Question

我有一个只包含日期时间的表，如下所示：

           datetime
2016-05-16 10:00:00
2016-05-16 10:30:00
2016-05-16 11:00:00
2016-05-16 11:30:00
2016-05-16 12:00:00
2016-05-16 12:30:00
2016-05-16 13:00:00
2016-05-16 13:30:00
2016-05-16 14:00:00
2016-05-16 14:30:00
2016-05-16 15:00:00
2016-05-16 15:30:00
2016-05-16 16:00:00
2016-05-16 16:30:00
2016-05-16 17:00:00
2016-05-16 17:30:00
2016-05-16 18:00:00
2016-05-16 18:30:00
2016-05-16 19:00:00
2016-05-16 19:30:00

我希望将此表连接到一个表，该表将某些时间间隔标记为closed：

              start                stop closed
2016-05-04 16:56:07 2016-05-04 17:26:20   TRUE
2016-05-13 09:29:23 2016-05-14 21:27:17   TRUE
2016-05-14 21:00:39 2016-05-15 20:48:37   TRUE
2016-05-16 10:26:56 2016-05-16 11:48:24   TRUE
2016-05-16 11:33:42 2016-05-16 12:24:31   TRUE
2016-05-16 13:08:15 2016-05-16 14:27:42   TRUE
2016-05-16 15:46:07 2016-05-16 16:19:38   TRUE
2016-05-16 16:33:49 2016-05-16 17:52:31   TRUE
2016-05-16 16:34:12 2016-05-16 17:52:50   TRUE
2016-05-16 16:35:00 2016-05-16 17:53:33   TRUE

如您所见，其中一些区间重叠，因此当我使用

连接两个表时

SELECT d.*, c.closed 
FROM dates d 
LEFT JOIN closures c
ON c.start <= d.datetime AND d.datetime <= c.stop

我在结果表中得到重复的行。有没有办法将左表中的日期时间标记为已关闭，如果至少有一个间隔跨越它而不重复任何行？

Answer 1

您可以使用完全符合要求的export function start(options: ReplOptions): events.EventEmitter;：

EXISTS()

或者只使用SELECT d.*, CASE WHEN EXISTS(SELECT 1 FROM closures c WHERE c.start <= d.datetime AND d.datetime <= c.stop) THEN 'TRUE' ELSE 'FALSE' END as Your_Ind FROM dates d：

DISTINCT

如果您只对已关闭的记录感兴趣，请在SELECT DISTINCT d.*, c.closed FROM dates d LEFT JOIN closures c ON c.start <= d.datetime AND d.datetime <= c.stop子句中使用EXISTS()。

WHERE

Answer 2

您可以使用$ kontena app deploy：

exists

这应该比使用select d.*, (case when exists (select 1 from closed c where c.start <= d.datetime AND d.datetime <= c.stop and d.closed = TRUE ) then 1 else 0 end) as closed from dates;或group by不同select关闭（开始，停止，关闭）`的版本更快。

范围连接没有重复

2 个答案: