Question

我想合并内部和位置基础上的行。我想查看一辆车停在特定位置的时间。在 1 里面，车辆进入特定位置， 0 表示离开特定位置。如果车辆在特定日期进入并离开相同位置，则计算 TimeLast 。例如 HD 和 BP 是车辆进入和离开的位置。

oId    Number   Id       Location            TimeFirst               Inside
1639    0664    1        Hd        2016-05-29 05:48                0
1639    0664    2        Kp        2016-05-29 18:46                1
1640    0359    1        Hd        2016-05-29 13:12                0
1640    0359    1        Hd        2016-05-29 13:42                1
1640    0359    1        Hd        2016-05-29 13:46                0
1643    0812    3        Lr        2016-05-29 11:41                0
1644    0806    4        Mn        2016-05-29 06:56                0
1644    0806    5        Bp        2016-05-29 09:54                1
1644    0806    5        Bp        2016-05-29 17:43                0

输出

oId    Number     Id    Location      TimeFirst                TimeLast       
1639    0664      1     Hd        2016-05-29 05:48         
1639    0664      2     Kp        2016-05-29 18:46         
1640    0359      1     Hd        2016-05-29 13:12         
1640    0359      1     Hd        2016-05-29 13:42          2016-05-29 13:46    
1643    0812      3     Lr        2016-05-29 11:41         
1644    0806      4     Mn        2016-05-29 06:56         
1644    0806      5     Bp       2016-05-29 09:54           2016-05-29 17:43

预期输出

{{1}}

Answer 1

这对您有所帮助：

private string _color;
public string Color
{
    get { return _color; }
    set { _color = value; }
}

Answer 2

正如您在Sql Server 2012上一样，您可以使用窗口聚合函数lag和lead来执行此操作，这样您可以将数据与上一个或下一个记录中的数据进行比较：

select   oId, Number, Id, Location, TimeFirst,
         CASE WHEN inside = 1 THEN DepartureTime END DepartureTime
from     (  select   zs.oId, o.Number, z.Id, z.Location, zs.TimeFirst, 
                     zs.inside,
                     lag(zs.inside) over (partition by o.oId, z.Id 
                                          order by zs.TimeFirst) as prev_inside,
                     lead(zs.inside) over (partition by o.oId, z.Id
                                           order by zs.TimeFirst) as next_inside,
                     lead(zs.TimeFirst) over (partition by o.oId, z.Id
                                              order by zs.TimeFirst) as DepartureTime
            from     zs 
            join     z on zs.Id = z.Id 
            join     o on zs.oId=o.oId
            where    z.GroupId in (1094,1095,1096) 
            and      CONVERT(Date,TimeFirst)='2016/5/29' 
         ) as base
where    (inside = 1 or coalesce(prev_inside,0) = 0)
order by oId, TimeFirst

SQL fiddle

如何在状态库上动态合并行

2 个答案: