Digging into several crashing tests revealed that committing
transactions is a fallible operation. Propagating those errors led to
exposing many new errors caused by the IDBRequest implementation
assuming that all successful responses contained a structured clone. The
end result is a bunch of new test failures that were previously hidden.
Testing: Existing test coverage is sufficient.
---------
Signed-off-by: Josh Matthews <josh@joshmatthews.net>
Introduce a safe wrapper trait for the unsafe `ToJSValConvertible`, and
use it in `script/dom` where the default `T` implementation works.
Part of https://github.com/servo/servo/issues/37951
---------
Signed-off-by: gterzian <2792687+gterzian@users.noreply.github.com>
These changes introduce a new kind of task that uses a variation of the
`task!` syntax. Existing `task!` usages create task structs that have a
`Send` bound, which requires the use of `Trusted<T>` to reference a DOM
object T inside of the task closure. The new syntax replaces the `Send`
bound with a `JSTraceable` bound, which requires explicit capture
clauses with types. This looks like:
```rust
task!(ScriptPrepare: {script: DomRoot<HTMLScriptElement>} |script| {
script.prepare(CanGc::note());
}),
```
The capture clauses must list every value that will be referenced from
the closure's environment—these values are moved into fields in the
generated task structure, which allows them to be traced by the GC as
part of a generated JSTraceable implementation. Since the closure itself
is not a `move` closure, any attempts to reference values not explicitly
captured will generate a borrow checker error since the closure requires
the `'static` lifetime.
Testing: Existing WPT tests exercise these code paths. I also attempted
to write incorrect tasks that capture references or use values not
explicitly captured, and the compiler correctly errors out.
Fixes: part of #35517
Signed-off-by: Josh Matthews <josh@joshmatthews.net>
1. Some rounding was wrongly used which can be a source of inaccuracy.
2. Use `window.InnerWidth` instead of `body.ClientWidth` according to
spec.
Testing: Many new passing cases.
Fixes: Part of #38042.
---------
Signed-off-by: Euclid Ye <euclid.ye@huawei.com>
This refactors some of the cookie retrieval mechanism to be less
repetitive and separates the cookie-list from the cookie-string which
will also be needed for Cookie Store.
Testing: No new behavior, should be covered by existing WPT tests.
Signed-off-by: Sebastian C <sebsebmc@gmail.com>
Renames the files to pythonic file names. Also replaces the bespoke
caching function with the functools caching function which is
functionally equivalent.
---------
Signed-off-by: Ashwin Naren <arihant2math@gmail.com>
BroadcastChannel uses 2 hash maps in the constellation and they were not
kept properly in sync. Also fixed a logic error where
`Option<HashMap<...>>` was not considered empty when being `None`.
Testing: These warnings should not happen anymore:
[2025-07-14T03:54:22Z WARN constellation::constellation] No sender for
broadcast router: (4,1)
[2025-07-14T03:33:59Z WARN constellation_traits::structured_data]
Attempt to broadcast structured serialized data including ImageBitmap
(should never happen).
[2025-07-14T03:33:59Z WARN constellation_traits::structured_data]
Attempt to broadcast structured serialized data including MessagePort
(should never happen).
[2025-07-14T03:33:59Z WARN constellation_traits::structured_data]
Attempt to broadcast structured serialized data including
OffscreenCanvas (should never happen).
[2025-07-14T03:33:59Z WARN constellation_traits::structured_data]
Attempt to broadcast structured serialized data including ReadableStream
(should never happen).
[2025-07-14T03:33:59Z WARN constellation_traits::structured_data]
Attempt to broadcast structured serialized data including WritableStream
(should never happen).
[2025-07-14T03:33:59Z WARN constellation_traits::structured_data]
Attempt to broadcast structured serialized data including
TransformStream (should never happen).
Signed-off-by: webbeef <me@webbeef.org>
WPT tests are expected to create screenshots as soon as everything is
loaded. If an animation is happening adding the "reftest-wait" class to
the root element is appropriate way to delay the screenshot. Previously,
the
test harness was waiting for all animations to finish, but that is just
leading to many timeouts. Removing that code fixes the timeouts.
Two Servo-specific tests are also updated as they were written with
Servo's previous behavior in mind.
Testing: There are test result updates for this change. Many TIMEOUTS
now either correctly
PASS OR FAIL.
Fixes: #36931.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
This change aims to reduce the scope of incremental box tree
construction. Previously, when text content changed, the parent node
would always be marked as requiring box tree reconstruction; now, it is
adjusted to only mark the parent node as needing to recollect its box
tree children.
