@caleb.macdonaldblack this doesn't really answer your question, but you should check out https://github.com/nathanmarz/specter

It was built specifically to solve querying and transforming of deeply nested data structures

i bumped into a hiccup with file i/o and laziness and was hoping someone could point me in the right direction: I’m using a reader to (.readLines reader) into a lazy-seq, something like

(let [rdr (io/reader (io/resource path))]
    (letfn [(read-next []
              (lazy-seq
                (if-let [line (.readLine rdr)]
                  (cons line (read-next))
                  (do (.close rdr) nil))))]
      (read-next))))

(-> line
      (clojure.string/trim)
      ;; clean up leading/trailing space around newlines
      (clojure.string/replace #"\s*\n+\s*" " ")
      ;; use consistent \space char
      (clojure.string/replace #"\s" " ")
      ;; split into codepoints (chars as strings)
      (clojure.string/split #"")))

@caleb.macdonaldblack AFAIK you still can't use clojure.walk/postwalk because of the bug @seancorfield filed: https://dev.clojure.org/jira/browse/CLJ-2031

@lwhorton Your code seems to work fine for me...

Does anyone know a function xxx which does this?

user=> (xxx [a '(1 2 3) b '(2 3 4)] {:a a :b b})
;; '({:a 1 :b 2}
;;   {:a 2 :b 3}
;;   {:a 3 :b 4})

it’s a list comprehension. for

Thank you @gonewest818!

user=> (for [a [1 2 3] b [2 3 4]] {:a a :b b})
;; ({:a 1, :b 2} {:a 1, :b 3} {:a 1, :b 4} {:a 2, :b 2} {:a 2, :b 3} {:a 2, :b 4} {:a 3, :b 2} {:a 3, :b 3} {:a 3, :b 4})

@blackawa If you supply multiple collections to map it does what you want:

(map (fn [a b] {:a a :b b}) [1 2 3] [2 3 4])
=> ({:a 1, :b 2} {:a 2, :b 3} {:a 3, :b 4})

@curlyfry Thanks! This is what I want!

with the risk of starting a bar fight, would anybody care to enumerate some of the upsides/tradeoffs when choosing between component, mount, and integrant (or others)?

from my very subjective view I think Integrant feels more “functional” and “simple”.

@mbjarland see https://clojurians-log.clojureverse.org/clojure/2017-07-21.html#inst-2017-07-21T22:32:57.901156Z

@mbjarland My poorly researched and also very subjective view is that: - Component is nicer than Mount in that it is more easily composable and testable + doesn't overload the namespace construct. - Integrant has similar upsides to Component, but is more data-oriented, and imo simpler + easier to understand.

I originally started using Component, but switched to mount mainly since component requires more tie-in and mount feels less like a framework and more like a tool for managing state

I'd say either one is fine, I guess I just prefer the simplicity of mount over component.

Any thoughts on how to identify the the point where it starts to make sense to consider either?

Hard to say... I found that mount had a much leaner barrier of entry, and it made more sense to me. It leverages it's dependency management through clojure namespaces, so it feels closer to home

Component on the other hand wrote it's own dependency management scheme, and uses a complex map to represent it's state

I'm not familiar with integrant

Component on the other hand represents it's lifecycle in a much cleaner way, the main differences are that Component leverages protocols to represent state. If you typically use Records and Protocols to represent your data, you'd love Component

however, if you are more interested in a data-oriented approach with the base datatypes, go with mount

again, I don't know anything about integrant, I believe it came later, so maybe it has some advantages over the other two

/ listening in - this is exactly the kind of discussion I was hoping for. Very useful /

Glad I could help 🙂

just watched the presentation on integrant (https://skillsmatter.com/skillscasts/9820-enter-integrant-a-micro-framework-for-data-driven-architecture-with-james-reeves). Have to say it looks pretty promising

Is there a standard way to say whether the java version is 9 or 8 or 7. Closest I can think of is (Integer/parseInt (first (clojure.string/split (System/getProperty "java.version") #"\."))) . I want to run some tests only on JDK 8 and others on JDK 9 . In python's pytest we can mark the tests to be run only for certain versions of python. Do I need to roll this out on my own ? I am sure there will be some way since this is a common problem.

