(remove #(= 1 (:id %)) my-sequence-of-items)

🙏

How about this? (take 10 (map first (iterate (fn [[a b]] [b (+ a b)]) [1 2]))) You can always destructure inside the iteration fn.

That's similar to what I landed on. It feels a little too much like I'm playing code golf though. For me at least it is a little difficult to see what that is doing at a glance. Maybe the destructuring obscures the b-for-a swap a little bit.

though to be fair, putting this in a (defn fib-iter ...) probably would be a big hint

yeah

thanks

No

no plans to add it at this time

you can make a reader macro that does parse readable number formatting

#num "$1,000.00"

seems better than underscores

This, from Mastering Clojure Macros, also fails.

$ lein try org.clojars.trptcolin/riddley "0.1.7.1"
'try' is not a task. See 'lein help'.

Did you mean this?
         new
         run
         pom
         jar
         do
         test
         vcs

❯ clj -Sdeps '{:deps {riddley/riddley {:mvn/version "0.2.0"}}}'
Clojure 1.11.1
user=> (doto 'riddley.walk require in-ns)
riddley.walk
riddley.walk=> (walk-exprs number? inc '(let [n 1] (+ n 1)))
(let* [n 2] (. clojure.lang.Numbers (add n 2)))
riddley.walk=>

don’t lein install it. Just use the artifact

but if you really want to have a local copy built, change this line to target java 1.8 instead of 1.6 i think

I changed the line to use 1.8 and then

❯ lein uberjar
Compiling 1 source files to /private/tmp/riddley/target/classes
warning: [options] bootstrap class path not set in conjunction with -source 8
1 warning
Created /private/tmp/riddley/target/riddley-0.2.0.jar
Created /private/tmp/riddley/target/riddley-0.2.0-standalone.jar

riddley on  master [!] via ☕ v17.30 on ☁️  metabase-query took 6s
❯ clj -Sdeps '{:deps {riddley/riddley {:local/root "target/riddley-0.2.0-standalone.jar"}}}'
Clojure 1.11.1
user=> (doto 'riddley.walk require in-ns)
riddley.walk
riddley.walk=> (walk-exprs number? inc '(let [n 1] (+ n 1)))
(let* [n 2] (. clojure.lang.Numbers (add n 2)))
riddley.walk=>

i’m able to build and use the resulting jar

@U11BV7MTK Trying to catch up with your idioms. Haven't used clj at the command line, just Cider (Emacs mode).

that’s ok. i’m just using the basic tools without any editor tooling so you can see what is going on

Are you wanting to develop with riddley or just use it in another project?

@U11BV7MTK Develop with it.

All I’m demonstrating • in the first example, you don’t need to build it or install it locally. You can just use it the way you would use any other library: by using its coordinates in maven. Riddley is in https://central.sonatype.dev/artifact/pro.juxt.clojars-mirrors.ztellman/riddley/0.2.0 • if you want to change riddley, I’m demonstrating that i can run lein uberjar which builds the jars. And then use the artifact it created

@U65G9MPCK do you mean make changes to riddley itself? Or just use it in another project?

Just use it.

then just declare a dependency on it with [riddley "0.2.0"] for a leiningen project or riddley/riddley {:mvn/version "0.2.0"} for a clojure deps project

example • leiningen: https://github.com/clj-commons/manifold/blob/master/project.clj#L10

@U11BV7MTK I'd been trying something like that, ...

(defproject template-matcher "0.1"
  :description "FIXME: write description"
  :url ""
  :license {:name "EPL-2.0 OR GPL-2.0-or-later WITH Classpath-exception-2.0"
            :url ""}
  :dependencies [[org.clojure/clojure "1.11.1"]
                 [riddley "0.2.0"]
                 ]
  :repl-options {:init-ns template-matcher.core})

(ns template-matcher.core
  (:require [clojure.string :as str]
            [clojure.pprint :refer [cl-format]]
            [riddley.walk :refer [walk-exprs]]))

Execution error (FileNotFoundException) at template-matcher.core/eval6256$loading (core.clj:1).
Could not locate riddley/walk__init.class, riddley/walk.clj or riddley/walk.cljc on classpath.

@U11BV7MTK The book says this should work, after the lein try thing.

(require '[riddley.walk :as walk])

@U11BV7MTK Cleaning up the install as you suggested didn't help (same error).

