137687 - Computed CSSDeclarations need to implement GetCssText

Status: NEW

(Core :: DOM: CSS Object Model, defect, P3)

Description

[Boris Zbarsky [:bzbarsky]]

nsComputedDOMStyle does not implement GetCssText....

Comment 1

[Christopher Aillon (sabbatical, not receiving bugmail)]

Attachment: Fix

This was originally part of my fix to spin out the queryable properties into
the static map, but I left it out of that patch and I forgot about it until
now.

Comment 2

[Boris Zbarsky [:bzbarsky]]

So... we need to canonicalize and do shorthands and the like if we do this,
imo.... (though we could land that as a start, I suppose).

Comment 3

[Boris Zbarsky [:bzbarsky]]

No plans to work on this any time in the foreseeable future, so to default owner.

Comment 5

[Kartikaya Gupta (email:[e-mail])]

Updating title to make this bug slightly easier to find.

Comment 6

[willmarquardt]

A fix for this would be very useful. In the meantime, here's how to work around it: https://gist.github.com/johnkpaul/1754808

Comment 7

[Chris Lord [:cwiiis]]

The attached gist doesn't work, but this does:

function getComputedStyleCssText(element) {
  var style = window.getComputedStyle(element), cssText;
 
  if (style.cssText != "") {
    return style.cssText;
  }
 
  cssText = "";
  for (var i = 0; i < style.length; i++) {
    cssText += style[i] + ": " + style.getPropertyValue(style[i]) + "; ";
  }
  
  return cssText;
}

Comment 8

[Boris Zbarsky [:bzbarsky]]

This looks like it's causing web compat issues.  See https://webcompat.com/issues/3822

Xidorn, do you want to try resurrecting the patch here?  Looks to me like Blink doesn't do any interesting shorthand stuff, so maybe we can just get away with the simple and silly implementation...

Comment 9

[Karl Dubost💡 :karlcow]

Thanks Boris for clarifying/unmixing the two issues

Comment 10

[Xidorn Quan [:xidorn] UTC+11]

This is crazy... Why do they need that? Couldn't we contact google and ask them not to do that?

I can work on the patch... but how should the spec say? Edge returns empty string for computed style's cssText as well. Could we just make the spec do this and ask Blink to change their behavior?

Comment 11

[Boris Zbarsky [:bzbarsky]]

> Why do they need that?

They're trying to copy the computed style of one node to the inline style of another afaict.

> Couldn't we contact google and ask them not to do that?

Karl has.  He asked them to do the getPropertyValue/setPropertyvalue thing.

> but how should the spec say? 

Spec says to do https://drafts.csswg.org/cssom/#serialize-a-css-declaration-block but that's ... complicated and annoying.

> Could we just make the spec do this and ask Blink to change their behavior?

And Safari (they match Blink; presumably this is still WebKit code)?  We could try, but in the meantime Google drive is slow.  :(

Comment 12

[Xidorn Quan [:xidorn] UTC+11]

Files a spec issue for this: https://github.com/w3c/csswg-drafts/issues/1033

> And Safari (they match Blink; presumably this is still WebKit code)?  We
> could try, but in the meantime Google drive is slow.  :(

The quickest way should be asking Google drive to fix I think....

Comment 13

[Karl Dubost💡 :karlcow]

>(In reply to Xidorn Quan [:xidorn] (UTC+10) from comment #12)
> The quickest way should be asking Google drive to fix I think....

This has been done today. This doesn't guarantee it will be. Some issues are sometimes never fixed, sometimes fixed one year later or more (because of a redesign). Let's give them one week.

Comment 14

[Firefox Bug Husbandry Bot]

Migrating Webcompat whiteboard priorities to project flags. See bug 1547409.

Comment 15

[Firefox Bug Husbandry Bot]

See bug 1547409. Migrating whiteboard priority tags to program flags.