conftest.py file: It is a very important thing in the pytest framework, which can realize the self-correspondence of fixture objects and automatically apply the cross-template and cross-file application operations, so that the definition of fixture objects is more flexible and convenient;

1 . 1 . 2. Example of use

1. Create a directory demo7, create two directories (test_case1 and test_case2) and a file (congtest.py) in this directory; create test files (test_1.py and test_2.py) in directory test_1, and create test files in test_2 ( test_3.py);

2. Contents of the file conftest.py:

import pytest

@pytest.fixture()
def get_data():
    a = '-------1-------'
    b = 2
    print(a)
    return b

3. The content of the file test_1.py: (test_1_1 and test_1_3 refer to the general get_data function of the conftest.py file)

import  pytest
def test_1_1(get_data):
    sun = 1+get_data
    print(sun)
    print('-----test_1_1----')

def test_1_2():
    print('-----test_1_2----')


def test_1_3(get_data):
    sun = 2+get_data
    print(sun)
    print('-----test_1_3----')
if __name__ == '__main__':
    pytest.main()

4. Contents of the file test_2.py: (test_2_1 refers to the general get_data function of the conftest.py file)

import  pytest
def test_2_1(get_data):
    sun = 10+get_data
    print(sun)
    print('-----test_2-1----')

def test_2_2():
    print('-----test_2-2----')


def test_2_3():
    print('-----test_2-3----')
if __name__ == '__main__':
    pytest.main()

5. Contents of the file test_3.py: (test_3_1 and test_3_3 refer to the general get_data function of the conftest.py file)

import  pytest
def test_3_1(get_data):
    sun = 100+get_data
    print(sun)
    print('-----test_3_1----')

def test_3_2():
    print('-----test_3_2----')


def test_3_3(get_data):
    sun = 200+get_data
    print(sun)
    print('-----test_3_3----')
if __name__ == '__main__':
    pytest.main(['-s', ['test_3.py']])

6. Execute three test files (execute all test files in the demo7 directory)

7. Test results:

From the results, it can be seen that the get_data function in conftest.py is referenced across directories and files

1. 2. 3. Scope of conftest.py

1. There can be multiple conftest.py files in a test project. If there are more than one, the scope will be different according to the location of conftest.py; in the root project directory, this file will play a global role. If in In a subdirectory, the file is only valid for the objects under the file and the subdirectories under the current level;

2. The following situations are as follows:

2.1. Add or delete a conftest.py file in the test_case1 directory: (the function name in this file is the same as the function name in conftest.py in the root project directory ( demo 7) )

2.2. File content:

import pytest

@pytest.fixture()
def get_data():
    a = '-------data2-------'
    b = 2
    print(a)
    return b

2.3. Execute all test files under demo7 again, and the test results are as follows: ( The test files in test_case1 refer to the functions in conftest.py in your own directory )

2.4. Summary: When the function name in the root project directory (demo7) conftest.py is the same as the function name in the subdirectory (test_case1) conftest.py, the test file in the test_case1 directory is given priority to use conftest.py in its own directory; and only Objects in the current level and subdirectories do not apply to test_case2; test_case2 uses conftest.py in the demo7 directory (acts globally);

3.1. Add or delete a conftest.py file in the test_case1 directory: (The function name in this file is different from the function name in conftest.py in the root project directory ( demo 7) ) The file test_1.py in test_case1 uses two function;

3.2. Modify test_1.py as follows:

import  pytest
def test_1_1(get_data):   # 引用根工程目录（demo7）下的conftest.py的函数
    sun = 1+get_data
    print(sun)
    print('-----test_1_1----')

def test_1_2():
    print('-----test_1_2----')


def test_1_3(get_data2):  # 引用test_case1下的conftest.py的函数
    sun = 2+get_data2
    print(sun)
    print('-----test_1_3----')
if __name__ == '__main__':
    pytest.main()

3.3 . Execute the test_1.py file, the result is as follows:

1. 2. Multiple fixtures

1. The sample code of the conftest.py file is as follows:

import pytest

@pytest.fixture()
def get_data1():
    a = '-------100-------'
    b = 20
    print(a)
    return b


