How to write concise code in Clojure

During code reviews I’ve seen the following repetition pattern a lot. I am going to use Clojure to illustrate, but it also happens in other programming languages:

(ns myapp.butterfly)

(defn create-butterfly
  [butterfly-attributes]
  ..)

(defn calculate-butterfly-wing-size
  [butterfly-type]
  ..)

Notice how butterfly is being repeated. Imagine how using unnecessary long symbols over and over again will lengthen the code, and slowly shroud the purpose of the function in unnecessary noise.

butterfly-attributesare already in the context of a create-butterfly function, which in turn already resides in a .butterfly namespace.

I will argue that the following is better. The code is more concise without loosing its meaning because the namespace provides a meaningful context.

(ns myapp.butterfly)

(defn create
  [attr]
  ..)

(defn calc-wing-size
  [type]
  ..)

I took the liberty to shorten attributes and calculate with common abbreviations, just like Clojure does with concat over “concatenate”.

Code from a different namespace, would have looked like the following:

(ns myapp.other-ns
  (:require [myapp.butterfly :refer [create-butterfly]]))

(create-butterfly {:name "Brimstone" :color "green"})

and now it can look like this:

(ns myapp.other-ns
  (:require [myapp.butterfly :as butterfly]))

(butterfly/create {:name "Brimstone" :color "green"})

It is not about making the code as short as possible… but it is. Just not at the cost of context / readability. Clever use of namespaces can help with that.