Let's look at the different flavors of function composition in Clojure.
There is a function in clojure.core called
comp that takes a set of functions and returns a function that is a composition of those functions:
(defn doubler [x] (* x 2)) (defn incrementer [x] (+ x 1)) (def doubler-and-incrementer (comp incrementer doubler)) (map doubler-and-incrementer [1 2 3 4]) ; => [3 5 7 9]
clojure.core/incexists, but I recreate it here to be explicit
Note that when composing functions with comp, they are applied right to left to their arguments:
(def incrementer-and-doubler (comp doubler incrementer)) (map incrementer-and-doubler [1 2 3 4]) ; => [4 6 8 10]
Another function in clojure.core is
partial that takes a function and some arguments to that function and returns a new function that will accept the remaining number of arguments. This can be useful when combined with comp and writing a custom reducer:
(def incrementer-and-doubler (comp (partial map doubler) (partial map incrementer))) (incrementer-and-doubler [1 2 3 4]) ; => [4 6 8 10]
Clojure 1.7 will introduce transducers, which will likely become the idiomatic way of accomplishing this same task.
A more popular method of composition are the thread-first and thread-last macros.
The thread-first macro evaluates the first expression then passes the result into first argument of the next form, and the result of that into the first argument of the next form and so on.
In other words, this:
(doubler (incrementer 2))
Is the same as this:
(-> 2 incrementer doubler)
Thread-last works the same way, but it passes the result of each expression into the last element of each successive form.
(->> [1 2 3 4] (map incrementer) (map doubler))
Is the same as:
(map doubler (map incrementer [1 2 3 4]))