Re^4: Parsing Motorola S-Rec file

[Tux]

Why not take the last step too?

$ cat test.pl
use 5.16.2;
use warnings;
use Benchmark  qw( cmpthese );
use List::Util qw( reduce sum );

my $count   = 12;
my $address = "A42187B56F";
my $data    = "39AD0D96D51CD3";

my $r = sub {
    reduce { $a + $b }
    map    { hex $_  }
    ($count, $address =~ m/../g, $data =~ m/../g);
    };

my $s = sub {
    sum map { hex } $count,
                   (unpack "(A2)*", $address),
                   (unpack "(A2)*", $data);
    };

say $r->();
say $s->();

cmpthese (-2, { LUreduce => $r, LUsum => $s });
$ perl test.pl
1487
1487
             Rate LUreduce    LUsum
LUreduce 216392/s       --     -40%
LUsum    359162/s      66%       --

[AppleFritter]

Ah, good point! I'm so used to using reduce for everything that sum didn't even occur to me. (And I've never been as comfortable with (un)pack as with regular expressions; time to remedy that, perhaps.)

[gri6507]

I'm sorry, but my Perl understanding is not quite up to par to understand what's going on here, or why this is better. Can you please clarify?

[Tux]

Using sum plus unpack "(A2)*" is 66% faster than using reduce plus m/../g.