hi, trying to follow a d3 example, not sure why the following translation doesn’t work... https://github.com/d3/d3-scale js:

var x = d3.scaleLinear()
    .domain([10, 130])
    .range([0, 960]);

x(20); // 80
x(50); // 320

(require '[cljsjs.d3])
(def x
  (-> js/d3
      (.scaleLinear)
      (.domain [10, 130])
      (.range [0, 960])))

(x 20) ;; nil
(x 50) ;; nil

ƒ scale(x) {
    return (output || (output = piecewise(domain, range$$1, clamp ? deinterpolateClamp(deinterpolate) : deinterpolate, interpolate$$1)))(+x);
  }

@mbossenbroek try using the #js tag before those vectors, D3 likely expects JS arrays https://cljs.github.io/api/syntax/js-literal

perfect! I never would have guessed that one. 🙂 Thanks!

(require '[clojure.tools.reader.edn :as edn])
;=> nil
(edn/read-string "#inst \"2010-11-12T13:14:15.666\"")
;=> #error {:message "No reader function for tag inst.", :data {:type :reader-exception, :ex-kind :reader-error}}

I thought this should create an #inst?

@phill It does, (fiddlyness with records being a mess from my perspective). Lots of unobvious steps that I had to do. I was thinking of turning it into a macro, but I don’t have too much experience writing those, so I was worried I’d end up putting together some ghastly thing that would invariably trip me up in future…

@lilactown the end reader doesn't have any reader tags registered by default, as it doesn't use the global var but takes a map parameter

that matches the clojure.edn api so we're not going to diverge there

@bronsa AFAICT both clojure.tools.reader.edn and clojure.edn parse #inst just fine

$ rebel-clj 'org.clojure/tools.reader {:mvn/version "1.3.2"}'
[Rebel readline] Type :repl/help for online help info
user=> (require '[clojure.tools.reader.edn :as edn])
nil
user=> (edn/read-string "1")
1
user=> (edn/read-string "#inst \"2010-11-12T13:14:15.666\"")
#inst "2010-11-12T13:14:15.666-00:00"
user=> (require 'clojure.edn)
nil
user=> (clojure.edn/read-string "#inst \"2010-11-12T13:14:15.666\"")
#inst "2010-11-12T13:14:15.666-00:00"

I’m basing this off of the example usage on the clojure.tools.reader lib anyway

@thheller maybe stupid question, but how did you make shadow-cljs popular on the beginning of the story? How did you spread information to people about your project?

Just thinking how to tell people about library, while it can’t be find in google on the beginning of the story 🙂

Has anyone here used devcards with generative tests? If so, whast’s your setup?

i don't understand why you would mix the two. Are you using devcards as your test runner? We only use devcards as a ui test

We've used devcards as a tet runner for a little bit but switched to bruce's cljs test runner. Devcards had some issues with async tests and the cljs test runner is really just a test reporter and uses the normal testing to run so everything is supported

@dpsutton yup I am using it as my test runner

if you switch to https://github.com/bhauman/cljs-test-display i think you will enjoy it more. I think i remember with devcards you need to navigate to the card namespace to run those tests

also the favicon and os notification features of cljs test display make it a real pleasure

Ok, I’ll try that right away; thanks!

@hmaurer I absolutely agree with @dpsutton

My main reason to use devcards as a test runner was to have a visual interface to look at tests, and it seems like what you linked me provides just that

yes. i think you will be tweeting about your better testing experience soon 🙂

😄

and for my generative tests, I would like to do something like “test passed if I can stest/check 1000 examples on my function and no error is thrown”

also it should be trivial to drop in since devcards exposes the same deftest and is forms i think ? you just replace the require statement and should be up and runnin

how can I express this with clojure.test?

yup t does; I basically just need to figure out how to set this up with my shadow-cljs tooling, but it shouldn’t be too long

https://github.com/clojure/test.check#clojuretest-integration looks promising

i haven't used test check in anger

ah, thanks!

in anger? 😄

in a real project versus just reading and poking in a repl

also while you are doing this, if you are using figwheel-main you should look into extra mains. Running tests, devcards, and app simultaneously is a really nice way to work

Is there a way to know which dependency is having big influences on the minified size?

@frozenlock if you’re using shadow-cljs, you can generate a build report: https://shadow-cljs.github.io/docs/UsersGuide.html#_build_report

Not using it 😞

@frozenlock node modules? cljsjs?

cljsjs

oh cljsjs is full size of the library, it can increase app very fast

even if you use 10% of libraries files

cljsjs still gets run through Closure and should get dead-code eliminated

I found a stackoverflow asking this in April and it looks like the only answer was “use shadow-cljs” 😕

Another solution is to use fighweel-main with {:npm {:bundles {"dist/index_bundle.js" "src/js/index.js"}}}

I know only this two solutions

https://github.com/kwladyka/form-validator-cljs/tree/doc here is an example of figwheel-main + clj as it is something new and not everybody know about it

https://github.com/kwladyka/form-validator-cljs/blob/doc/src/js/index.js - this is probably the most interesting file

Everything else is automated, but this file has to be coded manually

Thank you for the info

I'm still on the old figwheel. I guess I should update anyway. I could use this as an opportunity to try shadow-cljs.

shadow-cljs is cool tool too

Wait... does shadow-cljs allow you to avoid relying on cljsjs?

yes, both of this solution

you shouldn’t use clsjs with them

Interesting... well then, I know where my afternoon is going 😛

yeah, cljsjs can still be used for externs but 90% of the time is irrelevant

also AFAICT even figwheel-main doesn’t give you any mechanism to get a fine-grained bundle size report

I suppose lein-externs can do the job? (I already use it for library minified not on cljsjs.)

Or are externs somehow irrelevant with shadow-cljs?

IME shadow-cljs does a really good job of externs-inference, but there are ways to configure it: https://shadow-cljs.github.io/docs/UsersGuide.html#externs

I think you have better chance to find something which will show you node modules size

@lilactown not true that cljsjs goes through Closure

foreign libs never go through Closure, it cannot work

my mistake

and it doesn’t matter what tool you use really

ClojureScript, Figwheel, or shadow

stuff that cannot survive advanced compilation just doesn’t work

to avoid any confusion the rule is simple

if you did not write it in ClojureScript, it cannot take advantage of Closure

the only other way is to write Closure compatible JS, like transit-js - but hardly anyone is doing that

so if you’re using something from NPM, or some random JS dep - it’s going to bloat your application, plain and simple

and nothing is going to fix that (at least not with Webpack or whatever)

maybe folks should be looking more closely at ClojureScript + https://rollupjs.org/guide/en

I think shadow-cljs runs npm deps through simple optimizations IIRC

which doesn’t do that much

but it’s definitely not standard

and yeah, there’s only so much you can do

the thing you really want tree-shaking only works in advanced