将所有 VARCHAR 列更改为 NVARCHAR

Question

I have about 200 columns in my database with VARCHAR type which is unable to store the rupee symbol. Now I have to change the datatype of all the columns from VARCHAR to NVARCHAR .

Can anyone please tell me the short way to accomplish this? And why does the VARCHAR support the pound sign and not the rupee sign? I'm asking because I have to change the pound symbol to rupee symbol.

Answer 1

You can view all the columns, their data types and the table they belong using the ff query:

SELECT 
    t.name AS table_name,
    c.name AS column_name,
    tp.name AS data_type,
    c.max_length,
    c.is_nullable
FROM sys.tables AS t
INNER JOIN sys.columns c 
    ON t.OBJECT_ID = c.OBJECT_ID
INNER JOIN sys.types tp
    ON tp.user_type_id = c.user_type_id
WHERE tp.name = 'varchar'

From the query above, you want to generate a dynamic sql that will change all your VARCHAR columns to NVARCHAR .

DECLARE @sql AS VARCHAR(MAX) = ''

SELECT @sql = @sql 
    + 'ALTER TABLE ' + table_name 
    + ' ALTER COLUMN ' + column_name + ' NVARCHAR(' 
        + CASE WHEN max_length <> - 1 THEN CAST(max_length AS VARCHAR(10))  ELSE 'MAX' END + ')'
        + CASE WHEN is_nullable = 1 THEN ' NULL' ELSE '' END
        + ';' + CHAR(10)
FROM (
    SELECT 
        t.name AS table_name,
        c.name AS column_name,
        tp.name AS data_type,
        c.max_length,
        c.is_nullable
    FROM sys.tables AS t
    INNER JOIN sys.columns c 
        ON t.OBJECT_ID = c.OBJECT_ID
    INNER JOIN sys.types tp
        ON tp.user_type_id = c.user_type_id
    WHERE tp.name = 'varchar'
)t

PRINT @sql
EXEC(@sql)

Answer 2

I met this issue today and Felix's answer worked quite well with one exception- if the column has default value defined. I researched a little bit and came up with the following script that drops and restores the default value.

DECLARE @sql AS VARCHAR(MAX) = ''

SELECT @sql = @sql
    + CASE WHEN default_id>0 THEN 'ALTER TABLE '+table_name+' DROP CONSTRAINT '+OBJECT_NAME(default_id) +';'+CHAR(10) ELSE '' END
    + 'ALTER TABLE ' + table_name 
    + ' ALTER COLUMN ' + column_name + ' NVARCHAR(' 
        + CASE WHEN (max_length <> - 1 AND max_length<4001) THEN CAST(max_length AS VARCHAR(10))  ELSE 'MAX' END + ')'
        + CASE WHEN is_nullable = 1 THEN ' NULL' ELSE '' END
        + ';' + CHAR(10)
    + CASE WHEN default_id>0 THEN 'ALTER TABLE '+table_name+' ADD DEFAULT '+default_value+' FOR '+column_name +';'+CHAR(10) ELSE '' END
FROM (
    SELECT 
        t.name AS table_name,
        c.name AS column_name,
        tp.name AS data_type,
        c.max_length,
        c.default_object_id AS default_id,
        OBJECT_DEFINITION(c.default_object_id) AS default_value,
        c.is_nullable
    FROM sys.tables AS t
    INNER JOIN sys.columns c 
        ON t.OBJECT_ID = c.OBJECT_ID
    INNER JOIN sys.types tp
        ON tp.user_type_id = c.user_type_id
    WHERE tp.name = 'varchar'
)t

PRINT @sql
EXEC(@sql)

Answer 3

Normally in sqlserver VARCHAR datatype allows only the ANSI Characters, But NVARCHAR allows UNICODE Character sets also, Pound symbol(156-ascii number), $ symbols are comes under ANSI character set.

So VARCHAR allows those.

