True == 1 weirdness

Jussi Piitulainen harvesting at makes.email.invalid
Thu Sep 17 22:44:06 CEST 2015


Chris Angelico writes:
> On Fri, Sep 18, 2015 at 4:49 AM, Ian Kelly wrote:
>> On Thu, Sep 17, 2015 at 1:06 AM, Jussi Piitulainen
>> wrote:
>>> Ordinary binary operators not only combine things of the same type, they
>>> also produce a thing of that same type. So 'in' does not fit among them
>>> either.
>>>
>>> I feel it's _more_ at home among comparison operators. (Hm. That's
>>> 'operator' in a different sense.)
>>
>> Comparison operators *are* binary operators. All that "binary" means
>> is that it takes two arguments.
>
> I think what Jussi is saying is that int+int yields int, and
> float*float yields float, and so on - but even that is true only of
> the arithmetic operators, and not all of them (int/int -> float in
> Py3). But generalizing from "arithmetic operators" to "ordinary
> operators" is a little unfair, unless you assume that the sole purpose
> of programming is to represent algebra.

Yes, that's what I was trying to say, though I should have used the word
"operation" not "operator". The operators that denote something like
operations are routinely used to feed their values back to the same or
related operations; doing that with truth-valued operators does not
often make sense; their results are combined with and, or, and not
instead.

I can easily make up special cases myself, but now I'm trying to think
of typical uses and say that Python got this quite right.


More information about the Python-list mailing list