根据两个日期之间的周/年计算 OT

Question

我们每月支付两次。支付期为 1-15 日和 16-EOM。因此，支付期的结束可以在任何一天结束。

我们在周一至周日支付超过 40 小时的加班费。

如果工资期在星期六结束并且员工有 47 小时，我们会在该支票上支付 7 小时。如果员工在星期天工作，而现在一周的总小时数是 52..我们将支付 5 小时他们的下一次检查.. 现在是手动计算过程。

我正在努力思考如何编写查询以使我获得额外的结转……

这是我上一个支付周期 2019 年 9 月 1 日至 2019 年 9 月 15 日的每日工作总量的输出，因为第一个是星期天。我需要从 8 开始计算一周的加班时间/ 2019 年 2 月 26 日。在这个特殊的员工身上，他周末随叫随到。在 2019 年 8 月 16 日至 2019 年 8 月 31 日期间，他获得了 6.28 小时的加班费，在 2019 年 9 月 1 日加班了两个小时，因此 2 小时的加班需要结转到 9 月 1 日/ 2019 年至 2019 年 9 月 15 日检查。

ID HRS WK CDATE STU02 8.16 35 2019-08-26 00:00:00.000 STU02 9.37 35 2019-08-27 00:00:00.000 STU02 9.07 35 2019-08-28 00:00:00.000 STU02 7.91 35 2019-08-29 00:00:00.000 STU02 9.12 35 2019-08-30 00:00:00.000 STU02 2.65 35 2019-08-31 00:00:00.000 STU02 2.00 35 2019-09-01 00:00:00.000 STU02 4.17 36 2019-09-02 00:00:00.000 STU02 9.40 36 2019-09-03 00:00:00.000 STU02 8.80 36 2019-09-04 00:00:00.000 STU02 8.90 36 2019-09-05 00:00:00.000 STU02 8.93 36 2019-09-06 00:00:00.000 STU02 2.56 36 2019-09-07 00:00:00.000 STU02 2.00 36 2019-09-08 00:00:00.000 STU02 8.66 37 2019-09-09 00:00:00.000 STU02 9.14 37 2019-09-10 00:00:00.000 STU02 9.07 37 2019-09-11 00:00:00.000 STU02 9.29 37 2019-09-12 00:00:00.000 STU02 9.94 37 2019-09-13 00:00:00.000 STU02 2.00 37 2019-09-15 00:00:00.000

我很感激这个人对我尝试过的许多不同的事情的帮助。

** 使用表和数据更新 **

/****** Object:  Table [dbo].[DLI_TEST_DATE]    Script Date: 9/18/2019 3:50:50 PM ******/
SET ANSI_NULLS ON
GO

SET QUOTED_IDENTIFIER ON
GO

CREATE TABLE [dbo].[DLI_TEST_DATE](
    [EMPLOYEE_ID] [nvarchar](15) NULL,
    [REG_TOTAL] [float] NULL,
    [WEEK_NUM] [int] NULL,
    [CDATE] [datetime] NULL,
    [DAYOFWK] [int] NULL
) ON [PRIMARY]

GO

INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 2, 35, CAST(N'2019-08-25 00:00:00.000' AS DateTime), 1)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 8.16, 35, CAST(N'2019-08-26 00:00:00.000' AS DateTime), 2)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 9.37, 35, CAST(N'2019-08-27 00:00:00.000' AS DateTime), 3)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 9.07, 35, CAST(N'2019-08-28 00:00:00.000' AS DateTime), 4)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 7.91, 35, CAST(N'2019-08-29 00:00:00.000' AS DateTime), 5)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 9.12, 35, CAST(N'2019-08-30 00:00:00.000' AS DateTime), 6)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 2.65, 35, CAST(N'2019-08-31 00:00:00.000' AS DateTime), 7)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 2, 36, CAST(N'2019-09-01 00:00:00.000' AS DateTime), 1)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 4.17, 36, CAST(N'2019-09-02 00:00:00.000' AS DateTime), 2)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 9.4, 36, CAST(N'2019-09-03 00:00:00.000' AS DateTime), 3)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 8.8, 36, CAST(N'2019-09-04 00:00:00.000' AS DateTime), 4)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 8.9, 36, CAST(N'2019-09-05 00:00:00.000' AS DateTime), 5)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 8.93, 36, CAST(N'2019-09-06 00:00:00.000' AS DateTime), 6)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 2.56, 36, CAST(N'2019-09-07 00:00:00.000' AS DateTime), 7)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 2, 37, CAST(N'2019-09-08 00:00:00.000' AS DateTime), 1)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 8.66, 37, CAST(N'2019-09-09 00:00:00.000' AS DateTime), 2)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 9.14, 37, CAST(N'2019-09-10 00:00:00.000' AS DateTime), 3)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 9.07, 37, CAST(N'2019-09-11 00:00:00.000' AS DateTime), 4)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 9.29, 37, CAST(N'2019-09-12 00:00:00.000' AS DateTime), 5)
GO
INSERT [dbo].[DLI_TEST_DATE] ([EMPLOYEE_ID], [REG_TOTAL], [WEEK_NUM], [CDATE], [DAYOFWK]) VALUES (N'STU02', 9.94, 37, CAST(N'2019-09-13 00:00:00.000' AS DateTime), 6)
GO

Answer 1

如果适用，子查询将给定一周的小时数分成 2 个不同的支付期。 where 条款将仅包括加班时间，并且分为两个支付期。 选择包括加班的数学。

declare @DLI_TEST_DATA TABLE (
    [EMPLOYEE_ID] [nvarchar](15) NULL,
    [REG_TOTAL] [float] NULL,
    [WEEK_NUM] [int] NULL,
    [CDATE] [datetime] NULL,
    [DAYOFWK] [int] NULL
) 

