add Marpa::R2::ASF::rh_values() and Marpa::R2::Scanless::R::rule_closure() methods and tests for them

[rns]

rh_values() is a shortcut for

my $length = $glade->rh_length();
my @return_value = map { $glade->rh_value($_) } 0 .. $length - 1;

as used in sl_panda.t.

Also, if I got Deyan's idea right, it could be useful for rule_closure():

my $rule_id     = $glade->rule_id();
my $closure     = $grammar->rule_closure($rule_id);

my @semantic_values = $closure->( $glade->rh_values() );

which I plan to test in sl_panda1.t.

If that's what's been meant, I can start writing rule_closure() to convert all

my $penn_tag = penn_tag($symbol_name);

to

my $penn_tag = $panda_recce->rule_closure( $rule_id )->( $glade->rh_values() );

$panda_recce is here for a reason because I think rule_closure() needs to be a method of Scanless::R, not Scanless::G because actions are resolved in Scanless::R and rule_closure() must return resolved actions.

[jeffreykegler]

Looks llike a good idea. I'll look it over more closely during the next inter-phase.

And you are right that rule_closuure() need to be a Scanless::R method, not a Scanless::G method.

[rns]

sl_panda1.t will not benefit much from rule_closure() as it has no action closures so sl_panda.t looks like a better place to test rh_values().

Update: but we still need to test rule_closure() with ASF traversers, hence the above commit was reverted, sl_panda1.t restored and closures added to it.

[rns]

This is suggested like a solution for this issue #78 — Deyan was informed on IRC and will look into it as time permits — http://irclog.perlgeek.de/marpa/2014-01-23#i_8163300.

[jeffreykegler]

@dginev @rns : Reading the conversation, I am not 100% sure where this and the related issue #78 stand. Are you both still working this out, or is it ready for me to review, having in mind a merge into the master development branch?

[rns]

rh_values() and rule_closure() are both done, tested in sl_panda1.t and thus ready for review.

@deyan provided an example and I added the code to show how they work -- http://gist.github.com/rns/8579375

[jeffreykegler]

Great! I'll get to work on this, probably tomorrow California time.

[jeffreykegler]

I have a couple of follow-up issues, but Github auto-closed this when I merged and re-opening is not allowed. So I've moved the discussion to issue #78