Infer fixed-size tuples from infer type parameters with extends clauses at variadic positions #51157
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
…es at variadic positions
Andarist
commented
Oct 12, 2022
src/compiler/checker.ts
Outdated
Comment on lines
23330
to
23338
| if (elementFlags[startLength] & elementFlags[startLength + 1] & ElementFlags.Variadic && isTupleType(source)) { | ||
| // Middle of target is [...T, ...U] and source is tuple type | ||
| const impliedArity = getInferenceInfoForType(elementTypes[startLength])?.impliedArity; | ||
| if (impliedArity !== undefined) { | ||
| // Infer slices from source based on implied arity of T. | ||
| inferFromTypes(sliceTupleType(source, startLength, endLength + sourceArity - impliedArity), elementTypes[startLength]); | ||
| inferFromTypes(sliceTupleType(source, startLength + impliedArity, endLength), elementTypes[startLength + 1]); | ||
| } | ||
| } |
There was a problem hiding this comment.
this case was already here, it just got indented more since I put all 3 cases (1 old, 2 new ones) within the if (middleLength === 2) block
Andarist
commented
Oct 12, 2022
src/compiler/checker.ts
Outdated
Comment on lines
23358
to
23359
| const trailingSlice = createTupleType(getTypeArguments(source).slice(startIndex, endIndex), source.target.elementFlags.slice(startIndex, endIndex), | ||
| /*readonly*/ false, source.target.labeledElementDeclarations && source.target.labeledElementDeclarations.slice(startIndex, endIndex)); |
There was a problem hiding this comment.
This is based on the content of the sliceTupleType. However, I wasn't able to reuse that function as the purpose of this slice is different here and sliceTupleType wouldn't work for it - I didn't find another instance of such a "trailing slice" in the codebase so I've just inlined the thing here and adjusted for the needs here.
There was a problem hiding this comment.
To give more context behind this - I can't reuse sliceTupleType as it creates array types when index > target.fixedLength:
https://github.dev/microsoft/TypeScript/blob/acf854b636e0b8e5a12c3f9951d4edfa0fa73bcd/src/compiler/checker.ts#L15758-L15764
So it's not a literal "slice" - it's not just slicing between the startIndex and endIndex - it also performs a type conversion conditionally.
If not that condition, I could reuse the sliceTupleType here as follows:
const impliedArity = constraint.target.fixedLength;
const targetEndLength = getEndElementCount(target.target, ElementFlags.Fixed);
inferFromTypes(sliceTupleType(source, sourceArity - targetEndLength - impliedArity, targetEndLength), elementTypes[startLength + 1]);
weswigham
approved these changes
Nov 7, 2022
# Conflicts: # src/compiler/checker.ts
|
@weswigham I've synced this with main. You can trigger those extended test suites now. |
|
@typescript-bot pack this |
|
Heya @weswigham, I've started to run the perf test suite on this PR at f6840d5. You can monitor the build here. Update: The results are in! |
|
Heya @weswigham, I've started to run the extended test suite on this PR at f6840d5. You can monitor the build here. |
|
Heya @weswigham, I'm starting to run the diff-based user code test suite on this PR at f6840d5. Hold tight - I'll update this comment with the log link once the build has been queued. |
|
Heya @weswigham, I've started to run the parallelized Definitely Typed test suite on this PR at f6840d5. You can monitor the build here. |
|
Heya @weswigham, I've started to run the tarball bundle task on this PR at f6840d5. You can monitor the build here. |
|
Heya @weswigham, I'm starting to run the diff-based top-repos suite on this PR at f6840d5. Hold tight - I'll update this comment with the log link once the build has been queued. |
|
Hey @weswigham, I've packed this into an installable tgz. You can install it for testing by referencing it in your and then running There is also a playground for this build and an npm module you can use via |
|
@weswigham Here they are:
CompilerComparison Report - main..51157
System
Hosts
Scenarios
TSServerComparison Report - main..51157
System
Hosts
Scenarios
StartupComparison Report - main..51157
System
Hosts
Scenarios
Developer Information: |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
gabritto
requested changes
Nov 24, 2022
src/compiler/checker.ts
Outdated
| if (constraint && isTupleType(constraint) && !constraint.target.hasRestElement) { | ||
| const impliedArity = constraint.target.fixedLength; | ||
| inferFromTypes(sliceTupleType(source, startLength, sourceArity - (startLength + impliedArity)), elementTypes[startLength]); | ||
| } |
There was a problem hiding this comment.
I believe we also need to infer into the rest element here, if the source arity allows (i.e. there are left over elements after inferring into the tuple constraint). Right now, we're only inferring into the variadic type, so we don't infer into the rest element in a case like this:
type SubTup2Variadic2<T extends unknown[]> = T extends [
...infer B extends [any, any],
...(infer C)[]
]
? [...B, ...[C]]
: never;
type SubTup2Variadic2Test = SubTup2Variadic2<[a: 0, b: 1, ...c: number[]]>; // Currently [0, 1, unknown], should be [0, 1, number]
src/compiler/checker.ts
Outdated
| const impliedArity = constraint.target.fixedLength; | ||
| const endIndex = sourceArity - getEndElementCount(target.target, ElementFlags.Fixed); | ||
| const startIndex = endIndex - impliedArity; | ||
| const trailingSlice = createTupleType(getTypeArguments(source).slice(startIndex, endIndex), source.target.elementFlags.slice(startIndex, endIndex), |
There was a problem hiding this comment.
Is there some reason to use createTupleType instead of sliceTupleType here?
There was a problem hiding this comment.
I've mentioned this here but perhaps without going much into the relevant details.
Right now... I'm not exactly sure why I had to do it but I had some reason as I've used sliceTyupleType in other branches and I've called it out in the comment that I wasn't able to use it here. I can try to refresh my memory on this one later and provide the appropriate details
There was a problem hiding this comment.
I've added the requested clarification here: #51157 (comment)
src/compiler/checker.ts
Outdated
| const trailingSlice = createTupleType(getTypeArguments(source).slice(startIndex, endIndex), source.target.elementFlags.slice(startIndex, endIndex), | ||
| /*readonly*/ false, source.target.labeledElementDeclarations && source.target.labeledElementDeclarations.slice(startIndex, endIndex)); | ||
| inferFromTypes(trailingSlice, elementTypes[startLength + 1]); | ||
| } |
There was a problem hiding this comment.
Same as in the case above, we should also infer into the rest type if possible, for cases like this:
type SubTup2TrailingVariadic2<T extends unknown[]> = T extends [
...(infer C)[],
...infer B extends [any, any],
]
? [C, ...B]
: never;
type SubTup2TrailingVariadic2Test = SubTup2TrailingVariadic2<[...a: number[], b: 1, c: 2]>; // Currently [unknown, 1, 2], should be [number, 1, 2].
|
@gabritto thank you for the additional test cases! I will take a look into supporting them in the coming days. |
|
@gabritto I pushed out the requested improvements (inferring for the rest elements appearing with the variadic elements with those fixed-size tuple constraints). Could you take another look and re-review it? |
gabritto
approved these changes
Jan 3, 2023
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.
Fixes #51138