Split out in columns from 1 field in SQL2012

Here's a way...

declare @t1 table (mult varchar(500))
insert into @t1 (mult) values ('£ 68.25 (£ 68.25) to SI 38816'),
('£ 360.00 (£ 360.00) to SI 26231'),
('£ 360.00 (£ 360.00) to SI 38151'),
('£ 165.72 (€ 210.00) to SI 38814')

select  
SUBSTRING(mult,1,1) as Col1
,ltrim(rtrim(SUBSTRING(mult,2,charindex('(',mult)-2))) as Col2
,SUBSTRING(mult,charindex('(',mult)+1,1) as Col3
,SUBSTRING(mult,charindex('(',mult)+2,CHARINDEX(')',mult)-charindex('(',mult)-2) as Col4
,SUBSTRING(mult,CHARINDEX(') to ',mult)+5,2) as Col5
,RIGHT(mult, CHARINDEX(' ', REVERSE(mult)) - 1)as Col6
from @t1

Select allOpen in new window

You'd just use the select statement, replacing 'mult' with the name of your field.

/*
--run one time to set up test data
CREATE TABLE #test (
    details varchar(100)
    )
insert into #test ( details )
select '£ 68.25 (£ 68.25) to SI 38816' union all
select '£ 360.00 (£ 360.00) to SI 26231' union all
select '£ 360.00 (£ 360.00) to SI 38151' union all
select '£ 165.72 (€ 210.00) to SI 38814' union all
select '£ 360.00 (£ 360.00) to SIG 381517' --test value longer than 2 bytes after " to", in case it's possible

*/

SELECT
    LEFT(details, 1) AS Value1,
    LTRIM(RTRIM(SUBSTRING(details, 2, pos_of_left_paren - 2))) AS Value2,
    SUBSTRING(details, pos_of_left_paren + 1, 1) AS Value3,
    LTRIM(SUBSTRING(details, pos_of_left_paren + 2, pos_of_right_paren - pos_of_left_paren - 2)) AS Value4,
    SUBSTRING(details, pos_of_space_to + 4, CHARINDEX(' ', details, pos_of_space_to + 4) - (pos_of_space_to + 4)) AS Value5,
    RIGHT(details, CHARINDEX(' ', REVERSE(details)) - 1) AS Value6
FROM #test t
CROSS APPLY (
    SELECT CHARINDEX('(', details) AS pos_of_left_paren,
        CHARINDEX(')', details) AS pos_of_right_paren,
        CHARINDEX(' to', details) AS pos_of_space_to
) AS assign_alias_names

@Scott: Much better! I knew I should have gone the extra mile and dealt with the possibility of longer values in column 5!

Does every row of the data you're passing to the select statement look like the sample you posted in your question?

I'd guess that you have a line without at least one of the key identifying strings in it, or not in the position suggested by your sample data  -
'('          - cannot be the first character of the string
')'
' to'

You might want to add a where clause to the end of the SELECT statement to avoid trying to process rows with strings that don't conform to the expected pattern.

    CROSS APPLY (
SELECT CHARINDEX('(', details) AS pos_of_left_paren,
        CHARINDEX(')', details) AS pos_of_right_paren,
        CHARINDEX(' to', details) AS pos_of_space_to
) AS assign_alias_names 
WHERE details like '_%(_%) to% _%[^ ]'

Select allOpen in new window

You can use a separate query

SELECT * from CPMAuditUsage where details NOT LIKE '_%(_%) to% _%[^ ]'

Select allOpen in new window

To see which rows do not suit the current solution.

Be aware that with both Scott's proposed solution and my own, trailing spaces in your string currently prevent Col6 from being populated correctly. If your column is nvarchar, the LIKE pattern will omit those, but if it is char or varchar they would be included but have empty string for col6 value.

That can obviously be fixed by adding a RTRIM to this line.

  RIGHT(details, CHARINDEX(' ', REVERSE(rtrim(details))) - 1) AS Value6

Select allOpen in new window

but the point is that if your data isn't properly represented by the sample you included, the solution isn't likely to work entirely as expected.

Sounds like some data doesn't have a (.

Can you list sample data in other formats and briefly state/show how they should be interpreted.

Good luck with future qs.

this only parttially solved it and I think I found the answer in an earlier question to my issue