fail positives with "no-named-as-default"

[Mister-Hope]

Any ideas why this is reported?

[JounQin]

It's said you should use import { HeroInfo } instead

[Mister-Hope]

Stop closing my issue, I am not a noob as I am in charge for a few 2k+ star projects. I already told you I found a fail positive.

This is the generated dts file from npm:

So why should I use import { HeroInfo }

Edited: Sorry for my attitude with the first edition, I am at covid with fever, so in a bad mood. Please reopen this, if you need a reproduction I can try to provide one.

[JounQin]

Stop closing my issue, I am not a noob as I am in charge for a few 2k+ star projects. I already told you I found a fail positive.

You didn't provide any useful info in your initial issue.

---

It seems like a limitation, HeroInfo is already declared as an interface, so importing it as default would make HeroInfo interface unavailable any longer. The property usage should use another name instead of HeroInfo for the default importing identifier.

[Mister-Hope]

Thanks for the reply, BTW I updated my second reply, just a remind in case you miss it.

[SukkaW]

It seems like a limitation, HeroInfo is already declared as an interface, so importing it as default would make the HeroInfo interface unavailable any longer. The property usage should use another name instead of HeroInfo for the default importing identifier.

eslint-plugin-import-x treats type export and value export identically when tracking modules. While it's possible to distinguish whether an export is type-only through AST, it would be quite challenging.

It's undoubtedly bad practice to have a named type-only export and a default value-only export share the same name. I don't know whether we should address this issue or not. If someone wants to open a PR addressing this, I would be more than happy to review and merge it.

[meteorlxy]

IMO, to avoid naming the default export as the same name of a type is also the purpose of the rule. So I'd treat it as a valid warning.

[SukkaW]

IMO, to avoid naming the default export as the same name of a type is also the purpose of the rule. So I'd treat it as a valid warning.

But what about third-party libraries? I don't want to make this rule ignore all third-party libraries.

[AndreaPontrandolfo]

Related #136