Package org.elasticsearch.cluster.coordination

Interface LinearizabilityChecker.SequentialSpec

All Known Subinterfaces:
LinearizabilityChecker.KeyedSpec
Enclosing class:
LinearizabilityChecker
public static interface LinearizabilityChecker.SequentialSpec
Sequential specification of a datatype. Used as input for the linearizability checker. All parameter and return values should be immutable and have proper equals / hashCode implementations

  • Method Summary

    Modifier and Type Method Description
    java.lang.Object initialState()
    Returns the initial state of the datatype
    java.util.Optional<java.lang.Object> nextState​(java.lang.Object currentState, java.lang.Object input, java.lang.Object output)
    Next-state function, checking whether transitioning the datatype in the given state under the provided input and output is valid.
    default java.util.Collection<java.util.List<LinearizabilityChecker.Event>> partition​(java.util.List<LinearizabilityChecker.Event> events)
    For compositional checking, the history can be partitioned into sub-histories

  • Method Details

    • initialState

      java.lang.Object initialState()
      Returns the initial state of the datatype

    • nextState

      java.util.Optional<java.lang.Object> nextState​(java.lang.Object currentState, java.lang.Object input, java.lang.Object output)
      Next-state function, checking whether transitioning the datatype in the given state under the provided input and output is valid.
      Parameters:
      currentState - the current state of the datatype
      input - the input, associated with the given invocation event
      output - the output, associated with the corresponding response event
      Returns:
      the next state, if the given current state, input and output are a valid transition, or Optional.empty() otherwise

    • partition

      default java.util.Collection<java.util.List<LinearizabilityChecker.Event>> partition​(java.util.List<LinearizabilityChecker.Event> events)
      For compositional checking, the history can be partitioned into sub-histories
      Parameters:
      events - the history of events to partition
      Returns:
      the partitioned history