Rails - Feb 2006 - Bizarre ActiveRecord::Errors/validation problem

I could be going mad and missing something really obvious here but I''m
getting an unexpected result from the following.

The schema:

create_table :services, :force => true do |t|
    t.column :id, :primary_key
    t.column :name, :string, :ilmit => 5, :null => false
    t.column :port, :integer, :null => false
    t.column :status, :string
end


The model:

class Service < ActiveRecord::Base
    validates_presence_of :name, :port, :status
end

The unit tests:

class ServiceTest < Test::Unit::TestCase
  def test_new_service
        @service = Service.new
        @service.name = ''www''
        @service.port = 80
        @service.status = ''Unknown''
        assert @service.save
  end

    def test_invalid_service
        @service = Service.new
        assert !@service.save
        assert_equal "can''t be blank",
@service.errors[''name'']
        assert_equal "can''t be blank",
@service.errors[''status'']
        assert_equal "can''t be blank",
@service.errors[''port'']
    end
end

The first test passes fine, but the second test is failing on that last
assertion with:

=> <"can''t be blank"> expected but was
=> <nil>.

Any idea why the validate_presence_of method isn''t picking up the :port
column and providing an appropriate error message like it is with
name/status? Its not a reserved word or something is it?

To aid debugging, here is the output of:

@service.errors.each_full { |m| puts m }
=> Name can''t be blank
=> Status can''t be blank

As you can see, no sign of my error message for the port column.

Cheers
Luke
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://wrath.rubyonrails.org/pipermail/rails/attachments/20060210/769c7c25/attachment.html

Did you copy and paste this? If so, you''ve got a typo here:
t.column :name, :string, :ilmit => 5, :null => false

"ilmit" instead of ''limit''

On 2/10/06, Luke Redpath <contact@lukeredpath.co.uk>
wrote:
> I could be going mad and missing something really obvious here but
I''m
> getting an unexpected result from the following.
>
> The schema:
>
> create_table :services, :force => true do |t|
>     t.column :id, :primary_key
>     t.column :name, :string, :ilmit => 5, :null => false
>     t.column :port, :integer, :null => false
>     t.column :status, :string
> end
>
>
> The model:
>
> class Service < ActiveRecord::Base
>     validates_presence_of :name, :port, :status
> end
>
> The unit tests:
>
> class ServiceTest < Test::Unit::TestCase
>   def test_new_service
>         @service = Service.new
>         @ service.name = ''www''
>         @service.port = 80
>         @service.status = ''Unknown''
>         assert @service.save
>   end
>
>     def test_invalid_service
>         @service = Service.new
>         assert !@ service.save
>         assert_equal "can''t be blank",
@service.errors[''name'']
>         assert_equal "can''t be blank",
@service.errors[''status'']
>         assert_equal "can''t be blank", @service.errors
[''port'']
>     end
> end
>
> The first test passes fine, but the second test is failing on that last
> assertion with:
>
> => <"can''t be blank"> expected but was
> => <nil>.
>
> Any idea why the validate_presence_of method isn''t picking up the
:port
> column and providing an appropriate error message like it is with
> name/status? Its not a reserved word or something is it?
>
> To aid debugging, here is the output of:
>
> @service.errors.each_full { |m| puts m }
> => Name can''t be blank
> => Status can''t be blank
>
> As you can see, no sign of my error message for the port column.
>
> Cheers
> Luke
>
> _______________________________________________
> Rails mailing list
> Rails@lists.rubyonrails.org
> http://lists.rubyonrails.org/mailman/listinfo/rails
>
>
>