Question

我有以下列和列的数据类型

DailyWorkingHours as Daily (money)
case_completion_hour as casecomp (money)
xdate as month (datetime)

我有以下查询

select 
sum(cpw.DailyWorkingHours) as Daily, 
sum(ca.case_completion_hour) / 60 as casecomp, 
DateName( month , DateAdd( month , month(ca.xdate) , 0 ) - 1 ) as month

from reseller_user_profiles rup

join cases c on c.reseller_user_ID=rup.reseller_user_ID
join case_action ca on ca.case_ID=c.case_ID
join tblCaseProjectWorkers cpw on c.reseller_user_ID=cpw.resellerUserID 
join tblCaseProjectsNew cpn on cpn.ID=c.project_ID

group by 
DateName( month , DateAdd( month , month(ca.xdate) , 0 ) - 1 ) 

order by  DateName( month , DateAdd( month , month(ca.xdate) , 0 ) - 1 )

如果我运行查询，它将返回如下数据

Daily (column)  -  casecomp (column)  -  month (column)

1088,00           -  0,3333             -  February

544,00            -  0,3344             -  February

321,00            -  0,3377             -  February

150,00            -  0,3387             -  January

332,00            -  0,3330             -  January

658,00            -  4,3331             -  April

问题：

如何选择查询结果如下（即使月份为空，我想显示12个月）

Daily (column)  -  casecomp (column)  -  month (column)

150,00            -  0,3387             -  January

332,00            -  0,3330             -  

1088,00           -  0,3333             -  February

544,00            -  0,3344             -  

321,00            -  0,3377             -  

NULL              -  NULL               -  March

658,00            -  4,3331             -  April

NULL              -  NULL               -  May

NULL              -  NULL               -  June

NULL              -  NULL               -  July

NULL              -  NULL               -  August

NULL              -  NULL               -  September

NULL              -  NULL               -  October

NULL              -  NULL               -  November

NULL              -  NULL               -  December

在结果查询中按月内列进行月份列分组。根据实现这个，我的查询代码中要改变什么？

Answer 1

试一试。

    CREATE TABLE #Months (Num int, Name varchar(15))

    INSERT INTO #Months SELECT 1, 'January'
    INSERT INTO #Months SELECT 2, 'February'
    INSERT INTO #Months SELECT 3, 'March'
    INSERT INTO #Months SELECT 4, 'April'
    INSERT INTO #Months SELECT 5, 'May'
    INSERT INTO #Months SELECT 6, 'June'
    INSERT INTO #Months SELECT 7, 'July'
    INSERT INTO #Months SELECT 8, 'August'
    INSERT INTO #Months SELECT 9, 'September'
    INSERT INTO #Months SELECT 10, 'October'
    INSERT INTO #Months SELECT 11, 'November'
    INSERT INTO #Months SELECT 12, 'December'


    SELECT *
    FROM 
        #Months m
        LEFT JOIN 
            (
            select 
            sum(cpw.DailyWorkingHours) as Daily, 
            sum(ca.case_completion_hour) / 60 as casecomp, 
            DateName( month , DateAdd( month , month(ca.xdate) , 0 ) - 1 ) as month

            from reseller_user_profiles rup

            join cases c on c.reseller_user_ID=rup.reseller_user_ID
            join case_action ca on ca.case_ID=c.case_ID
            join tblCaseProjectWorkers cpw on c.reseller_user_ID=cpw.resellerUserID 
            join tblCaseProjectsNew cpn on cpn.ID=c.project_ID

            group by 
            DateName( month , DateAdd( month , month(ca.xdate) , 0 ) - 1 ) 


            ) yours
            ON yours.month = m.Name
ORDER BY m.Num

Answer 2

试试这个：

;with Months as (
  select 1 As Months_Number, 
         DateName(month, '2015-01-01') As Months_Name
  UNION ALL
  select Months_Number + 1 , DateName(month, DateAdd(month, Months_Number, '2015-01-01')) As Months_Name
  FROM Months
  WHERE Months_Number < 12
)

select 
sum(cpw.DailyWorkingHours) as Daily, 
sum(ca.case_completion_hour) / 60 as casecomp, 
Months_Name

from Months 
left join (
    reseller_user_profiles rup
    join cases c on c.reseller_user_ID=rup.reseller_user_ID
    join case_action ca on ca.case_ID=c.case_ID
    join tblCaseProjectWorkers cpw on c.reseller_user_ID=cpw.resellerUserID 
    join tblCaseProjectsNew cpn on cpn.ID=c.project_ID) on month(ca.xdate) = Months_Number 

group by Months_Name
order by Months_Name

兴趣点：

递归CTE为您提供从0到11的计数。这样可以实现简单而准确的left join条件。（参见我对EarlOfEnnui的回答的评论）
递归CTE还会为您提供月份名称。这允许更简单的group by和order by条款。
left join的{{1}}与您的所有其他表格相同（请注意，所有其他表格都有括号

T sql group by only column

2 个答案: