You can find a standard program on the Internet (generally available). If not, write a violent code that can ensure correctness, and then write a program that randomly generates data. Use a bat file to constantly create small data for your code and calibration (violence) to run and proofread the answers. (The above methods are commonly known as injectivity)

Bat file is as follows

: 1

Production _ data

a

a_

fc a.out a_。 be away from home

If errorlevel== 1 pause

Go to 1

A is the name of your program, and a_ is the scale or violence. If fc doesn't work, you can find it on disk C and put it next to the program.

Because the data is generated randomly, if the code has obvious loopholes, it is easy to be photographed (especially some details). Of course, some code performs very well in the case of random data, but it will get stuck in the constructed data. You can try to build extreme data for testing.

Generally speaking, the correct code is not necessarily correct. When you encounter a mistake, you'd better sort out your own ideas first, then follow your own code, and then use the right one.