我在这篇文章中回答了这个问题，但已被建议重新提出这个问题：
Group Non-Contiguous Dates By Criteria In Column
但是，我注意到虽然解决方案大部分时间都有效，但是有一些错误可以解决它似乎无法正确组合的问题。
例如，
数据：
DECLARE @TempTable TABLE([CUSTOMER_ID] INT
,[TEAM] VARCHAR(1)
,[TYPE] VARCHAR(1)
,[START_DATE] DATETIME
,[END_DATE] DATETIME
,[GROUP_DAYS_CRITERIA] INT)
INSERT INTO @TempTable VALUES (1,'A','A','2013-08-07','2013-12-31',28)
,(2,'B','A','2015-05-15','2015-05-28',28)
,(2,'B','A','2015-05-15','2016-05-12',28)
,(2,'B','A','2015-05-28','2015-05-28',28)
,(3,'C','A','2013-05-27','2014-07-23',28)
,(3,'C','A','2015-01-12','2015-05-28',28)
,(3,'B','A','2015-01-12','2015-05-28',28)
,(3,'C','A','2015-05-28','2015-05-28',28)
,(3,'C','A','2015-05-28','2015-12-17',28)
,(4,'A','B','2013-07-09','2014-04-21',7)
,(4,'A','B','2014-04-29','2014-08-01',7)
,(5,'A','A','2014-05-15','2015-04-24',28)
,(5,'A','A','2014-05-15','2015-04-24',28)
,(5,'A','A','2014-05-15','2014-05-15',28)
,(5,'A','A','2015-04-24','2015-05-13',28)
,(5,'A','B','2014-05-15','2014-05-15',7)
,(5,'A','B','2014-06-13','2015-04-24',7)
,(5,'A','B','2014-06-13','2015-04-24',7)
,(5,'A','B','2015-04-24','2015-05-13',7)
,(6,'A','A','2015-02-17','2015-04-28',28)
,(6,'A','A','2015-02-17','2015-04-28',28)
,(6,'A','A','2015-04-10','2015-04-28',28)
,(6,'A','A','2015-04-10','2015-04-28',28)
,(6,'A','A','2015-04-28','2015-06-04',28)
,(6,'A','A','2015-04-28','2015-08-03',28)
,(6,'A','A','2015-05-22','2015-08-03',28)
,(7,'A','A','2015-03-30','2015-04-28',28)
,(7,'A','A','2015-03-30','2015-04-28',28)
,(7,'A','A','2015-03-30','2015-04-28',28)
,(7,'A','A','2015-03-30','2015-04-28',28)
,(7,'A','A','2015-04-28','2015-11-17',28)
,(7,'A','A','2015-04-28','2015-11-17',28)
,(7,'A','A','2015-05-12','2015-11-17',28)
,(7,'A','A','2015-05-12','2015-11-17',28)
看起来像这样：
+-------------+------+------+------------+------------+---------------------+
| CUSTOMER_ID | TEAM | TYPE | START_DATE | END_DATE | GROUP_DAYS_CRITERIA |
+-------------+------+------+------------+------------+---------------------+
| 1 | A | A | 07/08/2013 | 31/12/2013 | 28 |
| 2 | B | A | 15/05/2015 | 28/05/2015 | 28 |
| 2 | B | A | 15/05/2015 | 12/05/2016 | 28 |
| 2 | B | A | 28/05/2015 | 28/05/2015 | 28 |
| 3 | C | A | 27/05/2013 | 23/07/2014 | 28 |
| 3 | C | A | 12/01/2015 | 28/05/2015 | 28 |
| 3 | B | A | 12/01/2015 | 28/05/2015 | 28 |
| 3 | C | A | 28/05/2015 | 28/05/2015 | 28 |
| 3 | C | A | 28/05/2015 | 17/12/2015 | 28 |
| 4 | A | B | 09/07/2013 | 21/04/2014 | 7 |
| 4 | A | B | 29/04/2014 | 01/08/2014 | 7 |
| 5 | A | A | 15/05/2014 | 24/04/2015 | 28 |
| 5 | A | A | 15/05/2014 | 24/04/2015 | 28 |
| 5 | A | A | 15/05/2014 | 15/05/2014 | 28 |
| 5 | A | A | 24/04/2015 | 13/05/2015 | 28 |
