If I have a collection of collections [[1 2 3] [4 5 6]] How do I get that into just [1 2 3 4 5 6] I've tried

(map #(conj [] %) collection)

(map #(into [] %) collection) 

(into [] cat [[1 2 3] [4 5 6]])

Thank you! and I think this is the first time I've used a transducer directly. I still haven't wrapped my mind around them.

Also

user=> (reduce into [[1 2 3] [4 5 6] [7 8 9]])
[1 2 3 4 5 6 7 8 9]

Also http://chouser.n01se.net/apply-concat/ [=

I've been learning Clojure for about a month now. I'm at a point now where, when programming Java at work, I'm getting quite frustrated not having a REPL to quickly asses my logic :face_with_cowboy_hat:

Are there any plans to support jdk 13? Or does it already work?

Fwiw, I’ve used 13 since it came to Homebrew and haven’t had any problems.

@sfyire Im using Clojure 1.10 with jdk 13

no problems so far

Thank you, I was guiding someone else through the process but discovered the jdk wasn't opening due to apple security constraints, at first it didn't appear in security and privacy

but later on it was in there, and making an exception for it worked

Could anyone point me to a youtube video or something about how to TDD with the repl using cursive. meaning I would like to convert somehow my manual testing in the repl to a test.

without the need of copy pasting etc. Its my understanding there is a way. or maybe a flow for doing this

This is kind of what im looking for

https://vimeo.com/230220635

I cant see what keys he presses to acheive lots of what he is doing

BTW i got this video from this slack

and he doest convert it to an actual test. would love to do that

Hi, I have a vector of [100 0 0 0 0 ...]. I would like to transform it to [100 101 102 103 104 ...]. Any ideas how I could achieve this?

is there a pattern for creating a collection of all numbers that satisfy a predicate?

Could you filter a range? (I'm a beginner too, so you may want to wait for a better answer)

(filter even? (range 1 10))

😁 yes, I think that's exactly what I am looking for, thanks!

what collection and what pred?

So the predicate is a function that checks weather a square on a game-board is empty and the collection would be coordinates. I think it's as simple as filtering a range for my case though, the only difference would be that I'm technically filtering pairs of numbers.

I want to basically pick empty squares from the board at random

and I figure it will be nicer code to choose a random element of a vector/set than to use iteration like I would in php or something

something like (->> board-squares (remove filled-in-squares) rand-nth)

if they are sets (rand-nth (set/difference positions filled-in-squares))

I see, so this way works with a collection of all squares as a starting point.

yeah. depending on how you are representing your state there are lots of ways. you could operate on the actual representation rather than a proxy like coordinates

yeah, the game-boards start out as json files, and I've been interacting with them mostly via coordinates, but the actually data structure is a vector, so that might be cleaner in this case

Hi all. Probably a dumb question, but I'm trying to use a React navigation-bar component (written in Javascript) that takes two sets of children (left and right) as props. I can't seem to figure this out. Here's the simplest case I could come up with: (defn right [] [:div "Right"]) (defn left [] [:div "Left"]) (defn navigation [] [:div.header_container [:> NavigationHeader {:leftChildren left :rightChildren right}]]) When I try to use this, I get a React error: React.createElement: type is invalid -- expected a string (for built-in components) or a class/function (for composite components) but got: undefined. You likely forgot to export your component from the file it's defined in, or you might have mixed up default and named imports. I've tried (left) and [:left] too. What am I missing?

based on that error message, I'd suspect NavigationHeader is undefined

because "Left" and "Right" are strings, so shouldn't get that message

I've required it the same way I've required other components from the same library: ["@bbc/igm-navigation-header" :refer [NavigationHeader], but maybe this is set up differently. I'll take a look at the source.

Looks like a normal default export, and no different than the others. It's going to be a typo, probably.

Changing the import to ["@bbc/igm-navigation-header" :as NavigationHeader] (:as rather than :refer) changes the error message to Uncaught Invariant Violation: Objects are not valid as a React child (found: object with keys {ns, name, fqn, _hash, cljs$lang$protocol_mask$partition0$, cljs$lang$protocol_mask$partition1$}). If you meant to render a collection of children, use an array instead.

do you mean import or require? these are different things

import doesn't support :as

I meant require. Sorry, JavaScript habits

with :as you are naming the entire lib, not just one thing in it

Got it. I needed to wrap left and right with (reagent.core/reactify-component ...)

oh, weird

I think it's a bit inconsistent that 'some' returns either true or nil. Is there are reason why it doesn't return false?

Some actually returns the result of the function you pass it or nil

user=> (some #{:a} [:a])
:a
user=>

even? returns true false. (some even? [1 3]) returns nil

that is the 'or nil'

if you need false instead of nil, you probably aren't writing idiomatic clojure code

Inconsistent with predicate functions, most or all of which have a ? at the end of their name? some doesn't, of course, which would have been inconsistent with that function naming convention.

nil works fine in my code, just wondering why this was

There are many Clojure functions that can return nil for which it is an idiom to use them as conditional expressions, with the knowledge that nil is treated the same as false in that context.

Are you wondering about the origin of that practice? Or something more specific to some in particular?

Yeah it makes sense how you put it. So all predicates have true/false then?

I believe all functions in the core Clojure library whose name ends in ? return either true or false, but that naming convention is not enforced anywhere in the language.

Using nil to mean logical false originated with Common Lisp, and one or more flavors of Lisp that influenced Common Lisp before that.

Interesting

So I should say "came into Clojure from the influence of Common Lisp", not "originated with Common Lisp".

true and false exist in Clojure primarily because of Java interop, according to early Rich Hickey talks on Clojure.

cl has nil but no false, scheme has #f but no nil, we get both false and nil and both act as "falsey" values in conditionals

also in cl () is just a sugar for NIL

From statements he made in some of those talks, it sounds like he would not have introduced true and false at all, if he could have figured out a clean way to do Java interop while converting Java false to/from Clojure nil in both directions, without missing any, but I suspect that would not be possible if nil is also used for Java null

If you are interested in those statements in particular, the early talks I am referring to are mostly from 2008, with transcripts and videos linked from here: https://github.com/matthiasn/talk-transcripts/tree/master/Hickey_Rich

imo if a function returns true it should return false, if it returns an object then it should return nil (the non-presence of an object)

which afaik is how most clojure functions behave

With the sometimes-nagging detail that true and false are objects.

And if you get deeply involved in Java interop, there are weirdnesses in both Java (and thus also in Clojure) involving freshly constructed Java Boolean false values treated as logical true in both languages. There be dragons, thankfully which rarely rear their ugly heads: <https://clojuredocs.org/clojure.core/if>

also java serialisation/deserialisation of booleans can yield some pretty surprising weirdness

Default serialization does the right thing

It’s usually people trying to be clever that do the wrong thing

yeah I guess it wouldn't have been java serialization. it would've been something weird used by mapreduce or apache beam or something like that

"clever" stuff