352459 - Incorrect decompilation for 4..@x++

Status: RESOLVED WORKSFORME

(Core :: JavaScript Engine, defect)

Description

[Jesse Ruderman]

js> function() { return 4..@x++; }     
function () {
    return 4.@x++;
}

The 4 wants some parens around itself.

Comment 1

[Brendan Eich [:brendan]]

Sorry, I have to call this minor and leave it in the [e-mail] pile.

Jeff, feel free to grab it.

/be

Comment 2

[Blake Kaplan (:mrbkap) (inactive)]

This seems fixed now:
js> function() { return 4..@x++; }
function () {
    return (4).@x++;
}

Marking as WFM (Jeff, feel free to figure out which bug fixed this).

Comment 3

[Bob Clary [:bc] (inactive)]

It is ok for this to convert the descendant operator ".." to property access "." ?

/cvsroot/mozilla/js/tests/e4x/decompilation/regress-352459.js,v  <--  regress-352459.js
initial revision: 1.1

Comment 4

[Jeff Walden [:Waldo]]

(In reply to comment #3)
> It is ok for this to convert the descendant operator ".." to property access
> "." ?

|4..@x++| is equivalent to |(4.).@x++| which is equivalent to |(4).@x++|; the key is that 4. is a numeric literal, and since the next token isn't a number but is rather .@ the attribute access acts upon (4.).  Consequently, tho it looks like it this isn't the descendant operator.

Consider also |3.141592654.toPrecision(3)| or |4..toString()| for very slightly more understandable cases.

I agree this isn't particularly intuitive.  :-)

Comment 5

[Bob Clary [:bc] (inactive)]

(In reply to comment #4)

Thanks Jeff, I <3 e4x even less now if at all possible.