I have a date column with dates stored as strings, such as 20120817. Unfortunately, the text form field that populates this column is free text, so I cannot guarantee that an occasional “E” or “whatever” shows up in this column. And more than a few already have.
What I need to do is convert the string column into a date column. Of course the convert will reject the random string characters. Is there any way to create a derived column that will not only convert the strings but exclude the non-date convertible strings?
If there were no non-date convertible strings in the table, the following would work:
ADD [convertedDate] AS CONVERT(DATE, [stringDate], 102)
And it does work perfectly in a test table I created. But when I introduce other non-convertible strings, I receive the dreaded “Conversion failed when converting date and/or time from character string” error for obvious reasons.
Is there a function that will catch non-convertible elements that I can add on to this derived column code? Or is a view or function the only – or best – way to handle this? I played around with IsDate() with little luck.
Thank you!
There’s a function called
ISDATE(date), maybe you can use it in a CASE statement or in the WHERE part of the query… It depends on how you’re doing it, maybe something like this