-Removed: BooleanSet classes are gone. (Unused anyways)
-Removed: Boolean2xMaps are also now gone. (Unused Anyways)
-Added: GlobalFlags that are shared across packages.
Each List implementation can be now turned off.
Or all Lists can be turned off.
ListIterator can't be turned of since it is a IndexBasedIterator and not
a List and a few implementation use them.
This will be a small rewrite of the "Builder".
Why this is done?
Basically this library is really big and having the Option to turn of
modules that you don't need is really important.
And right now enabling/disabling modules is really difficult since
everything is merged together.
So the first goal is to modularized where everything can be turned on or
of as wished.
- Added: ISizeProvider interface (Optimization Helper)
- Added: ISizeProvider into most Iterable implementations
(Distinct/Filter/FlatMap/ArrayFlatMap don't support it, for obvious
reasons)
- Added: ToArray function into Iterable which uses ISizeProvider to
reduce overhead of duplicating arrays.
- Fixed: putIfAbsent now replaces defaultValues
-Added: Better System to decide what code should be generated.
-Added: Unit Tests are now included in the main branch.
-Added: Automatic System to test the main branch and ensure that tests work.
-Added: Pull Requests now run tests too.
- Added: Coverage Badge
- Updated: Changelog.
- Changed: Changelog now has info how to obtain the sourcecode.
- Added: Over 11 Million Unit Tests to this library to ensure quality.
- Added: ArrayList size constructor now throws IllegalStateException if
the size parameter is negative
- Added: EnumMap specialized forEach implementation.
- Added: AbstractMap.remove now delegates to its primitive counterpart.
- Added: ConcurrentHashMap now implements ITrimmable
- Refactor: Removed a lot of disabled code from ArraySet.
- Removed: LinkedList.addAll(index, List) now delegates to
LinkedList.addAll(index, Collection) due to no special optimization
required.
- Fixed: AbstractList.SubList.get/set/swapRemove didn't calculate their
List index Properly
- Fixed: AbstractList.SubList chains now properly if you create SubLists
within SubLists.
- Fixed: AbstractList.Iterator.add now respects
Immutable/UnmodifiableLists.
- Fixed: AbstractList.Iterator.skip/back now keep track of the last
returned value for remove function to work properly.
- Fixed: CopyOnWriteArrayList.extract/removeElements(int, int) does now
proper range checks and remove elements properly.
- Fixed: CopyOnWriteArrayList.SubList now works properly. (Reimplemented
entirely)
- Fixed: CopyOnWriteArrayList.Iterator.previous() was returning the
wrong values.
- Fixed: CopyOnWriteArrayList.Iterator.skip now skips the right amount
of elements and stops where it should.
- Fixed: LinkedList.first/last/dequeue/dequeueLast now throws
NoSuchElementException when empty instead of IllegalStateException.
- Fixed: LinkedList had an edge case where the entire reverse iterator
would break if the wrong element was removed.
- Fixed: LinkedList.extractElement now returns the correct values.
- Fixed: AbstractMap.entrySet().remove(Object) now returns true if
defaultReturnValue elements were removed.
- Fixed: ConcurrentHashMap.remove(Object, Object) checks if the type
matches before comparing against null Values.
- Fixed: LinkedHashMap.clearAndTrim() was checking the wrong value for
determining the full reset or clearing of a Map.
- Fixed: HashMap.trim/clearToTrim() was using the wrong value to
determin if something should be done.
- Fixed: HashMap now compares empty values (0) against nullKeys when
Object Variants of the type are used.
- Fixed: ImmutableMap now compares empty values (0) against nullKeys
when Object Variants of the type are used.
- Fixed: ArrayMap.iterator(key) now throws NoSuchElementException when
the element wasn't found.
- Fixed: Linked/EnumMap array constructor was creating the wrong size
values array.
- Fixed: LinkedEnumMap.getAndMoveToFirst/Last was moving elements even
if the element wasn't present.
- Fixed: AVL/RBTreeMap.getFirst/LastKey was not throwing a
NoSuchElementException if the map was empty.
- Fixed: Map.Builder wasn't throwing a IllegalStateException when
creating a negative size builder.
- Fixed: AVL/RBTreeSet.DecendingSet.subSet(from, fromInclusive, to,
toInclusive) was creating a corrupt asending subset.
- Fixed: ArraySet throws now a IllegalStateException when trying to
create it with a negative size.
