The Design And Practice Of HTTP Information Filter Based On Server

Nowadays, Internet has become an indispensable part in our daily life as well as in socialdevelopment. However, characters, like openness, equality, unboundedness, have led to anunlimited abuse of the network, which is, as a result, flooded with lots of bad and sensitiveinformation. Webpage is an important medium in network information, which works in avariety kinds of servers. HTTP information filtering to server port, which can remove bad andsensitive information as well as reducing the negative effects the network has brought, hasbecome a hotspot in network field and is of great significance.Based on the background talked above, we choose HTTP as our study subject. On thebasis of general information filtering principle and common HTTP filtering technology, wedesign and present a server-port-based HTTP information filtering system. The main worksand accomplishments are as follows:First of all, we give a brief introduction of HTTP, analyze the feasibility of HTTPinformation filtering, including methods like process monitoring and manual control of TCPconnection, and achieve the goal of network monitoring by capturing data packages with thehelp of winpcap data package. Then, we come up with methods which improve monitorperformance.Secondly, in order to raise efficiency of matching and enhance real-time capacity. Wegive a detailed URL filtering method----process the URL with the hash method, reducing thememory space and making URL matching much faster. For the purpose of dealing with hashcollision and optimizing hash method, we adopt data structure “LinkList” to process hashcollision and speed up URL node searching with bit-mapping, making URL matching moreefficient on the whole.Thirdly, in order to match the key word information in the HTTP package faster, wemake some optimization based on WM multi-mode matching method and come up with ahigh-speed method which can match key word in Chinese, making key-word matching moreefficient on the whole.At last, combining methods talked above, we design and present a server–port-basedHTTP information filtering system(Prototype) which is then validated and tested in the operation. The validation of this method shows that compared with the original one, ourmethod sees a great improvement in efficiency. The system testing result tells that this systemprocesses relatively good accuracy, feasibility, stability and adaptability and works very wellconsidering the speed of filtering and accuracy.