Stephan Schroevers

Nice @mlrprananta! I don't think this can be a Refaster rule. Rather, it sounds like a slight extension of what's being worked on in #450. WDYT @rickie @cernat-catalin?

Hi @Marcono1234! For `StatementSwitchToExpressionSwitch` such skip-if-unsupported logic is already [in place](https://github.com/google/error-prone/blob/7acf133dc9fd0366afb57ee8e4582a78c808b73d/core/src/main/java/com/google/errorprone/bugpatterns/StatementSwitchToExpressionSwitch.java#L148-L150); did you observe that not working as expected? Assuming that logic does work for you: the other checks could...

> It didn't fail for me, so it might be the case that this test doesn't fully use the features of Lombok? That's correct; for Lombok to work in this...

Hey @almibarss, thanks for reaching out! I have ~zero experience with Gradle, but it sounds like (a) the Refaster rule compiler is properly picked up, since `.refaster` files are generated...

@almibarss happy it worked out! You're absolutely right about documentation! If you don't mind I'll reopen this issue to track that task. (Otherwise I can file a separate ticket and...

(I didn't check the comments before reviewing; sorry.) > In addition to class variables within the same compilation unit, should we also match on constants imported from other classes ?...

Rebased and resolved conflicts. What does the Error Prone team think of this PR?

I have rebased the branch and squashed in a small change to also expose the initialization time recorded since 8ddb7cbb05dc4b278836fd5019751c4e225323d9. NB: I see thumbs-up from Uber, Red Hat and Picnic...

> Is the intended use-case here that if you wanted to dig into the performance of a build, you'd add `-verbose`, and then review the output from Error Prone? Yes,...

> Would using a new Error Prone-separate flag instead of `-verbose` work for your use-case of this feature, I think it would work for us: based on some tests, `Log#printVerbose`...