c:\@Work\Perl\monks>perl -wMstrict -MData::Dump -le
"my @ra = (
   'keyword1 data1 data2 data3',
   'keyword2 data1 data2 data3',
   '    data4 data5',
   '    data6',
   'keyword1 data1 data2 data3 data4',
   'keyword3 data1',
   );
 ;;
 my $rx_continuation = qr{ \A \s+ \w }xms;
 ;;
 my $current_line;
 LINE:
 for my $line (@ra) {
   if ($line =~ $rx_continuation) {
     $current_line .= $line;
     next LINE;
     }
   ;;
   _process($current_line) if $current_line;
   $current_line = $line;
   }
 ;;
 _process($current_line) if $current_line;
 ;;
 dd \@ra
 ;;
 ;;
 sub _process {
   print qq{'$_[0]' ->}, map qq{ '$_'}, split ' ', $_[0];
   }
"
'keyword1 data1 data2 data3' -> 'keyword1' 'data1' 'data2' 'data3'
'keyword2 data1 data2 data3    data4 data5    data6' -> 'keyword2' 'data1' 'data2' 'data3' 'data4' 'data5' 'data6'
'keyword1 data1 data2 data3 data4' -> 'keyword1' 'data1' 'data2' 'data3' 'data4'
'keyword3 data1' -> 'keyword3' 'data1'
[
  "keyword1 data1 data2 data3",
  "keyword2 data1 data2 data3",
  "    data4 data5",
  "    data6",
  "keyword1 data1 data2 data3 data4",
  "keyword3 data1",
]