Although file is a standard UNIX utility and therefore preferable to mimetype, a Perl utility, I find it to generally be less reliable than mimetype for correctly identifying file types such as, for example, PDFs if the header is missing
