I don't think it will be possible to avoid splitting fragments in the
presence of `vertical-align`, because one `ScannedTextFragment` could
potentially be split into arbitrary many fragments, each having its own
vertical position that can influence layout of other fragments.
This code also removes parts of `Range` that were no longer used.
r? @glennw
This makes layout more idempotent, which is important for incremental
layout.
Also converts `is_root` to a set of flags and fixes a `TODO` concerning
percentage inline heights of images.
r? @glennw
splitting fragments.
I don't think it will be possible to avoid splitting fragments in the
presence of `vertical-align`, because one `ScannedTextFragment` could
potentially be split into arbitrary many fragments, each having its own
vertical position that can influence layout of other fragments.
This code also removes parts of `Range` that were no longer used.
writing it to a temporary location and overwriting it.
This makes layout more idempotent, which is important for incremental
layout.
Also converts `is_root` to a set of flags and fixes a `TODO` concerning
percentage inline heights of images.
This also makes command line options available as a global. If we're happy with that change I will go through the rest of the code and update it to avoid passing and cloning the Opts structure.
Extra size from margins should be included in block size, so that
layers are large enough to include the entire block. This is typically
hidden by large tile sizes (512x512), but fitted tiles makes the issue
a lot more common.
According to the documentation for Fragment::position, the inline axis
should include margin size, so we include it for blocks. Also fix
place_float which assumed that it was not included and
assign_inline_sizes which overrode the size set in
set_inline_size_constraint_solutions.
Typically this issue was hidden by large tile sizes, but fitted tiles
makes it more common.
This is missing a lot of parts, so it doesn't really make any real-world form submission work yet. It provides a skeleton on which the missing bits can be filled in.
What works:
- `<input>` elements except for `type=file`
- GET/POST methods
- URLencoded `enctype`s (default)
- Submission via `<form>.submit()` only
Stuff that needs to be done for most simple real-world cases to work:
- [Working text input](https://github.com/servo/servo/pull/3585)
- Click handlers for `<submit>`
- Possibly `<textarea>` support
- Support for the other two enctypes (#3649)
Todo:
- Correctly implement [planned navigation](https://html.spec.whatwg.org/multipage/forms.html#planned-navigation) using `TrustedFormAddress` (#3648)
- [Correctly implement form owners.](https://github.com/servo/servo/issues/3553) Requires html5ever and some discussion of the spec.
- `<input type=file>` support
- Image submit support
- Browsing contexts/targets
- Support for non-`<input>` controls
- Validation (?)
- Dirname (?)
stretching.
This preserves the usage of the Bloom filter throughout style recalc,
but the implementation is rewritten. Provides a 15% improvement on
Guardians of the Galaxy.
stretching.
This preserves the usage of the Bloom filter throughout style recalc,
but the implementation is rewritten. Provides a 15% improvement on
Guardians of the Galaxy.
DOM traversals and Flow traversals look very similar. This patch unifies them
with the preorder/postorder pattern. Hopefully, it also opens the door for writing
the traversal code only once, instead of the duplication we have today.
This patch puts in the initial framework for incremental reflow. Nodes' styles
are no longer recalculated unless the node has changed.
I've been hacking on the general problem of incremental reflow for the past
couple weeks, and I've yet to get a full implementation that actually passes all
the reftests + wikipedia + cnn. Therefore, I'm going to try to land the different
parts of it one by one.
This patch only does incremental style recalc, without incremental flow
construction, inline-size bubbling, reflow, or display lists. Those will be coming
in that order as I finish them.
At least with this strategy, I can land a working version of incremental reflow,
even if not yet complete.
r? @pcwalton
This patch puts in the initial framework for incremental reflow. Nodes' styles
are no longer recalculated unless the node has changed.
I've been hacking on the general problem of incremental reflow for the past
couple weeks, and I've yet to get a full implementation that actually passes all
the reftests + wikipedia + cnn. Therefore, I'm going to try to land the different
parts of it one by one.
This patch only does incremental style recalc, without incremental flow
construction, inline-size bubbling, reflow, or display lists. Those will be coming
in that order as I finish them.
At least with this strategy, I can land a working version of incremental reflow,
even if not yet complete.
r? @pcwalton
