🐛 False positive for `lint/correctness/noUnusedPrivateClassMembers` when using `??=`

[lgarron]

Environment information

CLI:
  Version:                      2.1.4
  Color support:                true

Platform:
  CPU Architecture:             aarch64
  OS:                           macos

Environment:
  BIOME_LOG_PATH:               unset
  BIOME_LOG_PREFIX_NAME:        unset
  BIOME_CONFIG_PATH:            unset
  BIOME_THREADS:                unset
  NO_COLOR:                     unset
  TERM:                         xterm-256color
  JS_RUNTIME_VERSION:           v24.5.0
  JS_RUNTIME_NAME:              node
  NODE_PACKAGE_MANAGER:         bun/1.2.19

Biome Configuration:
  Status:                       Loaded successfully
  Path:                         biome.json
  Formatter enabled:            true
  Linter enabled:               true
  Assist enabled:               true
  VCS enabled:                  false

Workspace:
  Open Documents:               0

What happened?

Repro link: https://biomejs.dev/playground/?indentStyle=space&indentWidth=4&bracketSameLine=true&code=YwBsAGEAcwBzACAAQQBwAHAAIAB7AAoAIAAgACMAcABlAHIAcwBpAHMAdABlAG4AYwBlAFIAZQBxAHUAZQBzAHQAOgAgAFAAcgBvAG0AaQBzAGUAPABiAG8AbwBsAGUAYQBuAD4AIAB8ACAAdQBuAGQAZQBmAGkAbgBlAGQAOwAKACAAIABzAGEAdgBlAEQAYQB0AGEAKAApACAAewAKACAAIAAgACAAdABoAGkAcwAuACMAcABlAHIAcwBpAHMAdABlAG4AYwBlAFIAZQBxAHUAZQBzAHQAIAA%2FAD8APQAgAG4AYQB2AGkAZwBhAHQAbwByAC4AcwB0AG8AcgBhAGcAZQAuAHAAZQByAHMAaQBzAHQAKAApADsACgAgACAAIAAgAC8ALwAgACYgCgAgACAAfQAKAH0ACgAKAG4AZQB3ACAAQQBwAHAAKAApADsA

Code:

class App {
  #persistenceRequest: Promise<boolean> | undefined;
  saveData() {
    this.#persistenceRequest ??= navigator.storage.persist();
    // …
  }
}

new App();

This results in a lint/correctness/noUnusedPrivateClassMembers error claiming that the .#persistenceRequest is field is unused. However, it is accessed on the line with ??= to gate a call to a function with side effects.

In fact, the linter will no longer error if you change it to essentially equivalent code:

this.#persistenceRequest = this.#persistenceRequest ?? navigator.storage.persist();

Expected result

No error.

Code of Conduct

• I agree to follow Biome's Code of Conduct

[vladimir-ivanov]

@lgarron this is a valid failure, consistent with eslint, please see:

https://eslint.org/play/#eyJ0ZXh0IjoiLyplc2xpbnQgbm8tdW51c2VkLXByaXZhdGUtY2xhc3MtbWVtYmVyczogXCJlcnJvclwiKi9cblxuY2xhc3MgQXBwIHtcbiAgI3BlcnNpc3RlbmNlUmVxdWVzdDtcbiAgc2F2ZURhdGEoKSB7XG4gICAgdGhpcy4jcGVyc2lzdGVuY2VSZXF1ZXN0ID8/PSBuYXZpZ2F0b3Iuc3RvcmFnZS5wZXJzaXN0KCk7XG4gIH1cbn1cbiIsIm9wdGlvbnMiOnsicnVsZXMiOnt9LCJsYW5ndWFnZU9wdGlvbnMiOnsicGFyc2VyT3B0aW9ucyI6eyJlY21hRmVhdHVyZXMiOnt9fX19fQ==

[lgarron]

@lgarron this is a valid failure, consistent with eslint, please see:

https://eslint.org/play/#eyJ0ZXh0IjoiLyplc2xpbnQgbm8tdW51c2VkLXByaXZhdGUtY2xhc3MtbWVtYmVyczogXCJlcnJvclwiKi9cblxuY2xhc3MgQXBwIHtcbiAgI3BlcnNpc3RlbmNlUmVxdWVzdDtcbiAgc2F2ZURhdGEoKSB7XG4gICAgdGhpcy4jcGVyc2lzdGVuY2VSZXF1ZXN0ID8/PSBuYXZpZ2F0b3Iuc3RvcmFnZS5wZXJzaXN0KCk7XG4gIH1cbn1cbiIsIm9wdGlvbnMiOnsicnVsZXMiOnt9LCJsYW5ndWFnZU9wdGlvbnMiOnsicGFyc2VyT3B0aW9ucyI6eyJlY21hRmVhdHVyZXMiOnt9fX19fQ==

I don't follow. Why would eslint's behaviour make this a valid failure?

[vladimir-ivanov]

because the assignment is never used anywhere else. Only if a return statement is added it will think that the intention is to be used elsewhere. Hope this makes sense?

[lgarron]

because the assignment is never used anywhere else. Only if a return statement is added it will think that the intention is to be used elsewhere. Hope this makes sense?

No, it doesn't make sense to me. 😕

As I wrote above:

This results in a lint/correctness/noUnusedPrivateClassMembers error claiming that the .#persistenceRequest is field is unused. However, it is accessed on the line with ??= to gate a call to a function with side effects.

In fact, the linter will no longer error if you change it to essentially equivalent code:

this.#persistenceRequest = this.#persistenceRequest ?? navigator.storage.persist();

[vladimir-ivanov]

ok, so the bug is essentially -> this.#persistenceRequest = this.#persistenceRequest ?? navigator.storage.persist();

[vladimir-ivanov]

https://eslint.org/play/#eyJ0ZXh0IjoiLyplc2xpbnQgbm8tdW51c2VkLXByaXZhdGUtY2xhc3MtbWVtYmVyczogXCJlcnJvclwiKi9cblxuY2xhc3MgQXBwIHtcbiAgI3BlcnNpc3RlbmNlUmVxdWVzdDtcbiAgc2F2ZURhdGEoKSB7XG4gICAgdGhpcy4jcGVyc2lzdGVuY2VSZXF1ZXN0ID0gdGhpcy4jcGVyc2lzdGVuY2VSZXF1ZXN0ID8/IG5hdmlnYXRvci5zdG9yYWdlLnBlcnNpc3QoKTtcbiAgfVxufVxuIiwib3B0aW9ucyI6eyJydWxlcyI6e30sImxhbmd1YWdlT3B0aW9ucyI6eyJwYXJzZXJPcHRpb25zIjp7ImVjbWFGZWF0dXJlcyI6e319fX19 this also seems consistent with eslint - no error

[vladimir-ivanov]

Personally I tend to agree with you, but the rule suggest we do the same behaviour as eslint:

sources: &[RuleSource::Eslint("no-unused-private-class-members").same()],

[vladimir-ivanov]

to me the second case should also be marked as invalid - possibly should be reported as bug on eslint :-)
@arendjr, @ematipico your opinion on this please?