use Benchmark qw(cmpthese);

my @strings = (
    'Some string about this long or so, maybe this long',
    'I like pizza this long or so, maybe this long',
    'this long or so, maybe this French Fries long',
    'This Sugar Rush Rocks. maybe this do not stop the clock.',
);

my %hash = (
    'Sugar Rush Rocks' => 'whatever',
    'long'             => 'itsgood',
    'this long'        => 'ilikeit',
    'maybe this'       => 'itsokay',
    'Some String'      => 'loooveit'
);

my @keys_sorted_by_length_desc =
  sort { length $b <=> length $a } keys %hash;

cmpthese(
    -5,
    {
        'Regex' => \&use_regex,
        'Index' => \&use_index,
    }
);

sub use_regex {
    foreach my $string (@strings) {
        foreach my $key (@keys_sorted_by_length_desc) {
            my $key_re = quotemeta($key);

            if ( $string =~ /(:?\A|\s)($key_re)\s*/i ) {
                #print "Found '$key' in '$string'\n";
                last;
            }
        }
    }
}

sub use_index {
    foreach my $string (@strings) {
        foreach my $key (@keys_sorted_by_length_desc) {
            my $lcstring = lc $string;
            my $lckey    = lc $key;
            if ( ( index $lcstring, $lckey ) > -1 ) {
                #print "Found '$key' in '$string'\n";
                last;
            }
        }
    }
}

__END__
         Rate Regex Index
Regex  7539/s    --  -86%
Index 52227/s  593%    --