* Add repeatCount to RawKeyEventDataAndroid
Adds a new field to RawKeyEventDataAndroid and sets it to the value
passed from the engine. It is the value returned by
KeyEvent.getRepeatCount(). This allows us to differentiate between
events generated by a new keypress and repeated events when a key is
held down.
* Add test for RawKeyEventDataAndroid.repeatCount
In order to enable keyboard navigation to more controls, I've added focus nodes to switches, checkboxes, and radio buttons. In addition, this change enables mouse hover over these controls.
- Added tests for focus, hover, and shortcut activation for Checkbox, Radio, and Switch.
Empirical measurements indicate on the network file system we use
internally, using `FileStat.stat` on thousands of files is much
faster than using `FileStat.statSync`. (It can be slower for files
on a local SSD, however.)
Add a flag to `ProjectFileInvalidator.findInvalidated` to let it
use `FileStat.stat` instead of `FileStat.statSync` when scanning for
modified files. This can be enabled by overriding `HotRunnerConfig`.
I considered creating a separate, asynchronous version of
`findInvalidated`, but that led to more code duplication than I
liked, and it would be harder to avoid drift between the versions.
This PR adds a new property on RawKeyEventDataAndroid to hold the deviceId from which that event was generated. The deviceId is been sent from the engine on the changes from this PR: flutter/engine#12958
With this, we will be able to identify from where the event came from, adding the ability to support local multiplayer games, with multiple gamepads.
This adds a canRequestFocus and skipTraversal argument to FocusScope and FocusScopeNode, so that a scope can prevent being traversed.
This allows a fix for a problem in the gallery where the focus while traversing the list of items would sometimes appear to disappear, since it would be focusing things that were in the backdrop that were part of the tree, but were not visible.
Related Issues
Fixes#42955
This adds a Focus node to the DropdownButton widget, allowing it to receive keyboard focus, and to show a focus highlight. In addition, I added the ability to activate the dropdown using the "enter" key binding (which is bound to ActivateAction in the WidgetsApp).
Related Issues
Fixes#42646Fixes#43008Fixes#42511
Flutter tool's service client library tries to "upgrade" sub-components of service responses with higher level model objects. It does this for all members of the response, rather than only those documented in the version it targets, and it errors out when encountering a new type. This causes it to fail when a new field is introduced with a new typed. Clients are required to ignore this.
This re-implements keyboard text selection so that it will work on platforms other than Android (e.g. macOS, Linux, etc.).
Also, fixed a number of bugs in editing selection via a hardware keyboard (unable to select backwards, incorrect conversion to ASCII when cutting to clipboard, lack of support for CTRL-SHIFT-ARROW word selection, etc.).
Did not address the keyboard locale issues that remain, or add platform specific switches for the bindings. All that will need some more design work before implementing them.
Related Issues
Fixes#31951
In #42533, I disabled the up/down arrows for focus navigation in text fields, but we thought of a better way to do it, so this is that better way.
This change reverts the other change, and instead it tests the context of the node in the action to see if it's an EditableText node. If so, then it doesn't do the navigation action.
The _computeChildren[Width/Height]With[Min/Max]Intrinsics methods in
RenderParagraph overwrite the text painter's placeholder dimensions
in order to compute potential widths and heights for the paragraph.
The placeholder dimensions should be restored to the values computed by
_layoutChildren before painting the text.
This optimizes certain paths in the FocusManager, FocusNode, and FocusScopeNode classes in order to fix a regression in stock_animation_open_first_frame_average when I added more focus nodes to the tree to do focus traversal.
Mainly I removed some remaining sync* iterators, and also started caching the computation of descendants and ancestors, since those are iterated over fairly often.
This improves stock_animation_open_first_frame_average by about 2.8% overall (so about half of a 4.9% regression, both averaged over 10 runs).
Addresses #42564
This disables the arrow key focus navigation for text fields. This was non-standard behavior for text fields, although it remains useful for other kinds of controls.
