Update: handle parentheses in multiline-ternary (fixes #13195)

[mdjermanovic]

Prerequisites checklist

• I have read the contributing guidelines.

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

[X] Bug fix

fixes #13195

This bug fix can produce more warnings for options "always" and "always-multiline", and fewer warnings for option "never".

What changes did you make? (Give an overview)

The rule will now treat parentheses around test, consequent, and alternate as a part of them.

The following added invalid tests for "always" and "always-multiline" would be valid without this change:

"(a\n) ? b\n: c"
"((a)\n) ? b\n: c"
"a ? (\nb)\n: c"
"a ? (\n(b))\n: c"
"a\n? (b\n): c"
"a\n? ((b)\n): c"
"a\n? b : (\nc)"
"a\n? b : (\n(c))"
"(a\n) ? (\nb\n) : (\nc)"
"((a)\n) ? (\n(b)\n) : (\n(c))"

The following added valid tests for "never would be invalid without this change:

"(a\n) ? b : c"
"((a)\n) ? b : c"
"a ? (\nb) : c"
"a ? (\n(b)) : c"
"a ? (b\n) : c"
"a ? ((b)\n) : c"
"a ? b : (\nc)"
"a ? b : (\n(c))"
"(a\n) ? (\nb\n) : (\nc)"
"((a)\n) ? (\n(b)\n) : (\n(c))"

All other added test cases are regression tests only (they were already valid/invalid).

Also, the reported location will now include parentheses.

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

"never" does not actually enforce that the entire ternary expression is on one line. This added test was already valid before this change:

 { code: "a  +\n b ? c +\n d : e + \nf", options: ["never"] }

I'll open a separate issue for this.