Add an overload to Object.freeze that preserves literal types

[weswigham]

This way when you write something like

const LogType = Object.freeze({
	error: "error", // message, c style arguments
	warn: "warn", // message, c style arguments
	info: "info", // message, c style arguments
	log: "log", // message, c style arguments
	debug: "debug", // message, c style arguments

	trace: "trace", // no arguments

	group: "group", // [label]
	groupCollapsed: "groupCollapsed", // [label]
	groupEnd: "groupEnd", // [label]

	profile: "profile", // [profileName]
	profileEnd: "profileEnd", // [profileName]

	time: "time", // name, time as [seconds, nanoseconds]

	clear: "clear", // no arguments
	status: "status" // message, arguments
});

(snippet from webpack) you get something with literally-typed members as a result, if possible. The intuition seems to be that it should just work like that already, but it does not currently.

[weswigham]

@typescript-bot test this

[typescript-bot]

Heya @weswigham, I've started to run the extended test suite on this PR at a5cb1a6. You can monitor the build here. It should now contribute to this PR's status checks.

[weswigham]

@typescript-bot user test this

[typescript-bot]

Heya @weswigham, I've started to run the parallelized community code test suite on this PR at a5cb1a6. You can monitor the build here. It should now contribute to this PR's status checks.

[typescript-bot]

The user suite test run you requested has finished and failed. I've opened a PR with the baseline diff from master.

[sandersn]

@typescript-bot user test this inline

[typescript-bot]

Heya @sandersn, I've started to run the diff-based community code test suite on this PR at 89cfe35. You can monitor the build here.

Update: The results are in!

[typescript-bot]

@sandersn
Great news! no new errors were found between main..refs/pull/33767/merge

[sandersn]

@weswigham this seems like a plausible change even though it's old. It didn't create any new errors in the user tests. Think it's worth trying out?

[weswigham]

Yeah, I think it's still a reasonable change.

[Andarist]

Out of curiosity... if I hover over the call using this signature I might see something like this:

function freeze<{
    a: 1;
}, string | number | bigint | symbol>(o: {
    a: 1;
}): Readonly<{
    a: 1;
}>

In here, the U was not "computed" but instead its constraint is displayed. What kind of TS behavior/trait is at play here? On what foundation/principles this PR is built? 🤔

[weswigham]

We assign positions literal types when they're contextually typed by a generic with literal bases in the constraint.