| 5 | A | B | 15/05/2014 | 15/05/2014 | 7 |
| 5 | A | B | 13/06/2014 | 24/04/2015 | 7 |
| 5 | A | B | 13/06/2014 | 24/04/2015 | 7 |
| 5 | A | B | 24/04/2015 | 13/05/2015 | 7 |
| 6 | A | A | 17/02/2015 | 28/04/2015 | 28 |
| 6 | A | A | 17/02/2015 | 28/04/2015 | 28 |
| 6 | A | A | 10/04/2015 | 28/04/2015 | 28 |
| 6 | A | A | 10/04/2015 | 28/04/2015 | 28 |
| 6 | A | A | 28/04/2015 | 04/06/2015 | 28 |
| 6 | A | A | 28/04/2015 | 03/08/2015 | 28 |
| 6 | A | A | 22/05/2015 | 03/08/2015 | 28 |
| 7 | A | A | 30/03/2015 | 28/04/2015 | 28 |
| 7 | A | A | 30/03/2015 | 28/04/2015 | 28 |
| 7 | A | A | 30/03/2015 | 28/04/2015 | 28 |
| 7 | A | A | 30/03/2015 | 28/04/2015 | 28 |
| 7 | A | A | 28/04/2015 | 17/11/2015 | 28 |
| 7 | A | A | 28/04/2015 | 17/11/2015 | 28 |
| 7 | A | A | 12/05/2015 | 17/11/2015 | 28 |
| 7 | A | A | 12/05/2015 | 17/11/2015 | 28 |
+-------------+------+------+------------+------------+---------------------+
目前这样出现：
+-------------+------+------+------------+------------+---------------------+
| Customer_Id | Team | Type | Start_Date | End_Date | Group_Days_Criteria |
+-------------+------+------+------------+------------+---------------------+
| 1 | A | A | 07/08/2013 | 31/12/2013 | 28 |
| 2 | B | A | 15/05/2015 | 12/05/2016 | 28 |
| 3 | B | A | 12/01/2015 | 28/05/2015 | 28 |
| 3 | C | A | 27/05/2013 | 23/07/2014 | 28 |
| 3 | C | A | 12/01/2015 | 28/05/2015 | 28 |
| 4 | A | B | 09/07/2013 | 21/04/2014 | 7 |
| 4 | A | B | 29/04/2014 | 01/08/2014 | 7 |
| 5 | A | A | 15/05/2014 | 24/04/2015 | 28 |
| 5 | A | B | 15/05/2014 | 15/05/2014 | 7 |
| 5 | A | B | 13/06/2014 | 24/04/2015 | 7 |
| 5 | A | A | 24/04/2015 | 13/05/2015 | 28 |
| 6 | A | A | 17/02/2015 | 28/04/2015 | 28 |
| 7 | A | A | 30/03/2015 | 28/04/2015 | 28 |
+-------------+------+------+------------+------------+---------------------+
我需要它像这样出来：
+-------------+------+------+------------+------------+---------------------+
| CUSTOMER_ID | TEAM | TYPE | START_DATE | END_DATE | GROUP_DAYS_CRITERIA |
+-------------+------+------+------------+------------+---------------------+
| 1 | A | A | 07/08/2013 | 31/12/2013 | 28 |
| 2 | B | A | 15/05/2015 | 28/05/2015 | 28 |
| 3 | C | A | 27/05/2013 | 23/07/2014 | 28 |
| 3 | C | A | 12/01/2015 | 17/12/2015 | 28 |
| 3 | B | A | 12/01/2015 | 28/05/2015 | 28 |
| 4 | A | B | 09/07/2013 | 21/04/2014 | 7 |
| 4 | A | B | 29/04/2014 | 01/08/2014 | 7 |
| 5 | A | A | 15/05/2014 | 13/05/2015 | 28 |
| 5 | A | B | 15/05/2014 | 15/05/2014 | 7 |
| 5 | A | B | 13/06/2014 | 13/05/2015 | 7 |
| 6 | A | A | 17/02/2015 | 03/08/2015 | 28 |
| 7 | A | A | 30/03/2015 | 17/11/2015 | 28 |
