Question

I’m trying to touch all potential dealer hands in blackjack, but when I kept blowing the stack, I realized things weren’t as depth first as expected. So I tried similar code in ruby and it performed differently.

this code,

(def d [2 3 4 5 6 7 8 9 10 10 10 10 11])
(defn dig1 [lst depth tot]
    (do 
        (print depth)
        (if (< tot 17) (map #(dig1 (conj lst %) (+ depth 1) (+ tot %)) d)) ))

(dig1 [0] 0 0)

produces: 011111111111112222222222222…

I expected map to execute the function on d[0] and dig down rather than to see everything executed at a given level. I obviously don’t understand what’s going on. Do I need to make something lazy(er)? map produces lazy sequences, but apparently chunks them in groups of 32.

in contrast,

@d = [2,3,4,5,6,7,8,9,10,10,10,10,11]
def dig(lst, depth, tot)
  p depth
  @d.map{|e| dig(lst.dup.push(e),depth+1,tot+e)} if tot < 17
end

produces what I would expect: 0123456789999999999999888888888888

If anyone could tell me how to make the clojure output look like the ruby output, I’d appreciate it.

Thanks, John

Answer 1

Generally map isn’t used when you don’t want the returned values back and are only evaluating the sequence for side effects. Something like doseq is preferable.

(def d [2 3 4 5 6 7 8 9 10 10 10 10 11])
(defn dig1 [lst depth tot]
  (print depth)
  (when (< tot 17)
    (doseq [i d]
      (dig1 (conj lst i)
            (inc depth)
            (+ tot i)))))
(dig1 [0] 0 0)

Produces: 012345678999....