Replace `deepAccess` with optional chaining

[gwhigs]

Type of issue

Question

Description

I see mixed use of both optional chaining (obj?.prop1?.prop2) and the deepAccess function in Prebid.js.

Is optional chaining preferred? If so, would a PR replacing some usage of deepAccess be appropriate?

I have a list of such patches for certain modules and would like to submit them. However, my internal build target includes native support for optional chaining. I'm unsure whether this is the case more generally.

Replacing deepAccess entirely — if desirable — would be a larger effort considering how deeply integrated some usages are.

[dgirardi]

No objections to such PRs. All the browsers we target now have native support for it, and it's part of our "standard" build output, but we still have babel set up so that it would get transpiled away for those that wish to build for older browsers. I believe the reason for deepAccess is that the project is old enough that these features were missing or experimental even for babel at the time.

[patmmccann]

Keeping open for a linter warning suggesting the use of optional chaining