I do like the idea of extending the line length to greater than 80. On Tue, Feb 21, 2017 at 6:29 PM, Justin Leet <[email protected]> wrote:
> Correct. It's pretty easy to do, I tested the line length limit property > on my own (to be explicit, nobody has ever agreed to anything other than > sun conventions + two spaces, but I was playing around with things. If we > want something else, we'll need to come to an agreement). Their Google > checkstyle does manage indenting, but I'm not sure if we need to mess with > some of the properties there. > > https://github.com/checkstyle/checkstyle/blob/master/src/ > main/resources/sun_checks.xml > will be the base > https://github.com/checkstyle/checkstyle/blob/master/src/ > main/resources/google_checks.xml#L152 > is at least close to what we want for indents. > > On Tue, Feb 21, 2017 at 2:45 PM, Michael Miklavcic < > [email protected]> wrote: > > > Justin, Just to clarify, the default Sun conventions don't abide fully to > > our style guide, correct? As you mention above, we would need to create a > > custom checkstyle.xml to handle "extended the character limit of a line > > past 80 and made it two space indents" > > > > On Tue, Feb 21, 2017 at 3:22 PM, Justin Leet <[email protected]> > > wrote: > > > > > I'm also +1 to blanket formatting. > > > > > > As a note, we'll also need to set most of the sun checkstyle violations > > to > > > warnings and only fail the build on error. Even the quick and dirty > > > reformat I tried only got rid of several thousand violations, leaving > > well > > > over 10k violations. We'll need to set it to only fail code styling > > > violations. If we're clever, we could probably even set it up so that > > test > > > classes don't even throw warnings on things like magic numbers, but > > that's > > > not at all necessary for right now. > > > > > > The current PR also does not include a custom checkstyle.xml. It's just > > > reporting on the default Sun conventions as a start. We can pretty > > easily > > > take that and modify it, though. I'd also say we should include the > > > IntelliJ codestyle template along with instructions for installing in > > that > > > portion of effort. I could be pretty easily persuaded to include it in > > this > > > PR also though, if we want. > > > > > > That checking is not part of the current PR (checkstyle is only in the > > > reporting element, not the build element). It's easy enough to setup > > > though. That setup is discussed at maven-checkstyle-plugin > > > <https://maven.apache.org/plugins/maven-checkstyle-plugin/usage.html>. > > > > > > > > > On Tue, Feb 21, 2017 at 2:13 PM, Otto Fowler <[email protected]> > > > wrote: > > > > > > > I would also like the idea configuration. > > > > > > > > > > > > On February 21, 2017 at 17:06:09, Casey Stella ([email protected]) > > > wrote: > > > > > > > > +1 to blanket reformat as well. > > > > > > > > On Tue, Feb 21, 2017 at 1:25 PM, Otto Fowler < > [email protected]> > > > > wrote: > > > > > > > > > +1. I agree with Michael’s points. > > > > > > > > > > > > > > > On February 21, 2017 at 16:23:21, Michael Miklavcic ( > > > > > [email protected]) wrote: > > > > > > > > > > +1 to a blanket reformat, failed build for improper formatting, and > > > > > automated formatting. I strongly prefer to remove "thinking" from > my > > > code > > > > > formatting and it has worked very well for me on large projects in > > the > > > > > past. There is capability now in IntelliJ to work with Checkstyle > as > > > > well. > > > > > https://youtrack.jetbrains.com/issue/IDEA-61520#comment=27-1292600 > > > > > https://plugins.jetbrains.com/idea/plugin/1065-checkstyle-idea > > > > > > > > > > A quick search didn't yield any obviously robust tools for > automating > > > the > > > > > formatting other than an older non-maintained project named > Jalopy. I > > > > think > > > > > the checkstyle integration with IntelliJ and Eclipse should suffice > > > since > > > > > the Maven plugin would give devs the ability to