fuchsia.ui.observation.geometry

The minimal position of the view's bounding box.

The maximal position of the view's bounding box.

The minimal and maximal points of a view's bounding box, in the coordinate system of that view.

The conversion ratio from physical pixels (of a display) to logical pixels (of the coordinate system of the view).

The offset data for the view's bounding box, in the coordinate system of that view.

The origin point of V's bounding box, in W's coordinate system.

The width of V's bounding box (along the direction where V's x axis increases), in W's coordinate system.

The height of V's bounding box (along the direction where V's y axis increases), in W's coordinate system.

The clockwise rotation about the origin, in degrees.

1

This view's fuchsia.ui.views.ViewRef koid.

2

This view's origin, logical size, pixel scale, and inset data, in the view's own coordinate system.

Limitations. Data consistency between server and client depend on the specific graphics API. Some APIs provide weak consistency, where the server-side data (this data) and the client-side data (in the view's UI client) are allowed to diverge for some time.

3

This view's extent, in the context view's coordinate system. It does NOT describe the child view's logical size.

This describes the "ground truth" position of this view within the context view, regardless of view tree depth, or specific layout state of intermediate views.

Limitations. It does NOT describe whether the view is "visible" (e.g., whether the view has opacity applied, or is not occluded by another view), and it does NOT describe whether the view is "hittable" (e.g., whether the view is positioned fully inside of every ancestor view's bounding box).

4

The space occupied within the parent view's coordinate system. It does NOT describe the child view's logical size.

5

The list of child views, in the order known to the graphics API.

Each integer in this vector refers to the child's position in the |views| or |incomplete| vector that the parent is in.

The identity, position, and size of each child view. Position and size are described by the extent of the child view within the parent view's coordinate system.

The view tree topology is reliable. A child placed here is equivalent to the parent view receiving a "child view connected" signal.

Limitations. A child's view boundary is described in the parent view's coordinate system, which is subject to weak consistency (depending on the graphics API). That is, when a parent view has a change in size or metrics, the context view may observe a "jump" as the parent view incorporates those data. In such cases, a new ViewTreeSnapshot is issued to describe the change in position, relative to the context view.

1

When the snapshot was taken. Timebase is monotonic time.

2

The context view (at element 0) and a complete list of its descendant views.

If MAX_VIEW_COUNT is exceeded, this field is not set, and an error is reported in Error.

1

When the response is sent. Timebase is monotonic time.

2

A list of most recent updates for a particular view.

3

Only set if an error condition is detected. If unset, the client may assume that updates has complete information over its epoch.

CHANNEL_OVERFLOW

Set to true when appending a ViewTreeSnapshot in WatchResponse.updates would exceed the limit of the FIDL channel. That snapshot is dropped, along with older snapshots.

BUFFER_OVERFLOW

Set to true when appending a ViewTreeSnapshot in WatchResponse.updates would exceed BUFFER_SIZE. That snapshot is dropped, along with older snapshots.

VIEWS_OVERFLOW

Set to true when the size of views in a ViewTreeSnapshot exceeds MAX_VIEW_COUNT. We represent this situation in the ViewTreeSnapshot with an unset views field.

The maximum number of ViewTreeSnapshots a client can expect in a Watch call's response. The number of ViewTreeSnapshots can be less than BUFFER_SIZE when the size of a Watch call's response exceeds the limit of a FIDL channel.

There is a limit on the number of ViewTreeSnapshots that can be included in a Watch call's response due to the size restrictions of a FIDL channel. This limit was calculated by finding out the maximum number of ViewTreeSnapshots which can be included in a Watch call's response when MAX_VIEW_COUNT is 1.

Note: This value would have to adjusted when modifying ViewTreeSnapshot or ViewDescriptor.

The maximum number of ViewDescriptors a client can expect in a ViewTreeSnapshot.

There is a limit on the number of ViewDescriptors that can be included in a ViewTreeSnapshot due to the size restrictions of a FIDL channel. This limit was calculated by finding out the maximum number of ViewDescriptors which can be included in a ViewTreeSnapshot when the BUFFER_SIZE is 1.

Note: This value would have to adjusted when modifying ViewDescriptor.

The ratio from physical pixels (of a display) to logical pixels (of the coordinate system of a view).

• The values are placed in (x, y) order.