Improve contextual typing of ending tuple elements #53036

ahejlsberg · 2023-03-01T00:30:49Z

This PR improves the precision of contextual types originating in ending elements of contextual tuple types. For example:

declare function test(...args: [...((arg: number) => void)[], (arg: string) => void]): void;

test(
    a => a,  // number
    b => b,  // number
    c => c   // string
);

const funcs: [...((arg: number) => void)[], (arg: string) => void] = [
    a => a,  // number
    b => b,  // number
    c => c   // string
];

Previously, the arrow function parameters would just be given an implicit any type.

Fixes #43122.
Fixes #52846.

ahejlsberg · 2023-03-01T00:44:52Z

@typescript-bot test this
@typescript-bot user test this inline
@typescript-bot run dt
@typescript-bot perf test faster
@typescript-bot test top100

typescript-bot · 2023-03-01T00:44:55Z

Heya @ahejlsberg, I've started to run the diff-based top-repos suite on this PR at b2f0543. You can monitor the build here.

Update: The results are in!

typescript-bot · 2023-03-01T00:44:55Z

Heya @ahejlsberg, I've started to run the extended test suite on this PR at b2f0543. You can monitor the build here.

typescript-bot · 2023-03-01T00:44:55Z

Heya @ahejlsberg, I've started to run the diff-based user code test suite on this PR at b2f0543. You can monitor the build here.

Update: The results are in!

typescript-bot · 2023-03-01T00:44:55Z

Heya @ahejlsberg, I've started to run the abridged perf test suite on this PR at b2f0543. You can monitor the build here.

typescript-bot · 2023-03-01T00:44:55Z

Heya @ahejlsberg, I've started to run the parallelized Definitely Typed test suite on this PR at b2f0543. You can monitor the build here.

typescript-bot · 2023-03-01T00:57:28Z

@ahejlsberg Here are the results of running the user test suite comparing main and refs/pull/53036/merge:

Everything looks good!

typescript-bot · 2023-03-01T00:59:16Z

Heya @ahejlsberg, I've run the RWC suite on this PR - assuming you're on the TS core team, you can view the resulting diff here.

typescript-bot · 2023-03-01T02:21:14Z

@ahejlsberg Here are the results of running the top-repos suite comparing main and refs/pull/53036/merge:

Everything looks good!

ahejlsberg · 2023-03-01T04:10:16Z

@typescript-bot perf test faster

typescript-bot · 2023-03-01T04:10:19Z

Heya @ahejlsberg, I've started to run the abridged perf test suite on this PR at b2f0543. You can monitor the build here.

Update: The results are in!

typescript-bot · 2023-03-01T04:25:27Z

@ahejlsberg
The results of the perf run you requested are in!

Here they are:

