanks for the reply. I am aware that I could make a wholesale change of columns from varchar, char and text to nvarchar, nchar and ntext. However, I would also have to check and change all relevant stored procedures. Also, there may be stuff "beyond my reach" in the front end app... So I'll keep that possibility as a last resort. Also, at this stage the application is not the issue - I am performing tests directly through EM.
If I simply change the database collation to one of the standard Polish collations and then use a script to reset all columns to the (new) database default collation, would that be enough do you think?
I am somewhat perplexed by what I see in the current configuration, because Polish characters do exist in the table. That tells me that varchar is OK to store those characters. Why can I not enter new data containing Polish characters though??
December 2, 2006 at 5:21 am
You need to do it now because character conversion happen with browser interaction with your program, I think you need to do it now when the database is small before it grows. I have not seen accent sensitive language that works with only database level collation. You can check the SQL Server 2005 BOL( books online) every language in Europe a separate collation and code page. Hope this helps.
Kind regards,
Gift Peddie
Viewing 4 posts - 1 through 3 (of 3 total)
You must be logged in to reply to this topic. Login to reply