Elixir Drops is a collection of small modules that provide useful extensions and functions that can be used to work with data effectively.
This package can be installed by adding drops to your list of dependencies in mix.exs:
def deps do [ {:drops, "~> 0.2.0"} ] end
Documentation can be found at https://hexdocs.pm/drops.
You can use Drops.Contract to define data coercion and validation schemas with arbitrary validation rules.
Here's an example of a simple UserContract which defines two required keys and expected types:
defmodule UserContract do use Drops.Contract schema do %{ name: string(), email: string() } end end UserContract.conform(%{name: "Jane", email: "jane@doe.org"}) # {:ok, %{name: "Jane", email: "jane@doe.org"}} {:error, errors} = UserContract.conform(%{email: 312}) # {:error, # [ # %Drops.Validator.Messages.Error.Type{ # path: [:email], # text: "must be a string", # meta: %{args: [:string, 312], predicate: :type?} # }, # %Drops.Validator.Messages.Error.Key{ # path: [:name], # text: "key must be present", # meta: %{args: [:name], predicate: :has_key?} # } # ]} Enum.map(errors, &to_string/1) # ["email must be a string", "name key must be present"] {:error, errors} = UserContract.conform(%{name: "Jane", email: 312}) # {:error, # [ # %Drops.Validator.Messages.Error.Type{ # path: [:email], # text: "must be a string", # meta: %{args: [:string, 312], predicate: :type?} # } # ]} Enum.map(errors, &to_string/1) # ["email must be a string"]
Contract's schemas are a powerful way of defining the exact shape of the data you expect to work with. They are used to validate the structure and the values of the input data. Using schemas, you can define which keys are required and whic are optional, the exact types of the values and any additional checks that have to be applied to the values.
Schema attributes are required by defaults and attributes that are optional must be marked explicitly using optional. Here's an example:
defmodule UserContract do use Drops.Contract schema do %{ optional(:name) => string(), :name => string() } end end UserContract.conform(%{email: "janedoe.org"}) # {:ok, %{email: "janedoe.org"}} UserContract.conform(%{name: "Jane", email: "janedoe.org"}) # {:ok, %{name: "Jane", email: "janedoe.org"}}
If preferred, you can also use required to be more explicit. This schema is equivalent to the above:
schema do %{ optional(:name) => string(), required(:name) => string() } end
You can define the expected types of the values using string, integer, float, boolean, atom, map, list, any and maybe functions. Here's an example:
defmodule UserContract do use Drops.Contract schema do %{ name: string(), age: integer(), active: boolean(), tags: list(:string), settings: map(:string), address: maybe(:string) } end end
You can define types that must meet additional requirements by using built-in predicates. Here's an example:
defmodule UserContract do use Drops.Contract schema do %{ name: string(:filled?), age: integer(gt?: 18) } end end UserContract.conform(%{name: "Jane", age: 21}) # {:ok, %{name: "Jane", age: 21}} UserContract.conform(%{name: "", age: 21}) # {:error, # [ # %Drops.Validator.Messages.Error.Type{ # path: [:name], # text: "must be filled", # meta: %{args: [""], predicate: :filled?} # } # ]} UserContract.conform(%{name: "Jane", age: 12}) # {:error, # [ # %Drops.Validator.Messages.Error.Type{ # path: [:age], # text: "must be greater than 18", # meta: %{args: [18, 12], predicate: :gt?} # } # ]}
Schemas can be nested, including complex cases like nested lists and maps. Here's an example:
defmodule UserContract do use Drops.Contract schema do %{ user: %{ name: string(:filled?), age: integer(), address: %{ city: string(:filled?), street: string(:filled?), zipcode: string(:filled?) }, tags: list(%{ name: string(:filled?), created_at: integer() }) } } end end UserContract.conform(%{ user: %{ name: "Jane", age: 21, address: %{ city: "New York", street: "Broadway", zipcode: "10001" }, tags: [ %{name: "foo", created_at: 1_234_567_890}, %{name: "bar", created_at: 1_234_567_890} ] } }) # {:ok, # %{ # user: %{ # name: "Jane", # address: %{city: "New York", street: "Broadway", zipcode: "10001"}, # age: 21, # tags: [ # %{name: "foo", created_at: 1234567890}, # %{name: "bar", created_at: 1234567890} # ] # } # }} UserContract.conform(%{ user: %{ name: "Jane", age: 21, address: %{ city: "New York", street: "Broadway", zipcode: "" }, tags: [ %{name: "foo", created_at: 1_234_567_890}, %{name: "bar", created_at: nil} ] } }) # {:error, # [ # %Drops.Validator.Messages.Error.Type{ # path: [:user, :address, :zipcode], # text: "must be filled", # meta: %{args: [""], predicate: :filled?} # }, # %Drops.Validator.Messages.Error.Type{ # path: [:user, :tags, 1, :created_at], # text: "must be an integer", # meta: %{args: [:integer, nil], predicate: :type?} # } # ]}
