voger

voger

How to test config options?

I have a library that uses option parser to format the output. The input is a string coming from the code itself, not a shell.

I am using this code to merge the switches given from the user with some predefined switches. The predefined switches are:

  • The default hard coded in the module
  • If are there any coming from config.exs merge them with the hard coded

If the user provides own switches, they will be merged to the predefined switches

defmodule Clouseau.Switches do
  #...
  @predefined_switches [
    file: true,
    full_path: false,
    module: true,
    line: true,
    text: true,
    border: false,
    colors: false
  ]

  #...
  @default_switches Keyword.merge(
                      @predefined_switches,
                      Application.get_env(
                        :clouseau, 
                        :default_switches, 
                        @predefined_switches
                        )
                    )

 #...
  defp apply(switches) do
    Keyword.merge(@default_switches, switches)
  end
end

The @default_switches module attribute is set during compile and Application.put_env(:clouseau, :default_switches, file: false) has no effect.

How can I test that changing the config options changes the behavior of my libray?

So far I came up with this test file.

defmodule SwitchesTest do
  use ClouseauCase

  setup_all  do
    # Save current config if exists
    env_switches = Application.get_env(:clousau, :default_switches)
    # Put new config 
    Application.put_env(:clouseau, :default_switches, file: false)
    # Recompile the file to refresh the @default_switches module attribute
    Kernel.ParallelCompiler.compile(["lib/switches.ex"])

    # On exit revert :default_switches config
    on_exit(fn ->
      # If it existed before put back the old one
      if env_switches do
        Application.put_env(:clouseau, :default_switches, env_switches)
      # if didn't exist before just delete the current one
      else
        Application.delete_env(:clouseau, :default_switches)
      end
      
      # Regenerate the @default_switches module attribute
      Kernel.ParallelCompiler.compile(["lib/switches.ex"])
    end)
  end
end

In the setup_all block I save the current config, recompile the switches module, and on_exit I revert back whatever config was there and recompile once again the switches module.

This works so far but it shows warnings during the test

warning: redefining module Clouseau.Switches (current version loaded from _build/test/lib/clouseau/ebin/Elixir.Clouseau.Switches.beam)
  lib/switches.ex:1

.warning: redefining module Clouseau.Switches (current version defined in memory)
  lib/switches.ex:1

Is there any better way to test the config options? Or maybe a better way to design the switches module but avoiding merging the hard coded switches with the configured every time in runtime?

Most Liked

LostKobrakai

LostKobrakai

LostKobrakai

LostKobrakai

I can see why keeping config in compile time can be desired, but if you can build your library to be configurable at runtime you can test it without being dependent on compile time. If you really want to stay at compile time you might want to look at ex_cldr, which does a bunch of compile time optimazations as well.

OvermindDL1

OvermindDL1

Yeah that’s my issue with the general config style. I still think we need a staged configuration system like I’ve used in other languages. Essentially you have global configuration, which only affects code-gen and acts as defaults for the later stages. Then there is the load-time configs, which should only affect things that happen at load time, in addition to acting as defaults for later stages. Then run-time configuration, which when changed then updates the configuration (users of this should either access it every time they need it, or for something like a database link it should receive an event when it changes so it can update as necessary), which of course acts as defaults for later stages. And finally there is the scoped stage which is where you pass in the configuration overrides to a specific call or instance of something.

This is similar to how I treat my TaskAfter library. I keep meaning to make a StagedConfiguration library for Elixir sometime, it would simplify a lot of my own work…

Where Next?

Popular in Questions Top

nobody
Hi! In PHP: $_SERVER[‘SERVER_ADDR’] - in Elixir? Searched the docs for ip address and the web, no good results. Thanks!
New
hariharasudhan94
I would like to know what is the best IDE for elixir development?
New
jay1
Why is it that the mnesia database isn’t the most preferred database for use in Elixir/Phoenix?
New
9mm
I am constructing a JSON object (map) and I need to conditionally set a field. I’m trying to write proper elixir-way code… and I’m at a l...
New
stefanluptak
Hello everybody, usually, I use a 29" ultra-wide monitor for VSCode which can easily accomodate explorer (files panel) + file with code ...
New
siddhant3030
Hi, I have to write a raw query for one of my project. But till now I have used ecto queries and don’t have much experience writing raw ...
New
freewebwithme
Using vs code and installed ElixirLS: support and debugger. And I got an error popped up on start up says Failed to run ‘elixir’ comma...
New

Other popular topics Top

hariharasudhan94
I would like to know what is the best IDE for elixir development?
New
JakeBecker
TL;DR: I’ve just released an implementation of Microsoft’s IDE-independent Language Server Protocol for Elixir. It adds language support ...
1144 55125 245
New
ashish173
I am using Ecto timestamps with postgres, I can see the timestamps() use the :naive_dateime but for my use case I wanted to store the ti...
New
sorentwo
Hello! tl;dr Announcing Oban, an Ecto based job processing library with a focus on reliability and historical observability. After spen...
985 44778 311
New
alice
Hey, Just curious what are the main benefits of Elixir compared to Clojure? When is Elixir more useful than Clojure and vice versa? Th...
New
sergio
Kind of like when jquery came out, it was super necessary. Existing drag and drop libraries have a bunch of baggage to support old browse...
New

We're in Beta

About us Mission Statement