Python类内赋值局部变量后修改为何同步改变self关联对象？

Hey there! Let's clear up this confusion—you're right that Python doesn't have pointers like C/C++, but it does use object references, which is exactly what's causing this behavior. Let's break it down step by step:

1. Python's Variable Model: References, Not Values #

In Python, every variable acts as a reference to an object, not a direct container for a value. When you run this line:

var = test.test_json

You're not copying the SimpleNamespace object that test.test_json points to—you're just making var point to the exact same object in memory. So var and self.test_json are two different names tied to one shared object.

That's why when you modify var.test_id = 6, you're actually changing the test_id attribute of that single shared object. Any other variable pointing to that object (like self.test_json) will automatically see the updated value.

2. Your Code's Misstep #

You mentioned wanting var to copy just the test_id value, but your code is assigning the entire SimpleNamespace object to var, not just the integer value of test_id. If you only wanted to copy the test_id value, you'd do this instead:

var_test_id = test.test_json.test_id  # Copies the integer value (5)
var_test_id = 6  # This won't affect self.test_json.test_id at all

Since integers are immutable in Python, assigning a new value to var_test_id just makes that variable point to a new integer object (6), leaving the original 5 completely untouched.

3. Fixing the Object Copy Issue #

If you need to create a separate copy of the entire SimpleNamespace object (so modifying var doesn't impact the original), use Python's copy module to make a shallow or deep copy:

from types import SimpleNamespace
import json
import copy  # Add this import

class Test:
    def __init__(self, json_data):
        self.tests = json_data.tests
        self.test_json = ""

    def do_something(self):
        for test in self.tests:
            # Make a shallow copy of the SimpleNamespace object
            var = copy.copy(test.test_json)
            self.test_json = test.test_json  # Points to the original object
            print("The self.test_json is: " + str(self.test_json.test_id))
            print("test_json is: " + str(var.test_id))
            var.test_id = 6
            print("(should be 5, unchanged) The self.test_json is: " + str(self.test_json.test_id))
            print("(should be 6, changed) test_json is: " + str(var.test_id))

if __name__ == '__main__':
    json_data = b'{"tests": [{"test_json": {"test_id":5 }}]}'
    data_as_simple_namespace = json.loads(json_data, object_hook=lambda d: SimpleNamespace(**d))
    t = Test(data_as_simple_namespace)
    t.do_something()

Running this modified code will give you the output you expected:

The self.test_json is: 5
test_json is: 5
(should be 5, unchanged) The self.test_json is: 5
(should be 6, changed) test_json is: 6

• Use copy.copy() for shallow copies (works great for flat objects like your SimpleNamespace).
• Use copy.deepcopy() if your object has nested mutable objects (like another SimpleNamespace inside it), to ensure all nested layers are duplicated.

内容的提问来源于stack exchange，提问作者avico34