Contextual type of methods only in noImplicitAny or JS #50542
Conversation
|
@typescript-bot run dt |
|
Heya @sandersn, I've started to run the diff-based user code test suite on this PR at 04c2fbd. You can monitor the build here. Update: The results are in! |
|
@typescript-bot test this |
sandersn
changed the title
|
User tests show a difference only in webpack, which uses checkJS -- there are two new errors where a parameter now inherits the base type but assumes a derived type. And there are many removed noImplicitAny errors. The RWC diff is empty. I hope that means there are no diffs. But I don't know how many RWC projects have noImplicitAny turned on. |
| @@ -9134,6 +9134,26 @@ namespace ts { | |||
| if (type) { | |||
| return addOptionality(type, /*isProperty*/ false, isOptional); | |||
| } | |||
| if (isMethodDeclaration(func) && isClassLike(func.parent) | |||
| && (noImplicitAny || isInJSFile(func.parent))) { // TODO: Might want to fiddle with restrictions here | |||
There was a problem hiding this comment.
out of curiosity - what's the reasoning to only apply this logic with noImplicitAny (and in JS files)? 🤔
There was a problem hiding this comment.
- noImplicitAny: untyped parameters will get an error at the same time they get
any, so this change can't cause code with 0 errors to compile with >0 errors. - JS files: in JS, errors aren't reported anyway, so any improvement in typing should show up as improvement in completions. (this might need to change to exclude checkjs)
|
@typescript-bot pack this |
|
Heya @RyanCavanaugh, I've started to run the tarball bundle task on this PR at 04c2fbd. You can monitor the build here. |
|
@typescript-bot test top100 |
|
Heya @sandersn, I've started to run the diff-based user code test suite on this PR at 04c2fbd. You can monitor the build here. Update: The results are in! |
|
@sandersn Here are the results of running the top-repos suite comparing Everything looks good! |
|
This would be fantastic. Hope it gets merged! |
Fixes #23911, maybe