+-------------+------+------+------------+------------+---------------------+
关于如何在保持正确输出标准的同时解决这个问题的任何想法？
丹尼尔
答案 0 :(得分：2)
Select A.Customer_ID
,A.Team
,A.Type
,Group_Days_Criteria
,Start_Date = B.MinD
,End_Date = max(B.MaxD)
From @TempTable A
Cross Apply (
Select MinD=min(Start_Date),MaxD=Max(End_Date)
From @TempTable
Where (Start_Date between A.Start_Date and A.End_Date or
End_Date between A.Start_Date and A.End_Date)
and Customer_ID = A.Customer_ID
and Team = A.Team
and Type = A.TYPE
and Group_Days_Criteria = A.Group_Days_Criteria
) B
Group By
A.Customer_ID
,A.Team
,A.Type
,A.Group_Days_Criteria
,B.MinD
返回
Customer_ID Team Type Group_Days_Criteria Start_Date End_Date
1 A A 28 2014-05-15 2015-05-13
1 A B 7 2014-05-15 2014-05-15
1 A B 7 2014-06-13 2015-05-13
答案 1 :(得分：1)
这是一个比我刚开始玩它时所假设的更复杂的问题。我不得不提出一种完全不同的做事方式，而不是我之前的回答。该答案将回顾一两行来比较日期。这是不够的，因为我们的磋商范围不规则。最后，正如我们所看到的，前一个答案的查询在现实世界中非常脆弱。
我们需要针对每个客户，团队和类型分组进行每次咨询，并比较我们的咨询日期。该解决方案通过分组将每次咨询与第一次咨询进行比较。如果手头的行在该咨询的日期范围内，它会更新第一次咨询的结束日期。如果手头的行超出该日期范围，则插入一行。继续我们的循环，我们将不再考虑第一次咨询。将对我们插入的行进行未来的比较，直到协商不再属于第二个日期范围。
这在我的机器上完成了一秒钟。
我添加了一个新的答案，这样你就可以看到我的第一道思路从一开始就不起作用。请让我知道，如果你有任何问题。
设置和返回结果：
declare @TempTable table
(
[CUSTOMER_ID] INT
,[TEAM] VARCHAR(1)
,[TYPE] VARCHAR(1)
,[START_DATE] DATETIME
,[END_DATE] DATETIME
,[GROUP_DAYS_CRITERIA] INT
)
INSERT INTO @tempTable
VALUES
(1,'A','A','2013-08-07','2013-12-31',28)
,(2,'B','A','2015-05-15','2015-05-28',28)
,(2,'B','A','2015-05-15','2016-05-12',28)
,(2,'B','A','2015-05-28','2015-05-28',28)
,(3,'C','A','2013-05-27','2014-07-23',28)
,(3,'C','A','2015-01-12','2015-05-28',28)
,(3,'B','A','2015-01-12','2015-05-28',28)
,(3,'C','A','2015-05-28','2015-05-28',28)
,(3,'C','A','2015-05-28','2015-12-17',28)
,(4,'A','B','2013-07-09','2014-04-21',7)
,(4,'A','B','2014-04-29','2014-08-01',7)
,(5,'A','A','2014-05-15','2015-04-24',28)
,(5,'A','A','2014-05-15','2015-04-24',28)
,(5,'A','A','2014-05-15','2014-05-15',28)
,(5,'A','A','2015-04-24','2015-05-13',28)
,(5,'A','B','2014-05-15','2014-05-15',7)
,(5,'A','B','2014-06-13','2015-04-24',7)
,(5,'A','B','2014-06-13','2015-04-24',7)
,(5,'A','B','2015-04-24','2015-05-13',7)
,(6,'A','A','2015-02-17','2015-04-28',28)
,(6,'A','A','2015-02-17','2015-04-28',28)
,(6,'A','A','2015-04-10','2015-04-28',28)
,(6,'A','A','2015-04-10','2015-04-28',28)
,(6,'A','A','2015-04-28','2015-06-04',28)
,(6,'A','A','2015-04-28','2015-08-03',28)
,(6,'A','A','2015-05-22','2015-08-03',28)
,(7,'A','A','2015-03-30','2015-04-28',28)
,(7,'A','A','2015-03-30','2015-04-28',28)
,(7,'A','A','2015-03-30','2015-04-28',28)
,(7,'A','A','2015-03-30','2015-04-28',28)
,(7,'A','A','2015-04-28','2015-11-17',28)
,(7,'A','A','2015-04-28','2015-11-17',28)
