EctoResource

Eliminate boilerplate involved in defining basic CRUD functions in a Phoenix context or Elixir module.

Rationale

When using Context modules in a Phoenix application, there's a general need to define the standard CRUD functions for a given Ecto.Schema. Phoenix context generators will even do this automatically. Soon you will notice that there's quite a lot of code involved in CRUD access within your contexts.

This can become problematic for a few reasons:

  • Boilerplate functions for CRUD access, for every Ecto.Schema referenced in that context, introduce more noise than signal. This can obscure the more interesting details of the context.
  • These functions may tend to accumulate drift from the standard API by inviting edits for new use-cases, reducing the usefulness of naming conventions.
  • The burden of locally testing wrapper functions, yields low value for the writing and maintainence investment.

In short, at best this code is redundant and at worst is a deviant entanglement of modified conventions. All of which amounts to a more-painful development experience. EctoResource was created to ease this pain.

Features

Generate CRUD functions for a given Ecto.Repo and Ecto.Schema

EctoResource can be used to generate CRUD functions for a given Ecto.Repo and Ecto.Schema. By default it will create every function needed to create, read, update, and delete the resouce. It includes the ! version of each function (where relevant) that will raise an error instead of return a value.

Allow customization of generated resources

You can optionally include or exclude specific functions to generate exactly the functions your context requires. There's also two handy aliases for generating read functions and read/write functions.

Automatic pluralization

For methods that return a list of records, it seems natural to use a plural name. For example, take a function named MyContext.all_schema. While this works, it makes the grammar a bit awkward and distracts from the intent of the function. EctoResource uses Inflex when generating functions to create readable english function names automatically. For example, given the schema Person, a function named all_people/1 is generated.

Generate documentation for each generated function

Every function generated includes documentation so your application's documentation will include the generated functions with examples.

Reflection metadata

A function is generated for each resource defined by EctoResource to list all the functions generated for each Ecto.Repo and Ecto.Schema. A mix task is included to provide easy access to this information.

Supports any module

While EctoResource was designed for Phoenix Contexts in mind, It can be used in any Elixir module.

Installation

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

    def deps do
      [
        {:ecto_resource, "~> 1.1.0"}
      ]
    end

Usage

Basic usage - generate all EctoResource functions

defmodule MyApp.MyContext do
  alias MyApp.Repo
  alias MyApp.Schema
  use EctoResource

  using_repo(Repo) do
    resource(Schema)
  end
end

This generates all the functions EctoResource has to offer:

  • MyContext.all_schemas/1
  • MyContext.change_schema/1
  • MyContext.create_schema/1
  • MyContext.create_schema!/1
  • MyContext.delete_schema/1
  • MyContext.delete_schema!/1
  • MyContext.get_schema/2
  • MyContext.get_schema!/2
  • MyContext.get_schema_by/2
  • MyContext.get_schema_by!/2
  • MyContext.update_schema/2
  • MyContext.update_schema!/2

Explicit usage - generate only given functions

defmodule MyApp.MyContext do
  alias MyApp.Repo
  alias MyApp.Schema
  use EctoResource

  using_repo(Repo) do
    resource(Schema, only: [:create, :delete!])
  end
end

This generates only the given functions:

  • MyContext.create_schema/1
  • MyContext.delete_schema!/1

Exclusive usage - generate all but the given functions

defmodule MyApp.MyContext do
  alias MyApp.Repo
  alias MyApp.Schema
  use EctoResource

  using_repo(Repo) do
    resource(Schema, except: [:create, :delete!])
  end
end

This generates all the functions excluding the given functions:

  • MyContext.all_schemas/1
  • MyContext.change_schema/1
  • MyContext.create_schema!/1
  • MyContext.delete_schema/1
  • MyContext.get_schema/2
  • MyContext.get_schema_by/2
  • MyContext.get_schema_by!/2
  • MyContext.get_schema!/2
  • MyContext.update_schema/2
  • MyContext.update_schema!/2

Alias :read - generate data access functions

defmodule MyApp.MyContext do
  alias MyApp.Repo
  alias MyApp.Schema
  use EctoResource

  using_repo(Repo) do
    resource(Schema, :read)
  end
end

This generates all the functions necessary for reading data:

  • MyContext.all_schemas/1
  • MyContext.get_schema/2
  • MyContext.get_schema!/2

Alias :read_write - generate data access and manipulation functions, excluding delete

defmodule MyApp.MyContext do
  alias MyApp.Repo
  alias MyApp.Schema
  use EctoResource

  using_repo(Repo) do
    resource(Schema, :read_write)
  end
end

This generates all the functions except delete_schema/1 and delete_schema!/1:

  • MyContext.all_schemas/1
  • MyContext.change_schema/1
  • MyContext.create_schema/1
  • MyContext.create_schema!/1
  • MyContext.get_schema/2
  • MyContext.get_schema!/2
  • MyContext.update_schema/2
  • MyContext.update_schema!/2

Resource functions

The general idea of the generated resource functions is to abstract away the Ecto.Repo and Ecto.Schema parts of data access with Ecto and provide an API to the context that feels natural and clear to the caller.

