Avatar of akirchner333

akirchner333's solution

to ETL in the Elixir Track

Published at Jun 16 2019 · 1 comment
Test suite


This exercise has changed since this solution was written.

We are going to do the Transform step of an Extract-Transform-Load.


Extract-Transform-Load (ETL) is a fancy way of saying, "We have some crufty, legacy data over in this system, and now we need it in this shiny new system over here, so we're going to migrate this."

(Typically, this is followed by, "We're only going to need to run this once." That's then typically followed by much forehead slapping and moaning about how stupid we could possibly be.)

The goal

We're going to extract some scrabble scores from a legacy system.

The old system stored a list of letters per score:

  • 1 point: "A", "E", "I", "O", "U", "L", "N", "R", "S", "T",
  • 2 points: "D", "G",
  • 3 points: "B", "C", "M", "P",
  • 4 points: "F", "H", "V", "W", "Y",
  • 5 points: "K",
  • 8 points: "J", "X",
  • 10 points: "Q", "Z",

The shiny new scrabble system instead stores the score per letter, which makes it much faster and easier to calculate the score for a word. It also stores the letters in lower-case regardless of the case of the input letters:

  • "a" is worth 1 point.
  • "b" is worth 3 points.
  • "c" is worth 3 points.
  • "d" is worth 2 points.
  • Etc.

Your mission, should you choose to accept it, is to transform the legacy data format to the shiny new format.


A final note about scoring, Scrabble is played around the world in a variety of languages, each with its own unique scoring table. For example, an "E" is scored at 2 in the Māori-language version of the game while being scored at 4 in the Hawaiian-language version.

Running tests

Execute the tests with:

$ elixir etl_test.exs

Pending tests

In the test suites, all but the first test have been skipped.

Once you get a test passing, you can unskip the next one by commenting out the relevant @tag :pending with a # symbol.

For example:

# @tag :pending
test "shouting" do
  assert Bob.hey("WATCH OUT!") == "Whoa, chill out!"

Or, you can enable all the tests by commenting out the ExUnit.configure line in the test suite.

# ExUnit.configure exclude: :pending, trace: true

If you're stuck on something, it may help to look at some of the available resources out there where answers might be found.


The Jumpstart Lab team http://jumpstartlab.com

Submitting Incomplete Solutions

It's possible to submit an incomplete solution so you can see how others have completed the exercise.


if !System.get_env("EXERCISM_TEST_EXAMPLES") do
  Code.load_file("etl.exs", __DIR__)

ExUnit.configure(exclude: :pending, trace: true)

defmodule TransformTest do
  use ExUnit.Case

  # @tag :pending
  test "transform one value" do
    old = %{1 => ["WORLD"]}
    expected = %{"world" => 1}

    assert ETL.transform(old) == expected

  @tag :pending
  test "transform more values" do
    old = %{1 => ["WORLD", "GSCHOOLERS"]}
    expected = %{"world" => 1, "gschoolers" => 1}

    assert ETL.transform(old) == expected

  @tag :pending
  test "more keys" do
    old = %{1 => ["APPLE", "ARTICHOKE"], 2 => ["BOAT", "BALLERINA"]}

    expected = %{
      "apple" => 1,
      "artichoke" => 1,
      "boat" => 2,
      "ballerina" => 2

    assert ETL.transform(old) == expected

  @tag :pending
  test "full dataset" do
    old = %{
      1 => ~W(A E I O U L N R S T),
      2 => ~W(D G),
      3 => ~W(B C M P),
      4 => ~W(F H V W Y),
      5 => ~W(K),
      8 => ~W(J X),
      10 => ~W(Q Z)

    expected = %{
      "a" => 1,
      "b" => 3,
      "c" => 3,
      "d" => 2,
      "e" => 1,
      "f" => 4,
      "g" => 2,
      "h" => 4,
      "i" => 1,
      "j" => 8,
      "k" => 5,
      "l" => 1,
      "m" => 3,
      "n" => 1,
      "o" => 1,
      "p" => 3,
      "q" => 10,
      "r" => 1,
      "s" => 1,
      "t" => 1,
      "u" => 1,
      "v" => 4,
      "w" => 4,
      "x" => 8,
      "y" => 4,
      "z" => 10

    assert ETL.transform(old) == expected
defmodule ETL do
  @doc """
  Transform an index into an inverted index.

  ## Examples

  iex> ETL.transform(%{"a" => ["ABILITY", "AARDVARK"], "b" => ["BALLAST", "BEAUTY"]})
  %{"ability" => "a", "aardvark" => "a", "ballast" => "b", "beauty" =>"b"}
  @spec transform(map) :: map
  def transform(input), do: Map.to_list(input) |> transform(%{})

  def transform([], acc), do: acc
  def transform([h | t], acc) do
    transform(t, Map.merge(acc, add_entries(h, acc)))

  defp add_entries({_, []}, acc), do: acc
  defp add_entries({value, [h | t]}, acc) do
    add_entries({value, t}, Map.merge(acc, %{String.downcase(h) => value}))

Community comments

Find this solution interesting? Ask the author a question to learn more.
Avatar of francescm
  • The Map.merge in transform/2 is useless, isn't it?
  • The Map.merge in add_entries/2 can just be a Map.put, that has a more steamlined syntax.

By the way: in order to convert a Map to something where you can apply pattern match, you can also just Enum.map that converts the Map in a Tuple behind the curtains.

What can you learn from this solution?

A huge amount can be learned from reading other people’s code. This is why we wanted to give exercism users the option of making their solutions public.

Here are some questions to help you reflect on this solution and learn the most from it.

  • What compromises have been made?
  • Are there new concepts here that you could read more about to improve your understanding?