I just copied your project.clj from above:

❯ cat project.clj
(defproject template-matcher "0.1"
  :description "FIXME: write description"
  :url ""
  :license {:name "EPL-2.0 OR GPL-2.0-or-later WITH Classpath-exception-2.0"
            :url ""}
  :dependencies [[org.clojure/clojure "1.11.1"]
                 [riddley "0.2.0"]
                 ]
  )

❯ lein repl
nREPL server started on port 58069 on host 127.0.0.1 - 
REPL-y 0.5.1, nREPL 0.8.3
Clojure 1.11.1
OpenJDK 64-Bit Server VM 17.0.1+12-LTS
    Docs: (doc function-name-here)
          (find-doc "part-of-name-here")
  Source: (source function-name-here)
 Javadoc: (javadoc java-object-or-class-here)
    Exit: Control+D or (exit) or (quit)
 Results: Stored in vars *1, *2, *3, an exception in *e

user=> (doto 'riddley.walk require in-ns)
riddley.walk
riddley.walk=> (walk-exprs number? inc '(let [n 1] (+ n 1)))
(let* [n 2] (. clojure.lang.Numbers (add n 2)))
riddley.walk=>

@U11BV7MTK Ok, that works fine for me, too. But I'm still getting the error in my Cider REPL.

template-matcher.core> (doto 'riddley.walk require in-ns)
Execution error (FileNotFoundException) at template-matcher.core/eval6294 (form-init13654649044716525103.clj:324).
Could not locate riddley/walk__init.class, riddley/walk.clj or riddley/walk.cljc on classpath.

Have you restarted your repl?

Many times.

@U11BV7MTK Time to ask about this in #C0617A8PQ?

at the top of the repl you should see some information about “startup form”. Can you paste that here?

@U11BV7MTK

;; Connected to nREPL server - 
;; CIDER 1.5.0 (Strasbourg), nREPL 0.9.0
;; Clojure 1.11.1, Java 17.0.5
;;     Docs: (doc function-name)
;;           (find-doc part-of-name)
;;   Source: (source function-name)
;;  Javadoc: (javadoc java-object-or-class)
;;     Exit: <C-c C-q>
;;  Results: Stored in vars *1, *2, *3, an exception in *e;

there should be the phrase startup form just below this?

@U11BV7MTK I don't see that. Here's the whole thing.

*** Closed on Tue Jan  3 11:53:45 2023 ***
;; Connected to nREPL server - 
;; CIDER 1.5.0 (Strasbourg), nREPL 0.9.0
;; Clojure 1.11.1, Java 17.0.5
;;     Docs: (doc function-name)
;;           (find-doc part-of-name)
;;   Source: (source function-name)
;;  Javadoc: (javadoc java-object-or-class)
;;     Exit: <C-c C-q>
;;  Results: Stored in vars *1, *2, *3, an exception in *e;
;; ======================================================================
;; If you're new to CIDER it is highly recommended to go through its
;; user manual first. Type <M-x cider-view-manual> to view it.
;; In case you're seeing any warnings you should consult the manual's
;; "Troubleshooting" section.
;;
;; Here are a few tips to get you started:
;;
;; * Press <C-h m> to see a list of the keybindings available (this
;;   will work in every Emacs buffer)
;; * Press <,> to quickly invoke some REPL command
;; * Press <C-c C-z> to switch between the REPL and a Clojure file
;; * Press <M-.> to jump to the source of something (e.g. a var, a
;;   Java method)
;; * Press <C-c C-d C-d> to view the documentation for something (e.g.
;;   a var, a Java method)
;; * Print CIDER's refcard and keep it close to your keyboard.
;;
;; CIDER is super customizable - try <M-x customize-group cider> to
;; get a feel for this. If you're thirsty for knowledge you should try
;; <M-x cider-drink-a-sip>.
;;
;; If you think you've encountered a bug (or have some suggestions for
;; improvements) use <M-x cider-report-bug> to report it.
;;
;; Above all else - don't panic! In case of an emergency - procure
;; some (hard) cider and enjoy it responsibly!
;;
;; You can remove this message with the <M-x cider-repl-clear-help-banner> command.
;; You can disable it from appearing on start by setting
;; 'cider-repl-display-help-banner' to nil.
;; ======================================================================
Execution error (FileNotFoundException) at template-matcher.core/eval6289$loading (core.clj:1).
Could not locate riddley/walk__init.class, riddley/walk.clj or riddley/walk.cljc on classpath.
template-matcher.core> (doto 'riddley.walk require in-ns)
Execution error (FileNotFoundException) at template-matcher.core/eval6294 (form-init13654649044716525103.clj:324).
Could not locate riddley/walk__init.class, riddley/walk.clj or riddley/walk.cljc on classpath.
template-matcher.core> 

for instance, I started up CIDER from the project.clj file i copied from you earlier and see this

