Research On The Method Of Implementing Python Data Types In PAR Platform

In order to solve the critical problem of software crisis better,improve the efficiency and reliability of software development actually,Professor Xue Jinyun proposed a general and applicable method of developing Algorithmic programs formally — PAR(Partition-And-Recur)method with the support of several national projects.The mechanism of data abstraction is used in the the method for designing programs.Various auto-generating program systems have been developed on PAR platform which is the support platform of PAR method.Apla ? Python auto-generating program system can convert Apla algorithm programs into executable Python programs.The system consists of converter and Python reusable component library.Python component library refers to the concept of software reuse,and its basic idea is to avoid duplication of effort in developing software.Through software reuse,the efficiency and quality of software development can be improved.The component library includes five basic categories of components: sequence,collection,package,tree and figure.The main goal of this paper is to design five categories of components in Python component library correctly.That is to say,using Python language to realize the five kinds of abstract data types: sequence,collection,package,tree and figure in Apla language(Abstract Programming Language that is an abstract language is used to define and design algorithm programs).First,to compare the data types of Apla language to that of python language analytically.Then,realizing the idea of generic programming through which setting a formal parameter that is called datatype with the method of Python class defines type variable.Finally,the two storage methods of components that is named sequential storage and chained storage were realized by array and chained list.Python component library with high reliability was installed in Apla ? Python auto-generating program system so that the high reliability of the generation system could be realized perfectly.Two test cases that refer to fast sort and preorder traversing binary tree were used to test and verify Python reusable component library.That is to say,the Apla programs of fast sort and preorder traversing binary tree were first of all input into the system.Then,the Python program was generated with the support of the component library,and was run.Finnally,check its actual operating results.By the test,the expected and satisfactory results had been achieved ideally.Apla ? Python auto-generating program system with the guidance of PAR method fully guarantees the correctness and reliability of the arithmetic programs.The implementation of Python reusable component library in the system has filled in the theoretical gap in the field of PAR method.It also can improve the efficiency of developing and debugging programs.With the development of software technology,the improvement of PAR method and PAR platform,and the expansion of Python component library,the functions of the generation system will be more and more abundant and robust.In the near future,the system will be comprehensively applied to all stages of software development.people will gain proficiency and get heavy use from the system eventually.