It turns out that other engines also prefer <h1 id=x> over <a name=x>.
So we can just use get_element_by_id() directly without worrying about
the type of element we find.
It turns out that other engines prefer <a id> over <a name> when
deciding which anchor element to jump to.
This patch aligns LibHTML's behavior with WebKit and Gecko.
Thanks to "/cam 2" for bringing this up. :^)
These will be useful for implementing various things. They don't do any
caching at the moment, but that might become valuable in the future.
To facilitate this change, I also made it possible to abort a tree walk
with for_each_in_subtree() by returning IterationDecision::Break from
the callback.
After the splitting-into-lines pass, remove any trailing whitespace
from all of a block's line boxes.
This improves the appearance of text-align: justify/right :^)
In order for this to work nicely, I made the line box classes use float
instead of int for its geometry information.
Justification works by distributing all of the whitespace on the line
(including the trailing whitespace before the line break) evenly across
the spaces in-between words.
We should probably use floating point (or maybe fixed point?) for all
the layout metrics stuff. But one thing at a time. :^)
Remove the Document pointer from HtmlView and always get to it through
the main Frame instead.
The idea here is to move towards HtmlView being higher-level than the
DOM stuff (as much as possible and practical.)
This patch implements basic support for <a href="#foo"> fragment links.
To figure out where we actually want to scroll to, we have to do
something different based on the layout node's box type. So if it's a
regular LayoutBox we can just use the LayoutBox::position().
However, if it's an inline layout node, we use the position of the
first line box fragment in the containing block contributed by this
layout node or one of its descendants.
Basically the same exact fix as I did for replaced elements. There's no
point in inserting a line break at the start of a line if all you're
trying to achieve is make more horizontal space for something.
HtmlView will now invoke the on_link_hover hook when the cursor enters
or leaves a DOM node that has an enclosing link element.
This patch also updates the meaning of Node::enclosing_link_element()
to find the nearest HTMLAnchorElementAncestor *with an href attribute*.
Use a zero-timer to schedule a style update after invalidating style
on any node. Nodes now have a needs_style_update flag which helps us
batch and coalesce the work.
We also start style updates at the root and work our way through the
document, updating any node that has the needs_style_update flag set.
This is slower than what we were doing before, but far more correct.
There is a ton of room for improvement here. :^)
It's now possible to set a page background image via <body background>.
Also, HtmlView now officially handles rendering the body element's
background (color, image or both.) LayoutBox is responsible for all
other background rendering.
Note that it's not yet possible to use CSS background-image properties
directly, since we can't parse them yet. :^)
If the current line box already has zero width, there's no point in
inserting a line break to make space, since we'll just be at x=0 after
breaking as well.
This removes an ugly unnecessary line break before images wider than
their containing block. :^)
Capture a weak pointer to the element and pass that to the load finish
callback in HTMLImageElement::load_image(). This allows us to ignore
completed loads if the <img> element is no longer around.
When loading a URL that ends in ".png", we now construct a simple
DOM document to contain the image. It also shows the image dimensions
in the document title.
Because we use <img src> to load the image into the synthetic document,
we end up loading the image resource twice. This issue will go away
once we have a smarter, caching, loader mechanism.
The FontCache caches the result of font lookups. The cache key is a
simple object called FontSelector which consists of the font family
and font weight (both strings.)
This drastically reduces time spent in font lookup.
Some inline stylesheets use HTML comments like "<!--blah blah-->".
The HTML parser currently generates a comment child node of the <style>
element whenever this happens, and we don't want the comment itself to
be interpreted as part of the stylesheet.
This is a total hack to get around the auto-detection mechanism for
whether a block has inline or block children. We'll say that tables
never have inline children for now, and then anything that actually
turns out to be an inline child will just be ignored by layout.
Instead of computing whether a block's children are inline based on the
first child, make it an imperatively-set flag.
This gives us some flexibility to ignore things like text nodes inside
a <table>, for example. I'm still unsure what the "correct" way to deal
with those will be. We'll find out sooner or later. :^)
This class introduces LayoutTable, LayoutTableRow and LayoutTableCell.
These are produced by "display" values table, table-row and table-cell
respectively.
Note that there's no layout happening yet, I'm just adding the classes.
We now wait until the pixels are actually needed before fully decoding
images in <img> elements.
This needs some more work and is currently a bit memory-wasteful since
we'll hang on to the raw image data forever.
If a LayoutNode is split into line box fragments, we need to walk our
fragments and invalidate them. It was not enough to do this only for
LayoutBox nodes.
To streamline the layout tree and remove irrelevant data from classes
that don't need it, this patch adds two new LayoutNode subclasses.
LayoutNodeWithStyleAndBoxModelMetrics should be inherited by any layout
node that cares about box model metrics (margin, border, and padding.)
LayoutBox should be inherited by any layout node that can have a rect.
This makes LayoutText significantly smaller (from 140 to 40 bytes) and
clarifies a lot of things about the layout tree.
I'm also adding next_sibling() and previous_sibling() overloads to
LayoutBlock that return a LayoutBlock*. This is okay since blocks only
ever have block siblings.
Do also note that the semantics of is<T> slightly change in this patch:
is<T>(nullptr) now returns true, to facilitate allowing to<T>(nullptr).
