Hi Stuart,
Thanks for checking that out for me! Sorry for not realizing in the
first place.
I of course would be happy to submit a patch. Should I submit that
here or over on the assembla page?
On Jul 13, 9:10 am, Stuart Halloway <stuart.hallo...@gmail.com> wrote:
> Hi Cam,
>
> Your tests aren't testing the interesting part without a doall.
>
> That said, my quick tests with doall show your approach faring even better.
> :-) Also, I think what my-flatten does with Java arrays is intuitive (and the
> current flatten not so much).
>
> A patch that preserves the semantics of the existing flatten (except for
> working with Java arrays) would be welcome.
>
> Thanks!
> Stu
>
>
>
> > Another flatten thread! Sorry..
>
> > Hello all, before I realized there was a flatten in the master branch
> > (and before I looked at contrib) I wrote this pretty standard code:
>
> > (defn my-flatten [coll]
> > (lazy-seq
> > (when-let [coll (seq coll)]
> > (let [x (first coll)]
> > (if (sequential? x)
> > (concat (my-flatten x) (my-flatten (next coll)))
> > (cons x (my-flatten (next coll))))))))
>
> > (There's very similar versions on the boards. I'm not claiming this is
> > anything amazing or unique.)
>
> > It's not as elegant as what's in core, but in my micro benchmarks (ran
> > on my laptop; 2.26 core 2 and 4gb ram) it seems to perform a bit
> > better, _especially_ in the already flattened case. It behaves just
> > like core/flatten except that it doesn't return an empty list when
> > passed a map or set, it just returns whatever you gave it but with the
> > top level converted to a seq. I'm pretty much a clojure noob, so are
> > there any hidden detractors of this implementation as opposed to the
> > version introduced in 1.2?
>
> > Also, quick note, if you swap the call to sequential? with seqable?
> > from contrib/core, it flattens maps and sets like you'd expect as
> > well.
> > Here is how it looks
> > user=> (my-flatten #{1 {2 3} 4 [5 6 7 #{8 {9 10}}]})
> > (1 2 3 4 5 6 7 9 10 8)
>
> > And for the micro-benchmarks (using "sequential?"):
>
> > user=> (time (dotimes [_ 1e7] (flatten [1 2 3 4])))
> > "Elapsed time: 14,661.592 msecs"
> > nil
>
> > user=> (time (dotimes [_ 1e7] (my-flatten [1 2 3 4])))
> > "Elapsed time: 922.268 msecs"
> > nil
>
> > user=> (time (dotimes [_ 1e7] (flatten [1 [2 [3 [4 [5 [6 [7 [8]
> > [[[9]]] 10 [11] 12 [13 14 [15]]]]]]]]])))
> > "Elapsed time: 18,147.959 msecs"
> > nil
>
> > user=> (time (dotimes [_ 1e7] (my-flatten [1 [2 [3 [4 [5 [6 [7 [8]
> > [[[9]]] 10 [11] 12 [13 14 [15]]]]]]]]])))
> > "Elapsed time: 6,088.914 msecs"
> > nil
>
> > user=> (time (dotimes [_ 1e7] (flatten [[1 2 3 4 5 6 7 8 9 10]])))
> > "Elapsed time: 11,696.693 msecs"
> > nil
>
> > user=> (time (dotimes [_ 1e7] (my-flatten [[1 2 3 4 5 6 7 8 9 10]])))
> > "Elapsed time: 1,533.983 msecs"
> > nil
>
> > Thoughts?
>
> > --
> > 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
--
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
