Instead of firing up a network request and synchronously blocking for it
to finish via a nested event loop, we now start an asynchronous request
when encountering <script src>.
Once the script load finishes (or fails), it gets executed at one of the
synchronization points in the HTML parser.
This solves some long-standing issues with random unexpected events
getting dispatched in the middle of parsing.
The HTML spec tells has some special rules for <body> and <frameset>
elements' onfoo event handler attributes. In some cases, the implicitly
generated event listeners should end up on the relevant global object
instead of the element itself.
This patch implements the first part of that behavior.
This logic was kept in the GlobalEventHandlers mixing for sharing
between Document and HTMLElement, but there are other interfaces who
need to support `onfoo` attribute event listeners as well.
Rather than destroying and rebuilding the entire document layout tree in
every call to `ComputedCSSStyleDeclaration::property()`, we now just
make sure that the layout tree exists.
This speeds up the DOM Inspector significantly, from taking several
seconds to select an element, to almost instant. :^)
The previous implementation was about a half implementation and was
tied to Element::innerHTML. This separates it and puts it into
HTMLDocumentParser, as this is in the parsing section of the spec.
This provides a near finished HTML fragment serialisation algorithm,
bar namespaces in attributes and the `is` value.
Since we don't have a direct connection to WindowServer, this is slighly
more naive implementation than what we were doing for single-process
mode.
Basically, there's a global RequestAnimationFrameDriver object that
has a 16ms single-shot timer. Whenever someone registers for a RAF
callback, we start the timer (if needed).
This is not ideal, but it's better than nothing. :^)
This namespace will be used for all interfaces defined in the URL
specification, like URL and URLSearchParams.
This has the unfortunate side-effect of requiring us to use the fully
qualified AK::URL name whenever we want to refer to the AK class, so
this commit also fixes all such references.
getComputedStyle(element) now returns a ComputedCSSStyleDeclaration
object, which is a live view of the computed style of a given element.
This works by ComputedCSSStyleDeclaration being a wrapper around an
element pointer. When you ask it for a CSS property, it gets the latest
computed style values from the element and returns them as a
CSS::StyleProperty object.
This first cut adds support for computed 'color' and 'display'.
In case the element doesn't have a corresponding node in the layout
tree, we fall back to using specified style instead. This is achieved by
performing an on-the-fly style resolution for the individual element and
then grabbing the requested property from that resolved style.
This patch moves the CSS property+value storage down to a new subclass
of CSSStyleDeclaration called PropertyOwningCSSStyleDeclaration.
The JavaScript wrapper for CSSStyleDeclaration now calls virtual
functions on the C++ object.
This is preparation for supporting computed style CSSStyleDeclaration
objects which won't have internal property storage, but rather an
internal element pointer. :^)
The spec allows us to optionally return from these for any reason.
Our reason is that we don't have all the infrastructure in place yet to
implement them.
This just returns an empty CSSStyleDeclaration for now. The real thing
needs to be a live object that provides a view onto the computed style
of a given element. This is far from that, but it's something. :^)
Let's err on the side of caution and use a WeakPtr instead of a
DOM::Document&. Even if the object lifetimes involved here should be
well-defined, they are fairly complicated.
This will be used by the HTML parser to determine whether it's okay to
start running a script.
Note that we don't actually count the script-blocking style sheets yet.
This patch only adds a the checking mechanism for the parser.
This roughly models the "queue an element task" algorithm from the spec.
For safety, this captures a strong reference to the element, and then
bundles that with a callback into a HTML::Task (that we then queue up.)
The current implementation felt a bit ad-hoc and notably allowed
textContent to operate on all node types. It also only returned the
child text content of the Node instead of the descendant text content.
