Charles Harvey wrote: > > I have a Mac client that is grabbing pictures and writing them > to a netatalk share, and the files prefix with klee and then are > one-up serialized (not my design, I just do networking). The PC > client grabs the images via samba share and displays them. This > is part of an museum exhibit (http://nano.arts.ucla.edu) that > runs till September 2004. On a busy day, the system grabs > around 4000 images. > > We are archiving the pictures, so I have a cron script that > copies the directory, tars the copy and sends it to a backup > server. That all works, but I need to empty the source and copy > directories. Previously, I was just calling rm -rf > /test/directory/*, but I get "Argument too long" errors. So, I > need to load the directory listing into an array, and rm or > unlink them one at a time. When I use readdir(DIR, $testdir), I > only get half of the entries. I tried explicitly declaring DIR > as an array, but I guess handles don't do that. How do I load > all of the directory into an array, or more specifically why am > I only getting half? I'm using Perl 5.8 on Mac OSX Panther.
Hi Charles. You could forget about diagnosing the problem and just use my $n = unlink grep -f, glob '/test/directory/klee*'; print "$n files deleted\n"; HTH, Rob -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] <http://learn.perl.org/> <http://learn.perl.org/first-response>