合并我的数据透视表查询中的行与多个数据透视表
以下SQL查询应该显示教师可用性。有4种可能的预订类型 - 上午，下午，全天和每小时。如果有AM预订，单元格中的文本应显示PM，如果有PM预订，则应显示AM，如果有全天预订，或者AM和PM预订，则应显示'xxx'。每小时预订 - 可以安全地假设每天只有2小时预订，AM为1（开始时间<= 12pm），PM为1（结束时间> 12pm），这意味着我们应该显示'XXX'。一切正常。
当有AM预订然后是PM每小时预订，或PM预订时，每小时预订时，我遇到的问题是显示'XXX'。
WITH Bookings AS
( SELECT TeacherID,
[WeekDay] = DATENAME(WEEKDAY, BookingDate),
[0], [1], [2], [3],
[Status] = CASE
WHEN ([0] > 0 AND [1] > 0) THEN 'XXX'
WHEN [2] > 0 THEN 'XXX'
WHEN [0] > 0 THEN 'PM'
WHEN [1] > 0 THEN 'AM'
WHEN [3] > 0 AND StartTime <= CONVERT(TIME, '12:00:00') AND EndTime >= CONVERT(TIME, '12:00:00') THEN 'XXX'
WHEN [3] > 0 AND EndTime <= CONVERT(TIME, '12:00:00') THEN 'PM'
WHEN [3] > 0 AND StartTime >= CONVERT(TIME, '12:00:00') THEN 'AM'
END
FROM ( SELECT TeacherID,
BookingDate,
BookingDuration,
StartTime = CASE WHEN BookingDuration = 3 THEN CAST(MIN(StartTime) OVER(PARTITION BY TeacherID, BookingDate, BookingDuration) AS TIME) ELSE NULL END,
EndTime = CASE WHEN BookingDuration = 3 THEN CAST(MAX(EndTime) OVER(PARTITION BY TeacherID, BookingDate, BookingDuration) AS TIME) ELSE NULL END,
[x] = 1
FROM BookingDays
WHERE (Status = 0 OR Status IS NULL)
) BookingDays
PIVOT
( SUM(x)
FOR BookingDuration IN ([0], [1], [2], [3])
) pvt
WHERE BookingDate >= DATEADD(ww, DATEDIFF(ww,0,'05/06/2013'), 0) AND BookingDate <= DATEADD(ww, DATEDIFF(ww,0,'05/06/2013'), 6)
), PivotedBookings AS
( SELECT *
FROM Bookings
PIVOT
( MAX([Status])
FOR [WeekDay] IN ([Monday], [Tuesday], [Wednesday], [Thursday], [Friday])
) pvt
)
SELECT ID,Firstname,Surname,Band,'£' + CONVERT(varchar(50),DefaultChargeRateDaily) + '/' + '£' + CONVERT(varchar(50), DefaultPayRateDaily) as 'BandRates',Telephone,Mobile,Teacher,TeacherAssistant,KeyStage,MAX(Monday) Monday,MAX(Tuesday) Tuesday,MAX(Wednesday) Wednesday,MAX(Thursday) Thursday,MAX(Friday) Friday, Notes
FROM (
SELECT t.ID,
t.Firstname,
t.Surname,
tb.Band,
t.DefaultChargeRateDaily,
t.DefaultPayRateDaily,
t.Telephone,
t.Mobile,
t.Teacher,
t.TeacherAssistant,
CASE WHEN t.Nursery > 0 THEN 'NUR' WHEN t.Reception > 0 THEN 'REC' WHEN t.Year1 > 0 THEN 'Y1' WHEN t.Year2 > 0 THEN 'Y2' WHEN t.Year3 > 0 THEN 'Y3' WHEN t.Year4 > 0 THEN 'Y4' WHEN t.Year5 > 0 THEN 'Y5' WHEN t.Year6 > 0 THEN 'Y6' WHEN t.Year7 > 0 THEN 'Y7' WHEN t.Year8 > 0 THEN 'Y8' WHEN t.Year9 > 0 THEN 'Y9' WHEN t.Year10 > 0 THEN 'Y10' WHEN t.Year11 > 0 THEN 'Y11' WHEN t.ALevel > 0 THEN 'ALevel' END + ' - ' + CASE WHEN t.ALevel > 0 THEN 'ALevel' WHEN t.Year11 > 0 THEN 'Y11' WHEN t.Year10 > 0 THEN 'Y10' WHEN t.Year9 > 0 THEN 'Y9' WHEN t.Year8 > 0 THEN 'Y8' WHEN t.Year7 > 0 THEN 'Y7' WHEN t.Year6 > 0 THEN 'Y6' WHEN t.Year5 > 0 THEN 'Y5' WHEN t.Year4 > 0 THEN 'Y4' WHEN t.Year3 > 0 THEN 'Y3' WHEN t.Year2 > 0 THEN 'Y2' WHEN t.Year1 > 0 THEN 'Y1' WHEN t.Reception > 0 THEN 'REC' WHEN t.Nursery > 0 THEN 'NUR' ELSE '' END as 'KeyStage',
