Dec 8, 2009 at 11:44 AM
Edited Dec 8, 2009 at 11:46 AM
I ran into the same issue, working with NVARCHAR columns and cyrillic letters (Bulgarian). After looking into the SQL Profiler for a while, I found out that the SQL generated by SLAM doesn't manage NVARCHAR conversion as it should. It executes INSERT/UPDATE
statements without casting the NVARCHAR fields to NVARCHAR (should prefix the literal in T-SQL with N).
However, there is a way of going around this in some cases.
The solution is to change the collation of the SLAM DB to one compatible with the charset you want to store in the DB.
Following is what I've done in order to do this:
1. Use MSSQL Server Management Studio to generate a script of the SLAM Db.
2. Append the collation to the "CREATE DATABASE " statement like this:
CREATE DATABASE [SLAM_DB] COLLATE cyrillic_general_ci_as
3. Recreate the rest of the DB. After collating the DB itself with a default collation, the rest of the objects will inherit it.