Evil Kraig F (6/27/2012)

---

Alright, it's not pretty but if you actually need multiple unique datasets returned...

MyStr, MyStr1 and MyStr2 in my version are just the names of the computed fields which you can substitute for your actual column as needed.

I'm not GSquared (nor do I pretend to be), but this may work for you:

;WITH TestData (MyStr) AS (

SELECT '\Big_Fam_Beverages\Soda Pop\Mountain Dew'

UNION ALL SELECT '\Lcal_Beverages\Soda Pop'

UNION ALL SELECT...

I like this problem and wanted to contribute. Alas my query contribution (as follows) barks like the dog that it is.

;WITH Booking AS (

SELECT ID,...

But if you're wedded to CASE you can do it like this:

SELECT @CMD=

case

when @model = 'Global' then

'SELECT '+@model+' AS Model

, Scenario_ID

FROM lp_Scenario AS a'

Else

'SELECT '+@model+' AS Model

, a.Scenario_ID

,...

dwilliscp (6/27/2012)

---

That is a simplistic version of the code.. the problem I have with IF statements, is that you have to create a very long IF THEN Else IF... version...

tinausa (6/27/2012)

---

The table is as below

[id] [subid] [pct]

abc w 10

abc e 20

abc g ...

Amen(!) to that.

Hmmm. Must've missed that. So then...

Les Cardwell (6/27/2012)

---

Dwain,

I must be doing something wrong here. I'm using the CTE to attempt the same logic as SW7 to resolve those new Sweden districts with two old districts...

Have you tried removing dbs.DBAvailMB from the GROUP BY?

Use GROUP BY on SQL Server and App name fields and SUM the fields you want to add up.

I am not sure why you're thinking UNION ALL.

All you need to do is replace SampleData after FROM in Chris's query with your table name (and get rid of his...

Chris has parsed the availability date only for you.

You'd need to apply the CHARINDEX steps to parse out the other fields you need from the InputString in a similar fashion.

His...

My expectation was that Chris's query would be faster. Shall we check?

Test harness (5000 rows):

SET STATISTICS TIME ON

;WITH Tally AS (

SELECT TOP 1000 n=ROW_NUMBER() OVER...