Add Eliud's eggs exercise

[kahgoh]

This is one of the 48in24 exercises

[tasxatzial]

Would you mind moving its config entry after complex-numbers? There are two PRs open that modify the main config, and merging it as-is will create conflicts in both.

Also, adding 'basics' as a prerequisite would be nice to have.

[tasxatzial]

@kahgoh Since the instructions mention a bit-count functionality that doesn’t actually exist, I’ve added a small appendix to ensure there’s no confusion.

[kahgoh]

Hey @tasxatzial, I just had a look at the instructions.append.md. I thought the note in the instructions is aimed to discourage students from using a function that actually counts the number of 1 bits in the input number (which would solve entire exercise with just one function). For example, in Crystal, Integers have a popcount function, x86-64 assembly has a popcnt instruction and Python has bit_count that does the same. By the way, Java has Integer.bitCount. I think that's what we're trying to discourage them from.

I don't think we need stop students from using the other bitwise operations. For example, another strategy is to use bit shifting (to divide by two) and bitwise & to work out when to increment the count (for example in this C solution or this Python one).

[tasxatzial]

I see, so a solution that looks like so would be considered ok?

(defn egg-count [number]
  (->> (Integer/toBinaryString number)
       count
       range
       (map #(bit-test number %))
       (filter true?)
       count))

This converts to a binary string (via Java interop), constructs a list from 0 to the number of its digits - 1, maps bit-test over that list to check if the bit at each index is true or false, and finally counts the number of true values.

[tasxatzial]

Okay, I'll allow the bit-* functions for now and re-evaluate once community solutions start coming in.

Side note: This appears to be a great exercise for exploring different approaches.

[tasxatzial]

I’m also thinking that 'don't use bit-count functionality' could easily be interpreted as 'don't use bit-* and count functions,' since bit-count isn't a Clojure function. I'll be monitoring the solutions closely, but I'll most likely add an appendix to prevent confusion.

Then again, mentioning the bit-* functions might make people lean towards using them instead of thinking of alternative ways to solve the problem.

Forum thread:
https://forum.exercism.org/t/eliuds-eggs-instructions-restrictions-are-too-specific/13672