Discovering code structure problems through UT

Through ut for a period of time, it is found that UT is not only used to guaranteeCodeIn addition to the correct logic, it can also be used to discover the design defects of the Code. Of course, this is not to discover design defects by running ut, but to write ut. Instead of using a test-driven development model, we write some code to do ut. In this way, if you find that ut for a piece of code is difficult to write, it usually means that the design of this piece of code has the following: the interface is not clear enough, the class responsibility is not clearly divided, and the source file has unnecessary compilation dependencies .... To put it bluntly, it is a "testability" problem. This sensitivity should be maintained when writing ut, and such design problems should be found and corrected in a timely manner.