是否可以在sql中创建:
拥有记录表:
ID LaborID OrderNr OrderStatusID OrderStatusDate
12990 3731573 OPT1814378 2 2018-05-28 09:35:30.123
13105 230687389 OPT1814378 1 2018-05-29 10:32:14.850
13106 230687389 OPT1814378 2 2018-05-29 10:52:14.403
13123 230480202 OPT1814378 1 2018-05-29 13:18:05.233
13130 230480202 OPT1814378 0 2018-05-29 13:29:17.360
12837 3731573 OPT1814089 2 2018-05-25 20:28:24.817
12906 10138504 OPT1814089 1 2018-05-26 10:41:18.680
12909 10138504 OPT1814089 2 2018-05-26 10:57:40.733
12913 10138504 OPT1814089 1 2018-05-26 11:41:48.387
12920 10138504 OPT1814089 0 2018-05-26 12:15:48.590
,其中
OrderStatusID
0 - End
1 - Begin
2 - pause
需要计算工作时间from begin to pause
(1->2)
或from begin to end
(1->0)
。
我的问题是我必须遵守一些条件:
如果第一条记录为2
,则忽略
工作始于1
但可以有更多停顿(1->2)
每次使用0
这种情况下的结果将是:
OPT1814378 230687389 00:20:00
OPT1814378 230480202 00:11:12
OPT1814089 10138504 00:16:12
OPT1814089 10138504 00:34:00
答案 0 :(得分:2)
希望这不是那么难看。
gluster-heketi
; with
cte as
(
-- CTE for generating a sequence no
select *, rn = row_number() over (partition by OrderNr
order by OrderStatusDate)
from @table
),
cte2 as
(
-- Clean up invalid any rows and regenerate new sequence no
select ID, LaborID, OrderNr, OrderStatusID, OrderStatusDate,
rn = row_number() over (partition by OrderNr
order by OrderStatusDate)
from cte
where (rn = 1 and OrderStatusID = 1)
or rn >= 2
)
select OrderNr, LaborID,
convert(varchar(10),
dateadd(second,
datediff(second,
min(OrderStatusDate),
max(OrderStatusDate)),
0),
108)
from cte2
group by OrderNr,
LaborID,
(rn - 1) / 2
将给出值0,0,1,1,2,2等,用于将行逐个分组。
答案 1 :(得分:0)
这个也有效
;WITH CTE
AS (
SELECT row_number() OVER ( PARTITION BY ordernr ORDER BY id ) RN ,* FROM test_ti
)
,cte2
AS (
SELECT *
FROM cte c1
WHERE NOT EXISTS (
SELECT *
FROM cte c2
WHERE c1.id = c2.id
AND c2.rn = 1
AND c2.orderstatusid = 2
)
)
SELECT OrderNr
,LaborId
,TimeInterval
FROM ( SELECT DateDiff(MI, TIME, NxtTm) TimeInterval ,*
FROM (
SELECT * ,lead(TIME) OVER ( ORDER BY id ) NxtTm
FROM cte2
) x
WHERE orderstatusid <> 0
) y
WHERE orderstatusid = 1