Treat fragment parser documents as not connected

[shannonbooth]

Matching the behaviour of engines. For: whatwg/html#11023, but needs a small patch in HTML spec also, which I will raise if directionally this is okay.

• At least two implementers are interested (and none opposed):
  • Chromium (already implements)
  • Webkit (already implements)
  • Firefox (already implements)
• Tests are written and can be reviewed and commented upon at:
  • https://github.com/web-platform-tests/wpt/blob/0d0249113dc429fbbb4fce7778caeceae0466739/domparsing/createContextualFragment.html#L85
  • https://github.com/web-platform-tests/wpt/blob/0d0249113dc429fbbb4fce7778caeceae0466739/html/semantics/forms/the-input-element/radio-disconnected-group-owner.html#L186
• Implementation bugs are filed:
  • Chromium: N/A
  • Gecko: N/A
  • WebKit: N/A
  • Deno (only for aborting and events): …
  • Node.js (only for aborting and events): …
• MDN issue is filed: N/A
• The top of this comment includes a clear commit message to use.

(See WHATWG Working Mode: Changes for more details.)

---

Preview | Diff

[annevk]

I don't like changing the meaning of connected for this. It feels a bit like a layering violation. Also, as discovered in whatwg/html#11669 (comment) I don't think this will actually do the trick long term as not all browsers create an intermediate document.

I wonder if @noamr and @foolip have ideas on how to address this properly.

[noamr]

I don't like changing the meaning of connected for this. It feels a bit like a layering violation. Also, as discovered in whatwg/html#11669 (comment) I don't think this will actually do the trick long term as not all browsers create an intermediate document.

I wonder if @noamr and @foolip have ideas on how to address this properly.

I think a more consistent approach would be to insert elements directly to the DocumentFragment rather than inserting them to the inert document and then moving them to the fragment in bulk.
Then they wouldn't be "connected" in the first place and we don't need a DOM change.

[shannonbooth]

Indeed it feels a bit yuck, though it doesn't seem possible to change in a nice way as long as connected is defined in terms of document and fragment parsing requires a document. If we can wire up fragment parsing without a document that indeed seems much nicer, I can try and investigate that approach (unless this is already being worked on based on what is mentioned above?)

[noamr]

Indeed it feels a bit yuck, though it doesn't seem possible to change in a nice way as long as connected is defined in terms of document and fragment parsing requires a document. If we can wire up fragment parsing without a document that indeed seems much nicer, I can try and investigate that approach (unless this is already being worked on based on what is mentioned above?)

An element inserted to a DocumentFragment is never connected so I think it would solve this.

Note that multiple in-flight and upcoming changes like whatwg/html#11818 are doing this work through redirecting the insertion target (see https://html.spec.whatwg.org/multipage/parsing.html#appropriate-place-for-inserting-a-node). This would probably be the place to fix it as well. I would prefer if we do this after these in-flight changes are complete though.

[WebReflection]

An element inserted to a DocumentFragment is never connected so I think it would solve this.

if persistent document fragment will ever exist, the element is connected "if the fragment is" so please consider the fact in the future there could be live/persistent document fragments, thank you!

[noamr]

An element inserted to a DocumentFragment is never connected so I think it would solve this.

if persistent document fragment will ever exist, the element is connected "if the fragment is" so please consider the fact in the future there could be live/persistent document fragments, thank you!

Sure, but in this case it wouldn't be that future type of fragment, but a normal disconnected fragment.

[shannonbooth]

Indeed it feels a bit yuck, though it doesn't seem possible to change in a nice way as long as connected is defined in terms of document and fragment parsing requires a document. If we can wire up fragment parsing without a document that indeed seems much nicer, I can try and investigate that approach (unless this is already being worked on based on what is mentioned above?)

An element inserted to a DocumentFragment is never connected so I think it would solve this.

Note that multiple in-flight and upcoming changes like whatwg/html#11818 are doing this work through redirecting the insertion target (see https://html.spec.whatwg.org/multipage/parsing.html#appropriate-place-for-inserting-a-node). This would probably be the place to fix it as well. I would prefer if we do this after these in-flight changes are complete though.

Sounds good, I'll hold off!