Stuff with the Perl Foundation. A couple of patches in the Perl core. A few CPAN modules. That about sums it up.
One problem I sometimes get hit with is needing to order things in a rather arbitrary fashion. This is annoying so I wrote the following code to order a list of hashes by a particular key according to their position in another list:
#!/usr/bin/perl
use strict;
use warnings;
use Data::Dumper;
sub sort_order {
my @names = @_;
my %order_by = map { $names[$_] => $_ } 0.. $#names;
if ( keys %order_by ne @names ) {
warn "duplicate names found";
}
return \%order_by;
}
my @list = (
{ name => 'foo' },
{ name => 'bar' },
{ name => 'baz' },
{ name => 'alpha' },
);
my $order_by = sort_order( qw/bar alpha foo baz/ );
@list = sort { $order_by->{$a->{name}} <=> $order_by->{$b->{name}} } @list;
print Dumper \@list;
This is rather useful, but I really want to be able to generalize this. However, sort takes a subname or a block, not an anonymous subroutine. I could do nasty things like have the sub install a sub in a symbol table and return the fully qualified name for the sort routine. Hmm, never thought of that.
Hell, now that I think of it, I could make what's returned an object. When used as a string it's the sort sub name and when it falls out of scope the subroutine is deleted. I'd just have to ensure this goes into some namespace which is documented to not be safe to use. In fact, this might be a nice, general purpose technique for getting around the "no anonymous sub" limitation of sort.
fwiw ... (Score:1)
# should probably be
if ( keys %order_by != @names ) {
If you didn't mind the performance hit of a tied hash, you could use my Tie::Hash::Sorted
You could combine your order sub and your sort routine in 1 by using lexicals and references. This way all you would need to do to change the sort order would be to change the array containing the list of keys - the rest would automagically happen. It would also allow you to add/delete/modify keys and values and always have the resul
Re:fwiw ... (Score:2)
Yup. That's a bug in my code, thanks :)
I can't use Tie::Hash::Sorted. I'm getting the hash from an external source. Thanks for the idea, though.
Re:fwiw ... (Score:1)
It does allow you to use an existing hash without damaging it in any way (it makes a copy).
Sounds like.... (Score:2)
File::Sort (Score:2)
Passing sort subs (Score:1)
osfameron
Anon subs work in sort as of 5.6.2 (Score:1)
As far as I understand perlfunc, sort accepts refs to anonymous subs. And I use such code all the time:
sort $sorter{$field}, @objs;Re:Anon subs work in sort as of 5.6.2 (Score:1)
Re:Anon subs work in sort as of 5.6.2 (Score:2)
Oops. You're quite right.
I tried it along time ago and found out it didn't work. I suppose I should have tried again :)