6.5. Summary

We’ve covered the topics of immutability, persistence, and laziness in this chapter. Clojure’s core composite data types are all immutable and persistent by default, and though this fact might presuppose fundamental inefficiencies, we’ve shown how Clojure addresses them. The implementation of a persistent sorted binary tree demonstrated how structural sharing eliminated the need for full copy-on-write. But structural sharing isn’t enough to guarantee memory efficiency, and that’s where the benefits of laziness come into the fold. The implementation of a lazy, tail-recursive quicksort demonstrated that laziness guarantees that sequences won’t be fully realized in memory at any given step.

In the next chapter, we’ll dive into Clojure’s notion of functional programming. Along the way, you’ll notice that much of the shape of functional implementations in Clojure will be influenced by the topics discussed in this chapter.