1084001 - Promise.name is an empty string, it should be "Promise"

Status: RESOLVED FIXED

(Core :: DOM: Core & HTML, defect)

Description

[Marco Castelluccio [:marco]]

I think this is a regression because I remember testing it when fixing bug 1028987. I'm not entirely sure though.

Comment 1

[Marco Castelluccio [:marco]]

It's not a regression.

Comment 2

[Tom S. (please needinfo tschuster)]

This has more to do with how the bindings set up the "function" Promise.

Comment 3

[Boris Zbarsky [:bzbarsky]]

This is similar to https://www.w3.org/Bugs/Public/show_bug.cgi?id=22392 but not identical.  Needs a Web IDL spec issue, I believe; right now per web IDL interface objects do not have .name.

Then again, I see nothing in ES6 that says that Promise should have a .name either, though I could be missing it.

Comment 4

[Boris Zbarsky [:bzbarsky]]

Ah, here's the relevant es6 bit, courtesy of Tom:

  Every built-in Function object, including constructors, that is not identified as an
  anonymous function has a name property whose value is a String. Unless otherwise
  specified, this value is the name that is given to the function in this specification.

We should just align both our impl and Web IDL with this.

Comment 5

[Boris Zbarsky [:bzbarsky]]

And:

  Unless otherwise specified, the name property of a built-in Function object, if it
  exists, has the attributes { [[Writable]]: false, [[Enumerable]]: false,
  [[Configurable]]: true }.

Comment 6

[Boris Zbarsky [:bzbarsky]]

Attachment: Set the .name of DOM interface objects to the name of the interface

Comment 7

[Boris Zbarsky [:bzbarsky]]

Oh, and a note:  The JSPROP_PERMANENT went away to align with the ES6 spec.  Spidermonkey has outstanding bugs to do so too.

Comment 8

[Boris Zbarsky [:bzbarsky]]

And another note: evilpie suggested maybe interning instead of copying here, since we'll atomize anyway when we go to define on the global.

Comment 9

[Peter Van der Beken [:peterv]]

Comment on attachment 8506376 [details] [diff] [review]
Set the .name of DOM interface objects to the name of the interface

Review of attachment 8506376 [details] [diff] [review]:
-----------------------------------------------------------------

::: dom/bindings/BindingUtils.cpp
@@ +464,5 @@
> +      return nullptr;
> +    }
> +
> +    JS::Rooted<JSString*> nameStr(cx,
> +                                  JS_NewStringCopyZ(cx, name));

Yeah, interning seems fine.

Comment 10

[Boris Zbarsky [:bzbarsky]]

Did the interning and filed/fixed https://www.w3.org/Bugs/Public/show_bug.cgi?id=27170 on the spec.

https://hg.mozilla.org/integration/mozilla-inbound/rev/35582d2331a8

Comment 11

[Boris Zbarsky [:bzbarsky]]

And https://hg.mozilla.org/integration/mozilla-inbound/rev/4506af8e95fd to back out the JSPROP_PERMANENT change, because imptests are silly.

I filed https://bugzilla.mozilla.org/show_bug.cgi?id=1089079 to track doing that.

Comment 12

[Tom S. (please needinfo tschuster)]

Fixed as far as I can tell
https://hg.mozilla.org/mozilla-central/rev/35582d2331a8
https://hg.mozilla.org/mozilla-central/rev/4506af8e95fd