Fixes#42259
This fixes a problem with reparenting of focus nodes where it would remove the node from the scope's focused children when reparented, even if it was being reparented to another place under the same scope. This only occurred if the scope in question didn't have focus.
This caused nodes to not get autofocused when they were a child of another node within the scope, and were reparented, and then the scope was given focus (as when a route is pushed).
This keeps the node in the scope's child list where it was if the node is reparented under a parent within the same scope.
- Added a test that tries to autofocus a TextField when the route is pushed and there is another
FocusNode above the text field. This was how this was first noticed: the autofocus got ignored in
this configuration.
- Added a test to focus_manager_test that tests for the specific case of reparenting a node when
it's in the focused children of the scope.
* Updated the BottomAppBar to use elevation overlays when a dark theme is used.
* Moved the code for calculating the overlay color to a new utility
ElevationOverlay class and made it private by not including in the material package.
* Allow showModalBottomSheet to present bottom sheet that is not dismissible by tapping on the scrim
* Add guards, improve styling and tests for BottomSheet
This changes the DropDownButton so that instead of dismissing itself when any metrics change occurs, it only dismisses itself when the orientation changes.
This gets around the fact that we can't currently have a dropdown and a text field on the same page because the keyboard disappearing when the dropdown gets focus causes a metrics change, and the dropdown immediately disappears when activated.
It still will cause the keyboard to jump up and down between controls, but that's a larger issue. At least now we can use the two together again.
...because otherwise, processes that think they're manipulating your
filesystem will be doing crazy things the test is ignoring, leading to
(at best) failures and (at worst) flakes or disk corruption.
This attempts to reland #40186 and #41220, that were reverted in #41945.
The main modifications from the original PRs are that I predefine the shortcuts and actions maps instead of defining them inline in the build function, and I use a new mapEquals to do a deep comparison so that we don't rebuild modified things if the contents of the map haven't changed.
I also eliminated an operator== and hashCode that were defined on the Actions widget, since widgets shouldn't have those. (it's too bad though: I get an 85% speedup if we leave this in! Too bad it prevents rebuilding of the children...)
Fixes#40101
Currently the surface of a platform view is only created only one when the state of PlatformViewLink is created. When the PlatformViewLink widget changes, the PlatformViewController in the corresponding state is also updated. Just the surface is not updated even though it depends on the controller.
This PR changes this behavior to recreate the surface whenever the controller is updated.
* Add smoke test for the new Android embedding
* Update AndroidManifest.xml in app template
* Update test README.md
* Remove widget_test.dart
* Update pubspec.yaml
* Force GeneratedPluginRegistrant.java
The Material Dark Theme spec has been updated (or soon will be) to clarify that the elevation overlay that is applied to surface colors to indicate elevation should be based off of the colorScheme.onSurface color and not white. This commit implements this change.
Restructure ProjectFileInvalidator.findInvalidated a bit
I plan to modify `ProjectFileInvalidator.findInvalidated` to allow
it to use `FileStat.stat` instead of `FileStat.statSync`.
Restructure `findInvalidated` a bit so that `FileStat.statSync` is
called in only one place.
Note that this change now always counts the `.packages` file as one
of the files scanned, even if it does not exist. I think that this
is okay since it more accurately reflects the number of times that we
hit the filesystem with `stat()`, and it is consistent with how other
files are counted.
In preparation for some refactoring that I will be doing to
`ProjectFileInvalidator.findInvalidated`, make its code a bit
clearer:
* Indicate which arguments may be null.
* Don't bother calling `FileStat.statSync` on the `.packages` file
for the initial load. This makes the checks for the `.packages`
file consistent with those for other files.
* Use `DateTime.isAfter()` instead of comparing microseconds
ourselves.
While I was touching this file, I also removed some unnecessary
comparisons to `false`. (`_ManifestAssetBundle.wasBuiltOnce()` is the
only implementation I can find of `AssetBundle.wasBuiltOnce()`, and
it never returns `null`.)
