Cloud Warehousing

Abstract

The exact validation of streaming XML documents can be realised by using visibly push-down automata (VPA) that are defined by Extended Document Type Def- initions (EDTD). It is straightforward to represent such an automaton as an Abstract State Machine (ASM). In doing so we enable computations on abstract states that are defined by a certain class of automata, in this case VPAs. In this paper we elaborate on this approach by taking also updates of XML documents into account. In this way the ASM-approach combines vertical refinements, which first make states explicit and then instantiate by a specific EDTD, with horizontal refinements, which replace streaming XML documents by stored ones and then add updates. Furthermore, as the EDTD appears as part of the abstract state, updating it is another natural extension by hor- izontal refinement. In this way we obtain consistently integrated updates and schema updates for XML documents, which can even be extended to become fault-tolerant by taking at most k errors in the document into consideration. It further provides an example of ASM-based computation with automata-defined states.