Parse error in Java parser when using type annotation with varags

[domagojcc]

Used version

rewrite-java 8.56.1

The gist

Object @Nullable ... other is interpreted as Object @Nullable[] ... other

while e.g. the following is interpreted correctly:
Object @Nullable[] ... other
Object ... other

Issue fully replicated in test:

package de.domagoj.or.test;

import static org.openrewrite.java.Assertions.java;

import org.junit.jupiter.api.Test;
import org.openrewrite.java.JavaParser;
import org.openrewrite.test.RecipeSpec;
import org.openrewrite.test.RewriteTest;

public class MyTest implements RewriteTest {
        @Test
        void shouldParse() {
            rewriteRun(
                java(
                    """
package de.domagoj.or.test.example;

import static java.lang.annotation.ElementType.TYPE_USE;

import java.lang.annotation.Documented;
import java.lang.annotation.Retention;
import java.lang.annotation.RetentionPolicy;
import java.lang.annotation.Target;

public interface Itf {
    @Documented
    @Retention(RetentionPolicy.CLASS)
    @Target({ TYPE_USE })
    public @interface NonNull {
    }
    @Documented
    @Retention(RetentionPolicy.CLASS)
    @Target({ TYPE_USE })
    public @interface Nullable {
    }


    @NonNull
    Object exampleMethod(@NonNull Object @Nullable ... other);
}
                    """
                )
            );
        }

}

[domagojcc]

I posted the issue to the wrong project, it should have been rewrite-java. However, I stumbled upon it initially using the rewrite-maven-plugin. It is only after isolating the issue in the test that it became obvious that it concerns the Java parser. I forgot to switch to the right project before posting the issue, sorry.

[timtebeek]

Thanks for the report here @domagojcc ! Indeed looks like a parsing or printing issue. Did you happen to already explore a potential parser fix?

[domagojcc]

Sorry, I did not have time to look into the sources. Still I took the time to extract the minimal test that reproduces the issue and I hoped that it would be easy for you to deduct what code is responsible but of course it depends on how old the code is even if your wrote it yourself. (At least my experience.)