rubocop-rails VS constant_sandbox

1) Auto-correcting a whole (large) codebase at once with tons of offenses and dozens of active branches should be used with caution. Merge conflicts, blame pollution (ok, can be solved with .git-blame-ignore-revs, though can hardly remember any project using it). Though, the most important argument is that auto-correct can introduce bugs. Unfortunately, even safe autocorrect can be unsafe. Recently, I broke one popular project (with a decent, but not 99.999% test coverage) with a single "safe" auto-correction commit 🙂 (This issue).

It's also possible to extend RuboCop through additional linters and formatters. You can build your own extensions or take advantage of existing ones if they are relevant to your project. For example, a Rails extension is available for the purpose of enforcing Rails best practices and coding conventions.

I'm still getting used to writing these. Still, this article from Evil Martians has been a big help. The rubocop-rails codebase also had some cops similar to what I wanted to put together. The cop we've put together checks if the class inherits from ActiveModel::Serializer and adds an offence to that line.

Let's take a glance at the action_filter cop real quick here, but just a quick part of it:

Following on from this, I highly recommend setting up your editor to automatically lint Ruby files with RuboCop and its Rails extension and start adapting your code to adhere to the Ruby Style Guide.

I just would like to point out that even though that is the most sane way, it comes with it owns set of problems. One of them is when developers start to code to cheat the linter, or they complicate the code just to "make the linter happy", another is when the linting rule introduces problems/errors like https://github.com/rubocop-hq/rubocop-rails/issues/418