Question

首先，非常感谢您对此进行调查。

我们有一个名为measure的表，它记录了实验室中的员工活动，如下所示：

╔════╦═════════╦═════════════════════╦═════════════════════╦═════════╦══════════════╗
║ id ║ staf_id ║     start_time      ║      stop_time      ║ task_id ║ fruit_tested ║
╠════╬═════════╬═════════════════════╬═════════════════════╬═════════╬══════════════╣
║  1 ║     123 ║ 2016-06-01 10:00:00 ║ 2016-06-01 13:15:00 ║       1 ║           90 ║
║  2 ║     123 ║ 2016-06-01 15:20:00 ║ 2016-06-01 18:30:00 ║       2 ║           60 ║
║  3 ║     333 ║ 2016-06-02 10:30:30 ║ 2016-06-02 15:45:00 ║       1 ║           30 ║
║  4 ║     333 ║ 2016-06-03 09:00:00 ║ 2016-06-03 09:45:00 ║       1 ║           60 ║
║  5 ║     555 ║ 2016-06-01 07:00:00 ║ 2016-06-01 11:15:00 ║       1 ║           90 ║
║  6 ║     555 ║ 2016-06-01 11:30:00 ║ 2016-06-01 13:00:00 ║       1 ║           30 ║
║  7 ║     123 ║ 2016-06-02 10:00:00 ║ 2016-06-02 14:00:00 ║       3 ║           30 ║
║  8 ║     333 ║ 2016-06-03 10:20:00 ║ 2016-06-03 13:15:15 ║       1 ║           30 ║
║  9 ║     333 ║ 2016-06-04 07:50:20 ║ 2016-06-04 12:30:50 ║       1 ║           60 ║
║ 10 ║     555 ║ 2016-06-03 05:30:00 ║ 2016-06-03 10:00:00 ║       1 ║           60 ║
║ 11 ║     123 ║ 2016-06-03 06:00:00 ║ 2016-06-03 10:30:30 ║       4 ║           90 ║
║ 12 ║     123 ║ 2016-06-03 12:15:04 ║ 2016-06-03 12:45:50 ║       5 ║           90 ║
║ 13 ║     123 ║ 2016-06-03 17:10:00 ║ 2016-06-03 19:00:00 ║       1 ║           30 ║
║ 14 ║     123 ║ 2016-06-04 12:15:04 ║ 2016-06-04 15:00:00 ║       1 ║           30 ║
║ 15 ║     123 ║ 2016-06-04 15:10:00 ║ 2016-06-04 15:30:00 ║       2 ║           90 ║
╚════╩═════════╩═════════════════════╩═════════════════════╩═════════╩══════════════╝

目前，我们的real timeClock表存在问题，作为临时解决方案，我想从度量表中检索每天每个员工的时钟详细信息，其中clock_in将是第一个measure.start_time用于day和clock_out将是该员工当天的最后一次measure.stop_time，如下所示：

╔══════════╦═════════════════════╦═════════════════════╗
║ staff_id ║      clock_in       ║      clock_out      ║
╠══════════╬═════════════════════╬═════════════════════╣
║      123 ║ 2016-06-01 10:00:00 ║ 2016-06-01 18:30:00 ║
║      123 ║ 2016-06-02 10:00:00 ║ 2016-06-02 14:00:00 ║
║      123 ║ 2016-06-03 06:00:00 ║ 2016-06-03 19:00:00 ║
║      123 ║ 2016-06-04 12:15:04 ║ 2016-06-04 15:30:00 ║
║      333 ║ 2016-06-02 10:30:30 ║ 2016-06-02 15:45:00 ║
║      333 ║ 2016-06-03 09:00:00 ║ 2016-06-03 13:15:15 ║
║      333 ║ 2016-06-04 07:50:20 ║ 2016-06-04 12:30:50 ║
║      555 ║ 2016-06-01 07:00:00 ║ 2016-06-01 13:00:00 ║
║      555 ║ 2016-06-03 05:30:00 ║ 2016-06-03 10:00:00 ║
╚══════════╩═════════════════════╩═════════════════════╝

我如何实现这一结果？任何建议都将深表感谢。

Answer 1

我认为这很容易，只是一个聚合：

select staff_id, min(start_time) as clock_in, max(stop_time) as clock_out
from measures
group by staff_id, date(start_time)
order by staff_id, min(start_time);

Answer 2

select staff_id,
    min(start_time) clock_in,
    max(stop_time) clock_out
from measures
group by staff_id,cast(start_time as date)
order by 1,2

用于从度量表

2 个答案: