1020975 - No WebIDL overload allowed when extended attributes differ

Status: RESOLVED FIXED

(Core :: DOM: Core & HTML, defect, P5)

Description

[Milan Sreckovic [:milan] (needinfo for best results)]

Can't do something like this:

interface CanvasPattern {
  [Throws, LenientFloat]
  void setTransform(double a, double b, double c, double d, double e, double f);

  [Throws]
  void setTransform(SVGMatrix matrix);
};

the error is:

Extended attributes differ on different overloads of <unresolved scope>::setTransform...

Ehsan suggested this is a bug, so here we go.

Comment 1

[Boris Zbarsky [:bzbarsky]]

This probably is We have no particular plans to change this behavior at this time.  It would require storing separate extended attribute sets for each signature of the overload set, and many extended attributes don't make sense on a per-overload basis.

In this case, you can simply put LenientFloat on the second overload, of course.

Comment 2

[Boris Zbarsky [:bzbarsky]]

Er, ignore the "This probably is" bit above.

Comment 3

[Milan Sreckovic [:milan] (needinfo for best results)]

Yes, not really blocking any work.  On a side note, can't actually put LenientFloat on the second override, we get a different error :)

Comment 4

[Boris Zbarsky [:bzbarsky]]

> we get a different error

Er... which one?

Comment 5

[(no longer active)]

(In reply to Boris Zbarsky [:bz] from comment #4)
> > we get a different error
> 
> Er... which one?

"[LenientFloat] used on an operation with no restricted float type arguments" from WebIDL.py.

Comment 6

[Boris Zbarsky [:bzbarsky]]

> "[LenientFloat] used on an operation with no restricted float type arguments"

Ah.  We should relax that for the overload case, I guess, and either do that check once we have all the overloads or allow [LenientFloat] to be specified on some overloads but not others and make sure it ends up on the operation if it's on any of the overloads.

Comment 7

[Firefox Bug Husbandry Bot]

https://bugzilla.mozilla.org/show_bug.cgi?id=1472046

Move all DOM bugs that haven’t been updated in more than 3 years and has no one currently assigned to P5.

If you have questions, please contact :mdaly.

Comment 8

[Kagami Rosylight [:saschanaz] (they/them)]

Attachment: Bug 1020975: Allow LenientFloat to be only in a specific overload

This assumes that LenientFloat won't break float-less overload behavior.

Comment 9

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/5161e20b698e
Allow LenientFloat to be only in a specific overload r=bzbarsky

Comment 10

[Andrei Ciure[:aciure]]

https://hg.mozilla.org/mozilla-central/rev/5161e20b698e