Suggestions for improving code quality

Programmers are like craftsmen from all walks of life , All rely on their own works . So if you want to get high income and recognition in the industry, you must constantly learn to improve your business level . Here I have summarized some suggestions for programmers to improve code quality, hoping to be helpful to your advanced level .

1. Improve code reuse

Whether it's writing new code or refactoring old code , We must pay attention to improving the reusability of code . Many Xiaobai only care about the implementation of business functions when writing code , Often the code is written as a lump . The next time there is a similar requirement, it has to be rewritten , The previous code is not used at all , This leads to a lot of repetitive work . So when we write code, we must pay attention to the reusability of code , Divide the code effectively , Achieve high cohesion and low coupling . Independent modules thus decoupled , It can be reused directly in other similar businesses . Reusable code generally has the following two characteristics :

1. The code is decoupled independently , Independent of other modules

2. The code is stateless , Repeated calls have no side effects

2. Remove unnecessary dependencies

Many novices are writing code , habit Copy Code , Sometimes, in order to realize a small function, you need to go elsewhere Copy A lot of dependent header files come into the business , Whether these header files can be used or not . Over time, the code becomes smelly and long , Very difficult to maintain . When we were writing code , There must be a sense of simplification , The header file is used and then imported , Remove what you can't use . At the same time, the third-party library that cannot be used must be removed in time , This can not only improve the quality of the code, but also improve the efficiency of the code .

3. Multithreaded applications consider thread safety

In multithreaded applications, if the data is read-only , Generally, there is no problem , However, if multithreading is involved in writing data at the same time, thread safety should be considered , Prevent data from being written repeatedly during multithreading . Be sure to lock when synchronizing data with multiple threads , Prevent conflicts between threads , At the same time, we should also consider the performance problems caused by locking in multithreading process . At the same time, the synchronization problem between threads should be avoided as far as possible . When introducing multithreading, you should also consider whether it will cause thread deadlock .

4. Adopt the scheme of delayed allocation

Many people write code for convenience , Declare the instance of the dependent class directly in the member of the class , The advantage of this is that you don't need to control the life cycle of member variables , The disadvantage is that the performance becomes worse when the class is instantiated . So if a member variable of a class is complex , Initialization takes time , I suggest adopting the scheme of delayed allocation . Deferred allocation is to declare the pointer of the corresponding dependency in the member variable of the class , Then instantiate when calling . Destroy the modified object when destructing or destroying . And when the class is called , We must avoid frequent creation and destruction of objects, which is a serious loss of performance .

5. Reduce memory allocation 、 Copy and release frequency

If the function input parameter passes a complex or memory consuming object , It is recommended to pass the reference of the class , This can avoid the creation and destruction of temporary variables , So as to improve the efficiency of function execution . When writing code, we must try to reduce the frequency of memory allocation, copy and release , These are time-consuming operations , If you can use the previous cache , Use the previous cache as much as possible .

6. class 、 Method 、 Variable usage private Modifier

Improve class encapsulation , Avoid exposing redundant interfaces and member variables , Use... As much as possible private jurisdiction , Can improve code security .

7. Is there any sensitive information in the abnormal information

Most of the time, for the convenience of debugging , Some sensitive content will be output to the log in the exception information , Including the running logic of the program and some user's personal sensitive information . Little imagine , This information may be obtained by some people with ulterior motives and used to destroy our system . So when releasing the program again , Be sure to check whether there is sensitive information in the abnormal information .

8. Control the permission domain of the code

Some code is privileged when writing products , For example, some codes can operate the server-side database , Delete the database . Some codes have administrator privileges , It can affect the user's operating system . When writing this kind of code , We must be very careful , Try to limit the privileged code to a limited range , Control it . At the same time, the authorization of other function call modules must adhere to the minimum principle , Prevent abuse of authority .

9. Pay attention to the handling of sensitive information

In product development , We must protect sensitive information , Prevent information leakage .

What information is sensitive ？

1. Personal sensitive information

Identity information 、 Privacy information 、 Business information, etc

2. Business sensitive information

Employee information 、 Business data 、 Process information, etc

Who has access to the corresponding sensitive information ？

The principle of minimum Authorization , The operation modules of sensitive information should be reduced as much as possible , Reduce the authorization of sensitive information .

Is there any sensitive information in the application log ？

Do not output sensitive information in the log , There are often problems in this aspect .

Does object serialization exclude sensitive information ？

Previously, there was a problem of leakage during the serialization and sending of sensitive information , So try to avoid sending user sensitive information through serialization .

Is there any special treatment for the storage of sensitive information ？

For example, the user's password cannot be stored in clear text , Try to store the hash value of the password with salt .

Sensitive information must be cleared in time when it is not used .

10. Check the safety and effectiveness of data

When there is no verification point for external data , We must verify the data on the transmission path , Including security verification and validity verification , Prevent unauthenticated data from invading the system and causing damage to the system .