Relands https://github.com/flutter/flutter/pull/149739 which was reverted due to https://github.com/flutter/flutter/issues/149851
---
fixes [`SegmentedButton` doesn't clip properly when one of the segments has `ColorFiltered`](https://github.com/flutter/flutter/issues/144990)
### Code sample
<details>
<summary>expand to view the code sample</summary>
```dart
import 'package:flutter/material.dart';
void main() => runApp(const MyApp());
class MyApp extends StatelessWidget {
const MyApp({super.key});
@override
Widget build(BuildContext context) {
return MaterialApp(
home: Scaffold(
body: Center(
child: SegmentedButton<int>(
segments: const <ButtonSegment<int>>[
ButtonSegment<int>(
value: 0,
label: ColorFiltered(
colorFilter:
ColorFilter.mode(Colors.amber, BlendMode.colorBurn),
child: Text('Option 1'),
),
),
ButtonSegment<int>(
value: 1,
label: Text('Option 2'),
),
],
onSelectionChanged: (Set<int> selected) {},
selected: const <int>{0},
),
),
),
);
}
}
```
</details>
### Before
### After
*Replace this paragraph with a description of what this PR is changing or adding, and why. Consider including before/after screenshots.*
*List which issues are fixed by this PR. You must list at least one issue. An issue is not required if the PR fixes something trivial like a typo.*
*If you had to change anything in the [flutter/tests] repo, include a link to the migration guide as per the [breaking change policy].*
Adds `mouseCursor` property in `Radio` to `CupertinoRadio` and `Radio.adaptive`.
The `mouseCursor` property added is of type `MouseCursor` and not `WidgetStateProperty<MouseCursor>` to match `Radio`'s `mouseCursor`.
## Description
This PRs fixes `DropdownMenu` behaviors when disabled.
Before this PR the `DropdownMenu` value can be changed when `DropdownMenu.enabled` was false because the inner `IconButton` was not disabled so it can get focus (for instance using tab key).
After this PR, the inner `TextField` and the inner `IconButton` are disabled when `DropdownMenu.enabled` is false.
## Related Issue
Fixes https://github.com/flutter/flutter/issues/149598.
## Tests
Updates 1 test.
Adds 1 test.
The new service extension `getRootWidgetTree` can be used instead of the existing:
* `getRootWidgetSummaryTree` --> use`getRootWidgetTree` with parameters `isSummaryTree=true`
* `getRootWidgetSummaryTreeWithPreviews` --> use `getRootWidgetTree` with parameters `isSummaryTree=true` and `withPreviews=true`
This new service extension will enable Flutter DevTools to combine the widget summary tree with the widget details tree by calling `getRootWidgetTree` with `isSummary=false` and `withPreviews=true`.
Closes https://github.com/flutter/devtools/issues/7894
Currently, the error message displayed to regenerate the lockfiles gives a Unix-like command ./gradlew, which will be incorrect for Windows environments. This PR uses globals.platform.isWindows to give the appropriate command.
closes#136763
This sentence didn't make a lot of sense as written. It looks like
it was probably based on the doc of invokeMapMethod, and then
incompletely adapted. So, fix it.
Reverts: flutter/flutter#150015
Initiated by: vashworth
Reason for reverting: Still failing tree on goldens
Original PR Author: dkwingsmt
Reviewed By: {chunhtai}
This change reverts the following previous change:
Relands #149568, which was reverted in https://github.com/flutter/flutter/pull/149998 due to unverified golden tests post-commit. (Honestly I don't know what happened but I guess resubmitting should resolve it.)
No code is changed.
Dart2Wasm doesn't officially support `dart:ffi` (only a small sketchy
subset needed in flutter web engine). We have seen user reporting issues
where existing packages don't work with dart2wasm due to using
`dart.library.ffi` in `bool.fromEnvironment` or in conditional imports
and it doing the wrong thing. So we're going to make `dart.library.ffi`
`false` in the compiler.
We therefore update the code to detect whether it runs under wasm via a
new `dart.tool.dart2wasm` environment variable.
This is a preparation for the change in dart2wasm which will start
making `const bool.fromEnvironment('dart.library.ffi')` return `false`
instead of `true`.
Notifies the engine when `PointerSignalEvents` have been ignored by the framework, through the `ui.PointerData.respond` method.
This allows the web to "preventDefault" (or not) on `wheel` events.
## Issues
* Fixes (partially): https://github.com/flutter/flutter/issues/139263
## Tests
* Added tests to ensure `respond` is called at the right time, with the right value.
## Demo
* https://dit-multiview-scroll.web.app
<details>
<summary>
## Previous versions
</summary>
1. Modified `PointerScrollEvent`, not shippable.
2. Modified when events were handled, instead of the opposite.
</details>
This PR addresses an issue with TextPainter's caret position calculation for text containing full-width spaces. Currently, the caret position is not accurately calculated for strings with full-width spaces. To resolve this, the following changes have been made:
Corrected the logic for caret position calculation when full-width spaces are present in the text.
Added and updated test cases to ensure accurate caret position calculation.
