Convert NativeObject from IdScriptableObject to lambda constructor.

[aardvark179]

@gbrail I'm creating this as a draft PR for now. There are 24 test failures from doc tests due to small changes in the toString results for various functions on object, and I'm not sure if we should work to reduce those differences or if we're fine with changing the tests.

I'll update this PR with benchmark comparisons when I've had time to run the full suite.

[gbrail]

This is directionally the right thing to do, and the right way to do it. As you can see from the test results, there are a bazillion legacy behaviors that need to be fixed when making a change like this, unfortunately. FWIW I was going to tackle fundamental classes like this later, but since you already did most of the work it'd be great to figure out what's holding back those tests that are now failing.

Also, after that's all done, please run:

./gradlew :tests:test --tests Test262SuiteTest -DupdateTest262properties

to regenerate the properties file!

[aardvark179]

I’ve got benchmarks. Some better, some worse, but I think I understand why. Probably won’t get the test fixes done till the weekend.

[aardvark179]

@gbrail Everything is passing nicely except for one proxy test which looks at the result of toString on Proxy. I fixed the doc tests by making our lambdas functions look more like the old id functions when converted to strings, but that broke the NativeProxy test.

That would be simple to fix by just changing the test, but the better long term solution might be to make our function toString output match the 262 spec. I just didn't fancy change all the occurrences in the docs.

[rbri]

That would be simple to fix by just changing the test, but the better long term solution might be to make our function toString output match the 262 spec.

i guess #1300 is exactly that story. maybe we have to address this in a separate PR

[aardvark179]

@gbrail, @rbri Right, done the function toString changes in #1840, and rebased this PR on top of that.