Coerce the result of a function call to the function's return type

[nrc]

No description provided.

[qa-wolf]

QA Wolf here! As you write new code it's important that your test coverage is keeping up.
Click here to request test coverage for this PR!

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name	Status	Preview	Updated (UTC)
modeling-app	✅ Ready (Inspect)	Visit Preview	Apr 21, 2025 9:24pm

[codecov]

Codecov Report

Attention: Patch coverage is 92.06349% with 10 lines in your changes missing coverage. Please review.

Project coverage is 85.16%. Comparing base (30ee547) to head (5a061d4).
Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
rust/kcl-lib/src/execution/exec_ast.rs	91.45%	10 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #6309   +/-   ##
=======================================
  Coverage   85.15%   85.16%           
=======================================
  Files         108      108           
  Lines       46269    46348   +79     
=======================================
+ Hits        39402    39470   +68     
- Misses       6867     6878   +11     

Flag	Coverage Δ
rust	85.16% <92.06%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[codspeed-hq]

CodSpeed Walltime Performance Report

Merging #6309 will improve performances by 18.15%

_{Comparing nrc-coerce-return (e5111d6) with main (941911e)}

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

Summary

⚡ 44 improvements
✅ 48 untouched benchmarks

Benchmarks breakdown

	Benchmark	BASE	HEAD	Change
⚡	parse_ball-bearing	3.8 ms	3.4 ms	+12.4%
⚡	parse_bench	3.7 ms	3.2 ms	+14.86%
⚡	parse_bracket	4.6 ms	4.2 ms	+11.51%
⚡	parse_car-wheel-assembly	398.6 µs	360.8 µs	+10.46%
⚡	parse_color-cube	2.4 ms	2.1 ms	+12.67%
⚡	parse_cycloidal-gear	4.1 ms	3.7 ms	+12.31%
⚡	parse_dodecahedron	7.4 ms	6.5 ms	+13.82%
⚡	parse_dual-basin-utility-sink	15.6 ms	13.7 ms	+13.66%
⚡	parse_enclosure	11.7 ms	10.3 ms	+13.92%
⚡	parse_exhaust-manifold	9.3 ms	8 ms	+15.26%
⚡	execute_flange	2.1 s	1.9 s	+10.1%
⚡	parse_flange	1.5 ms	1.3 ms	+11.69%
⚡	parse_focusrite-scarlett-mounting-bracket	8.6 ms	7.6 ms	+13.28%
⚡	parse_food-service-spatula	12.6 ms	10.9 ms	+15.7%
⚡	parse_french-press	11.7 ms	10.1 ms	+15.95%
⚡	parse_gear-rack	2.6 ms	2.3 ms	+12.28%
⚡	parse_gear	9.4 ms	8.4 ms	+12.22%
⚡	parse_gridfinity-baseplate-magnets	16.3 ms	14.5 ms	+12.42%
⚡	parse_gridfinity-baseplate	10.3 ms	8.9 ms	+14.59%
⚡	parse_gridfinity-bins-stacking-lip	22.1 ms	19.6 ms	+12.63%
...	...	...	...	...

ℹ️ Only the first 20 benchmarks are displayed. Go to the app to view all benchmarks.

[jtran]

This makes it so that a return type annotation on a function can have a runtime effect? What's a motivating use case for this? Maybe we can add a simple unit test.

[nrc]

This makes it so that a return type annotation on a function can have a runtime effect? What's a motivating use case for this? Maybe we can add a simple unit test.

The use case is something like fn foo(): number(mm) { return 42 } so that the returned value has the mm type rather than the default type. I'll add it as a test (I was being lazy and this is tested by the std lib functions already,but you're right it should have a unit test).

[nrc]

This makes it so that a return type annotation on a function can have a runtime effect? What's a motivating use case for this? Maybe we can add a simple unit test.

The use case is something like fn foo(): number(mm) { return 42 } so that the returned value has the mm type rather than the default type. I'll add it as a test (I was being lazy and this is tested by the std lib functions already,but you're right it should have a unit test).

Adding the test was easy, but in fixing the failing test, I had to do a bunch of work because types were not being properly exported (and had to factor out a std::types module to avoid import cycles in std).

[codspeed-hq]

CodSpeed Instrumentation Performance Report

Merging #6309 will not alter performance

_{Comparing nrc-coerce-return (5a061d4) with main (f8ca6ad)}

Summary

✅ 54 untouched benchmarks