Skip to content

SWiss ARmy Knife - Your everyday clojure toolbelt!

License

Notifications You must be signed in to change notification settings

verberktstan/swark

 
 

Repository files navigation

Swark

SWiss ARmy Knife - Your everyday clojure toolbelt!

Clojars Project

This library contains functionality that you might need every single day as a happy clojure(script) developer. The aim is to provide composable functions that you can use everyday, but are just not as trivial as (some (comp #{42} :answer) answers).

Most functionality should work in both Clojure and Clojurescript.

Basic usage

Clojure CLI/deps.edn

Add this dependency to the :deps map in deps.edn:

org.clojars.stanv/swark {:mvn/version "0.1.5"}

Require swark.core in your ns form:

(:require [swark.core :as swark])

Then you can use the Swark utility functions:

(swark/key-by :id [{:id 1} {:id 2}]) => {1 {:id 1} 2 {:id 2}}

(swark/map-vals count {:a [:b :c] :d [:e]}) => {:a 2 :d 1}

(swark/jab / 10 0) => nil

Little tour of Swark utilities

swark.core

  • key-by: Returns a map where all items are keyed by the result of calling (f item)
  • map-vals: Returns a map where f is applied to all the input map's values.
  • filter-keys: Returns a map containing only the map-entries whose key returns logical true when supplied to a predicate fn.
  • select-namespaced: Returns a map containing only those map-entries whose key's namespace is equal to the supplied namespace.
  • jab: Try and fail silently, returning nil when any kind of error or exception is thrown.
  • ->str: Returns input coerced to a (trimmed) string. Support (namesapced) keywords etc.
  • unid: Return a unique id string.
  • ->keyword: Returns input coerced to a keyword, replacing whitespace with dashes.
  • invalid-map?: Minimalistic spec checker, returns logical true if the input does not respect the spec-map. Spec map is simply a map with predicates as vals.
  • valid-map?: Complement of invalid-map?
  • memoir: Like memoize, but with flushing. Flush the complete cache, or specific parts.

swark.authom

Atomic authorization. Generate a token for a map, in conjunction with a password and optional secret. And of course check if it matches.

  • with-token: Returns a map item with a hashed token in it's metadata.
  • check: Checks the password (and optional secret) given a map item.
  • disclose: Returns map item with it's token associated with ::authom/token. Useful for serializing the hashed token.
  • conceal: Returns the map item with it's token moved to it's metadata. Useful for parsing a persisted record.

swark.cedric

  • Mem.: Creates a new instance of the in-memory implementation
  • Csv.: Creates a new instance the implementation of the csv backend
  • make-connection: Starts an atomic interface connection for a database.
  • upsert-items: Creates or updates items in the database
  • find-by-entity: Returns one database record found by its entity e.g. [:user/id 123]
  • find-by-primary-key: Returns database records found by its primary key e.g. #{:user/id}
  • read-items: Returns all (filtered) records from the database
  • archive-items: Marks items as archived

swark.atomic

  • atomic: Returns a map with in- and output async channels to provide atomic interactions for side-effecting functionality.
  • put!: Puts an instruction on the atomic's input channel, blocks and returns the response.
  • close!: Closes the atomic's channels and stops the internal go-loop.

Example - Integrate swark.authom & swark.cedric

Note: Cedric's CSV implementation is currently clj only! In cljs you can actually use the in-memory implementation (see swark.cedric/Mem)

Let's say you want to store a user record, some credentials and check their credentials. You can use swark.cedric for the persistence part, and swark.authom for the authentication part.

  1. Let's create/connect to a database via the Csv implementation and store db props related to users.
(ns my.ns
   (:require [swark.authom :as authom]
             [swark.cedric :as cedric]
             [swark.core   :as swark])
   (:import [swark.cedric Csv]))

(def DB (cedric/Csv. "/tmp/db.csv"))
(def PROPS (merge authom/CEDRIC-PROPS {:primary-key :user/id}))
  1. Create a new user record with cedric/upsert-items:
(def USER (-> DB (cedric/upsert-items PROPS [{:user/name "Readme User"}]) first))
  1. Store credentials (generated with authom/with-token) by upserting the user with cedric/upsert-items
(let [user (authom/with-token USER :user/id "pass" "SECRET")]
    (cedric/upsert-items DB PROPS [user]))
  1. Retrieve the user with cedric/find-by-primary-key and check their credentials with authom/check:
(let [user (-> DB (cedric/find-by-primary-key #{:user/id} {:where (comp #{"Readme User"} :user/name)}) first)]
    (-> user (authom/check :user/id "pass" "SECRET") assert))
  1. Since the csv file might change in the mean while, it is advised to execute all db actions as an asynchronous transaction. You can make use of cedric/make-connection like so:
(let [{::cedric/keys [transact! close!]} (-> "/tmp/db.csv" cedric/Csv. cedric/make-connection)]
    (transact! cedric/upsert-items {:primary-key :id} [{:test "data"} {:more "testdata" :something 123}]) ; Returns the upserted items.
    (transact! cedric/read-items   {}) ; Returns all items read.
    (close!)) ; Don't forget to close the async connection.

Tests

Run the tests with clojure -X:test/run

Development

Start a repl simply by running clojure -M:repl/basic command in your terminal. You can connect your editor via nrepl afterwards, e.g. from emacs; cider-connect-clj Or create a repl from your editor, e.g. from emacs; cider-jack-in-clj

Jar creation

Create an uberjar with clj -X:uberjar :jar swark-0.1.5.jar

Local installation

Install Swark locally with clj -X:install

Deploy to Clojars

env CLOJARS_USERNAME=username CLOJARS_PASSWORD=clojars-token clj -X:deploy

License

Swark by Stan Verberkt is marked with CC0 1.0 Universal