| The text in the web page is easy to be reused, taking up little space but containing a huge amount of information. It takes up the most percentage of web page information, and, most frequently, people obtain information just from the texts in the web page. At present, web study is in the developing stage with theories of all schools formed in the field of text extraction.However, for some kind of web pages with scripts, a series of problems have arisen by the opening of the web page and the implementation of scripts, like the intrusion of computer's software system by malicious codes; the annoyance of advertisements in browsing texts in a web page; the problem of copying and extraction of texts restricted by scripts. But, relatively speaking, texts in such a special web page are even more valuable than those in the ordinary ones. especially when they are used by criminals for secretiveness, corresponding technology should be used by network inspectors.In this paper, compilation and working theories of the scripts are analyzed profoundly, and the problem of filtering jammers has been solved. Besides, a program for the text extraction and consolidation has been put forward by the technology of RTTI (Run-Time Type Information) and script expanding, from the perspective of both manual and automatic extraction. The main conclusions are:Study has found that restrictions on all special pages are from its source code in the script. Such restrictions can be shielded by expanding scripts or generalizing the special pages. Jammers in a web page are all characters which intervene in the page and with the same color with that of the background. We expand scripts and make the jammers disable in order to filter them; RTTI technology has been proposed for software development, by which malicious scripts could be bypassed and texts could be automatically extracted; the similarity of multi-pages have been studied for the automatic extraction and merger of multi-page texts, by recording the extracting point of the first page and the searching the link address; a software model has been developed and tested of its ability on both ordinary and special pages in automatic extracting text and filtering jammers.Experimental results show that all methods which have been proposed in this paper can effectively resist the restrictions on the text extraction, and RTTI-based software technology is the most powerful automatic extraction method.The analysis of special pages and text extraction methods in this paper has a practical significance in the research of web information technology and the application of networks. |
