feat(lib/es2020): Add Promise.allSettled(…)
#34065
Conversation
|
LGTM |
|
I guess this code can be simplified similarly to #33707 ? |
|
Nope, typings for node still need to support 2.1 until further notice. |
|
I first need #33707 to be merged, so that I can get access to the |
src/lib/es2020.promise.d.ts
Outdated
| * @param values An array of Promises. | ||
| * @returns A new Promise. | ||
| */ | ||
| allSettled<T extends readonly any[] | readonly [any]>(values: T): Promise<{ -readonly [P in keyof T]: PromiseSettledResult<T[P] extends PromiseLike<infer U> ? U : T[P]> }>; |
There was a problem hiding this comment.
Should we make these readonly unknown[] | readonly [unknown]? (@RyanCavanaugh @rbuckton)
src/lib/es2020.promise.d.ts
Outdated
| @@ -0,0 +1,29 @@ | |||
| interface PromiseResolvedResult<T> { | |||
There was a problem hiding this comment.
Hi. Drive-by code review here. Shouldn’t this be named something like PromiseFulfilledResult to adhere to the states and fates terminology? I think the reason calling it PromiseResolvedResult is incorrect here is that while all fulfilled promises are resolved, not all resolved promises are fulfilled. A promise can be resolved to a pending promise, in which case it has not settled, and it can also be resolved to a rejected promise, in which case it has not fulfilled.
If my interpretation of states and fates terminology is correct, shouldn’t the correct name of this interface be PromiseFulfilledResult?
ExE-Boss
force-pushed
the
lib/es2020/add-promise-allsettled
branch
from
ec3628c to
2929aad
Compare
|
Thank you! |
|
Still feel like it should've used |
|
LGTM |
Depends on #33707(I’ve now inlined theAwaited<T>type)Fixes #31083
Closes #31085
review?(@Kingwl)