http://www.robohornet.org gives a score of 101.36 on master,
and 102.68 with this PR. The latter is slightly better,
but probably within noise level.
So it looks like this PR does not affect DOM performance.
This is expected since `Box::new` is defined as:
```rust
impl<T> Box<T> {
#[inline(always)]
pub fn new(x: T) -> Box<T> {
box x
}
}
```
With inlining, it should compile to the same as box syntax.
In a later PR, DomRoot<T> will become a type alias of Root<Dom<T>>,
where Root<T> will be able to handle all the things that need to be
rooted that have a stable traceable address that doesn't move for the
whole lifetime of the root. Stay tuned.
This should make it somewhat easier to experiment with alternative
representations in the future. To reduce churn, this commit leaves the String
field public, though.
Also, this will allow us to use the default String type to represent the IDL
USVString type, which explicitly forbids unpaired surrogates, ans as such is
a better match to the Rust String type.
This trait is used to hold onto the downcast and upcast functions of all
castable IDL interfaces. A castable IDL interface is one which either derives
from or is derived by other interfaces.
The deriving relation is represented by implementations of marker trait
DerivedFrom<T: Castable> generated in InheritTypes.
/^[ ]*use dom::bindings::codegen::InheritTypes::.*(Base|Cast|Derived)/ {
/::[a-zA-Z]+(Base|Cast|Derived);/d
s/([{ ])[a-zA-Z]+(Base|Cast|Derived), /\1/g
s/([{ ])[a-zA-Z]+(Base|Cast|Derived), /\1/g
s/, [a-zA-Z]+(Base|Cast|Derived)([},])/\2/g
s/, [a-zA-Z]+(Base|Cast|Derived)([},])/\2/g
/\{([a-zA-Z]+(Base|Cast|Derived))?\};$/d
s/\{([a-zA-Z_]+)\};$/\1;/
}
s/([a-zA-Z]+)Cast::from_ref\(\&?\**([a-zA-Z_]+)(\.r\(\))?\)/\2.upcast::<\1>()/g
s/([a-zA-Z]+)Cast::from_ref\(\&?\**([a-zA-Z_]+)(\.[a-zA-Z_]+\(\))?\)/\2\3.upcast::<\1>()/g
s/\(([a-zA-Z]+)Cast::from_ref\)/\(Castable::upcast::<\1>\)/g
s/([a-zA-Z]+)Cast::from_root/Root::upcast::<\1>/g
s/([a-zA-Z]+)Cast::from_layout_js\(\&([a-zA-Z_.]+)\)/\2.upcast::<\1>()/g
s/([a-zA-Z]+)Cast::to_ref\(\&?\**([a-zA-Z_]+)(\.r\(\))?\)/\2.downcast::<\1>()/g
s/([a-zA-Z]+)Cast::to_ref\(\&?\**([a-zA-Z_]+)(\.[a-zA-Z_]+\(\))?\)/\2\3.downcast::<\1>()/g
s/\(([a-zA-Z]+)Cast::to_ref\)/\(Castable::downcast::<\1>\)/g
s/([a-zA-Z]+)Cast::to_root/Root::downcast::<\1>/g
s/([a-zA-Z]+)Cast::to_layout_js\(&?([a-zA-Z_.]+(\(\))?)\)/\2.downcast::<\1>()/g
s/\.is_document\(\)/.is::<Document>()/g
s/\.is_htmlanchorelement\(\)/.is::<HTMLAnchorElement>()/g
s/\.is_htmlappletelement\(\)/.is::<HTMLAppletElement>()/g
s/\.is_htmlareaelement\(\)/.is::<HTMLAreaElement>()/g
s/\.is_htmlbodyelement\(\)/.is::<HTMLBodyElement>()/g
s/\.is_htmlembedelement\(\)/.is::<HTMLEmbedElement>()/g
s/\.is_htmlfieldsetelement\(\)/.is::<HTMLFieldSetElement>()/g
s/\.is_htmlformelement\(\)/.is::<HTMLFormElement>()/g
s/\.is_htmlframesetelement\(\)/.is::<HTMLFrameSetElement>()/g
s/\.is_htmlhtmlelement\(\)/.is::<HTMLHtmlElement>()/g
s/\.is_htmlimageelement\(\)/.is::<HTMLImageElement>()/g
s/\.is_htmllegendelement\(\)/.is::<HTMLLegendElement>()/g
s/\.is_htmloptgroupelement\(\)/.is::<HTMLOptGroupElement>()/g
s/\.is_htmloptionelement\(\)/.is::<HTMLOptionElement>()/g
s/\.is_htmlscriptelement\(\)/.is::<HTMLScriptElement>()/g
s/\.is_htmltabledatacellelement\(\)/.is::<HTMLTableDataCellElement>()/g
s/\.is_htmltableheadercellelement\(\)/.is::<HTMLTableHeaderCellElement>()/g
s/\.is_htmltablerowelement\(\)/.is::<HTMLTableRowElement>()/g
s/\.is_htmltablesectionelement\(\)/.is::<HTMLTableSectionElement>()/g
s/\.is_htmltitleelement\(\)/.is::<HTMLTitleElement>()/g
Given codegen now generates the various TypeId enums, it seems pointless to
still have to write their respective values in every DOM struct inheriting from
Node just to set the initial IS_IN_DOC flag in Document and IN_ENABLED_STATE in
form controls.
NodeTypeId is supposed to reflect the WebIDL inheritance hierarchy.
All of Text/ProcessingInstruction/Comment inherit from CharacterData,
which inherits from Node. There should be a CharacterDataTypeId value
that differentiates between those, instead.
