Ruby Grape

What is Grape?

Grape is a REST-like API framework for Ruby. It's designed to run on Rack or complement existing web application frameworks such as Rails and Sinatra by providing a simple DSL to easily develop RESTful APIs. It has built-in support for common conventions, including multiple formats, subdomain/prefix restriction, content negotiation, versioning and much more.

Stable Release

You're reading the documentation for the stable release of Grape, which should be 4.1.0. The current stable release is 4.0.1.

Project Resources

• Grape Website
• Documentation
• Need help? Open an Issue
• Follow us on Twitter

Grape for Enterprise

Available as part of the Tidelift Subscription.

The maintainers of Grape are working with Tidelift to deliver commercial support and maintenance. Save time, reduce risk, and improve code health, while paying the maintainers of Grape. Click here for more details.

Installation

Ruby 3.3.1 or newer is required.

Grape is available as a gem, to install it run:

bundle add grape

Basic Usage

Grape APIs are Rack applications that are created by subclassing Grape::API. Below is a simple example showing some of the more common features of Grape in the context of recreating parts of the Twitter API.

module Twitter
  class API < Grape::API
    version 'v1', using: :header, vendor: 'twitter'
    format :json
    prefix :api

    helpers do
      def current_user
        @current_user ||= User.authorize!(env)
      end

      def authenticate!
        error!('401 Unauthorized', 401) unless current_user
      end
    end

    resource :statuses do
      desc 'Return a public timeline.'
      get :public_timeline do
        Status.limit(20)
      end

      desc 'Return a personal timeline.'
      get :home_timeline do
        authenticate!
        current_user.statuses.limit(20)
      end

      desc 'Return a status.'
      params do
        requires :id, type: Integer, desc: 'Status ID.'
      end
      route_param :id do
        get do
          Status.find(params[:id])
        end
      end

      desc 'Create a status.'
      params do
        requires :status, type: String, desc: 'Your status.'
      end
      post do
        authenticate!
        Status.create!({
          user: current_user,
          text: params[:status]
        })
      end

      desc 'Update a status.'
      params do
        requires :id, type: String, desc: 'Status ID.'
        requires :status, type: String, desc: 'Your status.'
      end
      put ':id' do
        authenticate!
        current_user.statuses.find(params[:id]).update({
          user: current_user,
          text: params[:status]
        })
      end

      desc 'Delete a status.'
      params do
        requires :id, type: String, desc: 'Status ID.'
      end
      delete ':id' do
        authenticate!
        current_user.statuses.find(params[:id]).destroy
      end
    end
  end
end

Rails 7.1

Grape's deprecator will be added to your application's deprecators automatically as :grape, so that your application's configuration can be applied to it.

Mounting

All

By default Grape will compile the routes on the first request, but it is possible to pre-load routes using the compile! method.

Twitter::API.compile!

This can be added to your config.ru (if using rackup), application.rb (if using rails), or any file that loads your server.

Rack

The above sample creates a Rack application that can be run from a rackup config.ru file with rackup:

run Twitter::API

(With pre-loading you can use)

Twitter::API.compile!
run Twitter::API

And would respond to the following routes:

GET /api/statuses/public_timeline
GET /api/statuses/home_timeline
GET /api/statuses/:id
POST /api/statuses
PUT /api/statuses/:id
DELETE /api/statuses/:id

Grape will also automatically respond to HEAD and OPTIONS for all GET, and just OPTIONS for all other routes.

Alongside Sinatra (or other frameworks)

If you wish to mount Grape alongside another Rack framework such as Sinatra, you can do so easily using Rack::Cascade:

# Example config.ru

require 'sinatra'
require 'grape'

class API < Grape::API
  get :hello do
    { hello: 'world' }
  end
end

class Web < Sinatra::Base
  get '/' do
    'Hello world.'
  end
end

use Rack::Session::Cookie
run Rack::Cascade.new [Web, API]

Note that order of loading apps using Rack::Cascade matters. The grape application must be last if you want to raise custom 404 errors from grape (such as error!('Not Found',404)). If the grape application is not last and returns 404 or 405 response, cascade utilizes that as a signal to try the next app. This may lead to undesirable behavior showing the wrong 404 page from the wrong app.

Rails

Place API files into app/api. Rails expects a subdirectory that matches the name of the Ruby module and a file name that matches the name of the class. In our example, the file name location and directory for Twitter::API should be app/api/twitter/api.rb.

Modify config/routes:

mount Twitter::API => '/'

Zeitwerk

Rails's default autoloader is Zeitwerk. By default, it inflects api as Api instead of API. To make our example work, you need to uncomment the lines at the bottom of config/initializers/inflections.rb, and add API as an acronym:

ActiveSupport::Inflector.inflections(:en) do |inflect|
  inflect.acronym 'API'
end

Modules

You can mount multiple API implementations inside another one. These don't have to be different versions, but may be components of the same API.

class Twitter::API < Grape::API
  mount Twitter::APIv1
  mount Twitter::APIv2
end

You can also mount on a path, which is similar to using prefix inside the mounted API itself.

class Twitter::API < Grape::API
  mount Twitter::APIv1 => '/v1'
end

Declarations as before/after/rescue_from can be placed before or after mount. In any case they will be inherited.

class Twitter::API < Grape::API
  before do
    header 'X-Base-Header', 'will be defined for all APIs that are mounted below'
  end

  rescue_from :all do
    error!({ "error" => "Internal Server Error" }, 500)
  end

  mount Twitter::Users
  mount Twitter::Search

  after do
    clean_cache!
  end

  rescue_from ZeroDivisionError do
    error!({ "error" => "Not found" }, 404)
  end
end

Remounting

You can mount the same endpoints in two different locations.

class Voting::API < Grape::API
  namespace 'votes' do
    get do
      # Your logic
    end

    post do
      # Your logic
    end
  end
end

class Post::API < Grape::API
  mount Voting::API
end

class Comment::API < Grape::API
  mount Voting::API
end

Assuming that the post and comment endpoints are mounted in /posts and /comments, you should now be able to do get /posts/votes, post /posts/votes, get /comments/votes and post /comments/votes.

Mount Configuration

You can configure remountable endpoints to change how they behave according to where they are mounted.

class Voting::API < Grape::API
  namespace 'votes' do
    desc "Vote for your #{configuration[:votable]}"
    get do
      # Your logic
    end
  end
end

class Post::API < Grape::API
  mount Voting::API, with: { votable: 'posts' }
end

class Comment::API < Grape::API
  mount Voting::API, with: { votable: 'comments' }
end

Note that if you're passing a hash as the first parameter to mount, you will need to explicitly put () around parameters:

# good
mount({ ::Some::Api => '/some/api' }, with: { condition: true })

# bad
mount ::Some::Api => '/some/api', with: { condition: true }

You can access configuration on the class (to use as dynamic attributes), inside blocks (like namespace)

If you want logic happening based on a configuration, you can use the helper given.

class ConditionalEndpoint::API < Grape::API
  given configuration[:some_setting] do
    get 'mount_this_endpoint_conditionally' do
      configuration[:configurable_response]
    end
  end
end

If you want a block of logic running every time an endpoint is mounted (within which you can access the configuration Hash)

class ConditionalEndpoint::API < Grape::API
  mounted do
    YourLogger.info "This API was mounted at: #{Time.now}"

    get configuration[:endpoint_name] do
      configuration[:configurable_response]
    end
  end
end

More complex results can be achieved by using mounted as an expression within which the configuration is already evaluated as a Hash.

class ExpressionEndpointAPI < Grape::API
  get(mounted { configuration[:route_name] || 'default_name' }) do
    # some logic
  end
end

class BasicAPI < Grape::API
  desc 'Statuses index' do
    params((configuration[:entity] || API::Entities::Status).documentation)
  end
  params do
    requires :all, using: (configuration[:entity] || API::Entities::Status).documentation
  end
  get '/statuses' do
    statuses = Status.all
    type = current_user.admin? ? :full : :default
    present statuses, with: (configuration[:entity] || API::Entities::Status), type: type
  end
end

class V1 < Grape::API
  version 'v1'
  mount BasicAPI, with: { entity: mounted { configuration[:entity] || API::Entities::Status } }
end

class V2 < Grape::API
  version 'v2'
  mount BasicAPI, with: { entity: mounted { configuration[:entity] || API::Entities::V2::Status } }
end

Versioning

You have the option to provide various versions of your API by establishing a separate Grape::API class for each offered version and then integrating them into a primary Grape::API class. Ensure that newer versions are mounted before older ones. The default approach to versioning directs the request to the subsequent Rack middleware if a specific version is not found.

require 'v1'
require 'v2'
require 'v3'
class App < Grape::API
  mount V3
  mount V2
  mount V1
end

To maintain the same endpoints from earlier API versions without rewriting them, you can indicate multiple versions within the previous API versions.

class V1 < Grape::API
  version 'v1', 'v2', 'v3'

  get '/foo' do
    # your code for GET /foo
  end

  get '/other' do
    # your code for GET /other
  end
end

class V2 < Grape::API
  version 'v2', 'v3'

  get '/var' do
    # your code for GET /var
  end
end

class V3 < Grape::API
  version 'v3'

  get '/foo' do
    # your new code for GET /foo
  end
end

Using the example provided, the subsequent endpoints will be accessible across various versions:

GET /v1/foo
GET /v1/other
GET /v2/foo # => Same behavior as v1
GET /v2/other # => Same behavior as v1
GET /v2/var # => New endpoint not available in v1
GET /v3/foo # => Different behavior to v1 and v2
GET /v3/other # => Same behavior as v1 and v2
GET /v3/var # => Same behavior as v2

There are four strategies in which clients can reach your API's endpoints: :path, :header, :accept_version_header and :param. The default strategy is :path.

Strategies

Path

version 'v1', using: :path

Using this versioning strategy, clients should pass the desired version in the URL.

curl http://localhost:9292/v1/statuses/public_timeline

Header

version 'v1', using: :header, vendor: 'twitter'

Currently, Grape only supports versioned media types in the following format:

vnd.vendor-and-or-resource-v1234+format

Basically all tokens between the final - and the + will be interpreted as the version.

Using this versioning strategy, clients should pass the desired version in the HTTP Accept header.

curl -H Accept:application/vnd.twitter-v1+json http://localhost:9292/statuses/public_timeline

By default, the first matching version is used when no Accept header is supplied. This behavior is similar to routing in Rails. To circumvent this default behavior, one could use the :strict option. When this option is set to true, a 406 Not Acceptable error is returned when no correct Accept header is supplied and the :cascade option is set to false. Otherwise a 404 Not Found error is returned by Rack if no other route matches.

When an invalid Accept header is supplied, a 406 Not Acceptable error is returned if the :cascade option is set to false. Otherwise a 404 Not Found error is returned by Rack if no other route matches.

Grape will evaluate the relative quality preference included in Accept headers and default to a quality of 1.0 when omitted. In the following example a Grape API that supports XML and JSON in that order will return JSON:

curl -H "Accept: text/xml;q=0.8, application/json;q=0.9" localhost:1234/resource

Accept-Version Header

version 'v1', using: :accept_version_header

Using this versioning strategy, clients should pass the desired version in the HTTP Accept-Version header.

curl -H "Accept-Version:v1" http://localhost:9292/statuses/public_timeline

By default, the first matching version is used when no Accept-Version header is supplied. This behavior is similar to routing in Rails. To circumvent this default behavior, one could use the :strict option. When this option is set to true, a 406 Not Acceptable error is returned when no correct Accept-Version header is supplied and the :cascade option is set to false. Otherwise a 404 Not Found error is returned by Rack if no other route matches.

Param

version 'v1', using: :param

Using this versioning strategy, clients should pass the desired version as a parameter in the URL query string. A version given in the request body is not read.

curl http://localhost:9292/statuses/public_timeline?apiver=v1

The default name for the query parameter is 'apiver' but can be specified using the :parameter option.

version 'v1', using: :param, parameter: 'v'

curl http://localhost:9292/statuses/public_timeline?v=v1

Linting

You can check whether your API is in conformance with the Rack's specification by calling lint! at the top level of the API you serve, or through configuration. Every response the API gives is then checked once with Rack::Lint, including the 404 for a path no route matches and the responses of Rack apps mounted with mount. A lint! inside a namespace, or in an API mounted into another one, has no effect.

class Api < Grape::API
  lint!
end

Grape.configure do |config|
  config.lint = true
end

Grape.config.lint = true

Bug in Rack::ETag under Rack 3.X

If you're using Rack 3.X and the Rack::ETag middleware (used by Rails), a bug related to linting has been fixed in 3.1.13 and 3.0.15 respectively.

Describing Methods

You can add a description to API methods and namespaces. The description would be used by grape-swagger to generate swagger compliant documentation.

Note: Description block is only for documentation and won't affect API behavior.

desc 'Returns your public timeline.' do
  summary 'summary'
  detail 'more details'
  params  API::Entities::Status.documentation
  success API::Entities::Entity
  failure [[401, 'Unauthorized', 'Entities::Error']]
  default_response code: 500, message: 'InvalidRequest', model: Entities::Error
  named 'My named route'
  headers XAuthToken: {
            description: 'Validates your identity',
            required: true
          },
          XOptionalHeader: {
            description: 'Not really needed',
            required: false
          }
  hidden false
  deprecated false
  is_array true
  nickname 'nickname'
  produces ['application/json']
  consumes ['application/json']
  tags ['tag1', 'tag2']
end
get :public_timeline do
  Status.limit(20)
end

• detail: A more enhanced description
• params: Define parameters directly from an Entity
• success: (former entity) The Entity to be used to present the success response for this route.
• failure: (former http_codes) A definition of the used failure HTTP Codes and Entities.
• default_response: (former default) The definition and Entity used to present the default response for this route.
• named: A helper to give a route a name and find it with this name in the documentation Hash
• headers: A definition of the used Headers
• Other options can be found in grape-swagger

Configuration

Use Grape.configure to set up global settings at load time. Currently the configurable settings are:

• param_builder: Sets the Parameter Builder, defaults to :hash_with_indifferent_access.
• lint: Checks every response against Rack's specification, defaults to false. See Linting.
• warn_on_helper_overrides: Warns when a helper method masks a Grape::Endpoint instance method, defaults to false.
• raise_rendering_errors: Lets an error response that cannot be rendered propagate out of the middleware stack instead of being answered with a failsafe 500, defaults to false. See When the error response itself cannot be rendered.
• raise_rescue_handler_errors: Lets an exception raised inside a rescue_from block propagate out of the middleware stack instead of being redispatched, defaults to false. See Re-raising from inside a rescue_from block.

