New Rule: no-unnecessary-spread - a rule to prevent syntactically unnecessary object, array and JSX attribute spreads

Rule details

This rule would detect provably useless object literal, array literal, and JSX attribute spreads

Related ECMAScript feature

It’s not new, quite old - but I thought there was high value in this check based on my anecdotal experience across several large codebases.

What type of rule is this?

Warns about a potential problem

Example bad code

const x = { ...{} };
const y = { ...{ foo: 1 } };
const z = {
  a: 1,
  b: 2,
  ...{
    c: 3,
    d: 4
  },
  e: 5,
};

const a = [ ...[] ];
const b = [ ...[1] ];
const c = [
  1,
  2,
  ...[
    3,
    4
  ],
  5,
];

const j = <div {...{}} />;
const k = <div {...{ foo: 1 }} />;
const l = <div
  a={1}
  b={2}
  {...{
    c: 3,
    d: 4
  }}
  e={5}
/>;

Good versions of the above "bad" code

const x = { };
const y = { foo: 1 };
const z = {
  a: 1,
  b: 2,
  c: 3,
  d: 4
  e: 5,
};

const a = [ ];
const b = [ 1 ];
const c = [
  1,
  2,
  3,
  4,
  5,
];

const j = <div />;
const k = <div foo={1} />;
const l = <div
  a={1}
  b={2}
  c={3}
  d={4}
  e={5}
/>;

Why should this rule be in the core instead of a plugin?

I’ve worked across a few codebases and found this issue occurs at an alarmingly high rate in each codebase. I think it would be of great benefit to the community if it was part of the core rules and part of the recommended set to help cleanup obviously and syntactically unnecessary, dead code.

Participation

• I am willing to submit a pull request to implement this rule.

Additional comments

I have a prototype of the rule ready to go that I’ve tested on a large codebase (10s of thousands of files) and it includes a fixer that handles the majority of cases (in this codebase 97% of the cases were auto-fixable)

[aladdin-add]

I’m 👍 to the rule.

[aladdin-add]

yes, I’m 100% supportive if the syntax is new added. imho, the purpose of our policy change is to provide only rules that are sufficiently general(and useful), and I believe this proposal is well met. maybe we could make an exception here?