Comments on: Ruby’s case statement uses ===

By: zoo

zoo — Fri, 09 Dec 2011 12:39:20 +0000

Nope, that returns nil.

By: Bug

Bug — Wed, 14 Sep 2011 02:16:57 +0000

The point, then, is that the method ‘===’ is arbitrarily named, and might have been named more aptly. For example, ‘case_of?’ conveys that the object is being tested as belonging to a group of objects of which it is a special case. 2 is a special case of 2, just as it is a special case of Fixnum or Numeric, or even (1..7).

By: Bug

Bug — Wed, 14 Sep 2011 02:11:44 +0000

The point of confusion is that ‘===’ is a poor mnemonic for the particular kind of binary relation that it represents. It gives the impression that it acts as an identity check or some other equivalence relation, with which Ruby’s conception of case checking is incompatible.

By: Juan

Juan — Tue, 29 Dec 2009 17:09:04 +0000

Do this =>

my_type = Fixnum
case my_type
when Fixnum:
“Fixed number”
end

By: Rahim

Rahim — Sun, 30 Aug 2009 23:52:42 +0000

I think misread your initial post, I thought you were saying your second example didn’t work as expected (which looking at it again looks like a pretty common pattern which made it all the more surprising).

Googling gives a few people who’ve discussed the same behavior (http://www.justskins.com/forums/doing-a-case-on-class-23663.html, http://blog.jayfields.com/2007/03/ruby-operator.html), both of whom quote the ri docs for Module#=== which explains my observations :

——————————————- Module#===
mod === obj => true or false
—————————————————–
Case Equality—Returns +true+ if _anObject_ is an instance of
_mod_ or one of _mod_’s descendents. Of limited use for modules,
but can be used in +case+ statements to classify objects by class.

By: Rahim

Rahim — Sun, 30 Aug 2009 23:38:13 +0000

I read the docs and still couldn’t see when that choice would help for case statements (perhaps because I was biased by your example…)

Another, at first glance odd, result this leads to is eg:

>> Fixnum === 2
=> true
>> 2 === Fixnum
=> false

By: john

john — Sun, 30 Aug 2009 22:45:02 +0000

Rahim, it’s possible that it’s optimized for case statement usage, which is pretty neat to be honest. Was just unexpected to me until I understood what was going on.

In fact, the docs suggest that:

http://www.ruby-doc.org/core/classes/Object.html#M000345

By: Rahim

Rahim — Sun, 30 Aug 2009 20:14:52 +0000

Any idea of the original rationale for the crazed implementation of Fixnum.=== ?

>> Fixnum === Fixnum
=> false

doesn’t seem like a good move to me!