USING: grouping io.encodings.ascii io.files kernel math
prettyprint sequences sets ;

"input.txt" ascii file-contents [ 4 ] [ 14 ] bi
[ tuck <clumps> [ all-unique? ] find drop + . ] 2bi@