Are you using java.time.Instant to get the original UTC?

Using moment preferably

This is the code I am trying to convert (.utc moment date) I already take the date from the backend

Sorry, I don't know how moment works 🤷

Here it is using java.time:

(import [java.time ZoneId Instant])

(-> (Instant/parse "2022-11-24T11:50:26Z")
    (.atZone (ZoneId/of "Asia/Jerusalem")))

Moment is js right?

yes

https://momentjs.com/timezone/

you'll want to install the additional moment-timezone package

specifically, here's the docs you'll use https://momentjs.com/timezone/docs/#/using-timezones/converting-to-zone/

May also want to call https://momentjs.com/timezone/docs/#/using-timezones/getting-country-zones/ to get the zone names for Israel

Okay lets say I wanna use this exactly how do I convert it to clojure (sorry Im a beginner): ((date, "America/New_York"))

So I saw the link for timezones, hw do I translate it to Clojure exactly I dont know

???

(moment/tz date "America/New_York")

or (.tz moment date "America/New_York")

the moment reference may be different depending on how you required the libraries

Both give me same result: (println (.utc moment date))

(println (.tz moment date "Israel/Jerusalem"))

When I run that I get

Moment Timezone has no data for Israel/Jerusalem. See .

The TZ name is just "Israel"

Okay thanks turns out its a bug with moment.js How do I do this then?

moment(startdate).add(2, 'hours')

I just used the fn and regex matched against israel

(.add (moment startdate) 2 "hours")

Yeah I found it. Its a bug with ".fromNow date"

Whenever we use .fromNow it doesnt take timezone into acount

I'd suggest reading this post that can help you understand how to access properties and call functions on js objects in clojurescript https://www.verypossible.com/insights/clojurescript-and-javascript-interoperability-a-comprehensive-guide

Thanks alot everyone!

I feel like parens is the recommended one, but can't remember where I saw that

I'm improving clojure-lsp to add imports automatically, and not sure I should add a flag for those or just go with the most recommended way...

I use brackets for all imports, java and clojure

So no tradeoffs, it's just visually in the end? maybe we should mention it somewhere on the style guide

I think How to ns by Sierra mentioned parens, because the first symbol was special, but you could also say: avoid parens since they remind of a function call. 🤷

oh, there was where I saw that mention, thanks Yeah, TBH I don't have a personal preference, so I was expecting someone say: always use that because of X, kind of annoying need to support both by a flag 😔

I personally prefer ( in that case

I feel a configuration option coming up ;)

I use brackets too

i think it varies quite a bit depending on ppl

At least we agree it shouldn't be a set ;)

@U064WBEPL’s https://stuartsierra.com/2016/clojure-how-to-ns.html#vectors-or-lists-1 for reference

It is defined in the style guide (via examples), i.e., https://guide.clojure.style/#sort-requirements-and-imports

And here too: https://guide.clojure.style/#line-break-ns-declaration

I see no :import mention on that example @U11EL3P9U

see just pasted

I always use [foo Bar] , never foo.Bar even when I import a single class

Indeed, I use [foo bar] for import too 🙂

even in clojure itself it all mixed up

Yeah I saw that one, but didn't convinced me as it doesn't explicitly tell about those

Its also referened here: https://clojure.org/guides/learn/namespaces

yeah, ATM clojure-lsp is using full-package-import sym so no vectors or list 😂 (I think I did that in the past to not need to choose between any of those 😂)

Yeah, indeed there are more examples showing [

so maybe go with the [ by default and add a option to use (

👍 (for me!)

thanks all!

in core, the same person uses one or the other depending on the ns 🙂

could be from contrib migration

in 1.1 (?) days

clojure core contributions usually don't change style things, not even whitespace. you can't just hit format in emacs and then write a patch (found this out the painful way of course)

does lsp rewrite the extreme cases like this in require? https://github.com/clj-commons/aleph/blob/master/src/aleph/http.clj#L9-L13

Nope, I know that exists but I have no idea what will happen TBH 😅 , it's not officially supported and IIRC it's not recommended anymore right?

it's quite uncommon tho, I think I have only seen ztellman use that

yeah, maybe it just works by accident

I believe it's not recommended

ah nested libspecs, yeah avoid those ;)

clj-kondo does understand them, but they are painful to work with

it is supported by clojure, not by accident. SCI / bb also supports them (because of compatibility)

"Programming Clojure" 2e by Halloway & Bedra uses parentheses on p.44 and square brackets on p.72

i wanted to use those in clojurescript because typing my-app.components.pages or my-app.api.middleware 67 times is a pain but sadly it's not in cljs 😞

I do [] and have heard Rich say the same

something I noticed yesterday while playing with both styles + clojure-lsp: [a.b c\nd] will get c indented under a, and with (...) c and d will form a column

(:import (io.helidon.common.http Http$HeaderValue
                                 Http$Status))

(:import [io.helidon.common.http Http$HeaderValue
          Http$Status))

I find the former more readable personally, is that something we can/could customize if one would like to use [] and have the same indentation as with ()

the latter should hot happen, clojure-lsp should move all to the same line or all to the next line, BTW I made some improvements on that yesterday, maybe test with nightly #C032YH7P0R2

great, I ll have a look, thanks

on clojure-lsp 2022.11.24-22.12.04-nightly it's unchanged

oh no nevermind

clean namespace shuffles things like you said, my remark was about format-buffer

hum, so cljfmt is doing that

yup

I suppose it's possible to fix with a custom indent rule in cljfmt

maybe, one thing I'm considering doing is calling cljfmt format after a clean-ns, to keep consistency, needs more research to avoid issues like that one

it's tricky. if you have format buffer on-save-hook or something it gets a bit heavy

ah, my idea was to do that programaticaly on clojure-lsp side:

1. user reuqests clean-ns refactoring on editor 
2. clojure-lsp does the clean-ns
3. clojure-lsp call cljfmt passing only the ns code to format
4. clojure-lsp returns the clean-ns code for editor apply

you'd have to settle on a "bunching" style for when there are multiple classes aliased for a package tho

whereas now it just accepts whatever happily

within some formating rules still

yes, we would need to check if there is any config we always want to send + the specified by user on settings.