testfile.clj

;; http://pragprog.com/magazines/2012-03/comparing-java-and-scalas-expressiveness
;; first attempt for impl in clojure.. a lot to be improved! :)

(let [lines (map #(clojure.string/split % #" ")
                 (line-seq (clojure.java.io/reader (clojure.java.io/file "textfile.txt"))))
      grouped  (group-by #(second %) lines)
      parse-long (fn [log] (Long/parseLong (first log)))
      diff-times (fn [longs] (for [x (range (count longs)) :let [ts (nth longs x)] ] 
                              (if (= 0 x) ts
                                  (- ts (nth longs (- x 1))))))
      times (map (comp diff-times #(map parse-long %)) (vals grouped))]
  
  (print (zipmap (keys grouped) times)))

It is just 1000 times better than the two samples from the article. I am not questioning - and should not - if its funcitonal or non-functional. But the basic patterns of good code are much more present here than there. You dont have useless code, you have split the algorithms into parts and put them together.

Aniche likes to say that people know how to break code into pieces but do not know how to put them together. Thats what you do here... thats what he didnt do there (even breaking it...)

for sure, the Java version can be a lot better.
but C#'s LINQ is the best for now https://gist.github.com/2007893 :)

I've been using google collection/lambdaj/totallylazy/lazyrecords in the java situation to help too

…

On Fri, Mar 9, 2012 at 5:08 PM, Pedro Henriques dos Santos Teixeira ***@***.*** wrote: for sure, the Java version can be a lot better. but C#'s LINQ is the best _for now_ https://gist.github.com/2007893  :) --- Reply to this email directly or view it on GitHub: https://gist.github.com/2008267