| As the fast development of Internet, more and more people are exchanging their data through Internet, but the differences among computers (different hardware, different OS, different database) become more and more outstanding. To make data exchanging among different computers easier, middleware was proposed. It has shielded many lower tedious things by a friendly mode, to help programmers concentrate on the business logic, which has improved the efficiency and code re-using rate. The work mainly includes the following conclusions:1. A kind of architecture of two layers data access is designed for our middleware. The first layer is between client application and data access middleware, which is designed for organizing data and calling data access functions of middleware. The second layer is data access middleware, which is for exchanging data between database and client application.2. In the firs layer, we shield the process of dealing returned string from middleware by web data access component. So client application code could be run anywhere. Code re-using rate is valuable.3. While developing data access middleware, elastic data accessing patterns are pointed out, according to design pattern. Then some comparisons are made among these patterns and finally we pick out the most adaptable one and we pointed out every function of the class, their inter-connection and the way we could expand the class.4. The middleware use stored procedure to execute some SQL commands,which would decrease the exchanging times between database and client programs and shorten the response time. We use XML data to pass parameters into stored procedure. All did a good work.The middleware benefits from design pattern in architecture. In the entire data accessing middleware, the shielding database connection as well as the database operation means the application procedure is independent to the data source, which guarantee data accessing middleware having a stronger auto-adapted ability.
|
PDF Full Text Request