Rather than have delayed calls to super.didPop(), which raises my
eyebrow every time I see it, this provides a separate finished()
function to call, and uses the convention that if you want to call it
yourself, you just don't call super.didPop().
- `Scaffold.of(context).showBottomSheet(widget);`
- Returns an object with .closed Future and .close() method.
- Uses a StateRoute to handle back button.
- Take the Navigator logic out of the BottomSheet widget.
- Support showing a sheet while an old one is going away.
- Add Navigator.remove().
Many of the widgets that use CustomPaint were spamming repaints because
CustomPaint repaints when the identity of the onPaint callback changes, which
it does every build for StatelessComponents.
This patch changes CustomPaint to use a CustomPainter, similar to the new
custom layout widgets. The CustomPainter has a `shouldRepaint` function along
with its `paint` function. This function gives clients explicit control over
when the custom paint object repaints.
"showSnackBar()" is now a feature of a Scaffold. To get to a Scaffold
you either use a global key (`scaffoldKey.currentState.showSnackBar(...)`),
or you use `Scaffold.of(context)`.
Snack bars no longer have a route. They are entirely managed by the
Scaffold. Fixes#432.
Snack bars now queue up when you have several of them. Fixes#374.
Snack bars now auto-size themselves around their contents. This is step
one towards implementing multiline snack bars.
Snack bars now self-dismiss after some per-snackbar configurable period.
The self-dismissing pauses while a dialog is up above the snackbar (or
anything that uses ModalRoute). To enable this, there's now a
`ModalRoute.of(context)` API that returns the current ModalRoute, and
you will be rebuilt if you asked for this and the route's "current"
status changes. To implement this, the Navigator now rebuilds
unconditionally any time it pushes or pops a route.
Snack bars now use the curves that Android uses for snack bars.
Snack bar contents now fade in.
Shadows now render as three seprate MaskFilter.blur components per the most recent Material spec.
The shadows Map was replaced by a similar Map called elevationToShadow with entries that match the 10 elevations specifed by http://www.google.com/design/spec/what-is-material/elevation-shadows.html.
The "level" property (many classes) is now called "elevation", to match the Material spec.
BoxShadow now includes a spreadRadius parameter - as in CSS box-shadow. Renamed the BoxShadow blur property to blurRadius to further align BoxShadow with CSS box-shadow.
A common pattern is to use a Positioned with a Sized box to give both an offset
from the edge as well as a fixed size to the child. This patch rolls into into
the Stack layout algorithm for simplicity.
Fixes#250
PageRoute is now MaterialPageRoute.
This also changes the following:
- Now the HeroController is a Navigator observer, rather than a feature
of HeroPageRoutes, which are gone. This means heroes can work between
any kind of ModalRoute now.
- ModalPageRoute is moved from modal_barrier.dart to routes.dart.
- It allows routes to opt-out of their modal barrier being a shortcut to
popping the route.
- Features of PageRoute that aren't Material-specific get promoted to
ModalRoute features: storage, the subtree key, offstageness...
The AnimatedModalBarrier is still a ModalRoute feature.
- force the time dilation to 1.0 for the Widget tests, so that a local
change doesn't break all the tests during development.
- add missing license block to all the files.
- set ui.window.onBeginFrame to null when you use WidgetTester, so that
the engine doesn't trigger any confusing frames after our fake frames.
Prior to this patch, MultiChildLayoutDelegate couldn't be re-entered because it
cleared _idToChild when unwinding its stack. Now we restore the previous value
of _idToChild when we unwind.
Make sure to send tapcancel when the primary pointer fails because of
slop, even if the gesture won by default.
Also, minor cleanup and clarification of an invariant.
When we reactivate a subtree that had a global key, we weren't updating the
parent data because:
1) The child wasn't in the tree when we updated the parent data element.
2) The activated child didn't go through mount (just through update).
This patch moves the parent data update work to when we attach the render
object, which we do both during mount and when reactivating a child.
Fixes#345
Turns out that ignoring all error lines that match the empty string is a
poor way to go.
Also, we have to update all the example packages now too, since we
analyze them. So just have travis use our update script.
Also, remove flutter_tools' old travis stuff. It's now part of a bigger
repo.
Also, make travis use the dev Dart SDK, since we need the new analyzer.
Stable is way too out of date, e.g. it still complains about libraries
not having names and mixins using 'super', and the strong mode hints are
even more aggressive than on dev.
Other changes in this patch:
- Make the 'flutter' tool say "Updating flutter tool..." when it calls
pub get, to avoid confusion about what the pub get output is about.
- Make the bash flutter tool call pub get when the revision has
changed. (This was already happening on Windows.)
- Fix a raft of bugs found by the analyzer.
- Fix some style nits in various bits of code that happened to be near
things the analyzer noticed.
- Remove the logic in "flutter test" that would run "pub get", since
upon further reflexion it was determined it didn't work anyway.
We'll probably have to add better diagnostics here and say to run the
updater script.
- Remove the native velocity tracker script, since it was testing code
that has since been removed.
Notes on ignored warnings:
- We ignore warnings in any packages that are not in the Flutter repo or
in the author's current directory.
- We ignore various irrelevant Strong Mode warnings. We still enable
strong mode because even though it's not really relevant to our needs,
it does (more or less accidentally) catch a few things that are
helpful to us.
- We allow CONSTANTS_LIKE_THIS, since we get some of those from other
platforms that we are copying for sanity and consistency.
- We allow one-member abstract classes since we have a number of them
where it's perfectly reasonable.
- We unfortunately still ignore warnings in mojom.dart autogenerated
files. We should really fix those but that's a separate patch.
- We verify the actual source file when we see the 'Name non-constant
identifiers using lowerCamelCase.' lint, to allow one-letter variables
that use capital letters (e.g. for physics expressions) and to allow
multiple-underscore variable names.
- We ignore all errors on lines that contain the following magic
incantation and a "#" character:
// analyzer doesn't like constructor tear-offs
- For all remaining errors, if the line contains a comment of the form
// analyzer says "..."
...then we ignore any errors that have that "..." string in them.
Add a unit test for persistent bottom sheets
ShowBottomSheet now clears its placeholder after the bottom sheet has been dismissed.
Added support for generating fling gestures to WidgetTester.
Factored OverlayRoute out of the modal and persistent bottom sheet clases, since the bottom sheet classes need to drive the performance.
Added a bottom sheet to the stocks demo: long-press on a stock shows a modal bottom sheet.
Made AnimatedModalBarrier public.
- animation_bench.dart. This benchmark measures the full main-thread pipeline
for ticking the drawer entrance and exit animation.
- build_bench.dart. This benchmark measures a full app rebuild when there's no
state change.
- layout_bench.dart. This benchmark measures a full relayout.
This script runs the Flutter unit tests. By default, the script assumes you
have compiled a SkyShell in an "engine/src" that's a peer to the "flutter"
directory.
Now a RenderBox is considered hit if one of its children are hit or it itself
decides that it's hit. In particular, empty space inside a flex won't be hit
because none of the children are located there and a RenderFlex doesn't
consider itself hittable.
Fixes#53Fixes#1221
Previously, we'd leave the old values in the parent data if the types matches,
but not all render objects would reset these values during layout. For example,
RenderProxyBox doesn't set the position field because it doesn't read the
position field. However, leaving the old data there violates the invariants of
the box protocol and can cause trouble (e.g., localToGlobal giving the wrong
result).
Fixes#1939
Document the Gesture Detector constants.
Remove kEdgeSlop since it's obsolete on Android.
Add a test that verifies that when the first tap is canceled, the second
tap can become the first tap of a subsequent two-tap sequence.