run checks locally > > and > > > in > > > > > Github via Travis. > > > > > > > > > > > > > > > On Tue, Feb 21, 2017 at 12:32 PM, Nick Allen <[email protected]> > > > wrote: > > > > > > > > > > > I would be in favor of a blanket, reformat. Whether that is for > the > > > > > entire > > > > > > code base or one project at a time. Might be able to conquer and > > > divide > > > > > > some of the heavy-lifting of testing, if we do a project at a > time. > > > But > > > > > > whichever way you think is easier. I'd be glad to help. > > > > > > > > > > > > On Tue, Feb 21, 2017 at 1:57 PM, Justin Leet < > > [email protected]> > > > > > > wrote: > > > > > > > > > > > > > I already tried a blanket, manual reformat the other day, > through > > > > > > > IntelliJ. I did every file matching *.java in the project and > it > > > was > > > > > > > pretty quick. I didn't validate everything looked perfect > > > afterwards, > > > > > > but I > > > > > > > did click into a few files and things looked fine. I'm not > quite > > > sure > > > > > > what > > > > > > > the lifecycle of our autogenerated stuff is, so we'd want to > > regen > > > > > > > afterwards, but it's a pretty trivial thing to do. > > > > > > > > > > > > > > I'm sure there's more nuance (and definitely more testing) than > > > that, > > > > > but > > > > > > > off the top of my head I'm not sure what it would be. Either > > way, I > > > > > don't > > > > > > > think there's a huge amount of effort to just do the reformat, > > but > > > > we'd > > > > > > > still want to spin everything up and test it and so on. It's > > > probably > > > > > > more > > > > > > > work for everybody to rebase onto the (vastly) reformatted code > > > than > > > > > > > anything else, which will vary pretty significantly. > > > > > > > > > > > > > > For (slight) context, the changes are enough to eliminate ~5k > > > > > checkstyle > > > > > > > warnings (and there might be more if we have to tweak anything > in > > > the > > > > > > code > > > > > > > formatting). > > > > > > > > > > > > > > On Tue, Feb 21, 2017 at 10:34 AM, Casey Stella < > > [email protected] > > > > > > > > > > wrote: > > > > > > > > > > > > > > > Any idea, with those modifications to checkstyle, how much > > effort > > > > it > > > > > > will > > > > > > > > take to reformat the code to conform? > > > > > > > > > > > > > > > > On Tue, Feb 21, 2017 at 8:23 AM, Justin Leet < > > > > [email protected]> > > > > > > > > > > > > wrote: > > > > > > > > > > > > > > > > > As part of: > > > > > > > > > https://issues.apache.org/jira/browse/METRON-726 > > > > > > > > > https://github.com/apache/incubator-metron/pull/459 > > > > > > > > > > > > > > > > > > I integrated checkstyle into the mvn:site command, and have > > > > > > checkstyle > > > > > > > > > reports being run as part of the mvn:site reporting. I > expect > > > to > > > > be > > > > > > > > > celebrating hitting 25k checkstyle warnings soon. > > > > > > > > > > > > > > > > > > I tested out creating a code formatting setup in IntelliJ, > > > with a > > > > > > > couple > > > > > > > > > slight modifications of the default Sun conventions > (extended > > > the > > > > > > > > character > > > > > > > > > limit of a line past 80 and made it two space indents). > Given > > > > that > > > > > > > > > checkstyle includes it as a default option, it's probably > > > > > reasonably > > > > > > > > close > > > > > > > > > to the Sun conventions. I'm thinking we probably also at > > least > > > > > create > > > > > > > an > > > > > > > > > Eclipse profile, to open up ease of development. > > > > > > > > > > > > > > > > > > There's probably also a discussion about how exactly we > want > > to > > > > > > enforce > > > > > > > > it. > > > > > > > > > Is it just something we add to the PR checklist and have > > > > reviewers > > > > > > > give a > > > > > > > > > glance, do we setup a hook to autoformat code, etc? > > > > > > > > > > > > > > > > > > Justin > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > >
