@babel/plugin-transform-spread regression of spread array

[nihgwu]

v7 Regression

First check out: https://new.babeljs.io/docs/en/next/v7-migration.html
Also a partial upgrade tool: https://github.com/babel/babel-upgrade

Potential Commit/PR that introduced the regression
If you have time to investigate, what PR/date introduced this issue.

#6763

before that PR [...array] was transpiled to [].concat(array) now it's array.concat(), this change breaks immutable array(seamless-immutable), for example

import Immutable from `seamless-immutable`
const array = Immutable([1, 2, 3]);
console.log(Immutable.isImmutable([...array]);

it prints false in console of latest Chrome and in Node v8.11.3, but it prints true with Babel@7, it should be false, but as it's transpiled to array.concat(), and an immutable array returns immutable array too with builtin methods(concat, map, reduce...)

Describe the regression
A clear and concise description of what the regression is.

Input Code

var your => (code) => here;

Babel Configuration (.babelrc, package.json, cli command)

{
  "your": { "config": "here" }
}

Expected behavior/code
A clear and concise description of what you expected to happen (or code).

Environment

• Babel version(s): [e.g. v6.0.0, v7.0.0-beta.34]
• Node/npm version: [e.g. Node 8/npm 5]
• OS: [e.g. OSX 10.13.4, Windows 10]
• Monorepo [e.g. yes/no/Lerna]
• How you are using Babel: [e.g. cli, register, loader]

[babel-bot]

Hey @nihgwu! We really appreciate you taking the time to report an issue. The collaborators
on this project attempt to help as many people as possible, but we're a limited number of volunteers,
so it's possible this won't be addressed swiftly.

If you need any help, or just have general Babel or JavaScript questions, we have a vibrant Slack
community that typically always has someone willing to help. You can sign-up here
for an invite.

[danez]

Thanks for reporting this and thanks for linking to the PR that introduced the problem. Sorry for the late reply. I will now take care of this issue and probably partially revert some of the optimizations in the linked PR.