Questions

12 of 38

1What is the Iterator Protocol in JavaScript? What two things must an object implement to be a valid iterator?

2What is the difference between an iterable and an iterator? Can something be both?

3What built-in JavaScript data structures are iterable by default? How does for...of work under the hood with them?

4What does Symbol.iterator do? How would you make a plain object iterable from scratch?

5What is a 'lazy' iterator and why is it important for performance? Give a practical example.

6How would you implement an infinite iterator (e.g., an infinite counter)? How do you safely consume it?

7What happens if you forget to return { done: true } in a custom iterator? What are the consequences?

8How would you implement a reusable iterable (one that can be iterated multiple times)?

9Can you chain or compose custom iterators? Implement a map and filter for a custom iterator without converting to an array.

10What is a generator function? How is it different from a regular function in terms of execution flow?

11Explain what yield does. What does calling .next() return before and after a yield?

12What is the difference between yield and return inside a generator?

13What does yield* do? How is it different from a regular yield?

14Can you pass a value into a generator using .next(value)? How does that work and what is the practical use case?

15Explain generator.return(value) and generator.throw(error). When would you use each in production?

16What happens to a try/finally block inside a generator when .return() is called externally?

17How do generators handle errors? How does .throw() interact with try/catch inside a generator?

18Are generators lazy? How do they differ from eager evaluation with arrays in terms of memory and performance?

19What is the difference between a regular iterator and an async iterator? What protocol does an async iterator follow?

20What does Symbol.asyncIterator do? How does for await...of use it?

21Implement an async generator that fetches paginated API data, yielding one page at a time:

22What are the pitfalls of using for await...of with an async generator that makes network calls? How do you handle errors and cancellation?

23How would you implement a readable stream as an async iterable in Node.js?

24How would you use a generator to implement redux-saga-style side effect management? What makes generators a good fit for this?

25How can generators be used to implement coroutines or cooperative multitasking in JavaScript?

26Implement a take(n) utility that takes the first n values from any iterable — including infinite ones:

27When would you choose a generator over returning an array? What are the memory trade-offs?

28In a data pipeline processing millions of records, how would you use generators to avoid loading everything into memory?

29What are the debugging challenges with generators (e.g., in stack traces and async flows)? How do you mitigate them?

30How do generators compose? Implement a pipeline of generator-based transformations (like RxJS operators but synchronous).

31What are the limitations of generators? What problems are they not a good fit for?

32What are Iterator Helpers (.map(), .filter(), .take(), .drop() on iterators natively)? What stage are they at in the TC39 proposal pipeline?

33How does Array.from() use the iterator protocol? What's the difference between passing an iterable vs an array-like object?

34How do Map, Set, Array, and String expose their iterators? Are they the same object or different?

35What is the difference between map.keys(), map.values(), and map.entries()? What do they return?

36How does destructuring and spread (...) use the iterator protocol internally?

37How would you implement Promise-based async/await using generators and a runner function? (This is essentially how Babel transpiled async/await early on.)

38How would you use a generator to implement a tree traversal (DFS) without recursion stack concerns?

12 / 38

What is the difference between yield and return inside a generator?

yield pauses the generator and can be called multiple times, while return ends the generator and sets done: true; further calls to next() return { done: true }.

yield is used to produce intermediate values without terminating the generator; you can have many yield statements. return terminates the generator; the returned value is the final value and done becomes true. If no return is present, the generator implicitly returns undefined. After a return, the generator cannot be resumed.

Difficulty: 6/10

Topics: generators, yield vs return, iteration control

Scenario Questions

0-2 years experience

1
Write a simple generator that yields numbers 1 to 3 and then returns the string 'done'. What will a for...of loop output when iterating over it?
2
If you call next() on a generator after it has executed a return statement, what value do you get back and why?

2-5 years experience

1
You have a function that streams lines from a large file using a generator. A teammate changed a yield to return to signal end-of-file, and the consumer now only gets the first line. Explain why this happened and how to fix it.
2
During debugging, you notice that a generator used in an async pipeline never releases a file handle. The code uses return inside a try...finally. How does return affect the generator’s cleanup compared to yield?

5-8 years experience

1
Design a data‑processing component that can pause, resume, and early‑exit based on back‑pressure, using generators. Discuss how you would use yield for incremental results and return for signalling completion or error, and the trade‑offs for memory usage.
2
In a high‑throughput server, you replace a hand‑rolled iterator with a generator to lazily fetch DB rows. How would you ensure that early termination via return correctly closes the DB cursor, and what pitfalls could arise if you mistakenly used yield for termination?

8+ years experience

1
Our platform is migrating legacy callback‑based streaming APIs to generator‑based pipelines across multiple services. As the lead architect, outline a strategy for handling the semantic shift from return‑based termination to yield‑based flow, considering backward compatibility, error propagation, and observability.
2
When designing a cross‑team library for composable data streams, you need to define a convention for when a generator should return a final value versus when it should yield a sentinel. What factors (performance, API ergonomics, debugging) influence this decision, and how would you document and enforce it?

Follow-up Questions

❓What happens if you omit the `return` statement entirely?
❓Can a generator have multiple `return` statements, and how does that affect iteration?
❓How does `yield*` interact with `return` in a delegated generator?