After reading, it looks like it's a matter of setting a :ring-handler option. I'm going to proceed under the assumption that I have no idea how to do something so complicated that it's not a "simple" ring server anymore...

Why here a class exception :

(fn [s] (reduce (fn[acc n] (if (not= n (last acc)) (conj n acc))) [(first s)] s )) 

@roelofw Do you want to just know the mistake, or learn how to find it?

Last one. Learn how to find it. @rauh

Otherwise I have to ask every time again

Do you have access to the full stacktrace?

(pst) on the repl

I saw this error message on the 4clojure editor but I can copy it on a repl

moment

You may want to invest into getting an editor setup, Cursive is super simple and can print the last stacktrace very nicely

I have cursive

It has the command "Print last exception", which you should def. bind to something if it's not already

Wierd, the code

(fn [s] (reduce (fn[acc n] (if (not= n (last acc)) (conj n acc))) [(first s)] s ))
=> #object[user$eval1273$fn__1274 0x16b6478 "user$eval1273$fn__1274@16b6478"]  

works fine in repl

It's just a function definition, nothing gets evaluated

Run it with example data, you can wrap it in 2 more () and then pass it some vector

oke, now I see the same error message

print the stacktrace with Cursive

and with pst I see this error message :

ClassCastException java.lang.Long cannot be cast to clojure.lang.IPersistentCollection
	clojure.core/conj--4345 (core.clj:82)
	clojure.core/conj--4345 (core.clj:82)
	user/eval1525/fn--1526/fn--1527 (form-init7761432892955560103.clj:1)
	clojure.lang.PersistentVector.reduce (PersistentVector.java:341)
	clojure.core/reduce (core.clj:6544)
	clojure.core/reduce (core.clj:6527)
	user/eval1525/fn--1526 (form-init7761432892955560103.clj:1)
	user/eval1525 (form-init7761432892955560103.clj:1)
	user/eval1525 (form-init7761432892955560103.clj:1)
	clojure.lang.Compiler.eval (Compiler.java:6927)
	clojure.lang.Compiler.eval (Compiler.java:6890)
	clojure.core/eval (core.clj:3105)
=> nil

so it looks that my conj is not good

Yup 🙂

back to the manual why ?

So it expects a collection, but it was given a Long for one of it's arguments

just take a good look at the conj, and look up its documentation

I did , I mixed up acc and n

now I can work on making the function work

thanks

I have to compress ( 1 2 2 3 3 2 2) to ( 1 2 3 2)

I thought when I did [ (first acc) ] I see a vector but it looks not. conj places the new entries on the front and not in the end

anyone a tip why this is not working as expected

Try the Cursive debugger

Can I look step for step in a reduce function ?

When I do debug I see no output at all

@roelofw If conj adds to the front, then your collection is a list. If conj adds to the end, then your collection is a vector.

yep, I found it, I have forget to add a else statement

When I change that everything works fine

except for strings then I see this as output : '[\L \e \r \o \y] ` instead of "Leroy"

So I have to think how to make some changes here :

((fn [s] (reduce (fn[acc n]
                   (if (not= n (last acc)) (conj acc n) acc))
                 [(first s)] s )) 

I could use a if then for strings but then I see a lot of overhead

and I cannot use two anymous functions

some one a tip how to solve this without doing the reduce two times ?

Your problems is NOT "how to solve this problem for strings", you already solved it. Your problem is "how to turn [\a \b \c] into "abc"" in case the input is a string

You can say it also that way

@roelofw: he is giving you the tip you asked for

I know that I can use apply str to do this but then I have the reduce part two times

and I wonder if I can do it with one reduce function

No you don't need reduce twice

(let [res (reduce ...)] (if (string? input) ... ...), also cond->> if you wanna get fancy

Why would you need to do reduce twice?

I was thinking about this :

(if (string? input) (apply str( reduce ..... )) (reduce ......) 

I thougt I could use let only with static things

learned another thing

@seancorfield I do not want to get fancy. I still have to study how --> works

the same with partial instead of fn[x] ( .......)

:thumbsup::skin-tone-2: always learning!

(cond->> (reduce ...) (string? input) (apply str)) is rather nice @rauh

Im still a beginner so I have to learn a lot

@seancorfield that means that the reduce is executed and the apply str only if the input is a string ?

