CODEWITHSUNDEEP
language-agnostic
zounds
zounds

Reputation: 783

Languages where ¬(a = b) and (a ≠ b) can be different

In C++ one can overload the == and != operators for user types, but the language doesn't care how you do it. You can overload both to return true no matter what, so !(a==b) and (a!=b) don't necessarily have to evaluate to the same thing. How many other languages have a situation where ¬(a = b) and (a ≠ b) can be different? Is it a common thing?

It's not just an issue of overloads, but of strange corner cases even for primitive types. NaN in C and C++ doesn't compare equal to anything, including NaN. It is true that NaN != NaN in C, but maybe there are similar cases in other languages that cause ¬(a = b) and (a ≠ b) to be different?

Upvotes: 2

Views: 127

Answers (2)

Ross Patterson
Ross Patterson

Reputation: 9570

Guy L. Steele famously said

...the ability to define your own operator functions means that a simple statement such as x=a+b; in an inner loop might involve the sending of e-mail to Afghanistan.

And as corsiKa says, just because you can do it, doesn't make it a good idea.

Upvotes: 2

tsm
tsm

Reputation: 3658

I know for a fact that Python and Ruby can and Java and PHP can not. (In Java == determines if two objects are the same thing in memory, not just semantically equivalent values. In PHP...never mind.) I'd also imagine that Lisp and JS can and C can not, but that's a bit more speculative.

It's nothing unusual to be able to overload operators. It is very rare for !(a==b) and (a!=b) to have different results though.

Upvotes: 1

Related Questions