* Deprecates `BinaryMessages` in favor of a default instance of `BinaryMessenger`, called `defaultBinaryMessenger`
* Platform channels use the `defaultBinaryMessenger` for their binaryMessenger default argument.
Replacing the algorithm for finding the first focusable item in the focus tree. Somehow it was a kind of gibberish before, and really didn't work or make sense.
* Adding sample and animation for Flow widget
* Updated dart templates for layout in API docs
* Updated the sample code for a clearer example.
* make clean
Splits Xcode validation out of the iOS validator and into a stand-alone
validator, and groups the CocoaPods validator with that top-level
validator instead of the iOS validator. iOS now validates only the
iOS-specific tools (e.g., ideviceinstaller).
Reorganizes many of the associated clases so that those that are used by
both macOS and iOS live in macos/ rather than ios/. Moves some
validators to their own files as part of the restructuring.
This is the macOS portion of #31368
Updates documentation and non-public API to use American spellings for
consistency with the rest of the codebase.
No changes to behaviour... other than how it's spelt.
Adds a new macosPrefix, which serves the same purpose as iosPrefix but
for macOS plugins.
It is not yet used by the tooling, but this allows for plugins to start
to be written using it in preparation for tooling support for plugins.
Part of #32718
* Implement ExpandIcon custom color, expandedColor, and disabledColor
* Update to use pumpAndSettle instead of hard-coded duration
* Update colors to unfocused state, added dark mode test to active state
* Fix Colors.white30 doc opacity value
* Add links to Material Design specifications to color, expandedColor and disabledColor
* Update API docs to reference dark theme material page
* Clean up some flutter_tools tests
* Remove arbitrary retry that happens even for fundamental errors, and generally clean up _DevFSHttpWriter.
* Update dependencies (requires fixes; see next commit)
* Fixes for new dependencies.
This is another step towards supporting image loading
progress notification at the widgets layer.
This adds an `ImageChunkEvent` class along with associated
`ImageChunkListener` callback signature and an `onChunk`
property to `ImageStreamListener`. The events serve to
notify registered listeners when byte chunks are received
while loading an image.
https://github.com/flutter/flutter/issues/32374
This just fixes up some comments for DefaultFocusTraversal, and fixes a minor bug when setting the onKey on a FocusNode on creation before attaching to it.
The current API was broken in that you registered multiple
callbacks at once, but when you removed listeners, only the
primary listener was used to determine what was removed.
This led to unintuitive cases where the caller could get
unexpected behavior.
This updates the API to add and remove listeners using
a newly introduced [ImageStreamListener] object, a value
object that has references to the individual callbacks
that may fire.
flutter/flutter#24722flutter/flutter#32374flutter/flutter#32935
Resets the enabled color for outline text field default borders to "On Surface 38%" instead of "On Surface 12%" to match spec. Also fixes the hover overlay to be "On Surface 12%" to match spec.
This will allow us to plumb the chunks in a chunked response
up to the higher levels of the framework to notify interested
parties of network loading progress.
https://github.com/flutter/flutter/issues/32374
Currently, the fact that NetworkImage uses a static HttpClient
makes it impossible to properly test, as a mock in one test will
be reused in another test. This change fixes that.
https://github.com/flutter/flutter/issues/32374
* Moved Radio documentation line to be above sample
* Added LabeledRadio sample
* Add LabeledCheckbox sample
* Add LabeledSwitch sample
* Added LinkedLabelRadio sample to RadioListTile
* Added LinkedLabelCheckbox sample to CheckboxListTile
* Added LinkedLabelSwitch sample to SwitchListTile
* Added reference to Semantics docs
* Improve simple SwitchListTile, RadioListTile and CheckboxListTile samples
* Added assets to all SwitchListTile, RadioListTile and CheckboxListTile samples
This adds support for an animated focusColor and hoverColor to InputDecorator. This color will blend with the background over a fade in period whenever the InputDecorator is focused or hovered, respectively.
It also adds a Listener to the TextField to listen for hover events.
Until this change, Focus.of would return a FocusScopeNode if it found a FocusScope widget. This isn't really all that useful, and can easily lead to bad situations where many widgets think that the scope they are in (or the root scope!) is their indication of being focused.
