[Fix #4179] Prevent Rails/Blank cop from breaking when LHS of is a naked falsiness check

[rrosenblum]

This fixes #4179. This is a similar fix to what was done to fix #4171. This issue does not seem to impact Rails/Present.

[Drenmi]

Kind of an awkward miss by me. I assumed too hastily that the problem was the "not-not" check of the RHS. 😅

[rrosenblum]

Kind of awkward miss by me as well. I feel like it is such a simple example. I feel like I have been testing less edge cases since I started using NodePattern more. I think this is because NodePattern typically creates a more resilient check. I guess I will have to beef up my testing.

[Drenmi]

I feel like I have been testing less edge cases since I started using NodePattern more. I think this is because NodePattern typically creates a more resilient check.

Same. I think NodePattern is amazing for that reason. It's very hard to work with an AST that has so many permutations without it. I agree that we can't rely on it over tests, though. 🙂

Another thing I like about RuboCop is the testing strategy. "This literal code should (or should not) generate an offense" is super easy to understand and, I must admit, RuboCop is the only project for which I always write the tests first. 😶

[bbatsov]

👍

[esambo]

Would be nice to release a new rubocop version with this, as 0.48.0 is broken because of this

[joevandyk]

bump on new rubocop version :)

[bbatsov]

Likely on Monday. The weekend is a good opportunity to fix more bugs. :-)

On Fri, Mar 31, 2017 at 21:56 Joe Van Dyk ***@***.***> wrote: bump on new rubocop version :) — You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub <#4190 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAGVyoOle7U--gviGUZo5j83__80o7Aeks5rrUxLgaJpZM4Mq2TG> .

-- Best Regards, Bozhidar Batsov http://www.batsov.com