一种基于随机文件的C#与非托管C代码交互模式

The C# and Unmanaged C Code Interaction Model Based on Random File

在大型科学计算过程中往往会出现海量中间数据,需要在不同开发平台或语言之间传递,此时中间数据会消耗大量系统资源,引起系统性能急剧下降;此外,中间数据长度的不确定性和跨平台的移植性都对科学计算算法提出了较高的接口要求;为保证系统的稳定性和可移植性,文中提出一种基于随机文件读写进行数据传递的简单实用方法,来实现在C#.NET平台下的托管代码中调用C语言开发的非托管代码的交互模式。

There will often be magnanimous amount of middle data in the large-scale science computational process, which needs transfer between different development platforms or languages. Thus the middle data will consume massive system resources and will hamper the system performance rapidly. In addition, the middle data length＇s uncertainty and the cross platform＇s portability set the high connection demand of the science computation algorithm. To guarantee system＇s stability and the probability, this article proposed a simple practical means that carries on the data transfer based on the random file read-write device that realizes the non-trust code interactive pattern accessing nontrust code developed by C language under the C#.NET platform transfers trust code.