This changes Focus.of to throw an exception if it doesn't find a Focus widget before reaching the nearest FocusScope widget, or the root of the widget hierarchy.
It also adds a nullOk optional bool to Focus.of so that if a caller expects to not find a Focus widget, it can deal with that as it sees fit. I modified InkWell to use this new behavior.
This fixes an unreported issue that widgets using an InkWell will be drawn as focused the first time they are visited.
This adds a reverseDuration parameter to AnimationController so that the animation has a different duration when going in reverse as it does going forward.
Date picker tests have been flaky on Windows for a long time, and repeatedly require restarting of bots. In the interest of productivity, disabling them on Windows.
Related to #19696
Instead of requiring a name_output.sh, expect a file called
.app_filename in the macos/Flutter directory containing just the name of
the application. The expectation is that the Xcode build will create
that file with a script.
This is not intended as a long-term solution, but it's a substantial
improvement over name_output.sh:
- name_output.sh required constructing the full build output path; this
made sense when it was coupled with build.sh, but now that the
decision for where build output goes lives in flutter_tool, that logic
should as well.
- Changing the name of the application required also updating
name_output.sh, which is error-prone. With .app_filename, it can be
created using $PRODUCT_NAME, which means that the usual way of setting
the application name will automatically update this flow as well.
Part of #30706
Allows Windows builds to use the same structure and script as Linux
builds now use, calling into tool_backend to manage copying resources to
the project directory and building the bundle.
Also switches from expecting name_update.bat to expecting flutter\exe_filename
to be written during the build, as with the recent changes to the macOS build, to
reduce the amount of boilerplate needed in a windows\ project directory.
The PlatformViewsService listens for `viewFocused` calls on the
platform_views system channel and fires a callback that focuses the
focus node for the relevant AndroidView widget.
Correct docs and error message for UiKitViewController to refer to iOS
view rather than Android views. Minor cleanup to the docs for
PlatformViewsService, which previously mentioned Android only.
This implements focus and hover handling for Material buttons. It inserts Focus widgets into the tree in order to allow buttons to be focusable via keyboard traversal (a.k.a. TAB traversal), and Listener widgets into the InkWell to allow the detection of hover states for widgets.
Addresses #11344, #1608, and #13264.
This fixes propagation of needsCompositing from child widgets.
When needsCompositing is turned on by a child widget, it necessarily sets the needsCompositing bit of its parent widget, but RenderPointerListener was ignoring that piece of information and only turning on compositing if it thought it needed it for itself.
This corrects that, and adds a test for the condition, and updates a test that was affected by the change.
Fixes#32525 (again)
Eliminates the need for a build.bat in the Windows build workflow, adding
preliminary support for building using msbuild. The handling of
vcvars64.bat may be refined in the future, but this serves as a starting point.
This fixes#32525, because it now marks the compositing bits as needing to be recalculated if the mouse tracker changes its idea of whether or not a mouse is attached.
This bug occurred because the test framework was leaking state from one test to the next (the state about whether a mouse pointer was active), and so even though there was a "passing" test when run in order with the other tests in the file, when the test was run individually (or first), it would have failed and caught the bug.
This adds an assert to make sure that after each test there are no simulated mouse pointers connected, and now calls removePointer in all of the tests where this was a problem.
Some parts of the appbundle build process were based on the logic for building
APK packages. However, these steps (copying to a directory shared by all
build variants, and calculating a SHA) are not necessary for an appbundle.
- Removes SYMROOT from the Generated.xcconfig. Having it causes current
versions of Xcode to switch the project's build output to "Legacy",
which causes anything not overridden to use a project-relative build
directory instead of a shared directory in DerivedData, breaking
anything with subprojects that it depends on.
This means that `flutter run` and builds from Xcode will use
completely different build directories, but that each should be
internally consistent.
- Moves the FlutterMacOS.framework to $SRCROOT/Flutter. This is
consistent with the approach we're moving to for all desktop
platforms, and avoids issues finding it now that SYMROOT doesn't match
for the two different build modes.
