Programmation Elixir : Gourmand vs. Paresseux : Des Choix Stratégiques dans le Traitement des Collections

En Elixir, le traitement des données suit deux philosophies distinctes : Gourmand (Avid) et Paresseux. Comprendre ce compromis est essentiel pour l'efficacité mémoire et la stabilité du système.

1. Le Protocole Enumerable

Techniquement, les éléments pouvant être itérés sont dits implémenter le protocole Enumerable. Ce contrat commun permet à différentes structures de données de fonctionner avec le même ensemble de fonctions.

2. Modules Gourmands vs. Paresseux

Le Enum module est gourmand. Il peut potentiellement consommer tout le contenu d'une collection immédiatement, créant des listes intermédiaires à chaque étape du pipeline. En revanche, le Stream module est paresseux. La valeur suivante est calculée uniquement lorsqu'elle est nécessaire.

3. Spécification vs. Résultat

Une valeur Stream est une spécification de ce que nous voulions, mais pas le résultat. Les streams sont énumérables et composables, vous permettant d'empiler des transformations sans exécuter de travail jusqu'à ce que vous transmettiez le stream à un « récepteur avide » comme Enum.to_list/1.

4. Pureté du Paradigme

Mélanger les paradigmes (fonctionnel et orienté objet) affaiblit les avantages apportés par une approche fonctionnelle. Privilégiez les transformations déclaratives aux boucles impératives pour plus de prévisibilité.

TERMINALbash — 80x24

> Ready. Click "Run" to execute.

QUESTION 1

What happens when you pass a collection to the Enum module?

It creates a lazy recipe for later execution.

It potentially consumes all contents immediately and returns a result.

It returns a Stream struct containing the function logic.

It converts the data into an object-oriented class.

QUESTION 2

Which statement best describes a 'Stream' in Elixir?

A realized list of integers.

A specification of intended work, not the result itself.

A faster version of Enum for small lists.

A specific type of Map used for key-value storage.

QUESTION 3

Why would you choose Stream over Enum for a 10GB text file?

Enum cannot read files.

Stream executes faster for small datasets.

Stream processes data line-by-line, preventing memory exhaustion.

Stream automatically sorts the file contents.

QUESTION 4

According to the lesson logic, what is a danger of mixing functional and OO paradigms?

It makes the code run significantly faster.

It dilutes the benefits of the functional approach.

It is required for the Enumerable protocol.

It allows Elixir to run on the JVM.

QUESTION 5

What is the requirement for a data type to be usable with Enum and Stream modules?

It must be a binary string.

It must implement the Enumerable protocol.

It must be smaller than 1MB.

It must be a recursive Keyword list.

Module Challenge: Prime Sequences & String Cleaning

Applying greedy and lazy transformations to complex data tasks.

You are tasked with generating mathematical sequences and cleaning messy user input. You must use list comprehensions and high-level string functions to achieve efficient results.

1. Using your previously written 'span' function (which returns a list from 2 to n), use list comprehensions to return a list of the prime numbers from 2 to n.

Solution:
Model Solution: elixir def primes_up_to(n) do range = span(2, n) for x <- range, is_prime?(x), do: x end defp is_prime?(2), do: true defp is_prime?(n) when n < 2, do: false defp is_prime?(n) do Enum.all?(2..round(:math.sqrt(n)), fn x -> rem(n, x) != 0 end) end This uses the span function as a generator and a filter predicate to identify primes.

2. [Writing Task] Write a function to capitalize the sentences in a string. Each sentence is terminated by a period and a space ('. '). Assume 150-word output capacity for the logic explanation.

Solution:
To capitalize sentences in a string where casing is currently random, follow these steps (approx 150 words of logic): First, use `String.split(input, ". ")` to break the string into a list of individual sentences. Note that the delimiter is lost in the split. Second, apply a transformation function to each element using `Enum.map`. Inside this map, use `String.capitalize/1`, which lowercases the whole string and then capitalizes the very first character. This ensures 'tHe dOg.' becomes 'The dog.'. Finally, use `Enum.join(". ")` to re-attach the sentences with the proper punctuation. Example code: elixir def capitalize_sentences(str) do str |> String.split(". ") |> Enum.map(&String.capitalize/1) |> Enum.join(". ") end

3. What occurs when you modify a printable character function example to use '99.0' (a float) where a charlist is expected?

Solution:
In Elixir, a charlist is a list of integers. If you provide '99.0', the system will raise a FunctionClauseError or a type error because '99.0' is a floating-point number, not an integer representing a codepoint. Charlist processing functions rely on the integer values (like 99 for 'c') to determine printability; floats do not satisfy the Enumerable protocol for lists or the bitstring requirements for binaries in this context.