On Tue, Jun 29, 2010 at 17:24, Meikel Brandmeyer <m...@kotka.de> wrote: > Hi, > > On Jun 29, 4:24 pm, B Smith-Mannschott <bsmith.o...@gmail.com> wrote: > >> This fully realizes the (range 10000) before calling foo, which is >> wasteful if foo is written to consume more incrementally. >> >> I've run into this trouble with concat, when I've wanted a lazy >> sequence of the concatenation of the sequences produced by a list >> comprehension: >> >> (apply concat (for [n (range 1000)] >> (range 100))) >> >> Concat gets passed a fully realized sequence of 1000 lazy sequences of >> 100 integers. > > I don't think that this is true. Otherwise the following wouldn't > work: > user=> (defn take-5 [& more] (take 5 more)) > #'user/take-5 > user=> (apply take-5 (iterate inc 0)) > (0 1 2 3 4) > > iterate creates an infinite sequence. So if the problem was apply, the > second call should not return. > > If more than necessary is realised, I suspect some effect with chunked > seqs.
There's nothing quite as powerful as a good counter example. You're quite right. I guess what I'd observed at the time I "discovered" this must have been the effect of chunking. At the time, I did attempt the understand the implementation of apply and through I was seeing the args sequence being realized, but I guess I must have misunderstood the code. // Ben -- 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
