map on a hash map will process it as a sequence of key/value pairs. You can call key on a key/value pair to get the key and val to get the value. So the mapping function you want is essentially (fn [pair] (exp (key pair) (val pair)))

But you can also use destructuring in the function arguments to extract the key and value: (fn [[k v]] (exp k v))

Or you can use an anonymous function without destructuring: #(exp (key %) (val %)) Whichever you find easiest to read.

However, since you want to process all those values down to a single result, you really want a reduce -- and for reducing a hash map, there's reduce-kv which automatically extracts the key and value for each pair.

So you could write

(reduce-kv (fn [acc k v] (+ acc (exp k v))) 0 data)

Amazing, thank you! Will play around with each option 🙂

Changed to:

(reduce-kv (fn [acc k v] (* acc (exp k v))) 1 data)

Oops! That's what I get for not trying stuff in the REPL when I'm distracted watching TV 🙂 Glad you got there!

BTW, 2^2 = 4 and 4 * 5 = 20, not 10 so I think you mean 20 would be represented by {2 2 5 1} and 10 would be represented by {2 1 5 1}?

Yes yes yes sorry

PDF viewing it not specific clojurescript feature. You can include any js library to you cjs project an use it. For start i recommend for you use reagent. It is full feature library but more easy than others. https://reagent-project.github.io/

An example of such a library would be PDF.js correct? If so I'll begin by learning how to include that in a reagent project

Sorry i don't have idea about which lib use, recommend you select any with active support(commits). About include. You must include lib not in reagent. Any lib(reagent, pdf and other) include in clojureacript dependencies and after include you can call it in you project

Understood. I'm still learning about how the project structure works for Clojure and ClojureScript.

But js lib you can include in you html in tag script. And call it directly like in js. Maybe it is more easy solution for you.

https://gist.github.com/joelturnbull/1624148 https://squirrel.pl/blog/2012/10/16/hello-clojurescript-with-jquery/

For js libs which not have clj wrapper i simple called their by js-invoke https://cljs.github.io/api/cljs.core/js-invoke

Thank you for the links, I have so much to learn 😄

To be honest, this is a pretty difficult project. PDFs are very hard to work with and manipulate no matter what the language you use is, it's just inherent to PDFs.

Cors it is common http feature. For it correct working, you must return proper cors headers in response to browser. It should look like {:body "" :headers {"Access-Control-Allow-Origin" "*" "Access-Control-Allow-Methods" "*" "Access-Control-Allow-Headers" "*"}}

https://developer.mozilla.org/ru/docs/Web/HTTP/CORS

My example above allow for browsers send request from any domains to you domain

You can use Ring wrap-cors middleware. See: https://github.com/prestancedesign/todo-backend-reitit/blob/master/src/todo_backend/core.clj#L61

Not tested but you could do something like this:

user=> (take x (drop y (clojure.string/split-lines (slurp file))))

(subvec (-> (slurp "...")
            (s/split-lines))
        start end)

that one's better though 😅

Will that be efficient for large files?

I don't think so. I'm guessing it's more efficient to just call out to sed?

this seems like a viable option instead of slurp https://clojuredocs.org/clojure.core/line-seq

Though also I'm guessing the Java file reading utilities are pretty fast and efficient, but that is speculation and nothing more

Since line-seq returns a lazy sequence, does (take x (drop y ...) happen in constant time?

That's above my pay grade unfortunately. But they both return LazySeq

Okay, thanks guys. I'll try it out and compare against sed.

@UPWHQK562 a very ad-hoc experiment on my part but this example offered in the clojure docs is an order of magnitude faster than the code snippet I initially gave

(with-open [rdr ( "/tmp/foo.txt")]
    (reduce conj [] (line-seq rdr)))

you can modify this to grab which lines you want

Thanks @U03QBKTVA0N. There's a nice bit on that approach in this article https://blog.michielborkent.nl/transducing-text.html, I just wasn't sure if skipping lines would be an efficient operation.

There is no efficient way to seek to a random line unless you know the byte offset of that line

@U050ECB92 do you know if doing it through Clojure is less efficient than calling sed?

Is your program doing more work than this step?

sed also has that same limitation, it finds lines by consuming from the beginning

(The limitation is in the problem, not the tool)

> Is your program doing more work than this step? Not sure what you mean exactly. > (The limitation is in the problem, not the tool) Okay, so unless you know the byte index of the starting line, there isn't a way to skip the correct amount of bytes by providing a line number, because each line is an unknown amount of bytes. That makes sense now.

@U03QBKTVA0N

(vec (line-seq rdr))

(keep in mind that making a vector will read all lines into memory)

You can't read only specific line in file. This is not depend on language. Because at first you need find end of line \n or \r\n symbols. It is possible only by sequentially read file bytes. For large file it is efficient use lazy api, it read from file only when you need access to particular part on file. https://stackoverflow.com/a/25953671

Isn't it just a function returning a function?

you can have a look at how it's done here: https://github.com/arttuka/reagent-material-ui also: don't forget that #js does not work for nested objects, use clj->js or repeat #js (e.g.: #js {:a #js {:b 1}}

If you call js directly from clj by js-invoke, you must pass to js function js object, not clj object. Maybe it is you mistake. Below info about convert clj obj to js obj. https://cljs.github.io/api/cljs.core/clj-GTjs

It's also happening in the CI/CD environment

do you have the repo public anywhere?

and some background: uberjars need a main class declared and they run the -main function defined in that class. Do you know the main entrypoint you have declared for your uberjar?

https://gitlab.com/technomancy/leiningen/blob/master/sample.project.clj#L226

Yes, https://github.com/davisuga/gremlin-of-cypher

Got, will try it

What should exactly be in the main function? The https://github.com/weavejester/compojure/wiki/Getting-Started doesn't tell about it

Manifest-Version: 1.0
Created-By: Leiningen 2.9.5
Built-By: dan
Build-Jdk: 11.0.11
Leiningen-Project-ArtifactId: hello-world
Leiningen-Project-GroupId: hello-world
Leiningen-Project-Version: 0.1.0-SNAPSHOT
Main-Class: clojure.main

I see you have a lein-ring plugin but I don't see you starting a webserver in your code. As such your uberjar defines how to handle web requests but doesn't seem to hook the handler up to a running webserver

What is the recommended way to start the server?

It's not clear in the doc 😅

https://github.com/ring-clojure/ring/wiki

Oh, I was looking the wrong doc lol

Thanks

👍

I thought Neptune supported Cypher natively now

Oh, I didn't know that. I just needed CosmosDB, which doesn't support, but I added Neptune just because its easy

Cool project!