From @cognominal

$ cat mmd.pm6
multi sub a([]) { say "[]" }
multi sub a([$i]) { say "[$i]" }
a(); a([1]);
$ perl6 mmd.pm6
No applicable candidates found to dispatch to for 'a'. Available candidates are​:
:(Positional ())
:(Positional (Any $i))

  in main program body at line 3​:mmd.pm6
$

This makes impossible shape based pattern matching.

--
cognominal stef

From @moritz

Am 13.07.2010 02​:13, schrieb Stephane Payrard (via RT)​:

$ cat mmd.pm6
multi sub a([]) { say "[]" }
multi sub a([$i]) { say "[$i]" }
a(); a([1]);
$ perl6 mmd.pm6
No applicable candidates found to dispatch to for 'a'. Available candidates are​:
:(Positional ())
:(Positional (Any $i))

in main program body at line 3​:mmd.pm6
$

This is expected behaviour. Both signatures require at least one
argument, you pass in none.

If you mean to also allow no arguments at all, change your first multi to​:

multi sub a([]?) { say "[]" }

The RT System itself - Status changed from 'new' to 'open'

@moritz - Status changed from 'open' to 'rejected'

From @cognominal

Sorry. My example was wrong. Here another one. C<a([])> executes as
intended but
C<@​t> does not. I find that unintuitive.

cat mmd5.pm6
multi sub a(@​a) { say 1 ~ @​a.perl }
multi sub a([]) { say 2 ~ [].perl }
my @​t=(1,2);
a([]);
a(@​t)
$ perl6 mmd5.pm6
2[]
No applicable candidates found to dispatch to for 'a'. Available candidates are​:
:(@​a)
:(Positional ())

  in main program body at line 5​:mmd5.pm6
$

@moritz - Status changed from 'rejected' to 'open'

From @coke

On Wed Jul 14 15​:42​:09 2010, cognominal wrote​:

Sorry. My example was wrong. Here another one. C<a([])> executes as
intended but
C<@​t> does not. I find that unintuitive.

cat mmd5.pm6
multi sub a(@​a) { say 1 ~ @​a.perl }
multi sub a([]) { say 2 ~ [].perl }
my @​t=(1,2);
a([]);
a(@​t)
$ perl6 mmd5.pm6
2[]
No applicable candidates found to dispatch to for 'a'. Available
candidates are​:
:(@​a)
:(Positional ())

in main program body at line 5​:mmd5.pm6
$

This works now​:

$ cat foo
multi sub a(@​a) { say 1 ~ @​a.perl }
multi sub a([]) { say 2 ~ [].perl }
my @​t=(1,2);
a([]);
a(@​t)
$ ./perl6 foo
1Array.new()
1Array.new(1, 2)

Closable with tests.

--
Will "Coke" Coleda

From @moritz

Turns out that rakudo's answer is still wrong; it always calls the (@​a)
candidate, never the ([]) one​:

22​:44 < moritz> perl6​: multi sub a(@​a) { say 1 ~ @​a.perl }; multi sub
a([]) { say 2 ~ [].perl }; a []
22​:44 <+p6eval> rakudo d4dc7d​: OUTPUT«1Array.new()␤»
22​:44 <+p6eval> ..pugs b927740​: OUTPUT«*** ␤ Unexpected "[])"␤
expecting formal parameter or ")"␤ at /tmp/UwCYKvZkPH line 1, column
50␤»
22​:44 <+p6eval> ..niecza v15-2-gd19c478​: OUTPUT«2[]␤»
22​:44 < moritz> I kinda agree with niecza here
22​:45 < jnthn> Same...I'm thought we had a passing test for that case too
22​:45 < moritz> I was about to write on... and it didn't pass :/
22​:46 < jnthn> ah
22​:46 < jnthn> Something is off here.

From @moritz

Oh, and tests can now be found in S06-multi/unpackability.t

From @jnthn

On Fri Oct 21 19​:24​:38 2011, coke wrote​:

On Wed Jul 14 15​:42​:09 2010, cognominal wrote​:

Sorry. My example was wrong. Here another one. C<a([])> executes as
intended but
C<@​t> does not. I find that unintuitive.

cat mmd5.pm6
multi sub a(@​a) { say 1 ~ @​a.perl }
multi sub a([]) { say 2 ~ [].perl }
my @​t=(1,2);
a([]);
a(@​t)
$ perl6 mmd5.pm6
2[]
No applicable candidates found to dispatch to for 'a'. Available
candidates are​:
:(@​a)
:(Positional ())

in main program body at line 5​:mmd5.pm6
$

This works now​:

$ cat foo
multi sub a(@​a) { say 1 ~ @​a.perl }
multi sub a([]) { say 2 ~ [].perl }
my @​t=(1,2);
a([]);
a(@​t)
$ ./perl6 foo
1Array.new()
1Array.new(1, 2)

Closable with tests.

This wasn't the right answer, but we get it now​:

2[]
1Array.new(1, 2)

And moritz++ wrote tests, with I've now untodo'd, so resolving.

Merci,

/jnthn

@jnthn - Status changed from 'open' to 'resolved'