Define DocumentOrShadowRoot.activeElement

[rakina]

• At least two implementers are interested (and none opposed):
  • Blink
  • Gecko & Safari are reviewing the other PRs, so marking as interested
• Tests are written and can be reviewed and commented upon at:
  • HTML: Add WPT for DocumentOrShadowRoot.activeElement web-platform-tests/wpt#18447
• Implementation bugs are filed:
  • Chrome: already implemented in Chrome
  • Firefox: already implemented in Firefox
  • Safari: already implemented in Safari

Another part of #2013 and whatwg/dom#768. Moves Document.activeElement to DocumentOrShadowRoot.activeElement

---

/dom.html ( diff )
/index.html ( diff )
/infrastructure.html ( diff )
/interaction.html ( diff )

[rakina]

cc people involved in previous PRs @rniwa @annevk @domenic

[domenic]

I pushed some editorial tweaks, but I'd like your help fixing up the domintro for web developers.

[rniwa]

FWIW, we've had this implemented in WebKit for a while:
https://trac.webkit.org/browser/webkit/trunk/Source/WebCore/dom/DocumentOrShadowRoot.idl?annotate=blame#L36

[rakina]

FWIW, we've had this implemented in WebKit for a while:
https://trac.webkit.org/browser/webkit/trunk/Source/WebCore/dom/DocumentOrShadowRoot.idl?annotate=blame#L36

Thanks, updated the first post!

[EdgarChen]

We have had this implemented in Gecko: https://searchfox.org/mozilla-central/rev/30b01f4f60dbcbd6b01500a26b3100c28005cf62/dom/webidl/DocumentOrShadowRoot.webidl#28.

[rniwa]

I guess activeElement's behavior won't be affected by delegatesFocus?

[rniwa]

Hm... now that I'm going through some of the tests @rakina wrote, I see that ShadowRoot.activeElement wouldn't point to a slot even if the element inside the slot had a focus. This is kind of a shitty behavior but it'll probably pose too much compatibility risk to change that now. :(

[rniwa]

Oh I guess I should have waited for this PR to be merged before merging the WPT PR. But it looks like the remaining issues is really an editorial one so we should just fix that up & merge the PR.

[domenic]

Some algorithm tweaks needed. Also, I am curious about the fullscreen connection, e.g. maybe we could use retargeting but then return null if not in the same tree, like fullscreen does.

[rakina]

Some algorithm tweaks needed. Also, I am curious about the fullscreen connection, e.g. maybe we could use retargeting but then return null if not in the same tree, like fullscreen does.

Oh hey you're right, we can actually just retarget and check if the result's root is the DocumentOrShadowRoot! Thank you :D