Previously the webdriver do not wait for focus to complete, which can
cause some instability.
No matter interact as human or webdriver, the focus chain always goes
as: Embedder forwards -> Constellation (do some updates) -> Embedder (do
some updates).
---------
Signed-off-by: Euclid Ye <euclid.ye@huawei.com>
According to discussion
https://github.com/servo/servo/pull/34063#discussion_r2187147615, this
PR
1. adds `PIXEL_DELTA_FACTOR` to increase scroll speed for
`MouseScrollDelta::PixelDelta`, which is used by macOS and
high-precision mouse.
2. adds `LINE_WIDTH` to increase x-axis scroll speed for
`MouseScrollDelta::LineDelta`, which is used by Linux and Windows.
3. Increase mouse scroll speed in general to match other browsers
4. Reduce keyboard scroll offset to scroll exactly "one line", to match
other browsers.
Testing: Example in #35037 can now scroll in x-axis as fast as y-axis,
similar to other browsers.
Fixes: part of #38072.
---------
Signed-off-by: Euclid Ye <yezhizhenjiakang@gmail.com>
Signed-off-by: Euclid Ye <euclid.ye@huawei.com>
- Add `handler any user prompt` step for all commands.
- Enable webdriver tests which were blocked by `handle any user prompt`
step.
---------
Signed-off-by: batu_hoang <hoang.binh.trong@huawei.com>
We also need to bump egui-file-dialog and accesskit_winit. Changes are
done based on migration paths provided by egui.
Testing: No tests for servoshell
Fixes: #38117
Signed-off-by: sagudev <16504129+sagudev@users.noreply.github.com>
Sending all mouse button events (Left, Right, Middle, Back, Forward) to
the constellation for processing. Before, only left and right mouse
buttons were handled.
Testing: There is no new test for this. Manually testing was done. We
currently
do not have a good way to test user interaction in servoshell.
Fixes: #37996
Signed-off-by: abdelrahman1234567 <abdelrahman.hossameldin.awadalla@huawei.com>
1. Create `get_origin_relative_coordinates` according to
[spec](https://w3c.github.io/webdriver/#dfn-get-coordinates-relative-to-an-origin)
to be reused
2. Add previously missing offset for PointerOrigin::Element
3. Refactor code for perform pointermove/wheel to be closer to spec.
4. Handle some issues with spec:
https://github.com/w3c/webdriver/issues/1758
Testing: Several new passing cases as we are more precise with
coordinates now.
Fixes: Part of #38042.
---------
Signed-off-by: Euclid Ye <euclid.ye@huawei.com>
With helper functions introduced in #38020, this is now becoming
reality.
Testing: No behaviour change.
Fixes: #37937
Signed-off-by: Euclid Ye <euclid.ye@huawei.com>
After sending `GoBack` or `GoForward` command, webdriver wait for the
navigation complete.
It can be achieved by waiting for
`WebViewDelegate::notify_history_changed`
Testing:
`tests/wpt/meta/webdriver/tests/classic/back/back.py`
`tests/wpt/meta/webdriver/tests/classic/forward/forward.py`
---------
Signed-off-by: batu_hoang <longvatrong111@gmail.com>
Signed-off-by: Josh Matthews <josh@joshmatthews.net>
Signed-off-by: batu_hoang <hoang.binh.trong@huawei.com>
Co-authored-by: Josh Matthews <josh@joshmatthews.net>
I'm trying to use Servo as a daily driver. I like to pin the most often
used apps to the my taskbar (Dash to Panel in Gnome). I also launch apps
from the menu (Arc Menu). Adding a .desktop file make it show up in the
menu. I could not get pinning to work until I set the name with `winit`.
This is because the name is used to match it with the icon. I used the
Gnome Looking Glass (lg) command to determine that the name was not set.
Testing: I followed the instructions that I put in the .desktop file. I
copied the .desktop file to the right folder, set up a link to the svg
icon, and adjusted the paths to point to the compiled executable.
---------
Signed-off-by: Michael Mc Donnell <michael@mcdonnell.dk>
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>
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>
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>
Previously, we only consider OS decoration height. But when testing
#37960, I find that the decoration width is also non-zero.
Testing: Need to wait W3C spec change
https://github.com/web-platform-tests/wpt/pull/53421 related to
webdriver rectangle. When combined with #37960, this can fix at least
`window_resizeTo.html`.
Signed-off-by: Euclid Ye <yezhizhenjiakang@gmail.com>
This function is used to compute `screenLeft`, `screenTop`, `screenY`,
`screenX`. Previously, it wrongly computes viewport position instead of
window position, which didn't match
f7c86c4393/components/shared/embedder/lib.rs (L717-L720)
Testing: More webdriver test would pass once #37893 is merged.
Fixes: the position part of #37824.
