This issue covers the potential issue with MH_* strategies in StringConcatFactory that may end up calling toString() on arguments in wrong order. Non MH-strategies appear fine. See the evidence in the wild here:
https://www.guardsquare.com/en/blog/string-concatenation-java-9-conversion-confusion
This issue adds the relevant test for String concatenation. JDK 10 fails this new test. JDK 11+ passes the test, because the actual bug is JDK-8194554, which is fixed. In other words, StringConcatFactory implementation is actually correct, and it expects filterArguments to run in the specified order. Selectively applying fix from JDK-8194554 to JDK 10 makes it pass.