@alexmiller not sure where/how to report this but search at http://clojure.org seams broken - https://clojure.org/search?q=destructuring returns no results

Hi. I am working on a clojure library that's now compatible with the most popular (python) implementation of a standard. However I would also like to add the option to enable compatibility with the RFC. The implementation difference is trivial, however I am not sure how to expose the option to switch between both. Would you: * use a dynamic var that can be set to :python-compatible or :rfc-compatible? * pass the option explictely to the library's functions? * duplicate the functions (there are only 2), creating python- and rfc-compatible versions? maybe in separate namespaces?

Also should I default to one of the two, or throw if none is selected?

IMHO, option 3. Consumers can then pick the namespace they need for the behavior they want.

(Or bring in both if they need both for some reason.)

But if you already have something like keywords args or an option map, maybe option 2 (b/c in that case, it’s just another key/value pair)

Thanks, I was also leaning towards 3. The only problem I see is that any dependant libraries would also have to create 2 namespaces if they want to keep the flexibility. And, no, I don't have an options map yet, I just pass data to the functions.

Some of it will come down to a judgement about how important you think the ability to change this option is. The namespace approach implies a bit of a willingness to commit more strongly to one or the other of the two option values at a given call site…

(Stronger still, you could do something like two modules, and clients would pick the dependency they prefer.)

The only other thing I’d suggest is that you do what you can to make information on this setting obvious… (docs, good error messages if you can, etc.)

Makes sense. And I agree about the documentation. Probably more important than the decision itself. Thanks!

does anyone use http-kit? if so have you ever had an issue with {:as :stream} seemingly leaving file handles open in /proc/?

are you closing the streams?

@bhauman sorry for calling you out of lurk, but I saw that you were active in another room so might as well give it a shot 🙂 I’m really eager to try rebel-readline but I can’t figure out how to get it into and make it replace the default REPL in an existing Leiningen project. any help to get started or pointers in the right direction would be much appreciated 🙏

@schmee Add it to your lein dependencies, lein trampoline run -m rebel-readline.main

I'm going to add that info to the readme shortly

question about destructuring, it seems they are not lazy.

but this version didn't work, it is not lazy~

no

seems like the problem is in [p & lst], seems & is not lazy

it is lazy

your types don't match

every time you recursively call sieve you are adding another layer of seq nesting to lst

(for sieve-d)

my guess is you meant [[p & lst]] there

destructuring simply desugars to first/next:

(macroexpand-1 '(let [[a & b] [1 2 3]]))
(let* [vec__156 [1 2 3] seq__157 (clojure.core/seq vec__156) first__158 (clojure.core/first seq__157) seq__157 (clojure.core/next seq__157) a first__158 b seq__157])

@bhauman well, that was easy, thanks! 😄 is there a way to make vi mode the default? pressing ctrl+x,ctrl+v works but as soon as I press enter it reverts back to the regular bindings

so it cannot possibly make a difference

yes!! rookie mistake~

thanks guys~!!

I would recommend taking a closer look at the error messages you are getting

@schmee in ~/.clojure/rebel_readline.edn {:key-map :viins}

I will next time~ though what I observed was, it started to print an infinite list...

probably going to be the two most common questions

@schmee I edited the last one to remove the '.'

haha yeah, answering these two questions early in the readme will probably save you many, many hours 😅

can you set print-length by default?

@alexmiller not in the config

but with clojure.main -i opts

I mean can you just default it to something

you mean as the library author can I supply a default?

