Clojurians Log v2

Guardrails 1.1.3 is on Clojars. This release fixes a bug in async mode where errors were not being reported for Clojure. Clojurescript was not affected. https://github.com/fulcrologic/guardrails

:reveal: New version of #reveal — https://vlaaad.github.io/reveal/ `1.3.193` — is out! For this major release I focused on providing a way to interact with Reveal window by submitting commands, allowing to easily build better IDE integration. You can learn more in the new readme section: https://vlaaad.github.io/reveal/#interacting-with-reveal-from-code Here is a little demo using Cursive:

Released babashka 0.2.8 with additional built-in libraries: - core.match (much requested) - clojure.test.check (prep for including spec once it comes out of alpha) - hiccup (who doesn't use it?) https://github.com/babashka/babashka/blob/master/CHANGELOG.md#v028 Hop by in #babashka for questions and complaints ;)

Where is babashka.sh? 😃

@pez this is called babashka.process

Yeah, babashka.process is wonderful. I confused things and meant to lob for babashka.fs.

@pez That lib is not done yet. I'm stuck on the glob function :/ If you want to help me, that'd be great.

It is high on my list though

glob happens to be what I most want to have. 😃 I would love to help, even if I doubt I know how to. How are you stuck?

@pez Check issue 4 and 5.

I just have to share how I do it in lack of babashka.fs.

(defn glob [pattern]
  (-> (shell/sh "bash" "-c" (str "ls " pattern))
      :out
      (#(when-not (= % "")
          (string/split % #"\n")))))

so you are using the bash definition of glob. there are different defaults, recursive, not recursive

see https://github.com/babashka/fs/issues/5

Are there difficulties in implementing the recursive version or is it more a matter of deciding which one to use?

deciding which one to use as the default, also, include hidden dirs by default or not?

I have a work in progress version in the glob branch which can be run only from clojure currently

because I'm using java.nio.file.DirectoryStream which is not yet available in bb

but for testing, I would use that one

Could it be an option? I think non-recursive makes a good default, if I can opt in on the recursion.

@pez That's what I've done in the glob branch:

$ clj -A:babashka.fs/dev
Clojure 1.10.2-alpha2
user=>  (require '[babashka.fs :as fs])
nil
user=> (fs/glob "." "**/*.md")
[]
user=> (fs/glob "." "**/*.md" {:recursive true})
[#object[sun.nio.fs.UnixPath 0xdab48d3 "/Users/borkdude/Dropbox/dev/clojure/babashka/sci/CHANGELOG.md"]

but it might be a bit silly when searching for "**/*.md" to not have it recursive, since it won't match anything

Yeah. So that explains why some ignore files do not seem to work for me.