Perl regular expression question

I would like to use Perl split function on comma delimited lines and capture onl开发者_开发技巧y two or more words between the comma delimited into an array. Single words are not needed.

For example, this line -->> aaa, ccc ddd, eee, fff ggg uuu, I only want, ccc ddd and fff ggg uuu

while(<FH>)
{
   @ = split(/,/);  
}

demo script:

my @data = (
     'aaa, ccc ddd, eee, fff ggg uuu'
   , ' aaa bbb ,ccc,eee,fff ggg uuu '
   , 'aaa,ccc,eee,fff'
);
for my $line (@data) {
  printf "|%s| ==> \n", $line;
  $line =~ s/^\s+|\s+$//g;
  my @cut = grep { / / } split( /\s*,\s*/, $line );
  printf "|%s|\n\n", join( '|', @cut );
}

output:

|aaa, ccc ddd, eee, fff ggg uuu| ==>
|ccc ddd|fff ggg uuu|

| aaa bbb ,ccc,eee,fff ggg uuu | ==>
|aaa bbb|fff ggg uuu|

|aaa,ccc,eee,fff| ==>
||

less terse than Dallaylaen's solution, but may make it easy to check for/handle special cases.

UPDATE: Added " aaa," protection, and ", ," protection as well.

$line =~ s/^\s+|\s+$//g; # or you get false positives 
my @multiword = grep {/\s/} split /\s*,[,\s]*/, $line;

The split will eat up all space around commas, so anything in the array that contains a space is a multi-word.

It's as simple as:

(undef, undef, @parsedvalues) = split /,/

You esentialy throw away first two values you get from split.