SQL Server is set based. Sometimes we get data that isn't. And, when that happens, we need to convert it. One way to do this is using a simple function like this one:

-- =============================================
-- Author:        Josef Finsel
-- Create date: 2008-12-1
-- Description:    Split delimited list into
--              generic single column rows
--                based on passed in delmiter
-- =============================================
ALTER FUNCTION [dbo].[Delimited2Rows]
(
    @DelimitedString nvarchar(max),
    @Delimiter char(1)
)
RETURNS
@HoldingTable TABLE (
    IDColumn bigint identity primary key,
    DataColumn nvarchar(1024) )
AS
BEGIN
    while charindex(@Delimiter,@DelimitedString)>1
     begin
      INSERT INTO @HoldingTable VALUES(ltrim(rtrim(substring(@DelimitedString,1, charindex(@Delimiter,@DelimitedString)-1))))
      set @DelimitedString = substring(@DelimitedString, charindex(@Delimiter,@DelimitedString)+1, len(@DelimitedString))
     end
    if len(@DelimitedString)>0
      INSERT INTO @HoldingTable values(rtrim(ltrim(@DelimitedString)))
    RETURN
END

-- SELECT * FROM dbo.Delimited2Rows('testing;another row;WHat about me?!?;',';')

Now, SQL Server isn't really good at string manipulation, so this isn't something I'd want in a heavily used stored procedure. But, for converting data, it's handy. It doesn't support correctly parsing quoted text, so SELECT * FROM dbo.Delimited2Rows('"I like to eat cows; however, they don''t like to be eaten by me."',';') will return two rows, one containing "I like to eat cows and one containing however, they don't like to be eaten by me." Again, it's not designed to be a full blown parser, just a short step toward making life easier sometimes.