@spec async((() -> term()), opts :: [{:timeout, pos_integer()}]) ::
  {:middleware, Absinthe.Middleware.Async, term()}

Execute resolution field asynchronously.

This is a helper function for using the Absinthe.Middleware.Async.

Forbidden in mutation fields. (TODO: actually enforce this)

options

Options

• :timeout default: 30_000. The maximum timeout to wait for running the task.

example

Example

Using the Absinthe.Resolution.Helpers.async/1 helper function:

field :time_consuming, :thing do
  resolve fn _, _, _ ->
    async(fn ->
      {:ok, long_time_consuming_function()}
    end)
  end
end

@spec batch(
  Absinthe.Middleware.Batch.batch_fun(),
  term(),
  Absinthe.Middleware.Batch.post_batch_fun(),
  opts :: [{:timeout, pos_integer()}]
) :: {:middleware, Absinthe.Middleware.Batch, term()}

Batch the resolution of several functions together.

Helper function for creating Absinthe.Middleware.Batch

options

Options

• :timeout default: 5_000. The maximum timeout to wait for running a batch.

example

Example

Raw usage:

object :post do
  field :name, :string
  field :author, :user do
    resolve fn post, _, _ ->
      batch({__MODULE__, :users_by_id}, post.author_id, fn batch_results ->
        {:ok, Map.get(batch_results, post.author_id)}
      end)
    end
  end
end

def users_by_id(_, user_ids) do
  users = Repo.all from u in User, where: u.id in ^user_ids
  Map.new(users, fn user -> {user.id, user} end)
end

@spec dataloader(Dataloader.source_name()) :: dataloader_key_fun()

Resolve a field with a dataloader source.

This function is not imported by default. To make it available in your module do

import Absinthe.Resolution.Helpers

Same as dataloader/3, but it infers the resource name from the field name.

examples

Examples

field :author, :user, resolve: dataloader(Blog)

This is identical to doing the following.

field :author, :user, resolve: dataloader(Blog, :author, [])

@spec dataloader(Dataloader.source_name(), [dataloader_opt()]) :: dataloader_key_fun()

@spec dataloader(Dataloader.source_name(), any()) :: dataloader_key_fun()

Resolve a field with a dataloader source.

This function is not imported by default. To make it available in your module do

import Absinthe.Resolution.Helpers

Same as dataloader/3, but it infers the resource name from the field name. For opts see dataloader/3 on what options can be passed in.

examples

Examples

object :user do
  field :posts, list_of(:post),
    resolve: dataloader(Blog, args: %{deleted: false})

  field :organization, :organization do
    resolve dataloader(Accounts, use_parent: false)
  end

  field(:account_active, non_null(:boolean), resolve: dataloader(
      Accounts, callback: fn account, _parent, _args ->
        {:ok, account.active}
      end
    )
  )
end

@spec dataloader(Dataloader.source_name(), dataloader_key_fun() | any(), [
  dataloader_opt()
]) :: dataloader_key_fun()

Resolve a field with Dataloader

This function is not imported by default. To make it available in your module do

import Absinthe.Resolution.Helpers

While on_load/2 makes using dataloader directly easy within a resolver function, it is often unnecessary to need this level of direct control.

The dataloader/3 function exists to provide a simple API for using dataloader. It takes the name of a data source, the name of the resource you want to load, and then a variety of options.

basic-usage

Basic Usage

object :user do
  field :posts, list_of(:post),
    resolve: dataloader(Blog, :posts, args: %{deleted: false})

  field :organization, :organization do
    resolve dataloader(Accounts, :organization, use_parent: false)
  end

  field(:account_active, non_null(:boolean), resolve: dataloader(
      Accounts, :account, callback: fn account, _parent, _args ->
        {:ok, account.active}
      end
    )
  )
end

key-functions

Key Functions

Instead of passing in a literal like :posts or :organization in as the resource, it is also possible pass in a function:

object :user do
  field :posts, list_of(:post) do
    arg :limit, non_null(:integer)
    resolve dataloader(Blog, fn user, args, info ->
      args = Map.update!(args, :limit, fn val ->
        max(min(val, 20), 0)
      end)
      {:posts, args}
    end)
  end
end

In this case we want to make sure that the limit value cannot be larger than 20. By passing a callback function to dataloader/2 we can ensure that the value will fall nicely between 0 and 20.

If you have defined a custom batch query, you can instead return a Map with :batch key to specify the batch_key of Dataloader.load/4 / Dataloader.get/4 and an :item key to specify item_key argument of Dataloader.get/4.

# ...
resolve dataloader(Posts, fn user, _args, _info ->
  %{batch: {{:one, Post}, %{}}, item: [post_count: user]}
end)
# ...

options

Options

• :args default: %{}. Any arguments you want to always pass into the Dataloader.load/4 call. Resolver arguments are merged into this value and, in the event of a conflict, the resolver arguments win.
• :callback default: return result wrapped in ok or error tuple. Callback that is run with result of dataloader. It receives the result as the first argument, and the parent and args as second and third. Optionally can receive resolution as the fourth argument. Can be used to e.g. compute fields on the return value of the loader. Should return an ok or error tuple.
• :use_parent default: false. This option affects whether or not the dataloader/2 helper will use any pre-existing value on the parent. IE if you return %{author: %User{...}} from a blog post the helper will by default simply use the pre-existing author. Set it to true if you want to opt into using the pre-existing value instead of loading it fresh.

Ultimately, this helper calls Dataloader.load/4 using the loader in your context, the source you provide, the tuple {resource, args} as the batch key, and then the parent value of the field

def dataloader(source_name, resource) do
  fn parent, args, %{context: %{loader: loader}} ->
    args = Map.merge(opts[:args] || %{}, args)
    loader
    |> Dataloader.load(source_name, {resource, args}, parent)
    |> on_load(fn loader ->
      {:ok, Dataloader.get(loader, source_name, {resource, args}, parent)}
    end)
  end
end

Dataloader helper function

This function is not imported by default. To make it available in your module do

import Absinthe.Resolution.Helpers

This function helps you use data loader in a direct way within your schema. While normally the dataloader/1,2,3 helpers are enough, on_load/2 is useful when you want to load multiple things in a single resolver, or when you need fine grained control over the dataloader cache.

examples

Examples

field :reports, list_of(:report) do
  resolve fn shipment, _, %{context: %{loader: loader}} ->
    loader
    |> Dataloader.load(SourceName, :automatic_reports, shipment)
    |> Dataloader.load(SourceName, :manual_reports, shipment)
    |> on_load(fn loader ->
      reports =
        loader
        |> Dataloader.get(SourceName, :automatic_reports, shipment)
        |> Enum.concat(Dataloader.get(loader, SourceName, :manual_reports, shipment))
        |> Enum.sort_by(&reported_at/1)
      {:ok, reports}
    end)
  end
end