Front page | perl.perl5.porters |
Postings from January 2022
In this case, it's an inline recursive function of which the relevant
> parts look a bit like:
>
> sub do_stuff {
> my $prod = 1;
> my $recurse;
> $recurse = sub {
> ...
> for my $new (@interesting_values) {
> local $prod = $prod * $new; # aspirational
> $recurse->();
> }
> };
> $recurse->();
> }
>
> .. so shadowing doesn't help. Alternatives such as a) making it a
> one-element array, b) passing it as an argument or c) manually
> saving and restoring are all possible, but all likely to be slower,
> more error-prone, and add complexity to already-complex code.
>
okay here's my rewrite. I started with (c) then switched to (b).
sub do_stuff {
my $prod = shift;
my @interesting_values = wassup($prod);
for my $new (@interesting_values) {
do_stuff($prod * $new);
}
};
do_stuff(1);
I for one think passing an argument to a named recursive routine is less
error-prone and less complicated than trying to do arcane and fragile perl
four magic on a variable you're passing by shared outer scope to inner
iterations of an anonymous one.
It seems like it might be a good case for ditching recursion altogether in
favor of a generator/consumer pattern.
my @queue = (1);
while(@queue){
my $prod = shift @queue;
# push @queue, map { $prod * $_ } wassup($prod); # one way to
do it
push @queue, $prod * $_ for wassup($prod); # another way to
do it
}
--
"Lay off that whiskey, and let that cocaine be!" -- Johnny Cash
Thread Previous
|
Thread Next