But when u comes to the Rupee Symbol its recently invented so it comes under UNICODE character set, to achieve the rupee symbol we need to use rupees font or glyphs... Hope you understood y Pound comes under VARCHAR and Rupee Comes under NVARCHAR ...

Answer 4

Adding on to Felix Pamittan's answer if you need to include schema.

DECLARE @sql AS VARCHAR(MAX) = ''

SELECT @sql = @sql 
    + 'ALTER TABLE ' + table_schema_name + '.' + table_name 
    + ' ALTER COLUMN ' + column_name + ' NVARCHAR(' 
        + CASE WHEN max_length <> - 1 THEN CAST(max_length AS VARCHAR(10))  ELSE 'MAX' END + ')'
        + CASE WHEN is_nullable = 1 THEN ' NULL' ELSE '' END
        + ';' + CHAR(10)
FROM (
    SELECT 
        sc.name AS table_schema_name,
        t.name AS table_name,
        c.name AS column_name,
        tp.name AS data_type,
        c.max_length,
        c.is_nullable
    FROM sys.tables AS t
    INNER JOIN sys.columns c 
        ON t.OBJECT_ID = c.OBJECT_ID
    INNER JOIN sys.types tp
        ON tp.user_type_id = c.user_type_id
    INNER JOIN sys.schemas sc
        ON sc.schema_id = t.schema_id
    WHERE tp.name = 'varchar'
)t

PRINT @sql
EXEC(@sql)

Answer 5

SELECT 'ALTER TABLE' +QUOTENAME(TABLE_SCHEMA)+'.'+QUOTENAME(TABLE_NAME)+ 'ALTER COLUMN ' + COLUMN_NAME + ' NVARCHAR(100)'
FROM INFORMATION_SCHEMA.COLUMNS
WHERE TABLE_NAME = N'TAbleName'

Answer 6

This will loop through all tables and their columns, and will generate query for updating those which have datatype as varchar

SET NOCOUNT ON

DECLARE @TableName nvarchar(256), @ColumnName nvarchar(128), @Length nvarchar(128)
SET  @TableName = ''

WHILE @TableName IS NOT NULL

BEGIN
    SET @ColumnName = ''
    SET @TableName = 
    (
        SELECT MIN(QUOTENAME(TABLE_SCHEMA) + '.' + QUOTENAME(TABLE_NAME))
        FROM     INFORMATION_SCHEMA.TABLES
        WHERE         TABLE_TYPE = 'BASE TABLE'
            AND    QUOTENAME(TABLE_SCHEMA) + '.' + QUOTENAME(TABLE_NAME) > @TableName
            AND    OBJECTPROPERTY(
                    OBJECT_ID(
                        QUOTENAME(TABLE_SCHEMA) + '.' + QUOTENAME(TABLE_NAME)
                         ), 'IsMSShipped'
                           ) = 0
    )

    WHILE (@TableName IS NOT NULL) AND (@ColumnName IS NOT NULL)

    BEGIN
            SELECT @ColumnName=MIN(QUOTENAME(COLUMN_NAME)),@Length=MIN(CHARACTER_MAXIMUM_LENGTH)
            FROM     INFORMATION_SCHEMA.COLUMNS
            WHERE         TABLE_SCHEMA    = PARSENAME(@TableName, 2)
                AND    TABLE_NAME    = PARSENAME(@TableName, 1)
                --AND    DATA_TYPE IN ('char', 'varchar', 'nchar', 'nvarchar', 'int', 'decimal')
                AND    DATA_TYPE IN ('varchar')
                AND    QUOTENAME(COLUMN_NAME) > @ColumnName


        IF @ColumnName IS NOT NULL and @Length > 0

        BEGIN

            print (' ALTER TABLE ' +@TableName+ ' ALTER COLUMN ' + @ColumnName + ' NVARCHAR('+@Length+')')

            --exec (@dSql)

        END
    END    
END