Comparison Report - main..53036

_Metric	_main	₅₃₀₃₆	_Delta	_Best	_Worst	_p-value
_{Angular - node (v16.17.1, x64)}
_{Memory used}	_{359,068k (± 0.01%)}	_{359,288k (± 0.01%)}	_{+220k (+ 0.06%)}	_359,263k	_359,311k	_{p=0.005 n=6}
_{Parse Time}	_{3.71s (± 0.59%)}	_{3.72s (± 0.56%)}	_~	_3.68s	_3.74s	_{p=0.571 n=6}
_{Bind Time}	_{1.19s (± 0.53%)}	_{1.20s (± 0.43%)}	_~	_1.19s	_1.20s	_{p=0.091 n=6}
_{Check Time}	_{9.45s (± 0.58%)}	_{9.46s (± 0.38%)}	_~	_9.40s	_9.50s	_{p=0.568 n=6}
_{Emit Time}	_{7.89s (± 0.32%)}	_{7.90s (± 0.67%)}	_~	_7.87s	_8.01s	_{p=0.870 n=6}
_{Total Time}	_{22.24s (± 0.27%)}	_{22.28s (± 0.33%)}	_~	_22.21s	_22.42s	_{p=0.809 n=6}
_{Compiler-Unions - node (v16.17.1, x64)}
_{Memory used}	_{191,785k (± 0.68%)}	_{191,266k (± 0.03%)}	_~	_191,201k	_191,348k	_{p=0.936 n=6}
_{Parse Time}	_{1.56s (± 0.94%)}	_{1.57s (± 1.10%)}	_~	_1.55s	_1.60s	_{p=0.742 n=6}
_{Bind Time}	_{0.82s (± 0.92%)}	_{0.82s (± 0.00%)}	_~	_0.82s	_0.82s	_{p=0.598 n=6}
_{Check Time}	_{10.09s (± 0.57%)}	_{10.09s (± 0.40%)}	_~	_10.04s	_10.14s	_{p=0.747 n=6}
_{Emit Time}	_{2.99s (± 0.63%)}	_{3.01s (± 0.58%)}	_~	_3.00s	_3.04s	_{p=0.088 n=6}
_{Total Time}	_{15.46s (± 0.40%)}	_{15.49s (± 0.36%)}	_~	_15.42s	_15.57s	_{p=0.520 n=6}
_{Monaco - node (v16.17.1, x64)}
_{Memory used}	_{343,116k (± 0.01%)}	_{343,193k (± 0.00%)}	_{+77k (+ 0.02%)}	_343,171k	_343,219k	_{p=0.005 n=6}
_{Parse Time}	_{2.80s (± 0.53%)}	_{2.81s (± 0.35%)}	_~	_2.80s	_2.82s	_{p=0.216 n=6}
_{Bind Time}	_{1.08s (± 0.48%)}	_{1.09s (± 1.07%)}	_~	_1.08s	_1.11s	_{p=0.523 n=6}
_{Check Time}	_{7.68s (± 0.45%)}	_{7.68s (± 0.54%)}	_~	_7.64s	_7.76s	_{p=1.000 n=6}
_{Emit Time}	_{4.43s (± 0.82%)}	_{4.42s (± 0.90%)}	_~	_4.37s	_4.48s	_{p=0.748 n=6}
_{Total Time}	_{15.99s (± 0.41%)}	_{16.01s (± 0.42%)}	_~	_15.90s	_16.07s	_{p=0.629 n=6}
_{TFS - node (v16.17.1, x64)}
_{Memory used}	_{299,241k (± 0.01%)}	_{299,288k (± 0.00%)}	_{+47k (+ 0.02%)}	_299,278k	_299,314k	_{p=0.030 n=6}
_{Parse Time}	_{2.15s (± 0.48%)}	_{2.16s (± 0.62%)}	_~	_2.14s	_2.18s	_{p=0.652 n=6}
_{Bind Time}	_{1.24s (± 1.19%)}	_{1.25s (± 0.71%)}	_~	_1.24s	_1.26s	_{p=0.208 n=6}
_{Check Time}	_{7.18s (± 0.33%)}	_{7.16s (± 0.33%)}	_~	_7.14s	_7.20s	_{p=0.257 n=6}
_{Emit Time}	_{4.34s (± 0.62%)}	_{4.32s (± 0.48%)}	_~	_4.30s	_4.35s	_{p=0.327 n=6}
_{Total Time}	_{14.91s (± 0.30%)}	_{14.90s (± 0.40%)}	_~	_14.84s	_14.97s	_{p=0.748 n=6}
_{material-ui - node (v16.17.1, x64)}
_{Memory used}	_{475,646k (± 0.01%)}	_{475,711k (± 0.01%)}	_{+65k (+ 0.01%)}	_475,647k	_475,765k	_{p=0.045 n=6}
_{Parse Time}	_{3.29s (± 0.61%)}	_{3.30s (± 0.25%)}	_~	_3.29s	_3.31s	_{p=0.565 n=6}
_{Bind Time}	_{0.96s (± 0.93%)}	_{0.96s (± 0.66%)}	_~	_0.95s	_0.97s	_{p=1.000 n=6}
_{Check Time}	_{18.03s (± 0.21%)}	_{18.08s (± 0.45%)}	_~	_17.97s	_18.18s	_{p=0.334 n=6}
_{Emit Time}	_{0.00s (± 0.00%)}	_{0.00s (± 0.00%)}	_~	_0.00s	_0.00s	_{p=1.000 n=6}
_{Total Time}	_{22.28s (± 0.22%)}	_{22.33s (± 0.34%)}	_~	_22.23s	_22.43s	_{p=0.295 n=6}
_{xstate - node (v16.17.1, x64)}
_{Memory used}	_{545,709k (± 0.02%)}	_{545,736k (± 0.01%)}	_~	_545,666k	_545,771k	_{p=0.810 n=6}
_{Parse Time}	_{4.28s (± 0.50%)}	_{4.30s (± 0.65%)}	_~	_4.26s	_4.33s	_{p=0.195 n=6}
_{Bind Time}	_{1.76s (± 0.46%)}	_{1.76s (± 0.29%)}	_~	_1.76s	_1.77s	_{p=0.140 n=6}
_{Check Time}	_{2.98s (± 0.42%)}	_{2.99s (± 0.49%)}	_~	_2.97s	_3.01s	_{p=0.255 n=6}
_{Emit Time}	_{0.09s (± 0.00%)}	_{0.09s (± 0.00%)}	_~	_0.09s	_0.09s	_{p=1.000 n=6}
_{Total Time}	_{9.10s (± 0.27%)}	_{9.14s (± 0.25%)}	_{+0.04s (+ 0.48%)}	_9.11s	_9.18s	_{p=0.016 n=6}

