After this patch, if you try to use a widget that depends on being enclosed in
a material, you now get an assert and a debugPrint if you're not inside a
material.

Fixes #243

Factor out a reusable interface called Decoration from BoxDecoration.

Make all the consumers of BoxDecoration and the erstwhile BoxPainter
into consumers of Decoration.

Make a BoxPainter be something you get from a Decoration, rather than
something to which you pass a BoxDecoration.

Rename Shape to BoxShape now that it's documented specifically as
applying to boxes.

Move EdgeDims to its own file.

Move FractionalOffset up so that it's with the other helper classes in
its file rather than alone at the end.

Minor change to RenderClipOval's hit testing to avoid taking an
unnecessary square root.

Rename BoxDecorationPosition to DecorationPosition since
RenderDecoratedBox now takes any Decoration.

Implement hit testing for rounded rects.

Rename AnimatedBoxDecorationValue to AnimatedDecorationValue, and make
it support lerping across any Decoration (by deferring to the objects
involved).

This patch contains a number of improvements to the material ink response:

- The ink response now remains until you lift your finger
- When disappearing, the ink response now fades out
- The ink response is now the correct color (at least in the light theme)
- The ink response for IconButton now has a (circular) highlight
- The ink response for IconButton now repositions itself to be centered on the highlight

In addition, I've adjusted the various animation parameters to better match the
behavior of ink responses in the Java implementation of material.

Fixes #695

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 #53
Fixes #1221

Fixes #1807

Don't use a timeout to cancel tap tracking. Track only one primary pointer
and ignore non-primary pointers. Update tests to reflect desired behaviors.

Fixes #1779, #1780, #1781.

The goal is to follow the guidelines in
https://github.com/flutter/engine/blob/master/sky/specs/style-guide.md#packages

Fixes #1638

These are now part of material.dart.

Now we trigger the highlight from the InkWell so that it matches the tap.

Fixes #1525

In the vast majority of cases, folks should be interacting with the Widget
rather than its State.

Fixes #267

Fixes #1170

After this patch, InkWell is driven by gesture recognizers, which lets us
cleanly cancel splashes when the user actually scrolls.

I've also refactored all the clients of InkWell to use InkWell to detect
gestures instead of wrapping InkWell in a GestureDetector.

Fixes #1271

 * Rename GestureTapListener (and friends) To GestureTapCallback to match the
   other gesture callbacks.
 * Replace "ensureFoo" pattern with ??= operator.

All the use cases for EventDisposition are now addressed by the gesture
detection system.

Move the animation libraries into src/animation and change importers to use
package:sky/animation.dart. Also, move scheduler.dart into the animation
library so that the animation library can be self-contained.

Code outside of package:sky should import this code using

package:sky/rendering.dart
package:sky/widgets.dart

Moving this code into the "src" directory is a convention that signifies that
and it cleans up the generated dartdoc because the libraries in the src
directory aren't included in the generated documentation. Instead, the classes
are documented in the widgets.dart and rendering.dart libraries.

Sadly, box.dart has grown much longer than 1000 lines. This patch splits it up
into several files based on the class hierarchy. Fortunately, many of these
classes are loosely coupled to each other.

The name `root` is confusing because this value isn't the root of anything.
It's just the associated `RenderObject` instance.

We'll need this for compositing because we need to switch out the sky.Canvas
when we switch compositing layers.

And also Material Design of course.

@abarth

It was confusing to have both widget.dart and widgets.dart

This introduces the notion of event disposition and allows event
targets (widgets and render objects) to consume events that should not
be processed further. This is needed by the Switch component in the
Drawer in the stocks example. The Switch is embedded in a DrawerItem.
The Switch handles the gesture tap event to toggle its state and should
handle pointer events to allow swiping and draw its own radial
reaction. The DrawerItem also handles gesture taps to allow toggling
the switch value when tapping anywhere on the drawer and to draw its
own ink splash. When tapping on the switch, both the switch's render
object and the DrawerItem's listener are in the event dispatch path.
The Switch needs to signal in some fashion that it consumed the event
so the DrawerItem does not also try to toggle the switch's state.

This fixes some theoretical bugs whereby we were using hashCode to try
to get unique keys for objects, but really we wanted object identity.
It also lays the groundwork for a new GlobalKey concept.

I tried to keep the impact on the code minimal, which is why the "Key"
constructor is actually a factory that returns a StringKey. The code
has this class hierarchy:

```
   KeyBase
    |
   Key--------------+---------------+
    |               |               |
   StringKey    ObjectKey       UniqueKey
```

...where the constructors are Key and Key.stringify (StringKey),
Key.fromObjectIdentity (ObjectKey), and Key.unique (UniqueKey).

We could instead of factory methods use regular constructors with the
following hierarchy:

```
   KeyBase
    |
   LocalKey---------+---------------+
    |               |               |
   Key      ObjectIdentityKey   UniqueKey
```

...with constructors Key, Key.stringify, ObjectIdentityKey, and
UniqueKey, but I felt that that was maybe a more confusing hierarchy.
I don't have a strong opinion on this.