摘要： 提出了一种基于链式结构的XML文档生成方法，设计了一个利用Java中的stream tokenizer类实现HTML文档解析的算法，将解析得到的元素内容及文本内容生成的结点插入到相应的位置上，同步生成DOM解析树，对DOM解析树进行遍历，将遍历得到的信息以二叉链表的形式存储，采用改进的先根遍历算法对该二叉链表遍历，提取相应的信息构建DTD，完成整个转换生成的过程。

关键词: HTML, XML, DOM, 解析

Abstract: This paper puts forward the method of XML document based on linked-structure. It uses stream tokenizer to design an algorithm of HTML document parse. The element and text contents are inserted into the correct position to create DOM-parsing-tree as parsing. Lastly. After visiting the tree and storing the information into a binary-linked-list, it uses a modified preorder algorithm visiting the linked-list, then extracts corresponding information to build DTD and finish the whole generation process.

Key words: HTML, XML, Document object model(DOM), Parse