Make PersistentHashSet implement IReduce

Question

This improves speed of many reduce based operations on set which were falling back to seq-reduce, including code in clojure.set namespace such as clojure.set/union and (into (link: ) some-set).

I've included a few benchmarks I performed using simple-benchmark in a JavascriptCore environment (Planck REPL)

Answer 1

Comment made by: rohitaggarwal

I think the code currently is faithful to Clojure's implementation of PersistentHashSet. So any change from that would probably require more thought and/or history.

Also someone else also raised a similar issue on (link: https://groups.google.com/forum/#!topic/clojurescript/h-EYTTndvAY text: ClojureScript mailing list).

Answer 2

Reference: https://clojure.atlassian.net/browse/CLJS-1712 (reported by tmulvaney)