To change a setting value make sure that at some point during load time the following code runs

Grape.configure do |config|
  config.setting = value
end

For example, for the param_builder, the following code could run in an initializer:

Grape.configure do |config|
  config.param_builder = :hashie_mash
end

Available parameter builders are :hash, :hash_with_indifferent_access, and :hashie_mash. See params_builder.

You can also configure a single API:

API.configure do |config|
  config[key] = value
end

This will be available inside the API with configuration, as if it were mount configuration.

Parameters

Request parameters are available through the params hash object. This includes GET, POST and PUT parameters, along with any named parameters you specify in your route strings.

get :public_timeline do
  Status.order(params[:sort_by])
end

Parameters are automatically populated from the request body on POST, PUT, PATCH, DELETE and QUERY for form input, JSON and XML content-types.

The request:

curl -d '{"text": "140 characters"}' 'http://localhost:9292/statuses' -H Content-Type:application/json -v

The Grape endpoint:

post '/statuses' do
  Status.create!(text: params[:text])
end

Multipart POSTs and PUTs are supported as well.

The request:

curl --form image_file='@image.jpg;type=image/jpg' http://localhost:9292/upload

The Grape endpoint:

post 'upload' do
  # file in params[:image_file]
end

In the case of conflict between either of:

• route string parameters
• GET, POST and PUT parameters
• the contents of the request body on POST, PUT, PATCH, DELETE and QUERY

Route string parameters will have precedence.

Params Class

By default parameters are available as ActiveSupport::HashWithIndifferentAccess. This can be changed to, for example, Ruby Hash or Hashie::Mash for the entire API.

class API < Grape::API
  build_with :hashie_mash

  params do
    optional :color, type: String
  end
  get do
    params.color # instead of params[:color]
  end
end

The class can also be overridden on individual parameter blocks using build_with as follows.

params do
  build_with :hash
  optional :color, type: String
end

In the example above, params["color"] will return nil since params is a plain Hash.

Available parameter builders are :hash, :hash_with_indifferent_access, and :hashie_mash. See params_builder.

Declared

Grape allows you to access only the parameters that have been declared by your params block. It will:

• Filter out the params that have been passed, but are not allowed.
• Include any optional params that are declared but not passed.
• Perform any parameter renaming on the resulting hash.

Consider the following API endpoint:

format :json

post 'users/signup' do
  { 'declared_params' => declared(params) }
end

If you do not specify any parameters, declared will return an empty hash.

Request

curl -X POST -H "Content-Type: application/json" localhost:9292/users/signup -d '{"user": {"first_name":"first name", "last_name": "last name"}}'

Response

{
  "declared_params": {}
}

Once we add parameters requirements, grape will start returning only the declared parameters.

format :json

params do
  optional :user, type: Hash do
    optional :first_name, type: String
    optional :last_name, type: String
  end
end

post 'users/signup' do
  { 'declared_params' => declared(params) }
end

Request

curl -X POST -H "Content-Type: application/json" localhost:9292/users/signup -d '{"user": {"first_name":"first name", "last_name": "last name", "random": "never shown"}}'

Response

{
  "declared_params": {
    "user": {
      "first_name": "first name",
      "last_name": "last name"
    }
  }
}

Missing params that are declared as type Hash or Array will be included.

format :json

params do
  optional :user, type: Hash do
    optional :first_name, type: String
    optional :last_name, type: String
  end
  optional :widgets, type: Array
end

post 'users/signup' do
  { 'declared_params' => declared(params) }
end

Request

curl -X POST -H "Content-Type: application/json" localhost:9292/users/signup -d '{}'

Response

{
  "declared_params": {
    "user": {
      "first_name": null,
      "last_name": null
    },
    "widgets": []
  }
}

The returned hash is an ActiveSupport::HashWithIndifferentAccess.

The #declared method is not available to before filters, as those are evaluated prior to parameter coercion.

Include Parent Namespaces

By default declared(params) includes parameters that were defined in all parent namespaces. If you want to return only parameters from your current namespace, you can set include_parent_namespaces option to false.

format :json

namespace :parent do
  params do
    requires :parent_name, type: String
  end

  namespace ':parent_name' do
    params do
      requires :child_name, type: String
    end
    get ':child_name' do
      {
        'without_parent_namespaces' => declared(params, include_parent_namespaces: false),
        'with_parent_namespaces' => declared(params, include_parent_namespaces: true),
      }
    end
  end
end

Request

curl -X GET -H "Content-Type: application/json" localhost:9292/parent/foo/bar

Response

{
  "without_parent_namespaces": {
    "child_name": "bar"
  },
  "with_parent_namespaces": {
    "parent_name": "foo",
    "child_name": "bar"
  }
}

Include Missing

By default declared(params) includes parameters that have nil values. If you want to return only the parameters that are not nil, you can use the include_missing option. By default, include_missing is set to true. Consider the following API:

format :json

params do
  requires :user, type: Hash do
    requires :first_name, type: String
    optional :last_name, type: String
  end
end

post 'users/signup' do
  { 'declared_params' => declared(params, include_missing: false) }
end

Request

curl -X POST -H "Content-Type: application/json" localhost:9292/users/signup -d '{"user": {"first_name":"first name", "random": "never shown"}}'

Response with include_missing:false

{
  "declared_params": {
    "user": {
      "first_name": "first name"
    }
  }
}

Response with include_missing:true

{
  "declared_params": {
    "user": {
      "first_name": "first name",
      "last_name": null
    }
  }
}

It also works on nested hashes:

format :json

params do
  requires :user, type: Hash do
    requires :first_name, type: String
    optional :last_name, type: String
    requires :address, type: Hash do
      requires :city, type: String
      optional :region, type: String
    end
  end
end

post 'users/signup' do
  { 'declared_params' => declared(params, include_missing: false) }
end

Request

curl -X POST -H "Content-Type: application/json" localhost:9292/users/signup -d '{"user": {"first_name":"first name", "random": "never shown", "address": { "city": "SF"}}}'

Response with include_missing:false

{
  "declared_params": {
    "user": {
      "first_name": "first name",
      "address": {
        "city": "SF"
      }
    }
  }
}

Response with include_missing:true

{
  "declared_params": {
    "user": {
      "first_name": "first name",
      "last_name": null,
      "address": {
        "city": "SF",
        "region": null
      }
    }
  }
}

Note that an attribute with a nil value is not considered missing and will also be returned when include_missing is set to false:

Request

curl -X POST -H "Content-Type: application/json" localhost:9292/users/signup -d '{"user": {"first_name":"first name", "last_name": null, "address": { "city": "SF"}}}'

Response with include_missing:false

{
  "declared_params": {
    "user": {
      "first_name": "first name",
      "last_name": null,
      "address": { "city": "SF"}
    }
  }
}

Evaluate Given

By default declared(params) will not evaluate given and return all parameters. Use evaluate_given to evaluate all given blocks and return only parameters that satisfy given conditions. Consider the following API:

format :json

params do
  optional :child_id, type: Integer
  given :child_id do
    requires :father_id, type: Integer
  end
end

post 'child' do
  { 'declared_params' => declared(params, evaluate_given: true) }
end

Request

curl -X POST -H "Content-Type: application/json" localhost:9292/child -d '{"father_id": 1}'

Response with evaluate_given:false

{
  "declared_params": {
    "child_id": null,
    "father_id": 1
  }
}

Response with evaluate_given:true

{
  "declared_params": {
    "child_id": null
  }
}

It also works on nested hashes:

format :json

params do
  requires :child, type: Hash do
    optional :child_id, type: Integer
    given :child_id do
      requires :father_id, type: Integer
    end
  end
end

post 'child' do
  { 'declared_params' => declared(params, evaluate_given: true) }
end

Request

curl -X POST -H "Content-Type: application/json" localhost:9292/child -d '{"child": {"father_id": 1}}'

Response with evaluate_given:false

{
  "declared_params": {
    "child": {
      "child_id": null,
      "father_id": 1
    }
  }
}

Response with evaluate_given:true

{
  "declared_params": {
    "child": {
      "child_id": null
    }
  }
}

Parameter Precedence

Using route_param takes higher precedence over a regular parameter defined with same name:

format :json

params do
  requires :foo, type: String
end
route_param :foo do
  post do
    { value: params[:foo] }
  end
end

Request

curl -X POST -H "Content-Type: application/json" localhost:9292/bar -d '{"foo": "baz"}'

Response

{
  "value": "bar"
}

Parameter Validation and Coercion

You can define validations and coercion options for your parameters using a params block.

params do
  requires :id, type: Integer
  optional :text, type: String, regexp: /\A[a-z]+\z/
  group :media, type: Hash do
    requires :url
  end
  optional :audio, type: Hash do
    requires :format, type: Symbol, values: [:mp3, :wav, :aac, :ogg]
  end
  mutually_exclusive :media, :audio
end
put ':id' do
  # params[:id] is an Integer
end

When a type is specified an implicit validation is done after the coercion to ensure the output type is the one declared.

Optional parameters can have a default value.

params do
  optional :color, type: String, default: 'blue'
  optional :random_number, type: Integer, default: -> { Random.rand(1..100) }
  optional :non_random_number, type: Integer, default:  Random.rand(1..100)
end

Default values are eagerly evaluated. Above :non_random_number will evaluate to the same number for each call to the endpoint of this params block. To have the default evaluate lazily with each request use a lambda, like :random_number above.

Note that default values will be passed through to any validation options specified. The following example raises Grape::Exceptions::IncompatibleOptionValues when the API is defined, since the default is not one of the allowed values.

params do
  optional :color, type: String, default: 'blue', values: ['red', 'green']
end

The correct implementation is to ensure the default value passes all validations.

params do
  optional :color, type: String, default: 'blue', values: ['blue', 'red', 'green']
end

You can use the value of one parameter as the default value of some other parameter. In this case, if the primary_color parameter is not provided, it will have the same value as the color one. If both of them not provided, both of them will have blue value.

params do
  optional :color, type: String, default: 'blue'
  optional :primary_color, type: String, default: -> (params) { params[:color] }
end

Supported Parameter Types

The following are all valid types, supported out of the box by Grape:

• Integer
• Float
• BigDecimal
• Numeric
• Date
• DateTime
• Time
• Boolean
• String
• Symbol
• Rack::Multipart::UploadedFile (alias File)
• JSON

Custom Types and Coercions

Aside from the default set of supported types listed above, any class can be used as a type as long as an explicit coercion method is supplied. If the type implements a class-level parse method, Grape will use it automatically. This method must take one string argument and return an instance of the correct type, or return an instance of Grape::Validations::Types::InvalidValue which optionally accepts a message to be returned in the response.

class Color
  attr_reader :value
  def initialize(color)
    @value = color
  end

  def self.parse(value)
    return new(value) if %w[blue red green].include?(value)

    Grape::Validations::Types::InvalidValue.new('Unsupported color')
  end
end

params do
  optional :color, type: Color, default: Color.new('blue')
  requires :more_colors, type: Array[Color] # Collections work
  optional :unique_colors, type: Set[Color] # Duplicates discarded
end

get '/stuff' do
  # params[:color] is already a Color.
  params[:color].value
end

Alternatively, a custom coercion method may be supplied for any type of parameter using coerce_with. Any class or object may be given that implements a parse or call method, in that order of precedence. The method must accept a single string parameter, and the return value must match the given type.

params do
  requires :passwd, type: String, coerce_with: Base64.method(:decode64)
  requires :loud_color, type: Color, coerce_with: ->(c) { Color.parse(c.downcase) }

  requires :obj, type: Hash, coerce_with: JSON do
    requires :words, type: Array[String], coerce_with: ->(val) { val.split(/\s+/) }
    optional :time, type: Time, coerce_with: Chronic
  end
end

Note that, a nil value will call the custom coercion method, while a missing parameter will not.

Example of use of coerce_with with a lambda (a class with a parse method could also have been used) It will parse a string and return an Array of Integers, matching the Array[Integer] type.

params do
  requires :values, type: Array[Integer], coerce_with: ->(val) { val.split(/\s+/).map(&:to_i) }
end

Grape will assert that coerced values match the given type, and will reject the request if they do not. To override this behaviour, custom types may implement a parsed? method that should accept a single argument and return true if the value passes type validation.

class SecureUri
  def self.parse(value)
    URI.parse value
  end

  def self.parsed?(value)
    value.is_a? URI::HTTPS
  end
end

params do
  requires :secure_uri, type: SecureUri
end

Multipart File Parameters

Grape makes use of Rack::Request's built-in support for multipart file parameters. Such parameters can be declared with type: File:

params do
  requires :avatar, type: File
end
post '/' do
  params[:avatar][:filename] # => 'avatar.png'
  params[:avatar][:type] # => 'image/png'
  params[:avatar][:tempfile] # => #<File>
end

First-Class JSON Types

Grape supports complex parameters given as JSON-formatted strings using the special type: JSON declaration. JSON objects and arrays of objects are accepted equally, with nested validation rules applied to all objects in either case:

params do
  requires :json, type: JSON do
    requires :int, type: Integer, values: [1, 2, 3]
  end
end
get '/' do
  params[:json].inspect
end

client.get('/', json: '{"int":1}') # => '{"int" => 1}'
client.get('/', json: '[{"int":"1"}]') # => '[{"int" => 1}]'

client.get('/', json: '{"int":4}') # => HTTP 400
client.get('/', json: '[{"int":4}]') # => HTTP 400

Additionally type: Array[JSON] may be used, which explicitly marks the parameter as an array of objects. If a single object is supplied it will be wrapped.

params do
  requires :json, type: Array[JSON] do
    requires :int, type: Integer
  end
end
get '/' do
  params[:json].each { |obj| ... } # always works
end

For stricter control over the type of JSON structure which may be supplied, use type: Array, coerce_with: JSON or type: Hash, coerce_with: JSON.

