Question: are annotations in Java the same as decorators in Python?

Write it at the front

I just arrived at the company in the morning , During breakfast , Attracted by such a problem in the Group , As shown in the figure below ：

Then I thought again , I feel that what I said is not all right , So I want to make a corroboration from the perspective of procedure .

Of course , Before writing this article , I also consulted many articles , Some views on this issue , The summary is as follows ：

• Java Annotations are also called metadata , A code level description .Python Decorator is a kind of grammar sugar .
• Annotations are for others , Functionality is not just determined by annotations ; The decorator directly intercepts , Directly change the behavior of the decorated object ！
• annotation （Annotation）： Additional metadata support is provided only , It doesn't do anything . Need something else Scanner Perform operations based on metadata .
• Decorator （Decorator）： Only definition hijacking is provided , The ability to define classes and their methods does not provide any additional metadata .

Seriously, these conceptual things , I really see it in the clouds , I suggest reading Wikipedia or textbooks .

My personal point of view , It must be that annotations and decorators are not the same .

Don't talk much , Or go straight to the code , Let's use actual cases ！

One 、 In terms of expression

@zhujie(' Parameters ')

1、 The same thing ：

All are @ start , annotation 、 Decorators can be customized 、 Can take parameters 、 Can be executed before the code block is marked .

2、 Difference ：

• java Annotations can be written in classes 、 Method 、 Variable header ;
• python Decorators can be written in classes 、 On the way .

Two 、 Case contrast

1、Java annotation

The sample code is as follows ：

 @Override
public String toString() {
return "PlaylistInfo{" +
"curtime='" + curtime + '\'' +
", issmarter='" + issmarter + '\'' +
", xmusicnum='" + xmusicnum + '\'' +
", picurl=" + picurl +
", playlist=" + playlist +
", systemtime=" + systemtime +
'}';
}

@Override： The meaning of rewriting , If you are not satisfied with the parent class, you can implement it yourself , Generally, methods are checked during compilation .

Obviously , The annotation is placed above the method , Only responsible for compiling 、 Check , The content of the method and the function of the method have not been changed .

2、python Decorator

The example code is as follows ：

class TestClass():
@property
def myValue(self):
return self.value
if __name__=="__main__":
TestClass.myValue = ' Decorator ！'
print (TestClass.myValue)

@property: As an attribute

It is obvious that , Decorators directly change the function .

3、 Conclusion

From the above we can see that , The difference between annotations and decorators ：

• 1、 The comments are just a check 、 Calibration , The marked code will not be modified .
• 2、 Decorator can be marked in method , And change the modified code function .

Come here , Do you think , They are not the same thing at all , Because it's not the same at all .

Actually , stay java Annotation and reflection in can be implemented python The effect of the interior decorator .

Have you been cheated again ？ Don't worry. , Let's look back ！

Two 、 Annotation implementation

The benefits of annotation ： Without changing the source code , Implement new functions for source code . If there is a large area of code that needs to change the same function , You can use annotations on methods or classes to implement

1、 Annotation scenarios implemented

Use them separately python And Java The way , Realize the verification of program calculation , Write the abnormal results to error.log In file

2、Python How to achieve

The example code is as follows ：

# In this case, it is used as the result of write error
def check(func):
def wrapper(*args, **kwargs):
try:
res = func(*args, **kwargs)
return res
except Exception as err:
with open("error.log", mode="at", encoding='utf-8') as f:
f.write("start".center(50, '*'))
f.write('\n')
f.write(str(func))
f.write('\n')
f.write(str(err.__class__))
f.write('\n')
f.write("end".center(50, '*'))
f.write('\n')
return wrapper
@check
def add(a, b):
print(a + b)
@check
def divide(a, b):
print(a / b)
add(50, 50)
divide(100, 0)