Testing: This should not change observable behavior and is thus covered
by existing WPT tests.
Signed-off-by: sharpshooter_pt <ibluegalaxy_taoj@163.com>
Enable by default the DOM "ImageBitmap" interface previously gated
behind
"dom.imagebitmap.enabled" preference as experimental web platform
feature.
https://html.spec.whatwg.org/multipage/imagebitmap-and-animations.html#imagebitmap
NOTE: Non blocking non supported functionality:
- vector HTMImageElement, bitmap/vector SVGImageElement
- EXIF image orientation
Testing: Covered by existing WPT tests
Fixes: #34112
Signed-off-by: Andrei Volykhin <andrei.volykhin@gmail.com>
These warnings have been there for quite a while and annoying.
```
Compiling webxr v0.0.1 (D:\servo\components\webxr)
warning: variable does not need to be mutable
--> components\webxr\openxr\mod.rs:557:17
|
557 | let mut context = contexts
| ----^^^^^^^
| |
| help: remove this `mut`
|
= note: `#[warn(unused_mut)]` on by default
warning: field `profile_type` is never read
--> components\webxr\openxr\interaction_profiles.rs:50:9
|
49 | pub struct InteractionProfile<'a> {
| ------------------ field in this struct
50 | pub profile_type: InteractionProfileType,
| ^^^^^^^^^^^^
|
= note: `InteractionProfile` has derived impls for the traits `Debug` and `Clone`, but these are intentionally ignored during dead code analysis
= note: `#[warn(dead_code)]` on by default
```
Testing: No behaviour change.
Signed-off-by: Euclid Ye <yezhizhenjiakang@gmail.com>
This PR fixes the issue where underlines weren't appearing on
whitespaces. This was due to whitespace being ignored in the `glyphs`
function of `components/layout/display_list/mod.rs` when the fragment
didn't have a selection. I added in a check to include the whitespace if
there's a selection or if there are any line decorations. I also renamed
the field from `ignore_whitespace` to `include_whitespace` to make it a
bit clearer since it was being reversed everywhere it was used anyway.
**Before:**
<img width="1235" height="169" alt="image"
src="https://github.com/user-attachments/assets/51d47781-355f-4915-8100-f3a7db81027f"
/>
**After:**
<img width="1235" height="169" alt="image"
src="https://github.com/user-attachments/assets/9b44fe77-d600-4080-9f3a-2c9b33924f51"
/>
Testing: `/css/css-text/white-space/pre-wrap-018.html` is now passing.
Also verified manually by running
`data:text/html;base64,PGRpdiBzdHlsZT0idGV4dC1kZWNvcmF0aW9uOiB1bmRlcmxpbmU7Ij5IZWxsbyBXb3JsZCE8L2Rpdj4=
`
Fixes: https://github.com/servo/servo/issues/33463
---------
Signed-off-by: Leo Ring <leoring03@gmail.com>
Signed-off-by: leo030303 <59373587+leo030303@users.noreply.github.com>
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
Similar to #37960, previously, `AvailHeight`, `AvailWidth`, `Height`,
`Width` ask compositor for screen metrics. This PR moves the request to
embedder.
This simplifies code, and reduces workload of compositor, which is
busier most of time.
Testing: No behaviour change. Updated some tests. `Width/Height` matches
other browsers.
---------
Signed-off-by: Euclid Ye <yezhizhenjiakang@gmail.com>
These changes reduce the number of times we need to query layout for the
same information when creating mouse/pointer events.
Testing: No new tests required for maintaining existing behaviour.
Signed-off-by: Josh Matthews <josh@joshmatthews.net>
This change allows the compiler to recognize if any read-only operations
are missing an implementation when processing a readonly transaction.
Testing: The existing behaviour is unchanged, so current tests suffice.
The new code is unused and cannot be tested.
Fixes: part of #6963
---------
Signed-off-by: Josh Matthews <josh@joshmatthews.net>
This was a bit confusing at first, but the report-only only
had an effect if it was used in conjunction with the regular
CSP header. This is incorrect, as the report-only header
can be present on its own.
Additionally, there was double-logic for parsing the CSP list
values, since we can only concatenate CSP lists if we have
an initial value, which requires a concrete policy value.
Therefore, abstract that way by looping over both headers and
handling the case where initially it is `None` and, if the
CSP header is not present, still `None` when we parse
the `report-only` header.
Additionally, update a WPT test. It was expecting the image
to load, yet was showing the fail image.