Multiple Allowed Types

Variant-type parameters can be declared using the types option rather than type:

params do
  requires :status_code, types: [Integer, String, Array[Integer, String]]
end
get '/' do
  params[:status_code].inspect
end

client.get('/', status_code: 'OK_GOOD') # => "OK_GOOD"
client.get('/', status_code: 300) # => 300
client.get('/', status_code: %w(404 NOT FOUND)) # => [404, "NOT", "FOUND"]

As a special case, variant-member-type collections may also be declared, by passing a Set or Array with more than one member to type:

params do
  requires :status_codes, type: Array[Integer,String]
end
get '/' do
  params[:status_codes].inspect
end

client.get('/', status_codes: %w(1 two)) # => [1, "two"]

Multiple Hash Schemas with oneof

A Hash parameter that may take one of several different shapes can be declared with the oneof: option. Each variant is a Proc that uses the normal params DSL — so the full validator surface (requires, optional, regexp:, values:, allow_blank:, nested Hash/Array, etc.) is available inside.

params do
  requires :value, type: Hash, oneof: [
    proc { requires :fixed_price, type: Float },
    proc do
      requires :time_unit, type: String
      requires :rate, type: Float
    end
  ]
end
post '/pricing' do
  params[:value]
end

Both of these requests succeed:

curl -d '{"value":{"fixed_price":100.0}}' -H 'Content-Type: application/json' /pricing
curl -d '{"value":{"time_unit":"hour","rate":50.0}}' -H 'Content-Type: application/json' /pricing

A request that doesn't match any variant returns 400 with value does not match any of the allowed schemas.

Variants are tried in declaration order; the first variant that validates without errors wins, and any coercions it performed (e.g. "150.5" → 150.5) are applied to the request params. Nested hash structures inside variants work the same as elsewhere in Grape:

params do
  requires :options, type: Hash, oneof: [
    proc do
      requires :form, type: Hash do
        requires :colour, type: String
        optional :size, type: Integer
      end
    end,
    proc do
      requires :api, type: Hash do
        requires :authenticated, type: Grape::API::Boolean
      end
    end
  ]
end

oneof: requires type: Hash. The variants array must be non-empty and each entry must be a Proc; violating either raises an ArgumentError at definition time.

Validation of Nested Parameters

Parameters can be nested using group or by calling requires or optional with a block. In the above example, this means params[:media][:url] is required along with params[:id], and params[:audio][:format] is required only if params[:audio] is present. With a block, group, requires and optional require an additional option type, which can be either Array or Hash; a block without one raises Grape::Exceptions::MissingGroupType. Depending on the value, the nested parameters will be treated either as values of a hash or as values of hashes in an array.

params do
  optional :preferences, type: Array do
    requires :key
    requires :value
  end

  requires :name, type: Hash do
    requires :first_name
    requires :last_name
  end
end

Dependent Parameters

Suppose some of your parameters are only relevant if another parameter is given; Grape allows you to express this relationship through the given method in your parameters block, like so:

params do
  optional :shelf_id, type: Integer
  given :shelf_id do
    requires :bin_id, type: Integer
  end
end

In the example above Grape will use blank? to check whether the shelf_id param is present.

given also takes a Proc with custom code. Below, the param description is required only if the value of category is equal foo:

params do
  optional :category
  given category: ->(val) { val == 'foo' } do
    requires :description
  end
end

You can rename parameters:

params do
  optional :category, as: :type
  given category: ->(val) { val == 'foo' } do
    requires :description
  end
end

Note: given takes the name a param is declared with, not the one as renames it to. In the example, it is category, not type: the renaming applies only to declared(params), and naming type raises Grape::Exceptions::UnknownParameter.

Group Options

Parameters options can be grouped. It can be useful if you want to extract common validation or types for several parameters. Within these groups, individual parameters can extend or selectively override the common settings, allowing you to maintain the defaults at the group level while still applying parameter-specific rules where necessary.

The example below presents a typical case when parameters share common options.

params do
  requires :first_name, type: String, regexp: /\w+/, desc: 'First name', documentation: { in: 'body' }
  optional :middle_name, type: String, regexp: /\w+/, desc: 'Middle name', documentation: { in: 'body', x: { nullable: true } }
  requires :last_name, type: String, regexp: /\w+/, desc: 'Last name', documentation: { in: 'body' }
end

Grape allows you to present the same logic through the with method in your parameters block, like so:

params do
  with(type: String, regexp: /\w+/, documentation: { in: 'body' }) do
    requires :first_name, desc: 'First name'
    optional :middle_name, desc: 'Middle name', documentation: { x: { nullable: true } }
    requires :last_name, desc: 'Last name'
  end
end

You can organize settings into layers using nested with blocks. Each layer can use, add to, or change the settings of the layer above it. This helps to keep complex parameters organized and consistent, while still allowing for specific customizations to be made.

params do
  with(documentation: { in: 'body' }) do  # Applies documentation to all nested parameters
    with(type: String, regexp: /\w+/) do  # Applies type and validation to names
      requires :first_name, desc: 'First name'
      requires :last_name, desc: 'Last name'
    end
    optional :age, type: Integer, desc: 'Age', documentation: { x: { nullable: true } }  # Specific settings for 'age'
  end
end

Renaming

You can rename parameters using as, which can be useful when refactoring existing APIs:

resource :users do
  params do
    requires :email_address, as: :email
    requires :password
  end
  post do
    User.create!(declared(params)) # User takes email and password
  end
end

The value passed to as will be the key when calling declared(params).

Built-in Validators

allow_blank

Parameters can be defined as allow_blank, ensuring that they contain a value. By default, requires only validates that a parameter was sent in the request, regardless its value. With allow_blank: false, empty values or whitespace only values are invalid.

allow_blank can be combined with both requires and optional. If the parameter is required, it has to contain a value. If it's optional, it's possible to not send it in the request, but if it's being sent, it has to have some value, and not an empty string/only whitespaces.

params do
  requires :username, allow_blank: false
  optional :first_name, allow_blank: false
end

values

Parameters can be restricted to a specific set of values with the :values option.

params do
  requires :status, type: Symbol, values: [:not_started, :processing, :done]
  optional :numbers, type: Array[Integer], default: [1], values: [1, 2, 3, 5, 8]
end

Supplying a range to the :values option ensures that the parameter is (or parameters are) included in that range (using Range#include?).

params do
  requires :latitude, type: Float, values: -90.0..+90.0
  requires :longitude, type: Float, values: -180.0..+180.0
  optional :letters, type: Array[String], values: 'a'..'z'
end

Note endless ranges are also supported with ActiveSupport >= 6.0, but they require that the type be provided.

params do
  requires :minimum, type: Integer, values: (10..)
  optional :maximum, type: Integer, values: ..10
end

Note that both range endpoints have to be a #kind_of? your :type option (if you don't supply the :type option, it will be guessed to be equal to the class of the range's first endpoint). So the following is invalid:

params do
  requires :invalid1, type: Float, values: 0..10 # 0.kind_of?(Float) => false
  optional :invalid2, values: 0..10.0 # 10.0.kind_of?(0.class) => false
end

An exclusive upper bound is expressed with the standard Ruby ... range (three dots). Ruby has no range syntax for an exclusive lower bound; an arity-one :values Proc predicate (see below) is the recommended way to express one, though excluding a specific endpoint (e.g. zero) via :except_values also works:

params do
  requires :score,    type: Float, values: 0.0...5.0                    # 0.0 <= score < 5.0
  requires :positive, type: Float, values: ->(v) { v > 0.0 }            # amount > 0.0 (recommended)
  requires :other,    type: Float, values: 0.0.., except_values: [0.0]  # amount > 0.0 (alternative)
end

Combined with :type, ranges (open or closed) cover common numeric bound checks without a dedicated validator:

params do
  requires :quantity, type: Integer, values: (1..)     # must be a positive integer
  requires :discount, type: Float, values: 0.0..100.0  # must be between 0 and 100
  requires :rating,   type: Integer, values: 5..5      # must be exactly 5
  requires :numbers,  type: [Integer], values: (1..)   # every element must be positive
end

The :values option can also be supplied with a Proc, evaluated lazily with each request. If the Proc has arity zero (i.e. it takes no arguments) it is expected to return either a list or a range which will then be used to validate the parameter.

For example, given a status model you may want to restrict by hashtags that you have previously defined in the HashTag model.

params do
  requires :hashtag, type: String, values: -> { Hashtag.all.map(&:tag) }
end

Alternatively, a Proc with arity one (i.e. taking one argument) can be used to explicitly validate each parameter value. In that case, the Proc is expected to return a truthy value if the parameter value is valid. The parameter will be considered invalid if the Proc returns a falsy value or if it raises a StandardError.

params do
  requires :number, type: Integer, values: ->(v) { v.even? && v < 25 }
end

While Procs are convenient for single cases, consider using Custom Validators in cases where a validation is used more than once.

When using optional together with :values, a missing key, a nil value, and any value that coerces to nil (such as "" for type: Symbol) all pass validation — optional collapses "key may be absent" with "value may be nil". To reject blank values while still allowing the key to be absent, add allow_blank: false:

params do
  optional :state, type: Symbol, values: [:active, :inactive], allow_blank: false
end

With requires, blank values are already rejected: requires enforces presence and :values rejects nil.

except_values

Parameters can be restricted from having a specific set of values with the :except_values option.

The except_values validator behaves similarly to the values validator in that it accepts either an Array, a Range (including open/beginless/endless ranges), or a Proc. Unlike the values validator, however, except_values only accepts Procs with arity zero.

params do
  requires :browser, except_values: [ 'ie6', 'ie7', 'ie8' ]
  requires :port, type: Integer, except_values: { value: 0..1024, message: 'is not allowed' }
  requires :negative, type: Integer, except_values: ..-1
  requires :hashtag, except_values: -> { Hashtag::FORBIDDEN_LIST }
end

same_as

A same_as option can be given to ensure that values of parameters match.

params do
  requires :password
  requires :password_confirmation, same_as: :password
end

length

Parameters with types that support #length method can be restricted to have a specific length with the :length option.

The validator accepts :min or :max or both options or only :is to validate that the value of the parameter is within the given limits.

params do
  requires :code, type: String, length: { is: 2 }
  requires :str, type: String, length: { min: 3 }
  requires :list, type: [Integer], length: { min: 3, max: 5 }
  requires :hash, type: Hash, length: { max: 5 }
end

regexp

Parameters can be restricted to match a specific regular expression with the :regexp option. If the value does not match the regular expression an error will be returned. Note that this is true for both requires and optional parameters.

params do
  requires :email, regexp: /.+@.+/
end

The validator will pass if the parameter was sent without value. To ensure that the parameter contains a value, use allow_blank: false.

params do
  requires :email, allow_blank: false, regexp: /.+@.+/
end

mutually_exclusive

Parameters can be defined as mutually_exclusive, ensuring that they aren't present at the same time in a request.

params do
  optional :beer
  optional :wine
  mutually_exclusive :beer, :wine
end

Multiple sets can be defined:

params do
  optional :beer
  optional :wine
  mutually_exclusive :beer, :wine
  optional :scotch
  optional :aquavit
  mutually_exclusive :scotch, :aquavit
end

Warning: Never define mutually exclusive sets with any required params. Two mutually exclusive required params will mean params are never valid, thus making the endpoint useless. One required param mutually exclusive with an optional param will mean the latter is never valid.

exactly_one_of

Parameters can be defined as 'exactly_one_of', ensuring that exactly one parameter gets selected.

params do
  optional :beer
  optional :wine
  exactly_one_of :beer, :wine
end

Note that using :default with mutually_exclusive will cause multiple parameters to always have a default value and raise a Grape::Exceptions::Validation mutually exclusive exception.

at_least_one_of

Parameters can be defined as 'at_least_one_of', ensuring that at least one parameter gets selected.

params do
  optional :beer
  optional :wine
  optional :juice
  at_least_one_of :beer, :wine, :juice
end

all_or_none_of

Parameters can be defined as 'all_or_none_of', ensuring that all or none of parameters gets selected.

params do
  optional :beer
  optional :wine
  optional :juice
  all_or_none_of :beer, :wine, :juice
end

Nested mutually_exclusive, exactly_one_of, at_least_one_of, all_or_none_of

All of these methods can be used at any nested level.

params do
  requires :food, type: Hash do
    optional :meat
    optional :fish
    optional :rice
    at_least_one_of :meat, :fish, :rice
  end
  group :drink, type: Hash do
    optional :beer
    optional :wine
    optional :juice
    exactly_one_of :beer, :wine, :juice
  end
  optional :dessert, type: Hash do
    optional :cake
    optional :icecream
    mutually_exclusive :cake, :icecream
  end
  optional :recipe, type: Hash do
    optional :oil
    optional :meat
    all_or_none_of :oil, :meat
  end
end

Namespace Validation and Coercion

Namespaces allow parameter definitions and apply to every method within the namespace.

namespace :statuses do
  params do
    requires :user_id, type: Integer, desc: 'A user ID.'
  end
  namespace ':user_id' do
    desc "Retrieve a user's status."
    params do
      requires :status_id, type: Integer, desc: 'A status ID.'
    end
    get ':status_id' do
      User.find(params[:user_id]).statuses.find(params[:status_id])
    end
  end
end

The namespace method has a number of aliases, including: group, resource, resources, and segment. Use whichever reads the best for your API.

