When writing more involved Babashka scripts, we usually want to extract some common functionality out into library files. Yet there seems to be no established default, where Babashka is looking for those. So if you want to require a lib-namespace, you can set it up like this:

• Create a lib.bb (or lib.clj) file; if both exist, the .bb one is used; add (ns lib) to the top.
• Add . to the class-path.

The later can be done via:

• Run Babashka with the -cp parameter:

bb -cp . run.bb

• Create a bb.edn file and set :paths:

{:paths ["."]}

Slightly more involved ways can be found in the project setup section of the Babashka docs

Started a project to put together the more intricate correlations that make the Vaadin Flow framework tick.

Vaadin Flow Cheat sheets and flow diagrams

First one is the navigation life cycle:

Since Maven Central changed the search I had to fix my helper script. Time to try out Babashka.

#!/usr/bin/env bb
(defn artifact [{:keys [g a v latestVersion]}]
  (str g ":" a ":" (or v latestVersion)))

(defn query [q]
  (let [query-params {:start 0 :rows 50}]
    (if (str/includes? q ":")
      (let [[g a] (str/split q #":" 2)]
        (assoc query-params
               :q (str "g:" g " AND " "a:" a)
               :core "gav"))
      (assoc query-params :q q))))

(defn request [q]
  (-> (org.httpkit.client/get
        "https://search.maven.org/solrsearch/select"
        {:content-type "application/json"
         :query-params (query q)})
      deref
      :body
      (cheshire.core/parse-string true)
      :response
      :docs))

(assert (= 1 (count *command-line-args*)))

(let [[q] *command-line-args*]
  (->> (request q)
       (map artifact)
       (run! prn)))

Same as with Groovy beforehand:

(defmacro map-of
  [& xs]
  `(hash-map ~@(mapcat (juxt keyword identity) xs)))

(let [a 1
      b 2]
  (println (map-of a b)))
; {:b 2, :a 1}