Fixed regression in reverse mapped type inference caused by cache leak #59232
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
This PR doesn't have any linked issues. Please open an issue that references this PR. From there we can discuss and prioritise. |
typescript-bot
added
the
For Uncommitted Bug
Andarist
commented
Jul 10, 2024
|
This looks good but the tests are out of date, it seems! |
The fix is correct but causes test tests/cases/compiler/mappedTypeRecursiveInference.ts to stack overflow, which is the main problem. |
|
I think that my theory about 2 deep stacks being stacked on top of each other here might be correct. This I captured this target and printed it after the stack was cleared and it succeeded. The result isn't pretty but it did "work": typeToString(target)This isn't particularly useful display... so ye, definitely something could be done to optimize how this gets printed. That said, maybe making the type print lazy in diagnostics would be something worth exploring either way? If it's only a feasible idea it could improve the performance of some of those pathological cases as some of the created error messages end up being discarded. Interestingly, this doesn't print the same expansion of this type in the error message: // @strict: false
// @strictNullChecks: false
type Deep<T> = { [K in keyof T]: Deep<T[K]> };
const test: {
globalThis: Deep<typeof globalThis>;
} = { globalThis };I don't know where the difference is between those 2. |
Adding my explanation of things to this:
Putting all of those things together means that this PR causes the test to stack overflow. |
…thub.com/Andarist/TypeScript into fix/reverse-mapped-inference-same-source
|
thanks @gabritto for fixing that printing issue ❤️ |
|
@typescript-bot test it |
|
Hey @gabritto, the results of running the DT tests are ready. Everything looks the same! |
|
@gabritto Here are the results of running the user tests with tsc comparing Everything looks good! |
|
@gabritto Here they are:
tscComparison Report - baseline..pr
System info unknown
Hosts
Scenarios
Developer Information: |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
@gabritto Here are the results of running the top 400 repos with tsc comparing Everything looks good! |
gabritto
approved these changes
Jul 12, 2024
weswigham
approved these changes
Jul 12, 2024
|
@typescript-bot cherry-pick to release-5.5 |
DanielRosenwasser
pushed a commit
that referenced
this pull request
Jul 16, 2024
…e-5.5 (#59258) Co-authored-by: Mateusz Burzyński <mateuszburzynski@gmail.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR fixes regression from #57837 that was discovered here: reduxjs/redux-toolkit#4500
Unfortunately, it also conflicts with a PR that landed later on: #58318 . With this fix
tests/cases/compiler/mappedTypeRecursiveInference.tsruns into:This is specifically caused by the newly introduced
reportRelationErrorhere: https://github.com/microsoft/TypeScript/pull/58318/files#diff-d9ab6589e714c71e657f601cf30ff51dfc607fc98419bf72e04f6b0fa92cc4b8R21710This gets called by
isPropertySymbolTypeRelatedin an already deep stack created by this infinitely expanding mapped type. Then that type (or its slice) gets converted into a string eagerly within this function bygetTypeNamesForErrorDisplayand this recurses quite a bit too.I think that this isn't a new problem. Nothing is fundamentally wrong with the PR that added this
reportRelationError. It's just that two big stacks call were accidentally "concatenated" in a way that they were not before.I'm looking for some guidance as to what could be a potential fix for this. My current best idea is to lazily convert types to strings when actually needed/when handing over to the TS Server as a final diagnostic. This would affect a bunch of places though and I don't want to follow that rabbit hole before getting initial thoughts from the team about this.
cc @gabritto