Use the
finddepth
function from File::Find, shown in
Example 9.3
.
#!/usr/bin/perl
# rmtree1 - remove whole directory trees like rm -r
use File::Find qw(finddepth);
die "usage: $0 dir ..\n" unless @ARGV;
*name = *File::Find::name;
finddepth \&zap, @ARGV;
sub zap {
if (!-l && -d _) {
print "rmdir $name\n";
rmdir($name) or warn "couldn't rmdir $name: $!";
} else {
print "unlink $name";
unlink($name) or warn "couldn't unlink $name: $!";
}
}
Or use
rmtree
from File::Path, as shown in
Example 9.4
.
#!/usr/bin/perl
#
rmtree2 - remove whole directory trees like rm -r
use File::Path;
die "usage: $0 dir ..\n" unless @ARGV;
foreach $dir (@ARGV) {
rmtree($dir);
}
WARNING:
These programs remove an entire directory tree. Use with extreme caution!
The File::Find module exports both a
find
function, which traverses a tree in the (essentially random) order the files occur in the directory, as well as a
finddepth
function, which is guaranteed to visit all the files underneath a directory before visiting the directory itself. This is exactly what we need to remove a directory and its contents.
We have to use two different functions,
rmdir
and
unlink
. The
unlink
function deletes only files, and
rmdir
only deletes empty directories. We need to use
finddepth
to make sure that we've first removed the directory's contents before we
rmdir
the directory itself.
Check first that the file isn't a symbolic link before determining if it's a directory.
-d
returns true for both a directory and a symbol link to a directory.
stat
,
lstat
, and the file test operators like
-d
all use the operating system call
stat
(2), which returns all the information kept about a file in an inode. These functions and operators retain that information and let you do more tests on the same file with the special underscore (
_
) filehandle. This avoids redundant system calls that would return the same information, slowly.