actually, it doesn't sort them as last. It seems to sort them somewhere between read*-`reae*`.

(:require [p :as p]
          [q :as q]
          [r :as r]
          [ra :as ra]
          [rb :as rb]
          [rc :as rc]
          [rd :as rd]
          [re :as re]
          [rea :as rea]
          [read :as read]
          #?(:clj [foobar :as f])
          [reader :as reader]
          [reae :as reae]
          [rf :as rf]
          [s :as s]
          [t :as t]))

I think you are the first one to complain about that, so not sure if this is a common need. The reader conditional between read and reader makes no sense to me, do you know why that?

No idea. My original plan was to fix all the unsorted namespaces in our project, and while looking at the dry run output I noticed that those reader conditional rows were sorted in a funny way

Then I duplicated it in emacs using the normal clean ns code action

Hum, I see, it's clojure-lsp that is moving to that position, I thought you said your project had it in that way on purpose

Nope, we typically have those reader conditional rows first

Alright, A issue makes sense, not sure if we should always sort the reader conditionals as last or first, but the current way looks weird indeed

I can test this a bit more and make an issue

Thanks!

I think this is a clojure-lsp bug, we only consider var-definitions on the outline

Please open an issue on clojure-lsp with detailed report

Ok, thanks.

🙏

I’m actually not sure what to ask for. There’s both the method and the keyword, (of course, it doesn’t have to be a keyword).