3、Java How to achieve

The sample code is as follows ：

public class Calculator {
@Check
public void add() {
System.out.println(50 + 50);
}
@Check
public void divide() {
System.out.println(100 / 0);
}
}

import java.lang.annotation.ElementType;
import java.lang.annotation.Retention;
import java.lang.annotation.RetentionPolicy;
import java.lang.annotation.Target;
@Retention(RetentionPolicy.RUNTIME)
@Target(ElementType.METHOD)
public @interface Check {
}

import java.io.BufferedWriter;
import java.io.FileWriter;
import java.io.IOException;
import java.lang.reflect.Method;
public class CheckDemo {
public static void main(String[] args) throws IOException {
Calculator test = new Calculator();
Class<? extends Calculator> c = test.getClass();
Method[] methods = c.getMethods();
BufferedWriter bw = new BufferedWriter(new FileWriter("input.txt"));
for (Method m :
methods) {
if (m.isAnnotationPresent(Check.class)) {
try {
m.invoke(test);
} catch (Exception e) {
bw.newLine();
bw.write(e.getCause().getMessage()+"----");
bw.newLine();
bw.write(String.valueOf(e.getCause().getClass()));
}
}
}
bw.flush();
bw.close();
}
}

Run their compilers separately , The results are shown in the following figure ：

4、 Conclusion

It can be seen from the above that ,Java Annotation and reflection in can be implemented python The effect of the interior decorator .

3、 ... and 、 An example of interface development

1、 Demand scenarios

Enter the user password , Return to the user information interface

2、Python How to achieve

The sample code is as follows ：

from flask import Flask
import json
from flask import request
app = Flask(__name__) # start-up
@app.route('/rongrong/getUserInfo',methods=['GET']) # Request path 、 Request mode
def login():
username = request.args.get("username") # obtain url In the parameter “username” Value
password = request.args.get("password") # obtain url In the parameter “password” Value
if username and password: # If the value passed in is true, print the following information
data = json.dumps({
"username":username,
"password":password,
"code":"200",
"message":" success "
},ensure_ascii=False) # Solve the problem of Chinese garbled code
else: # If the transmission parameter is empty, print the following information
data = json.dumps({
"message":" Please pass the parameter "
},ensure_ascii=False)
return data
if __name__ == "__main__":
app.run() # function

3、Java How to achieve

The sample code is as follows ：

 @RequestMapping(value = "/rongrong/getUserInfo", method = RequestMethod.GET)
public Result<List<Student>> findStudentByName(HttpServletRequest request) {
String username = request.getParameter("username");
String password = request.getParameter("password");
JsonRootBean jsonRootBean = new JsonRootBean();
jsonRootBean.setUsername(username);
jsonRootBean.setPassword(password);
jsonRootBean.setCode("200");
jsonRootBean.setMessage(" success ");
Result result = ResultUtils.success(jsonRootBean);
return result;
}

4、 Start respective services , The operation results are as follows ：

python result ：

Java result :

5、 Conclusion

Python The decorator is very simple , It is to modify the behavior of a function through a layer of shell , and @decorator_func It's just a grammar sugar , The writing method used to beautify ornaments .
Java The comments in are different , It is from the language level for the classes in the code , function , Add some metadata that can be read by runtime to the field , The annotation provider should read these metadata at run time , And deal with it accordingly .

Four 、 At the end

The author's lack of talent and learning , Writing this article is entirely out of technical itch , Naturally, I have consulted a lot of articles , Only then has this article .

The following is a personal opinion only ：

• Look like , But there are two species , But it can make them behave similarly ;
• Python The ornament of is just like his name , Very straightforward , Is to realize the decorator mode （ A grammar sugar of ）[email protected] Part corresponds to a function that returns as a function , You can input the objective function 、 Output filter , And other interventions 、 packing ;
• Java There are several annotations , According to the action period, there are compilation periods 、 Operation period, etc , Just mark the class or method , In the corresponding period, your program can read them through reflection ;
• Java On the surface, it seems useless , But less is more , A little packaging can be achieved with Python Decorators have the same function , The premise is how to call the target class and method , As long as the annotation is explained in the calling wrapper , Just OK 了 ;

Through various means, they can become one thing , So in terms of results , you 're right , It can be regarded as one thing .

let me put it another way , Sometimes I feel that the decorator is more like Java A design pattern of .

Do you think? ？ Welcome to comment area for discussion ！