@pytest.fixture()
def get_data2():
    a = '-------200-------'
    c = 30
    print(a)
    return c

2. The test file test_4.py is as follows:

import  pytest
def test_3_1(get_data1, get_data2):
    sun = get_data1+get_data2
    print(sun)
if __name__ == '__main__':
    pytest.main(['-s', 'test_4.py'])

3. The result of executing the test_4.py file is as follows:

1.3. Mutual calls of fixtures

1. The sample code of the conftest.py file is as follows: (data2 calls data1)

import pytest

@pytest.fixture()
def data1():
    print('----data1---')
    a = 1
    return a


@pytest.fixture()
def data2(data1):    # 调用data1的数据
    print('----data2---')
    b = 2+data1
    return b

2. The test file test_1.py is as follows:

import pytest

def test_1(data2):
    sum = 100+data2
    print(sum)
    assert sum==103

if __name__ == '__main__':
    pytest.main()

3. Execution result:

1. 4. The scope of the fixture

1 . 4 . 1. Introduction

1. There is a parameter scope in the fixture, which can control the scope of the fixture object, and its values include session, module, class, function, package, etc.

1 . 4 . 2. The scope difference between class and function in scope

1. The function in the scope is equivalent to the setup_function at the function level and the setup_method at the class level

2. Let's test the difference between class and function in scope:

2.1, scope='class', the sample code is as follows:

import pytest

@pytest.fixture(scope='class')
# @pytest.fixture(scope='function')
def data1():
    a = 2
    print('----data1----')
    return a

def test_1(data1):
    sum = 1 + data1
    print(sum)
    print('----test1----')

class Test_1():
    def test_2(self, data1):
        sum = 2 + data1
        print(sum)
        print('---test2_1---')

    def test_1(self, data1):
        sum = 3+data1
        print(sum)
        print('---test2_2---')

def test_2(data1):
    sum = 4+data1
    print(sum)
    print('----test2----')

if __name__ == '__main__':
    pytest.main()

2.2. Execution result: (class type data is only executed once data1, but the data can be passed to all functions in the class; function type use cases will be executed once)

3.1, scope='function', the sample code is as follows:

import pytest

# @pytest.fixture(scope='class')
@pytest.fixture(scope='function')
def data1():
    a = 2
    print('----data1----')
    return a

def test_1(data1):
    sum = 1 + data1
    print(sum)
    print('----test1----')

class Test_1():
    def test_2(self, data1):
        sum = 2 + data1
        print(sum)
        print('---test2_1---')

    def test_1(self, data1):
        sum = 3+data1
        print(sum)
        print('---test2_2---')

def test_2(data1):
    sum = 4+data1
    print(sum)
    print('----test2----')

if __name__ == '__main__':
    pytest.main()

3.2. The execution results are as follows: (acting on functions, all functions in the type test case will be executed once; function type test cases will also be executed once)

4. Summary: The difference between function and class: if the two are applied to the test function, both are equivalent, and both represent the creation of a new firmware object; if it is for the test method in the class, function represents each test The method creates a brand new object, and class means that all test cases in the test class share a test firmware;

1 . 4 . 3. The difference between the scope of session and module in scope

1. The difference between session and module: module means that for a module, the same firmware object is used in each module, and the firmware objects created by different modules are different; while session means the entire session, and the objects used by different modules are also different;

1.5, fixture use case management

1 . 5 . 1. Skip use case skip /skipif

1. @pytest.mark.skip(reason='do not execute')

2. Test file:

import pytest

@pytest.mark.skip(reason='不执行')
def test_01():
    print('---test1---')

def test_02():
    print('---test2---')

if __name__ == '__main__':
    pytest.main()

3. Execution result:

4. @pytest.mark.skipif(2<3,reason='skip use case'); test file:

import pytest

@pytest.mark.skipif(2<3,reason='跳过用例')
def test_01():
    print('---test1---')

def test_02():
    print('---test2---')

if __name__ == '__main__':
    pytest.main()

