http: fix close return value mismatch between doc and implementation

[kylo5aby]

Fixes: #51787

[nodejs-github-bot]

Review requested:

• @nodejs/crypto
• @nodejs/http
• @nodejs/net

[ShogunPanda]

We need a test for this.

@nodejs/http I'm a little uncertain on this: since the docs state this should have returned a instance rather than void, is this change semver-major or not (and in that case minor or patch)?

[climba03003]

I'm a little uncertain on this: since the docs state this should have returned a instance rather than void

I am not the one you are asking, but from my point of view, it is unintentional change (so, it should be patch) comes from #41263

The behavior of net.Server is not changed until now.

node/lib/net.js

Lines 2236 to 2273 in 17187dd

	Server.prototype.close = function(cb) {
	if (typeof cb === 'function') {
	if (!this._handle) {
	this.once('close', function close() {
	cb(new ERR_SERVER_NOT_RUNNING());
	});
	} else {
	this.once('close', cb);
	}
	}
	if (this._handle) {
	this._handle.close();
	this._handle = null;
	}
	if (this._usingWorkers) {
	let left = this._workers.length;
	const onWorkerClose = () => {
	if (--left !== 0) return;
	this._connections = 0;
	this._emitCloseIfDrained();
	};
	// Increment connections to be sure that, even if all sockets will be closed
	// during polling of workers, `close` event will be emitted only once.
	this._connections++;
	// Poll workers
	for (let n = 0; n < this._workers.length; n++)
	this._workers[n].close(onWorkerClose);
	} else {
	this._emitCloseIfDrained();
	}
	return this;
	};

[kylo5aby]

My perspective is that return instance from net.server.close() is possibly for the sake of facilitating chaining call, based on this, http.server.close() should perhaps maintain the same behavior.

[kylo5aby]

@ShogunPanda I have added test cases, and retriggered git Action, but there occurs errors on Windows/macOS as shown below, I think it's doesn't because wrongly run the test cases, can you take a look?

[ShogunPanda]

@ShogunPanda I have added test cases, and retriggered git Action, but there occurs errors on Windows/macOS as shown below, I think it's doesn't because wrongly run the test cases, can you take a look?

@climba03003 Guess who introduced that patch? XD I agree, it's a involuntary change. I would go for semver-patch.

@nodejs/build I don't think it's related to this PR. Can you please investigate?

[kylo5aby]

@ShogunPanda I have added test cases, and retriggered git Action, but there occurs errors on Windows/macOS as shown below, I think it's doesn't because wrongly run the test cases, can you take a look?

@climba03003 Guess who introduced that patch? XD I agree, it's a involuntary change. I would go for semver-patch.

@nodejs/build I don't think it's related to this PR. Can you please investigate?

Yeah, maybe this because of git action error instead of this PR or build problem

[mcollina]

lgtm

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/57344/

[nodejs-github-bot]

Landed in 60ce078