In perl that is by no means a guarantee to also be the fastest code :)

If your dataset is large/huge, you'd notice that the unpack solution posted by an Anonymous Monk was way faster than all the others. And once you are used to that, it will be much easier to extend and maintain than regular expressions.