,(7,'A','A','2015-05-12','2015-11-17',28)
,(7,'A','A','2015-05-12','2015-11-17',28)
-- renamed work table columns for ease of typing
-- note: try to avoid using reserved words (type) for column names
declare @prepTable table -- temp table with row number for loops, padded end date for comparison
(
[CustomerId] INT
,[Team] VARCHAR(1)
,[ConsultType] VARCHAR(1)
,[StartDate] DATETIME
,[EndDate] DATETIME
,[GroupDaysCriteria] INT
,ConsultEndDate datetime
,rn int
);
declare @ConsolidateConsults table -- consult work table
(
[CustomerId] INT
,[Team] VARCHAR(1)
,[ConsultType] VARCHAR(1)
,[StartDate] DATETIME
,[EndDate] DATETIME
,[GroupDaysCriteria] INT
,ConsultEndDate datetime
,ConsultGroup int
);
declare @CustomerTeamType table -- temp lookup table for loops
(
[CustomerId] INT
,CustomerNumber int
,[Team] VARCHAR(1)
,TeamId int
,[ConsultType] VARCHAR(1)
,ConsultTypeId int
,ConsultCount int
);
insert into @CustomerTeamType
select Customer_Id,
null,
Team,
null as TeamId,
[Type],
null as ConsultTypeId,
count(*)
from @TempTable
group by Customer_Id, Team, [Type];
update c
set CustomerNumber = ctt.CustomerNumber,
TeamId = ctt.TeamId,
ConsultTypeId = ctt.ConsultTypeId
from @CustomerTeamType c
inner join
(
select ct.CustomerId,
ct.Team,
ct.ConsultType,
dense_rank() over (partition by 1 order by ct.CustomerId) as CustomerNumber,
dense_rank() over (partition by ct.CustomerId order by ct.Team) as TeamId,
dense_rank() over (partition by ct.CustomerId, ct.Team order by ct.ConsultType) as ConsultTypeId
from @CustomerTeamType ct
) as ctt
on c.CustomerId = ctt.CustomerId
and c.Team = ctt.Team
and c.ConsultType = ctt.ConsultType;
insert into @prepTable
select Customer_Id,
Team,
[Type],
[Start_Date],
[End_Date],
Group_Days_Criteria,
dateadd(day, Group_Days_Criteria, [End_Date]),
row_number() over (partition by Customer_Id, Team, [Type] order by [Start_Date] asc, [End_Date] desc)
from @tempTable;
insert into @ConsolidateConsults
select CustomerId,
Team,
ConsultType,
StartDate,
EndDate,
GroupDaysCriteria,
ConsultEndDate,
1 -- rn of one indicates a sure consult begin date
from @prepTable
where rn = 1;
--prepare loop counters and end points
declare @consultGroup int = 1;
declare @customer int = 1;
declare @team int = 1;
declare @consultType int = 1;
declare @maxCustomer int = (select max(CustomerNumber) from @CustomerTeamType);
declare @maxTeam int;
declare @maxConsultType int;
declare @totalConsults int;
-- loop through each consult in each consult type in each consult team for each customer
while @customer <= @maxCustomer
begin
-- Get total conult teams for customer x
set @maxTeam = (select max(TeamId)
