一個可延伸且易於維護的XML文件資料庫系統

Abstract

隨著XML逐漸成為網際網路上資料交換的標準，XML文件的管理也成為一項重要的課題。欲管理大量的資料，使用資料庫來處理應該是無庸置疑的。目前使用關連式資料庫來處理XML文件的研究中，都存在著資料庫裡所儲存的XML文件不易更新的問題，然而實際上，一份XML文件的內容往往是經常在變動的。 在這篇論文中，我們提出一個新的方法，能夠很容易地更新資料庫中的XML文件。這個方法首先定義了一個與XML文件結構無關的關連式綱目(relational schema)，利用這個綱目跟一個維護演算法(maintenance algorithm)來保持XML文件與資料庫內容的一致性。此外，由於XML的各項相關技術大多尚未標準化，未來隨實有可能會出現新的XML技術，因此，為了保持我們XML資料庫系統的彈性，我們也定義了一套延伸系統功能的架構，可以在不修改原始系統的原則下，利用各種plug-in讓我們的XML資料庫系統增加新的功能。根據我們實驗的結果，我們的方法在查詢XML文件有不錯的表現，並且在維護經常有小幅度變更的XML文件有特別好的效果。

XML is becoming a standard for data exchange over the Internet. XML’s proliferation raises the question of how XML documents are managed. It is natural to use a DBMS due to the vast quantity of XML documents. There have been several approaches to handling XML documents with a relational database. However, they share the common question that it is difficult to update the XML documents. Unfortunately, an XML document is changing all the time. In this thesis, we propose a new approach to easily updating XML documents in a relational database. Firstly, the approach defines a relational schema which is independent of the structure of the XML documents. The consistency between the XML documents and the database contents is preserved with this schema and a maintenance algorithm. Furthermore, most of the XML related technologies have not been standardized yet. It is quite possible that new technologies will emerge in the future. Therefore, our database system proposes an extensible architecture for keeping the flexibility of the system. It uses plug-ins to extend its functionalities without modifying the existing system. According to our experiment, our approach has a good performance in the query retrieval. In particular, it has an obvious improvement in maintaining frequent, small updates of XML documents.