For some applications your position "I can't have any incorrect data in the system, if it aint 100% sure I'll reject it" is definitely the right one. All I'm trying to say is that for me in this particular case to be safe means to import everything I possibly can even if it means that there might be an incorrect character now and then.

There is not one correct way to handle errors. The right way depends on the application.

Edit by castaway: Closed small tag in signature