Schemaless XML document management in object-oriented databases

XML has emerged as the universal format for data exchange and is being widely used in technical documentation of manufacturing companies, book publishing, and Web publishing. It is also being used in contexts that are mission-critical, such as banking and e-commerce. One of the main reasons for the popularity of XML is its semi-structured nature and the flexibility of its component parts. Schemaless XML documents, which do not have a schema or Document Type Definition (DTD), have the highest level of flexibility, as they are not bound to any kind of structure. However, support for schemaless XML documents in the major relational and object-oriented databases is limited to storing each of them as a CLOB (Character Large Object). This kind of storage requires the usage of associative search criteria to query the document, and does not support piecewise update and multiple document queries. In this research, we developed a middleware layer over an Oracle DBMS to support the storage, retrieval, query and update of schemaless XML documents. We also propose an algorithm by which we can store the logical structure of the XML document in addition to storing its content. The final part of this paper presents an algorithm that can rebuild XML documents by using the stored logical structures and the content of the XML documents. The rebuilt XML document is a well-formed document and is similar to the original document.