(apply a-fn input-1 (apply concat the-map)).

if you're writing a-fn consider using (defn a-fn [input-1 & {:as input-pairs}] ...) that way you can use both: https://clojure.org/news/2021/03/18/apis-serving-people-and-programs

a-fn actually was a macro 😊

You can't do it with a macro then. Unless you're doing it in another macro.

If it's the latter, then (a-fn input-1 ~@(apply concat the-map))` or something like that - depending on the outer macro.

You can use #'identity only because vars are callable.

The Var class implements IFn and delegates the implementation to the wrapped function, if it's a function.

You can still use @#'xs.

So, given that:

user> (def s #{1})
;; => #'user/s
user> (filter #'s [1 2 3])
;; => (1)
user> (filter #'s #'s)
Error printing return value (IllegalArgumentException) at clojure.lang.RT/seqFrom (RT.java:557).
Don't know how to create ISeq from: clojure.lang.Var
user> 

Thanks a lot for the explanation 🙂

Is there like a tome of "Curious Implementation Details" with stuff like this for Clojure? Or are you it? 😄

I think I recall Alex Miller saying something about a Clojure Puzzlers book, in the vein of the Java Puzzlers one 😉

https://clojure.org/reference/special_forms

Got it. So we can't resolve them. Is there any way to determine a symbol is special form?

i’d look into using clj-kondo programatically to get the analysis

@U03UFHD3W1H https://clojuredocs.org/clojure.core/special-symbol_q

@U11BV7MTK I didn't know clj-kondo can be used as parser/analyzer. Should play with it)

@U0739PUFQ This will help me, thanks!

So, after a day of debugging it looks like this is most likely an issue with Loom more than anything. Calling set-agent-send-off-executor! will cause pmap to use that executor. Deeply nested spawning of threads seems to lock up loom easily (ie. a few level of pmap nesting). Below is a code snippet that can be used to see it:

(defn ppmap
  [f coll]
  (->> coll
       (mapv
         (fn [x]
           (let [cf (java.util.concurrent.CompletableFuture.)]
             (Thread/startVirtualThread
               (bound-fn []
                 (try
                   (.complete cf (f x))
                   (catch Exception e
                     (.completeExceptionally cf e)))))
             cf)))
       (map deref)))

(let [n 4]
   (->> (range n)
        (ppmap
          (fn [_]
            (->> (range n)
                 (ppmap (fn [_]
                          (->> (range n)
                               (ppmap (fn [_]
                                        (ppmap identity (range n))))
                               (apply concat)
                               (into #{}))))
                 (apply concat)
                 (into #{}))))
        (apply concat)
        (into #{})))

(haystack contributor here) I just found this :) Interesting insights! This seems another reason why we shouldn't add pmap in Haystack and sibling projects. Don't hesitate to create issues in Haystack/etc as you find them.

it returns the first item that matches a predicate. It doesn’t really pass items on in a chain. Not sure what the transducer would do. Does the cgrand lib have a version?

Right, I was confused. https://github.com/cgrand/xforms/blob/master/src/net/cgrand/xforms.cljc#L769 Here is what I meant, some is the reducing function, not a part of the transducers. Thanks!