-
Notifications
You must be signed in to change notification settings - Fork 12.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Do not expand type references in keyof and index access #58715
Do not expand type references in keyof and index access #58715
Conversation
…h the keyof operator.
}; | ||
export declare const fnExp: (v: O["prop"], p: Omit<O, "prop">, key: keyof O, p2: Omit<O, keyof I>) => void; | ||
//// [b.d.ts] | ||
export declare const f: (v: string, p: Omit<{ |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'm a bit surprised we don't emit this as
export declare const f: typeof import("./a").fn;
since we're willing to emit the import("./aExp").O["prop"]
et al.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
typeof
is currently only used for function declarations. I think leaning into typeof
for variables is an options that should be considered. I'm not sure why only function declarations use typeof
(although I haven't really though too deeply about it)
src/compiler/checker.ts
Outdated
} | ||
|
||
if (isTypeReferenceNode(node)) { | ||
return tryVisitTypeReference(node) ?? serializeExistingTypeNode(context, node); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Are the explicit fallbacks needed with the automatic fallback added in #58720?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yeah. You are right. Better to lean on the new fallback mechanism. As it currently is we could end up trying to print the type two times (when we fail). Pushed the change
@@ -34,7 +34,7 @@ verify.quickInfos({ | |||
func?: Function; | |||
}, obj: T, key: "name"): "name"`, | |||
4: `function foof<Error, "name">(settings: (row: Error) => { | |||
value: Error["name"]; | |||
value: string; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Weird - why's this one change? Isn't Error
in scope?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It's not about Error
. Before when I saw a type parameter that needed mapping, I would fall back to type serialiation. Then type serialization would map T
to Error
and then print Error
. Now when I see the T
in T["name"]
and that it can't be reused without falling back to serialization, I fall back to serializaing T["name"]
insead which gets us string
.
* upstream/main: (37 commits) Added NoTruncation flag to completions (microsoft#58719) Clone node to remove location even when it has been modified if needed (microsoft#58706) Properly account for `this` argument in intersection apparent type caching (microsoft#58677) Fix: Include Values of Script Extensions for Unicode Property Value Expressions in Regular Expressions (microsoft#58615) In `reScanSlashToken` use `charCodeChecked` not `codePointChecked` (microsoft#58727) Shorten error spans for errors reported on constructor declarations (microsoft#58061) Mark file as skips typechecking if it contains ts-nocheck (microsoft#58593) Fixed an issue with broken `await using` declarations in `for of` loops (microsoft#56466) Do not expand type references in keyof and index access (microsoft#58715) Improve the performance of isolatedDeclarations quickfix (microsoft#58722) Unwrap `NoInfer` types when narrowing (microsoft#58292) Recover from type reuse errors by falling back to inferred type printing (microsoft#58720) Fixing self import (microsoft#58718) Enable JS emit for noCheck and noCheck for transpileModule (microsoft#58364) Revert PR 55371 (microsoft#58702) Update dependencies (microsoft#58639) Fix baselines after PR 58621 (microsoft#58705) Do not infer `yield*` type from contextual `TReturn` (microsoft#58621) `await using` normative changes (microsoft#58624) Handling statements from a known source file (microsoft#58679) ...
Fixes #58713