chore: convert indexers to internal iteration

[triceo]

This will allow complex filtering that will be necessary to make containingAnyOf() work.

[Copilot]

Pull request overview

This pull request converts the indexer's external iteration pattern (using UniqueRandomSequence.pick() and remove()) to an internal iteration pattern using the standard Java Iterator interface. This refactoring prepares the codebase for implementing complex filtering required by the containingAnyOf() functionality.

Changes:

• Replaced UniqueRandomSequence interface and its implementations with UniqueRandomIterator implementing the standard Iterator interface
• Updated all indexers to provide randomIterator() methods instead of requiring external access patterns
• Moved iterator implementations from neighborhood.stream.enumerating.common to bavet.common.index package
• Simplified the API by removing ElementAccessor abstraction layers (ListBasedElementAccessor and IndexerBasedElementAccessor)

Reviewed changes

Copilot reviewed 24 out of 24 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
DefaultUniqueRandomSequenceTest.java	Deleted - replaced with reduced test coverage in DefaultUniqueRandomIteratorTest.java
SelectionProbabilityTest.java	Moved to bavet.common.index package and updated to use new iterator API
FilteredUniqueRandomIteratorTest.java	Moved and renamed from FilteredUniqueRandomSequenceTest.java, updated for iterator pattern
DefaultUniqueRandomIteratorTest.java	New test file with basic iterator tests, significantly reduced coverage compared to original
UniLeftDatasetInstance.java	Updated parameter name from rightSequenceStoreIndex to rightIteratorStoreIndex
UniLeftDataset.java	Updated parameter name to match instance class
LeftTerminalUniEnumeratingStream.java	Refactored to inline store index reservations
ListBasedElementAccessor.java	Deleted - no longer needed with direct iterator access
IndexerBasedElementAccessor.java	Deleted - no longer needed with direct iterator access
FilteredUniqueRandomSequence.java	Deleted - replaced by FilteredUniqueRandomIterator
AbstractRightDatasetInstance.java	Changed from building sequences to returning iterators directly
AbstractLeftDatasetInstance.java	Simplified by removing UnwrappingIterator and using direct iterator access
BiRandomMoveIterator.java	Updated to use iterator pattern instead of sequence pattern
UniqueRandomIterator.java	New interface replacing UniqueRandomSequence, implementing standard Iterator
RandomAccessIndexerBackend.java	Added randomIterator methods returning new iterator types
LinkedListIndexerBackend.java	Added randomIterator methods that throw UnsupportedOperationException
Indexer.java	Added randomIterator method signatures to interface
FilteredUniqueRandomIterator.java	New implementation of filtered random iteration using Iterator pattern
EqualIndexer.java	Implemented randomIterator methods delegating to downstream indexers
DefaultUniqueRandomIterator.java	New implementation replacing DefaultUniqueRandomSequence with Iterator pattern
ContainingIndexer.java	Implemented randomIterator methods delegating to downstream indexers
ContainingAnyOfIndexer.java	Added randomIterator methods throwing UnsupportedOperationException
ContainedInIndexer.java	Implemented randomIterator methods with iterator function abstraction
ComparisonIndexer.java	Implemented randomIterator methods with support for single and multiple indexers

Comments suppressed due to low confidence (1)

core/src/main/java/ai/timefold/solver/core/impl/bavet/common/index/UniqueRandomIterator.java:55

• The javadoc for the remove() method is misleading. It states "Removes a random element in the underlying list which has not already been removed." However, this is standard Iterator.remove() behavior - it removes the element that was most recently returned by next(), not a random element. The documentation should clarify that it removes the last element returned by next(), consistent with the standard Iterator contract.

[sonarqubecloud]

Quality Gate failed

Failed conditions
63.9% Coverage on New Code (required ≥ 70%)

See analysis details on SonarQube Cloud