From: Bob Barrows on
I'm still not entirely sure what you tried. Did you use the Derived
Column task to create a new column using the expression you referred to
in the earlier post? That should certainly handle the 00/00/0000
problem. What it won't handle is the invalid dates problem and the regex
solution seems to be the right path for that.
Again, I'm not entirely sure what you have tried, but I do suggest you
use the message board at the SQLIS site to ask for help with the tools
they provide.


robboll wrote:
> I'm not giving up because I need a production solution now -- which to
> me means DTS imported into SSIS (what a shame). Hey! It works!!!
>
> The simple solution in DTS is to comment out the bad imput row in the
> Transformations:
> 'DTSDestination("CHANGE_DATE") = DTSSource("Col004")
>
> And use the IsDate function as you mentioned earlier in a simple IF/
> Else statement:
>
> If IsDate(DTSSource("Col004")) then
> DTSDestination("CHANGE_DATE") = DTSSource("Col004")
> else
> DTSDestination("CHANGE_DATE") = null
> end if
>
> I was hoping to be able to use a script task to do this when I found
> what I thought was the solution on SQLIS.com. Looked promising at the
> time but still no luck. I know to use import tables but SSIS is
> supposed to be an improvement to DTS. So I am determined though to
> find the SSIS solution. Will keep digging.

--
HTH,
Bob Barrows