Proposal of new Object design

[jedel1043]

This PR closes #591.

Currently some tests fail because we aren't considering the many overrides of Object internal functions that ExoticObjects do.

The discussion of this issue is in #591, and the proposed solution is to implement every override that the spec requires directly inside Object, considering every exotic and special object implementation. This works, but it's a little detrimental for the modularity of the design.

This PR is an alternative design based on dynamic function pointers that solves the same problem.
The rationale on why I used function pointers is in this line:

boa/boa/src/object/internal_methods/mod.rs

Lines 599 to 600 in 4114e39

	} else if proto.borrow().data.internal_methods.__get_prototype_of__ as usize
	!= ordinary_get_prototype_of as usize

the spec states: "If p.[[GetPrototypeOf]] is not the ordinary object internal method defined in 10.1.1, set done to true.", and I couldn't find a way to express this check without using function pointers.

Please let me know what you think can be improved.

[Razican]

Test262 conformance changes:

Test result	master count	PR count	difference
Total	78,897	78,897	0
Passed	27,936	27,936	0
Ignored	15,616	15,616	0
Failed	35,345	35,345	0
Panics	0	0	0
Conformance	35.41%	35.41%	0.00%

[jedel1043]

Not gonna lie, I expected a lot worse benckmarks with this new design, but there are some small gains and moderate losses. Maybe I overlooked some optimization so I'm still going to try to lower the perf diff.
@HalidOdat could I hear your opinion on this design? Should we sacrifice performance for api ergonomics?

[jedel1043]

Ok, I made some optimizations and now the benchmarks have very few moderate regressions and some small to moderate improvements, so maybe we don't need to sacrifice performance with this design.

[jedel1043]

@HalidOdat Rebased 😁

[jedel1043]

Test262 conformance changes:

Test result	master count	PR count	difference
Total	80,685	80,685	0
Passed	31,809	31,821	+12
Ignored	15,818	15,818	0
Failed	33,058	33,046	-12
Panics	2	2	0
Conformance	39.42%	39.44%	+0.01%

Fixed tests (12):

test/built-ins/Object/keys/15.2.3.14-5-11.js [strict mode] (previously Failed)
test/built-ins/Object/keys/15.2.3.14-5-11.js (previously Failed)
test/built-ins/Object/create/properties-arg-to-object-non-empty-string.js [strict mode] (previously Failed)
test/built-ins/Object/create/properties-arg-to-object-non-empty-string.js (previously Failed)
test/built-ins/Object/assign/Source-String.js [strict mode] (previously Failed)
test/built-ins/Object/assign/Source-String.js (previously Failed)
test/built-ins/Object/entries/primitive-strings.js [strict mode] (previously Failed)
test/built-ins/Object/entries/primitive-strings.js (previously Failed)
test/built-ins/Object/values/primitive-strings.js [strict mode] (previously Failed)
test/built-ins/Object/values/primitive-strings.js (previously Failed)
test/built-ins/String/numeric-properties.js [strict mode] (previously Failed)
test/built-ins/String/numeric-properties.js (previously Failed)

[jedel1043]

Ok, this should be good now

[Razican]

Wow impressive work. I'll continue the review later, as this is a huge diff!

[jedel1043]

@Razican Ok, did a big documentation cleanup. Now every internal method and exotic internal method is documented. Also did a cleanup of some todos. Let me know it there's something else missing 😃

[Razican]

Looking good! I noticed there is this file that should probably not be committed: boa/my_trace.mm_profdata. Maybe we could add the extension to the gitignore?

I was also missing a link to the spec, but for the rest, it's ready to be merged :)

[jedel1043]

Looking good! I noticed there is this file that should probably not be committed: boa/my_trace.mm_profdata. Maybe we could add the extension to the gitignore?

Oops, let me add it to the .gitignore.

[Razican]

Everything looks great now!