Binary page implementation of a canonical native storage for XML

Abstract

<p>XML is a simple and very flexible text format, originally designed to meet the challenges of large-scale electronic publishing. Great as XML is for representing data, many XML-based query processors and storage managements have been proposed. With the classical memory problem of DOM parsers when an XML document is mapped onto an internal tree structure, many implementations handle a rather small document size. CanStoreX with textual page implementation approaches the problem by breaking an XML document into smaller pieces, stored into pages. It preserves the structure of the original XML document as well as does not require the whole document to be loaded into the main memory at once. Its binary page implementation removes major memory problems. This allows CanStoreX to parse XML documents of size 100 gigabytes or larger without any conspicuous problems. This shows that CanStoreX is scalable in terms of storage requirement, memory management, and query processing. The only two bottlenecks, encoding and decoding processes, can be diminished by embedding them into a computer chip, which will further bring CanStoreX to its primal state.</p>