Monday = CASE WHEN an.Date = DATEADD(ww, DATEDIFF(ww,0,'05/06/2013'), 0) AND an.TeacherID = t.ID THEN an.Text WHEN t.Status = 0 THEN 'XXX' ELSE COALESCE(pb.Monday, '') END,
Tuesday = CASE WHEN an.Date = DATEADD(ww, DATEDIFF(ww,0,'05/06/2013'), 1) AND an.TeacherID = t.ID THEN an.Text WHEN t.Status = 0 THEN 'XXX' ELSE COALESCE(pb.Tuesday, '') END,
Wednesday = CASE WHEN an.Date = DATEADD(ww, DATEDIFF(ww,0,'05/06/2013'), 2) AND an.TeacherID = t.ID THEN an.Text WHEN t.Status = 0 THEN 'XXX' ELSE COALESCE(pb.Wednesday, '') END,
Thursday = CASE WHEN an.Date = DATEADD(ww, DATEDIFF(ww,0,'05/06/2013'), 3) AND an.TeacherID = t.ID THEN an.Text WHEN t.Status = 0 THEN 'XXX' ELSE COALESCE(pb.Thursday, '') END,
Friday = CASE WHEN an.Date = DATEADD(ww, DATEDIFF(ww,0,'05/06/2013'), 4) AND an.TeacherID = t.ID THEN an.Text WHEN t.Status = 0 THEN 'XXX' ELSE COALESCE(pb.Friday, '') END,
Notes
FROM Teachers t
LEFT JOIN PivotedBookings pb
ON pb.TeacherID = t.ID
LEFT JOIN TeacherBands tb
ON tb.ID = t.Band
LEFT JOIN AvailabilityNotes an
ON t.ID = an.TeacherID
WHERE t.Active = 0 and (t.Status = 1 or t.Status = 0) and t.PrimarySchool = 1 and t.ID = 9094
) T1
GROUP BY ID,Firstname,Surname,Telephone,Mobile,Teacher,TeacherAssistant,KeyStage,Notes,DefaultChargeRateDaily,DefaultPayRateDaily,Band
ORDER BY Surname,Firstname asc
第一部分生成以下内容 -
SELECT TeacherID,
[WeekDay] = DATENAME(WEEKDAY, BookingDate),
[0], [1], [2], [3],
[Status] = CASE
WHEN ([0] > 0 AND [1] > 0) THEN 'XXX'
WHEN [2] > 0 THEN 'XXX'
WHEN [0] > 0 THEN 'PM'
WHEN [1] > 0 THEN 'AM'
WHEN ([3] > 0 AND EndTime <= CONVERT(TIME, '12:00:00')) AND ([0] > 1) THEN 'XXX'
WHEN ([3] > 0 AND EndTime <= CONVERT(TIME, '12:00:00')) AND ([1] > 1) THEN 'XXX'
WHEN [3] > 0 AND StartTime <= CONVERT(TIME, '12:00:00') AND EndTime >= CONVERT(TIME, '12:00:00') THEN 'XXX'
WHEN [3] > 0 AND EndTime <= CONVERT(TIME, '12:00:00') THEN 'PM'
WHEN [3] > 0 AND StartTime >= CONVERT(TIME, '12:00:00') THEN 'AM'
END
FROM ( SELECT TeacherID,
BookingDate,
BookingDuration,
StartTime = CASE WHEN BookingDuration = 3 THEN CAST(MIN(StartTime) OVER(PARTITION BY TeacherID, BookingDate, BookingDuration) AS TIME) ELSE NULL END,
EndTime = CASE WHEN BookingDuration = 3 THEN CAST(MAX(EndTime) OVER(PARTITION BY TeacherID, BookingDate, BookingDuration) AS TIME) ELSE NULL END,
