使用SSIS以不同方式填充从CSV到表的日期列

Question

I have CSV file

**date column**
06/04/1999
06/04/2000
.
.
.06/04/2017

I'm just loading this file from Csv file to Database table using SSIS

Flat file source to OLEDB Destination

where date column is absolutely DT_STR and SQL Table date column is varchar(50)

But it is behaving differently when I load sometimes it is populating like

  **date column**
    06/04/1999
    06/04/2000

and some other times like

 **date column**
    1999-04-06
    2000-04-06

I'm just trying to load the same what ever is there in Flat file but why it is giving "-" sometimes . Can any one please suggest me why it behaving differently each time . It doesn't have any conversions(like Derived Columns) in SSIS

Answer 1

You have to convert these values to DateTime then to String with desired format yyyy-MM-dd using a ScriptComponent Transformation

1. Create a Flat File connection manager and an OLEDB Connection Manager (Source and Destination)
2. Add a DataFlow Task
3. In the DataFlow Task add a Flat File Source , Script Component , OLEDB Destination
4. your DataFlow should look like Source --> Script --> Destination
5. In the Script component Mark the Time column as input( assuming it's name is inDateColumn ) and add an Output Column (ex: OutDateColumn ) of String DT_STR

6. In the Script Write The Following Code: (using Vb.net)

    Public Overrides Sub Input0_ProcessInputRow(ByVal Row As Input0Buffer) If Not Row.inDateColumn_IsNull AndAlso _ Not String.IsNullOrEmpty(Row.inDateColumn.Trim) Then ' You can add more formats inside the following method Row.OutDateColumn = DateTime.ParseExact(Row.inDateColumn.Trim,New String(){"yyyy-MM-dd","dd/MM/yyyy"},New System.Globalization.CultureInfo("En-GB"), System.Globalization.DateTimeStyles.None).ToString("yyyy-MM-dd") Else Row.OutDateColumn_IsNull = True End If End Sub 

7. Map OutDateColumn to the destination column