from @CustomerTeamType
where CustomerNumber = @customer
);
while @team <= @maxTeam
begin
-- get total consult types for customer x, team y
set @maxConsultType = (select max(ConsultTypeId) from @CustomerTeamType where CustomerNumber = @customer and TeamId = @team)
while @consultType <= @maxConsultType
begin
-- get total consults for customer x, team y, consult type z
set @totalConsults = (select ConsultCount from @CustomerTeamType where CustomerNumber = @customer and TeamId = @team and @consultType = ConsultTypeId)
-- reset to first consult
declare @consult int = 1
while @consult <= @totalConsults
begin
declare @row table ( -- table to hold one consult at a time
CustomerId int,
Team varchar(1),
ConsultType varchar(1),
StartDate datetime,
EndDate datetime,
GroupDaysCriteria int,
ConsultEndDate datetime,
ConsultGroup int
)
insert into @row
select pt.CustomerId,
pt.Team,
pt.ConsultType,
pt.StartDate,
pt.EndDate,
pt.GroupDaysCriteria,
pt.ConsultEndDate,
@consultGroup
from @prepTable pt
inner join @CustomerTeamType ctt
on pt.CustomerId = ctt.CustomerId
and pt.Team = ctt.Team
and pt.ConsultType = ctt.ConsultType
and ctt.CustomerNumber = @customer
and ctt.ConsultTypeId = @consultType
and ctt.TeamId = @team
and pt.rn = @consult
-- if row at hand falls within previous consult, update end date to greater of end dates
if exists (
select * from @row r
inner join @ConsolidateConsults cc
on r.CustomerId = cc.CustomerId
and r.ConsultType = cc.ConsultType
and r.Team = cc.Team
and cc.ConsultGroup = @consultGroup
and r.StartDate >= cc.StartDate
and r.StartDate <= cc.ConsultEndDate
)
Begin
update cc
set cc.ConsultEndDate = case when cc.ConsultEndDate > r.ConsultEndDate then cc.ConsultEndDate else r.ConsultEndDate end,
cc.EndDate = case when cc.EndDate > r.EndDate then cc.EndDate else r.EndDate end
from @ConsolidateConsults cc
inner join @row r
on r.CustomerId = cc.CustomerId
and r.Team = cc.Team
and @consultGroup = cc.ConsultGroup
and r.ConsultType = cc.ConsultType
and cc.ConsultGroup = @consultGroup
--set @updates = @updates + 1
end
-- if row at hand falls after existing consultation, create a new row and grouping
else
begin
-- increment consult grouping so next loop considers proper consultation
set @consultGroup = @consultGroup + 1
insert into @ConsolidateConsults
select CustomerId,
Team,
ConsultType,
StartDate,
EndDate,
GroupDaysCriteria,
ConsultEndDate,
@consultGroup
from @row
end
set @consult = @consult + 1 -- move to next record
delete @row -- delete record we have just compared
end -- finished individual consultations for customer x, team y, consult type z
set @consultType = @consultType + 1 --move to next consult type
