Clojurians Log v2

> ~"foo"
-1
> ~NaN
-1
> ~["foobar"]
-1

I can't recall ever using ~ operator in JS

yes

well, ~{} is used in js* ;)

oh you could probably do the same for conj and vectors - in my datastar-expr library I had these implemented in terms of merge and concat (I had the benefit of assuming an even thinner translation barrier than squint's and narrowing the semantics of these functions, ie no clever inference needed)

(mapv emit `[(merge) (merge m1) (merge m1 m2)
             (concat) (concat v1) (concat v1 v2)])
;; => ["{}" "{...m1}" "{...m1,...m2}"
;;     "[]" "[...v1]" "[...v1,...v2]"]


(mapv emit `[(merge m {:a 1}) (merge {:a 1} m) (merge {} m {})
             (concat xs [1 2 3]) (concat [1 2 3] xs) (concat [] xs [])])
;; => ["{...m,'a':1}" "{'a':1,...m}" "{...m}"
;;     "[...xs,1,2,3]" "[1,2,3,...xs]" "[...xs]"]

(emit `(assoc m :k1 v1 k2 v2)) ;; => "{...m,'k1':v1,[k2]:v2}"
(emit `(conj xs 1 2)) ;; => "[...xs,1,2]"
(emit `(cons 1 xs)) ;; => "[1,...xs]"
(emit `(list* 1 2 3 xs)) ;; => "[1,2,3,...xs]"

(emit `(get (assoc {} :a 1) :a)) ;; => "({'a':1}).a"

yes, planning to do that next

yeah thanks for the merge idea, that's neat

feel free to post these examples here: https://github.com/squint-cljs/squint/issues/728

I mean, only the merge and concat ones, the rest isn't going to work in squint since it has a distinct list type

yeah I guess it depends how faithful to clojure's semantics you want to be re. concat returning a lazy seq

(oh nice, I realised my current impl for (get m k nf) => "m[k]??nf" , the in operator is definitely a better way to go)

yes, in :)

cc @chris358

Every time I see a Eucalypt example I have the same feeling: "I can't believe this works".

https://github.com/squint-cljs/squint/issues/732

thanks!