我正在努力构建一个UPSERT脚本，该脚本在发生冲突时（1）修改现有记录以避免冲突，然后（2）插入新记录。我正在尝试建立一个在保持空间的同时保持时间点保真度的表。从理论上讲，与捕获不变的重复数据切片相比，正确捕获数据中的更改效率更高。
我已经建立了一个示例来帮助指出我要做什么。在特定的时间段内，有三名赛车手在特定的赛道上。其中一名驾驶员决定改变赛道，而另两名则决定留在原地。在稍后的某个时间点，第四个驾驶员进入图片。
我希望能够查询表格并说“今天谁在哪个赛道上？”
希望这是有道理的。
谢谢。
----------------------Set up an example table
create table drivers(
firstName varchar(30)
,lastName varchar(30)
,raceTrack varchar(30)
,effectiveDate date
,expirationDate date
);
insert into drivers
select 'ricky', 'bobby', 'talladega', cast('2019-01-01' as date), cast('9999-12-31' as date)
union
select 'cal', 'naughton', 'martinsville', cast('2019-01-01' as date), cast('9999-12-31' as date)
union
select 'jean', 'girard', 'daytona', cast('2019-01-01' as date), cast('9999-12-31' as date);
----------------------Expire the daytona record for Jean
----This is the UPDATE portion of the UPSERT and triggers a conflict with the UPSERT since firstname and lastname match.
update drivers set expirationdate = cast('2019-02-10' as date)
where firstname = 'jean' and lastname = 'girard';
----------------------Insert new circuit de monaco record for Jean
----This is the INSERT portion of the UPSERT
insert into drivers
select 'jean', 'girard', 'circuit de monaco', cast('2019-02-11' as date), cast('9999-12-31' as date);
----------------------A new driver comes into the data
----This is an example of the expected behavior without a conflict
insert into drivers
select 'reese', 'bobby', 'greenville pickens', cast('2019-02-12' as date), cast('9999-12-31' as date);
----------------------Query all of the driver and track records for some day in January 2019
select *
from drivers
where cast('2019-01-11' as date) between effectivedate and expirationdate;
----------------------Query all of the driver and track records for some day in February 2019
select *
from drivers
where cast('2019-02-11' as date) between effectivedate and expirationdate;
select *
from drivers
where cast('2019-02-12' as date) between effectivedate and expirationdate;