set @consultGroup = 1 -- reset conultation grouping
end -- finished comparing for consult type
set @consultType = 1 -- reset consult type
set @consultGroup = 1 -- reset conultation grouping
set @team = @team + 1 -- move to next team
end -- finished consult teams
set @team = 1 -- reset consult team
set @consultGroup = 1 -- reset conultation grouping
set @customer = @customer + 1 -- move to next customer
end -- finished customer
-- get output
select CustomerId as Customer_Id,
Team,
ConsultType as [Type],
StartDate as [Start_Date],
EndDate as [End_Date],
from @ConsolidateConsults
order by CustomerId, Team, ConsultType, StartDate
查询返回：
╔═════════════╦══════╦══════╦═════════════════════════╦═════════════════════════╦═════════════════════╗
║ Customer_Id ║ Team ║ Type ║ Start_Date ║ End_Date ║ Group_Days_Criteria ║
╠═════════════╬══════╬══════╬═════════════════════════╬═════════════════════════╬═════════════════════╣
║ 1 ║ A ║ A ║ 2013-08-07 00:00:00.000 ║ 2013-12-31 00:00:00.000 ║ 28 ║
║ 2 ║ B ║ A ║ 2015-05-15 00:00:00.000 ║ 2016-05-12 00:00:00.000 ║ 28 ║
║ 3 ║ B ║ A ║ 2015-01-12 00:00:00.000 ║ 2015-05-28 00:00:00.000 ║ 28 ║
║ 3 ║ C ║ A ║ 2013-05-27 00:00:00.000 ║ 2014-07-23 00:00:00.000 ║ 28 ║
║ 3 ║ C ║ A ║ 2015-01-12 00:00:00.000 ║ 2015-12-17 00:00:00.000 ║ 28 ║
║ 4 ║ A ║ B ║ 2013-07-09 00:00:00.000 ║ 2014-04-21 00:00:00.000 ║ 7 ║
║ 4 ║ A ║ B ║ 2014-04-29 00:00:00.000 ║ 2014-08-01 00:00:00.000 ║ 7 ║
║ 5 ║ A ║ A ║ 2014-05-15 00:00:00.000 ║ 2015-05-13 00:00:00.000 ║ 28 ║
║ 5 ║ A ║ B ║ 2014-05-15 00:00:00.000 ║ 2014-05-15 00:00:00.000 ║ 7 ║
║ 5 ║ A ║ B ║ 2014-06-13 00:00:00.000 ║ 2015-05-13 00:00:00.000 ║ 7 ║
║ 6 ║ A ║ A ║ 2015-02-17 00:00:00.000 ║ 2015-08-03 00:00:00.000 ║ 28 ║
║ 7 ║ A ║ A ║ 2015-03-30 00:00:00.000 ║ 2015-11-17 00:00:00.000 ║ 28 ║
╚═════════════╩══════╩══════╩═════════════════════════╩═════════════════════════╩═════════════════════╝
答案 2 :(得分：0)
我想我已经使用套装破解了它（虽然很麻烦）......
DECLARE @TempTable TABLE([CUSTOMER_ID] INT
,[TEAM] VARCHAR(1)
,[TYPE] VARCHAR(1)
,[START_DATE] DATETIME
,[END_DATE] DATETIME
,[GROUP_DAYS_CRITERIA] INT)
INSERT INTO @TempTable VALUES (1,'A','A','2013-08-07','2013-12-31',28)
,(2,'B','A','2015-05-15','2015-05-28',28)
,(2,'B','A','2015-05-15','2016-05-12',28)
,(2,'B','A','2015-05-28','2015-05-28',28)
,(3,'C','A','2013-05-27','2014-07-23',28)
,(3,'C','A','2015-01-12','2015-05-28',28)
,(3,'B','A','2015-01-12','2015-05-28',28)
,(3,'C','A','2015-05-28','2015-05-28',28)
,(3,'C','A','2015-05-28','2015-12-17',28)
,(4,'A','B','2013-07-09','2014-04-21',7)
,(4,'A','B','2014-04-29','2014-08-01',7)
,(5,'A','A','2014-05-15','2015-04-24',28)
,(5,'A','A','2014-05-15','2015-04-24',28)
