Is there a way to enable wildcard imports but enfo...

# ktlint

[David Kubecka]

Is there a way to enable wildcard imports but enforce the rules set by ij_kotlin_name_count_to_use_star_import?

[Paul Dingemans]

If you want to allow wildcard imports than you will need to disable the rule that forbids them. See docs.

[David Kubecka]

I read the docs and I know that I can enable wildcard imports. I'm asking whether I can set additional restrictions to the wildcarding behaviour, such as name count.

[Paul Dingemans]

Yes you can set them, but ktlint is not using them. But they will still affect the formatting done by the default Intellij IDEA formatter.

[David Kubecka]

Would you consider making ktlint using them?

[Paul Dingemans]

I don't know because it is not clear to me what you expect. But in general I am hesitant about adding configuration option for minor optimization (https://pinterest.github.io/ktlint/latest/faq/#can-a-new-toggle-be-added-to-optionally-enabledisable-format-code-in-a-particular-way)

[David Kubecka]

I expect that a linter would cover and enforce all agreed rules. So if the team agrees that we should use wildcard imports if there are 3 or more imports from the same package I would like to enforce that. Currently, that's not possible with ktlint (while it's possible -although not in the enforcing mode - in IDEA)

[David Kubecka]

Since IDEA itself has an option to fine-tune wildcard imports (with quite a bad UX, though) I wouldn't consider this just a minor option.

[phldavies]

As I understand it, ktlint only has access to a model of the code in the file. As such, it cannot count the number of imports used from java.util.*, for example. The only enforcement it could do is to replace >n imports from a package with an import package.* but it wouldn't be able to unpack the import package.* to two distinct imports if you were to remove one of the usages.

[David Kubecka]

I see, makes sense. I would say that for me even packing >n imports would be useful but I understand that it's not fully consistent.