Part of #4577
Signed-off-by: Tim van der Lippe <tvanderlippe@gmail.com>
This PR removes existing path(segment) abstractions in favor of
`kurbo::BezPath`, well actually wrapped `kurbo::BezPath`, to ensure
building of valid paths. This allows us better Path2D building in script
and doing all validation and segmentation there and also allows us
remove blocking is_point_in_path on Path2D as we can now do this in
script. Current path is still done on canvas thread side as it will be
harder to move to script (will be done as a follow up), but it now uses
this new path abstraction.
Using kurbo also allows us to ditch our manual svgpath parser with the
one provided by kurbo.
Same code is stolen from: https://github.com/servo/servo/pull/36821.
Testing: Existing WPT tests
Fixes: #37904
wpt run: https://github.com/sagudev/servo/actions/runs/16172191716
---------
Signed-off-by: sagudev <16504129+sagudev@users.noreply.github.com>
Add Constraint DOMException to handle the name conflict error in
`IDBDatabase::createObjectStore` method.
Testing: `./mach test-wpt tests/wpt/tests/IndexedDB/`, but it seems
there are many test failures even in main branch already ([related
comment](https://github.com/servo/servo/pull/37605#issuecomment-2993889163)).
Fixes: #37571
---------
Signed-off-by: iamlockon <xdddxyyyxzzz123@gmail.com>
Found and manually tested by comparing rendering between Firefox and
Servo.
Testing: Existing WPT coverage is sufficient.
Fixes: #37782
Signed-off-by: Josh Matthews <josh@joshmatthews.net>
Previously, `screenX`, `screenY`, `outerHeight`, `outerWidth`, `moveBy`,
`resizeBy` ask compositor for window rectangle, which then return
"inner" rectangle after consulting Embedder.
This PR
1. removes `GetClientWindowRect` from compositor, and directly let
script ask embedder.
2. add `window_size` to `ScreenGeometry`
3. add a lot of docs to `ScreenGeometry`
Testing: `tests\wpt\mozilla\tests\mozilla\window_resizeTo.html` can now
pass for Headed Window.
Fixes: #37824
---------
Signed-off-by: Euclid Ye <yezhizhenjiakang@gmail.com>
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
Laying out a grid item will now use the same logic regardless of whether
it's replaced or not.
This reduces the amount of code, and should have no observable effect
(but hard to say since and I don't understand Taffy).
Testing: Unneeded (no behavior change)
This part of #37942
Signed-off-by: Oriol Brufau <obrufau@igalia.com>
There is a new default cargo clippy lint, `unsafe_op_in_unsafe_fn`,
which requires unsafe code to be wrapped in unsafe blocks, even inside
functions marked as unsafe. The lint is disabled as much of our code
doesn't fulfill this contract. The thing itself is pretty useful in
order to gradually remove unsafety, so this change starts adding
`unsafe` blocks so we can eventually enable this lint.
Testing: This doesn't change behavior so existings tests should suffice.
Fixes: This is part of #35955.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Console messages on OHOS/Android targets were duplicated.
Originally messages were on the debug level directed to stdout and then
redirected to the logger. https://github.com/servo/servo/pull/37912
changed this behavior recently.
Testing: Tested by looking at the debug log of certain things.
ohos-speedometer for example.
Signed-off-by: Narfinger <Narfinger@users.noreply.github.com>
The scripted_caller only has information if the context is coming
from a script. If an element fetch listener processes CSP
violations, then this information doesn't exist. Instead, we should
use the global URL and the line number. WPT tests don't appear
to expect a column number, as they are all zero. Not all elements
are updated, as I am not actually sure all of them need it.
The source position remains an Option, since there are also code
paths that don't correspond to element or script sources. Maybe
in the future we can always determine the source position, but
let's take small steps towards that.
Part of #4577
Signed-off-by: Tim van der Lippe <tvanderlippe@gmail.com>
Co-authored-by: Josh Matthews <josh@joshmatthews.net>
The iterator produced by Node::inclusive_ancestors roots the items but
rev_version just drop and unroot them right away. This patch makes it
possible to work on the Node references instead to avoid rooting.
*Describe the changes that this pull request makes here. This will be
the commit message.*
Testing: *Describe how this pull request is tested or why it doesn't
require tests*
Fixes: *Link to an issue this pull requests fixes or remove this line if
there is no issue*
Signed-off-by: webbeef <me@webbeef.org>
Before we only handled no-repeat for rect, this means we rendered
https://sagudev.github.io/briefcase/no-repeat.html incorrectly (like
firefox). Now if one of the axis is bounded (does not repeat) we clip it
and let other axis be unbounded (technically we clip it to end of
canvas).
This is also needed for vello backend.
