← All modules

mkdir -p, rm -r and "empty this directory" as three Raku subs — the whole-tree operations the core deliberately leaves out.

Version

0.2 zef:raku-community-modules

Depends

nothing outside the core

License

Artistic-2.0

Its own test suite

1 file, green

Checked

2026-08-28 against Raku++ 3.20.1 and Rakudo 2026.08

Where it lives

raku.land · source

Install it #

$ rakupp install File::Directory::Tree

zef install File::Directory::Tree writes the same store; either installer leaves the module usable by both engines.

What it is for #

Core Raku gives you mkdir and rmdir, one level at a time, and rmdir refuses a directory with anything in it. The shell verbs you actually reach for — mkdir -p, rm -r — operate on trees, and this module is those verbs: mktree creates a whole path of directories, rmtree removes a directory with everything below it, and empty-directory clears the contents while keeping the directory. Twenty distributions depend on it; if you have used File::Temp, its cleaned-up-for-you tempdir is this module working.

use File::Directory::Tree;

my $root = $*TMPDIR.add("fdt-demo-$*PID");
mktree $root.add('deep/deeper/deepest');
say $root.add('deep/deeper/deepest').d;

$root.add('deep/note.txt').spurt('hello');
say rmtree($root.add('deep'));
say $root.add('deep').e;
say $root.d;
rmtree $root;
say $root.e;

True
True
False
True
False

Both subs take anything Cool that names a path — a string or an IO::Path — and return True on success. mktree also takes :mask(0o700) for the permissions of every directory it creates, which is the polite way to make a private cache directory in one call.

Emptying without removing #

empty-directory is the third verb, for the log directory or spool whose existence other code relies on — everything inside goes, including subtrees, and the directory itself stays:

use File::Directory::Tree;

my $work = $*TMPDIR.add("fdt-empty-$*PID");
mktree $work.add('cache/a/b');
$work.add('cache/one.txt').spurt('x');
$work.add('cache/a/two.txt').spurt('y');

say empty-directory($work.add('cache'));
say $work.add('cache').d;
say $work.add('cache').dir.elems;
rmtree $work;

True
True
0

Symlinks inside the tree are unlinked, not followed — a link out to somewhere precious does not make rmtree delete the precious thing.

What the edges do #

The semantics at the edges are the ones that make scripts idempotent, and they are worth knowing rather than guessing:

use File::Directory::Tree;

my $work = $*TMPDIR.add("fdt-edges-$*PID");
mktree $work;

say rmtree($work.add('never/existed'));

my $file = $work.add('plain.txt');
$file.spurt('z');
say rmtree($file).so;
say $file.e;

rmtree $work;

True
False
True

Removing a tree that is not there is True — the state you asked for is the state you have, so a cleanup path can run twice without guarding. Pointing rmtree at a plain file, though, is a Failure — this module removes directory trees, not files, and it will not quietly become rm -rf for whatever it is handed. The file is untouched. (A Failure is falsey but throws if used carelessly; .so, as here, or an if/with disarms it.)

Where the two engines differ #

Nothing on this page. Tree creation, removal, the emptying, and both edge behaviours answer identically under Raku++ and Rakudo.

What was run to put this page here #

1. Parse — every file of the distribution is parsed by Raku++ itself.
2. Install — rakupp install File::Directory::Tree; it depends on nothing outside the core.
3. Test — the distribution's own suite: 1 file, green.
4. Run — every example on this page, twice under each engine, as the site is built.

Nothing had to be fixed anywhere — which for this module means recursive directory walking, unlink vs rmdir dispatch, symlink detection with .l, and Failure objects all agree between the engines. It earned the quietest kind of page.