Class StringTrieBuilder.LinearMatchNode

java.lang.Object
com.ibm.icu.util.StringTrieBuilder.Node
com.ibm.icu.util.StringTrieBuilder.ValueNode
com.ibm.icu.util.StringTrieBuilder.LinearMatchNode
Enclosing class:
StringTrieBuilder
private static final class StringTrieBuilder.LinearMatchNode extends StringTrieBuilder.ValueNode

  • Field Details

    • strings

      private CharSequence strings

    • stringOffset

      private int stringOffset

    • length

      private int length

    • next

      private StringTrieBuilder.Node next

    • hash

      private int hash

  • Constructor Details

  • Method Details

    • hashCode

      public int hashCode()
      Overrides:
      hashCode in class StringTrieBuilder.ValueNode

    • equals

      public boolean equals(Object other)
      Overrides:
      equals in class StringTrieBuilder.ValueNode

    • add

      public StringTrieBuilder.Node add(StringTrieBuilder builder, CharSequence s, int start, int sValue)
      Description copied from class: StringTrieBuilder.Node
      Recursive method for adding a new (string, value) pair. Matches the remaining part of s from start, and adds a new node where there is a mismatch.
      Overrides:
      add in class StringTrieBuilder.ValueNode
      Returns:
      this or a replacement Node

    • register

      public StringTrieBuilder.Node register(StringTrieBuilder builder)
      Description copied from class: StringTrieBuilder.Node
      Recursive method for registering unique nodes, after all (string, value) pairs have been added. Final-value nodes are pre-registered while add()ing (string, value) pairs. Other nodes created while add()ing registerNode() themselves later and might replace themselves with new types of nodes for write()ing.
      Overrides:
      register in class StringTrieBuilder.Node
      Returns:
      The registered version of this node which implements write().

    • markRightEdgesFirst

      public int markRightEdgesFirst(int edgeNumber)
      Description copied from class: StringTrieBuilder.Node
      Traverses the Node graph and numbers branch edges, with rightmost edges first. This is to avoid writing a duplicate node twice. Branch nodes in this trie data structure are not symmetric. Most branch edges "jump" to other nodes but the rightmost branch edges just continue without a jump. Therefore, write() must write the rightmost branch edge last (trie units are written backwards), and must write it at that point even if it is a duplicate of a node previously written elsewhere. This function visits and marks right branch edges first. Edges are numbered with increasingly negative values because we share the offset field which gets positive values when nodes are written. A branch edge also remembers the first number for any of its edges. When a further-left branch edge has a number in the range of the rightmost edge's numbers, then it will be written as part of the required right edge and we can avoid writing it first. After root.markRightEdgesFirst(-1) the offsets of all nodes are negative edge numbers.
      Overrides:
      markRightEdgesFirst in class StringTrieBuilder.Node
      Parameters:
      edgeNumber - The first edge number for this node and its sub-nodes.
      Returns:
      An edge number that is at least the maximum-negative of the input edge number and the numbers of this node and all of its sub-nodes.

    • write

      public void write(StringTrieBuilder builder)
      Overrides:
      write in class StringTrieBuilder.ValueNode

    • setHashCode

      private void setHashCode()