A heterogeneous multi-platform code generation method based on IMCL model

Abstract

The invention relates to a heterogeneous multi-platform code generation method based on IMCL model. The method comprises the following steps: adopting IMCL language to characterize the function of a target system and obtaining the function model of the target system; adopting a formal method to decompose the functional model to obtain one or more platform models corresponding to one or more platforms in the target system; mapping the one or more platform models to corresponding platforms, thereby supplementing the platform models in detail; based on the abstract syntax tree, according to the characteristics of the one or more platforms, designing corresponding transformation rules, so as to transform the one or more platform models into corresponding platform codes. The invention solves the problem of the prior art that it is difficult to generate multi-platform codes at the same time, the function model is decomposed to be used for automatic generation of the target platform codes, and the invariance of the constraint relationship between the function and the function is ensured through the program slicing method and the decomposition method based on the formal system dependency diagram.

Description

technical field [0001] The invention relates to the field of automatic generation of computer codes, and particularly relates to a heterogeneous multi-platform code generation method based on the IMCL model, which is used to realize the automatic generation of codes. Background technique [0002] Model-driven architecture was proposed in 2001. It is an approach to software design that uses a platform-independent model to define system functionality through an appropriate domain-specific language (DSL). The advantage is that by abstracting the DSL model, the system architecture can be expressed with less DSL code than other programming languages ​​(Java, C++, C#). In addition, programs written in a domain-specific language are easier to understand, and even the DSL code can be expressed in a graphical language, and the difference between business logic and technical implementation can be eliminated. Since model-driven architecture is platform-independent, many interpreters ...

AI Technical Summary

Problems solved by technology

No matter which model conversion method is adopted, it is unavoidable to avoid the problems brought about in the process of model conversion, including how to realize the conversion between different expression models, how to ensure the consistency of the two models, etc.

In addition, in the later stage of model-driven development, in order to ensure that the implementation meets the requirements, the programmer needs to combine the model and specific deployment to code the physical equipment separately. After many simulations and tests, the code is repeatedly adjusted. This process consumes a lot of time.

Images