Elm collections: How they work, and when to use them.

Most applications, in any language, requires the use of collections. Elm has several collection types built-in: Lists, Arrays, Sets and Dicts. When is it better to use a List instead of an Array, and when do you use Sets? In this talk, Robin explains how the different collections are implemented under the hood, with the hope that this knowledge will make it painfully obvious when to use the different collection types.

Transcript

1. Elm Collections How they work And when to use them [ ]
2. Who am i? • Robin Heggelund Hansen • Noteworthy projects: • elm-community/dict-extra • Skinney/elm-array-exploration
3. Purpose of this talk • Learn how the four core collection types (Lists, Arrays, Dicts and Sets) are implemented. • Because it’s fun! • It makes it clear which collection to use, in any given use case.
4. Warning The company, characters, and implementations portrayed in this talk are fictitious. No identification with actual persons (living or deceased), companies and implementations is intended or should be inferred.
5. List type List a
 = Empty
 | Cons a (List a)
6. List [] [3] [1, 2, 3] [2, 3] Empty Cons 3 Cons 2 Cons 1
7. List Empty Cons 3 Cons 2 Cons 1 Empty Cons 3 Cons 2 Cons 1
8. List Summary • Ideal when operating at the head of the list. • Ideal for iterating from left to right. • Ideal when the most important thing is to add elements. • Everything else is sub-optimal.
9. Dict type Dict comparable v
 = Empty
 | Node 
 comparable
 v
 (Dict comparable v)
 (Dict comparable v)
10. Dict Bananas 5 Apples 2 Oranges 3 Mango 3
11. Dict Bananas 5 Apples 2 Oranges 3 Mango 3
12. Dict Bananas 5 Apples 2 Oranges 3 Mango 3
13. Dict summary • Quick access by key. • No duplicate keys. • Sorted order. • Requires comparable keys.
14. Set type alias Set comparable
 = Dict comparable comparable
15. Set summary • Requires comparable values.
16. Array 1 0 2 4 3 6 5 7 0 1 0
17. Array 143 00 111 000 000 000 000 000 000 000 010 001 7 1 2 0 0 0 0 0 0 0 0
18. Array summary • Fast get/set at any index. • Fast when adding to the end. • Fast map, filter, foldl, foldr.
19. Summary • Use List when operating at the head, or when iterating through all elements from left to right. • Use Array when you need fast random access, operating at the tail, or when you require iteration from right to left. • Use Dict when you need to access elements by key. • Use Sets when you need to operate with unique values (can also use dictionary).
20. Summary • Remember. Elm tries to help you by not exposing non-optimal API (like get/set for Lists). • Try to avoid converting between collection types. Compromise on the best collection for your use case.
21. Questions?