Manu Sridharan

The issue is that you call `this.properties.username()` (and `password()`) inside a lambda body. NullAway doesn't know when the lambda function will be invoked, and it does not assume that `username()`...

This is a different case. I don't think the `KnownInitializers` flag is being used correctly; that flag is for (most likely inherited) methods that behave like constructors almost and run...

Thanks for this report! This is not intentional behavior; it's quite surprising in fact. Let me dig a bit to see if I can fix.

Based on https://github.com/projectlombok/lombok/issues/3346 I figured out the right config required here: ``` lombok.copyableAnnotations+=org.jspecify.annotations.NonNull lombok.copyableAnnotations+=org.jspecify.annotations.Nullable ``` If you want Lombok to use JSpecify annotations for generated code you can also do:...

I opened a PR to fix this upstream in Lombok: https://github.com/projectlombok/lombok/pull/3613

Fixed upstream in projectlombok/lombok#3608

@commonquail thanks for the detailed report! My guess is for many of these cases we need to model more `Optional` APIs. (The `or()` case looks trickier.) We would welcome a...

FYI I am going to hold off on landing this, as I realized it doesn't completely fix the problems we were seeing. Once we figure out a plan for moving...

> I think this might end up being the safest bet. The other alternatives are: a) inconsistent semantics for `@Nullable` seemingly only dependent on the annotation FQN, or b) "ignored"...

Looking at [the code doing the bad list access](https://github.com/uber/NullAway/blob/3a47fee24a81f7f9f7877f7a1b5356ce16b62825/nullaway/src/main/java/com/uber/nullaway/NullAway.java#L1629-L1629) it seems this is not exactly about the regex for package names, but it's some other issue. We'd love to track...