解析代码自动生成框架的设计与实现

Parsing original data correctly is the first important step in each project. In the actual project, the diversify data structures and the similarity of analytic methods make the traditional manual coding produces a lot of repetitive work, which reduces the development efficiency as well as causing errors, maintenance difficulties and other issues. This paper designs and develops a PAGF (Parsing code Automatically Generated Framework), which designs and combines a series of functional modules based on templates to generate code for data parsing. PAGF automatically generates executable code for data parsing by parsing a description file written in a custom language called SDFL (Source Data Format Language). PAGF has a strong scalability, the code can be combined with a variety of tools, applied to a variety of scenarios. PAGF also allows users to follow the guidance through visual encapsulation, the users can get the code for data parsing automatically by writing a small amount of data description file.Verified by a real case based on Hadoop, the generated code has the same performance with the manual code.?????