* Set the node to not opaque if it’s wrapping an UIActivityIndicatorView
* Add check for subclass of UIVisualEffectView to see if the node needs to be opaque
* New way to informing interaction delegate if ASCellNode get’s a setNeedsLayout call
* Remove nodeDidGetNeedsLayout as optional and better check if cell updates need to happen
* Differentiate between letting a normal display node and a cell node know that the size could have changed
* Check for _interactionDelegate is nil
* Add nullability to out pointer
* Adress comments and some renaming
Resolves#3037
This is a do-over of c85aa11, with the following differences:
- Fixed `_ASTableViewCell` properties **always** being mutated to match corresponding `ASCellNode` passthrough properties, even when the hosted `ASCellNode` is currently `nil`. Now only inheriting passthrough properties when there’s an actual instance of `ASCellNode` to read them from.
- Corrected spelling of `ASCellNode`’s `separatorInset` property.
- Reverted `_ASCollectionViewCell` inheriting the hosted cell node’s `backgroundColor` & `clipsToBounds` properties. This seems to be surprising and unwanted behavior to some, as seen in #3053 and #3044.
- Moved passthrough of `UITableViewCell`’s `selectedBackgroundView` into `_ASTableViewCell`’s `-setNode:` for consistency.
* Clean up snapshot test cases:
- Drop 32bit snapshots
- Update iOS 9 snapshots and move some others to the correct directory
- Remove `setUp` overrides in subclasses of `ASLayoutSpecSnapshotTestCase` that set `recordMode` flag. Doing so in multiple places make it difficult to enable/dis
able the flag for all layout spec test cases. It can always be overridden locally if need
to.
- Fix BUCK build
* Fix testThatOnDidLoadThrowsIfCalledOnLoadedOffMain of ASDisplayNodeTests
These changed recently with the addition of the ability to provide nodeBlocks.
It might be more strict to assert that nil can't be provided unless the datasource supports
inter-op, but I think it is also pretty reasonable to just use an empty ASCellNode in this case.
Either way, I'm really amazed / surprised to see all the improvements that appeared in ASDataController!
* Build ASElementMap and use it
* Keep building!
* Clean up
* Restrict ASDataController subclassing
* Clean up more
* More rearranging
* Dear lord it's time for bed
* Make things clearer
* Sadly remove subclassing restriction for ASDataController
* Remove dead initializer
* Fix deadlock occurs during layout of ASDisplayNode that triggered a change set update in ASDataController
* Access _transitionInProgress ivar directly if already locked