You can conveniently define a route parameter as a namespace using route_param.

namespace :statuses do
  route_param :id do
    desc 'Returns all replies for a status.'
    get 'replies' do
      Status.find(params[:id]).replies
    end
    desc 'Returns a status.'
    get do
      Status.find(params[:id])
    end
  end
end

You can also define a route parameter type by passing to route_param's options.

namespace :arithmetic do
  route_param :n, type: Integer do
    desc 'Returns in power'
    get 'power' do
      params[:n] ** params[:n]
    end
  end
end

Custom Validators

class AlphaNumeric < Grape::Validations::Validators::Base
  def validate_param!(attr_name, params)
    return if params[attr_name].match?(/\A[[:alnum:]]+\z/)

    validation_error!(attr_name, 'must consist of alpha-numeric characters')
  end
end

params do
  requires :text, alpha_numeric: true
end

You can also create custom classes that take parameters.

class MaxLength < Grape::Validations::Validators::Base
  def validate_param!(attr_name, params)
    return if params[attr_name].length <= @options

    validation_error!(attr_name, "must be at the most #{@options} characters long")
  end
end

params do
  requires :text, max_length: 140
end

A validator is registered under its class name, underscored and without a Validator suffix, for every API in the process. Pick a name that no built-in validator uses: a class named Length would replace the built-in length validator everywhere.

You can also create custom validation that use request to validate the attribute. For example if you want to have parameters that are available to only admins, you can do the following.

class Admin < Grape::Validations::Validators::Base
  def validate(request)
    # return if the param we are checking was not in request
    # @attrs is a list containing the attribute we are currently validating
    # in our sample case this method once will get called with
    # @attrs being [:admin_field] and once with @attrs being [:admin_false_field]
    return unless request.params.key?(@attrs.first)
    # check if admin flag is set to true
    return unless @options
    # check if user is admin or not
    # as an example get a token from request and check if it's admin or not
    validation_error!(@attrs, 'Can not set admin-only field.') unless request.headers['X-Access-Token'] == 'admin'
  end
end

And use it in your endpoint definition as:

params do
  optional :admin_field, type: String, admin: true
  optional :non_admin_field, type: String
  optional :admin_false_field, type: String, admin: false
end

Each validator is instantiated once at route definition time and frozen. Any setup (option parsing, message building) should happen in initialize, not in validate_param! or validate.

Available helpers

The following protected/private helpers are available in any Grape::Validations::Validators::Base subclass:

Helper	Description
default_message_key(key)	Class-level macro. Declares the default I18n key for validation_error!. A per-option :message override still takes precedence.
validation_error!(attr_name_or_params, message = @exception_message)	Raises Grape::Exceptions::Validation. Accepts a single attribute name or a pre-computed array of full param names.
@options	The validator option value, deep-frozen at initialization.
@attrs	Frozen array of attribute names this validator applies to.
@scope	The ParamsScope — use @scope.full_name(attr_name) for the fully-qualified param name.
option_value	Returns @options[:value] if present, otherwise @options.
options_key?(key)	Returns true if @options is a hash with a non-nil key.
hash_like?(obj)	Returns true if obj responds to key?.
scrub(value)	Returns value with invalid byte sequences scrubbed.
translate(key, **opts)	I18n lookup with :en fallback and grape.errors.messages scope. Called at request time to respect per-request locale.

Use default_message_key for a fixed I18n key. The message is resolved once at route definition time via message, so a per-option :message override still wins:

class SpecialValidator < Grape::Validations::Validators::Base
  default_message_key :special

  def validate_param!(attr_name, params)
    return if valid?(params[attr_name])

    validation_error!(attr_name)
  end
end

For interpolated messages that must respect per-request locale, call translate directly inside validate_param!:

class SpecialValidator < Grape::Validations::Validators::Base
  def validate_param!(attr_name, params)
    return if valid?(params[attr_name])

    validation_error!(attr_name, translate(:special, min: 2, max: 10))
  end
end

Validation Errors

Validation and coercion errors are collected and an exception of type Grape::Exceptions::ValidationErrors is raised. If the exception goes uncaught it will respond with a status of 400 and an error message. The validation errors are grouped by parameter name and can be accessed via Grape::Exceptions::ValidationErrors#errors.

The default response from a Grape::Exceptions::ValidationErrors is a humanly readable string, such as "beer, wine are mutually exclusive", in the following example.

params do
  optional :beer
  optional :wine
  optional :juice
  exactly_one_of :beer, :wine, :juice
end

You can rescue a Grape::Exceptions::ValidationErrors and respond with a custom response or turn the response into well-formatted JSON for a JSON API that separates individual parameters and the corresponding error messages. The following rescue_from example produces [{"params":["beer","wine"],"messages":["are mutually exclusive"]}].

format :json
subject.rescue_from Grape::Exceptions::ValidationErrors do |e|
  error! e, 400
end

Grape::Exceptions::ValidationErrors#full_messages returns the validation messages as an array. Grape::Exceptions::ValidationErrors#message joins the messages to one string.

For responding with an array of validation messages, you can use Grape::Exceptions::ValidationErrors#full_messages.

format :json
subject.rescue_from Grape::Exceptions::ValidationErrors do |e|
  error!({ messages: e.full_messages }, 400)
end

Grape returns all validation and coercion errors found by default. To skip all subsequent validation checks when a specific param is found invalid, use fail_fast: true.

The following example will not check if :wine is present unless it finds :beer.

params do
  requires :beer, fail_fast: true
  requires :wine
end

The result of empty params would be a single Grape::Exceptions::ValidationErrors error.

Similarly, no regular expression test will be performed if :blah is blank in the following example.

params do
  requires :blah, allow_blank: false, regexp: /blah/, fail_fast: true
end

I18n

Grape supports I18n for parameter-related error messages, but will fallback to English if translations for the default locale have not been provided. See en.yml for message keys.

In case your app enforces available locales only and :en is not included in your available locales, Grape cannot fall back to English and will return the translation key for the error message. To avoid this behaviour, either provide a translation for your default locale or add :en to your available locales.

Custom validators that inherit from Grape::Validations::Validators::Base have access to a translate helper (see Grape::Util::Translation) and should use it instead of calling I18n directly. It applies the same :en fallback as built-in validators, defaults scope to 'grape.errors.messages', and handles interpolation without needing format:

# Good — scope defaults to 'grape.errors.messages', interpolation forwarded automatically
translate(:special, min: 2, max: 10)

# Bad — format is unnecessary and risks conflicting with I18n reserved keys
format I18n.t(:special, scope: 'grape.errors.messages'), min: 2, max: 10

Example custom validator using an interpolated i18n message:

class SpecialValidator < Grape::Validations::Validators::Base
  def validate_param!(attr_name, params)
    return if valid?(params[attr_name])

    validation_error!(attr_name, translate(:special, min: 2, max: 10))
  end
end

Custom Validation messages

Grape supports custom validation messages for parameter-related and coerce-related error messages.

presence, allow_blank, values, regexp

params do
  requires :name, values: { value: 1..10, message: 'not in range from 1 to 10' }, allow_blank: { value: false, message: 'cannot be blank' }, regexp: { value: /^[a-z]+$/, message: 'format is invalid' }, message: 'is required'
end

same_as

params do
  requires :password
  requires :password_confirmation, same_as: { value: :password, message: 'not match' }
end

length

params do
  requires :code, type: String, length: { is: 2, message: 'code is expected to be exactly 2 characters long' }
  requires :str, type: String, length: { min: 5, message: 'str is expected to be at least 5 characters long' }
  requires :list, type: [Integer], length: { min: 2, max: 3, message: 'list is expected to have between 2 and 3 elements' }
end

all_or_none_of

params do
  optional :beer
  optional :wine
  optional :juice
  all_or_none_of :beer, :wine, :juice, message: "all params are required or none is required"
end

mutually_exclusive

params do
  optional :beer
  optional :wine
  optional :juice
  mutually_exclusive :beer, :wine, :juice, message: "are mutually exclusive cannot pass both params"
end

exactly_one_of

params do
  optional :beer
  optional :wine
  optional :juice
  exactly_one_of :beer, :wine, :juice, message: { exactly_one: "are missing, exactly one parameter is required", mutual_exclusion: "are mutually exclusive, exactly one parameter is required" }
end

at_least_one_of

params do
  optional :beer
  optional :wine
  optional :juice
  at_least_one_of :beer, :wine, :juice, message: "are missing, please specify at least one param"
end

Coerce

params do
  requires :int, type: { value: Integer, message: "type cast is invalid" }
end

With Lambdas

params do
  requires :name, values: { value: -> { (1..10).to_a }, message: 'not in range from 1 to 10' }
end

Pass symbols for i18n translations

You can pass a symbol if you want i18n translations for your custom validation messages.

params do
  requires :name, message: :name_required
end

# en.yml

en:
  grape:
    errors:
      format: ! '%{attributes} %{message}'
      messages:
        name_required: 'must be present'

Overriding Attribute Names

You can also override attribute names.

# en.yml

en:
  grape:
    errors:
      format: ! '%{attributes} %{message}'
      messages:
        name_required: 'must be present'
      attributes:
        name: 'Oops! Name'

Will produce 'Oops! Name must be present'

With Default

You cannot set a custom message option for Default as it requires interpolation %{option1}: %{value1} is incompatible with %{option2}: %{value2}. You can change the default error message for Default by changing the incompatible_option_values message key inside en.yml

params do
  optional :name, values: { value: -> { (1..10).to_a }, message: 'not in range from 1 to 10' }, default: 5
end

Using dry-validation or dry-schema

As an alternative to the params DSL described above, you can use a schema or dry-validation contract to describe an endpoint's parameters. This can be especially useful if you use the above already in some other parts of your application. If not, you'll need to add dry-validation or dry-schema to your Gemfile.

Then call contract with a contract or schema defined previously:

CreateOrdersSchema = Dry::Schema.Params do
  required(:orders).array(:hash) do
    required(:name).filled(:string)
    optional(:volume).maybe(:integer, lt?: 9)
  end
end

# ...

contract CreateOrdersSchema

or with a block, using the schema definition syntax:

contract do
  required(:orders).array(:hash) do
    required(:name).filled(:string)
    optional(:volume).maybe(:integer, lt?: 9)
  end
end

The latter will define a coercing schema (Dry::Schema.Params). When using the former approach, it's up to you to decide whether the input will need coercing.

The params and contract declarations can also be used together in the same API, e.g. to describe different parts of a nested namespace for an endpoint.

Headers

Request

Request headers are available through the headers helper or from env in their original form.

get do
  error!('Unauthorized', 401) unless headers['Secret-Password'] == 'swordfish'
end

get do
  error!('Unauthorized', 401) unless env['HTTP_SECRET_PASSWORD'] == 'swordfish'
end

Header Case Handling

The above example may have been requested as follows:

curl -H "secret_PassWord: swordfish" ...

The header name will have been normalized for you.

• In the headers helper names will be coerced into a downcased kebab case as secret-password, and looked up case-insensitively, so headers['Secret-Password'] finds it too. On Rack < 3, a standard header such as User-Agent keeps its usual capitalization.
• In the env collection they appear in all uppercase, in snake case, and prefixed with 'HTTP_' as HTTP_SECRET_PASSWORD

The header name will have been normalized per HTTP standards defined in RFC2616 Section 4.2 regardless of what is being sent by a client.

Response

You can set a response header with header inside an API.

header 'X-Robots-Tag', 'noindex'

When raising error!, pass additional headers as arguments. Additional headers will be merged with headers set before error! call.

error! 'Unauthorized', 401, 'X-Error-Detail' => 'Invalid token.'

Routes

To define routes you can use the route method or the shorthands for the HTTP verbs. To define a route that accepts any route set to :any. Parts of the path that are denoted with a colon will be interpreted as route parameters.

route :get, 'status' do
end

# is the same as

get 'status' do
end

# is the same as

get :status do
end

# is NOT the same as

get ':status' do # this makes params[:status] available
end

# This will make both params[:status_id] and params[:id] available

get 'statuses/:status_id/reviews/:id' do
end

To declare a namespace that prefixes all routes within, use the namespace method. group, resource, resources and segment are aliases to this method. Any endpoints within will share their parent context as well as any configuration done in the namespace context.

The route_param method is a convenient method for defining a parameter route segment. If you define a type, it will add a validation for this parameter.

route_param :id, type: Integer do
  get 'status' do
  end
end

# is the same as

namespace ':id' do
  params do
    requires :id, type: Integer
  end

  get 'status' do
  end
end

Optionally, you can define requirements for your named route parameters using regular expressions on namespace or endpoint. The route will match only if all requirements are met.

get ':id', requirements: { id: /[0-9]*/ } do
  Status.find(params[:id])
end

namespace :outer, requirements: { id: /[0-9]*/ } do
  get ':id' do
  end

  get ':id/edit' do
  end
end

A requirement can also name one of the capture types of Mustermann, the pattern library Grape routes with, instead of a regular expression. Besides constraining the match, a capture type that carries a converter hands the endpoint the converted value rather than a String.

get ':id', requirements: { id: Integer } do
  params[:id] # => 23, an Integer, for GET /23 — GET /michael is a 404
end

get 'events/:on', requirements: { on: :date } do
  params[:on] # => #<Date: 2020-01-02>, for GET /events/2020-01-02
end

Integer, Float, Symbol, Date and Gem::Version convert, as do their symbol spellings :integer, :float, :symbol, :date and :version. :uuid, :slug and :locale constrain the match without converting. A name Mustermann has no capture type for raises when the API is defined.

A params block still owns what the endpoint sees, since validation runs after the router: declaring requires :id, type: String alongside requirements: { id: Integer } hands the endpoint "23".

route_param names the parameter itself, so it takes the constraint on its own — route_param :id, requirements: Integer — and refuses a Hash, which would name the parameter a second time. A namespace or an endpoint can carry several captures, so requirements there are the Hash form keyed by param name.

route_param :id defines the parameter on its own — params[:id] reaches the endpoint, and the route documents it as a path capture. type: and requirements: are two ways of typing it, and they work at different layers. type: Integer declares requires :id, type: Integer: the value is coerced, appears in declared(params) and carries its type into the route's documented params, and a value the path let through but the type rejects is a 400. requirements: Integer types the route instead: a segment that does not match is not this route at all, so it is a 404, and the value arrives already converted by the router, undeclared and documented only as a capture. A declared type: Integer also narrows the path, but to digits alone, so /-7 is a 404 there while the requirement matches it. Given both, the requirement decides what the route matches and the declared type decides what the endpoint sees.

