[TODO] config/gen/makefiles/dynpmc_pl.in sub gather_groups_and_libs()

[jkeenan]

This XXX flag comment generated a t/codingstd/perlcritic.t failure tonight. (Why tonight? After sitting in the file for so long? Don't know.)

# XXX: ordering of libs might be crucial
sub gather_groups_and_libs {
    my @pmcs = @_;

    my ( %group_files, %group_libs, %pmc_group, %pmc_libs );
    for my $pmc (@pmcs) {
        our $class = retrieve("$pmc.dump");

        # there can be many libs
        my %libs = %{ $class->{flags}{lib} || {} };
        $pmc_libs{$pmc} = \%libs;

        # There should be at most a single group
        my $group = $class->{flags}{group}
          or next;
        $pmc_group{$pmc} = $group;
        push @{ $group_files{$group} }, $pmc;
        $group_libs{$group} ||= {};
        foreach my $lib ( keys %libs ) {
            $group_libs{$group}->{$lib} = 1;
        }
    }

    return (\%group_files, \%group_libs, \%pmc_group, \%pmc_libs);
}

So I'm converting this to a Trac ticket: Can anyone opine as to whether the ordering of the libs might indeed be crucial?

Thank you very much.
kid51

[fperrad]

Why tonight ? see r36169 & r36198, before this file wasn't checked against our coding standard.

[rurban]

The ordering is crucial on less forgiving linking platforms as win32 and cygwin.

But the pmc ordering is already preprocessed to order parents, so the XXX can be safely removed

[jkeenan]

Thanks, Reini and François, for looking at this. Eliminated the comment in r36520. Closing ticket.