Just to be sure I understand: non-toplevel, non-ASCII file names will not be scanned properly, right? -- Pierre Neidhardt https://ambrevar.xyz/