Dots in a path segment

A path capture matches [^/?#.]+: a dot introduces the format suffix, so a captured segment never contains one. A dotted value therefore either fails to route or arrives truncated, and the declared type makes no difference — type: String and type: Float compile the same pattern as an untyped parameter, only type: Integer narrows it to digits.

class WithFormat < Grape::API
  format :json
  route_param :name do
    get { params[:name] } # GET /report.pdf => 404, the segment cannot hold the dot
  end
end

class WithoutFormat < Grape::API
  route_param :name do
    # GET /report.pdf => 200, params[:name] is "report" and params[:format] is "pdf"
    get { params[:name] }
  end
end

The second one is the one to watch: the extension is taken as the format, so the endpoint is handed a truncated value and the client still gets a 200.

Percent-encoding the dot routes it through — GET /report%2Epdf hands the endpoint "report.pdf" — and so does a requirement that admits dots:

route_param :name, requirements: /.+/ do
  get { params[:name] } # => "report.pdf" for GET /report.pdf, "1.2.3" for GET /1.2.3
end

Such a regexp claims the extension as well, so GET /a.b.json hands the endpoint "a.b.json". The response format is still read off the end of the path, though, whatever the route captured: with no format of its own, the API above answers GET /a.b.json in JSON, and GET /a.b.xml with a 500, since the XML formatter cannot render a String. Declaring format settles the response format whatever the extension. A narrower constraint leaves the suffix alone: get ':n', requirements: { n: Float } matches /4.2 and still reads .json as the format.

The QUERY Method

Grape supports QUERY, a safe and idempotent method that carries its query in the request content rather than in the URI. It is the method to reach for when a query is too large, too structured, or too sensitive to encode into a query string.

params do
  requires :q, type: String
  optional :limit, type: Integer
end
query '/search' do
  Status.search(params[:q], limit: params[:limit])
end

curl -X QUERY 'http://localhost:9292/search' -H Content-Type:application/json -d '{"q": "grape", "limit": 10}'

The content is parsed into params exactly as it is for POST, and a successful response defaults to 200, not 201. Because the content is the query, a QUERY request that arrives without a Content-Type is rejected with 400 instead of falling back to the API's default format.

Helpers

You can define helper methods that your endpoints can use with the helpers macro by either giving a block or an array of modules.

module StatusHelpers
  def user_info(user)
    "#{user} has statused #{user.statuses} status(s)"
  end
end

module HttpCodesHelpers
  def unauthorized
    401
  end
end

class API < Grape::API
  # define helpers with a block
  helpers do
    def current_user
      User.find(params[:user_id])
    end
  end

  # or mix in an array of modules
  helpers StatusHelpers, HttpCodesHelpers

  before do
    error!('Access Denied', unauthorized) unless current_user
  end

  get 'info' do
    # helpers available in your endpoint and filters
    user_info(current_user)
  end
end

You can define reusable params using helpers.

class API < Grape::API
  helpers do
    params :pagination do
      optional :page, type: Integer
      optional :per_page, type: Integer
    end
  end

  desc 'Get collection'
  params do
    use :pagination # aliases: includes, use_scope
  end
  get do
    Collection.page(params[:page]).per(params[:per_page])
  end
end

You can also define reusable params using shared helpers.

module SharedParams
  extend Grape::API::Helpers

  params :period do
    optional :start_date
    optional :end_date
  end

  params :pagination do
    optional :page, type: Integer
    optional :per_page, type: Integer
  end
end

class API < Grape::API
  helpers SharedParams

  desc 'Get collection.'
  params do
    use :period, :pagination
  end

  get do
    Collection
      .from(params[:start_date])
      .to(params[:end_date])
      .page(params[:page])
      .per(params[:per_page])
  end
end

Helpers support blocks that can help set default values. The following API can return a collection sorted by id or created_at in asc or desc order.

module SharedParams
  extend Grape::API::Helpers

  params :order do |options|
    optional :order_by, type: Symbol, values: options[:order_by], default: options[:default_order_by]
    optional :order, type: Symbol, values: %i(asc desc), default: options[:default_order]
  end
end

class API < Grape::API
  helpers SharedParams

  desc 'Get a sorted collection.'
  params do
    use :order, order_by: %i(id created_at), default_order_by: :created_at, default_order: :asc
  end

  get do
    Collection.send(params[:order], params[:order_by])
  end
end

Path Helpers

If you need methods for generating paths inside your endpoints, please see the grape-route-helpers gem.

Parameter Documentation

You can attach additional documentation to params using a documentation hash.

params do
  optional :first_name, type: String, documentation: { example: 'Jim' }
  requires :last_name, type: String, documentation: { example: 'Smith' }
end

If documentation isn't needed (for instance, it is an internal API), documentation can be disabled.

class API < Grape::API
  do_not_document!

  # endpoints...
end

In this case, Grape won't create objects related to documentation which are retained in RAM forever.

Cookies

You can set, get and delete your cookies very simply using cookies method.

class API < Grape::API
  get 'status_count' do
    # a cookie the client sent back is a String, so convert before counting
    cookies[:status_count] = cookies[:status_count].to_i + 1
    { status_count: cookies[:status_count] }
  end

  delete 'status_count' do
    status_count = cookies[:status_count]
    cookies.delete(:status_count)
    { status_count: status_count }
  end
end

Use a hash-based syntax to set more than one value.

cookies[:status_count] = {
  value: 0,
  expires: Time.tomorrow,
  domain: '.twitter.com',
  path: '/'
}

cookies[:status_count][:value] +=1

Delete a cookie with delete.

cookies.delete :status_count

Specify an optional path.

cookies.delete :status_count, path: '/'

HTTP Status Code

By default Grape returns a 201 for POST-Requests, 204 for DELETE-Requests that don't return any content, and 200 status code for all other Requests. You can use status to query and set the actual HTTP Status Code

post do
  status 202

  if status == 200
     # do some thing
  end
end

You can also use one of status codes symbols that are provided by Rack utils

post do
  status :no_content
end

Redirecting

You can redirect to a new url temporarily (302) or permanently (301).

redirect '/statuses'

redirect '/statuses', permanent: true

Recognizing Path

You can recognize the endpoint matched with given path.

This API returns an instance of Grape::Endpoint.

class API < Grape::API
  get '/statuses' do
  end
end

API.recognize_path '/statuses'

Since version 2.1.0, the recognize_path method takes into account the parameters type to determine which endpoint should match with given path.

class Books < Grape::API
  resource :books do
    route_param :id, type: Integer do
      # GET /books/:id
      get do
        #...
      end
    end

    resource :share do
      # POST /books/share
      post do
      # ....
      end
    end
  end
end

API.recognize_path '/books/1' # => /books/:id
API.recognize_path '/books/share' # => /books/share
API.recognize_path '/books/other' # => nil

Allowed Methods

When you add a GET route for a resource, a route for the HEAD method will also be added automatically. You can disable this behavior with do_not_route_head!.

class API < Grape::API
  do_not_route_head!

  get '/example' do
    # only responds to GET
  end
end

When you add a route for a resource, a route for the OPTIONS method will also be added. The response to an OPTIONS request will include an "Allow" header listing the supported methods. If the resource has before, after and finally callbacks they will be executed, but no other callbacks will run.

class API < Grape::API
  get '/rt_count' do
    { rt_count: current_user.rt_count }
  end

  params do
    requires :value, type: Integer, desc: 'Value to add to the rt count.'
  end
  put '/rt_count' do
    current_user.rt_count += params[:value].to_i
    { rt_count: current_user.rt_count }
  end
end

curl -v -X OPTIONS http://localhost:3000/rt_count

> OPTIONS /rt_count HTTP/1.1
>
< HTTP/1.1 204 No Content
< Allow: OPTIONS, GET, PUT, HEAD

You can disable this behavior with do_not_route_options!.

If a request for a resource is made with an unsupported HTTP method, an HTTP 405 (Method Not Allowed) response will be returned. If the resource has before and finally callbacks they will be executed, but no other callbacks will run.

curl -X DELETE -v http://localhost:3000/rt_count/

> DELETE /rt_count/ HTTP/1.1
> Host: localhost:3000
>
< HTTP/1.1 405 Method Not Allowed
< Allow: OPTIONS, GET, PUT, HEAD

Raising Exceptions

You can abort the execution of an API method by raising errors with error!.

error! 'Access Denied', 401

Anything that responds to #to_s can be given as a first argument to error!.

error! :not_found, 404

You can also return JSON formatted objects by raising error! and passing a hash instead of a message.

error!({ error: 'unexpected error', detail: 'missing widget' }, 500)

You can set additional headers for the response. They will be merged with headers set before error! call.

error!('Something went wrong', 500, 'X-Error-Detail' => 'Invalid token.')

error! raises a Grape::Exceptions::Halt, which Grape renders as the response without handing it to rescue_from. Since it is an exception, a database transaction it leaves is rolled back. It also reaches Grape from a fiber or thread whose result the route waits for, such as an Async task.

Order.transaction do
  order = Order.create!(declared(params))
  error!('Out of stock', 409) unless order.product.in_stock? # the order is rolled back
end

It is a StandardError, so a rescue => e around error! catches it as well. Rescue Grape::Exceptions::Halt first and re-raise it to let it through.

You can present documented errors with a Grape entity using the grape-entity gem.

module API
  class Error < Grape::Entity
    expose :code
    expose :message
  end
end

The following example specifies the entity to use in the failure definition.

desc 'My Route' do
 failure [[401, 'Unauthorized', API::Error]]
end
error!({ message: 'Unauthorized' }, 401)

The following example specifies the presented entity explicitly in the error message.

desc 'My Route' do
 failure [[401, 'Unauthorized']]
end
error!({ message: 'Unauthorized', with: API::Error }, 401)

Default Error HTTP Status Code

By default Grape returns a 500 status code from error!. You can change this with default_error_status.

class API < Grape::API
  default_error_status 400
  get '/example' do
    error! 'This should have http status code 400'
  end
end

Handling 404

For Grape to handle all the 404s for your API, it can be useful to use a catch-all. In its simplest form, it can be like:

route :any, '*path' do
  error!('Not Found', 404) # or something else
end

It is very crucial to define this endpoint at the very end of your API, as it literally accepts every request.

Exception Handling

Grape can be told to rescue all StandardError exceptions and return them in the API format.

class Twitter::API < Grape::API
  rescue_from :all
end

This mimics default rescue behaviour when an exception type is not provided. Any other exception should be rescued explicitly, see below.

Grape can also rescue from all exceptions and still use the built-in exception handing. This will give the same behavior as rescue_from :all with the addition that Grape will use the exception handling defined by all Exception classes that inherit Grape::Exceptions::Base.

The intent of this setting is to provide a simple way to cover the most common exceptions and return any unexpected exceptions in the API format.

class Twitter::API < Grape::API
  rescue_from :grape_exceptions
end

If you want to customize the shape of grape exceptions returned to the user, to match your :all handler for example, you can pass a block to rescue_from :grape_exceptions.

rescue_from :grape_exceptions do |e|
  error!(e, e.status)
end

The opt-in takes precedence over a catch-all handler, whether that catch-all is written as rescue_from :all or as a class such as rescue_from StandardError. So a validation failure stays a 400 rather than becoming whatever the catch-all returns:

class Twitter::API < Grape::API
  rescue_from StandardError do
    error!('server error', 500)
  end
  rescue_from :grape_exceptions   # validation errors still answer 400
end

A handler registered for a specific Grape exception class is more precise than the opt-in and still wins, so rescue_from Grape::Exceptions::ValidationErrors keeps its own handler. The opt-in only outranks handlers that matched through a non-Grape ancestor.

You can also rescue specific exceptions.

class Twitter::API < Grape::API
  rescue_from ArgumentError, UserDefinedError
end

A class listed explicitly is rescued whether or not it inherits from StandardError; see Exceptions that should be rescued explicitly.

When several classes could match, the one registered first in a scope wins — as with the clauses of a Ruby rescue. Register the more specific class before the broader one, or the narrower handler never runs:

class Twitter::API < Grape::API
  rescue_from ArgumentError do ... end   # matched first
  rescue_from StandardError do ... end   # everything else
end

Grape warns when a rescue_from is registered for a class an earlier one in the same scope already covers. This is about ordering within a scope; a handler in a nested namespace or a mounted API always takes precedence over one inherited from an enclosing scope, whatever the classes are.

A rescue_from also has to be declared above the routes it should cover. A route captures the handlers registered before it, so one declared below a route never runs for it:

class Twitter::API < Grape::API
  rescue_from :all do ... end   # covers GET /statuses
  get :statuses do ... end
  rescue_from ArgumentError do ... end   # never runs for GET /statuses
end

The same goes for use, helpers and the before/after callbacks.

Notice that you could combine these two approaches (rescuing custom errors takes precedence). For example, it's useful for handling all exceptions except Grape validation errors.

class Twitter::API < Grape::API
  rescue_from Grape::Exceptions::ValidationErrors do |e|
    error!(e, 400)
  end

  rescue_from :all
end

The error format will match the request format. See "Content-Types" below.

