Re: Interesting line which I read from Monk's snippets

Greetings jesuashok,

It's a very curious line, and others have explained what it does already; however, it has a couple negatives: It's not easily understood, and it's slower than a more easily understood alternative.

use strict;
use warnings;
use Benchmark qw(cmpthese);

cmpthese(
    100000,
    {
        'ahash' => sub {
            my $pat = {qw(a - b = c ~ d ^)}->{substr($0, 0, 1)};
        }, 
        'tr'    => sub {
            (my $pat = substr($0, 0, 1)) =~ tr/abcd/-=~^/;
        },
    },
);
[download]

The above benchmark returns this:

          Rate ahash    tr
ahash  75120/s    --  -86%
tr    541712/s  621%    --
[download]

It's a nice trick, but I wouldn't use it anywhere except in an obfu.

gryphon
Whitepages.com Development Manager (WDDC)
code('Perl') || die;

Comment on Re: Interesting line which I read from Monk's snippets Select or Download Code

Replies are listed 'Best First'.
Re^2: Interesting line which I read from Monk's snippets by pKai (Priest) on Jan 31, 2006 at 08:30 UTC
To deliver the same result (i. e. `undef`) for not found input values, the `tr` solution has to be a bit more elaborate. Somewhat along the line of `my $found = (my $pat = substr($0,0,1)) =~ tr/abcd/-=~^/; $pat = $found ? $pat : undef;` [download]	[reply] [d/l] [select]