SQL CASE IF [VALUE] WHEN [Value] AND [VALUE] THEN答案

【问题标题】：SQL CASE IF [VALUE] WHEN [Value] AND [VALUE] THENSQL CASE IF [VALUE] WHEN [Value] AND [VALUE] THEN
【发布时间】：2017-03-08 16:52:20
【问题描述】：

我不确定我是否过于复杂了这么小的任务，但如果有人可以在正确的方向上帮助我一点，那将不胜感激。

我有两个具有相同字段的表，我创建了一个将这两个值合二为一的视图。

目前我的声明是：

    SELECT TOP 1000 [Name]
      ,[WorkOrder]
      ,[OnSiteStarted]
      ,[OnSiteSubmitted]

  FROM [DBName].[dbo].[sw_onsite_offsite]
  ORDER BY WorkOrder

结果集出来为(sn-p)：

Name       WorkOrder      OnSiteStarted          OnSiteSubmitted
On-Site     23022017    2017-02-23 20:31:49.000  2017-02-23 20:33:47.000
Off-Site      35        2017-02-15 18:32:26.000  2017-02-15 21:11:04.000
On-Site     64169086    2017-02-22 11:42:40.000  2017-02-22 15:10:17.000
Off-Site    91487689    2017-02-11 13:48:11.000  2017-02-11 14:32:37.000
On-Site     91494707    2017-02-13 14:39:42.000  2017-02-13 14:49:03.000
Off-Site    92810031    2017-02-17 09:30:59.000  2017-02-17 09:58:19.000
On-Site     92810031    2017-02-17 09:00:12.000  2017-02-17 09:30:48.000
Off-Site    92810034    2017-02-17 10:01:27.000  2017-02-17 10:24:45.000
On-Site     92810039    2017-03-03 12:33:41.000  2017-03-03 12:52:37.000
Off-Site    92810039    2017-03-03 12:44:52.000  2017-03-03 13:41:35.000

我所追求的最终结果是：当非现场和现场表单具有相同的 WorkOrder 编号时，这将仅与 DATEDIFF(MINUTES, [onsitesubmitted], [offsitesubmitted] 字段一起显示它们。

CASE 语句是否会以某种方式适用于此，还是其他方面？

【问题讨论】：

你能告诉我们想要的输出吗？
您使用的是哪个 DBMS？

标签： sql case

【解决方案1】：

我想你想要一个自我加入：

select t1.WorkOrder,
    DATEDIFF(MINUTES, t1.onsitesubmitted, t2.OnSiteSubmitted) diff
from [DBName].[dbo].[sw_onsite_offsite] t1
join [DBName].[dbo].[sw_onsite_offsite] t2 on t1.WorkOrder = t2.WorkOrder
where t1.name = 'On-Site'
    and t2.name = 'Off-Site'

这仅返回那些在场外和现场都存在相同 WorkOrder 的行。如果您希望它在不匹配的情况下也能产生结果，您可以使用适当的外连接。

【讨论】：

【解决方案2】：

SELECT onsite.[WorkOrder],
       DATEDIFF(MINUTES, onsite.OnSiteSubmitted, offsite.OnSiteSubmitted)
FROM   [DBName].[dbo].[sw_onsite_offsite] onsite join [DBName].[dbo].[sw_onsite_offsite] offsite on onsite.WorkOrder = offsite.WorkOrder
WHERE  onsite.Name = 'On-Site' AND
       offsite.Name = 'Off-Site'
ORDER BY WorkOrder

【讨论】：

【解决方案3】：

如果您的数据库支持，您可以使用 lead() 或 lag() 窗口函数（您尚未标记数据库，但它看起来像 sql-server）

select 
    Name
  , WorkOrder
  , OnSiteStarted
  , OnSiteSubmitted
  , datediff(minute
      , lead([onsitesubmitted]) over (
          partition by WorkOrder 
          order by Name 
        )
      , [onsitesubmitted]
     ) as Duration
from [dbo].[sw_onsite_offsite]
order by WorkOrder, OnSiteSubmitted

sql server 中的rextester 演示：http://rextester.com/OTNS68646

+----------+-----------+---------------------+---------------------+----------+
|   Name   | WorkOrder |    OnSiteStarted    |   OnSiteSubmitted   | Duration |
+----------+-----------+---------------------+---------------------+----------+
| Off-Site |        35 | 2017-02-15 18:32:26 | 2017-02-15 21:11:04 | NULL     |
| On-Site  |  23022017 | 2017-02-23 20:31:49 | 2017-02-23 20:33:47 | NULL     |
| On-Site  |  64169086 | 2017-02-22 11:42:40 | 2017-02-22 15:10:17 | NULL     |
| Off-Site |  91487689 | 2017-02-11 13:48:11 | 2017-02-11 14:32:37 | NULL     |
| On-Site  |  91494707 | 2017-02-13 14:39:42 | 2017-02-13 14:49:03 | NULL     |
| On-Site  |  92810031 | 2017-02-17 09:00:12 | 2017-02-17 09:30:48 | NULL     |
| Off-Site |  92810031 | 2017-02-17 09:30:59 | 2017-02-17 09:58:19 | 28       |
| Off-Site |  92810034 | 2017-02-17 10:01:27 | 2017-02-17 10:24:45 | NULL     |
| On-Site  |  92810039 | 2017-03-03 12:33:41 | 2017-03-03 12:52:37 | NULL     |
| Off-Site |  92810039 | 2017-03-03 12:44:52 | 2017-03-03 13:41:35 | 49       |
+----------+-----------+---------------------+---------------------+----------+

【讨论】：