The following examples will all assume a repo named Repo and a schema named Person.

all_people

Fetches a list of all %Person{} entries from the data store. Note: EctoResource will pluralize this function name using Inflex

iex> all_people()
[%Person{id: 1}]

iex> all_people(preloads: [:address])
[%Person{id: 1, address: %Address{}}]

iex> all_people(order_by: [desc: :id])
[%Person{id: 2}, %Person{id: 1}]

iex> all_people(preloads: [:address], order_by: [desc: :id]))
[
  %Person{
    id: 2,
    address: %Address{}
  },
  %Person{
    id: 1,
    address: %Address{}
  }
]

iex> all_people(where: [id: 2])
[%Person{id: 2, address: %Address{}}]

change_person

Creates a %Person{} changeset.

iex> change_person(%{name: "Example Person"})
#Ecto.Changeset<
  action: nil,
  changes: %{name: "Example Person"},
  errors: [],
  data: #Person<>,
  valid?: true
>

create_person

Inserts a %Person{} with the given attributes in the data store, returning an :ok/:error tuple.

iex> create_person(%{name: "Example Person"})
{:ok, %Person{id: 123, name: "Example Person"}}

iex> create_person(%{invalid: "invalid"})
{:error, %Ecto.Changeset}

create_person!

Inserts a %Person{} with the given attributes in the data store, returning a %Person{} or raises Ecto.InvalidChangesetError.

iex> create_person!(%{name: "Example Person"})
%Person{id: 123, name: "Example Person"}

iex> create_person!(%{invalid: "invalid"})
** (Ecto.InvalidChangesetError)

delete_person

Deletes a given %Person{} from the data store, returning an :ok/:error tuple.

iex> delete_person(%Person{id: 1})
{:ok, %Person{id: 1}}

iex> delete_person(%Person{id: 999})
{:error, %Ecto.Changeset}

delete_person!

Deletes a given %Person{} from the data store, returning the deleted %Person{}, or raises Ecto.StaleEntryError.

iex> delete_person!(%Person{id: 1})
%Person{id: 1}

iex> delete_person!(%Person{id: 999})
** (Ecto.StaleEntryError)

get_person

Fetches a single %Person{} from the data store where the primary key matches the given id, returns a %Person{} or nil.

iex> get_person(1)
%Person{id: 1}

iex> get_person(999)
nil

iex> get_person(1, preloads: [:address])
%Person{
    id: 1,
    address: %Address{}
}

get_person!

Fetches a single %Person{} from the data store where the primary key matches the given id, returns a %Person{} or raises Ecto.NoResultsError.

iex> get_person!(1)
%Person{id: 1}

iex> get_person!(999)
** (Ecto.NoResultsError)

iex> get_person!(1, preloads: [:address])
%Person{
    id: 1,
    address: %Address{}
}

get_person_by

Fetches a single %Person{} from the data store where the attributes match the given values.

iex> get_person_by(%{name: "Chuck Norris"})
%Person{name: "Chuck Norris"}

iex> get_person_by(%{name: "Doesn't Exist"})
nil

get_person_by!

Fetches a single %Person{} from the data store where the attributes match the given values. Raises an Ecto.NoResultsError if the record does not exist

iex> get_person_by!(%{name: "Chuck Norris"})
%Person{name: "Chuck Norris"}

iex> get_person_by!(%{name: "Doesn't Exist"})
** (Ecto.NoResultsError)

update_person

Updates a given %Person{} with the given attributes, returns an :ok/:error tuple.

iex> update_person(%Person{id: 1}, %{name: "New Person"})
{:ok, %Person{id: 1, name: "New Person"}}

iex> update_person(%Person{id: 1}, %{invalid: "invalid"})
{:error, %Ecto.Changeset}

update_person!

Updates a given %Person{} with the given attributes, returns a %Person{} or raises Ecto.InvalidChangesetError.

iex> update_person!(%Person{id: 1}, %{name: "New Person"})
%Person{id: 1, name: "New Person"}

iex> update_person!(%Person{id: 1}, %{invalid: "invalid"})
** (Ecto.InvalidChangesetError)

Caveats

This is not meant to be used as a wrapper for all the Repo functions within a context. Not all callbacks defined in Ecto.Repo are generated. EctoResource should be used to help reduce boilerplate code and tests for general CRUD operations.

It may be the case that EctoResource needs to evolve and provide slightly more functionality/flexibility in the future. However, the general focus is reducing boilerplate code.

Contribution

Bug reports

If you discover any bugs, feel free to create an issue on GitHub. Please add as much information as possible to help in fixing the potential bug. You are also encouraged to help even more by forking and sending us a pull request.

Issues on GitHub

Pull requests

  • Fork it (https://github.com/daytonn/ecto_resource/fork)
  • Add upstream remote (git remote add upstream git@github.com:daytonn/ecto_resource.git)
  • Make sure you're up-to-date with upstream main (git pull upstream main)
  • Create your feature branch (git checkout -b feature/fooBar)
  • Commit your changes (git commit -am 'Add some fooBar')
  • Push to the branch (git push origin feature/fooBar)
  • Create a new Pull Request

Nice to have features/improvements (:point_up::wink:)

  • Ability to override pluralization
  • Find functions (maybe?)

License

Apache 2.0