These changes ensure that the caret position for text with full-width spaces is computed correctly.
This issue was introduced by the commit [a0a854a](a0a854a78b).
Related Issue: [#149099](https://github.com/flutter/flutter/issues/149099)
- When `--web-renderer` is omitted, keep the value `null` until it later materializes to either `canvaskit` or `skwasm`.
- No more hardcoded defaults anywhere. We use `WebRendererMode.defaultForJs/defaultForWasm` instead.
- When in `--wasm` mode, the JS fallback is now `canvaskit` instead of `auto`.
- Add test for defaulting to `skwasm` when `--wasm` is enabled.
Fixes https://github.com/flutter/flutter/issues/149826
This reverts commit 32081aab69.
Reason for revert: Appears to be failing tests in tree
```
01:48 +2991 ~18: /Volumes/Work/s/w/ir/x/w/flutter/packages/flutter/test/cupertino/action_sheet_test.dart: Taps on a button can be slided to other buttons
âââ¡ EXCEPTION CAUGHT BY FLUTTER TEST FRAMEWORK âââââââââââââââââââââââââââââââââââââââââââââââââââââ
The following SkiaException was thrown while running async test code:
Skia Gold received an unapproved image in post-submit
testing. Golden file images in flutter/flutter are triaged
in pre-submit during code review for the given PR.
Visit https://flutter-gold.skia.org/ to view and approve
the image(s), or revert the associated change. For more
information, visit the wiki:
https://github.com/flutter/flutter/wiki/Writing-a-golden-file-test-for-package:flutter
Debug information for Gold --------------------------------
stdout: Given image with hash 601f421e3bb643f437cc12395de96152 for test
cupertino.cupertinoActionSheet.press-drag
Expectation for test: ce8ef79c146857d162663b1161fd6d5c (positive)
Expectation for test: 3bb57a8782f67836c6ad3ece7f00729a (positive)
Expectation for test: 3f8c592774caf3c760fbbd318a7dc5af (positive)
Expectation for test: 61863f38217aa3349c239eeb49b84930 (positive)
Expectation for test: 98af16e9b7eb3fb810b44c74bb18ffb9 (positive)
Untriaged or negative image:
https://flutter-gold.skia.org//detail?grouping=name%3Dcupertino.cupertinoActionSheet.press-drag%26source_type%3Dflutter&digest=601f421e3bb643f437cc12395de96152
stderr: Test: cupertino.cupertinoActionSheet.press-drag FAIL
result-state.json: No result file found.
```
This PR matches the various colors of `CupertinoActionSheet` more closely with the native one.
The following colors are changed.
* Sheet background color
* Pressed button color
* Cancel button color
* Pressed cancel button color
* Divider color
* Content text color
The resulting colors match with native one with deviation of at most 1 (in terms of 0~255 RGB).
The following are comparison (left to right: Native, Flutter after PR, Flutter current)
<img width="1295" alt="image" src="https://github.com/flutter/flutter/assets/1596656/3703a4a8-a856-42b1-9395-a6e14b1881ca">
<img width="1268" alt="image" src="https://github.com/flutter/flutter/assets/1596656/1eb9964e-41f1-414a-99ae-0a2e7da8d3fd">
_Note: The divider thickness is adjusted to `1/dpr` instead of 0.3 in both Flutter version to make them look more native, as will be proposed in https://github.com/flutter/flutter/pull/149636._
### Derivation
All the colors are derived through color picker and calculation. The algorithm is as followed:
* Assume all colors are translucent grey colors, i.e. having the same value `x` for R, G, and B, with an alpha `a`.
* Given the barrier color is `x_B1=0` when the background is black, and `x_B2=204` when the background is white.
* Pick the target color `x_t1` when the background is black, and `x_t2` when the background is white
* Solve the following equations for `x` and `a`
```
a * x + (1-a) * x_B1 = x_t1
a * x + (1-a) * x_B2 = x_t2
a = 1 - (x_t1 - x_t2) / (x_B1 - x_B2)
x = (x_t1 - (1-a) * x_B1) / a
```
These equations use a linear model for color composition, which might not be exact, but is close enough for an accuracy of (1/255).
The full table is as follows:
<img width="1091" alt="image" src="https://github.com/flutter/flutter/assets/1596656/0fb76291-c3cc-4bb5-aefa-03ac6ac9bf1f">
* The first two columns are colors picked from XCode.
* The 3~4 columns are the colors picked from the current Flutter. Notice the deviation, which is sometimes drastic.
* The 5~6 columns are the colors picked from Flutter after this PR. The deviation is at most 1.
* The last few columns are calculation.
* There are two rows whose calculation is based on adjusted numbers, since the original results are not accurate enough, possibly due to the linear composition.
During the calculation, I assumed these colors vary between light and dark modes, but it turns out that both modes use the same set of colors.
### Screenshots
This reverts commit 9d1de7b674.
Reverts due to log spam and crashing of the dependency version checker (which doesn't block the build but still isn't the desired outcome).
Work towards https://github.com/flutter/devtools/issues/7894
The Flutter DevTools team is working on a new Widget Inspector. As part of this work, we need to make adjusts to the widget inspector service API.