Custom error formatters for existing and additional types can be defined with a proc. The formatter receives a frozen Grape::Exceptions::ErrorResponse as error: (read its status, message, headers, backtrace and original_exception), plus three context kwargs — env:, include_backtrace:, include_original_exception:. Pull just the keys you need with ** to ignore the rest:

class Twitter::API < Grape::API
  error_formatter :txt, ->(error:, **) {
    "error #{error.status}: #{error.message} from #{error.backtrace}"
  }
end

You can also use a module or class.

module CustomFormatter
  def self.call(error:, **)
    { status: error.status, message: error.message, backtrace: error.backtrace }.to_json
  end
end

class Twitter::API < Grape::API
  error_formatter :custom, CustomFormatter
end

You can rescue all exceptions with a code block. The error! wrapper automatically sets the default error code and content-type.

class Twitter::API < Grape::API
  rescue_from :all do |e|
    error!("rescued from #{e.class.name}")
  end
end

Optionally, you can set the format, status code and headers.

class Twitter::API < Grape::API
  format :json
  rescue_from :all do |e|
    error!({ error: 'Server error.' }, 500, { 'Content-Type' => 'text/error' })
  end
end

Re-raising from inside a rescue_from block

A rescue_from block can re-raise an exception to invoke a different handler. This is useful for translating one exception class into another:

class Twitter::API < Grape::API
  rescue_from Grape::Exceptions::ValidationErrors do |e|
    raise Api::Exceptions::InvalidValueError, e.full_messages
  end

  rescue_from Api::Exceptions::InvalidValueError do |e|
    error!({ errors: e.message }, 422)
  end
end

The first handler re-raises; the second handler runs against the new exception.

If the re-raised exception has no registered rescue_from and is a Grape::Exceptions::Base subclass, it is rendered through the default Grape error path (using its own status and message). Anything else — typos, NoMethodError, an unrelated StandardError — is treated as an internal error: it is exposed on the rack env for upstream Rack middleware to observe, and rendered to the API consumer as a generic 500 Internal Server Error. This avoids leaking internal detail in the response body.

Because the exception is answered rather than raised, nothing above Grape catches it, so it is published under rack.exception — the key error trackers read for a handled exception — as well as Grape's own grape.exception. Grape does no logging of its own here; register a rescue_from :internal_grape_exceptions handler to take that over.

You can take control of the internal-error path by opting in with rescue_from :internal_grape_exceptions:

class Twitter::API < Grape::API
  rescue_from :internal_grape_exceptions do |e|
    Sentry.capture_exception(e)
    error!({ message: 'Something went wrong' }, 500)
  end
end

When this handler is registered the framework hands the original exception to you, and you own the response shape and any logging. The framework deliberately does not log internal errors itself — it has no way to know your preferred format or destination.

A second raise inside the redispatched handler is not redispatched again — it goes straight to the framework's generic 500. This bounds the chain at one redispatch and prevents loops.

To have an exception raised inside a rescue_from block propagate out of the middleware stack instead, so that Rack middleware above Grape can act on it:

Grape.configure do |config|
  config.raise_rescue_handler_errors = true
end

With this on the exception is re-raised untouched: it is not redispatched, so rescue_from :internal_grape_exceptions never sees it and neither rack.exception nor grape.exception is set. A block that calls error! still renders as usual.

You can also rescue all exceptions with a code block and handle the Rack response at the lowest level.

class Twitter::API < Grape::API
  rescue_from :all do |e|
    Rack::Response.new([ e.message ], 500, { 'Content-type' => 'text/error' })
  end
end

Or rescue specific exceptions.

class Twitter::API < Grape::API
  rescue_from ArgumentError do |e|
    error!("ArgumentError: #{e.message}")
  end

  rescue_from NoMethodError do |e|
    error!("NoMethodError: #{e.message}")
  end
end

By default, rescue_from will rescue the exceptions listed and all their subclasses.

Assume you have the following exception classes defined.

module APIErrors
  class ParentError < StandardError; end
  class ChildError < ParentError; end
end

Then the following rescue_from clause will rescue exceptions of type APIErrors::ParentError and its subclasses (in this case APIErrors::ChildError).

rescue_from APIErrors::ParentError do |e|
    error!({
      error: "#{e.class} error",
      message: e.message
    }, 500)
end

To only rescue the base exception class, set rescue_subclasses: false. The code below will rescue exceptions of type RuntimeError but not its subclasses.

rescue_from RuntimeError, rescue_subclasses: false do |e|
    error!({
      message: e.message
    }, 500)
end

Helpers are also available inside rescue_from.

class Twitter::API < Grape::API
  format :json
  helpers do
    def server_error!
      error!({ error: 'Server error.' }, 500, { 'Content-Type' => 'text/error' })
    end
  end

  rescue_from :all do |e|
    server_error!
  end
end

The rescue_from handler must return a Rack::Response object, call error!, or raise an exception (either the original exception or another custom one). An exception raised in rescue_from is dispatched to the handler registered for its class; when there is none, Grape answers it itself, as described in Re-raising from inside a rescue_from block.

Alternately, use the with option in rescue_from to specify a method or a proc.

class Twitter::API < Grape::API
  format :json
  helpers do
    def server_error!
      error!({ error: 'Server error.' }, 500, { 'Content-Type' => 'text/error' })
    end
  end

  rescue_from :all,          with: :server_error!
  rescue_from ArgumentError, with: -> { Rack::Response.new('rescued with a method', 400) }
end

Inside the rescue_from block, the environment of the original controller method(.self receiver) is accessible through the #context method.

class Twitter::API < Grape::API
  rescue_from :all do |e|
    user_id = context.params[:user_id]
    error!("error for #{user_id}")
  end
end

Rescuing exceptions inside namespaces

You could put rescue_from clauses inside a namespace and they will take precedence over ones defined in the root scope:

class Twitter::API < Grape::API
  rescue_from ArgumentError do |e|
    error!("outer")
  end

  namespace :statuses do
    rescue_from ArgumentError do |e|
      error!("inner")
    end
    get do
      raise ArgumentError.new
    end
  end
end

Here 'inner' will be result of handling occurred ArgumentError.

Unrescuable Exceptions

Grape::Exceptions::InvalidVersionHeader, which is raised when the version in the request header doesn't match the currently evaluated version for the endpoint, will never be rescued from a rescue_from block (even a rescue_from :all) This is because Grape relies on Rack to catch that error and try the next versioned-route for cases where there exist identical Grape endpoints with different versions.

Exceptions that should be rescued explicitly

Any exception that is not subclass of StandardError should be rescued explicitly. Usually it is not a case for an application logic as such errors point to problems in Ruby runtime. This is following standard recommendations for exceptions handling.

When the error response itself cannot be rendered

A rescue_from handler can build a payload its error formatter cannot serialize. The common case is echoing request-derived text back to the client:

class Missing < StandardError; end

class API < Grape::API
  format :json

  rescue_from(Missing) { |e| error!({ error: 'not_found', detail: e.message }, 404) }

  route_param(:id) { get { raise Missing, "no such thing: #{params[:id]}" } }
end

GET /%C3%28 puts an invalid UTF-8 byte in params[:id], and the JSON formatter raises when it reaches that byte in the message — after the handler has already returned, so rescue_from :all does not help.

Rather than let that exception escape the middleware stack, Grape answers 500: first retrying the API's own format with the framework's Internal Server Error message, then falling back to a bare text/plain body if even that cannot be rendered.

The exception is published on the rack env under rack.exception — the key error trackers read to find an exception that was handled rather than raised — and on Grape's own grape.exception. It is also written to rack.errors, so it reaches the log with no tracker installed.

To opt out and have the exception propagate out of the middleware stack instead:

Grape.configure do |config|
  config.raise_rendering_errors = true
end

With this on the failsafe never runs: the exception is re-raised untouched, so neither env key is set and nothing is written to rack.errors.

Logging

Grape::API provides a logger method which by default will return an instance of the Logger class from Ruby's standard library.

To log messages from within an endpoint, you need to define a helper to make the logger available in the endpoint context.

class API < Grape::API
  helpers do
    def logger
      API.logger
    end
  end
  post '/statuses' do
    logger.info "#{current_user} has statused"
  end
end

To change the logger level.

class API < Grape::API
  self.logger.level = Logger::INFO
end

You can also set your own logger.

class MyLogger
  def warning(message)
    puts "this is a warning: #{message}"
  end
end

class API < Grape::API
  logger MyLogger.new
  helpers do
    def logger
      API.logger
    end
  end
  get '/statuses' do
    logger.warning "#{current_user} has statused"
  end
end

A logger belongs to the API it is set on, and to the APIs mounted under it that set none of their own. An API that sets none, and is mounted under none that does, logs to a default shared by every such API.

For similar to Rails request logging try the grape_logging or grape-middleware-logger gems.

API Formats

Your API can declare which content-types to support by using content_type. If you do not specify any, Grape will support XML, JSON, BINARY, and TXT content-types. The default format is :txt; you can change this with default_format. Essentially, the two APIs below are equivalent.

class Twitter::API < Grape::API
  # no content_type declarations, so Grape uses the defaults
end

class Twitter::API < Grape::API
  # the following declarations are equivalent to the defaults

  content_type :xml, 'application/xml'
  content_type :json, 'application/json'
  content_type :binary, 'application/octet-stream'
  content_type :txt, 'text/plain'

  default_format :txt
end

If you declare any content_type whatsoever, the Grape defaults will be overridden. For example, the following API will only support the :xml and :rss content-types, but not :txt, :json, or :binary. Importantly, this means the :txt default format is not supported! So, make sure to set a new default_format.

class Twitter::API < Grape::API
  content_type :xml, 'application/xml'
  content_type :rss, 'application/rss+xml'

  default_format :xml
end

Serialization takes place automatically. For example, you do not have to call to_json in each JSON API endpoint implementation. The response format (and thus the automatic serialization) is determined in the following order:

• Use the file extension, if specified. If the file is .json, choose the JSON format.
• Use the value of the format parameter in the query string, if specified.
• Use the format set by the format option, if specified.
• Attempt to find an acceptable format from the Accept header.
• Use the default format, if specified by the default_format option.
• Default to :txt.

For example, consider the following API.

class MultipleFormatAPI < Grape::API
  content_type :xml, 'application/xml'
  content_type :json, 'application/json'

  default_format :json

  get :hello do
    { hello: 'world' }
  end
end

• GET /hello (with an Accept: */* header) does not have an extension or a format parameter, so it will respond with JSON (the default format).
• GET /hello.xml has a recognized extension, so it will respond with XML.
• GET /hello?format=xml has a recognized format parameter, so it will respond with XML.
• GET /hello.xml?format=json has a recognized extension (which takes precedence over the format parameter), so it will respond with XML.
• GET /hello.xls (with an Accept: */* header) has an extension, but that extension is not recognized, so it will respond with JSON (the default format).
• GET /hello.xls with an Accept: application/xml header has an unrecognized extension, but the Accept header corresponds to a recognized format, so it will respond with XML.
• GET /hello.xls with an Accept: text/plain header has an unrecognized extension and an unrecognized Accept header, so it will respond with JSON (the default format).

You can override this process explicitly by calling api_format in the API itself. For example, the following API will let you upload arbitrary files and return their contents as an attachment with the correct MIME type.

class Twitter::API < Grape::API
  post 'attachment' do
    filename = params[:file][:filename]
    content_type MIME::Types.type_for(filename)[0].to_s
    api_format :binary # there's no formatter for :binary, data will be returned "as is"
    header 'Content-Disposition', "attachment; filename*=UTF-8''#{CGI.escape(filename)}"
    params[:file][:tempfile].read
  end
end

You can have your API only respond to a single format with format. If you use this, the API will not respond to file extensions other than specified in format. For example, consider the following API.

class SingleFormatAPI < Grape::API
  format :json

  get :hello do
    { hello: 'world' }
  end
end

• GET /hello will respond with JSON.
• GET /hello.json will respond with JSON.
• GET /hello.xml, GET /hello.foobar, or any other extension will respond with an HTTP 404 error code.
• GET /hello?format=xml will respond with an HTTP 406 error code, because the XML format specified by the request parameter is not supported.
• GET /hello with an Accept: application/xml header will still respond with JSON, since it could not negotiate a recognized content-type from the headers and JSON is the effective default.

The formats apply to parsing, too. The following API will only respond to the JSON content-type and will not parse any other input than application/json, application/x-www-form-urlencoded, multipart/form-data, multipart/related and multipart/mixed. All other requests will fail with an HTTP 415 error code.

class Twitter::API < Grape::API
  format :json
end

When the content-type is omitted, Grape will return a 415 error code unless default_format is specified. The following API will try to parse any data without a content-type using a JSON parser.

class Twitter::API < Grape::API
  format :json
  default_format :json
end

If you combine format with rescue_from :all, errors will be rendered using the same format. default_error_formatter renders the errors of a format that has no error formatter of its own, such as one you declared yourself; it does not replace the error formatter of a built-in format like :json. To change how errors of a built-in format are rendered, register one for it with error_formatter.

class Twitter::API < Grape::API
  content_type :xls, 'application/vnd.ms-excel'
  format :xls

  default_error_formatter :json # :xls has no error formatter of its own
end

Custom formatters for existing and additional types can be defined with a proc.

class Twitter::API < Grape::API
  content_type :xls, 'application/vnd.ms-excel'
  formatter :xls, ->(object, env) { object.to_xls }
end

You can also use a module or class.

module XlsFormatter
  def self.call(object, env)
    object.to_xls
  end
end

class Twitter::API < Grape::API
  content_type :xls, 'application/vnd.ms-excel'
  formatter :xls, XlsFormatter
end

Built-in formatters are the following.

• :json: use object's to_json when available, otherwise call MultiJson.dump
• :xml: use object's to_xml when available, usually via MultiXml
• :txt: use object's to_txt when available, otherwise to_s
• :serializable_hash: use object's serializable_hash when available, otherwise fallback to :json
• :binary: data will be returned "as is"
• :jsonl: one line per element of an Array or Enumerator, otherwise one line, each rendered as :json renders it. Not a default content type, see JSON Lines
• :md, :markdown: use object's to_markdown when available, return a String "as is", otherwise raise Grape::Exceptions::InvalidFormatter. Not default content types, see Serving Markdown

Serving Markdown

:md and :markdown are not among the default content types. Declare either one to answer clients that ask for Markdown, such as LLM agents sending Accept: text/markdown.

class API < Grape::API
  content_type :json, 'application/json'
  content_type :md, 'text/markdown; charset=utf-8'
  default_format :json

  get '/pages/:id' do
    Page.find(params[:id]) # renders with Page#to_markdown
  end
end

• GET /pages/1.md, GET /pages/1.markdown, GET /pages/1?format=md, ?format=markdown and GET /pages/1 with an Accept: text/markdown header answer text/markdown.
• :md and :markdown name the same format: declaring either one answers both extensions, under the name declared. A request for .markdown on the API above negotiates :md, so a formatter :md of your own renders it too.
• to_markdown is the method Rails' render markdown: calls (Rails 8.1), so an object written for a Rails controller renders the same way in Grape.
• Grape does not convert anything else to Markdown. When the client negotiated Markdown, a body without to_markdown that is not a String answers in the default format instead; when the API pins format :md, it answers 500.
• RFC 7763 makes charset a required parameter of text/markdown, hence charset=utf-8.
• Errors are rendered as plain text, which is valid Markdown.

