SQL Server String Concatenation with Null

Question

I am creating a computed column across fields of which some are potentially null.

The problem is that if any of those fields is null, the entire computed column will

Answer 1

Use COALESCE. Instead of your_column use COALESCE(your_column, ''). This will return the empty string instead of NULL.

Answer 2

In Sql Server:

insert into Table_Name(PersonName,PersonEmail) values(NULL,'xyz@xyz.com')

PersonName is varchar(50), NULL is not a string, because we are not passing with in single codes, so it treat as NULL.

Code Behind:

string name = (txtName.Text=="")? NULL : "'"+ txtName.Text +"'";
string email = txtEmail.Text;

insert into Table_Name(PersonName,PersonEmail) values(name,'"+email+"')

Answer 3

From SQL Server 2012 this is all much easier with the CONCAT function.

It treats NULL as empty string

DECLARE @Column1 VARCHAR(50) = 'Foo',
        @Column2 VARCHAR(50) = NULL,
        @Column3 VARCHAR(50) = 'Bar';


SELECT CONCAT(@Column1,@Column2,@Column3); /*Returns FooBar*/

Answer 4

I just wanted to contribute this should someone be looking for help with adding separators between the strings, depending on whether a field is NULL or not.

So in the example of creating a one line address from separate fields

Address1, Address2, Address3, City, PostCode

in my case, I have the following Calculated Column which seems to be working as I want it:

case 
    when [Address1] IS NOT NULL 
    then (((          [Address1]      + 
          isnull(', '+[Address2],'')) +
          isnull(', '+[Address3],'')) +
          isnull(', '+[City]    ,'')) +
          isnull(', '+[PostCode],'')  
end

Hope that helps someone!

Answer 5

ISNULL(ColumnName, '')

Answer 6

This example will help you to handle various types while creating insert statements

select 
'insert into doc(Id, CDate, Str, Code, Price, Tag )' + 
'values(' +
      '''' + convert(nvarchar(50), Id) + ''',' -- uniqueidentifier
    + '''' + LEFT(CONVERT(VARCHAR, CDate, 120), 10) + ''',' -- date
    + '''' + Str+ ''',' -- string
    + '''' + convert(nvarchar(50), Code)  + ''',' -- int
    + convert(nvarchar(50), Price) + ',' -- decimal
    + '''' + ISNULL(Tag, '''''') + '''' + ')'  -- nullable string

 from doc
 where CDate> '2019-01-01 00:00:00.000'