->> means that every part is executed after each other ?

@seancorfield if I do this :

(fn [s] (cond ->> (reduce (fn[acc n] (if (not= n (last acc)) (conj acc n) acc)) [(first s)] s ) (string? s) (apply str))) 

I see this error message : java.lang.RuntimeException: Can't take value of a macro: #'clojure.core/->>, compiling:(NO_SOURCE_PATH:0) on 4 clojure

@roelofw you have a space between cond and ->>

Then I see this error : java.lang.RuntimeException: Unable to resolve symbol: cond->> in this context, compiling:(NO_SOURCE_PATH:0)

4clojure uses clojure 1.4 for some reason and cond->> is available since 1.5

oke, then I use a let binding

@moxaj thanks

solved another one

and see now I could have done it more easy with partition with

Now the next one : calculating factorials

I think I use reduce another time

hopefully some of you fellow clojurians can help me out

i have a map that has an event keyword as key and a fn as a value

in some of these pairs the i get this result :user.handler/confirm-invite #object[clojure.lang.Var$Unbound 0x31e50ab5 Unbound: #'users/confirm-invitation-handler]

other pairs in the same map, coming from the same namespace do not have these issues

I don't think we can answer that without knowing a lot more about how that map is built. Somehow you have an unbound var rather than a function

Hey, how would you advice to write tests with clojure.spec? Do I need to wrap it with some other test framework like test.check or midje? How do you do that, I use spec in repl, but I’d like those verifications to be run from CI as well...

I have to re-implement interleave . Someone who has a tip how I can do this. IM thinking about a loop with a acc

@alexmiller

(ns api.users ..)  (defn invite-user-handler []
  ...)

(def event-handlers {:frontend.users.profile.handler/user-profile                     get-user-handler
                     :frontend.users.list.handler/process-user-list                   get-users-handler
                     :frontend.users.list.handler/invite-user                         invite-user-handler})

(defn start []
  (doseq [[event handler] event-handlers]
    (println "event - handler" event handler)
    (ws/register event handler))) 

@alexmiller all in the same namespace

the println for the last event handler produces

:frontend.users.list.handler/invite-user #object[clojure.lang.Var$Unbound 0x31e50ab5 Unbound: #'api.users/invite-user-handler] 

i don't see confirm-invitation-handler in your code at all

sry, wrong copy paste

i get that output for more than one handler

but others are not unbound

same namespace, fn definetly exist

stacktrace:

:cause Attempting to call unbound fn: #'api.users/invite-user-handler
 :via
 [{:type java.lang.IllegalStateException
   :message Attempting to call unbound fn: #'api.users/invite-user-handler
   :at [clojure.lang.Var$Unbound throwArity Var.java 43]}]
 :trace
 [[clojure.lang.Var$Unbound throwArity Var.java 43]
  [clojure.lang.AFn invoke AFn.java 32]
  [ 

is there any importance that above you have user.handler and here you have api.users?

are you conflating different user namespaces?

no

today i don't seem to get naming right 😉 i just wanted to shorten the namespaces for readability and screwede that up

just assume it's (ns api.users)

was this working earlier and you are shortening the namespace?

what does that mean

i shortended in the snippet, not in actual code 😉

oh i see sorry

no biggie, my error, i edited in the messages

resolve invite-user-handler from your repl?

same thing

api.users=> (invite-user-handler ....)
IllegalStateException Attempting to call unbound fn: #'api.users/invite-user-handler  clojure.lang.Var$Unbound.throwArity (Var.java:43) 

another handler from the same namespace

api.users=> (get-users-handler "")
2016-11-23 15:11:46,877 WARN [....] 

ok ok, found the error... wrong parens -.-'

When I have to re-implement interleave , do I need to use a anymous function ?

I know no other way I can hold the outcome

How do I get the first item of s1 and the first item of s2 in the anymous function. Here my code so far : fn [s1 s2] (reduce [acc n ] (conj acc n1 ..) [] s1 s2 ) ?

your reduce form doesn't seem to have a function form inside of it

and it looks like you have the signature of reduce off

sorry , (reduce (fn [acc n1] (conj acc n2 ) [] s1)

what is n2?

n2 must be n1

so you are just appending to a vector?

you'll just end of with (vector s1)

n1 schould be a number of the first seq and n2 schould be the first number of the second seq which I want to interleave

I have to interleave 2 seq without using interleave

and I try some things out how to solve that

can you reduce over two collections?

