Thanks for the review! I think I'll add additional logic to be more conservative when an import is unresolved. The rule already handles unresolved type imports gracefully, thanks to the unresolved symbols of the symbol API. I think we should also be more intelligent with unresolved static imports.

Originally posted by @oowekyala in #3718 (comment)

Note: one FP caused by this is here: https://github.com/openjdk/jdk/blob/jdk-11%2B28/src/java.base/macosx/classes/sun/nio/ch/KQueuePort.java#L36

We should maybe introduce symbols for unresolved fields.