---------
Signed-off-by: Euclid Ye <yezhizhenjiakang@gmail.com>
Previously, we KeyboardAction will be forwarded to constellation by the
embedder. Now we use `webview.notify_input_event`, which will send
`WebDriverCommandMsg::ForwardInputEvent` for KeyboardAction
Fixes: part of https://github.com/servo/servo/issues/37370
---------
Signed-off-by: PotatoCP <kenzieradityatirtarahardja18@gmail.com>
Testing: Just add comment. It turns out difficult to achieve, see
https://github.com/rust-windowing/winit/issues/2494.
Fixes: Nothing. Provide some guidance to #37878
---------
Signed-off-by: Euclid Ye <yezhizhenjiakang@gmail.com>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
`toolbar_height` is already part of `inner_size`, caused wrongly
calculated `outer_size`. Even worse, it tried to `request_inner_size`
with the already wrong `outer_size`.
This PR make sure resize is accurate by first calculate the title/border
height, and then compute the `inner_size` for `request_inner_size`. This
is necessary because no direct `request_outer_size` is available.
Testing: As manually tested, set window size WebDriver command no longer
overshoot. This is also shared by
[window.resizeTo](https://drafts.csswg.org/cssom-view/#dom-window-resizeto)
JS method. WPT test would be necessary. (But that one is intermittent
TIMEOUT. So created new one in
https://github.com/servo/servo/pull/37856)
WebDriver test will be postponed after
https://github.com/web-platform-tests/wpt/pull/53421 is merged and
synced to Servo.
Fixes: Task 3 of https://github.com/servo/servo/issues/37804
---------
Signed-off-by: Euclid Ye <yezhizhenjiakang@gmail.com>
Move `webdriver_port` option from servo config options to servoshell
preference.
Testing: run `./mach run` with/without `--webdriver` option and see if
the webdriver server runs on the port (default: 7000)
Fixes: https://github.com/servo/servo/issues/37843
Signed-off-by: Jason Tsai <git@pews.dev>
In Firefox, clicking on the address bar selects all the text. This makes
Servo to do the same. I reworked the code so that the shortcut only
changes the focus. That means that either clicking or the shortcut
changes the focus which in turn selects the text.
Testing: I tested it manually by clicking in the address bar and using
the shortcuts and noticing that the text was selected.
Signed-off-by: Michael Mc Donnell <michael@mcdonnell.dk>
Previously, our Servo-specific spatial tree scroll offsets were opposite
to
that of WebRender and also the web platform. This is due to the fact,
likely, that `winit` wheel directionality is also flipped. This change
has both the Servo spatial tree and the API take offsets that are
consistent with the web.
Any possible changes to the meaning of wheel directionality will be
handled in a followup change.
This is a breaking change to the Servo API.
Testing: This change updates unit tests.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
1. Rename `GetWindowSize` to `GetWindowRect`
2. Return the WindowRect in device pixels correctly. Previously, it
returns `(0, 0, ScreenWidth, ScreenHeight)` which is a static value.
3. Add `fn window_rect` to `WindowPortsMethods`. Implement it for both
Headless Window and Headed Window.
Testing: Tested manually with powershell script. Result is now dynamic
and reflects the truth.
Fixes: Task 1 & 2 of https://github.com/servo/servo/issues/37804
---------
Signed-off-by: Euclid Ye <yezhizhenjiakang@gmail.com>
I usually use the alt-d in shortcut in Firefox to select the text in the
address bar. Firefox supports both alt-d and ctrl-l. This change adds
the alt-d shortcut so you can use either.
Testing: I tested it by running Servo and pressing alt-d. I don't know
if there is a way to create a test for it?
---------
Signed-off-by: Michael Mc Donnell <michael@mcdonnell.dk>
1. Remove the unnecessary new thread which use GetWindowRect command and
blocks for 500ms. Previously this is necessary because constellation
forward "resize" to embedder, and WebDriver wait for a constant
sufficient time to `GetWindowRect` in the new thread. This caused long
delay because there are many subtests and SetWindowRect is called
between each.
2. Remove `resize_timeout`
3. Return current dimension instead of 0 from embedder when it fails to
resize.
4. Do resizing as long as one of width/height is `Some`, according to
spec.
Testing: All Conformance test with new passing cases.
Fixes: https://github.com/servo/servo/pull/37663#issuecomment-2999120615
---------
Signed-off-by: Euclid Ye <yezhizhenjiakang@gmail.com>
- Move webdriver `GetViewportSize`, `LoadURL` and `Refresh` to
servoshell.
- Add `GoBack` and `GoFoward` commands.
Testing: Need to finish moving webdriver to servoshell then evaluate
again
Fixes: https://github.com/servo/servo/issues/37370
Signed-off-by: batu_hoang <longvatrong111@gmail.com>
These changes ensure that our browser shell can integrate with
screenreaders. We do not provide any accessibility information about
webview content yet, which requires further API design work in both
Servo, accesskit, and egui.
Testing: No a11y-specific testing at this point; just verifying that
existing tests continue to pass.
Fixes: part of #4344
---------
Signed-off-by: Josh Matthews <josh@joshmatthews.net>
Previously, `handle_webdriver_messages` only called in
`winit::ApplictionHandler`. This means the webdriver command only looped
in headed (desktop) event loop.
