8.4 KiB
Getting started with Clojure some of the basics.
Language syntax
The clojure syntax is super simple it is always a function call followed by arguments, this applies for things like standard conditionals and, or, not etc
All functions return a value, this is always the last statement called inside the function.
When ever you see an opening bracket the value after is always a function unless the bracket is preceded by a character in which case this is a macro a common macro is '(1 2) which is the same as typing (list 1 2) so a function is still the first parameter.
Basic data structures
In clojure you will find you spend most of your time dealing with lists, vectors, maps and sets each has its own bracket notation in the language you can also call vector, list, hash-map and hash-set functions to the same effect, how ever it is far more convenient to use the shorthand.
(ns demo.core)
;; lists are denoted with brackets, but remember to use '(
;; or (list function) else the first value is a function call.
'(1 2 3)
;; vectors are donoted with square brackets
[1 2 3 4]
;; hash-maps are denoted with curly braces
;; they must always contain key value pairs.
{:key-one 1 :key-two 2}
;; sets also use curly braces but start with a #
;; sets have to be unique so duplicates will throw an error
#{1 2 3}
;;#{1 2 1 3} how ever would through an error.Maths
All maths operators are also functions, meaning you start with the operation then the numbers to apply the operator against, opposed to many other languages where you would separate the numbers by operators.
This goes back to function first then arguments.
;; This is equivalend to 1 + 2 + 3 in other languages
(+ 1 2 3)
;; When using multiple operators just nest them.
(* 2 (+ 1 2 3))Variables
Variables in clojure are defined with def function, how ever unlike other languages you can not change these unless using some construct which allows this like an atom.
;; These define variable Which can not change
(def my-string "A string once set does not change")
(def my-number 3.14)
;; If you want to change your variables you need to define your variable as an atom
;; Where possible you want to avoid this, the language is immutable by design to help avoid bugs
;; any variable that can be arbitarily changed like atom's means any piece of code can change the value
;; and potentially cause issues when used else where in functions that did not expect the new value
(def my-atom-string (atom "A string once set does not change"))
(def my-atom-number (atom 3.14))
(def my-atom-hashmap (atom {:one 1 :two 2}))
;;One important thig to remember is that to access the values of an atom you need to prepend an @
;;when using functions that change an atom you do not need the @
;;@ is shorthand for the deref function
@my-atom-string
@my-atom-number
@my-atom-hashmap
;; Atoms are changed by calling functions which change the value in a thread safe fashion
;; reset! is to replace the value, swap! is used to update a value and takes a function to apply the update
(reset! my-atom-string "My new string")
;; Append to the existing string
(swap! my-atom-string str " appended text")
;; replace with empty hash map
(reset! my-atom-hashmap {})
;; add new key and value using assoc function
(swap! my-atom-hashmap assoc :key-one "value")
;; There is also defonce which is handy for hot reloading
;; when your code is reloaded all variables will be reset to the inital states
;; with defonce the values are maintained over reloads which helps when testing user flows
;; by maintining the state the user does not need to start again
(defonce my-atom-hashmap (atom {:ex-one "hi"}))
(swap! my-atom-hashmap assoc :key-one "value")Conditionals
Similar rules apply to if conditions if, and, or and not are also functions, it's also worth noting that all these function return values in other languages like python you would update a variable and the if would not directly return a value, try evaluating these to get a feel for this.
When needing multiple conditions look into, cond condp or case
;; OR AND NOT can be used by there own as function calls
(or nil 1)
(and nil 1)
(not true)
;; These can then be mixed with the if condition function like so
;; its important to note that the first parameter is the condition to check
;; then the code to call followed by the else function, you can only call a single function
(if (true? (not true))
1
2)
;; If you don't need the else your better of using when
(when (= 1 1)
(prn "True so print me!")
(prn "multiple statements aloowed in a when, only last returns a value"))
;; If you need to call multiple function when only one is allowed use the do function.
(if (true? (not true))
(do
(prn "first thing todo")
(prn "second thing todo, last statement will be the return")
1)
2)Looping
For the most part you will use map filter and reduce which apply a function for each item in a sequence of values, for the most part this is enough.
You can also use loop which takes your initial values as params inside the [] block and you call recur supplying the updated values in each iteration, not calling recur will end the loop.
;; set n to 0 initially, increment in each loop while n is less than 5 return n value when this is no longer the case
(loop [n 0]
(if (< n 5)
(recur (inc n)) n))
For loops are also available in a similar fashion to loop
;; loop over the sequence of numbers, assigning each to X until the end
;; run any functions on the value in the rest of the body
(for [x [1 2 3 4 5 6]]
(* x x))Hashmap's keyword's & De-structuring
One of the fundamentals to working with hashmap's in clojure is that keywords are function when ever you see :my-key you are actually calling a function this is very different to most other languages.
This has some really nice side effects, one being it is very easy to navigate your hasahmap's the example below shows how to pull out the value 2.
Basic value fetching
(def my-hashmap {:top {:first-key 1 :second-key 2}})
;; you can get the value in a number of ways.
;; using neted get return the result of one get to the next
(get (get my-hashmap :top) :second-key)
;; much nicer is to use get-in and specify the path
(get-in my-hashmap [:top :second-key])
;; You can also call the keywords as a function
(:second-key (:top my-hashmap))
;; or using something called a threading macro
;; push the map through the :top function then the result
;; into the :second-key function
(-> my-hashmap :top :second-key)
You can use de-structuring in clojure to explode out keys when passed to a function.
Using :or we can set default values if the key is missing, the :as keyword can be used to access the unstructured map.
(defn my-fn [{:keys [id name value missing] :or {missing "not set"} :as my-hashmap}]
[id name value missing my-hashmap])
(my-fn {:id 1 :name "bob" :value 456})Wierd symbol's
There is a good reference on the symbols in clojure below when you encounter one your not sure about. https://clojure.org/guides/weird_characters
;; : Colon indicates a keyword's
;; :: Double colon makes a namespaced keyword, what ever is defined at the top of your file under (ns)
;; will be pre pended to the keyword
;; , comma is just white space and is used only for readability to the user
;; #( is an annoymous function (fn [param1] (prn param1)) is equivalent to #(prn %) % being param1
;; you can also use %1 %2 %3 etc to refernce other params, longer form is prefered because the params are named
;; but for very short small functions this variant can be handy
;; -> is the threading macro, basically the result of each statement is passed to the next
;; as the first parameter, this works nicely with hashmaps because keywords are functions
;; ->> same as above but the result is the last parameter
;; when working with sequences you tend to use this one most functions Take
;; a sequence as the last parameter
;; '( this is the same as writting (list 1 2 3) the ' denotes we are using the list function
;; #_ this is the comment block it