Serving pre-rendered JSON

The :json and :serializable_hash formatters encode whatever the endpoint returned, so a String that already holds JSON gets encoded a second time. Wrap the body in Grape::PrecompiledJson to opt that response out of encoding.

class API < Grape::API
  format :json

  get '/cached' do
    body Grape::PrecompiledJson.new(Rails.cache.read('payload')) # => '{"id":1}', served as-is
  end
end

An Array is joined into a JSON array without its members being parsed, so a cached collection is spliced together rather than round-tripped. Members may themselves be Grape::PrecompiledJson instances.

Grape::PrecompiledJson.new(['{"id":1}', '{"id":2}']).to_s # => '[{"id":1},{"id":2}]'

Bodies that are not wrapped keep encoding exactly as before, so this is opt-in per response rather than a mode the whole API is in. Wrap only the whole body: a wrapper nested inside a Hash or Array that is then handed to an encoder is serialized as an ordinary object, because no encoder knows to unwrap it. A value that is neither a String nor an Array raises Grape::Exceptions::InvalidFormatter and answers 500, rather than handing Rack a body it cannot serve. Nothing checks that the String actually holds JSON — that is the caller's responsibility.

Errors are unaffected: they are rendered by the error formatter, so error! still returns properly encoded JSON.

If a body is present in a request to an API, with a Content-Type header value that is of an unsupported type a "415 Unsupported Media Type" error code will be returned by Grape.

Response statuses that indicate no content as defined by Rack here will bypass serialization and the body entity - though there should be none - will not be modified.

JSON Lines

:jsonl serves JSON Lines: one JSON value per line, each line ended by a newline. It is not among the default content types, so declare it, as application/jsonl, or application/x-ndjson if your clients expect that one.

class API < Grape::API
  content_type :jsonl, 'application/jsonl'
  format :jsonl

  get :users do
    User.all # {"id":1,"name":"Ada"}\n{"id":2,"name":"Grace"}\n
  end
end

An Array renders one line per element and any other object one line, each as the :json formatter renders it, so a Grape::PrecompiledJson element is served as its own line. A String is served as it is, taken to be JSON Lines already.

An Enumerator is streamed: each element is rendered only as the server reads the body, so a large export is never held in memory whole.

get :export do
  User.find_each.lazy.map { |user| user.slice(:id, :name) }
end

The status and headers of a streamed response are sent before its first line is rendered, so an exception raised while it streams is past rescue_from: the response just ends early.

A request body of application/jsonl is parsed into an Array of its lines' values, skipping blank lines. Like a JSON body that is an Array, it is available in env['api.request.body'] and is not merged into params. To validate the lines as params, wrap them in a Hash with a parser of your own.

parser :jsonl, ->(body, env) { { items: Grape::Parser::Jsonl.call(body, env) } }

params do
  requires :items, type: Array do
    requires :name, type: String
  end
end
post :import do
  declared(params)[:items]
end

An error renders as a single JSON line.

JSONP

Grape supports JSONP via Rack::JSONP, part of the rack-contrib gem. Add rack-contrib to your Gemfile.

require 'rack/contrib'

class API < Grape::API
  use Rack::JSONP
  format :json
  get '/' do
    'Hello World'
  end
end

CORS

Grape supports CORS via Rack::CORS, part of the rack-cors gem. Add rack-cors to your Gemfile, then use the middleware in your config.ru file.

require 'rack/cors'

use Rack::Cors do
  allow do
    origins '*'
    resource '*', headers: :any, methods: :get
  end
end

run Twitter::API

Content-type

Content-type is set by the formatter. You can override the content-type of the response at runtime by setting the Content-Type header.

class API < Grape::API
  get '/home_timeline_js' do
    content_type 'application/javascript'
    "var statuses = ...;"
  end
end

API Data Formats

Grape accepts and parses input data sent with the POST, PUT, PATCH, DELETE and QUERY methods as described in the Parameters section above. It also supports custom data formats. You must declare additional content-types via content_type and optionally supply a parser via parser unless a parser is already available within Grape to enable a custom format. Such a parser can be a function or a class.

With a parser, parsed data is available "as-is" in env['api.request.body']. Without a parser, data is available "as-is" and in env['api.request.input'].

The following example is a trivial parser that will assign any input with the "text/custom" content-type to :value. The parameter will be available via params[:value] inside the API call.

module CustomParser
  def self.call(object, env)
    { value: object.to_s }
  end
end

content_type :txt, 'text/plain'
content_type :custom, 'text/custom'
parser :custom, CustomParser

put 'value' do
  params[:value]
end

You can invoke the above API as follows.

curl -X PUT -d 'data' 'http://localhost:9292/value' -H Content-Type:text/custom -v

You can disable parsing for a content-type with nil. For example, parser :json, nil will disable JSON parsing altogether. The request data is then available as-is in env['api.request.body'].

JSON and XML Processors

Grape uses JSON and ActiveSupport::XmlMini for JSON and XML parsing by default. It also detects and supports multi_json and multi_xml. Adding those gems to your Gemfile and requiring them will enable them and allow you to swap the JSON and XML back-ends.

RESTful Model Representations

Grape supports a range of ways to present your data with some help from a generic present method, which accepts two arguments: the object to be presented and the options associated with it. The options hash may include :with, which defines the entity to expose.

Grape Entities

Add the grape-entity gem to your Gemfile. Please refer to the grape-entity documentation for more details.

The following example exposes statuses.

module API
  module Entities
    class Status < Grape::Entity
      expose :user_name
      expose :text, documentation: { type: 'string', desc: 'Status update text.' }
      expose :ip, if: { type: :full }
      expose :user_type, :user_id, if: ->(status, options) { status.user.public? }
      expose :digest do |status, options|
        Digest::MD5.hexdigest(status.text)
      end
      expose :replies, using: API::Entities::Status, as: :replies
    end
  end

  class Statuses < Grape::API
    version 'v1'

    desc 'Statuses index' do
      params API::Entities::Status.documentation
    end
    get '/statuses' do
      statuses = Status.all
      type = current_user.admin? ? :full : :default
      present statuses, with: API::Entities::Status, type: type
    end
  end
end

You can use entity documentation directly in the params block with using: Entity.documentation.

module API
  class Statuses < Grape::API
    version 'v1'

    desc 'Create a status'
    params do
      requires :all, except: [:ip], using: API::Entities::Status.documentation.except(:id)
    end
    post '/status' do
      Status.create! params
    end
  end
end

You can present with multiple entities using an optional Symbol argument.

  get '/statuses' do
    statuses = Status.all.page(1).per(20)
    present :total_page, 10
    present :per_page, 20
    present :statuses, statuses, with: API::Entities::Status
  end

The response will be

  {
    total_page: 10,
    per_page: 20,
    statuses: []
  }

In addition to separately organizing entities, it may be useful to put them as namespaced classes underneath the model they represent.

class Status
  def entity
    Entity.new(self)
  end

  class Entity < Grape::Entity
    expose :text, :user_id
  end
end

If you organize your entities this way, Grape will automatically detect the Entity class and use it to present your models. In this example, if you added present Status.new to your endpoint, Grape will automatically detect that there is a Status::Entity class and use that as the representative entity. This can still be overridden by using the :with option or an explicit represents call.

You can present hash with Grape::Presenters::Presenter to keep things consistent.

get '/users' do
  present({ id: 10, name: :dgz }, with: Grape::Presenters::Presenter)
end

The response will be

{
  id:   10,
  name: 'dgz'
}

It has the same result with

get '/users' do
  present :id, 10
  present :name, :dgz
end

Hypermedia and Roar

You can use Roar to render HAL or Collection+JSON with the help of grape-roar, which defines a custom JSON formatter and enables presenting entities with Grape's present keyword.

Rabl

You can use Rabl templates with the help of the grape-rabl gem, which defines a custom Grape Rabl formatter.

Active Model Serializers

You can use Active Model Serializers serializers with the help of the grape-active_model_serializers gem, which defines a custom Grape AMS formatter.

Sending Raw or No Data

In general, use the binary format to send raw data.

class API < Grape::API
  get '/file' do
    content_type 'application/octet-stream'
    File.binread 'file.bin'
  end
end

You can set the response body explicitly with body.

class API < Grape::API
  get '/' do
    content_type 'text/plain'
    body 'Hello World'
    # return value ignored
  end
end

Use body false to return 204 No Content without any data or content-type.

If you want to empty the body with an HTTP status code other than 204 No Content, you can override the status code after specifying body false as follows

class API < Grape::API
  get '/' do
    body false
    status 304
  end
end

You can also set the response to a file with sendfile. This works with the Rack::Sendfile middleware to optimally send the file through your web server software.

class API < Grape::API
  get '/' do
    sendfile '/path/to/file'
  end
end

To stream a file in chunks use stream

class API < Grape::API
  get '/' do
    stream '/path/to/file'
  end
end

If you want to stream non-file data use the stream method and a Stream object. This is an object that responds to each and yields for each chunk to send to the client. Each chunk will be sent as it is yielded instead of waiting for all of the content to be available.

class MyStream
  def each
    yield 'part 1'
    yield 'part 2'
    yield 'part 3'
  end
end

class API < Grape::API
  get '/' do
    stream MyStream.new
  end
end

Authentication

Basic Auth

Grape has built-in Basic authentication (the given block is executed in the context of the current Endpoint). Authentication applies to the current namespace and any children, but not parents.

http_basic do |username, password|
  # verify user's password here
  # IMPORTANT: make sure you use a comparison method which isn't prone to a timing attack
end

Register custom middleware for authentication

Grape can use custom Middleware for authentication. How to implement these Middleware have a look at Rack::Auth::Basic or similar implementations.

For registering a Middleware you need the following options:

• label - the name for your authenticator to use it later
• MiddlewareClass - the MiddlewareClass to use for authentication
• option_lookup_proc - A Proc with one Argument to lookup the options at runtime (return value is an Array as Parameter for the Middleware).

Example:

Grape::Middleware::Auth::Strategies.add(:my_auth, AuthMiddleware, ->(options) { [options[:realm]] } )


auth :my_auth, realm: 'Test Api' do |username, password|
  # lookup the user's password here
  { 'user1' => 'password1' }[username] == password
end

Use Doorkeeper, warden-oauth2 or rack-oauth2 for OAuth2 support.

You can access the controller params, headers, and helpers through the context with the #context method inside any auth middleware inherited from Grape::Middleware::Auth::Base.

Describing and Inspecting an API

Grape routes can be reflected at runtime. This can notably be useful for generating documentation.

Grape exposes arrays of API versions and compiled routes. Each route contains a prefix, version, namespace, method and params. You can add custom route settings to the route metadata with route_setting.

class TwitterAPI < Grape::API
  version 'v1'
  desc 'Includes custom settings.'
  route_setting :custom, key: 'value'
  get do

  end
end

Examine the routes at runtime.

TwitterAPI::versions # yields ['v1']
TwitterAPI::routes # yields an array of Grape::Route objects
TwitterAPI::routes[0].version # => 'v1'
TwitterAPI::routes[0].description # => 'Includes custom settings.'
TwitterAPI::routes[0].settings[:custom] # => { key: 'value' }

Note that Route#route_xyz methods have been deprecated since 0.15.0 and removed since 2.0.1.

Please use Route#xyz instead.

Note that difference of Route#options and Route#settings.

The options can be referred from your route, it should be set by specifying key and value on verb methods such as get, post and put. The settings can also be referred from your route, but it should be set by specifying key and value on route_setting.

Current Route and Endpoint

It's possible to retrieve the information about the current route from within an API call with route.

class MyAPI < Grape::API
  desc 'Returns a description of a parameter.'
  params do
    requires :id, type: Integer, desc: 'Identity.'
  end
  get 'params/:id' do
    route.params['id'] # yields the parameter description
  end
end

The current endpoint responding to the request is self within the API block or env['api.endpoint'] elsewhere. The endpoint has some interesting properties, such as source which gives you access to the original code block of the API implementation. This can be particularly useful for building a logger middleware.

class ApiLogger < Grape::Middleware::Base
  def before
    file = env['api.endpoint'].source.source_location[0]
    line = env['api.endpoint'].source.source_location[1]
    env['rack.logger']&.debug "[api] #{file}:#{line}"
  end
end

Before, After and Finally

Blocks can be executed before or after every API call, using before, after, before_validation and after_validation. If the API fails the after call will not be triggered, if you need code to execute for sure use the finally.

Before and after callbacks execute in the following order:

1. before
2. before_validation
3. validations
4. after_validation (upon successful validation)
5. the API call (upon successful validation)
6. after (upon successful validation and API call)
7. finally (always)

Steps 4, 5 and 6 only happen if validation succeeds.

If a request for a resource is made with an unsupported HTTP method (returning HTTP 405) only before and finally callbacks will be executed. The remaining callbacks will be bypassed.

If a request for a resource is made that triggers the built-in OPTIONS handler, only before, after and finally callbacks will be executed. The remaining callbacks will be bypassed.

