HTML Tree Parsing Algorithm Based on Pre-extracted Data

摘要

In the paper, a new method of extracting HTML Tree from web pages is proposed. Its main idea is that the parts of web pages which are not easy to parse including tags and attributes should be handled previously, then the remaining parts are tidied and parsed, and then both the two former extracted parts are deposited in the tree. As integrated the tidying process and the parsing process, the new method does not only keep the web data integrity but also simplify the complexity of algorithms. The test shows that it can parse all kinds of web pages and provide concrete fault tolerance mechanisms.