SQL查询按工作流条件对数据进行分组

我正在编写一个SQL查询来查看一些项目指标 我有两张桌子

WorkflowDefinition (In this the workflow definition is according to the   product land and is different for different productLaneID)
ProductLaneID | StepID | StepName | WorkflowRevisionID
4               1        RA         1
4               2        Design     1
4               3        ME         1
4               4        Tooling    1
4               5        Waiting    1
4               6        Manuf      1
4               7        Completed  1
5               1        RA         1
5               2        Concepts   1
5               3        Design     1
5               4        ME         1
5               5        Tooling    1
5               6        Waiting    1
5               7        Manuf      1
5               8        FieldTesting1
5               9        Completed  1

在这方面我想加入俱乐部

 Step 3&4 from ProductLaneID=4 as PLES
 StepID 5&6 from ProductLaneID = 4 as Manuf

StepID 2&3 from productLaneID =5 as Design
StepID 4&5 from productLaneID =5 as PLES
StepID 6&7 from productLaneID =5 as Manufacturing


ProjectScheduleHistory
ProjectID|FromStepID|ToStepID|EnteredStepDate|ExitedStepDate

现在，项目可以从步骤1-2和2-1开始，然后再到1-2 所以我想找到第一步进入的日期和最后一步退出的日期。 因此，它将根据设计、PLES和制造计算每个步骤的天数

结果是

ProjectID | Design CompletionDate | Days in Design| PLES Comp Date | Days in PLES| Manuf Comp Date | No of Days in Manuf.

我从下面的sql查询开始

select ProjectID,FromStepID,ToStepID,
[EnteredStepDate] = MIN(EnteredStepDate),
[ExitedStepDate] = MAX(ExitedStepDate),ProjectLaneID,
[DaysInStep] = SUM(DATEDIFF(DAY,EnteredStepDate,ExitedStepDate)),
REPLACE(REPLACE(REPLACE(REPLACE(StepName,'Manufacturing             Engineering','PLES'),'Tooling','PLES'),'Concept','Design'),'Waiting On   Order','Manufacturing')StepName
from lovWorkflowDefinition
left outer join dtaProjectScheduleHistory
on FromStepID = StepID
where (StepName = 'Design' or StepName = 'PLES' or StepName =  'Manufacturing') and (ProjectLaneID=4 or ProjectLaneID=5)
group by   ProjectID,FromStepID,ToStepID,EnteredStepDate,ExitedStepDate,ProjectLaneID,StepName

我想根据设计、PLES和制造对数据进行分组

---

.

使用条件

CASE/WHEN

语句和派生表子查询考虑以下内容：

SELECT ProjectID, Max([Step Name]) As [StepName],
       Max(CASE WHEN [Step Name] = 'Design' THEN ExitedStepDate END) As [Design Comp Date],
       Sum(CASE WHEN [Step Name] = 'Design' 
                THEN DATEDIFF(DAY, EnteredStepDate, ExitedStepDate) END) AS [Days in Design],

       Max(CASE WHEN [Step Name] = 'Ples' THEN ExitedStepDate END) As [PLES Comp Date],
       Sum(CASE WHEN [Step Name] = 'Ples' 
                THEN DATEDIFF(DAY, EnteredStepDate, ExitedStepDate) END) AS [Days in Design],

       Max(CASE WHEN [Step Name] = 'Manuf' THEN ExitedStepDate END) As [Manuf Comp Date],
       Sum(CASE WHEN [Step Name] = 'Manuf' 
                THEN DATEDIFF(DAY, EnteredStepDate, ExitedStepDate) END) AS [Days in Design],
FROM 
   (SELECT proj.ProjectID, proj.FromStepID, proj.ToStepID, 
           proj.EnteredStepDate, proj.ExitedStepDate,
           CASE 
             WHEN proj.StepID IN (3, 4) AND work.ProductLaneID = 4 THEN 'Ples'
             WHEN proj.StepID IN (5, 6) AND work.ProductLaneID = 4 THEN 'Manufacturing'
             WHEN proj.StepID IN (2, 3) AND work.ProductLaneID = 5 THEN 'Design'
             WHEN proj.StepID IN (4, 5) AND work.ProductLaneID = 5 THEN 'Ples'
             WHEN proj.StepID IN (6, 7) AND work.ProductLaneID = 5 THEN 'Manufacturing'
           END  As [Step Name]
   FROM WorkflowDefinition work
   LEFT OUTER JOIN ProjectScheduleHistory proj
   ON wrk.StepID = proj.FromStepID
   ) As dT
GROUP BY ProjectID

---

语法看起来像SQL Server，所以我为您删除了无关的数据库标记。嗨，Parfait，谢谢您的输入。我可以做到这一点。感谢你的努力。现在我想修改它来处理更多的情况，比如若项目从第3步直接转到第7步，那个么我们如何处理这种情况。这同样适用于其他步骤。查询应该仍然有效。从我收集的资料来看，

ProjectScheduleHistory

表维护了

FromStepID

和

ToStepID

。所以只需将3和7的新记录添加到表中，并通过

StepID

查询连接到

WorkflowDefinition

表中即可。除非出现

[Step Name]

的新条件，否则在派生表中添加一个额外的

CASE/WHEN

语句。Hi Parfait，因为我们在做WorkflowDefinition工作，所以在外部联接ProjectScheduleHistory时，每个ProjectLaneID都会得到一行，这是不需要的。我们只需要在projScheduleHistory中为from查询添加一个带有StepName的列。我们只需要ProjectScheduleHistory中基于新工作流定义的数据。您可以在派生表中添加一个

WHERE

子句条件，以过滤结果记录。因此，添加定义新WorkflowDefinition的内容。对于

[步骤名称]

，将其添加到select和group by lines中的外部主查询中。请参见编辑。执行此操作时，除ProjectID外，所有列都被删除。