5. Execution result:

1 . 5 . 2. Mark failure xfail

1. Code example:

import pytest

def test_01():
    print('---test1---')

@pytest.mark.xfail(reason='跳过')
def test_02():
    print('---test2---')
    assert 1==2

if __name__ == '__main__':
    pytest.main()

2. Execution result:

1 . 5 . 3. Repeat repeat

1. Install the pytest-repeat library first; execute pip install pytest-repeat

2. Sample code: ( specify the number of times the test case is executed by the method of mark mark )

import pytest

def test_01():
    print('---test1---')
@pytest.mark.repeat(5)
def test_02():
    print('---test2---')
    assert 1==1

if __name__ == '__main__':
    pytest.main(['-s','test_0_1.py'])

3. Execution result:

4. Sample code: ( executed by the parameter ' -- count =5 ' , all test cases in the test file will be executed 5 times in sequence )

import pytest

def test_01():
    print('---test1---')

# @pytest.mark.ios
# @pytest.mark.repeat(5)
def test_02():
    print('---test2---')
    assert 1==1

if __name__ == '__main__':
    pytest.main(['-s', 'test_0_1.py', '--count=5'])  # 会对该测试文件里的所有测试用例都执行5次

5. Execution result: ( after use case 1 is executed 5 times , use case 2 is executed 5 more times )

6. You can use the parameter ' --repeat-scope= session ' to specify the repeated scope ( similar to the scope of the fixture ) , sample code:

import pytest

def test_01():
    print('---test1---')

# @pytest.mark.ios
# @pytest.mark.repeat(5)
def test_02():
    print('---test2---')
    assert 1==1

if __name__ == '__main__':
    pytest.main(['-s', 'test_0_1.py', '--count=5', '--repeat-scope=session'])  # 整个文件的用例执行五次

7. Execution result: ( the use case of the entire file is executed five times )

8. Combining --count =10000 and -X parameters, in the specified number of repeated executions, if the first use case execution fails, it will stop and continue to execute ;

9. --repeat-scope Summary:

9.1, function: default parameter, the scope of action is to execute the next test case after repeated execution of each test case;

9.2, class: Indicates that the use cases in the class are repeatedly executed in units of use case sets;

9.3, module: Indicates that the use cases in the module are repeatedly executed in units of modules;

9.4, session: Indicates that the entire test session is used as a unit, that is, all collected tests are executed once, and then all use cases are executed once;

1.6 . Assertion mechanism of fixture

1 . 6 . 1. Common assertions

1. Common assertions:

1.1. assert XX: judge that XX is true;

1.2, assert not XX: judge that XX is not true;

1.3. assert a in b: judge that b contains a;

1.4, assert a==b: judge that a is equal to b;

1.5. assert a !=b: judge that a is not equal to b;

2. Sample code:

import pytest

class Test_a():

    def test_1(self):
        assert 1

    def test_2(self):
        assert 'sh' in 'shell'

    def test_3(self):
        assert 1==1

    def test_4(self):
        assert 'shell' != 'shell_1'

if __name__ == '__main__':
    pytest.main(['-s','test_1.py'])

3. Execution result:

1.7 . Use pytest.ficture decorator to carry parameters

1. 7. 1. Use the parameter params in the test firmware object

1. Sample code:

import pytest

class Calculater():
    def add(self,a,b):
        return a+b

@pytest.fixture(scope='function', params=[[1,2,3],[2,3,5]])
def get_data(request):
    # print(Calculater())
    # print(request.param)
    return Calculater(), request.param

def test_add(get_data):
    print(get_data[1][0], get_data[1][1], get_data[1][2])
    assert get_data[0].add(get_data[1][0], get_data[1][1]) == get_data[1][2]

if __name__ == '__main__':
    pytest.main(['-s', 'test_2.py'])

2. Test results:

3. Summary: If the params method is used to achieve parameterization , the parameter name passed in the firmware object is always request , and the parameter object must be returned , that is, request.params ;

4. The parameterized function can also be directly placed in conftest.py, which can separate the firmware initialization code and the parameterized code to realize two functions;

