The main purpose of this PR is to make it so that when you set the
initial route and it's a hierarchical route (e.g. `/a/b/c`), it
implies multiple pushes, one for each step of the route (so in that
case, `/`, `/a`, `/a/b`, and `/a/b/c`, in that order). If any of those
routes don't exist, it falls back to '/'.
As part of doing that, I:
* Changed the default for MaterialApp.initialRoute to honor the
actual initial route.
* Added a MaterialApp.onUnknownRoute for handling bad routes.
* Added a feature to flutter_driver that allows the host test script
and the device test app to communicate.
* Added a test to make sure `flutter drive --route` works.
(Hopefully that will also prove `flutter run --route` works, though
this isn't testing the `flutter` tool's side of that. My main
concern is over whether the engine side works.)
* Fixed `flutter drive` to output the right target file name.
* Changed how the stocks app represents its data, so that we can
show a page for a stock before we know if it exists.
* Made it possible to show a stock page that doesn't exist. It shows
a progress indicator if we're loading the data, or else shows a
message saying it doesn't exist.
* Changed the pathing structure of routes in stocks to work more
sanely.
* Made search in the stocks app actually work (before it only worked
if we happened to accidentally trigger a rebuild). Added a test.
* Replaced some custom code in the stocks app with a BackButton.
* Added a "color" feature to BackButton to support the stocks use case.
* Spaced out the ErrorWidget text a bit more.
* Added `RouteSettings.copyWith`, which I ended up not using.
* Improved the error messages around routing.
While I was in some files I made a few formatting fixes, fixed some
code health issues, and also removed `flaky: true` from some devicelab
tests that have been stable for a while. Also added some documentation
here and there.
* Revert "Revert "Change all ocurrences of '$runtimeType#$hashCode' to use the idAndType method. (#10871)" (#10880)"
This reverts commit ceb814aa647ecd310794a72a7a5c28820ec57a25.
* Change all instances of '$runtimeType#$hashCode' to use the describeIdentity method.
The describeIdentity method generates a shorter description with a consistent length
consisting of the runtime type and the a 5 hex character long truncated
version of the hash code.
Introduce CompositedTransformTarget and CompositedTransformFollower
widgets, corresponding render objects, and corresponding layers.
Adjust the way text fields work to use this.
Various changes I needed to debug the issues that came up.
* Add more tests of toStringDeep.
These tests will help ensure toStringDeep behavior is not regressed
by refactoring toStringDeep to support structured data.
* Rolls the engine to 75c74dc463d56e17be10315cfde409010fd8f90b.
* Adds framework support for the `AppLifecycleState.inactive` and `AppLifecycleState.suspending` states.
What are the new states?
------------------------
* `AppLifecycleState.inactive` is emitted on iOS only and corresponds to iOS's foreground inactive state. Current iOS state transitions are:
`resumed` <--> `inactive` <--> `paused`
* `AppLifecycleState.suspending` is currently emitted on Android only and corresponds to the
transition to Android's stopped state. Current Android state transitions are:
`resumed` <--> `paused` --> `suspending` --> `resumed`
These transitions may change in future.
This is a breaking change on iOS
--------------------------------
This changes the semantics of the `AppLifecycleState.paused` state on
iOS. The behaviour associated with this state is unchanged on Android.
For background on iOS application states see the state transition
diagram in the [App Programming Guide for iOS](https://developer.apple.com/library/content/documentation/iPhone/Conceptual/iPhoneOSProgrammingGuide/TheAppLifeCycle/TheAppLifeCycle.html#//apple_ref/doc/uid/TP40007072-CH2-SW6).
On iOS, prior to this change, `AppLifecycleState.paused` corresponded to
a transition to the _foreground inactive_ state. It now corresponds to a
transition to the _background state_. The newly-added
`AppLifecycleState.inactive` state now corresponds to entering the
_foreground inactive_ state, which (currently) has no exact analogue on
Android.
