sub weirdex {
  local $_ = shift;

  tr/a-zA-Z//dc;
  tr/a-zA-Z//s;
  m/[aeiou]/g and substr($_, pos) =~ s/[aieuo]//g;

  $_;
}

print "$_ - ", weirdex($_), $/
  for "giulienk", "larry wall", "etheroskedasticity";

__output__

giulienk - gilnk
larry wall - larywl
etheroskedasticity - ethrskdstcty
[download]

_________
broquaint

Thanks to yours and jmcnamara's reply: i totally forgot about the power of tr/// modifiers, especially c and s, which i never used before. :)


$|=$_="1g2i1u1l2i4e2n0k",map{print"\7",chop;select$,,$,,$,,$_/7}m{..}g

Nice. I was unhappy with your vowel stripping though. You use a looping match for offsets into $_ but really, you only have to find the *first* vowel and then no looping is required.

#  m/[aeiou]/g and substr($_, pos) =~ s/[aieuo]//g;
  /[aeiou]/ and substr( $_, $+[0] ) =~ tr/aeiou//d;
[download]

You use a looping match for offsets into $_ but really ...

Er, what looping? The /g matches the first vowel then saves the position of the match for substr which the replace then operates on. I didn't want to use the $+ variable because of the overhead it invokes.
HTH

_________
broquaint

    sub weirdex {
        local $_ = $_[0];

        tr/a-zA-Z//s;
        tr/a-zA-Z//cd;
        my $i; s/([aeiou])/$1if!$i++/eg;

        $_;
    }
[download]

--
John.

FYI, the algorithm sounds a lot like (heh) metaphone:

• Text::Metaphone
• Text::DoubleMetaphone
• Text::TransMetaphone

-- Rocco Caputo - rcaputo@pobox.com - poe.perl.org

use strict;
use Test::More 'no_plan';

sub weirdex {
        local $_ = shift;
        tr/A-Za-z//cd;
        tr/A-Za-z//s;
        /[aeiou]/g;
        substr($_, pos()) =~ s/[aeiou]+//g;
        return $_;
}

my %test = (
        'giulienk'           => 'gilnk',
        'larry wall'         => 'larywl',
        'etheroskedasticity' => 'ethrskdstcty',
);

my ($input, $result);

is(weirdex($input), $result, "$input") while ($input, $result) = each 
+%test;

__END__
ok 1 - etheroskedasticity
ok 2 - larry wall
ok 3 - giulienk
1..3
[download]

Makeshifts last the longest.

Y is a vowel, isn't it?

Sometimes, so it may not be necessary for this particular algorithm. It would also appear that two Welsh words in the dictionary use 'w' as a vowel.
HTH

_________
broquaint

Y and W are pseudo-vowels. They represent vowel sounds, but are not vowel letters. This is just one of the many contradictions between spoken and written language. I think W is a vowel in Croatian.

• All the solutions (except the original one) have the same problem: they don't care about upper-case vowels. All examples were in lower cases so nobody noticed it.
• The original implementation was wrong has it would consider consonant separated by a "non-first-vowel" as contigous ones and squash them if they are equal.
• The illustrious surname "Schwartz" with its consonant-to-vowel ratio of 7.0 render this algorithm pretty useless :)

$|=$_="1g2i1u1l2i4e2n0k",map{print"\7",chop;select$,,$,,$,,$_/7}m{..}g