The Design And Implementation Of Manageable Search Engine In Office Automation System

Through deep study of search engine's core technology, such as index building, parsing Chinese words, index searching. With the explosion of information, common vague match searing can not meet the user requirement in office automation system, also there are many logic coding for search in specific modules which lead to unmanageable, unable to extend. Moreover, due to the interface of searching database and searching index didn't integrate; developer may be difficult in use different interface. The object of the searching package is to build a manageable searching package which is irrelevant to specific module and data field in database, also it can be manageable, support parsing Chinese words and full text searching. It also integrates the interface of database searching and full text searching to make it easier to switch between database searching and full text searching.This paper designs and implements office automation system deployable search engine package, this package constructs on Compass and Lucene information retrieval tools, it also integrates OA system database searching, index searching and query interface, abstract OA system searching to an manageable and deployable package, it decuple system business module, reduce repeat coding and improve the structure of system, it also implements parsing Chinese words in full text searching system, optimizing the dictionary structure of parsing words.This package through the support of Spring and Hibernate to manage the instance of database searching and index searching, through the mapping Java Object of Hibernate, implements maintaining searching field and indexing table by the configure, it also maintain parsing and searching interface by the configure to implements make system more flexible. Through the integrating of Compass and Hibernate, it supply transaction index of database and synchronizing index with database to improve the speed of building index and maintaining index in transaction.This package also implements decuple business logic in OA system, through setting data transport object, system delay generate database query and index query in core level to implements the cooperation of tiers.This package could improve developing efficiency and saving developing period, it also reduce the coupling in the system, and decrease the costs on managing system to attach importance to develop business logic, it also enhance plain data field vague match searching to solve the problem of insufficient function. By personal configuration, we can make the search engine satisfy different user requirement to simplify developing and enhance system function.