Haskell in REPL:

    Prelude> :m Data.List
    Prelude Data.List> let f xs = [(head ys, length ys) | ys <- group xs]
    Prelude Data.List> f "aaaabbbcca"
    [('a',4),('b',3),('c',2),('a',1)]

Of course using the "group" function instead of programming it myself is cheating a bit ...

#(mapv (juxt (comp str first) count) (partition-by identity %))

Surprising that F# doesn't have a "chunkBy" function for sequences; it makes the solution in Elixir vastly more readable than the version in the article:

    "aaaabbbcca" |> String.codepoints() |> Enum.chunk_by(& &1) |> Enum.map(&{hd(&1), length(&1)})

There is a chunkBy, it's just not this nice. This would be a great project for the F#+ guys.

I did it this way:

  map (head &&& length) . group

(Although in an actual full program, there would be other considerations (what considerations those are depends on the program being written), and in a full program I probably would not be using Haskell, but rather C.)

I attempted this as a Haskell beginner. Positive it can be improved upon.

    runLengthEncode :: String -> [(Char, Int)]
    runLengthEncode [] = []
    runLengthEncode input =
        let 
            chFirst = head input
            zipped = zip input $ chFirst : input
        in 
            rle zipped (chFirst, 0)
        where 
            rle :: [(Char, Char)] -> (Char, Int) -> [(Char, Int)]
            rle [] (ch, count) = [(ch, count)]
            rle ((ch1, ch2) : cs) (ch, count) = 
                if ch1 == ch2 then
                    rle cs (ch1, count + 1)
                else 
                    (ch2, count) : rle cs (ch1, 1)

meanwhile python be out here like:

  [(m.group(1), len(m.group()))
   for m in re.finditer(r"([a-z])\1*", “aaaabbbcca")

  import itertools
  [(c, len(list(it))) for (c, it) in itertools.groupby("aaaabbbcca")]

It's not the first time I've wanted

  def ilen(it):
    return sum(1 for _ in it)

as part of itertools. Which would make the second line be:

  [(c, itertools.ilen(it)) for (c, it) in itertools.groupby("aaaabbbcca")]

Rejected many times, eg, from 2010: https://bugs.python.org/issue9547 .

newlisp:

(set 'L '()) (dolist (s (explode "aaaabbbcca")) (if (!= (ref-all s L) '((0 0))) (push (list s 1) L) (inc (L 0 1)))) (print (reverse L))