Test_runner: --experimental-test-coverage detects uncovered lines incorrectly.

[muturgan]

Version

20.15.0

Platform

Darwin macbook-WQW3312KW0 23.5.0 Darwin Kernel Version 23.5.0: Wed May  1 20:14:38 PDT 2024; root:xnu-10063.121.3~5/RELEASE_ARM64_T6020 arm64

Subsystem

No response

What steps will reproduce the bug?

1. git clone git@github.com:muturgan/fastest-express-validator.git
2. cd fastest-express-validator && npm ci && npm run build
3. npm run test_coverage
4. look at the uncovered lines colunm of a dist/index.cjs file

How often does it reproduce? Is there a required condition?

It reproduces every time. No special conditions are needed.

What is the expected behavior? Why is that the expected behavior?

Report of a dist/index.cjs file is NOT contains 53-55 uncovered lines (which corresponds to a defaultRequestValidatorHandler success branch). Because of I have a correct user request test case which not produces a validation error.

What do you see instead?

The coverage report of a dist/index.cjs file mark 53-55 lines as uncovered.

Additional information

No response

[RedYetiDev]

Thanks for the report! Could you provide a minimally reproducible example? One without any external dependencies / repos?

[muturgan]

RedYetiDev

I will try

[cjihrig]

Are you enabling source maps?

[muturgan]

Are you enabling source maps?

No, but how it should affects a test coverage? Tests running on already compiled and not minified code.

But affects in fact. I have tried to enable sourceMap in a tsconfig. In this case tests were passed successfully but coverage calculation failed with a message: Warning: Could not report code coverage. TypeError: Cannot read properties of undefined (reading 'line')

[cjihrig]

Hm, I tried running the tests and added some print statements. The if (Object.keys(errors).length === 0) statement in question was never entered for me. Maybe I'm doing something wrong?

[muturgan]

Hm, I tried running the tests and added some print statements. The if (Object.keys(errors).length === 0) statement in question was never entered for me. Maybe I'm doing something wrong?

I read the tests very carefully. Yeah, it turns out that this check looks appropriate in the context of a single function. But from the outside this function is called in such a way that this condition will never be fulfilled. A kind of overkill. I apologize for the time spent. It seems that everything fine with the coverage calculation.