View Revisions: Issue #6632

Back to Issue
Summary 0006632: Convert to Unicode-friendly NVARCHAR type all internal SymmetricDS tables for SQL Server
Revision 2024-10-31 17:58 by pbelov
Description Convert to NVARCHAR any and all columns which currently use VARCHAR (on all SymmetricDS internal tables) when using MS SQL Server database.
Also, set the following engine parameter to true by default:
mssql.use.ntypes.for.sync=true
Revision 2024-10-31 17:59 by pbelov
Description Convert to Unicode-friendly NVARCHAR any and all columns which currently use VARCHAR (on all SymmetricDS internal tables) when using MS SQL Server database.
Also, set the following engine parameter to true by default:
mssql.use.ntypes.for.sync=true
Revision 2024-10-31 17:58 by pbelov
Additional Information
Revision 2024-10-31 18:06 by pbelov
Additional Information CREATE TABLE SymQaRoot.dbo.test_unicode_nvarchar_max(
    id int NOT NULL
    ,nvarchar_max nvarchar(max) NULL
    ,nvarchar_4000 nvarchar(4000) NULL
    ,PRIMARY KEY (id)
);
-- Set up replication for this table, then insert data with Unicode characters:
INSERT INTO SymQaRoot.dbo.test_unicode_nvarchar_max VALUES (11, 'NOTHING_special!', 'NOTHING_special!0');
INSERT INTO SymQaRoot.dbo.test_unicode_nvarchar_max VALUES (22, N'Cyrillic=Привет', N'Cyrillic=Привет'); -- Eg. Cyrillic alphabet

-- Query to detect when Unicode characters get "dropped":
SELECT * FROM SymQaRoot.dbo.test_unicode_nvarchar_max
WHERE CAST(nvarchar_max AS VARCHAR(max)) <> [nvarchar_max];

-- Confirm that the CSV data in the staging area uses Unicode support (at the file level) and does not convert Unicode characters to question marks: '?'.