,(5,'A','A','2014-05-15','2014-05-15',28)
,(5,'A','A','2015-04-24','2015-05-13',28)
,(5,'A','B','2014-05-15','2014-05-15',7)
,(5,'A','B','2014-06-13','2015-04-24',7)
,(5,'A','B','2014-06-13','2015-04-24',7)
,(5,'A','B','2015-04-24','2015-05-13',7)
,(6,'A','A','2015-02-17','2015-04-28',28)
,(6,'A','A','2015-02-17','2015-04-28',28)
,(6,'A','A','2015-04-10','2015-04-28',28)
,(6,'A','A','2015-04-10','2015-04-28',28)
,(6,'A','A','2015-04-28','2015-06-04',28)
,(6,'A','A','2015-04-28','2015-08-03',28)
,(6,'A','A','2015-05-22','2015-08-03',28)
,(7,'A','A','2015-03-30','2015-04-28',28)
,(7,'A','A','2015-03-30','2015-04-28',28)
,(7,'A','A','2015-03-30','2015-04-28',28)
,(7,'A','A','2015-03-30','2015-04-28',28)
,(7,'A','A','2015-04-28','2015-11-17',28)
,(7,'A','A','2015-04-28','2015-11-17',28)
,(7,'A','A','2015-05-12','2015-11-17',28)
,(7,'A','A','2015-05-12','2015-11-17',28)
DECLARE @Holding TABLE([CUSTOMER_ID] INT
,[TEAM] VARCHAR(1)
,[TYPE] VARCHAR(1)
,[START_DATE] DATETIME
,[END_DATE] DATETIME
,[GROUP_DAYS_CRITERIA] INT
,[START_ROW_ORDER] INT
,[END_ROW_ORDER] INT)
INSERT INTO @Holding
SELECT TT.[CUSTOMER_ID]
,TT.[TEAM]
,TT.[TYPE]
,TT.[START_DATE]
,TT.[END_DATE]
,TT.[GROUP_DAYS_CRITERIA]
,ROW_NUMBER() OVER (PARTITION BY TT.[CUSTOMER_ID],TT.[TEAM],TT.[TYPE] ORDER BY TT.[START_DATE]) [START_ROW_ORDER]
,ROW_NUMBER() OVER (PARTITION BY TT.[CUSTOMER_ID],TT.[TEAM],TT.[TYPE] ORDER BY CASE WHEN TT.[END_DATE] IS NULL THEN 1 ELSE 0 END,TT.[END_DATE]) [END_ROW_ORDER]
FROM @TempTable TT
SELECT DISTINCT C.[CUSTOMER_ID]
,C.[TEAM]
,C.[TYPE]
,CASE WHEN C.[GROUP_ON_PREVIOUS] = 1 THEN LAG(C.[START_DATE]) OVER (PARTITION BY C.[CUSTOMER_ID],C.[TEAM],C.[TYPE] ORDER BY C.[START_DATE])
ELSE C.[START_DATE]
END [START_DATE]
,CASE WHEN C.[GROUP_ON_NEXT] = 1 THEN LEAD(C.[END_DATE]) OVER (PARTITION BY C.[CUSTOMER_ID],C.[TEAM],C.[TYPE] ORDER BY C.[START_DATE])
ELSE C.[END_DATE]
END [END_DATE]
FROM(SELECT A.[CUSTOMER_ID]
,A.[TEAM]
,A.[TYPE]
,A.[START_DATE]
,B.[END_DATE]
,CASE WHEN A.[START_DATE] <= DATEADD(DAY,A.[GROUP_DAYS_CRITERIA],LAG(B.[END_DATE]) OVER (PARTITION BY A.[CUSTOMER_ID],A.[TEAM],A.[TYPE] ORDER BY A.[START_ROW_ORDER])) THEN 1
ELSE 0
END [GROUP_ON_PREVIOUS]
,CASE WHEN DATEADD(DAY,A.[GROUP_DAYS_CRITERIA],B.[END_DATE]) >= LEAD(A.[START_DATE]) OVER (PARTITION BY A.[CUSTOMER_ID],A.[TEAM],A.[TYPE] ORDER BY A.[START_ROW_ORDER]) THEN 1
ELSE 0
END [GROUP_ON_NEXT]
FROM @Holding A
INNER JOIN @Holding B ON A.[CUSTOMER_ID] = B.[CUSTOMER_ID]
AND A.[TEAM] = B.[TEAM]
AND A.[TYPE] = B.[TYPE]
AND A.[START_ROW_ORDER] = B.[END_ROW_ORDER]) C
WHERE NOT (C.[GROUP_ON_PREVIOUS] = 1 AND C.[GROUP_ON_NEXT] = 1)
在不到一秒的时间内返回行