;; Connected to nREPL server - 
;; CIDER 1.2.0 (Nice), nREPL 0.9.0
;; Clojure 1.11.1, Java 17.0.1
;;     Docs: (doc function-name)
;;           (find-doc part-of-name)
;;   Source: (source function-name)
;;  Javadoc: (javadoc java-object-or-class)
;;     Exit: <C-c C-q>
;;  Results: Stored in vars *1, *2, *3, an exception in *e;
;;  Startup: /Users/dan/bin/lein update-in :dependencies conj \[nrepl/nrepl\ \"0.9.0\"\] -- update-in :plugins conj \[cider/cider-nrepl\ \"0.27.4\"\] -- repl :headless :host localhost
user> (doto 'riddley.walk require in-ns)
riddley.walk
riddley.walk> (walk-exprs number? inc '(let [n 1] (+ n 1)))
(let* [n 2] (. clojure.lang.Numbers (add n 2)))
riddley.walk> 

@U11BV7MTK I see I have a later version of CIDER (1.5.0), same version of nREPL. I also have a slightly newer version of Java (17.0.5).

fair enough.

can you remove the init-ns so that you don’t start up in the namespace of your project?

That way you can get a repl up and running without blowing up

then evaluate (System/getProperty "java.class.path") and paste it here

My suspicion is that you built a bad jar earlier and then installed it so you aren’t getting a correct copy of riddley

Now I get the startup info:

;;  Startup: /usr/local/bin/lein update-in :dependencies conj \[nrepl/nrepl\ \"0.9.0\"\] -- update-in :plugins conj \[cider/cider-nrepl\ \"0.28.5\"\] -- repl :headless :host localhost

what changed that you get startup info now?

When I removed :repl-options.

weird. ok

Acts better, now.

user> (doto 'riddley.walk require in-ns)
riddley.walk
riddley.walk> 

riddley.walk> (System/getProperty "java.class.path")
"/media/sf_home/code/Cogex/template-matcher/test:/media/sf_home/code/Cogex/template-matcher/src:/media/sf_home/code/Cogex/template-matcher/dev-resources:/media/sf_home/code/Cogex/template-matcher/resources:/media/sf_home/code/Cogex/template-matcher/target/classes:/home/bschrag/.m2/repository/cider/cider-nrepl/0.28.5/cider-nrepl-0.28.5.jar:/home/bschrag/.m2/repository/nrepl/nrepl/0.9.0/nrepl-0.9.0.jar:/home/bschrag/.m2/repository/org/clojure/clojure/1.11.1/clojure-1.11.1.jar:/home/bschrag/.m2/repository/org/clojure/spec.alpha/0.3.218/spec.alpha-0.3.218.jar:/home/bschrag/.m2/repository/org/clojure/core.specs.alpha/0.2.62/core.specs.alpha-0.2.62.jar:/home/bschrag/.m2/repository/riddley/riddley/0.2.0/riddley-0.2.0.jar:/home/bschrag/.m2/repository/org/nrepl/incomplete/0.1.0/incomplete-0.1.0.jar"

Sorry, got out of sync with your requests...

no worries

what happens if you evaluate (require 'template-matcher.core)?

riddley.walk> (require 'template-matcher.core)
Syntax error compiling at (template_matcher/core.clj:327:5).
Unable to resolve symbol: walk-exprs in this context

I can see the defn for walk-exprs in walk.clj (in riddley). But I am not a namespace expert. Let me try the book's form.

@U11BV7MTK Better now.

riddley.walk> (require 'template-matcher.core)
nil

what did you change?

[riddley.walk :as walk] --> [riddley.walk :refer [walk-exprs]] I had left walk-exprs unqualified, in source, when trying the former.

glad you are sorted

Thanks, but I'm still a little mystified.

you changed several things at once. We finally got it fixed but you had uses of walk-exprs in your code but didn’t introduce that

now i would restore the repl-options or whatever and see if it works again

I had tried this many times with the :refer form---it was a temporary inconsistency. Restoring :repl-options...

