Refactoring And Optimization Of Abstract Syntax Tree Based On COStream Compiler

With the development of multi-core architecture platform,people have put forward higher requirements on data and computing capabilities.As an efficient parallel programming model,the data stream programming language-COStream can leverage the parallelism of data,task and pipeline to maximize parallelism of program.It has the characteristics of fast speed,low power consumption and easy to use.However,with the development of technology,in the iterative update process of COStream,problems such as inflexible design,difficult to expand and maintain,low applicability have emerged.Aiming at the problems faced by COStream compiler,refactoring and optimization of the abstract syntax tree based on COStream compiler is designed and implemented.According to the structural characteristics of the abstract syntax tree,the syntax tree is designed by using the idea of object-oriented programming.The refactored part includes lexical analysis,syntax analysis,interface adaptation and code generation,which greatly improves the Flexibility and scalability of the syntax tree.We improved the memory performance and programming friendliness of the COStream by providing debugging information and misalignment positioning,symbol table management as well as memory optimization management mechanism.For the problem that COStream lacks library support,grammar and library extension process specification based on COStream compiler is proposed.We introduce the basic extension method of grammar based on COStream compilation system and expand the library on COStream compiler from the two aspects of local library extension and external library extension,which can effectively improve the generality and stronger applicability of the COStream.We take the X86 architecture multi-core processor as the target platform and use several typical algorithms in the multimedia processing domain as test cases.Refactored COStream is evaluated by the standard of the evaluation method,which shows the effectiveness of the reconstruction method.