sci 0.3.5 seems to have dropped default binding for str. noticed when bumping up deps on malli, so:
;; works on 0.3.4, not on 0.3.5
(require '[sci.core :as sci])
(sci/eval-string "(str/upper-case \"kikka\")")
Is this intentional? don’t see it on the https://github.com/babashka/sci/blob/master/CHANGELOG.md#v035. 🤔I can easily change the transformation docs & tests to use something else, but would like to document this to Malli CHANGELOG too.
Yes, it was intentional. Having the default str alias in the user namespace can be configured with:
{:aliases {'str 'clojure.string}}
Not sure where I removed it anymore
I think the default configuration of SCI should be unopiniated, which was basically the reason
That’s a good call. I first ported all examples to use the long form but the examples looked bad and added the alias into malli-sci default config. thanks!
Shaving off some size of the advanced compiled output: a 160kb win.
(but now self-hosted CLJS is unhappy..)
kinda wild, but using (or) with 20 clauses doesn’t work but 19 does in my SCI compiler
(or
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
;nil
true)
=> true(or
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
true)
=> nilIs this JVM or CLJS?
jvm
issue welcome
+1 let me confirm I can repro
I can
oh nice, I’ll open an issue
found the issue
Pushed the fix
Did you solve your issue with clojure.test the other day?
yea! way to remember. ended up being a mixture of clojure.test parallelization pain
what I ended up with was a redefs.clj namespace, and a macro that does all the things necessary to make clojure.test run in isolation
it just took a bit of trial and error to get the magic recipe, thank you so much again!
will this be autodeployed? should I monitor this page https://clojars.org/org.babashka/sci
you're not using git deps I assume?
leiningen
I can do another mvn release
🙏🏻
Will do in about 30 minutes
thank you!
30 minutes? What happened to customer service? SMH
🤡
Truly a remarkable individual, cant even come close to overstating it
0.3.32