XML Value Index Implementation

• Implemented with two B+ Trees

• Tree 1: Path index

  • Each distinct reverse path to a PathID
  • Example of reverse path: id/employee/dept
  • Enables queries such as: //id (all id elements)

• Tree 2: Value index

  • Data Entry: PathID, value, RowID, NodeID
  • Good for queries: /book/author/name
  • But not for queries: //name=”Maggie”

Notes: