Deep comparison of dicts - cmp() versus ==?

[Rustom Mody]

On Friday, March 20, 2015 at 8:27:26 AM UTC+5:30, Victor Hooi wrote:
> Hi Ben,
> 
> When I said "deep", I meant, as in, to an arbitrary level of nesting (i.e. dicts, containing dicts, containing dicts etc) - sorry if I got the terminology wrong.
> 
> The two dicts weren't equal by intention - the idea was that a comparison operator would return "False" for those two.
> 
> I was just curious why cmp() was phased out (as in, were there cases where "==" was better) - but if functionality they're the same, and it's just a nomenclature thing, that's also fine =).
> 
> Finally, so cmp()/== return true/false for comparison...

I guess you are confusing two orthogonal questions:
1. deep vs shallow comparison
2. binary vs trivalent comparison

Numbers (not complex) satisfy the trichotomy law: ie for any 2 numbers x,y:
x < y or x > y o x = y

Unfortunately most logic in programming languages is binary -- true or false
(well Fortran had an arithmetic if -- very unfashionable even in Fortran --
but thats because of the need to goto)

Likewise C's strcmp is trivalent. But a trivalent value is clumsy in a binary logic.  Unfortunately no switch statement for <, =, > -- so we are stuck with
if-else-if-else.

All this is unrelated to deep-vs=shallow: cmp is deep, == is deep
In fact what would/could a shallow comparison mean??
Shallow/deep only make sense for copy not comparison
[If at all: Languages like Haskell are explicitly constructed so that the
programmers is disallowed from making that distinction at all]