1243858 - Exponentiation Operator precendence update (ES2016/ES7, Stage 4)

Status: RESOLVED FIXED

(Core :: JavaScript Engine, defect)

Description

[Rick Waldron [:rwaldron]]

Exponentiation Operator reached stage 4 status today and the normative spec will be merged with Ecma 262 (https://github.com/tc39/ecma262/pull/318)

Comment 1

[Rick Waldron [:rwaldron]]

fscholz, I've updated the MDN docs for exp op: 

https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/Arithmetic_Operators#Exponentiation_(**)

Comment 2

[ziyunfei]

(In reply to Rick Waldron [:rwaldron] from comment #1)
> fscholz, I've updated the MDN docs for exp op: 
> 
> https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/
> Arithmetic_Operators#Exponentiation_(**)

Waldron, is (-2) ** 2 also a syntax error?

Comment 3

[Rick Waldron [:rwaldron]]

(In reply to ziyunfei from comment #2)
> (In reply to Rick Waldron [:rwaldron] from comment #1)
> > fscholz, I've updated the MDN docs for exp op: 
> > 
> > https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/
> > Arithmetic_Operators#Exponentiation_(**)
> 
> Waldron, is (-2) ** 2 also a syntax error?

No, it's not. 


Also, I think the ids on that page need to have "(" and ")" removed, so they don't break the fragment urls.

Comment 4

[Florian Scholz (Open Web Docs)]

(In reply to Rick Waldron [:rwaldron] from comment #3)
> Also, I think the ids on that page need to have "(" and ")" removed, so they
> don't break the fragment urls.
All operators on that page are also linked without their syntax, like this:
https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/Arithmetic_Operators#Exponentiation 
Unfortunately, I can't control what is in the TOC as it is generated automatically.

Comment 5

[Liu Xing]

Is it OK to use Exponentiation Operator on Firefox Stable version now?

I try on Firefox Stable and Developer Edition, both failed.
> 2**3
> SyntaxError: expected expression, got '*'

Comment 6

[Sebastian Zartner [:sebo]]

(In reply to Liu Xing from comment #5)
> Is it OK to use Exponentiation Operator on Firefox Stable version now?

It is currently only available on Nightly. See bug 1198453. I couldn't find a bug for letting it ride the trains (move to the other releases), though. Not sure whether this will be covered in this bug, too, or a new bug will be created.

Sebastian

Comment 7

[Liu Xing]

(In reply to Sebastian Zartner [:sebo] from comment #6)
> (In reply to Liu Xing from comment #5)
> > Is it OK to use Exponentiation Operator on Firefox Stable version now?
> 
> It is currently only available on Nightly. See bug 1198453. I couldn't find
> a bug for letting it ride the trains (move to the other releases), though.
> Not sure whether this will be covered in this bug, too, or a new bug will be
> created.
> 
> Sebastian

Thank you very much.

Comment 8

[André Bargull [:anba]]

Attachment: exp_op_unary.patch

I think this approach should suffice to meet the left-hand side restriction, but I'm not 100% sure, therefore I'm only asking for feedback at this point.

Comment 9

[Tooru Fujisawa [:arai]]

I'm missing the context.

where is the latest spec or proposal?
this, or somewhere else?
  http://rwaldron.github.io/exponentiation-operator/

Comment 10

[Jeff Walden [:Waldo]]

https://tc39.github.io/ecma262/#sec-exp-operator probably?

Comment 11

[Tooru Fujisawa [:arai]]

thanks :)

Comment 12

[Tooru Fujisawa [:arai]]

Comment on attachment 8792575 [details] [diff] [review]
exp_op_unary.patch

Review of attachment 8792575 [details] [diff] [review]:
-----------------------------------------------------------------

looks good.

::: js/src/js.msg
@@ +201,5 @@
>  MSG_DEF(JSMSG_BAD_INCOP_OPERAND,       0, JSEXN_REFERENCEERR, "invalid increment/decrement operand")
>  MSG_DEF(JSMSG_BAD_METHOD_DEF,          0, JSEXN_SYNTAXERR, "bad method definition")
>  MSG_DEF(JSMSG_BAD_OCTAL,               1, JSEXN_SYNTAXERR, "{0} is not a legal ECMA-262 octal constant")
>  MSG_DEF(JSMSG_BAD_OPERAND,             1, JSEXN_SYNTAXERR, "invalid {0} operand")
> +MSG_DEF(JSMSG_BAD_POW_LEFTSIDE,        0, JSEXN_SYNTAXERR, "invalid operand for '**' expression")

it would be nice to suggest adding parenthesis to either lhs or whole exponentiation expression, if it can be described in short.
(we could explain the detail in MDN document tho)

Comment 13

[André Bargull [:anba]]

Attachment: bug1243858.patch

The error message has been amended to explain why the operand is invalid ("unparenthesized unary expression").

Drive-by change: Updated JSMSG_ARRAY_INIT_TOO_BIG to use AE/OED instead of BE spelling ("initialiser" -> "initializer").

Comment 14

[André Bargull [:anba]]

Try run:
https://treeherder.mozilla.org/#/jobs?repo=try&revision=ffb3438b5f8a07a54364032fcb878a1f15690743

Comment 15

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/mozilla-inbound/rev/80683e82570c
Disallow unary expression on left-hand side of exponentiation operator. r=arai

Comment 16

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/80683e82570c

Comment 17

[Florian Scholz (Open Web Docs)]

Docs have already been updated (see comment 1 and following). Thanks all!

https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/Arithmetic_Operators#Exponentiation
https://developer.mozilla.org/en-US/Firefox/Releases/52#JavaScript