问题

I'm here to provide translations for your code. Here's the translated code:

我对pytest还很陌生，无法解决一个令人讨厌的问题。
我想要一个使用dataclass的基本测试值的fixture。对于某些测试函数，我想要在本地修改其中一些值，以测试特定情况。但我希望保持基本值不变，以便在下一个测试函数中使用！

这是我的代码：

```python
import pytest
from dataclasses import dataclass

@dataclass
class Def_Input:
    var1 = 3
    var2 = 4
    
@pytest.fixture
def def_inp():
    def_inp = Def_Input()
    return def_inp

def test_1(def_inp):
    temp = def_inp
    temp.var1 = 5
    assert def_inp.var1 == 3
    assert temp.var1 + temp.var2 == 9

第二个断言没有问题，但第一个有问题。我得到一个断言错误，说当我没有修改def_inp时，def_inp.var1 = 5，但我使用了一个临时变量...

您是否有任何解决方法的想法？

编辑：
@Steven97102 这对我提供的示例有效，但我正在处理一个稍微复杂一点的函数。在dataclass中，我从外部类创建对象。类似于这样：

import pytest
import dataclasses
from dataclasses import dataclass

class Tab:
    def __init__(self, name):
        self.name = name

@dataclass
class Def_Input:
    var1 = 3
    var2 = 4
    name = 'paul'
    tab = Tab(name)
    
@pytest.fixture
def def_inp():
    def_inp = Def_Input()
    return def_inp

def test_1(def_inp):
    temp = dataclasses.replace(def_inp)
    temp.var1 = 5
    temp.tab.name = 'john'
    assert def_inp.tab.name == 'paul'
    assert def_inp.var1 == 3
    assert temp.var1 + temp.var2 == 9

您是否知道如何使第一个断言生效？

<details>
<summary>英文:</summary>

I&#39;m new to pytest and I can&#39;t solve an annoying problem.
I want to have a fixture with base test values from a dataclass. And for certain test functions, I want to to modify some of these values locally to test some specificities. But I want to keep the base values unaltered for the next test functions !

Here&#39;s my code:

import pytest
from dataclasses import dataclass

@dataclass
class Def_Input:
var1 = 3
var2 = 4

@pytest.fixture
def def_inp():
def_inp = Def_Input()
return def_inp

def test_1(def_inp):
temp = def_inp
temp.var1 = 5
assert def_inp.var1 == 3
assert temp.var1 + temp.var2 == 9

The 2nd assert is not a problem, but the first one is. I get an AssertionError saying def_inp.var1 = 5 when I didn't modify def_inp, I used a temporary variable...

Would you have any ideas at to solve this?

Edit:
@Steven97102 this does work for the example I gave, but I work on a slightly more complex function. In the dataclass, I create objects from external classes. It's something like this:

import pytest
import dataclasses
from dataclasses import dataclass

class Tab:
def init(self, name):
self.name = name

@dataclass
class Def_Input:
var1 = 3
var2 = 4
name = 'paul'
tab = Tab(name)

@pytest.fixture
def def_inp():
def_inp = Def_Input()
return def_inp

def test_1(def_inp):
temp = dataclasses.replace(def_inp)
temp.var1 = 5
temp.tab.name = 'john'
assert def_inp.tab.name == 'paul'
assert def_inp.var1 == 3
assert temp.var1 + temp.var2 == 9

Would you know I could get the first assert to work?

</details>


# 答案1
**得分**: 0

Python使用引用模型来传递类实例。这意味着`temp = def_inp`仅仅创建一个指向与`def_inp`相同对象的变量`temp`。为了避免修改现有实例，你需要创建一个新的实例。

以下是制作`def_inp`的深度拷贝的解决方案：

```python
import pytest
import dataclasses
from dataclasses import dataclass

@dataclass
class Def_Input:
    var1 = 3
    var2 = 4
    
@pytest.fixture
def def_inp():
    def_inp = Def_Input()
    return def_inp

def test_1(def_inp):

    # 制作"def_inp"的深度拷贝
    temp = dataclasses.replace(def_inp)

    temp.var1 = 5
    assert def_inp.var1 == 3
    assert temp.var1 + temp.var2 == 9

希望这对你有所帮助。

import pytest
import dataclasses
from dataclasses import dataclass

@dataclass
class Def_Input:
    var1 = 3
    var2 = 4
    
@pytest.fixture
def def_inp():
    def_inp = Def_Input()
    return def_inp

def test_1(def_inp):

    # make a deep copy of "def_inp"
    temp = dataclasses.replace(def_inp)

    temp.var1 = 5
    assert def_inp.var1 == 3
    assert temp.var1 + temp.var2 == 9