Fixes#32494
This implements a DefaultFocusTraversal widget to describe the focus traversal policy for its children, defined by a FocusTraversalPolicy object from which custom policies may be created. Pre-defined policies include widget-order traversal, "reading order" traversal and directional traversal.
This fixes a benchmark regression introduced in #32350. The performance is improved by just reverting the Layer.find<S> routines to use their old definitions, instead of defining them in terms of the findAll<S> lazy iterators.
Fixes#32387
This changes Listener to trigger enter/move/exit in all Listeners below the pointer, not just the leaf region (the first region hit). This is because we need to allow listeners to be nested so that, say, a widget that handles changing color on hover, but also is wrapped in a Tooltip (that handles hover) can trigger both actions, not just one.
To that end, I added a findAll to Layer, similar to the existing find method that was previously used. It returns an iterator over annotated layers which match the given data type.
Since the findAll is implemented as returning an Iterable (and is sync*), I re-implemented the find routines as just returning the first result from findAll, since that should be just as efficient, and would then prevent duplication in the implementation.
I recently added some code to keep hover events from being propagated when a mouse wasn't attached. While that works, there are times when it can fire callbacks during the building of other components, since they can now be called from detach/attach. This is not ideal, since it will assert then. This changes the code so that it won't update the annotations during attach/detach, but also won't push the annotation layer unless a mouse is connected, achieving the same result as before, but with better semantics.
The basic problem is that in the detach for RenderPointerListener, it would detach the annotation, which could cause onExit to be called on the annotation, since the widget was disappearing under the mouse, and thus needs to receive an onExit, but that onExit might be (and probably will be) calling setState, which marks the owning widget as needing to be built, sometimes when it already has been.
The fix creates a new _ListenerElement that overrides activate and deactivate in order to tell the render object ahead of the detach that it might be detached, and so the onExit gets called before the detach instead of during it.
In addition, I now avoid scheduling more than one check for mouse positions per frame.
Instead of using a custom WidgetController, which is very brittle, we just use the usual infrastructure.
Also, use structured data instead of an array.
This adds offsetMoreOrLessEquals to handle small floating point errors in offsets.
- Increased the AM/PM, minute and hour buttons to at least 48x48
- Added InkWells to all of them
- Adjusted the landscape layout for the AM/PM buttons to be horizontal
- Added a test to ensure the regions are at least 48x48
Fixed several layout issues with the material date picker. Mostly just removed hard coded sizes to allow the grid view to scroll instead of overflowing.
Replace CupertinoSliverRefreshControl.onRefresh's then callback with whenCompleted callback, so when onRefresh completes with error the sliver refresh control retracts like when it completes with value.
* Add Flag to determine if pixels is set by viewport during disposal
* Add TODO to remove nested TabBarView workaround once unnecessary build/dispose issues are resolved
This is a re-land of #31561, after fixing performance regressions.
Added change listening to the MouseTracker so that the Listener and tooltip can react to whether or not a mouse is connected at all. Added a change check to make sure Listener only repaints when something changed.
Fixes#22817
* Revert "Revert "Remove pressure customization from some pointer events (#30414)" (#30873)"
This reverts commit f34c2ef0a3.
* Revert pressure removal of PointerUpEvent
* Replace PR with an issue
* Add tests for fromMouseEvent
This is the correct metric to report for compilation time benchmarks rather
than RunTime. Rename the 'gen_snapshot' value to merely 'snapshot' for
backwards compatibility and overall simplicity.
This change simplifies Dart's benchmarking of Flutter by making it easier to
adopt --report-timings (made for Dart to use), which makes the benchmarks
much more robust.
In #31614, I added an unfocus() to FocusNodes to allow giving up of focus, but it only worked on the primary focus. This changes that so that it will unfocus the entire chain, not just the primary focus. Now, if you call unfocus() on a FocusNode or FocusScopeNode, and their hasFocus returns true, then after calling unfocus(), it will return false. Before this change, it would only do that if hasPrimaryFocus was also true.
This also fixes a bug in the way setFirstFocus was implemented, making it conform more to the behavior of the previous implementation. It has simplified logic in reparent, and in when it requests focus for scope nodes that have had setFirstFocus called on them.