Testing: Tests in WPT exists and another test is added.
---------
Signed-off-by: sagudev <16504129+sagudev@users.noreply.github.com>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
The cookie-rs library parses the cookie expiry date based on the format
from RFC 2616 (for HTTP/1.1), which is stricter than the format from RFC
6265 (for HTTP cookie).
This patch implements the cookie expiry date algorithm from RFC 6265.
When Cookie::parse fails in parsing the expiry date, we try to parse the
expiry again with this algorithm, to provide extra compatibility with
legacy systems.
Testing: Pass a WPT test that was expected to fail before, and add a
unit test.
Fixes: #36452
---------
Signed-off-by: Kingsley Yung <kingsley@kkoyung.dev>
Workaround to prevent
`webdriver/tests/classic/perform_actions/invalid.py` from TIMEOUT. I
don't see the spec telling us to limit this number, so I assume the
correct fix is to handle it on parsing.
Testing: `webdriver/tests/classic/perform_actions/invalid.py`.
Signed-off-by: PotatoCP <kenzieradityatirtarahardja18@gmail.com>
A recent update in the spec (https://github.com/whatwg/dom/pull/1079)
introduced new rules for name validation of attribute, element, and
doctype. This PR implements the new name validation rules in
`components/script/dom/bindings/domname.rs`. The old XML name validation
rules are not fully removed because there remains a few usage of it in
`ProcessingInstructions` and `xpath`.
Testing: Covered by WPT tests
Fixes: #37746
---------
Signed-off-by: minghuaw <michael.wu1107@gmail.com>
Signed-off-by: Minghua Wu <michael.wu1107@gmail.com>
Co-authored-by: Xiaocheng Hu <xiaochengh.work@gmail.com>
When laying out a grid item, the tentative block size that is used to
determine the intrinsic inline sizes was different than the one used in
the containing block for the contents. The latter is now used for both
cases.
Testing: A test is now passing.
Signed-off-by: Oriol Brufau <obrufau@igalia.com>
Improves the implementation of keys to a point where key ranges can be
implemented as well.
Due to me making branching mistakes I'll have to cherry-pick out the
first commit (it's from #37682)
---------
Signed-off-by: Ashwin Naren <arihant2math@gmail.com>
Previously, we SendKeys will be forwarded to constellation by the
embedder. Now we use webview.notify_input_event, which will send
WebDriverCommandMsg::ForwardInputEvent for the KeyboardEvent and
CompositionEvent.
Fixes: part of https://github.com/servo/servo/issues/37370
---------
Signed-off-by: PotatoCP <kenzieradityatirtarahardja18@gmail.com>
Add support for APNG animated image decoding. Rework `load_from_memory`
image decoding api, to handle all the image format that currently
supported.
Testing: This change should allow `apng` and `webp` format image start
moving, and should make some WPT test related to APNG pass.
Partially address: https://github.com/servo/servo/issues/37493
[wpt try
run](https://github.com/rayguo17/servo/actions/runs/15840339570)
cc @xiaochengh
Signed-off-by: rayguo17 <rayguo17@gmail.com>
With the work to unify the layout logic for replaced and non-replaced
boxes (#37942), I think the `IndependentNonReplacedContents` enum does
no longer make much sense.
Therefore, this removes `IndependentNonReplacedContents`, merging its
values into `IndependentFormattingContextContents`.
The methods defined on `IndependentFormattingContextContents` can now be
on `IndependentFormattingContext`, in particular this implies that the
layout results of a replaced box will now be cached.
Testing: Unneeded (no behavior change)
This part of #37942
Signed-off-by: Oriol Brufau <obrufau@igalia.com>
Laying out a float or atomic inline will now use the same logic
regardless of whether it's replaced or not.
This reduces the amount of code, and should have no observable effect.
Testing: Unneeded (no behavior change)
This part of #37942
Signed-off-by: Oriol Brufau <obrufau@igalia.com>
Does not yet handle failures of endpoints, which requires us to update
metadata. I don't see that metadata being used anywhere, so I am not
sure if there is WPT coverage for it.
Part of #37238
Signed-off-by: Tim van der Lippe <tvanderlippe@gmail.com>
When a style change does not chang the structure of the box tree, it is
possible to skip box tree rebuilding for an element. This change adds
support for reusing old box trees when no element has that type of
damage. In order to make this happen, there needs to be a type of
"empty" `LayoutDamage` that just indicates that a fragment tree layout
is necessary.
This is the first step toward incremental fragment tree layout.