Testing: this will fix webdriver test on headless `test-wpt`, which
widely used in the CI.
Signed-off-by: PotatoCP <kenzieradityatirtarahardja18@gmail.com>
The compositor always does the same thing with these events regardless
of the phase, so I think it is completely unecessary.
Testing: This shouldn't change behavior at all, so is covered by
existing tests.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Moving `webdriver` to `servoshell`:
- Let `servoshell` manage lifecycle of `webdriver`
- One by one, move the handling of webdriver commands from
`constellation` to `embedder`
Partially fix: https://github.com/servo/servo/issues/37370
Partially fix webdriver test timeout with `no_top_browsing_context`
cc: @xiaochengh
---------
Signed-off-by: batu_hoang <longvatrong111@gmail.com>
Currently, the hover state will stay when the mouse moves out of the
webview, this PR fixes it
Testing: Hover on the `About` on servo.org, and then move the mouse up
to the browser UI, see the hover state resets
Signed-off-by: Tony <legendmastertony@gmail.com>
Implement action synchronization for wheel event. Previously only done
for pointer here https://github.com/servo/servo/pull/36932.
Testing:
`tests/wpt/meta/webdriver/tests/classic/perform_actions/wheel.py`
---------
Signed-off-by: PotatoCP <kenzieradityatirtarahardja18@gmail.com>
Split out `save_output_image_if_necessary` into its own file so the code
can be shared by servoshell on the desktop and ohos.
Additionally, hook it up to the loop in OHOS and have OHOS allow
exiting.
Testing: Manual testing on ohos and desktop.
---------
Signed-off-by: Narfinger <Narfinger@users.noreply.github.com>
The logic was opposite to intended in #37284. We ended up staying in
same tab for `window.open` but focus new tab for WebDriver.
Testing: This modifies whether or not tabs are activated in servoshell
and WebDriver. WebDriver is currently not tested and neither is
servoshell.
Signed-off-by: Euclid Ye <yezhizhenjiakang@gmail.com>
For Desktop port of `request_open_auxiliary_webview`, stay on the
original WebView if the request originates WebDriver.
This is to make sure `webdriver_server::handle_new_window` does not
focus the new window, according to spec. See
c7eba2dbba/tests/wpt/tests/webdriver/tests/classic/new_window/new_window.py (L31-L37)
**To clarify**: this won't change the behaviour when user interacts, but
only affects WebDriver [New
Window](https://w3c.github.io/webdriver/#new-window).
Testing: `./mach test-wpt -r --log-raw "D:/servo log/all.txt"
./tests/wpt/tests/webdriver/tests/classic --product servodriver` based
on 96b0973037
---------
Signed-off-by: Euclid Ye <yezhizhenjiakang@gmail.com>
Dropping the window while the rendering contexts are still around causes
a segmentation fault during shutdown. This is a very fragile change. I
added comments to hopes of making regressions less likely.
Testing: I don't think we have a way to write tests for this change
since it requires a wayland system ):
Fixes: https://github.com/servo/servo/issues/36711
Signed-off-by: Simon Wülker <simon.wuelker@arcor.de>
Not all winit platforms support getting the system theme properly. In
some places we fall back to the light theme, but it seems that the
default for `egui` is dark. This change makes it so that we fall back to
the light theme more consistently, meaning that servoshell on Wayland
will properly use the light theme.
Testing: This is difficult to test because we have no servoshell display
tests.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Implement missing synchronization in `dispatch_actions` of `WebDriver`.
https://w3c.github.io/webdriver/#dispatching-actions
> The user agent event loop has spun enough times to process the DOM
events generated by the last invocation of the >[dispatch tick
actions](https://w3c.github.io/webdriver/#dfn-dispatch-tick-actions)
steps.
- Add a way for `ScriptThread` to notify `WebDriver` about the
completion of input commands.
- Add a `webdriver_id` field for `InputEvent`. `ScriptThread` uses it to
distinguish WebDriver events and sends notification.
Tests:
`./mach test-wpt --product servodriver -r
tests\wpt\tests\webdriver\tests\classic\element_click\events.py` pass if
`hit_testing` pass. Check
[issue](https://github.com/servo/servo/issues/36676#issuecomment-2882917136)
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 change adds a shadow-tree widget for `<input type=color>` elements.
It also involves some changes to the way layout interacts with the DOM,
because currently all `input` and `textarea` elements are rendered as
plain text and their descendants are ignored. This obviously doesn't
work for `<input type={color, date, range, etc}>`.
<details><summary>HTML used for the screenshot above</summary>
```html
<input type=color>
```
</details>
Testing: I doubt that this affects WPT tests, because the appearance and
behaviour of the widget is almost entirely unspecified.
---------
Signed-off-by: Simon Wülker <simon.wuelker@arcor.de>
Instead of returning true / false it's better to return a Result (even
if we continue ignoring possible error).
Testing: Doesn't change any behavior
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
