Correct data errors in certain CSS interfaces.

[jpmedley]

The corrects a few errors with some CSS interfaces.

None of them have @@iterator which must be present in the the IDL. They are in neither the spec, nor Chrome's implemention of it:

• CSSNumericArray
• CSSTransformValue
• CSSUnparsedValue

The setters and getters listed in the IDL were misinterpreted as functions. For example, instead of there being a method named CSSNumericArray.CSSNumbericValue there is actually the ability to treat an instance of the interface as an array. In other words, CSSNumbericArray[n] would be a valid programming reference.

[jpmedley]

cc: @rachelandrew

[ddbeck]

None of them have @@iterator which must be present in the the IDL.

Do iterable declarations not imply the @@iterator symbol? It looks like all three have the declaration.

This possibly relates to #6367.

[foolip]

Do iterable declarations not imply the @@iterator symbol? It looks like all three have the declaration.

These three interfaces all have iterable<T> declarations in https://drafts.css-houdini.org/css-typed-om/, and https://heycam.github.io/webidl/#es-iterable is what says that results in a @@iterator property being added. (A bunch of other things are also implied, and it's worth knowing that which are different for iterable<T> and iterable<K, V>.)

[ddbeck]

Hi @jpmedley, do you plan to come back to this PR?

[jpmedley]

I shouldn't have removed @@iterator. The reason I thought it should be removed is because I'm still confused about. It doesn't look to me like you ever write this in a script. What you actually write is instanceVariable[Symbol.iterator](). Is so, then that's what needs to be in the BCD.

If we need to discuss that first, I'll happily put @@iterator back for now.

[ddbeck]

@jpmedley If you want to change the way BCD names the iterator symbol, then I think we need to take that up under a separate issue. The @@ notation for well-known symbols seems conventional within BCD and on MDN (see Array's @@iterator for an example); if we're going to change one, we ought to be consistent about it.