For example, using a simple before block to set a header.

before do
  header 'X-Robots-Tag', 'noindex'
end

You can ensure a block of code runs after every request (including failures) with finally:

finally do
  # this code will run after every request (successful or failed)
end

Namespaces

Callbacks apply to each API call within and below the current namespace:

class MyAPI < Grape::API
  get '/' do
    "root - #{@blah}"
  end

  namespace :foo do
    before do
      @blah = 'blah'
    end

    get '/' do
      "root - foo - #{@blah}"
    end

    namespace :bar do
      get '/' do
        "root - foo - bar - #{@blah}"
      end
    end
  end
end

The behavior is then:

GET /           # 'root - '
GET /foo        # 'root - foo - blah'
GET /foo/bar    # 'root - foo - bar - blah'

Params on a namespace (or whichever alias you are using) will also be available when using before_validation or after_validation:

class MyAPI < Grape::API
  params do
    requires :blah, type: Integer
  end
  resource ':blah' do
    after_validation do
      # if we reach this point validations will have passed
      @blah = declared(params, include_missing: false)[:blah]
    end

    get '/' do
      @blah.class
    end
  end
end

The behavior is then:

GET /123        # 'Integer'
GET /foo        # 404, the declared Integer narrows the path to digits

Versioning

When a callback is defined within a version block, it's only called for the routes defined in that block.

class Test < Grape::API
  resource :foo do
    version 'v1', :using => :path do
      before do
        @output ||= 'v1-'
      end
      get '/' do
        @output += 'hello'
      end
    end

    version 'v2', :using => :path do
      before do
        @output ||= 'v2-'
      end
      get '/' do
        @output += 'hello'
      end
    end
  end
end

The behavior is then:

GET /v1/foo       # 'v1-hello'
GET /v2/foo       # 'v2-hello'

Altering Responses

Using present in any callback allows you to add data to a response:

class MyAPI < Grape::API
  format :json

  after_validation do
    present :name, params[:name] if params[:name]
  end

  get '/greeting' do
    present :greeting, 'Hello!'
  end
end

The behavior is then:

GET /greeting              # {"greeting":"Hello!"}
GET /greeting?name=Alan    # {"name":"Alan","greeting":"Hello!"}

Instead of altering a response, you can also terminate and rewrite it from any callback using error!, including after. This will cause all subsequent steps in the process to not be called. This includes the actual api call and any callbacks

Anchoring

Grape by default anchors all request paths, which means that the request URL should match from start to end to match, otherwise a 404 Not Found is returned. However, this is sometimes not what you want, because it is not always known upfront what can be expected from the call. Rails solves this problem by using a anchor: false option in your routes. In Grape this option can be used as well when a method is defined.

For instance when your API needs to get part of an URL, for instance:

class TwitterAPI < Grape::API
  namespace :statuses do
    get '/(*:status)', anchor: false do

    end
  end
end

This will match all paths starting with '/statuses/'. There is one caveat though: the params[:status] parameter only holds the first part of the request url. Luckily this can be circumvented by using the described above syntax for path specification and using the PATH_INFO Rack environment variable, using env['PATH_INFO']. This will hold everything that comes after the '/statuses/' part.

Instance Variables

You can use instance variables to pass information across the various stages of a request. An instance variable set within a before validator is accessible within the endpoint's code and can also be utilized within the rescue_from handler.

class TwitterAPI < Grape::API
  before do
    @var = 1
  end

  rescue_from :all do
    puts @var # => 1
  end

  get '/' do
    puts @var # => 1
    raise
  end
end

The values of instance variables cannot be shared among various endpoints within the same API. This limitation arises due to Grape generating a new instance for each request made. Consequently, instance variables set within an endpoint during one request differ from those set during a subsequent request, as they exist within separate instances.

class TwitterAPI < Grape::API
  get '/first' do
    @var = 1
    puts @var # => 1
  end

  get '/second' do
    puts @var # => nil
  end
end

Using Custom Middleware

Grape Middleware

You can make a custom middleware by using Grape::Middleware::Base. It's inherited from some grape official middlewares in fact.

For example, you can write a middleware to log application exception.

class LoggingError < Grape::Middleware::Base
  def after
    return unless @app_response && @app_response[0] == 500
    env['rack.logger'].error("Raised error on #{env['PATH_INFO']}")
  end
end

Your middleware can overwrite application response as follows, except error case.

class Overwriter < Grape::Middleware::Base
  def after
    [200, { 'Content-Type' => 'text/plain' }, ['Overwritten.']]
  end
end

You can add your custom middleware with use, that push the middleware onto the stack, and you can also control where the middleware is inserted using insert, insert_before and insert_after.

class CustomOverwriter < Grape::Middleware::Base
  def after
    [200, { 'Content-Type' => 'text/plain' }, [@options[:message]]]
  end
end


class API < Grape::API
  use Overwriter
  insert_before Overwriter, CustomOverwriter, message: 'Overwritten again.'
  insert 0, CustomOverwriter, message: 'Overwrites all other middleware.'

  get '/' do
  end
end

You can access the controller params, headers, and helpers through the context with the #context method inside any middleware inherited from Grape::Middleware::Base.

Rails Middleware

Note that when you're using Grape mounted on Rails you don't have to use Rails middleware because it's already included into your middleware stack. You only have to implement the helpers to access the specific env variable.

If you are using a custom application that is inherited from Rails::Application and need to insert a new middleware among the ones initiated via Rails, you will need to register it manually in your custom application class.

class Company::Application < Rails::Application
  config.middleware.insert_before(Rack::Attack, Middleware::ApiLogger)
end

Remote IP

By default you can access remote IP with request.ip. This is the remote IP address implemented by Rack. Sometimes it is desirable to get the remote IP Rails-style with ActionDispatch::RemoteIp.

Add gem 'actionpack' to your Gemfile and require 'action_dispatch/middleware/remote_ip.rb'. Use the middleware in your API and expose a client_ip helper. See this documentation for additional options.

class API < Grape::API
  use ActionDispatch::RemoteIp

  helpers do
    def client_ip
      env['action_dispatch.remote_ip'].to_s
    end
  end

  get :remote_ip do
    { ip: client_ip }
  end
end

Writing Tests

Writing Tests with Rack

Use rack-test and define your API as app.

RSpec

You can test a Grape API with RSpec by making HTTP requests and examining the response.

describe Twitter::API do
  include Rack::Test::Methods

  def app
    Twitter::API
  end

  context 'GET /api/statuses/public_timeline' do
    it 'returns an empty array of statuses' do
      get '/api/statuses/public_timeline'
      expect(last_response.status).to eq(200)
      expect(JSON.parse(last_response.body)).to eq []
    end
  end
  context 'GET /api/statuses/:id' do
    it 'returns a status by id' do
      status = Status.create!
      get "/api/statuses/#{status.id}"
      expect(last_response.body).to eq status.to_json
    end
  end
end

There's no standard way of sending arrays of objects via an HTTP GET, so POST JSON data and specify the correct content-type.

describe Twitter::API do
  context 'POST /api/statuses' do
    it 'creates many statuses' do
      statuses = [{ text: '...' }, { text: '...'}]
      post '/api/statuses', statuses.to_json, 'CONTENT_TYPE' => 'application/json'
      expect(last_response.status).to eq 201
    end
  end
end

Airborne

You can test with other RSpec-based frameworks, including Airborne, which uses rack-test to make requests.

require 'airborne'

Airborne.configure do |config|
  config.rack_app = Twitter::API
end

describe Twitter::API do
  context 'GET /api/statuses/:id' do
    it 'returns a status by id' do
      status = Status.create!
      get "/api/statuses/#{status.id}"
      expect_json(status.as_json)
    end
  end
end

MiniTest

require 'test_helper'

class Twitter::APITest < Minitest::Test
  include Rack::Test::Methods

  def app
    Twitter::API
  end

  def test_get_api_statuses_public_timeline_returns_an_empty_array_of_statuses
    get '/api/statuses/public_timeline'
    assert last_response.ok?
    assert_equal [], JSON.parse(last_response.body)
  end

  def test_get_api_statuses_id_returns_a_status_by_id
    status = Status.create!
    get "/api/statuses/#{status.id}"
    assert_equal status.to_json, last_response.body
  end
end

Writing Tests with Rails

RSpec

describe Twitter::API do
  context 'GET /api/statuses/public_timeline' do
    it 'returns an empty array of statuses' do
      get '/api/statuses/public_timeline'
      expect(response.status).to eq(200)
      expect(JSON.parse(response.body)).to eq []
    end
  end
  context 'GET /api/statuses/:id' do
    it 'returns a status by id' do
      status = Status.create!
      get "/api/statuses/#{status.id}"
      expect(response.body).to eq status.to_json
    end
  end
end

In Rails, HTTP request tests would go into the spec/requests group. You may want your API code to go into app/api - you can match that layout under spec by adding the following in spec/rails_helper.rb.

RSpec.configure do |config|
  config.include RSpec::Rails::RequestExampleGroup, type: :request, file_path: /spec\/api/
end

MiniTest

class Twitter::APITest < ActiveSupport::TestCase
  include Rack::Test::Methods

  def app
    Rails.application
  end

  test 'GET /api/statuses/public_timeline returns an empty array of statuses' do
    get '/api/statuses/public_timeline'
    assert last_response.ok?
    assert_equal [], JSON.parse(last_response.body)
  end

  test 'GET /api/statuses/:id returns a status by id' do
    status = Status.create!
    get "/api/statuses/#{status.id}"
    assert_equal status.to_json, last_response.body
  end
end

Stubbing Helpers

Because helpers are mixed in based on the context when an endpoint is defined, it can be difficult to stub or mock them for testing. Grape::Endpoint.before_each allows you to define behavior on the endpoint that will run before every request.

This feature is provided by Grape::Testing, a standalone module intended for test environments only. Add it to your test helper:

require 'grape/testing'

Then use it in your tests:

describe 'an endpoint that needs helpers stubbed' do
  before do
    Grape::Endpoint.before_each do |endpoint|
      allow(endpoint).to receive(:helper_name).and_return('desired_value')
    end
  end

  after do
    Grape::Endpoint.reset_before_each
  end

  it 'stubs the helper' do

  end
end

Reloading API Changes in Development

Reloading in Rack Applications

Use grape-reload.

Reloading in Rails Applications

Rails 7+ (Zeitwerk)

Rails 7+ uses Zeitwerk as the default autoloader, which automatically handles reloading of code in development mode without any additional configuration.

If your API files are in app/api, Zeitwerk will automatically autoload and reload them. No additional configuration is needed.

If you encounter issues with reloading, ensure that:

1. Your API files follow Zeitwerk naming conventions (file names should match class names).
2. The config.enable_reloading is set to true in config/environments/development.rb (this is the default).

For troubleshooting autoloading issues, have a look at the Rails documentation.

See the Rails Autoloading and Reloading Constants guide for more information.

Rails 6 and Earlier

For Rails versions before 7, you need to configure reloading manually.

Add API paths to config/application.rb.

# Auto-load API and its subdirectories
config.paths.add File.join('app', 'api'), glob: File.join('**', '*.rb')
config.autoload_paths += Dir[Rails.root.join('app', 'api', '*')]

Create config/initializers/reload_api.rb.

if Rails.env.development?
  ActiveSupport::Dependencies.explicitly_unloadable_constants << 'Twitter::API'

  api_files = Dir[Rails.root.join('app', 'api', '**', '*.rb')]
  api_reloader = ActiveSupport::FileUpdateChecker.new(api_files) do
    Rails.application.reload_routes!
  end
  ActiveSupport::Reloader.to_prepare do
    api_reloader.execute_if_updated
  end
end

See StackOverflow #3282655 for more information.

Performance Monitoring

Active Support Instrumentation

Grape has built-in support for ActiveSupport::Notifications which provides simple hook points to instrument key parts of your application.

Hook Points

The following hook points are currently supported:

endpoint_run.grape

The main execution of an endpoint, includes filters and rendering.

• endpoint - The endpoint instance

endpoint_render.grape

The execution of the main content block of the endpoint.

• endpoint - The endpoint instance

endpoint_run_filters.grape

• endpoint - The endpoint instance
• filters - The filters being executed
• type - The type of filters (before, before_validation, after_validation, after)

endpoint_run_validators.grape

The execution of validators.

• endpoint - The endpoint instance
• validators - The validators being executed
• request - The request being validated

format_response.grape

Serialization or template rendering.

• env - The request environment
• formatter - The formatter object (e.g., Grape::Formatter::Json)

See the ActiveSupport::Notifications documentation for information on how to subscribe to these events.

Subscribe to Hooks

Once subscribed to the instrumentation, you can intercept the events reported above.

ActiveSupport::Notifications.subscribe(/<api_path>/) do |name, start, finish, id, payload|
  # your code to intercept the notification
end

The request data, the API’s internal data, and the response can be retrieved from the payload.

You can use payload.fetch(:endpoint) or directly payload[:endpoint].

The :endpoint contains the data currently being processed, and access to attributes such as body, request, params, headers, cookies and response_cookies

For example, payload[:endpoint].body provides the current state of the response.

ActiveSupport::Notifications.subscribe(/v1/) do |name, start, finish, id, payload|
  hook_record = {
    hook: name,
    status: payload[:env]&.dig("api.endpoint")&.status,
    format: payload[:env]&.dig("api.format"),
    body: payload[:endpoint]&.body,
    duration: (finish - start) * 1000
  }
  # your code to save the notification
end

Monitoring Products

Grape integrates with following third-party tools:

• New Relic - built-in support from v3.10.0 of the official newrelic_rpm gem, also newrelic-grape gem
• Librato Metrics - grape-librato gem
• Rails Performance - rails_performance gem
• Skylight - skylight gem, documentation
• AppSignal - appsignal-ruby gem, documentation
• ElasticAPM - elastic-apm gem, documentation
• Datadog APM - ddtrace gem, documentation

Contributing to Grape

Grape is work of hundreds of contributors. You're encouraged to submit pull requests, propose features and discuss issues.

See CONTRIBUTING.

Security

See SECURITY for details.

License

MIT License. See LICENSE for details.

Copyright

Copyright (c) 2010-2020 Michael Bleigh, Intridea Inc. and Contributors.