A department store uses the following table to track sales. The data model does not include information about the floor and the category of the goods been sold.
CREATE TABLE A1
(
date datetime,
category nvarchar(30),
turnover money
);
INSERT INTO A1 VALUES (SYSDATETIME(), 100, 1000);
INSERT INTO A1 VALUES (SYSDATETIME(), 201, 1700);
...
Data model/ table cannot be changed or edited. Information on the product group and floor can be diverted from the category.
CREATE VIEW sort_department_productgroups_view
AS
WITH sort_department_productgroups
AS (SELECT date,
category,
turnover,
CASE category
WHEN 100 THEN 1
WHEN 201 THEN 1
WHEN 303 THEN 1
WHEN 101 THEN 2
WHEN 102 THEN 2
ELSE 9
END
floor,
CASE category
WHEN 100 THEN 'sport'
WHEN 102 THEN 'sport'
WHEN 201 THEN 'leisure'
WHEN 303 THEN 'business'
WHEN 101 THEN 'sport'
WHEN 202 THEN 'leisure'
ELSE 'unknown'
END
productgroup
FROM a1)
SELECT *
FROM sort_department_productgroups;
go
Example query on the new view:
SELECT * FROM sort_department_productgroups_view where productgroup='sport';
Are there better ways to deal with such a task? Would this work on a big database?
This is really crying for a lookup table. But I must admit, that a category column of type nvarchar filled with numeric values makes me wonder...
Try it like this:
NVARCHAR(30)
as OP mentioned, that this is unchangeable...CREATE TABLE CATEGORY(ID INT IDENTITY CONSTRAINT PK_CATEGORY PRIMARY KEY
,category NVARCHAR(30)
,CategoryName VARCHAR(100)
,[Floor] INT );
INSERT INTO CATEGORY VALUES
(100,'sport',1)
,(101,'sport',2)
,(102,'sport',2)
,(201,'leisure',1)
,(202,'leisure',9)
,(303,'business',9)
--add more...
CREATE TABLE A1
(
date datetime,
category NVARCHAR(30),
turnover money
);
INSERT INTO A1 VALUES (SYSDATETIME(), 100, 1000);
INSERT INTO A1 VALUES (SYSDATETIME(), 201, 1700);
GO
CREATE VIEW sort_department_productgroups_view
AS
SELECT A1.date
,a1.turnover
,ISNULL(CATEGORY.Floor,9) AS [Floor]
,ISNULL(CATEGORY.CategoryName,'unknown') AS [productgroup]
FROM A1
LEFT JOIN CATEGORY ON CATEGORY.category=A1.category;
GO
SELECT * FROM sort_department_productgroups_view