Level 2 exam — Node.js

Exam — fresh questions

12 new questions: recall (25%) + application (50%) + arbitration (25%) on the Node.js core.
Answer everything, then click Check: your score and explanations will appear.

2.1 · Recall

1. What is a Node native module?

A project configuration file A package to install via npm before each use A built-in library, imported by name with no installation A browser plugin for running JavaScript

fs, path, os, http… ship with Node: require('fs') is enough, zero installation.

2.1 · Application

2. Salim reads reports on a server used by the whole team. Which fs version?

The sync version everywhere, easier to read A copy-paste of the file before each read A manual on-screen read The async version (fs.promises + await), non-blocking

On a server, always async: Sync blocks the event loop for every client.

2.1 · Arbitration

3. A colleague builds paths with __dirname + '/data/' + name. Salim's verdict?

Reject it: not portable (separator / vs \) and fragile — use path.join Accept it: it's the recommended method Replace it with hard-coded absolute paths Print it to check by eye

path.join picks the right separator for the system and handles edge cases — never concatenate.

2.2 · Recall

4. What do .on() and .emit() do on an EventEmitter?

.on() triggers and .emit() listens, in that order .on(name, callback) registers a listener, .emit(name, data) triggers it Both write to a log file Both install an npm package

.on registers the reaction, .emit triggers it with data — decoupled by name.

2.2 · Application

5. Salim creates an emitter for his import progress. What does he add first?

A loop that re-emits each event ten times A setTimeout delaying every emission A console.log instead of any emission A listener on the error event (otherwise, crash on the first problem)

error is special: emitted with no listener, it crashes the program — always listen for it.

2.2 · Application

6. The team wants three independent reactions to each imported file (display, log, counter). How?

Add three .on listeners on the same event, without touching the emitter Modify the import to call the three functions directly Merge the three displays into one text file Poll the import in a loop every second

Adding behaviour = adding a listener; the emitter stays unchanged.

2.3 · Recall

7. What is the principle of a stream?

Loading all data into memory before starting Processing data in chunks, as it flows Printing data before processing it Systematically compressing every file

In chunks, as it flows: constant memory whatever the total size.

2.3 · Application

8. Salim copies a 2 GB export. Which line does the job without saturating RAM?

A readFile followed by a writeFile of the full content A copy-paste of the file in the explorer fs.createReadStream('source').pipe(fs.createWriteStream('dest')) A loop concatenating the whole file into a string

.pipe() flows chunks from reading to writing, with flow control.

2.3 · Arbitration

9. A colleague wants to load everything into memory “because RAM will cope”. Verdict?

Good idea: it's always faster Acceptable for small scripts only Replace with paper printouts Reject for large volumes: saturation, slowness, crashes — streams process in chunks

Large volumes = streams (constant memory); readFile stays perfect for small files.

2.4 · Application

10. Salim creates his first native server. Which two lines are the foundation?

http.createServer((req, res) => {...}) then .listen(3000) An e-mail sent to each visitor A static HTML file opened by double-clicking it A spreadsheet shared with the team

createServer receives (req, res) on each request; listen opens the listening port.

2.4 · Application

11. The GET /rapports route must return the inventory as JSON. What does Salim do?

He returns an empty 200 in all cases res.statusCode = 200, Content-Type application/json, res.end(JSON.stringify(data)) He returns HTML with no header and no status He leaves the request unanswered

The right status + the right Content-Type + the serialised body, finished with end().

2.4 · Arbitration

12. Should you skip the native server to go straight to Express?

No: using Express in production is forbidden Yes: the native server is strictly useless No: the native server reveals the foundation (req/res, routing, statuses) and avoids the black box Yes: Express has nothing to do with the http module

Express rests on http: the native server shows what it simplifies and helps debugging.

What's next?

Score ≥ 80%: move on to Level 3 — Express, routes, middlewares and databases, on the native foundation gained here.
Score 60-79%: revisit the failed sections (2.1 async, 2.2 events, 2.3 streams, 2.4 req/res) and redo the 48h challenges.
Score < 60%: rebuild Salim's project step by step (inventory, events, archive, server), then retake the test.
In any case: keep a native reflex — async on servers, portable paths, listened errors, large volumes as streams.