pundit_custom_errors is an extension for the Pundit gem that enables the creation of custom error messages. This adds more flexibility to retrieve different kinds of messages in the same validation method, according to the nature of the error. As for the default error message, it is also set up to generate them by using a localization file (if existent).

Add this line to your application's Gemfile:

gem 'pundit_custom_errors'

And then execute:

$ bundle

Or install it yourself as:

$ gem install pundit_custom_errors

(Note: since this gem is intended to extend Pundit's capabilities, it is assumed that you're already familiar with it. If not, It is highly recommended that you take a look at Pundit's documentation first.)

For this example, let's suppose there's already a standard Policy class for an object (Post, in this case), called PostPolicy, containing a validation for the show action. It would look like this:

  class PostPolicy
    def show?
      # dummy validation code for the 'show' action
      false
    end
  end

Since the show? method is returning false, this example will always throw a Pundit::NotAuthorizedError, with Pundit's default error message inside. As said before, this behavior can be changed by either setting a message in an attribute inside the Policy class or creating default messages for the actions' validation methods inside a localization YAML file. Both approaches are explained below.

In order to set custom messages, the first thing to be done is to create the error_message attribute with attr_accessor permissions inside the desired Policy class. The PostPolicy class will look like this:

  class PostPolicy
    attr_accessor :error_message

    def show?
      # dummy validation code for the 'show' action
      false
    end
  end

By putting that attribute inside a class, every time a validation method returns false, Pundit will try to use the error_message. If there's something set there (like the example below), Pundit will use it as the error message.

  class PostPolicy
    attr_accessor :error_message

    def show?
      @error_message = "You're not allowed to see this. Better luck next time!"

      # dummy validation code for the 'show' action
      false
    end
  end

As the message is set, every time the show? method returns false, the message present inside error_message will be put inside the Pundit::NotAuthorizedError instance.

By running the command:

$ rails generate pundit_custom_errors:initialize

A file called pundit_custom_errors.en.yml will be generated inside the config/locales folder. It contains the default message, used if there's no messages for the given controller/action validation.

The policy keys must be the snake case representation of the given policy classes, containing action names as keys (e.g: show? and the message as value). Also, the policy hashes must be inside the pundit hash, inside the en (or the desired language abbreviation) hash.

In short, the YAML file structure should be similar as the example below:

en:
  pundit:
    default: "You have requested a page that you do not have access to."
    post_policy:
      show?: "You're not allowed to see this."

1. The gem will use the message present in the error_message attribute, that should be part of the given Policy class.
2. If there's no error message (or even the attribute, if PunditCustomErrors::Policy isn't being extended), it will use the message for the given action validation, present in the YAML file.
3. If there's no message for the given action validation in the YAML, it will use the YAML's default message inside pundit hash.
4. If there's no YAML default message, it will use the hardcoded message, behaving the same way as Pundit does today.

1. Fork it ( https://github.com/luisdaher/pundit_custom_errors/fork )
2. Create your feature branch (git checkout -b my-new-feature)
3. Commit your changes (git commit -am 'Add some feature')
4. Push to the branch (git push origin my-new-feature)
5. Create a new Pull Request