You can define custom type casting functions that will be applied to the input data before it's validated. This is useful when you want to convert the input data to a different format, for example, when you want to convert a string to an integer. Here's an example:
defmodule UserContract do use Drops.Contract schema do %{ count: cast(:string) |> integer(gt?: 0) } end end UserContract.conform(%{count: "1"}) # {:ok, %{count: 1}} UserContract.conform(%{count: nil}) # [ # %Drops.Validator.Messages.Error.Caster{ # error: %Drops.Validator.Messages.Error.Type{ # path: [:count], # text: "must be a string", # meta: %{args: [:string, nil], predicate: :type?} # } # } # ]} UserContract.conform(%{count: "-1"}) # {:error, # [ # %Drops.Validator.Messages.Error.Type{ # path: [:count], # text: "must be greater than 0", # meta: %{args: [0, -1], predicate: :gt?} # } # ]}
It's also possible to define a custom casting module and use it via caster option:
defmodule CustomCaster do @spec cast(input_type :: atom(), output_type :: atom(), any, Keyword.t()) :: any() def cast(:string, :string, value, _opts) do String.downcase(value) end end defmodule UserContract do use Drops.Contract schema do %{ text: cast(:string, caster: CustomCaster) |> string() } end end UserContract.conform(%{text: "HELLO"}) # {:ok, %{text: "hello"}}
You can define a schema that will atomize the input map using atomize: true option. Only keys that you specified will be atomized, any unexpected key will be ignored. Here's an example:
defmodule UserContract do use Drops.Contract schema(atomize: true) do %{ name: string(), age: integer(), tags: list(%{ name: string() }) } end end UserContract.conform(%{ "unexpected" => "value", "this" => "should not be here", "name" => "Jane", "age" => 21, "tags" => [ %{"name" => "red"}, %{"name" => "green"}, %{"name" => "blue"} ] }) # {:ok, # %{ # name: "Jane", # age: 21, # tags: [%{name: "red"}, %{name: "green"}, %{name: "blue"}] # }}
If built-in types are not enough, or if you want to reuse schema definitions, you can define custom types using Drops.Type. Here's an example:
defmodule Types.Age do use Drops.Type, integer(gteq?: 0) end defmodule Types.Name do use Drops.Type, string(:filled?) end defmodule UserContract do use Drops.Contract schema do %{ name: Types.Name, age: Types.Age } end end UserContract.conform(%{name: "Jane", age: 42}) # {:ok, %{name: "Jane", age: 42}} {:error, errors} = UserContract.conform(%{name: "Jane", age: -42}) Enum.map(errors, &to_string/1) # ["age must be greater than or equal to 0"] {:error, errors} = UserContract.conform(%{name: "Jane", age: "42"}) Enum.map(errors, &to_string/1) # ["age must be an integer"]
You can also define reusable schemas, since they are represented as map type:
defmodule Types.User do use Drops.Type, %{ name: string(:filled?), age: integer(gteq?: 0) } end defmodule UserContract do use Drops.Contract schema do %{ user: Types.User } end end UserContract.conform(%{user: %{name: "Jane", age: 42}}) # {:ok, %{user: %{name: "Jane", age: 42}}} {:error, errors} = UserContract.conform(%{user: %{name: "Jane", age: -42}}) Enum.map(errors, &to_string/1) # ["user.age must be greater than or equal to 0"] {:error, errors} = UserContract.conform(%{user: %{name: "Jane", age: "42"}}) Enum.map(errors, &to_string/1) # ["user.age must be an integer"]
Another handy custom type is a union:
defmodule Types.Price do use Drops.Type, union([:integer, :float], gt?: 0) end defmodule ProductContract do use Drops.Contract schema do %{ price: Types.Price } end end ProductContract.conform(%{price: 42}) # {:ok, %{price: 42}} ProductContract.conform(%{price: 42.3}) # {:ok, %{price: 42.3}} {:error, errors} = ProductContract.conform(%{price: -42}) Enum.map(errors, &to_string/1) # ["price must be greater than 0"] {:error, errors} = ProductContract.conform(%{price: "42"}) Enum.map(errors, &to_string/1) # ["price must be an integer or price must be a float"]
You can define arbitrary rule functions using rule macro. These rules will be applied to the input data only if it passed schema validation. This way you can be sure that rules operate on data that's safe to work with.
Here's an example how you could define a rule that checks if either email or login is provided:
defmodule UserContract do use Drops.Contract schema do %{ email: maybe(:string), login: maybe(:string) } end rule(:either_login_or_email, %{email: nil, login: nil}) do {:error, "email or login must be provided"} end end UserContract.conform(%{email: "jane@doe.org", login: nil}) # {:ok, %{email: "jane@doe.org", login: nil}} UserContract.conform(%{email: nil, login: "jane"}) # {:ok, %{email: nil, login: "jane"}} UserContract.conform(%{email: nil, login: nil}) # {:error, # [ # %Drops.Validator.Messages.Error.Rule{ # path: [], # text: "email or login must be present", # meta: %{} # } # ]}