Ruby Custom Error Classes: Inheritance of the Message Attribute

Ruby custom error classes: inheritance of the message attribute

raise already sets the message so you don't have to pass it to the constructor:

class MyCustomError < StandardError
  attr_reader :object

  def initialize(object)
    @object = object
  end
end

begin
  raise MyCustomError.new("an object"), "a message"
rescue MyCustomError => e
  puts e.message # => "a message"
  puts e.object # => "an object"
end

I've replaced rescue Exception with rescue MyCustomError, see Why is it a bad style to `rescue Exception => e` in Ruby?.

How to use custom error messages

Simply modify your script like this.

class MyCustomError < StandardError
  def message
    "oops... That was not a number"
  end
end

def print_a_number
  begin
    puts "chose a number"
    number = Float(gets) rescue false
    raise MyCustomError unless number.is_a? Numeric
    puts "The number you chose is #{number}"
  rescue MyCustomError => err
    puts err.message
    puts err.backtrace.inspect
  end
end

Ruby exception inheritance with dynamically generated classes

Ok, I'll try to help here :

First a module is not a class, it allows you to mix behaviour in a class. second see the following example :

module A
  module B
    module Error
      def foobar
        puts "foo"
      end
    end
  end
end

class StandardError
  include A::B::Error
end

StandardError.new.kind_of?(A::B::Error)
StandardError.new.kind_of?(A::B)
StandardError.included_modules #=> [A::B::Error,Kernel]

kind_of? tells you that yes, Error does possess All of A::B::Error behaviour (which is normal since it includes A::B::Error) however it does not include all the behaviour from A::B and therefore is not of the A::B kind. (duck typing)

Now there is a very good chance that ruby-aws reopens one of the superclass of NameError and includes Amazon::AWS:Error in there. (monkey patching)

You can find out programatically where the module is included in the hierarchy with the following :

class Class
  def has_module?(module_ref)
    if self.included_modules.include?(module_ref) and not self.superclass.included_modules.include?(module_ref)                      
        puts self.name+" has module "+ module_ref.name          
    else
      self.superclass.nil? ? false : self.superclass.has_module?(module_ref)
    end        
  end
end
StandardError.has_module?(A::B::Error)
NameError.has_module?(A::B::Error)

Regarding your second question I can't see anything better than

begin 
#do AWS error prone stuff
rescue Exception => e
  if Amazon::AWS::Error.constants.include?(e.class.name)
    #awsError
  else
    whatever
  end 
end

(edit -- above code doesn't work as is : name includes module prefix which is not the case of the constants arrays. You should definitely contact the lib maintainer the AWSError class looks more like a factory class to me :/ )

I don't have ruby-aws here and the caliban site is blocked by the company's firewall so I can't test much further.

Regarding the include : that might be the thing doing the monkey patching on the StandardError hierarchy. I am not sure anymore but most likely doing it at the root of a file outside every context is including the module on Object or on the Object metaclass. (this is what would happen in IRB, where the default context is Object, not sure about in a file)

from the pickaxe on modules :

A couple of points about the include statement before we go on. First, it has nothing to do with files. C programmers use a preprocessor directive called #include to insert the contents of one file into another during compilation. The Ruby include statement simply makes a reference to a named module. If that module is in a separate file, you must use require to drag that file in before using include.

(edit -- I can't seem to be able to comment using this browser :/ yay for locked in platforms)

Raising an an exception: string vs custom class

There is an important advantage of custom errors like ApiTokenExpired if you have to handle the error in an specific way somewhere in the call stack.

begin
  # ...
rescue ApiTokenExpired => error
  # handle the specific error
rescue => error
  # default error handling 
end

IMO the decision to create custom errors doesn't depend on the size of the project or future maintenance.

Since custom error classes cause an extra effort I don't use it by default. When a special error handling is needed then an error class should be introduced.

