Proton

A tool for data structure reuse. Or something. Easier to just illustrate. Suppose you need some relatively complex map, say as a slice of application state:

pizza = %{
  size: "large",
  verticality: "deep",
  crust: "cornmeal",
  sauce: "tomato",
  toppings: [
    %{name: "spinach", extra: 0},
    %{name: "giardiniera", spiciness: "hot", half: "left"}
  ],
  order: %{
    type: "pickup",
    placed_at: "19:30PM CST"
    customer: "Erik"
  }

Suppose you want to specify a bunch of these that share many things but also diverge in many particulars. Rather than writing that sort of thing over and over, it would be easier and more resistant to errors and the ravages of time if you could just do something like this:

%{
  protos: ["spinach_deep_dish", "size_large", "cornmeal_crust"]
  extras: ["left-hot-giardiniera"]
  order: %{
    type: "pickup",
    placed_at: "19:30PM CST"
    customer: "Erik"
  }
}

Note how we specify all of the real particulars of this order (customer, time, etc.) and pull in the rest attribute-by-attribute from other sources. Here protos is shorthand for "prototypes": i.e., data that we want to pull in, include, etc. Proton will expand those names into actual maps, (along with any prototypes they specify) flatten the list, and merge them into a single data structure for you.

Note: "extras" works that way too in this example, like a one-to-many relationship, though you'll have to expand those explicitly; Proton doesn't handle that for you. It's just a significant enough part of all the real world use cases I've come up with that it seemed reasonable to illustrate it. If we were talking about blog posts, "tags" might be a good case: You specify a list of tags as strings, and then use Proton to expand those into more intelligent entities that include, say, a URL for a semantic purposes.

The benefits are:

• Encapsulation: The first example above, where the pizza order is given as one big map literal, exposes the implementation of your pizza machinery because it exposes the raw data it consumes. What if you want to change formats, definitions, etc.?

• DRY: You probably sell a lot of deep dish spinach pizzas, and don't want to have to write it out each time.

• Extension/recombination: For all you know, "spinach_deep_dish" resolves to ["spinach", "deep_dish"]

And so on. Stay tuned for more examples from real world use.

Installation

If available in Hex, the package can be installed by adding proton to your list of dependencies in mix.exs:

def deps do
  [
    {:proton, "~> 0.2.0"}
  ]
end

Documentation can be generated with ExDoc and published on HexDocs. Once published, the docs can be found at https://hexdocs.pm/proton.