HellyerP wrote:
> This week I was fortunate enough to hear Damian speak twice, once on
> everything and once on Perl6. Damian, it was tremendous of you to come
> and speak to us in London - thank-you very much.
It was my pleasure.
> If he's coming to a city near you, book
> your seats now. The Munich YAPC is next, I think.
Not quite. Belfast next week. Zurich the week after. Then Munich.
> [1] subs with prebound arguments: will they behave as though their
> parameter list were declared exactly like their ancestor but with the
> bound parameter bodily removed?
Yes.
> [2] There was a rumour that we may be able to overload subroutines. How
> will this interact with prebinding?
>
> &foo = sub( ; $alpha, @beta ) { print "array" }
> &foo = sub( ; $alpha, %beta ) { print "hash" }
>
> foo( $x, @y ); # array
> foo( $x, %z ); # hash
> foo( $x ); # ambiguity error
> foo( @y ); # array, $alpha is undef
>
> &bar = &foo.assuming( beta => [ 1, 2 ] );
> bar(); # array, $alpha is undef
Yes.
> # will this create overloaded versions of baz?
> &baz = &foo.assuming( alpha => 0 );
I hope so. :-)
> [3] How is more than one argument prebound?
>
> &foo = sub ( $alpha, $beta, $gamma ) { ... }
>
> # chaining should work.
> &bar = &foo.assuming( alpha => 1 ).assuming( beta => 2 );
>
> # or ought that to be one assumption with a HASH?
> &bar = &foo.assuming( { alpha => 1, beta => 2 } );
>
> # or even just a list of PAIRs?
> &bar = &foo.assuming( alpha => 1, beta => 2 );
The third version is best. The first version also works.
The second probably doesn't work.
> [4] Fancy fors
>
> If we want to traverse two arrays, we can do this
>
> for @alpha, @beta { ... }
>
> and it will exhaust alpha before we get the first element of beta.
Yes.
> The only undefs seen in the block are those that exist in the arrays.
> Alternately, <groan> this
>
> for @alpha ; @beta { ... }
>
> will alternate between elements of @alpha, binding each to the magic
> topic $_ in turn. If the arrays are not of equal length, then the
> shorter one will be 'padded' with undef until the longer one is
> exhausted. So if I wanted the alternation without being affected by the
> padding, I would write something like this:
>
> for @alpha ; @beta {
> next unless defined;
> ...
> }
Probably.
But there might well be a special dispensation that two (or more)
streams feeding one variable only alternate between those streams that
are not exhausted.
Damian
