Optional Chaining Operator (Stage 1)

[hzoo]

Champion: @gisenberg
Spec Repo: https://github.com/gisenberg/proposal-optional-chaining
Stayed at Stage 1 at the July 2017 meeting: https://docs.google.com/presentation/d/1OcytQtyykmOZJwm-LFgOP0FpQccyeajAjLdOvBki9a0/
Stayed at Stage 1 at the Sept 2017 meeting: https://docs.google.com/presentation/d/1iiCtJSW42Z7lg0YlagOZOfI-FTrkN88OuiO3pySawSo/edit?usp=sharing

Syntax

a?.b

Implementation

• Optional Chaining Operator (Stage 1) babel#5813
• As part of https://github.com/babel/babel/releases/tag/v7.0.0-alpha.15
• Test https://babeljs.io/7/

[littledan]

Note that the issue of contention at TC39 was the "short-circuiting" semantics. See this issue for details: tc39/proposal-optional-chaining#10 . Previously, the specification was option 4. The specification was changed to option 2 based on online feedback.

The committee seemed to be advocating option 1, but it's unclear if they had a good understanding of the CoffeeScript data or the way other programming languages work here.

These differences matter for both parsing and the transform, though there may be some design that handles the parsing in a single way among all options.

Maybe Babel could implement multiple of these options and become a place where the differences could be explored interactively by programmers.

[hzoo]

@littledan if it moves the proposal forward (worth implementing both for someone) we could have options - to convince people one is better than the other.

[ljharb]

The only concern there is that it turns a likelihood of someone shipping untranspiled code using eventually-wrong semantics into a certainty.

[hzoo]

@ljharb this is for stage-0, and we are happy to break whenever if we aren't sure of the semantics. Maybe in these cases we have a warning. I don't think we do need to implement both actually, just show the input/output

[littledan]

I don't see this as having a higher risk than many of the other proposals that Babel has implemented at Stage 1. I think @hzoo has done a good job communicating clearly that Stage 1 and 2 are highly unstable.

[littledan]

Anyway, I'm not sure if anyone would really use the various alternatives, it just might be an interesting way to get feedback.

[hzoo]

Yeah and if we do implement multiple versions, it's definetely not an intention in any way to be used other than for feedback for the proposal itself.

[morsmodr]

Is there any progress in this? Currently from here: https://github.com/babel/proposals I see that the Readme.md has not been updated in a long while. The proposal is in stage 4 which means it should be part of preset-env but as per the proposals link I shared, optional chaining is supposed to be in babel-preset-stage-1?

Is there a way to confirm this has indeed been added to preset-env?

[JLHwung]

it should be part of preset-env

Yes, optional-chaining is already included in @babel/preset-env 7.12. Please refer to TC39 proposals repo for latest status: https://github.com/tc39/proposals the Babel proposals repo may be out of date.

Closing this issue as it has been supported by @babel/preset-env