Jeff Moden (11/29/2013)
Not that many people would notice but you can get the code to be almost a third faster by eliminating the (-1) subtraction. Last but not least, the simplification actually instills a guarantee that you won't have an accidental overrun sometimes caused by the very rare but annoying delay as to when the WHERE clause is executed when you join the results of this function in an external query (Itzik Ben-Gan published that correction to his code just a couple of days after he first published the cascading CTE method used here).
CREATE FUNCTION dbo.ExplodeDates
(
@pStartDate DATETIME
,@pEndDate DATETIME
)
RETURNS TABLE WITH SCHEMABINDING AS
RETURN WITH
E1(N) AS (
SELECT 1 UNION ALL SELECT 1 UNION ALL
SELECT 1 UNION ALL SELECT 1 UNION ALL
SELECT 1 UNION ALL SELECT 1 UNION ALL
SELECT 1 UNION ALL SELECT 1 UNION ALL
SELECT 1 UNION ALL SELECT 1
)
, E7(N) AS (SELECT 1 FROM E1 a, E1 b, E1 c, E1 d, E1 e, E1 f, E1 g)
SELECT TheDate = @pStartDate UNION ALL --This eliminates the need for subtraction on the Row_Number
SELECT TOP (DATEDIFF(dd,@pStartDate,@pEndDate)) --and addition here
TheDate = DATEADD(dd,ROW_NUMBER() OVER (ORDER BY (SELECT NULL)),@pStartDate)
FROM E7
;
How are you demonstrating the difference, Jeff? The results of a quick test show the opposite, but not by very much:
DECLARE @startdate DATE, @enddate DATE, @Blackhole DATE;
SET @startdate = '00010101'
SET @enddate = '99991231';
PRINT '==Slow inline tally table==';
PRINT CHAR(10)+'--Subtraction--';
SET STATISTICS TIME ON;
with
N0 as (SELECT 1 as n UNION ALL SELECT 1) -- 2 rows
,N1 as (SELECT 1 as n FROM N0 t1, N0 t2) -- 4 rows
,N2 as (SELECT 1 as n FROM N1 t1, N1 t2) -- 16 rows
,N3 as (SELECT 1 as n FROM N2 t1, N2 t2) -- 256 rows
,N4 as (SELECT 1 as n FROM N3 t1, N3 t2) -- 65,536 rows
,N5 as (SELECT 1 as n FROM N4 t1, N4 t2) -- 4,294,967,296 rows
,nums as (SELECT ROW_NUMBER() OVER (ORDER BY (SELECT 1)) as num FROM N5)
SELECT @Blackhole = thedate FROM (
SELECT thedate = DATEADD(day,num-1,@startdate)
FROM nums
WHERE num <= DATEDIFF(day, @startdate, @enddate) + 1
) d;
SET STATISTICS TIME OFF;
PRINT CHAR(10)+'--UNION--';
SET STATISTICS TIME ON;
with
N0 as (SELECT 1 as n UNION ALL SELECT 1) -- 2 rows
,N1 as (SELECT 1 as n FROM N0 t1, N0 t2) -- 4 rows
,N2 as (SELECT 1 as n FROM N1 t1, N1 t2) -- 16 rows
,N3 as (SELECT 1 as n FROM N2 t1, N2 t2) -- 256 rows
,N4 as (SELECT 1 as n FROM N3 t1, N3 t2) -- 65,536 rows
,N5 as (SELECT 1 as n FROM N4 t1, N4 t2) -- 4,294,967,296 rows
,nums as (SELECT ROW_NUMBER() OVER (ORDER BY (SELECT 1)) as num FROM N5)
SELECT @Blackhole = thedate FROM (
SELECT thedate = @startdate
UNION ALL
SELECT DATEADD(day,num,@startdate)
FROM nums
WHERE num <= DATEDIFF(day, @startdate, @enddate)
) d;
SET STATISTICS TIME OFF;
PRINT CHAR(10)+CHAR(10)+'==Fast IBG inline tally table==';
PRINT CHAR(10)+'--Subtraction--';
SET STATISTICS TIME ON;
WITH
E1(N) AS (
SELECT 1 UNION ALL SELECT 1 UNION ALL
SELECT 1 UNION ALL SELECT 1 UNION ALL
SELECT 1 UNION ALL SELECT 1 UNION ALL
SELECT 1 UNION ALL SELECT 1 UNION ALL
SELECT 1 UNION ALL SELECT 1
)
, E7(N) AS (SELECT 1 FROM E1 a, E1 b, E1 c, E1 d, E1 e, E1 f, E1 g)
SELECT @Blackhole = thedate FROM (
SELECT TOP (1+DATEDIFF(dd,@StartDate,@EndDate))
TheDate = DATEADD(dd,ROW_NUMBER() OVER (ORDER BY (SELECT NULL))-1,@StartDate)
FROM E7
) d;
SET STATISTICS TIME OFF;
PRINT CHAR(10)+'--UNION--';
SET STATISTICS TIME ON;
WITH
E1(N) AS (
SELECT 1 UNION ALL SELECT 1 UNION ALL
SELECT 1 UNION ALL SELECT 1 UNION ALL
SELECT 1 UNION ALL SELECT 1 UNION ALL
SELECT 1 UNION ALL SELECT 1 UNION ALL
SELECT 1 UNION ALL SELECT 1
)
, E7(N) AS (SELECT 1 FROM E1 a, E1 b, E1 c, E1 d, E1 e, E1 f, E1 g)
SELECT @Blackhole = thedate FROM (
SELECT TheDate = @StartDate
UNION ALL --This eliminates the need for subtraction on the Row_Number
SELECT TOP (DATEDIFF(dd,@StartDate,@EndDate)) --and addition here
TheDate = DATEADD(dd,ROW_NUMBER() OVER (ORDER BY (SELECT NULL)),@StartDate)
FROM E7
) d
;
SET STATISTICS TIME OFF;
PRINT '=================================================================================';
For fast, accurate and documented assistance in answering your questions, please read this article.
Understanding and using APPLY, (I) and (II) Paul White
Hidden RBAR: Triangular Joins / The "Numbers" or "Tally" Table: What it is and how it replaces a loop Jeff Moden