Re: [Emacs-diffs] master 6354e3c: Handle indentation of nested ternary operators in JS

Re: [Emacs-diffs] master 6354e3c: Handle indentation of nested ternary operators in JS

[Stefan Monnier]

> +  isSet
> +    ? (isEmpty ? 2 : 3)
> +    : 4

Not sure about JS, but at least in C, those parentheses are not needed
(so the indentation code should ideally DTRT even in the absence of
parentheses).


        Stefan

Re: [Emacs-diffs] master 6354e3c: Handle indentation of nested ternary operators in JS

[Davis Herring]

>> +  isSet
>> +    ? (isEmpty ? 2 : 3)
>> +    : 4
>
> Not sure about JS, but at least in C, those parentheses are not needed

JavaScript and C have the same ?: parsing.  Moreover, I think even PHP 
would get that one right without parentheses, because grouping the first 
? and : and the second ? and : would be self-intersecting.

(What PHP gets wrong is

   !isSet ? 4 : isEmpty ? 2 : 3

which it, unlike all sane languages, parses as

   (!isSet ? 4 : isEmpty) ? 2 : 3

aka

   !isSet || isEmpty ? 2 : 3

which is rarely what was intended.)

Davis

-- 
This product is sold by volume, not by mass.  If it appears too dense or 
too sparse, it is because mass-energy conversion has occurred during 
shipping.

Re: [Emacs-diffs] master 6354e3c: Handle indentation of nested ternary operators in JS

[Dmitry Gutov]

On 13.04.2017 22:24, Stefan Monnier wrote:
>> +  isSet
>> +    ? (isEmpty ? 2 : 3)
>> +    : 4
> 
> Not sure about JS, but at least in C, those parentheses are not needed
> (so the indentation code should ideally DTRT even in the absence of
> parentheses).

You're right, it's the same in JS. Luckily, most programmers don't feel 
comfortable leaving the inner group without parentheses (so the reporter 
went away satisfied).

Not exactly sure what logic to use to support the paren-less case, so 
patch welcome!