I do not think so

so your fundamental approach is off here then

so this idea can be trown into the trash can

not necessarily

you could put them into the same "thing" and then work over that single thing

just the point is that the signature of reduce only takes one parameter

so you could put it into a map, but this seems a little clunky

you could put them into a vector as element one and element two

but again, it seems like we are kinda twisting stuff

it seems like what we want to do is "for each" element in each list, combine them

so to me, i hear a map function

yep, that is why I try to find a way this can work

so if we had [a b c] [d e f] as our lists, if we wanted an intermediate step, we could easily end up with ([a d] [b e] [c f])

by just mapping a function over both collections

and if we had that form, is there a way to easily make it look like the final result that we want?

yep

which function?

my first quess schould be interleave and I m not allowed to use it

nope

I will take another look at the cheat scheet

which function does ([a d] [b e] [c f]) => (a d b e c f)

concat ?

flatten

@dpsutton yep flatten or (mapcat identity ....)

o, I was looking at zipmap

well, mapcat kinda does flatten

but i definitely would not have thought of that one

It's the equivalent of flatMap in other langs

so @roelofw if we know how to get from the intermediate step to the final step, do you know how to get to the intermediate step?

([a b c] [d e f]) => ([a d] [b e] [c f])

i.e. map a fn and concat the results

@dpsutton zipmap maybe ?

don't overcomplicate it

but think about it

and its a very common function but used in a way that's not used as much, ie, over two collections rather than over just one

oke, I will think about it

I will leave and think about it

perfect

these are better if you mull it over for a while

BTW Once you can do it in two steps- with the intermediate result ([a d] [b e] [c f]) it's easy to make it one step using something I mentioned above.

and i think your idea of a reduction makes sense, you just need to figure out how to begin. This should be nicer as it only traverses your collection in one pass

If anyone is interested in picking up Clojure Applied, it’s 40% until Dec 2 the Pragmatic Press’s annual Black Friday sale - https://pragprog.com/book/vmclojeco/clojure-applied - use code turkeysale2016

other Clojure books can be had with the same

@dpsutton can you any tips about the function I have to use for my problem

You want to "for each" over two collections

and that's a definite map

map can go over one collection, taking each element at a time

or if passed multiple collections, takes an element at a time from each and passes them in to your function

so if we want to make some vectors so that we get ([a d] [b e] [c f]) we are making vectors

oke, I never seen that map could take multiple collections

I will google for some examples of that

use your repl

and read the docstring

so [f c1 c2] does a function with 2 colls c1 and c2

yup

> Function f should accept number-of-colls arguments

and just make sure your mapping function can accept that many args

oke, then I have to think what f must be

what must f take and what must it output

f must take the 1 element of c1 and one of c2 and the output must be [ element c1, element c2]

cool

exactly

know of any functions like that? or must we build it

we just need to make a vector of our input args

that would be vector

the only thing I have to figure out is how to get the elements

correct

map does that for you

(map vector [1 3 5] [2 4 6])

normally I would do [acc , n] where acc is the output and n the current number

that's a reducing signature

map calls the function you provided with the successive elements of the collection(s)

so here, map is calling vector with (vector 1 2), then (vector 3 4) then (vector 5 6)

vector is varyadic and handles the function call with two arguments

And now flatten it and we did our two step process

There you go

And now that you get that two step process look up mapcat

To do it in "one" step

Hey guys, what’s difference between -> and ->>?

-> is the thread-first macro, where the result is put as the first argument of each subsequent expression

as in

whereas with ->> the result gets put as the LAST argument

as in

the commas are equivalent to whitespace, just to show where the result goes

more info at http://clojure.org/guides/threading_macros

@dpsutton thanks, time for thinking how to flatten a seq without using flatten

the easy challenges are becoming harder now

Well it kinda did anyways

That's why I said you're doing it in "one" step

You're using one form but it's calling concat underneath

That's why I like your reduction form from earlier

@dpsutton you mean the fn[s] reduce (fn [acc n] (conj acc n)) ?

I use that often when I have to calculate or count things and im not allowed to use count for example

If I have to flatten a seq and not allowed to use flatten, the only way is to use recursion with a if then ?

You could probably use a loop form