Briefly, the _foreground inactive_ state is the state entered when
switching from an app to the app switcher, receiving a phone call, or
responding to a TouchID request. Apps are permitted to continue
animating/updating views in this state if desired. From the _foreground
inactive_ state, the app may transition back to _active_ (e.g., entering
the app switcher then resuming the app), or to the _background_ state
(e.g., switching to the home screen or another app).
What to change
--------------
If your app does not handle the `AppLifecycleState.paused` state in a
`WidgetsBindingObserver.didChangeAppLifecycleState` implementation, no
changes are required.
If you do handle `AppLifecycleState.paused`, you may additionally wish
to also handle `AppLifecycleState.inactive`. For example, games should
probably pause on entering the app switcher rather than wait to be
backgrounded.
More details
------------
For background on Android application states see the state transition
diagram in the [Android Activity Lifecycle docs](https://developer.android.com/guide/components/activities/activity-lifecycle.html#alc).
Previously we used logical pixels. This made the accessibility metrics
tiny on modern devices, since the OS works in physical units.
Also add a bit more debugging info and some docs.
Fixes https://github.com/flutter/flutter/issues/8476
More detailed list of changes in this patch:
* Replaced the didTouch special logic with more generic logic that
uses Activities instead. Now instead when you tap down the
Scrollable calls `hold()` which begins a HoldScrollActivity which is
a hybrid of DragStartDetails and IdleScrollActivity and can be
canceled. When you let go, it gets canceled and that goes ballistic.
* Make DragGestureRecognizer more aggressive about grabbing pointers,
otherwise a second pointer in a situation with competing horizontal
and vertical recognizers always gets taken by the other one.
* Fixed the _GestureSemantics widget to call the "down" callbacks so
that it follows the same pattern as "real" interactions.
* Added tests for the above.
* Added a hashCode to ScrollActivity.toString (and subclasses).
* Added a toString to ScrollDragController, and include it in
DragScrollActivity's toString.
* s/coorindator/coordinator/
* Add a comment in DragStartDetails to distinguish it from the
otherwise identical DragDownDetails, so we're not tempted to merge
them.
This splits the frame pipeline into two, beginFrame and drawFrame.
As part of making this change I added some debugging hooks that helped
debug the issues that came up:
* I added debugPrintScheduleFrameStacks which prints a stack whenever
a frame is actually scheduled, so you can see why frames are being
scheduled.
* I added some toString output to EditableText and RawKeyboardListener.
* I added a scheduler_tester.dart library for scheduler library tests.
* I changed the test framework to flush microtasks before pumping.
* Some asserts that had the old string literal form were replaced by
asserts with messages.
I also fixed a few subtle bugs that this uncovered:
* setState() now calls `ensureVisualUpdate`, rather than
`scheduleFrame`. This means that calling it from an
AnimationController callback does not actually schedule an extra
redundant frame as it used to.
* I corrected some documentation.
* Adjust the defaults behaviour of scroll views.
Now, primary scroll views scroll by default. Others only scroll if necessary.
* apply suggested changes
* Add a text formatter interface used by EditingText. Provide some default implementations.
* self nits
* Handle -1 selection offsets
* review notes
* simplify regular expression
* Add whitelisting formatters. Use a custom phone number formatter in text demo.
* review notes
* not being able to addAll(null) is pretty annoying
* review notes
* partial tests
* Add tests
* didn’t end up needing mockito
* move to services
Also:
* Add three explicit sizing modes to Stack for non-positioned
children: loose, expand, and passthrough. (All three are used.)
* Fix a bug whereby layers would try to paint in the same frame as
they were removed from layout (but not detached).
* Fix a bug whereby Offstage wasn't properly marking the parent dirty
when changing its sizedByParent flag.
* Explicitly make Overlay expand non-positioned children.
* Explicitly have InputDecoration pass through the constraints from
its Row to its Stack children.