System

_{Machine Name}	_ts-ci-ubuntu
_Platform	_{linux 5.4.0-135-generic}
_Architecture	_x64
_{Available Memory}	_{16 GB}
_{Available Memory}	_{15 GB}
_CPUs	_{4 × Intel(R) Core(TM) i7-4770 CPU @ 3.40GHz}

Hosts

node (v16.17.1, x64)

Scenarios

Angular - node (v16.17.1, x64)
Compiler-Unions - node (v16.17.1, x64)
Monaco - node (v16.17.1, x64)
TFS - node (v16.17.1, x64)
material-ui - node (v16.17.1, x64)
xstate - node (v16.17.1, x64)

_Benchmark	_Name	_Iterations
_Current	₅₃₀₃₆	₆
_Baseline	_main	₆

Developer Information:

Download Benchmark

Andarist · 2023-03-01T11:57:58Z

src/compiler/checker.ts

+            }
+            // If element index is known and a contextual property with that name exists, return it. Otherwise return the
+            // iterated or element type of the contextual type.
+            return (!firstSpreadIndex || index < firstSpreadIndex) && getTypeOfPropertyOfContextualType(type, "" + index as __String) ||


Maybe it should be treated as a separate issue... but I feel that perhaps this feature here might directly complicate things for that issue so it's at least worth mentioning now.

Take a look at this TS playground. The problem there is that getTypeOfPropertyOfContextualType gets the getElementTypeOfSliceOfTupleType, so we end up with a union of all element types instead of types of individual tuple members.

I prepared a quick fix for this issue in this PR: #53042 but it doesn't handle the case that is being fixed here (ending tuple elements).

The case outlined there is currently a little bit silly (T[K & keyof T] within [K in keyof T2]: { ... }) but I partially encountered this need when working on inference for intersected mapped types (PR here). This case would be better if rewritten like this:

diff --git a/tests/cases/compiler/contextuallyTypedElementsOfGenericZippingTuples.ts b/tests/cases/compiler/contextuallyTypedElementsOfGenericZippingTuples.ts index a011321a26..c27c91930f 100644 --- a/tests/cases/compiler/contextuallyTypedElementsOfGenericZippingTuples.ts +++ b/tests/cases/compiler/contextuallyTypedElementsOfGenericZippingTuples.ts @@ -11,8 +11,8 @@ declare function test<T extends unknown[], T2 extends unknown[]>( ], b: [ ...{ - [K in keyof T2]: { - consume: (arg: T[K & keyof T]) => T2[K]; + [K in keyof T2 as K & keyof T]: { + consume: (arg: T[K]) => T2[K]; }; } ]

There are 2 problems with this code today though:

such a reverse mapped type isn't inferred at all (a PR fixing this problem can be found here)

this isn't recognized as a tuple type so TS reports A rest element type must be an array type.ts(2574). That probably could be improved as well.

@ahejlsberg Any concerns given the above?

