I'm getting a page not found error. Are you sure this repo isn't private?

I use for to create html structure (div rather than table) and pass a hash-map of data with the values used in that structure the generator functions in the Practical landing page use this approach If numbering is important it could be added to the data passed into the function and if necessary the data sorted by the number value before used in for. (not suggesting this is best, as best is relative to what best actually means and possibly the significance of the numbers https://github.com/practicalli/practicalli.github.io/blob/live/src/practicalli/books.cljs

I have found map-indexed is super useful with React because React wants each item in a list to have a distinct key. Of course keys representative of their item are most likely to help React avoid work if you make certain kinds of updates to the list, but a sequential counter is a convenient second-best. I.e., [:tr {:key (str i)} ...

thank you guys for your opinion. I would stick with map-indexed

What are the semantics of function f? Indeed, what is the real name of f? And what is the real name of sample-state? I do not think Clojure itself dictates any answer, the API being built decides that. So if f expects a list and knows about how you are encoding that, you can pass sample-state. If f is ... well, if (defn f [state] ...) is even an option, it must know about the data structure. So I would pass that rather than pull it apart and create a point of brittleness should the state structure have to change. my2 🤷

> "It is better to have 100 functions operate on one data structure than to have 10 functions operate on 10 data structures." This quote is from Alan Perlis' Epigrams on Programming (1982). So I usually peek one function form that meets my current needs best. And if I need another later, I can add one more function.

(defn f [head tail])

(defn f* [{:keys [head tail]}]
  (f head tail))

to the point of the question “what is the purpose of the function”, another way to think of it is in term of the architectural layers. a representing pattern being, 3 categories of functions, one that operates on scalar values, the next level up (that may or may not use the previous layer) operating on entities (or rather maps), the final being usually an aggregation point but that operates on sequences. so, the intent of the function, driving the type of parameter list it accepts.

order of parameters is also an important consideration, which can heavily influence the ergonomics of use later when using compositional mechanics like partial, comp, ->, transducers, etc. rarely going to be perfect on the first pass, but important to pay attention and try to get in a decent shape up front. from experience, in the long term in massive code bases, there can be a lot of pain/confusion until these considerations are addressed.

@U0PUGPSFR @U04V4KLKC @UE4SAM0A0 thanks a lot for your answers. I updated the question for more context, but I think I had my answer: in this case, just pass a map since it is going to be simpler to use with reduce

cool, np 👍

My only rule of thumb is to define functions that take a hash-map as an argument unless there is on obvious need not too I like to give my values context regarding their purpose as it makes the code easier to work with A hash-map also give a lot of flexibility as to what information can be received and allows the API of the code to grow with less chance of breaking changes or creating similar functions

clojure -J-Dmy.main.props=test/resources/cfg.properties -M -m my.main.ns

that should work

Ah cool. Thanks alex

it’s in clj --help > clj-opts: > -Jopt Pass opt through in java_opts, ex: -J-Xmx512m (i always go back and reference this so letting you know where you can find it)

bonus points for being pretty-printed

have you tried clojure.pprint/pprint?

usually, if you're writing edn, then you should use some variant of clojure.pprint/pprint , pr, or pr-str.

with the caveat that writing readable edn reliably is tricky and at the very least, want to explicitly set the various bindings that control printing. Here's an example:

(defn write-edn [w obj]
  (binding [*print-length* nil
            *print-level* nil
            *print-dup* false
            *print-meta* false
            *print-readably* true

            ;; namespaced maps not part of edn spec
            *print-namespace-maps* false

            *out* w]
    (pr obj)))

I’ve used pprint in the past, I’m just always thrown by some missing understanding around how side effects and lazy sequences interact

I thought that if I chucked a doall in there, it would make everything evaluate, and only then would the spit occur

but nope

> I thought that if I chucked a doall in there, it would make everything evaluate, and only then would the spit occur That is what is happening, but calling spit calls str on the content. The .toString method for for 's return value is always [email protected] regardless of whether it has been realized.

> (str (doall
        (for [i (range 3)]
          i)))
"[email protected]"

How do I apply your write-edn function to get a plain text file?

(I don’t mean, as distinct from edn…. I’m just trying to get data to paste to someone who isn’t running a Clojure notebook)

something like:

(require '[ :as io])
(with-open [w (io/writer "my-file.edn")]
  (write-edn w my-data))

thank you!!

is the secret using real writers, then?

you can also use spit with pr-str

I only use spit as a quick and dirty option. Usually, using the stuff is a similar amount of code and is easier to tweak to fit the use case.

same goes for slurp.

you can also just put it in a vector using vec instead of doall