Raise custom Exception with arguments

Solution:

class FooError < StandardError
  attr_reader :foo

  def initialize(foo)
   super
   @foo = foo
  end
end

This is the best way if you follow the Rubocop Style Guide and always pass your message as the second argument to raise:

raise FooError.new('foo'), 'bar'

You can get foo like this:

rescue FooError => error
  error.foo     # => 'foo'
  error.message # => 'bar'

If you want to customize the error message then write:

class FooError < StandardError
  attr_reader :foo

  def initialize(foo)
   super
   @foo = foo
  end

  def message
    "The foo is: #{foo}"
  end
end

This works well if foo is required. If you want foo to be an optional argument, then keep reading.

---

Explanation:

Pass your message as the second argument to raise

As the Rubocop Style Guide says, the message and the exception class should be provided as separate arguments because if you write:

raise FooError.new('bar')

And want to pass a backtrace to raise, there is no way to do it without passing the message twice:

raise FooError.new('bar'), 'bar', other_error.backtrace

As this answer says, you will need to pass a backtrace if you want to re-raise an exception as a new instance with the same backtrace and a different message or data.

Implementing FooError

The crux of the problem is that if foo is an optional argument, there are two different ways of raising exceptions:

raise FooError.new('foo'), 'bar', backtrace # case 1

and

raise FooError, 'bar', backtrace # case 2

and we want FooError to work with both.

In case 1, since you've provided an error instance rather than a class, raise sets 'bar' as the message of the error instance.

In case 2, raise instantiates FooError for you and passes 'bar' as the only argument, but it does not set the message after initialization like in case 1. To set the message, you have to call super in FooError#initialize with the message as the only argument.

So in case 1, FooError#initialize receives 'foo', and in case 2, it receives 'bar'. It's overloaded and there is no way in general to differentiate between these cases. This is a design flaw in Ruby. So if foo is an optional argument, you have three choices:

(a) accept that the value passed to FooError#initialize may be either foo or a message.

(b) Use only case 1 or case 2 style with raise but not both.

(c) Make foo a keyword argument.

If you don't want foo to be a keyword argument, I recommend (a) and my implementation of FooError above is designed to work that way.

If you raise a FooError using case 2 style, the value of foo is the message, which gets implicitly passed to super. You will need an explicit super(foo) if you add more arguments to FooError#initialize.

If you use a keyword argument (h/t Lemon Cat's answer) then the code looks like:

class FooError < StandardError
  attr_reader :foo

  def initialize(message, foo: nil)
   super(message)
   @foo = foo
  end
end

And raising looks like:

raise FooError, 'bar', backtrace
raise FooError(foo: 'foo'), 'bar', backtrace

Getting a list of existing Rails Error classes for re-use/inheritance

There's a mostly adequate solution here: http://www.ruby-forum.com/topic/158088

Since this question is unanswered, yet coming up at the top of Google search results, I decided to wrap Frederick Cheung's solution up in a rake task and post it here.

Drop the following in lib/tasks/exceptions.rake

namespace :exceptions do
  task :list => :environment do
    exceptions = []

    ObjectSpace.each_object(Class) do |k|
      exceptions << k if k.ancestors.include?(Exception)
    end

    puts exceptions.sort { |a,b| a.to_s <=> b.to_s }.join("\n")
  end
end

Run it with:

bundle exec rake exceptions:list

If you're still on Rails 2, or not using Bundler, leave off the bundle exec

This list is probably adequate, but not exhaustive. For example, ActiveResource defines several exceptions such as ActiveResource::ConnectionError and ActiveResource::TimeoutError that are not appearing when I run this task. Maybe someone else can enlighten me as to why.

Ruby's Exception Error classes

Define an initialize method, which takes the message as an argument with a default value. Then call StandardError's initialize method with that message (using super).

class MyError < StandardError
  def initialize(msg = "You've triggered a MyError")
    super(msg)
  end
end

Make error show conditional message?

where you have

 <% if object.errors.count = 1 %>

you need to have:

 <% if object.errors.count == 1 %>

note the double equal sign.

---

Related Topics