Josh Dukes wrote: > Of course this would detect unicode files as being binary and maybe > that's not what you want. How are you thinking about doing it in > perl exactly? There is no such thing as a unicode file. You most likely mean UTF-8 or UTF-16 coded text. Christian