This PR simply re-factors the `widget_inspector_test` in preparation for this work, to make it easier for us to modify the tests / add new tests.
Although this PR is large (`widget_inspector_test` is over 5000 lines!!), it does not change any of the current test logic.
Instead it:
- Moves test cases that were prefixed with "WidgetInspectorService" into the pre-existing group named "WidgetInspectorService"
- Moves tests cases in that group which were skipped with `!WidgetInspectorService.instance.isWidgetCreationTracked() // [intended] Test requires --track-widget-creation flag` into a group called "Requires flag --track-widget-creation"
- Adds two helper functions, `pumpWidgetTreeWithABC` and `findElementABC` and uses them in all test cases that were duplicating the same widget set-up.
If the service is disposed when we make calls on it, we need to bubble up an exception that is understood by the callers.
This is another speculative fix for https://github.com/flutter/flutter/issues/149238.
We have fiddled with this a bunch, and there is definitively no way to reliably include this, so I am removing it.
Since CI does not cover all possible values, we can't consistently look up images for local testing. We thought removing it from the look up would work fine, and it did for most tests, but there were still some that could not find an image without it.
A brief history on the abi key
- added in https://github.com/flutter/flutter/pull/143621
- disabled in https://github.com/flutter/flutter/pull/148023
- added back in https://github.com/flutter/flutter/pull/148072
- we thought there was only an issue with alphabetizing keys
- removed from local image look up in https://github.com/flutter/flutter/pull/149696
I updated the docs page to also discuss what makes a good key and what does not based on what we learned here.
This PR implements the `CupertinoActionSheet` part of https://github.com/flutter/flutter/issues/19786.
This PR creates a new kind of gesture "sliding tap", which can be a simple tap but also allows the user to slide to other buttons during the tap, and select the button that the slide ends in.
The following video shows the behavior on a button list (left to right: Native iOS, after PR, before PR):
https://github.com/flutter/flutter/assets/1596656/1718630d-6890-4833-908b-762332a39568
Notice:
1. When the tap starts on a button or on the content section, the gesture can slide into a button to highlight it or activate it.
2. When the user performs a quick tap on a button, the button is immediately highlighted. (Before the PR, the highlight waits until the time out of a tap gesture, causing a quick tap to not highlight anything at all.)
The following video shows the behavior when the actions section scrolls (left to right: Native iOS, after PR)
https://github.com/flutter/flutter/assets/1596656/5eb70bc1-ec25-4376-9500-2aaa12f0034b
Notice:
1. Moving left or right doesn't cancel sliding tap, but moving vertically prioritizes the scrolling.
2. Moving before the drag starts is also recognized as a sliding tap.
Also, multiple pointers interact with the button list following an algorithm of "using earliest pointer". I can't record videos about it but I've added unit tests.
A sliver that is pinned to the start of its `CustomScrollView` and reacts to scrolling by resizing between the intrinsic sizes of its min and max extent prototypes.
The minimum and maximum sizes of this sliver are defined by `minExtentPrototype` and `maxExtentPrototype`, a pair of widgets that are laid out once. You can use `SizedBox` widgets to define the size limits.
This sliver is preferable to the general purpose `SliverPersistentHeader` for its relatively narrow use case because there's no need to create a `SliverPersistentHeaderDelegate` or to predict the header's minimum or maximum size.
The sample shows how this sliver's two extent prototype properties can be used to create a resizing header whose minimum and maximum sizes match small and large configurations of the same header widget.
https://github.com/flutter/flutter/assets/1377460/fa7ced98-9d92-4d13-b093-50392118c213
Related sliver utility PRs: https://github.com/flutter/flutter/pull/143538, https://github.com/flutter/flutter/pull/143196, https://github.com/flutter/flutter/pull/127340.
This change adds a new property in Semantics widget that would take an integer corresponding to the heading levels defined by the ARIA heading role. This is necessary in order to get proper accessibility and usability in a website for users who rely on screen readers and other assistive technologies.
Issue fixed by this PR:
fixes https://github.com/flutter/flutter/issues/97894
Engine part:
https://github.com/flutter/engine/pull/41435
Unpin `archive` version and run `flutter update-packages --force-upgrade` to show there are no additional dependencies pulled in (url_launcher was updated, but unrelated).
Fixes https://github.com/flutter/flutter/issues/149427
Fixes https://github.com/flutter/flutter/issues/149435
CI does not comprehensively cover all of the possible abi keys, so some folks can't test locally if their machine's abi does not match CI. This just removes it from the local testing look up. The right image will be summoned, and in CI we can still keep track of it.
fixes https://github.com/flutter/flutter/issues/142393
The issue is that if routerdelegate mutate its currentConfiguration outside of Router's workflow, the change will be propagate back to routeinformationprovider at the end of the frame.
However if another things trigger router's dependencies change within the same frame. The dependencies will trigger a reparse of the current route which would end up override the currentConfiguration in routerDelegate.
This change introduce a transaction system that each operation will be add to the end of the transaction future so everything will be execute inorder
