What is the best practice for handling nil objects and properties?

Question

Say I have a User object, which has an email property, and I need the upper cased last letter of their email:

u = User.find(1)    
letter = u.email.upcase.last

If u or email is nil in this chain, then I get a NoMethodError: undefined method 'blah' for nil:Nilclass. I should be able to work around it in most cases, but sometimes, a nil gets where it shouldn't or its hard to contain. One way would be verbose:

u = User.find(1)
letter = nil
if u && u.email
 letter = u.email.upcase.last
end

But this gets annoying and hazardous in a view, or in a long chain of a.bunch.of.properties.down.a.hierarchy. I read about try method in Rails:

u = User.find(1)
letter = u.try(:email).try(:upcase).try(:last)

This is less verbose, but I feel icky writing all those tries. And once I put try in the chain, I have to use them all the way down. Is there a better way?

Answer 1

I like to use the Null Object Pattern. Avdi has a great post explaining this, but the basic idea is you have a little class that can stand in for an object and respond reasonably to the messages you might pass the original object. I've found these are useful not only for avoiding NoMethodErrors but also for setting default values/nice messages.

For instance, you could do:

class NilUser
  def email
    "(no email address)"
  end
end

u = User.find(1) || NilUser.new
u.email.upcase.last # => No errors!