New: no-unsafe-optional-chaining rule (fixes #13431)

[yeonjuan]

Prerequisites checklist

• I have read the contributing guidelines.

What is the purpose of this pull request? (put an "X" next to an item)

[ ] Documentation update
[ ] Bug fix (template)
[x] New rule (template)
[ ] Changes an existing rule (template)
[ ] Add autofixing to a rule
[ ] Add a CLI option
[ ] Add something to the core
[ ] Other, please explain:

What changes did you make? (Give an overview)

Added a new rule no-unsafe-optional-chaining which proposed by @mysticatea #13431

This rule will catch cases that may possibly TypeError when using optional chaining.

var obj = {};

// TypeError

(obj?.foo).bar;

(obj?.foo)();

(obj?.foo)`template`;

new (obj?.foo)();

class A extends obj?.foo {};

[...obj?.foo];

bar(...obj?.foo);

new Bar(...obj?.foo);

1 in obj?.foo;

var { a } = obj?.foo;

var [a] = obj?.foo;

// [edited]
foo instanceof obj?.prop

for (foo of obj?.prop);

[{ foo } = obj?.prop] = [];

with (obj?.prop);

// [edited] - handle logical result.

(obj?.foo ?? obj?.bar)();

(foo || obj?.foo)();

And by the option disallowArithmeticOperators: true it also catches the operation which possibly results in unexpected NaN. - comment

obj?.foo * bar;
obj?.foo - bar;
obj?.foo + bar;
obj?.foo % bar;
obj?.foo ** bar;

a *= obj?.foo;
a -= obj?.foo;
a += obj?.foo;
a %= obj?.foo;
a **= obj?.foo;

+obj?.foo;
-obj?.foo;

Is there anything you'd like reviewers to focus on?

Maybe there are some other cases we should handle?

[mdjermanovic]

Maybe there are some other cases we should handle?

I think all of these will also throw on undefined:

• foo instanceof obj?.prop
• for (foo of obj?.prop);
• [{ foo } = obj?.prop] = []; (AssignmentPattern#right if AssignmentPattern#left is a pattern)
• with (obj?.prop);

[yeonjuan]

@mdjermanovic Thanks 👍 I added handling those.

[mdjermanovic]

Looks great in general. I didn't review everything in detail yet, but noticed a couple of problems.

[yeonjuan]

@mdjermanovic Thanks
I fixed it and refactor the tests. :)

[yeonjuan]

@mdjermanovic Thanks for all reviews. It seems to be fixed but, I'll take a look once again and request a review. :)

[mdjermanovic]

LGTM, thanks! 🎉

[yeonjuan]

Hi @mdjermanovic I have a question :) Maybe this rule should handle await also?
IMO It should be, but want to know others' opinion.

(async function main() {
          (await (foo?.bar))?.baz // ok

          (await (foo?.bar)).baz; // Maybe error? It's lint free in current version.
})();

I marked it as do not merge because it seems to need a decision.

[mdjermanovic]

(await (foo?.bar)).baz; // Maybe error? It's lint free in current version.

Makes sense to me, we can handle await in function checkUndefinedShortCircuit since await undefined evaluates to undefined.

[mdjermanovic]

LGTM, thanks!

[btmills]

LGTM. Thanks @yeonjuan!