[x] = 1
FROM BookingDays
WHERE (Status = 0 OR Status IS NULL)
) BookingDays
PIVOT
( SUM(x)
FOR BookingDuration IN ([0], [1], [2], [3])
) pvt
WHERE BookingDate >= DATEADD(ww, DATEDIFF(ww,0,'05/06/2013'), 0) AND BookingDate <= DATEADD(ww, DATEDIFF(ww,0,'05/06/2013'), 6)
TeacherID | WeekDay | 0 | 1 | 2 | 3 | Status
9094 | Monday | NULL | NULL | 1 | NULL | XXX
9094 | Tuesday | NULL | NULL | NULL | 1 | AM
9094 | Wednesday| NULL | 1 | NULL | NULL | AM
9094 | Thursday | NULL | 1 | NULL | NULL | AM
9094 | Thursday | NULL | NULL | NULL | 1 | PM
9094 | Friday | NULL | NULL | 1 | 1 | XXX
预订时间 -
0 - AM 下午1点 2 - 全天 3 - 每小时
我们可以在这里看到，两个星期四的行需要合并并在状态列中显示为XXX，而不是2个单独的行。
下面的屏幕截图显示了一个示例。突出显示黄色，它显示PM，但它应显示XXX，因为每小时有1 PM和AM。
如果每小时，每小时，每小时和下午，我怎样才能显示XXX？
谢谢！
1 个答案:
答案 0 :(得分：2)
我会将复杂逻辑拆分为两个步骤，使用IntermediateBookings CTE将逻辑减少到两列has_am和has_pm。这些更容易计算，从这些，最终输出也很容易计算。理论上，您可以只改进原始语句的CASE表达式以包含更多AND和OR s，但这不是很容易维护。查询的开头应该是这样的：
WITH IntermediateBooking as (
SELECT TeacherID,
[WeekDay] = DATENAME(WEEKDAY, BookingDate),
[0], [1], [2], [3],
has_am = case when
[1] > 0
OR [2] > 0
OR ([3] > 0 AND StartTime <= CONVERT(TIME, '12:00:00'))
then 1
else null
end,
has_pm = case when
[0] > 0
OR [2] > 0
OR ([3] > 0 AND EndTime >= CONVERT(TIME, '12:00:00'))
then 1
else null
end
FROM ( SELECT TeacherID,
BookingDate,
BookingDuration,
StartTime = CASE WHEN BookingDuration = 3 THEN CAST(MIN(StartTime) OVER(PARTITION BY TeacherID, BookingDate, BookingDuration) AS TIME) ELSE NULL END,
EndTime = CASE WHEN BookingDuration = 3 THEN CAST(MAX(EndTime) OVER(PARTITION BY TeacherID, BookingDate, BookingDuration) AS TIME) ELSE NULL END,
[x] = 1
FROM BookingDays
WHERE (Status = 0 OR Status IS NULL)
) BookingDays
PIVOT
( SUM(x)
FOR BookingDuration IN ([0], [1], [2], [3])
) pvt
WHERE BookingDate >= DATEADD(ww, DATEDIFF(ww,0,'05/06/2013'), 0) AND BookingDate <= DATEADD(ww, DATEDIFF(ww,0,'05/06/2013'), 6)
),
Bookings AS (
SELECT TeacherID, [WeekDay],
case when sum(has_am) > 0 and sum(has_pm) > 0 then 'XXX'
WHEN sum(has_am) > 0 then 'AM'
WHEN sum(has_pm) > 0 then 'PM'
end as [Status]
FROM IntermediateBookings
GROUP BY TeacherID, [WeekDay]
)
其余的 - 从PivotedBookings开始 - 可以保持原样。
但是，由于我没有可用的表，因此无法检查语法错误以及结果是否符合要求。可能需要进行一些调整。
相关问题
最新问题
- 我写了这段代码，但我无法理解我的错误
- 我无法从一个代码实例的列表中删除 None 值，但我可以在另一个实例中。为什么它适用于一个细分市场而不适用于另一个细分市场？
- 是否有可能使 loadstring 不可能等于打印？卢阿
- java中的random.expovariate()
- Appscript 通过会议在 Google 日历中发送电子邮件和创建活动
- 为什么我的 Onclick 箭头功能在 React 中不起作用？
- 在此代码中是否有使用“this”的替代方法？
- 在 SQL Server 和 PostgreSQL 上查询，我如何从第一个表获得第二个表的可视化
- 每千个数字得到
- 更新了城市边界 KML 文件的来源？