On Sun, Nov 8, 2009 at 8:41 AM, Michael Jaaka <michael.ja...@googlemail.com> wrote: > Can any imperative code be transformed to functional equivalent? > (Give an answer in terms of the same way I can answer on recursion and > loops)
Short answer: Yes. Long answer: Yes, but sometimes the transformation can get rather unwieldy. Most of the time, such transformations are easy. For example, when writing the clojure-contrib combinatorics library, I was transforming highly imperative register-based algorithms from Knuth's books into loop/recur using Clojure's functional vectors. It's just as inscrutable to read as the original, but it was fairly easy to transform. The basic technique is to, rather than change something in place and "returning void", you return the changed version from a function. The problem is that sometimes, there are many different pieces of many different states that need to be updated, and you end up passing more and more things around in a functional way until eventually you are threading some monolithic "world state" through all your functions, basically simulating the imperative process. This is where monads can come in handy, to help hide the threading of the monolithic state. But at this point, it's not clear whether you're really any better off than doing it imperatively to begin with. That's why it's good that Clojure has a suite of mechanisms for dealing with state (agents, refs, vars, atoms...) --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en -~----------~----~----~----~------~----~------~--~---
