Incremental Refactoring

[thephoenixofthevoid]

I found out that many methods exist actually in 2 different implementations: for wildcard emitter and classic one, but for no reason got merged into single function. I am going to:

1. Separate implementations into two sets of functions
2. Make the module an ES6 module that supports tree-shaking
3. Improve quality of the code

I am starting this PR to show progress in real time and also discuss these changes. I don't insist on merging, but if you find this useful, it would be great.

[thephoenixofthevoid]

With given changes the performance of wildcard version doubled, but ordinary emitter's performance decreased. This is because I temporary removed many micro optimizations that made it hard to refactor the code. I plan to add them back at some point.

> eventemitter2-es@6.4.2 benchmark /home/feanor/Documents/ProgrammingResearch/Lvl60/EventEmitter2
> node test/perf/benchmark.js

Platform: linux, x64, 7878MB
Node version: v13.11.0
CPU: 1 x Intel(R) Core(TM) i5-8250U CPU @ 1.60GHz @ 3129MHz
 1 x Intel(R) Core(TM) i5-8250U CPU @ 1.60GHz @ 3342MHz
 1 x Intel(R) Core(TM) i5-8250U CPU @ 1.60GHz @ 3263MHz
 1 x Intel(R) Core(TM) i5-8250U CPU @ 1.60GHz @ 3040MHz
 1 x Intel(R) Core(TM) i5-8250U CPU @ 1.60GHz @ 3168MHz
 1 x Intel(R) Core(TM) i5-8250U CPU @ 1.60GHz @ 3294MHz
 1 x Intel(R) Core(TM) i5-8250U CPU @ 1.60GHz @ 3326MHz
 1 x Intel(R) Core(TM) i5-8250U CPU @ 1.60GHz @ 3198MHz
----------------------------------------------------------------
EventEmitterHeatUp x 8,168,953 ops/sec ±0.35% (92 runs sampled)
EventEmitter x 7,667,663 ops/sec ±0.55% (93 runs sampled)
EventEmitter2 x 21,699,789 ops/sec ±0.41% (95 runs sampled)
EventEmitter2 (wild) x 539,570 ops/sec ±0.28% (98 runs sampled)
EventEmitter2 (wild) using plain events x 1,048,870 ops/sec ±0.21% (95 runs sampled)
EventEmitter2 (wild) emitting ns x 3,271,452 ops/sec ±0.38% (92 runs sampled)
EventEmitter2 (wild) emitting a plain event x 5,379,542 ops/sec ±0.20% (97 runs sampled)
EventEmitter2Es x 11,466,351 ops/sec ±0.46% (93 runs sampled)
EventEmitter2Es (wild) x 934,198 ops/sec ±0.35% (97 runs sampled)
EventEmitter2Es (wild) using plain events x 1,945,518 ops/sec ±0.38% (94 runs sampled)
EventEmitter2Es (wild) emitting ns x 2,935,073 ops/sec ±0.47% (94 runs sampled)
EventEmitter2Es (wild) emitting a plain event x 4,538,065 ops/sec ±0.25% (97 runs sampled)
EventEmitter3 x 19,735,018 ops/sec ±1.09% (91 runs sampled)

Fastest is EventEmitter2

[SukkaW]

Will the PR being continued?

[thephoenixofthevoid]

Yeah, I am interested in continuing, but It's important to discuss changes. The project is much depended upon, changes I have proposed so far are massive. Long story short: need feedback.