Hi,
this is an attempt to avoid some illegal reflective access in RabbitAutoConfigurationTests. The idea is that rather than using reflection in order to get down to the used TrustManager we look for output that TrustEverythingTrustManager is producing when in use. It's not pretty, but it works and is reflection free. It's also the least invasive option without opening modules and/or changing the production code that I could find.
As this is part of the overall JDK 17 story under #26767, it should probably go into 2.5.x.
Let me know what you think. Cheers, Christoph
Comment From: snicoll
Thanks @dreis2211, I am not a big fan of asserting the console output but, looking at the previous version, I think it's an acceptable trade-off.
Comment From: dreis2211
Me neither, @snicoll . But thanks for accepting it.