Package org.apache.uima.internal.util.rb_trees

Class Int2IntRBT

java.lang.Object
org.apache.uima.internal.util.rb_trees.IntArrayRBTcommon
org.apache.uima.internal.util.rb_trees.Int2IntRBT
public class Int2IntRBT extends IntArrayRBTcommon
A map<int, int> uses IntArrayRBTcommon Int to Int Map, based on IntArrayRBT, used in no-duplicates mode Implements Map - like interface: keys and values are ints Entry set not (yet) impl no keySet() no values()

  • Field Details

    • values

      protected int[] values

  • Constructor Details

    • Int2IntRBT

      public Int2IntRBT()
      Constructor for IntArrayRBT.

    • Int2IntRBT

      public Int2IntRBT(int initialSize)

  • Method Details

    • setupArrays

      protected void setupArrays()
      Overrides:
      setupArrays in class IntArrayRBTcommon

    • copy

      public Int2IntRBT copy()

    • clear

      public void clear()

    • flush

      public void flush()
      Overrides:
      flush in class IntArrayRBTcommon

    • ensureCapacityKlrp

      protected void ensureCapacityKlrp(int requiredSize)
      Description copied from class: IntArrayRBTcommon
      There are two strategies for storing data, controlled by useklrp. If useklrp, then 4 elements are put together into one int vector, taking 4 words per element. Other elements are kept in their own vectors. The growth strategy for the 4-element one is to a) start at some minimum (a power of 2) b) grow by doubling up to 2 * 1024 *1024 c) grow by adding 2 *1024 * 1024, until d) reaching the maximum size (the max index will be 1 less) e) when that size is reached, the next int[] is set up with the minimum, and it grows as above. The test for growth and growing is made individually for the different parts. For color (a boolean), the size for stopping doubling is 32 * 2 * 1024 * 1024, so the # of words is the same.
      Overrides:
      ensureCapacityKlrp in class IntArrayRBTcommon
      Parameters:
      requiredSize - -

    • get

      public int get(int k)
      Get the value for a given key
      Parameters:
      k - -
      Returns:
      the value

    • getMostlyClose

      public int getMostlyClose(int k)

    • put

      public int put(int k, int v)
      adds a k, v pair. if k already present, replaces v. returns previous value, or 0 if no prev value
      Parameters:
      k - -
      v - -
      Returns:
      previous value or 0 if key not previously present

    • findKeyFast

      protected int findKeyFast(int k)
      Fast version of findKey Keeps the last node referenced *** NOT THREAD SAFE *** Tries to shorten the search path, conditionally
      Parameters:
      k - -
      Returns:
      -

    • keyIterator

      public IntListIterator keyIterator()

    • keyIterator

      public IntListIterator keyIterator(int aKey)

    • keyValueIterator

      public IntKeyValueIterator keyValueIterator()

    • keyValueIterator

      public IntKeyValueIterator keyValueIterator(int aKey)