How to get desired output in SSIS by grouping by and count?

问题

I'm creating an SSIS package wherein a user needs to be notified if there is a duplicate productcode for one productlabel. We retrieve the products thru a csv flat file being sent to us in a shared location.

In this example I have Productcode = 1a for both productlabel Physio Ormix and Nixen.

So basically the productcode = 1a was a duplicate because it was also used by the productlabel=Nixen. Hence, notify users for the duplicate ProductCode and Productlabel used. I tried creating an aggregate that performs group by ProductCode and Counts it.

Can someone give me tips on how to do this?

回答1:

I think you can use a script component and a conditional split to get the duplicates without all this logic:

1. Inside the data flow task add a Script Component
2. Add an output column of type DT_BOOL (example name is Flag)

3. Inside the script component write a similar script:

   using System.Collections.Generic;
   
   public class ScriptMain:  
       UserComponent  
   
   {  
   
       List<string> lstKey = new List<string>;
       List<string> lstKeylabel = new List<string>;
   
       public override void Input0_ProcessInputRow(InputBuffer0 Row)  
       {  
   
           if(!lstKey.Contains(Row.ProductCode){
   
               lstKey.Add(Row.ProductCode);
               lstKeylabel.Add(Row.ProductCode + ";" + Row.ProductLabel);
               Row.Flag = true;
   
           }else if(lstKeylabel.Contains(Row.ProductCode + ";" + Row.ProductLabel)) {
   
               Row.Flag = true;
   
           }else{
   
               Row.Flag = false;
   
           }
   
       }  
   
   }
   

4. Add a conditional split after the script component with a similar expression:

   [Flag] == true
   

5. All records that are passed thru the true path are unique, all rows passed in the false path are the duplicates.

来源：https://stackoverflow.com/questions/55036780/how-to-get-desired-output-in-ssis-by-grouping-by-and-count