Give browsing contexts an explicit container

[annevk]

This is the first step for #5020.

If #5089 is merged first I can remove all "nested through" usage and fix #4409 I think.

---

/browsers.html ( diff )
/parsing.html ( diff )
/webappapis.html ( diff )

[annevk]

@shivanigithub this could also help with top-level origin. That is, through a document's browsing context's container document we can get to the actual parent of a document. If we keep doing that we hit the top-level document. We can then get the true top-level origin that's not tainted by bfcache.

In particular, if you have document A1, which nests A2, which nests A3. And then you navigate A2 to B. And then you navigate A1 to C.

If you go through A1's (or B's) parent documents using the above pattern you'll get A1. If you use top-level browsing context's active document you'll get C, which is bad.

[domenic]

Seem to be some issues around disconnected iframes, and the related switch from nested -> child in various places.

[domenic]

Let's be sure to note in the commit message that this updates to account for shadow trees

[domenic]

Oh, shoot. I think I was under the mistaken impression that the nested browsing context stays set, even after an iframe gets removed. So a lot of my comments are wrong, from that perspective, and this PR does not make normative changes in the way I thought it did.

I think the main thing would be to straighten out the child BC definition, and then I can take another look at this with the correct frame of mind.

[annevk]

I'm not really sure what you're asking for at this point.

[domenic]

I've resolved all the comments that were resulting from my confusion and added an idea on how to make it less confusing.

[annevk]

Proposed commit message:

Give browsing context an explicit container

This generally sets out the infrastructure for a more secure model for working with browsing contexts, as outlined in #5020.

* Adds browsing context's container
* Adds browsing context's container document
* Makes "update the rendering" account for browsing contexts in shadow trees
* Removes nested through; closes #4409.

I think #1336 is also fixed now?

(Note to self: rebase #5095 once this lands.)

[domenic]

I don't think this fully takes care of everything in #1336. In particular there are still uses of "child browsing contexts" which are associated with a Document, not another BC.

I think these are the only two instances:

Discard all the child browsing contexts of document.

and

The term plugin refers to a user-agent defined set of content handlers used by the user agent that can take part in the user agent's rendering of a Document object, but that neither act as child browsing contexts of the Document nor introduce any Node objects to the Document's DOM.

(with the first being more problematic, since the second is somewhat informal).