raku-rosettacode/programming_tasks/A/ADFGVX_cipher.md

ADFGVX cipher

Slightly different results from every other entry so far. See discussion page for reasons. It is impossible to tell from casual observation which column comes first in the Raku example. In every other (so far), the sub group with 4 characters is the first column.

srand 123456; # For repeatability
 
my @header   = < A D F G V X >;
my $polybius = (flat 'A'..'Z', 0..9).pick(*).join;
 
my $key-length = 9;
my $key = uc 'unixdict.txt'.IO.words.grep( { (.chars == $key-length) && (+.comb.Set == +.comb) } ).roll;
 
my %cypher   = (@header X~ @header) Z=> $polybius.comb;
 
my $message = 'Attack at 1200AM';
 
use Terminal::Boxer;
say "Key: $key\n";
say "Polybius square:\n", ss-box :7col, :3cw, :indent("\t"), '', |@header, |(@header Z $polybius.comb.batch: 6);
say "Message to encode: $message";
say "\nEncoded: " ~ my $encoded = encode $message;
say "\nDecoded: " ~ decode $encoded;
 
sub encode ($text is copy) {
    $text = $text.uc.comb(/<[A..Z 0..9]>/).join;
    my @order = $key.comb.pairs.sort( *.value )».key;
    my @encode = %cypher.invert.hash{ $text.comb }.join.comb.batch($key-length).map: { [$_] };
    ((^$key-length).map: { @encode».[@order]».grep( *.defined )».[$_].grep( *.defined ).join }).Str;
}
 
sub decode ($text is copy) {
    my @text = $text.split(' ')».comb;
    my $chars = @text[0].chars;
    $_ = flat |$_, ' ' if .chars < $chars for @text;
    my @order = $key.comb.pairs.sort( *.value )».key.pairs.sort( *.value )».key;
    %cypher{ ( grep { /\w/ }, flat [Z] @order.map( { |@text.batch($key-length)».[$_] } ) ).batch(2)».join }.join;
}