Question

如何让sql查询在mysql的日期范围内每天返回一行？

我有一个具有这种结构的工作表：

job_id  |   station_id  |   start_date  |   end_date    |
--------+---------------+---------------+---------------+
    1   |   1           |   2017-06-01  |   2017-06-02  |
    2   |   2           |   2017-06-01  |   2017-06-03  |
    3   |   1           |   2017-06-02  |   2017-06-04  |
    4   |   4           |   2017-06-01  |   2017-06-02  |
    5   |   2           |   2017-06-04  |   2017-06-05  |
    6   |   1           |   2017-06-06  |   2017-06-08  |
    7   |   3           |   2017-06-01  |   2017-06-02  |

和一个电台表：

station_id |    type    |
-----------+------------|
    1      |    1       |
    2      |    1       |
    3      |    2       |
    4      |    2       |
    5      |    2       |

如果我想知道2017-06-01中有多少2型电台正在使用，我可以很容易地做到这一点：

SELECT COUNT(job_id)
FROM jobs
WHERE start_date <= '2017-06-01'
AND  end_date > '2017-06-01'
AND station_id IN (SELECT station_id FROM station WHERE type = 2);

这个例子告诉我们正在使用2个站（3和4）：

+-------------+
|COUNT(job_id)|
+-------------+
|           2 |
+-------------+

但是， 我希望这会在一系列日期中每天返回一行 。

到目前为止，我已经用多次运行第一个查询的python脚本解决了这个问题， 有没有办法直接用mysql执行此操作？ < / p>

编辑：根据user1960808的要求，范围2017-06-01 2017-06-03和类型2的预期结果将是：

+-----------+-------------+
|   date    |COUNT(job_id)|
+-----------+-------------+
|2017-06-01 |   2         |
|2017-06-02 |   0         |
|2017-06-03 |   0         |
+-----------+-------------+

Answer 1

也许最简单的方法是生成范围：

select d.dte, count(job_id)
from (select '2017-06-01' as dte union all
      select '2017-06-02' as dte 
     ) d left join
     jobs j
     on j.start_date <= d.dte and j.end_date > d.dte
where j.station_id in (select station_id from station s where s.type = 2)
group by d.dte;

如果你有很多日期，这不是最有效的解决方案。

Answer 2

检查

DELIMITER $$

Create Procedure GetJobRunningCount(dStartDate datetime,dEndDate datetime,dType int)
Begin

  DROP TABLE IF EXISTS CountSummary;
  CREATE TEMPORARY TABLE CountSummary(SelectedDate date,Count int);


  While (dStartDate<=dEndDate) do

  insert into CountSummary(SelectedDate,Count)
    SELECT dStartDate,COUNT(1)
FROM jobs jo
inner join station st on(st.station_id=jo.station_id and type=dType )
WHERE start_date <= dStartDate
AND  end_date > dStartDate;


 set dStartDate=Date_add(dStartDate,interval 1 Day);

   End while;

   select * from CountSummary;

END$$
DELIMITER ;

call GetJobRunningCount('2017-06-01','2017-06-08',2)

Answer 3

如果你可以创建一个如下面url所述的整数表，那么我认为可以通过以下查询来实现。

https://www.xaprb.com/blog/2005/12/07/the-integers-table/

创建整数表后的查询

SELECT job_id,station_id,start_date + INTERVAL i DAY, start_date,end_date
FROM job
INNER JOIN integers
       ON i <= DATEDIFF(end_date,start_date)
AND station_id IN (SELECT station_id FROM station WHERE type = 2);

如何让sql查询在mysql的日期范围内每天返回一行？

3 个答案: