Things are slow here today. Even for the Chum-Bucket!

So here's a Perl script and a shell script that solves jumbles.

#!/usr/bin/perl -wT

$ENV{'PATH'} = "";
use strict;
my $word = defined($_=shift) ? $_ : die "usage: $0 <word>\n"; 
chomp($word);

($word) = $word =~ /^(\w+)$/ or die 'Bad User!';
my @chars = split //, $word;

my $speller ="spelled";

sub spelled {
    my $word = shift;
    my $out=`$speller $word`;
    chomp($out);
    if ( length($out) > 0 ) {
        print "[$out]\n";
    }
}



sub swap {
    my $i = shift;
    my $j = shift;
    my $A = shift; # ref to array
    my $tmp = @{$A}[$i];
    @{$A}[$i] =  @{$A}[$j];
    @{$A}[$j] =  $tmp;
}

sub permute {
    my $i = shift;
    my $n = shift;
    my $T = shift; #ref to array

    if ( $i == $n ) {
                my $word = join( "", @{$T} );
        spelled( $word );
    } else {
        for my $j ( $i..$n) {
            swap( $i-1, $j-1, $T );
            permute( $i+1, $n, $T );
            swap( $i-1, $j-1, $T );
        }
    }
}
permute (1,$#chars+1,\@chars);
[download]

The above script use this shell script.

$ cat spelled
#!/bin/sh
ASPELL=/usr/bin/aspell
GREP=/usr/bin/grep
EGREP=/usr/bin/egrep
for i in $*
do
        correct=`echo $i | $ASPELL -a | $GREP -v '^@' | $GREP -v '^&' 
+| $EGREP  -v '^$'`
        if [ ! -z "$correct" ] 
        then
                echo "$i"
        fi
done
[download]

Here's some sample output:

$ jumble.pl cat
[cat]
[act]
[download]