- Fixed: ArraySet.addMoveToLast(key) was crashing when a key was already
present.
- Fixed: Immutable/LinkedHashSet now keep track of their iteration index
properly.
- Fixed: LinkedHashSet.moveToFirst/Last(key) would crash if the Set was
empty.
- Fixed: LinkedHashSet.clearAndTrim() was checking the wrong value for
determining the full reset or clearing of a Map.
- Fixed: HashSet.trim/clearToTrim() was using the wrong value to
determin if something should be done.
-Fixed: AbstractList/ImmutableList/ArraySet/ArrayMap skip/back implementation was causing crashes and didn't update the last returned value.
-Fixed: ArraySet/ArrayMap previous was not subtracting before returning value.
-Fixed: BidirectionalIterator back was calling the object variant instead of the TypeSpecific Variant.
-Fixed: TreeSets/Maps Iterator now fully supports backwards Iterating.
-Added: Specialized skip/back function to improve speed in ImmutableHashSet/LinkedHashSet/CustomLinkedHashSet
-Fixed: Maps.EmptyMap getOrDefault wasn't returning the default value.
-Fixed: Maps.UnmodifiableMap wasn't respecting the change of the DefaultReturnValue. (it has its own version of it so it doesn't break the rule of it, it wasn't returning the expected value.
-Added: A bunch of Methods for Unmodifiable/Empty maps that should fail right out but would still make it past a certain point. (Compute/Merge/Supply/etc)
-Fixed: CollectionWrapper.equals wasn't accounting for self.
-Fixed: MapWrapper.get didn't account for that it was a wrapper.
-Fixed: UnmodifiableMapWrapper was linking to synchronized maps due to a unknown reason.
-Added: UnmodifiableMapWrapper now has a lot more functions it right out says unsupported instead of indirect ways.
-Fixed: AbstractCollection.Iterator was changing the pointer before validating if the element was added. (ImmutableLists)
-Fixed: EmptyCollections containsAll should return true if it is compared against a emptyCollection.
-Fixed: HashCode/Equals for the Lists/Sets implementations.
-Changed: Iterators.wrap(Array) allows now dynamic arrays.
-Fixed: ArrayList/LinkedList extractElements crashing when 0 or less elements are desired.
-Fixed: TreeMap pollFirst/LastKey should return the defaultMin/max value instead of a Empty value.
-Fixed: TreeMap keySet implementation was missing the class type implementations to pass keySet tests.
-Fixed: TreeMap.SubMap Iterator (primitive Keys) was crashing because a Null was set on to a primitive.
-Fixed: AbstractCollection.retainAll didn't push removed values through the consumer.
-Fixed: AbstractCollection.toArray wouldn't reset the last entry if the input array was larger then the elements in the collection.
-Fixed: SubList didn't check for ranges properly or didn't use parent list to validate changes.
-Fixed: ArrayList.addElements didn't check input array fully and used the wrong variable to move the elements around.
-Fixed: LinkedList.addElements(EmptyInput) would crash.
-Fixed: LinkedList.swapRemove didn't account for if the removed element was the prelast one.
-Fixed: LinkedList.removeElements would break the implementation if the list was almost empty and the middle element was removed.
-Fixed: LinkedHashSet.addAndMoveToFirst wouldn't move elements to the first place.
-Added: Array only sorting function return the inputed array. This was done to allow for static final references to use the method in one go without having to make lambda wrappers. Cleaner code.
-Fixed: Crash with FIFOQueues peek function when loops were applied.
-Fixed: FIFOQueues clean function was doing unessesary extra work.
-Added: Stream Overrides functions now support sorted.
-Updated: Changelog.
-Added: A couple more badges because why not.
This though isn't finishing the rest because the interface still needs
to be implemented.
-Added: Rest of the ConcurrentMap implementation
-Fixed: LinkedOpenHashMap Count method was returning prematurely.
-Added: ArrayList.of(CLASS, size) function that allows to preallocate
the size of the generic list.
-Updated: ListTests suppressors only suprres what they need to.
-Added: Start of the ConcurrentHashMap implementation (based on Guavas
implementation)
Note that the ConcurrentHashMap implementation is just started not
finished and still needs a lot of work, but the base code - any
subclasses is technically finished.
The implementation is also using linkedMaps to have faster iteration
performance.
Speiger