Re: Why 'statement may have no effect' lint warning here?

[Wolfgang Laun]

On Wed, 14 Apr 2021 at 07:16, <arnold@skeeve.com> wrote:

>
> Rather, it does try to point out real semantic mistakes, such as
>
>         a + b   # useless statement
>
I am working to striaghten this out and will eventually post a patch.
>

We all agree with lint not being a code quality analysis tool.

If I understand you correctly, you intend to fix the current misbehaviour
so that only truly useless statements should be flagged.

I don't want to make life hard for you but can you please clarify your
intention. Which of the following will lint flag after your fix?
   a + b    # as above: useless
   a == b  # as above: useless
   a++ == a--    # ?
  f_without_side_effect(a);   # ?
  f_with_side_effect(b) == 2       # ?
  1 == 2 && a++     # ?
  1 == 1 || b--     # ?
  a = a # ?
  a *= 1 #?
  a += 0 #?
  a*a < 0 && b = 1001  # ?
and so on - you get the drift.

Will you be able to describe the situation after the patch in a single,
simple sentence?

I'd say that "any expression statement that contains a logical expression
that may inhibit the execution of subexpressions with side effects or
function calls" would be such a sentence, but, alas, it would require a ==
1 && b = 2 to be flagged.

Wolfgang Laun