* Generate projects using the new Android embedding
* Add comment about usesNewEmbedding:true
* Feedback
* Rework way to detect new embedding in new apps
This adds the default shortcuts and actions for keyboard-based focus traversal of apps.
This list of shortcuts includes shortcuts for TAB, SHIFT TAB, RIGHT_ARROW, LEFT_ARROW, UP_ARROW, DOWN_ARROW, and the four DPAD keys for game controllers (because the DPAD produces arrow key events).
It doesn't yet include functionality for triggering a control (e.g. SPACE, ENTER, or controller buttons), because that involves restructuring some of the Flutter controls to trigger animations differently, and so will be done in another PR (#41220)
* dont update last compiled time when compilation is rejected
* lets try flushing, thats a neat trick
* windows man
* Update hot_reload_test.dart
* Update hot_reload_test.dart
* Update devfs.dart
* Update hot_reload_test.dart
* Update hot_reload_test.dart
* add test that verifies when compile is good that time is updated
* Update devfs_test.dart
Adds an ActivateAction to controls that use InkWell. Make InkWell host the Focus for those controls, and add the top level binding to the ENTER key. This will make it possible to trigger a button using the enter key, and to get an ink response when the button is triggered.
This is a breaking change because it moves the Focus widget into the InkWell. If you have a component that uses the InkWell directly and you used to wrap that InkWell in a Focus widget to give it its notion of focus, it will now not look for that Focus ancestor for its focus state anymore. In order to control focus on the InkWell, you need to give it a FocusNode directly, via the new focusNode parameter. This should not affect users of widgets like OutlineButton or FloatingActionButton and the like, since those have been modified in this PR.
* Updates to slider_test.dart
Fixed error that required onChanged.
* Updated slider.dart and slider_test.dart debugFillProperties
Updated slider.dart debugFillProperties to include all properties and slider_test debugFillProperties to test all properties
* Updated Slider onChanged property
* Updated debugFillProperties to ignore testing methods that are null in test
Per feedback from Hans/Anthony on Flutter Material
* Updated spacing in debugFillProperties
Per feedback from Hans
* Updated range_slider.dart and range_slider_test.dart to have DiagnosticProperties tests for all properties
* Updated range_slider.dart and range_slider_test.dart per feedback
Preventing null labels from creating an error.
Added made RangeLabels and RangeValues constants in instantiation of Range Slider.
Fixed spacing.
* Updates to start and end labels in diagnostic property tests
Updated per feedback from @clocksmith
* Stop using build_runner for dart2js
* fixes to yield when computing hashes and to imports
* add missing await
* Update filecache_test.dart
* Fix paths in filecache test
* use file uri for import
* add test cases and configurable override
* remove test dep
* fix filepaths for windows
* test no longer failing
* fix paths for test cases
* fix typo
* address comments
* make a constant
* make filehash async and use constant
* fix silly logic error
Spawn no more than 1 iframe in web tests. Using multiple iframes in DDK mode times out tests. Also set concurrency to 1 to guarantee that the test runner does not even attempt to parallelize tests.
Currently Linux builds override the default BUILD mode by putting it in
the generated config. That makes it sticky for manual runs of make,
which is inconsistent with how other platforms work.
Instead, pass the build mode as a command-line override, the same way
someone would if building directly with make. This makes the flow of
controlling the mode less confusing.
Fixes#41528
* Fix visbility of web server device when Chrome is not available
* Add tests
* Update workflow test
* Fix tests to not rely on Chrome being on the underlying machine
This re-lands #41417 with a slight change that will hopefully not tickle the analyzer as it did before. The last time I tried to land this, the analyzer succeeded for the analyze step in Cirrus, and locally, but failed in an integration test.
ReorderableListView was constructing a GlobalObjectKey using
the child key as the value. This only had the intended behavior
if the child key was identical across build method invocations.
