Netted multilevel index matching method for publishing and subscribing

Abstract

The invention discloses a netted multilevel index matching method for publishing and subscribing, and belongs to the field of information technology processing. According to the method, an improved netted multilevel index structure is proposed for the problem that full-table scanning is needed when subscribed items of a user are deleted in an existing multilevel index matching algorithm, the subscribed items can be locked with link table header pointers of the subscribed items and then deleted under the condition that the indexing efficiency is ensured, full-table scanning of multilevel indexes in the deleting process is omitted, and thus the index deleting efficiency is improved; multiple subscribed items corresponding to subscribed item attributes are defined with a doubly-linked table, identical subscribed items in the indexes are connected, and thus the index maintaining efficiency is improved; indexes are established in levels for operate operators and attribute subscribed items respectively, querying conditions can be rapidly pruned and filtered, and thus the index querying and matching efficiency is improved.

Description

technical field [0001] The invention relates to the field of information technology processing, in particular to a publish-subscribe-oriented mesh multi-level index matching method. Background technique [0002] The message subscription / publishing mode can effectively reduce the coupling between the communication parties, completely decouple the publisher and the subscriber in time, space and synchronization, and reduce the dependence of the communication parties. It is an effective asynchronous communication mode. It has been widely used in large-scale distributed networks. In the message publishing / subscribing mode, the data is published and subscribed in the form of messages, and the subscribers indicate the type of events they are interested in and notify the proxy server; when the publisher publishes a new event, the server will use the matching algorithm to subscribe Events are sent to subscribers. Among them, the matching algorithm is responsible for efficiently fin...

AI Technical Summary

Problems solved by technology

[0005] Purpose of the invention: Aiming at the problem that the existing publish-subscribe index matching algorithm has duplication of matching process and low efficiency when deleting, the present invention proposes a publish-subscribe-oriented mesh multi-level index matching method, in which the same subscription The different attribute nodes of the entry are connected through the linked list, which significantly improves the indexing efficiency when the subscription entry is deleted in the publish-subscribe system

However, when a multi-level index deletes a subscription entry, because the multiple predicate attributes of a subscription entry are scattered in multiple index linked lists, it requires a full index scan when deleting. The algorithm complexity is high

Images