Implement property accessors

[tofpie]

This Pull Request closes #986.

It changes the following:

• Execution of property accessors

[codecov]

Codecov Report

Merging #987 (cef486e) into master (23bc476) will increase coverage by 0.24%.
The diff coverage is 80.18%.

@@            Coverage Diff             @@
##           master     #987      +/-   ##
==========================================
+ Coverage   59.79%   60.03%   +0.24%     
==========================================
  Files         167      167              
  Lines       11087    11134      +47     
==========================================
+ Hits         6629     6684      +55     
+ Misses       4458     4450       -8     

Impacted Files	Coverage Δ
boa/src/property/mod.rs	56.48% <ø> (+3.05%)	⬆️
...ax/ast/node/declaration/arrow_function_decl/mod.rs	60.86% <ø> (ø)
...c/syntax/ast/node/declaration/function_expr/mod.rs	61.29% <0.00%> (ø)
boa/src/context.rs	65.69% <41.93%> (+0.93%)	⬆️
boa/src/value/display.rs	85.00% <60.00%> (-3.64%)	⬇️
boa/src/builtins/map/mod.rs	72.02% <66.66%> (-0.44%)	⬇️
boa/src/environment/object_environment_record.rs	23.72% <66.66%> (+2.67%)	⬆️
boa/src/syntax/ast/node/object/mod.rs	58.92% <74.19%> (+21.78%)	⬆️
boa/src/object/gcobject.rs	65.00% <78.57%> (-0.24%)	⬇️
boa/src/builtins/object/mod.rs	67.26% <80.00%> (-0.41%)	⬇️
... and 29 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 23bc476...8a7ac4e. Read the comment docs.

[tofpie]

@RageKnify I continue the conversation started on the issue.

In fact, property accessors are implemented both in Value::get/set_field and GcObject::get/set. As this PR has modified all the calling sites for Value::get/set_field (adding context and ?), I could also replace all calls to Value::get/set_field by to_object(..)?.get/set, it would not be such a big effort. What do you think?

[HalidOdat]

Looks good!

Check my comments, some are unrelated but it would be nice to change in this PR :)

[RageKnify]

@RageKnify I continue the conversation started on the issue.

In fact, property accessors are implemented both in Value::get/set_field and GcObject::get/set. As this PR has modified all the calling sites for Value::get/set_field (adding context and ?), I could also replace all calls to Value::get/set_field by to_object(..)?.get/set, it would not be such a big effort. What do you think?

Your idea is that it will be a simple transition in the future, right? Makes sense.

We will also eventually remove Value::set_field and Value::get_field since only GcObject will have that kind of functionality, right?

[tofpie]

@HalidOdat said it is not as simple as I initially thought:

Maybe I could change all the calls to Value::get/set_field to to_object(..)?.get/set as you have suggested in #577. Now that context has been added to get/set_field, the bulk of the work would be a search and replace. What do you think?

It's not what needs to be done .to_object() does some other things and would not be spec compliant, the calls to .get/.set should be reinforced by the type system as it does not make much sense to have set/get for non-object (Value type) values as explained in #577.

we can merge this as it's main purpose was to add property accessors, and replace .get_field and .set_field builtin function by builtin function, through incremental changes (with many PRs), Allowing .get_field/set_field to take context is the first step, the end goal should be to completely remove them, it's really hard otherwise.

I guess I will leave as it is, just implementing the suggestions and add some unit tests.

[HalidOdat]

Looks good to me!

Once this is merged we can start removing the .set_field and .get_field functions from builtin functions :)

[RageKnify]

Some minor things can be cleaned up and I have some questions on Value::new_object current state.

[HalidOdat]

Test262 conformance changes:

Test result	master count	PR count	difference
Total	78,493	78,493	0
Passed	19,787	20,855	+1,068
Ignored	15,585	15,585	0
Failed	43,121	42,053	-1,068
Panics	396	394	-2
Conformance	25.21	26.57	+1.36%

The conformance increase is really nice :)