From eric.hodges@gmail.com

use v6;

rule test {test};

"test" ~~ /<test>/;
say '$/.keys => ', $/.keys.perl;
say '%($/).keys =&gt; ', %($/).keys.perl;

# outputs
# $/.keys => []
# %($/).keys => ["test"]

Same could be said for .values and .kv

It would be very DWIM for it to act like a hash in these cases by default.

From @coke

On Tue Jan 20 12​:58​:20 2009, eric256 wrote​:

use v6;

rule test {test};

"test" ~~ /<test>/;
say '$/.keys => ', $/.keys.perl;
say '%($/).keys =&gt; ', %($/).keys.perl;

# outputs
# $/.keys => []
# %($/).keys => ["test"]

Same could be said for .values and .kv

It would be very DWIM for it to act like a hash in these cases by default.

Here's a slightly updated version​:

$ more foo.p6
use v6;

our rule test {test};

"test" ~~ /(<&test>)/;
say '$/.keys => ', $/.keys.perl;
say '%($/).keys =&gt; ', %($/).keys.perl;

$ ./perl6 foo.p6
$/.keys => (0)
%($/).keys => ()

--
Will "Coke" Coleda

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

From @cognominal

I edited my garbled sentences.

[15​:28​:22] <cognominal> I know that a match is both an array and a
hash and that both containers have a .keys method. But I would expect
.keys for a Match to return a list of the named reductions
[15​:28​:37] <cognominal> rakudo​: grammar A { token TOP { <a> }; token
a { a } }; class A​::A { method TOP($/) { say $&lt;a>.keys }} ;
A.parse( 'a', :actions(A​::A))
[15​:28​:42] <p6eval> rakudo 4bd478​: OUTPUT«␤»
[15​:29​:09] <moritz_> the spec isn't clear on that point
[15​:31​:36] <cognominal> I think it can be useful on actions for a
rule that is a mere alternation list. method alternation($/) {
  make $/."{($/.keys)[0]}".ast } # if I got that right
[15​:32​:00] <masak> fwiw, I'd also expect the subrule names from .keys
[15​:32​:31] <moritz_> cognominal​: are you sure you don't want .caps for that?
[15​:33​:06] * masak .oO( .shower-caps )
[15​:33​:24] <moritz_> that said, I could very well live with
Match.keys returning (Match.list.keys, Match.hash.keys)
[15​:33​:41] <moritz_> and likewise .values and .kv

--
cognominal stef

From @FROGGS

TimToady​: a Match object should return both array and hash keys in the
absence of further guidance as to which is wanted (implementations tend
to store the numeric keys in the hash in any case for convenience)

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

From @Mouq

This appears to have been a resolved issue for some time, both in spec and in implementation. I've added a test in Roast commit​:

commit bb73feaaa6bdd3d0a89a701552c7e4ca5e41a2f4
Author​: Mouq <alexmoquin@​gmail.com>
Date​: Mon Feb 9 00​:59​:13 2015 -0500

  Add test for RT #​77146

Marking this ticket as resolved :)

On Sat May 11 01​:13​:10 2013, FROGGS.de wrote​:

TimToady​: a Match object should return both array and hash keys in the
absence of further guidance as to which is wanted (implementations tend
to store the numeric keys in the hash in any case for convenience)

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