@U11BV7MTK Working fine now with :repl-options. I'd thought recompiling riddley would overcome any former bad info. And maybe it did. I did realize just a few restarts ago that Cider's C-c M-r (restart REPL) doesn't clean things out as well as killing the nREPL buffer and redoing C-c M-j (new REPL). So, maybe both of those things. Anyway, now I seem good to go! Thanks so much!

seems like a good question to me! I think usually people mean #1 - changing the running function to a "fixed" version via a connected repl (and making the same change in the source for future deploys)

so the question then is how to start the repl server in your app. Clojure ships with the ability to act as a socket server by just adding some system properties (see https://clojure.org/reference/repl_and_main#_launching_a_socket_server) and maybe some people do that. Alternately, you can programmatically do the same thing so you might make some way to start a server on demand via any means of interaction you might put into your server

and as far as the "state sharing" - if you connect to the same Clojure runtime, you get that automatically. if you connect via a remote repl to the JVM running Clojure, there is only one namespace, and only one var at a specific name so doing an alter-var-root or equivalent will affect what the app is running (unless you have used direct linking and applied an uber jar, in which case, more drastic efforts may be required)

is any of this a good idea operationally? well....

Brilliant. I work in a tool for my day job which is actually based on Allegro CL and we have a socket REPL there too. It uses a bit of a different architecture and hence my assumption was also #1 because that's exactly what I do for our production issues 😁 . It was not clear to me while reading the docs about how the REPL spawns with the app. I guess I skipped the section on Socket REPLs thinking it was just an alternative for the nREPL. @U064X3EF3 your answer has helped me a lot to relieve me from what was bothering me since a few days. Great talk at reClojure BTW. I just watched it 2 hours back 🙂

I agree its scenario #1 Running Clojure always includes a REPL process, however it may not be configured to accept an external connection, e.g via nREPL or SocketServer. If a hot-fix approach is desirable, then the production app would be run with either an nREPL server or SocketServer and then code can be evaluated either at the REPL prompt or via a connected editor. Obviously changes like this should not be taken lightly and security should be in place so that changes cant be made without authorisation and should be done over a secure connections (probably connecting securely to the server first, then connecting to the REPL) If there are multiple instances of the service, then this hot-fix could cause some very interesting 'challenges'. So having an efficient CI / CD process is more consistent approach

@U023570MX0R I just recently blogged about how we do this with nREPL servers on QA and production: https://corfield.org/blog/2022/12/18/calva-joyride-portal/

@U05254DQM: what is the reason behind having a REPL process included if one never would want to use it? Should it not be more explicit to start the process because launching a shadow unaccessible REPL process sounds like a waste of processing power.

We've run Socket Servers in production for ages -- by specifying a JVM option to start it up when Clojure starts (per what Alex said) but recently decided we wanted the full "development" experience by adding nREPL (and Portal -- so I can visualize production data in a web browser connected to it).

The whole confusion began because of your excellent REPL driven development video I watched @U04V70XH6.

You did mentioned about the running REPL servers there and I was left scratching my head.. But I will go through the blog above

Haha... sorry to confuse 🙂 Hopefully that blog post helps.

We don't patch any production processes that way (we build uberjars with direct linking for deployment) but it's very helpful to debug production issues, especially being able to visualize data via Portal.

(we can easily & automatically deploy updated JARs to production so hot-fixing things isn't worth the risk/complexity at this point)

The REPL server is not a separate process - it is I believe a listening socket that can accept a connection, and if such a socket is created, then there is a thread running a REPL reading input from that socket, and sending output back to it.

That thread runs in the same process that is the JVM process you started to start your application.

> I might sound stupid asking it but I would like to ask it anyway.. "The man who asks a question is a fool for a minute, the man who does not ask is a fool for life." - Confucius.

I’ve been using socket repl for years, occasionally, to debug tricky production issues. It’s been an invaluable tool for me. I very rarely change stuff this way because it’s a bit dangerous and needs to be done on multiple machines

Paging @U051S5XR3 for production REPL stories :)

@U06BE1L6T I have been also doing it but in a different tool called Planisware which has an inbuilt REPL which I am used to. The tool is itself programmed in Allegro Common Lisp but provides a REPL to write code in an internal language called Planisware Script which is closely following ECMAScript which gets internally interpreted/compiled down to CL. It is one of the reasons why I feel at home in Clojure and had very little trouble with understand REPL-driven development 😀

^ here's one story: https://youtu.be/zsTXte3p0Uk?t=339

From the video, if you're @U051S5XR3, then this is what it looks like to be always-connected to eight production REPLs 🤓

I have used socket repls exclusively for the last two years or so. One thing I absolutely love about it is that there is essentially no difference between my “dev” experience with our repo and a repl into a production jar. From my point of view, there’s no difference between clj -A:dev:socket and java "$(socket-repl 6000)" -jar metabase.jar . They are functionally indistinguishable. I have a little shell function called socket-repl that expands out to the full incantation to set up a socket repl with the jvm option.