全复杂度介绍（Introduction to full complexity）.docVIP

全复杂度介绍（Introduction to full complexity） In the software industry, almost all developers are talking about the quality of the code, and everyone on the code quality has its own set of views. Even the term code (code smell) taste has entered the popular vocabulary, a description of the code need a way to improve the. The code is developed by us according to the taste of their accumulated some work experience to judge, some people think that code can reflect the code structure and quality, but also some people and that code is used to explain a mechanism that the code is too complex. Obviously, Javadocs? Is very useful, but how many embedded comments to maintain code? If the code is written well enough, it also needs to explain yourself? From these we can see that the code smell is a mechanism of subjective evaluation, in many cases below, although some of the worst possible code to others had written the best code. In our work, whether there are many such voices, is the first look a bit messy, but its good scalability. Therefore, we need a way to objectively evaluate the code quality, some can decisively tell whether we are viewing the code there is a risk. Whether you believe it or not, this kind of thing does exist! The mechanisms for objectively evaluating code quality has been a long time, but most developers ignore them. These mechanisms are called code metrics (code metric). At present, some companies such as HUAWEI, Primeton etc. have more stringent requirements on the quality of the code, using the CMMI5 standard to evaluate the quality of the code. According to their unit test coverage as a code quality assurance means. Unit test coverage type has the following: statement coverage, branch coverage, condition coverage, path coverage. In the unit test before the three cover is very easy to reach, but there will be some defects. In this article, I will not detail calculation method of the first three coverage, focus on to talk about the problem of path coverage.