For clarity, I introduce <LayoutDom<Element>>::parent_node_ref to contain
the remaining unsafety bits out of composed_parent_node_ref which is more
complex than just a field access.
They don't do anything fancy so there is no additional unsafety calling them
compared to using LayoutDom in the first place, the usual story of all
those changes.
Implement dirname support for form element
Added support for dirname in input on form submit
Added Dir getter / setter for HTMLElement
NOT YET Added get directionality according to https://html.spec.whatwg.org/multipage/dom.html#the-directionality
- [X] `./mach build -d` does not report any errors
- [x] `./mach test-tidy` does not report any errors
- [x] These changes fix#25379 (GitHub issue number if applicable)
Add SupportedPropertyNames to Document (also fix iframe getting)
Existing test of named-getting an iframe now succeeds. I added a new test for Object.getOwnPropertyNames(document) based on my understanding of the spec; that test could use a second opinion.
UPDATE: This was trying to do too many things in one PR as originally submitted. It is now using #25572 as a base, and I suggest reviewing that PR before this one to avoid duplicating review effort.
---
<!-- Thank you for contributing to Servo! Please replace each `[ ]` by `[X]` when the step is complete, and replace `___` with appropriate data: -->
- [X] `./mach build -d` does not report any errors
- [X] `./mach test-tidy` does not report any errors
- [X] These changes fix#7273 for all implemented named getters, fix#25146, and fix the iframe case only of #25145.
<!-- Either: -->
- [X] There are tests for these changes
<!-- Also, please make sure that "Allow edits from maintainers" checkbox is checked, so that we can help you if you get stuck somewhere along the way.-->
<!-- Pull requests that do not address these steps are welcome, but they will require additional verification as part of the review process. -->
Attr is a Node, with consequences for many Node methods
<!-- Please describe your changes on the following line: -->
Attr is now a Node, as current WHATWG specs require. I think I did some unidiomatic things to make compareDocumentPosition work and it could use a look by someone more familiar with how to write concise Rust code. I also think the new cases in compareDocumentPosition lack tests; it is possible to compare the position of two attributes, or of an attribute and an element, and I don't think any tests are exercising that functionality.
---
<!-- Thank you for contributing to Servo! Please replace each `[ ]` by `[X]` when the step is complete, and replace `___` with appropriate data: -->
- [X] `./mach build -d` does not report any errors
- [X] `./mach test-tidy` does not report any errors
- [X] These changes fix#25124
<!-- Either: -->
- [ ] There are tests for these changes (existing cases of Node methods are well-tested, and there is a WPT test specifically for whether Attr is Node, but I'm not sure about new Node method cases)
<!-- Also, please make sure that "Allow edits from maintainers" checkbox is checked, so that we can help you if you get stuck somewhere along the way.-->
<!-- Pull requests that do not address these steps are welcome, but they will require additional verification as part of the review process. -->
There are a few canvas2d-related dependencies that haven't updated, but they
only use euclid internally so that's not blocking landing the rest of the
changes.
Given the size of this patch, I think it's useful to get this landed as-is.
registered_mutation_observers() returns immutable references and does
not init mutation observers.
registered_mutation_observers_mut() lazily initializes raredata if it
does not exist.
Updated code that uses this methods to call appropriate method when
mutation is not necessary.