INSERT into @DLI_TEST_DATA
    VALUES (N'STU02', 2, 34, CAST(N'2019-08-25 00:00:00.000' AS DateTime), 1)
        ,(N'STU02', 8.16, 35, CAST(N'2019-08-26 00:00:00.000' AS DateTime), 2)
        ,(N'STU02', 9.37, 35, CAST(N'2019-08-27 00:00:00.000' AS DateTime), 3)
        ,(N'STU02', 9.07, 35, CAST(N'2019-08-28 00:00:00.000' AS DateTime), 4)
        ,(N'STU02', 7.91, 35, CAST(N'2019-08-29 00:00:00.000' AS DateTime), 5)
        ,(N'STU02', 9.12, 35, CAST(N'2019-08-30 00:00:00.000' AS DateTime), 6)
        ,(N'STU02', 2.65, 35, CAST(N'2019-08-31 00:00:00.000' AS DateTime), 7)
        ,(N'STU02', 2, 35, CAST(N'2019-09-01 00:00:00.000' AS DateTime), 1)
        ,(N'STU02', 4.17, 36, CAST(N'2019-09-02 00:00:00.000' AS DateTime), 2)
        ,(N'STU02', 9.4, 36, CAST(N'2019-09-03 00:00:00.000' AS DateTime), 3)
        ,(N'STU02', 8.8, 36, CAST(N'2019-09-04 00:00:00.000' AS DateTime), 4)
        ,(N'STU02', 8.9, 36, CAST(N'2019-09-05 00:00:00.000' AS DateTime), 5)
        ,(N'STU02', 8.93, 36, CAST(N'2019-09-06 00:00:00.000' AS DateTime), 6)
        ,(N'STU02', 2.56, 36, CAST(N'2019-09-07 00:00:00.000' AS DateTime), 7)
        ,(N'STU02', 2, 36, CAST(N'2019-09-08 00:00:00.000' AS DateTime), 1)
        ,(N'STU02', 8.66, 37, CAST(N'2019-09-09 00:00:00.000' AS DateTime), 2)
        ,(N'STU02', 9.14, 37, CAST(N'2019-09-10 00:00:00.000' AS DateTime), 3)
        ,(N'STU02', 9.07, 37, CAST(N'2019-09-11 00:00:00.000' AS DateTime), 4)
        ,(N'STU02', 9.29, 37, CAST(N'2019-09-12 00:00:00.000' AS DateTime), 5)
        ,(N'STU02', 9.94, 37, CAST(N'2019-09-13 00:00:00.000' AS DateTime), 6)

select WEEK_NUM,fullweek - (case when endfirstperiod+endsecondperiod <= 40 then 40.0 else endfirstperiod+endsecondperiod end) OvertimeCarriedOver
from (
    select week_num,sum(reg_total) fullweek
        ,sum(case when day(cdate) between 10 and 15 then reg_total else 0 end) endfirstperiod
        ,sum(case when day(cdate) between 16 and 21 then reg_total else 0 end) beginsecondperiod
        ,sum(case when day(cdate) between day(eomonth(cdate)) - 5 and day(eomonth(cdate)) then reg_total else 0 end) endsecondperiod
        ,sum(case when day(cdate) between 1 and 6 then reg_total else 0 end) beginfirstperiod
    from @DLI_TEST_DATA
    group by week_num
) basic
where fullweek > 40.0
    and beginfirstperiod+beginsecondperiod > 0
    and endfirstperiod+endsecondperiod > 0
order by week_num

Answer 2

核心思想是将所有内容按逻辑周分组，然后查看相关支付期与该周第一天不匹配的日期。

carryover_max 是那些日子的小时数。根据第 40 小时的工作时间，该数字可能过高。在最终输出中，它以该周的总加班时间为上限。

with weeks as (
    select week_start, week_end,
        case when sum(REG_TOTAL) > 40
             then sum(REG_TOTAL) - 40 else 0 end as overtime_total,
        case when sum(REG_TOTAL) > 40
             then sum(case when period <> week_period then REG_TOTAL else 0 end)
             else 0 end carryover_max
    from
        dbo.DLI_TEST_DATE cross apply (
            select
                dateadd(day, -datepart(weekday, dateadd(day, -1, cdate)) + 1, cdate)
        ) as v(week_start) cross apply (
            select
                dateadd(day, 6, week_start),
                case when datepart(day, week_start) <= 15 then 1 else 2 end,
                case when datepart(day, cdate) <= 15 then 1 else 2 end
        ) as v2(week_end, week_period, period)
    group by week_start, week_end
), periods as (
    select distinct period_start, period_end
    from
        dbo.DLI_TEST_DATE cross apply (
           select
                datefromparts(datepart(year, cdate), datepart(month, cdate),
                    case when datepart(day, cdate) <= 15 then 1 else 15 end),
                datefromparts(datepart(year, cdate), datepart(month, cdate),
                    case when datepart(day, cdate) <= 15 then 16 else datepart(day, eomonth(cdate)) end)
        ) v(period_start, period_end)
)
select period_start,
    sum(case when carryover_max > overtime_total then overtime_total else carryover_max end) as overtime_owed
from
    periods inner join
        weeks w on w.week_end >= period_start and w.week_end <= period_end
group by period_start;

https://rextester.com/TVXF30798

顺便说一句，您真的不想使用float 作为您的数据类型。

由于有关于周编号的问题，我刚刚计算了日期。无论如何，这在新年的几周内应该会更好。我还假设您的服务器设置在使用 datepart(weekday...) 时将星期日作为一周的第一天。