Sure:

my $rx = qr/
             ^            # At the beginning of the string  
             (            # Capture in $1
               (?:        # Paren for grouping (don't capture in $2)
                 \w       # match a "wordy" character
                 (?:      # paren for grouping,
                   \s+|\z # match at least one space or the end of the
+ string
                 )
               )*         # zero or more times
              )           # end of capture $1
            (.*)          # and match what comes after storing it in $
+2
          /x              # allow these comments
;
[download]

The rationale is to match at the beginning all the one-character strings, storing them in $1. The rest of the input will be catched in $2