That's frustrating. Are you seeing the results in target/checkstyle-result.xml for the relevant module? For larger modules, the reports are annoying to wade through, but a search for "<error" should help.
I confirmed that the new line change in that commit was not necessary for checkstyle (at least on my laptop...today...lol). I can open an issue to create a separate entry for jupiter/junit classes in imports if you'd prefer that. At this point, we define java -> other -> tika in imports (generally, there are some nuances for javax). Would we want java -> junit -> other -> tika? On Sun, Aug 13, 2023 at 12:03 AM Tilman Hausherr <thaush...@t-online.de> wrote: > The checkstyle failures are weird. Here's a fix of a failure: > > > https://github.com/apache/tika/commit/50a03d85835405338d47e29fb453c8bc274eac79 > > > It makes the code worse IMHO. What's even more difficult is that > checkstyle doesn't report where the error is. There is a report in the > "site" directory but that one has hundreds of failures. > > I tried to make changes in the config in the pom.xml but failed. I > didn't want to spend more time on this so I committed by changes slowly > (which is a habit of mine anyway) to see which file would fail. > > Tilman > >
