This change creates an `AppBarTheme` to be used with `AppBar` widgets. This allows for users to theme their AppBars separately from their overall Theme if they choose.
These are essentially self-inflicted race conditions. Instead of timeouts we're going to try a more verbose logging mechanism that points out when things are taking a long time.
There's no reason to disallow this as TextStyle can handle lerping between null parameters just fine. This change also adds tests to ensure the correct behavior
* [FAB] Adding FAB as class in tests.
* [Material] FAB theme tests corrected. Also deeper into Material for elevation testing.
* [Material] Comment clarification.
* [Material] TODO formatting.
* [Material] Formatting.
* [Material] Comment correction.
* [Material] Updated values.
* [Material] Removing comment, low-level check, and adding safety.
* Add more animation tests for rotation
* Test animation progress
* Progress on the tests
* Everything 'works'
* We don't need to guarantee the rotation of the fab as it exits.
* Remove print statement
* Cleanup extra lines
* Remove whitespace
Generally, we should always use default-less switches for
platform-specific behaviour so we're forced to make a conscious decision
on each instance of such behaviour any time we add support for a new
platform.
The lerp method for IconThemeData was not doing anything to prevent exceptions when null arguments are used. This change fixes that and adds a test for it.
* Include caret gaps in position calculation to avoid descender cutoff
* Calculate caret offset instead of exposing it
* Test that textfields scroll vertically by full line height
* Only add offset to end of caret to avoid going above 0.0
* Fix and test horizontal vs. vertical scrolling
* CR: docs and top caret offset logic
* Take max instead of assuming caretRect.top is always a full line down
* Always subtract caretOffset from caretTop since caretRect.top can be negative
Generally, we should always use default-less switches for
platform-specific behaviour so we're forced to make a conscious decision
on each instance of such behaviour any time we add support for a new
platform.
Previously, a disabled floating action button always had zero
elevation, which looks dumb.
This also fixes the issue whereby highlightElevation was not honoured
on floating action buttons.
This also fixes an issue I found during testing whereby setState was
being called during build when onHighlightChanged fired due to
onPressed becoming null while a gesture is ongoing (which triggers an
onTapCancel synchronously during build).