I guess that’s really not in the “read” area

its not

sorry, I’m probably not enough in the loop to be saying sensible things :) so I will stop

in fact the default main is just there for trying it out

@bhauman well, now I’ve been using it for almost a full minute, and all I can say is ohmygodthisthingisamazing

this is going to make Clojure coding so much more enjoyable for me! 😄

@alexmiller if you ever have any questions please let me know, but the rebel-readline.main can only really be launched from clojure.main and, it should have all the default root bindings etc.

also the rebel-readline.main right now is more of an example of what can be created with rebel-readline

I could make if a proxy for clojure.main, but at the moment it's not so simple to do that

@bhauman just actually trying it for the first time and this is amazing stuff

well thanks 🙂

just trying to bring Clojure to the people

I really would love to see if we could include this by default with clj

@alexmiller you could just add it add as dep and call a different main

how do you intend to distribute documentation? in the readme? github wiki?

readme to start

I want to start customizing the keybindings, and since I’ve already done some lein-related stuff as well it would be nice to have a place to write it down so other can benefit

I think it would be ideal if we bundled and included it in the clojure tools uberjar, would need to think about the main question though. the clj script can do whatever.

@schmee its not a magical land of customize the bindings how you see fit... Jline has some hard dependencies on some of the keybindings behavior

for non core bindings though it should work great

@schmee for keybinding stuff lets' put it in the Wiki 🙂

yeah, I’m thinking first and foremost of the inline source and doc commands

my vim-attuned brain can’t handle all the ctrls 😂

why would you need anything but emacs?

oh yeah! that's a good idea

@alexmiller my wrists beg to differ

Is there a way to return an input stream from this function instead of writing it into a file? I cant separate the logic to write into a file in this function.

(defn encrypt-bytes
  [bytes dest ^PGPPublicKey pub-key]
  (let [enc   (doto (BcPGPDataEncryptorBuilder. PGPEncryptedData/AES_256)
                (.setWithIntegrityPacket true)
                (.setSecureRandom (SecureRandom.)))
        gen   (doto (PGPEncryptedDataGenerator. enc)
                (.addMethod (BcPublicKeyKeyEncryptionMethodGenerator. pub-key)))]
    (with-open [out (.open gen (io/output-stream dest) (long (alength bytes)))]
      (.write out bytes))))

I would return a byte array

@schmee @noisesmith if there are better vim bindings for the clojure-behavior let me know in an issue we can discuss

like a baos

@schmee also don't change the indent binding 🙂

@bhauman if you add some lein-stuff to the readme, make sure you mention how to put it into your ~/.lein/profiles.clj so that you can have the awesomeness available at all times 👍

🙂

@bhauman is it possible to change the key binding via the config file? I’ve been looking around in the source a bit but I can’t seem to find anything

yeah this isn't possible yet you would have to construct your own repl

its possible to create a line-reader and then change the bindings and then use that for your repl, but probably better to ask me to change the vim bindings

roger, having more “vim-y” binding for the stuff in bind-clojure-widgets would be neat, I’ll think about it a bit and let you know if I stumble on something that feels natural

on a different note, I just had a baller idea: can I make a nodisassemble widget thing and plug that in somehow?

@schmee I'd start by experimenting with the core api, and then work your way up from there 😉

@bhauman haha, I just spent 30 minutes trying to lein install the repo locally, and I just realized that between I discovered the repo an hour ago and I cloned it the artifact name changed! 😂

well, at least I figured it out 😄

sorry about that

no worries, pre-release software is called pre-release for a reason 🙂

Has anyone experience with firebase-admin sdk? My aleph service crash when added.

the syntax highlighting just works and everything, holy crap

@schmee oh send a PR for that!

comin’ right up 👍

that wasn't too hard was it?

but I'm surprised you did that so fast ...

rebel-readline/project.clj                                |  5 +++--
 rebel-readline/src/rebel_readline/clojure/line_reader.clj | 24 +++++++++++++++++++++---
 rebel-readline/src/rebel_readline/core.clj                |  2 +-
 3 files changed, 25 insertions(+), 6 deletions(-)

exciting work!

Does this do syntax highlighting on the output? Haven’t played with it yet, but would love to write the whidbey-equivalent integration for puget

@bhauman I can only commend you for a very well-written and easy to understand library 👏

@schmee that diff isn't expanding for me

or is that just partial

I get it you were just giving me the stats