db2 查询多种情况

Question

我在下面有三个查询（除了 WHERE 子句的第一行之外相同），它们都可以在我的脚本中完美运行。第一个查询客户的订单，第二个查询分配给代表的所有订单，第三个查询整个公司的所有订单周期。

同样，它们都可以在给定各自变量的情况下工作（所有变量都来自同一页面），但我正在尝试为所有 3 种情况填充表格中的列。

有没有一种方法可以组合这些并创建一个查询，为每个子句提供相同的值？

所以，我希望一个查询返回所有 6 列。这是在 db2 上运行的，所以我不知道最好的方法，但我可以创建一个更大的基于 CASE 的查询吗？

//query on orders for this customer
SELECT
    count(*) as sales_180Cust,
    180/count(*) as velocityCust

FROM orders g
    inner join dates i
    on g.date1 = i.acyyyymmdd
WHERE g.cust = $customer
AND g.frm = $frm
AND g.cvr = $cvr
AND g.clr = $clr
AND i.aciso between current_Date - 180 DAY AND current_Date;

//orders belonging to representative
SELECT
    count(*) as sales_180Rep,
    180/count(*) as velocityRep

FROM orders g
    inner join dates i
    on g.date1 = i.acyyyymmdd
WHERE g.rep = $rep
AND g.frm = $frm
AND g.cvr = $cvr
AND g.clr = $clr
AND i.aciso between current_Date - 180 DAY AND current_Date;

//query across ALL orders
SELECT
    count(*) as sales_180Company,
    180/count(*) as velocityCompany

FROM orders g
    inner join dates i
    on g.date1 = i.acyyyymmdd
WHERE g.frm = $frm
AND g.cvr = $cvr
AND g.clr = $clr
AND i.aciso between current_Date - 180 DAY AND current_Date;

Answer 1

这将是另一种方法

WITH CTE AS (
    SELECT  g.cust
    ,       g.rep
    FROM orders g
        inner join dates i
        on g.date1 = i.acyyyymmdd
    WHERE 
        g.frm = $frm
    AND g.cvr = $cvr
    AND g.clr = $clr
)
SELECT
    count(*)     as sales_180Company
,   180/count(*) as velocityCompany
,   'cust' as query
FROM CTE
WHERE cust = $customer
UNION ALL
SELECT
    count(*)     as sales_180Company
,   180/count(*) as velocityCompany
,   'rep' as query
FROM CTE
WHERE rep = $rep
UNION ALL
SELECT
    count(*)     as sales_180Company
,   180/count(*) as velocityCompany
,   'all' as query
FROM CTE

返回例如

 SALES_180COMPANY VELOCITYCOMPANY QUERY
 ---------------- --------------- -----
                3              60 cust
                2              90 rep
                5              36 all

Answer 2

select q1.*, q2.*, q3.*
from 
  (select count(*) as sales_180Cust,    180/count(*) as velocityCust    from table(values 1) t(i)) q1
, (select count(*) as sales_180Rep,     180/count(*) as velocityRep     from table(values 1, 2) t(i)) q2
, (select count(*) as sales_180Company, 180/count(*) as velocityCompany from table(values 1, 2, 3) t(i)) q3

我修改了你的 FROM 和 WERE 子句来表达这个想法。

Answer 3

有很多方法可以做到这一点。 UNION 将是一种显而易见的方式。 GROUPING SETs 更聪明一点。

create table orders(date1 int, rep int, cust int,frm int, cvr int, clr int, aci int);
create table dates(acyyyymmdd int, aciso date);
create variable $frm int default 1;
create variable $cvr int default 1;
create variable $clr int default 1;
create variable $customer int default 1;
create variable $rep int default 1;
insert into orders values (1,0,0,1,1,1,1), (1,1,1,1,1,1,1), (1,2,1,1,1,1,1), (1,3,1,1,1,1,1), (1,1,2,1,1,1,1);
insert into dates values (1, current date); 

然后这个

SELECT
    count(*)     as sales_180Company
,   180/count(*) as velocityCompany
,   g.cust
,   g.rep 
FROM orders g
    inner join dates i
    on g.date1 = i.acyyyymmdd
WHERE 
    g.frm = $frm
AND g.cvr = $cvr
AND g.clr = $clr
AND i.aciso between current_Date - 180 DAY AND current_Date
GROUP BY GROUPING SETS ( (), (cust), (rep) )
HAVING (cust = $customer AND rep is null) 
OR     (cust is null AND rep = $rep)
OR     (cust is null AND rep is null)

给这个

 SALES_180COMPANY VELOCITYCOMPANY CUST REP
 ---------------- --------------- ---- ----
                5              36 NULL NULL
                3              60    1 NULL
                2              90 NULL    1

Answer 4

或者这个...在 SQL 中通常有多种方式来处理事情

WITH CTE AS (
    SELECT  g.cust
    ,       g.rep
    FROM orders g
        inner join dates i
        on g.date1 = i.acyyyymmdd
    WHERE 
        g.frm = $frm
    AND g.cvr = $cvr
    AND g.clr = $clr
) , CUST AS (
SELECT
    count(*)     as sales_180Company
,   180/count(*) as velocityCompany
FROM CTE
WHERE cust = $customer
) , REP AS (
SELECT
    count(*)     as sales_180Company
,   180/count(*) as velocityCompany
FROM CTE
WHERE rep = $rep
) , ALL AS (
SELECT
    count(*)     as sales_180Company
,   180/count(*) as velocityCompany
FROM CTE
)
SELECT * FROM CUST, REP, ALL

重新调整

 SALES_180COMPANY VELOCITYCOMPANY SALES_180COMPANY VELOCITYCOMPANY SALES_180COMPANY VELOCITYCOMPANY
 ---------------- --------------- ---------------- --------------- ---------------- ---------------
                3              60                2              90                5              36