The new strategy is to scope the child key's value to the
State object's identity, allowing child keys to have value
compare semantics while disambiguating among different list view
instances.
This changes the way ActionDispatchers are found by the Actions widget, so that by default it will look for dispatchers of the parent Actions widgets instead of just creating a default ActionDispatcher. This allows overriding of the ActionDispatcher at the top level: before, the custom action dispatcher would only be invoked if explicitly set on all the Actions widgets.
This is not a breaking change because there was a default value to the dispatcher parameter before that performed this function, and not specifying the dispatcher anywhere will still result in a default dispatcher being created.
The proposed change will change focus handling when pushing and popping routes so that the FocusScopeNode for the route receives focus when pushed, and that the FocusScopeNode in the navigator receives focus when the route is popped.
This means that the last setFirstFocus call on the scope is used to determine which control actually receives focus. When the focus scope receives focus, it traverses its children, trying to find a non-scope node that is the "first focus" of itself or a child node.
This is a breaking change, because the focus behavior has changed. If you push a route after this change, and had a 'first focus' set on a widget via FocusScopeNode.setFirstFocus, it won't currently receive focus immediately, but after this change it will. Similarly, if you pop a route after this change, the focus will go back to where it was before the route was pushed, which is correct, but different from what happens now.
Adds very preliminary support for Windows and Linux plugins:
- Adds those platforms to the new plugin schema, initially supporting just a plugin class.
- Adds C++ plugin registrant generation for any Windows or Linux plugins found.
This doesn't have yet have any build tooling for either platform, so anyone using the generated registrant still needs to do manual build configuration. This reduces the manual work, however, and creates a starting point for future tooling work.
As with all Windows and Linux work at this time, this is not final, and subject to change without warning in the future (e.g., Windows could potentially switch to a C# interface, or
'linux' may change to 'gtk' or 'linux_gtk' in pubspec.yaml).
* MouseRegion documentation claimed that onEnter and onExit
would track entry and exit regardless of whether the pointer was
down or up
* It did such, but when grabbing the value of `event.down` from
the passed event, the value was always `false`
* PointerEnterEvent and PointerExitEvent were overriding the value
passed from PointerEvent in constructors, even if the value was true
e.g. in invocations of .fromMouseEvent((PointerMoveEvent...))
* This change now passes the value along to PointerEnter/ExitEvents
while providing it a default of false, and updates documentation
Fixes#40637
This fixes the mouse hover code to not schedule frames with every mouse move.
Before this, it would schedule a post frame callback, and then schedule a frame immediately, even if there was nothing that needed to be updated. Now it will schedule checks for mouse position updates synchronously, unless there's a new annotation, and skip scheduling a new frame in all cases. It has to be async in the case of a new annotation (i.e. a new MouseRegion is added), since when the annotation is added, it hasn't yet painted, and it can't hit test against the new layer until after the paint, so in that case it schedules a post frame callback, but since it's already building a frame when it does that, it doesn't need to schedule a frame.
The code also used to do mouse position checks for all mice if only one mouse changed position. I fixed this part too, so that it will only check position for the mouse that changed.
* Carrying last size and transform information to TextInputConnection. Doing this we are making sure that the size/transform information from the previous connection will be removed, when connection changes.
* remove unused lastsize and lasttransform values
* Adding unit tests. Adressing comments.
There were four or five different implementations in various tests for sendFakeKeyEvent, which roughly all did the same thing. I was going to add yet another one, and decided that it needed to be generalized and centralized. This replaces those instances with something that just takes a LogicalKeyboardKey so that it's self-documenting, and can be used with multiple platforms.
This adds two functions to widget tester: sendKeyDownEvent and sendKeyUpEvent which simulate key up/down from a physical keyboard. It also adds global functions simulateKeyDownEvent and simulateKeyUpEvent that can be called without a widget tester. All are async functions protected by the async guard.
