我有一个具有以下结构的 SQLite 表:
正如您所看到的,
sampleDate
列每1分钟就有一个新行,但有时该过程会停止并且存在“时间间隙”,如下划线示例所示。我需要通过 SQL
查询找到这些差距,因为这个表很大。
可能的预期结果输出可能是:
2023-03-20 12:56:27
2023-03-29 10:46:46
这可能表示这两个日期之间存在差距。
您可以使用
LEAD()
函数检索下一个日期,然后计算差异以检索间隙:
with cte as (
select *,
lead(sampleDate, 1, sampleDate) over (order by sampleDate) as nextDate,
CAST(strftime('%M', lead(sampleDate, 1, sampleDate) over (order by sampleDate)) as integer)
- CAST(strftime('%M', sampleDate) as integer) as diff
from mytable
)
select sampleDate, nextDate
from cte
where diff > 1
或者您可以使用左反连接来获取没有下一分钟记录的行吗?下面提供的示例 -
with cte as (
select 4556 id, datetime('2023-03-20 12:54:27') sampleDate
union ALL
select 4557 id, datetime('2023-03-20 12:55:27') sampleDate
union ALL
select 4558 id, datetime('2023-03-20 12:56:27') sampleDate
union ALL
select 4559 id, datetime('2023-03-29 10:46:46') sampleDate
union ALL
select 4560 id, datetime('2023-10-06 10:52:47') sampleDate
union ALL
select 4561 id, datetime('2023-10-06 10:53:47') sampleDate
)
select a.* from cte a
left join cte b ON b.sampledate = datetime(a.sampleDate,'+1 minutes') -- if wanted the gap to be exact one minutes
-- ON b.sampledate between datetime(a.sampleDate,'+1 seconds') and datetime(a.sampleDate,'+1 minutes') -- if you wanted the gap to be anywhere under a minute
where b.id is NULL