No, that's a separate issue.

Yes, I just meant specifically the "this feature here might directly complicate things for that issue so it's at least worth mentioning now" bit.

# Conflicts: # src/compiler/checker.ts

ahejlsberg added 3 commits February 28, 2023 16:08

Improve contextual typing of ending tuple elements

5f947f8

Accept new baselines

adeafcc

Add tests

b2f0543

typescript-bot assigned ahejlsberg Mar 1, 2023

typescript-bot added Author: Team For Milestone Bug PRs that fix a bug with a specific milestone labels Mar 1, 2023

Andarist mentioned this pull request Mar 1, 2023

Fixed contextual types within generic tuple mapped types #53042

Open

Andarist reviewed Mar 1, 2023

View reviewed changes

ahejlsberg added 4 commits March 15, 2023 17:16

Merge branch 'main' into fix52846

a3d6017

# Conflicts: # src/compiler/checker.ts

Merge branch 'main' into fix52846

bad7048

Accept new baselines

53ad72c

Merge branch 'main' into fix52846

da43359

# Conflicts: # src/compiler/checker.ts

ahejlsberg requested review from DanielRosenwasser, jakebailey and RyanCavanaugh March 16, 2023 17:20

jakebailey approved these changes Mar 17, 2023

View reviewed changes

ahejlsberg merged commit 9b7f291 into main Mar 19, 2023

ahejlsberg deleted the fix52846 branch March 19, 2023 15:13

Andarist mentioned this pull request Dec 13, 2023

Contextual inference of functions in tuple types failures #47226

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Improve contextual typing of ending tuple elements #53036

Improve contextual typing of ending tuple elements #53036

ahejlsberg commented Mar 1, 2023 •

edited

Loading

ahejlsberg commented Mar 1, 2023

typescript-bot commented Mar 1, 2023 •

edited

Loading

typescript-bot commented Mar 1, 2023 •

edited

Loading

typescript-bot commented Mar 1, 2023 •

edited

Loading

typescript-bot commented Mar 1, 2023 •

edited

Loading

typescript-bot commented Mar 1, 2023 •

edited

Loading

typescript-bot commented Mar 1, 2023

typescript-bot commented Mar 1, 2023

typescript-bot commented Mar 1, 2023

ahejlsberg commented Mar 1, 2023

typescript-bot commented Mar 1, 2023 •

edited

Loading

typescript-bot commented Mar 1, 2023

Andarist Mar 1, 2023

jakebailey Mar 16, 2023

ahejlsberg Mar 17, 2023

jakebailey Mar 17, 2023

Improve contextual typing of ending tuple elements #53036

Improve contextual typing of ending tuple elements #53036

Conversation

ahejlsberg commented Mar 1, 2023 • edited Loading

ahejlsberg commented Mar 1, 2023

typescript-bot commented Mar 1, 2023 • edited Loading

typescript-bot commented Mar 1, 2023 • edited Loading

typescript-bot commented Mar 1, 2023 • edited Loading

typescript-bot commented Mar 1, 2023 • edited Loading

typescript-bot commented Mar 1, 2023 • edited Loading

typescript-bot commented Mar 1, 2023

typescript-bot commented Mar 1, 2023

typescript-bot commented Mar 1, 2023

ahejlsberg commented Mar 1, 2023

typescript-bot commented Mar 1, 2023 • edited Loading

typescript-bot commented Mar 1, 2023

Andarist Mar 1, 2023

Choose a reason for hiding this comment

jakebailey Mar 16, 2023

Choose a reason for hiding this comment

ahejlsberg Mar 17, 2023

Choose a reason for hiding this comment

jakebailey Mar 17, 2023

Choose a reason for hiding this comment

ahejlsberg commented Mar 1, 2023 •

edited

Loading

typescript-bot commented Mar 1, 2023 •

edited

Loading

typescript-bot commented Mar 1, 2023 •

edited

Loading

typescript-bot commented Mar 1, 2023 •

edited

Loading

typescript-bot commented Mar 1, 2023 •

edited

Loading

typescript-bot commented Mar 1, 2023 •

edited

Loading

typescript-bot commented Mar 1, 2023 •

edited

Loading