Editorial: Upstream outerHTML property from DOM Parsing

[lukewarlow]

Follow up to #10264

Corresponding DOM Parsing spec PR: w3c/DOM-Parsing#80

---

/dynamic-markup-insertion.html ( diff )
/index.html ( diff )
/infrastructure.html ( diff )
/scripting.html ( diff )

[lukewarlow]

cc @domenic just as an fyi this one should be ready to go. But no worries if you don't have review capacity for it :)

[domenic]

Excited to help finish the job!

[domenic]

I just realized I gave bad advice on creating the body element and appending it. Because appending the node will move it away from wherever it was originally. That's no good!

We may want to just revert to the fictitious node thing and fix it later. But I dug a bit deeper into how we could make this work. If you're up for it...

For HTML, we basically need step 5.2 "Element" case of #html-fragment-serialisation-algorithm extracted.

For XML, I think we use https://w3c.github.io/DOM-Parsing/#dfn-xml-serializing-an-element-node . Although I guess there's some state that is set up in https://w3c.github.io/DOM-Parsing/#dfn-xml-serialization and is normally implicitly passed, which would need to be passed through?

(Also I just realized that #serialising-xhtml-fragments seems to be a vague version of DOM-Parsing's algorithm with no references in the specification... ugh.)

[lukewarlow]

I think our best bet here is to revert to the fictious node for now, and then once more of DOM Parsing is upstreamed (working on insertAdjacentHTML next) we can go through and see how much we can consolidate and abstract between the HTML algorithms and the DOM Parsing ones.

[lukewarlow]

I just realized I gave bad advice on creating the body element and appending it. Because appending the node will move it away from wherever it was originally. That's no good!

We may want to just revert to the fictitious node thing and fix it later.

@domenic this should be done now. As I say I think once have more of the DOM Parsing spec in html we'll have a bigger picture on the best way to improve things.