Testing: This should not change observable behavior and thus is covered
by
existing WPT tests. Performance numbers to follow.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
Laying out a flex item will now use the same logic regardless of whether
it's replaced or not.
This reduces the amount of code, and should have no observable effect.
Testing: Unneeded (no behavior change)
This part of #37942
Signed-off-by: Oriol Brufau <obrufau@igalia.com>
Implement internal pseudo element, which would be resolved as a
"Implemented Pseudo Element" within style computation. This is an
concrete element that would has a primary style after the style
computation, but could match and style resolved like an pseudo element.
Therefore, it would have a different behavior compared to how does
`pseudo`s that `ServoLayoutNode` had. Where they would not have a
concrete element behind it. Note that, due to the nature of these pseudo
elements residing inside a UA widget, these pseudo elements would
therefore not be accessible in JavaScript by default.
This kind of element is required in order to implement the [form control
pseudo element](https://drafts.csswg.org/css-forms-1/#pseudo-elements)
like `::placeholder`, `::color-swatch`, `::field-text`, etc.
See [this docs](https://hackmd.io/@ChaKweTiau/BJ3zRdLQlg) for more
details of the implementation.
Then, the implemented pseudo element is utilized to implement style
matching for input `type=text`.
Servo's side of: https://github.com/servo/stylo/pull/212
Testing: No WPT regression.
---------
Signed-off-by: stevennovaryo <steven.novaryo@gmail.com>
`BoxFragment` had 2 fields that are only relevant for block-level boxes:
`clearance` and a boxed `block_margins_collapsed_with_children`.
This moves both pieces of data into a new `BlockLevelLayoutInfo` struct,
which is boxed.
As a result, the size of `BoxFragment` is reduced from 272 to 264 bytes.
Testing: Unneeded (no behavior change)
Signed-off-by: Oriol Brufau <obrufau@igalia.com>
During the object (de)serialization steps on structured cloning
https://html.spec.whatwg.org/multipage/#safe-passing-of-structured-data
it is possible to throw a JS exception.
`self.postMessage({ get whatever() { throw customError } }));`
Require to propagate a pending JS exception and not throw the default
"DataCloneError" DOM exception.
Testing: Improvements in the following tests
- html/infrastructure/safe-passing-of-structured-data/*
- html/webappapis/structured-clone/structured-clone.any.js*
- wasm/serialization/arraybuffer/transfer.window.js
- webmessaging/without-ports/026.html
- workers/semantics/structured-clone/dedicated.html
Signed-off-by: Andrei Volykhin <andrei.volykhin@gmail.com>
Let `WebDriverCommandMsg::ScriptCommand` goes through embedder first.
Give `embedder` the ability to release `webdriver` from waiting for a
response of `ExecuteScript`.
Tests: https://github.com/longvatrong111/servo/actions/runs/16071375821
No regression compared to CI run on main branch.
Fixes: https://github.com/servo/servo/issues/37370
cc: @xiaochengh
---------
Signed-off-by: batu_hoang <longvatrong111@gmail.com>
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
This PR updates the webidl files for `TextEncoder` and `TextDecoder`.
For `TextDecoder.webidl`, the type `AllowSharedBufferSource`, as defined
in the latest spec, is replaced with `BufferSource` as servo currently
does not support `ShareArrayBuffer`
Testing: This update does not introduce any change to the generated rust
binding, so the existing testing would suffice
---------
Signed-off-by: Minghua Wu <michael.wu1107@gmail.com>
Signed-off-by: Xiaocheng Hu <xiaochengh.work@gmail.com>
Co-authored-by: Xiaocheng Hu <xiaochengh.work@gmail.com>
The structured cloning with transfer list
https://html.spec.whatwg.org/multipage/#structuredserializewithtransfer
throws a "DataCloneError" DOM expection by default if
serialization/transferral
is not possible, but a platform object can throw a custom excepton on
its serialization/transfer steps.
One example is OffscreenCanvas, which can throw
an "InvalidStateError" exception if the context mode is not none on
transfer steps.
https://html.spec.whatwg.org/multipage/#the-offscreencanvas-interface:transfer-steps
Testing: Improvements in the following tests
-
html/canvas/offscreen/manual/the-offscreen-canvas/offscreencanvas.transferrable*
Fixes: #37919
Signed-off-by: Andrei Volykhin <andrei.volykhin@gmail.com>
Laying out an absolutely or fixedly positioned box will now use the same
logic regardless of whether it's replaced or not.
This reduces the amount of code, and should have no observable effect.
Testing: Unneeded (no behavior change)
This part of #37942
Signed-off-by: Oriol Brufau <obrufau@igalia.com>
