repl: convert to ES2015 class

[avivkeller]

This PR modifies the REPL to use an ES5 class.

It will (hopefully) be one of many in series to refactor the REPL's code, as IMO it is currently hard to read as is. My plan is to first convert the REPL to ECMAScript classes, and then move away from using the the domain module. IMO keeping the internal codebase up-to-date with deprecations and additions is a great way to maintain longevity.

[avivkeller]

Right now, despite this change, most of the logic for the REPL remains in the constructor. I plan to move those into class methods (but not all at once - as that broke things).

[avivkeller]

CC @nodejs/repl

[codecov]

Codecov Report

Attention: Patch coverage is 94.02439% with 49 lines in your changes missing coverage. Please review.

Project coverage is 87.62%. Comparing base (5a3c605) to head (67a09cf).
Report is 41 commits behind head on main.

Files with missing lines	Patch %	Lines
lib/repl.js	94.02%	47 Missing and 2 partials ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #54838 +/- ##
========================================
Coverage 87.61% 87.62% ========================================
Files 650 651 +1 Lines 182945 183326 +381 Branches 35394 35445 +51 ========================================
+ Hits 160282 160632 +350 - Misses 15918 15953 +35 + Partials 6745 6741 -4 

Files with missing lines	Coverage Δ
lib/repl.js	94.06% <94.02%> (-0.06%)	⬇️

... and 33 files with indirect coverage changes

[jasnell]

I'm not against this change but I think we need to take it through a deprecation cycle first given the breaking change.

[aduh95]

Typo in the commit title s/ES5/ES2015/: "ECMAScript [5] does not use classes such as those in C++, Smalltalk, or Java." (source: https://262.ecma-international.org/5.1/#sec-4.2.1). The class keyword is listed as "Future Reserved Words" in ES5, see https://262.ecma-international.org/5.1/#sec-7.6.1.2

[avivkeller]

semver-major, as this is a breaking change. See #54842 for the deprecation, which should land in a release line before this.

[marco-ippolito]

semver-major, as this is a breaking change. See #54842 for the deprecation, which should land in a release line before this.

doc deprecation needs to be followed by a runtime deprecation (emit a warning) which is semver major. The runtime deprecation needs to land before refactoring and removing the constructor without new which is a EOL deprecation.

[atlowChemi]

@redyetidev We might want to consider landing this (possibly temporarily) with a solution similar to what some of the test-runner reporters implemented to solve this semver-major issue:

node/lib/test/reporters.js

Lines 34 to 42 in a65105e

	spec: {
	__proto__: null,
	configurable: true,
	enumerable: true,
	value: functionvalue(){
	spec??=require('internal/test_runner/reporter/spec');
	returnReflectConstruct(spec,arguments);
	},
	},

node/lib/test/reporters.js

Lines 52 to 60 in a65105e

	lcov: {
	__proto__: null,
	configurable: true,
	enumerable: true,
	value: functionvalue(){
	lcov??=require('internal/test_runner/reporter/lcov');
	returnReflectConstruct(lcov,arguments);
	},
	},

This would potentially allow landing this PR without needing to wait for the entire deprecation cycle to complete, and handle the deprecation separately

[avivkeller]

The issue with that is:

constreporters=require("node:test/reporters");(newreporters.spec())instanceofreporters.spec// false

---

Sidenote: This is blocked by #54869

[avivkeller]

Closing until the runtime deprecation is done (which won't be for a while)