Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Consider nullable annotations in explicit nulls #21953

Open
wants to merge 1 commit into
base: main
Choose a base branch
from

Conversation

HarrisL2
Copy link
Contributor

Fixes #21629

Not sure if the tests are sufficient.

return as;
}

@Nullable
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We should test annotations at parameter positions as well: f(@NotNull String s), g(@Nullable String s)

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It seems like @NotNull is currently not working in parameter position.

Also, I am not sure about how to test @Nullable in parameter position, since we should be testing that the parameter is a (String | Null) instead of a String?.

I discussed with @olhotak, and one possible way is through overriding, where the following code would fail

public class JavaClass {
    public int f(@Nullable String s)
}
class ScalaClass extends JavaClass {
    override def f(s: String): Int
}

However we handle overriding logic through a special case for explicit nulls, so this currently doesn't fail. We might have to change the behavior in some other files.

Copy link
Contributor

@olhotak olhotak Dec 2, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I discussed offline with @noti0na1 . We agreed that we should try to remove the special cases in the overriding logic for explicit nulls. Flexible types would make them unnecessary. I've create a new issue #22071 about this.

After that's done, it should be possible to create such tests here using overriding.

tests/explicit-nulls/neg/i21629/J.java Show resolved Hide resolved
@@ -95,8 +98,17 @@ object JavaNullInterop {
* This is useful for e.g. constructors, and also so that `A & B` is nullified
* to `(A & B) | Null`, instead of `(A | Null & B | Null) | Null`.
*/
private class JavaNullMap(var outermostLevelAlreadyNullable: Boolean)(using Context) extends TypeMap {
def nullify(tp: Type): Type = if ctx.flexibleTypes then FlexibleType(tp) else OrNull(tp)
private class JavaNullMap(var outermostLevelAlreadyNullable: Boolean, explicitlyNullable: Boolean)(using Context) extends TypeMap {
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't think the NotNull/Nullable annotation on the symbol should affect any types nested inside.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm not sure what you mean by this, I add the explicitlyNullable flag here to distinguish the case between a flexible type and an OrNull. Are you referring to the places where we call this in apply?

@noti0na1 noti0na1 assigned HarrisL2 and unassigned noti0na1 Nov 19, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Consider nullable annotations in explicit nulls
3 participants