Phil Parkin (3/30/2010)
I do not believe that this is the most efficient way of working.I get the impression that most people here move files to an archive folder once they have been processed. If you always do this, you won't need to put file-select logic into your package, and that is how I would recommend you proceed.
You will also benefit from knowing that everything in the archive folder has, in fact, been loaded. You would not easily know this otherwise.
I'm going to disagree with you here Phil, this approach adds a element to the processing in the form of renaming the file to a common file name and also copying it to an archive (which is either a bunch of clumsy file system tasks or a manual process). With this method you would also need to the re name the file on the way into the archive so you know which days file it was.