Closed sbrannen closed 1 year ago
Thank you, @sbrannen. This is indeed a regression, and it affects all negated type patterns, not just !void
, i.e. also something like !String
(which then in turn should match void types, among many others). I have created a regression test already. When I can fix that depends on my free cycles outside of work.
@sbrannen, the problem should be fixed. Please try the latest 1.9.21-SNAPSHOT
in your Spring build. Without feedback, I am not confident to release 1.9.20.1 anytime soon.
<repositories>
<repository>
<id>ossrh-snapshots</id>
<url>https://oss.sonatype.org/content/repositories/snapshots</url>
<releases>
<enabled>false</enabled>
</releases>
<snapshots>
<enabled>true</enabled>
<updatePolicy>always</updatePolicy>
</snapshots>
</repository>
</repositories>
I am disappointed that there was no re-testing feedback. That lack of feedback kept me withholding the AspectJ 1.9.20.1 bugfix release for longer than necessary. Anyway, today I released the new version. Enjoy.
In the test suite for the Spring Framework, we have had the following pointcut for several years.
However, once we upgraded from AspectJ 1.9.19 to 1.9.20, our test using that pointcut began to [fail](https://ge.spring.io/s/v2js2afkh3zom/tests/task/:spring-aop:test/details/org.springframework.aop.aspectj.annotation.ReflectiveAspectJAdvisorFactoryTests/namedPointcutFromAspectLibrary()?top-execution=1).
Specifically, the above pointcut no longer matches a method such as
int getAge();
.To get our test to pass again, we changed the pointcut to
@Pointcut("execution(int get*())")
in https://github.com/spring-projects/spring-framework/commit/c72dd1ff66f841443e864925061d0ff01a1a59be.