Fix: implicit-arrow-linebreak adds extra characters (fixes #11268)

[ghost]

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

[ ] Documentation update
[ x ] Bug fix (template)
[ ] 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:

Fixes 11268

What changes did you make? (Give an overview)
The original auto-fix calculated white spaces into formatting comments and parentheses, which was the cause for extra characters.

The cause for extra parentheses was due to this line, which searched for arrow functions that followed the evaluated arrow function within the source code.

This PR removes the calculation of white spaces, and more accurately finds nested arrow functions within the evaluated node.

Is there anything you'd like reviewers to focus on?
Please provide/suggest any other test cases if necessary.

[not-an-aardvark]

Thanks for the PR! I have a few questions.

[not-an-aardvark]

Thanks for continuing to work on this!

I'm probably missing something given that the tests are passing, but could you clarify which part of the change prevents the extra closing parentheses from being added? I would have guessed the change happens due to one fewer loop iteration on line 111, but it seems like that only applies to nodes containing nested arrow functions, which isn't the case for the example from #11268.

[ghost]

@not-an-aardvark This line in the code prevents extra closing parentheses from being added by navigating to the body of the arrow function, and this line prevents the appending of extra parentheses.

[not-an-aardvark]

Both of those lines are in the addParentheses function. Is the JSDoc comment for that function out of date? The comment says that the function only applies to code containing nested arrow functions, but the example code from #11268 doesn't contain nested arrow functions.

[ghost]

@not-an-aardvark Thanks for clarifying! Given the example code from #11268

start()
  .then(() => 
    /* If I put a comment here, eslint --fix breaks badly */
    process && typeof process.send === 'function' && process.send('ready')
  )
  .catch(err => {
  	/* catch seems to be needed here */
       console.log('Error: ', err)
  })

when eslint --fix is applied, the output becomes

  /* If I put a comment here, eslint --fix breaks badly */
start()
   .then(() => process && typeof process.send === 'function' && process.send('ready')
    )
   .catch(err => {
    /* catch seems to be needed here */
   console.log('Error: ', err)
    })

And the output for this test case

hello(response =>
  // comment
  response, param => param)

becomes

// comment
hello(response => response, param => param)

Prior to this PR, The JSDoc for addParentheses was incorrect in finding any arrow functions that followed. The above linked line as well this one in autoFixBesides ensure that only nested arrow functions are targeted. Please let me know if any other test cases should be handled.

[not-an-aardvark]

LGTM, thanks!