* Add move detection and support to ASLayoutTransition
...and NSArray+Diffing.
Add some tests.
* Update CHANGELOG.md
* Update CHANGELOG.md
* Update ASLayout+IGListKit.h
* Update ASLayout+IGListKit.mm
* Use std collections to avoid NSNumber boxing
* Update ASLayoutTransition.mm
* Code review updates.
* Use `unordered_multimap` on stack instead of unordered_map<id,queue> on heap
* Remove notFound BOOL (use NSNotFound sentinel value) and put some vars inside the if (insertions/moves) loop
* Don't copy defaultCompare block (redundant under ARC)
* Whitespace
* Remove unneeded mutableCopy-s in ArrayDiffingTests
* Code review updates.
* Type _subnodeMoves pair.first to ASDisplayNode * instead of id
* C++ enumeration
* unowned refs for adding previousLayout nodes to _subnodeMoves
* Remove unreleated ASDynamicCast that is probably right though
* Add commentary to NSArray+Diffing.h; make multimap elements unowned
* Use std::make_pair, optimize ASLayout+IGListKit
* Oops I thought I had added these headers but nope
* Simplify simplify
* Diff subnodes instead of sublayouts
* Another randomized test with actual ASLayouts
This is because committing the layout transition (aka `-_completeLayoutTransition:`) results in subnode insertions and removals which must be called lock-free.
- Currently, there is a pair of mutex unlock and unlock that wraps around `-_u_measureNodeWithBoundsIfNecessary:` in `__layout`. That is because this method must be called without the lock.
- When an assertion occurs within that method, the runtime bails early without reacquire the lock (so the lock is free now). However, the runtime then hits the end of the outmost mutex locker scope and tries to release the lock that it no longer holds, causing another assertion in ASThread to be shown to user (#932). This makes it extremely hard to idenfity the root assertion.
- Fix by replacing the unlock/lock pair with a mutex unlocker.
* Adds support for having multiple interface state delegates.
Hopefully in a performant way.
* Switch to respondsToSelector for int del instead of separate object
* Add CHANGELOG
* Make ASDisplayNode+InterfaceState.h public
* Huy's comments
* Don't even bother removing since it's a weak hash table.
* fix SIMULATE_WEB_RESPONSE not imported #449
* Fix to make rangeMode update in right time
* remove uncessary assert
* avoid extra version log.
* check dictionary earlier
Fixed removing node from supernode after layout transition when automaticallyManagesSubnodes is disabled. In case if developer prefer to manage subnodes by himself then he wants to be sure that stack will not change automatically.
* fix SIMULATE_WEB_RESPONSE not imported #449
* Fix to make rangeMode update in right time
* Renew supplementary node on relayout.
* Add support for layoutDelegate (ASCollectionLayout).
* revert space
* Update change log
* fix build error
* refactor
* set default size
* return early when can delegate
* fix SIMULATE_WEB_RESPONSE not imported #449
* Fix to make rangeMode update in right time
* Match interface update closer to UIKit.
* allow the correct exiting sequence for thrashing
* refactor
* Fork dealloc queue in an experiment
* Fix and put back
* Use the right selector
* Go simpler
* Clarify name
* Type inference
* Use CFTypeRefs like a boss
* Improve comments
* License header
* [ASNetworkImageNodeTests] Add defaultImage test and improve image test. #trivial
* [ASImageNode+AnimatedImage] Fix early return when animatedImage is nil in setAnimatedImage. #trivial
* fix SIMULATE_WEB_RESPONSE not imported #449
* Fix to make rangeMode update in right time
* Fix pager node for interface coalescing
* Fix typo
* change log
* [ASTextNode2] Upgrade lock safety by protecting all ivars (including rarely-changed ones).
Although I don't know of any specific crashes caused by this, I think we should
lock all properties by default. There are also some indications of premature
optimization in keeping lock scope small, where it is actually important to
have transactional integrity, and also where the ASDisplayNode base class is
otherwise going to repeatedly re-lock the object anyway.
I think this will remain pretty efficient, especially with os_unfair_lock enabled.
* Use compare-assign macros
Because multiple threads can enter this allocWithZone: method around the same time, it is possible for one of them to setSuperclass first, and then the second thread would get stuck in an infinite loop. When climbing the inheritance heirarchy, ASTextNode2 would be encountered by this second thread, and it would continue until reaching c == Nil. Since there was no case to catch this, an infinite loop would result. Then the main thread can be deadlocked if a method like waitUntilAllUpdatesAreProcessed is called on ASCollectionView.