5. conftest.py file, sample code:

import pytest

class Calculater():
    def add(self, a, b):
        return a+b

@pytest.fixture(scope='function')
def get_data():
    return Calculater()

@pytest.fixture(params=[[1,2,3],[2,2,4]])
def get_data_params(request):
    return request.param

6. Test code:

import pytest


def test_add_1(get_data, get_data_params):
    print(get_data_params[0],get_data_params[1],get_data_params[2])
    assert get_data.add(get_data_params[0],get_data_params[1])==get_data_params[2]

if __name__ == '__main__':
    pytest.main(['-s', 'test_3.py'])

7. Test results:

8. It can also be implemented in the following form, sample code:

import pytest

class Calculater_2(object):
    def __init__(self, a, b):
        self.a = a
        self.b = b
    def add(self):
        return self.a+self.b

@pytest.fixture(scope='function', params=[[1,1,2],[3,3,6]])
def get_data_params_2(request):
    return Calculater_2(request.param[0], request.param[1]), request.param[2]

def test_sum(get_data_params_2):
    return get_data_params_2[0].add() == get_data_params_2[1]

if __name__ == '__main__':
    pytest.main(['-s', 'test_4.py', '--disable-warnings'])

9. Execution result:

1 . 7 . 1. Parse the yaml file to transfer data

1. The yaml file is as follows: data.yaml

data:
  - [1,1,2]
  - [2,2,4]
  - [3,3,6]

2. Parse the yaml file: read_yaml.py

import yaml


def readyaml():
    with open('data.yaml', encoding='utf-8') as f:
        get_yaml = yaml.load(f, Loader=yaml.FullLoader)
    print(get_yaml)
    print(type(get_yaml))
    return get_yaml['data']

# readyaml()

3. conftest.py file (change params=specific value to the value parsed from yaml)

import pytest
from demo10.read_yaml import *   # 导入解析yaml得到yaml值的文件
class Calculater():
    def add(self, a, b):
        return a+b

@pytest.fixture(scope='function')
def get_data():
    return Calculater()

@pytest.fixture(params=readyaml())
def get_data_params(request):
    return request.param

4. Test file: test_4.py

import pytest



def test_add_1(get_data, get_data_params):
    print(get_data_params[0],get_data_params[1],get_data_params[2])
    assert get_data.add(get_data_params[0],get_data_params[1])==get_data_params[2]

if __name__ == '__main__':
    pytest.main(['-s', 'test_3.py'])

5. Execute the test_4.py file to get the result:

1. 8. mark custom mark

1 . 8 . 1. Customize mark test cases and execute test cases according to marks

1. Code example: (5 test cases are marked differently, use case 1 is marked as: L1; use cases 2 and 3 are marked as: apple; use cases 4 and 5 are marked as: case1)

import pytest

@pytest.mark.L1
def test_1():
    print('---test1---')

@pytest.mark.apple
def test_2():
    print('---test2---')

@pytest.mark.apple
def test_3():
    print('---test3---')

@pytest.mark.case1
def test_4():
    print('---test4---')
@pytest.mark.case1
def test_5():
    print('---test5---')

if __name__ == '__main__':
    pytest.main(['-s', 'test_5.py', '-m case1', '--disable-warnings'])

2. Execution result: (The result below is the execution marked as: case1)

3. If you want to execute all use cases marked as case1; parameters in the file:

pytest.main(['-s', 'test_5.py', '-m not case1', '--disable-warnings'])

When executed in the terminal , the command line is as follows : pytest -s test_5.py -m "not case1"

Results of the:

3. Summary : directly add the parameter ' -m case1 ' or ' -m =case1 ' ; execute all use cases except the selected mark , when adding parameters in the test file ' -m not case1', when executing the command line in the terminal : -m "not case1"

Pytest test framework (4) --- use of decorators

1. Fixture decorator

1.1 Introduction and use of fixture's conftest.py file

1 . 1 . 1. Introduction

1 . 1 . 2. Example of use

1. 2. 3. Scope of conftest.py

1. 2. Multiple fixtures

1.3. Mutual calls of fixtures

1. 4. The scope of the fixture

1 . 4 . 1. Introduction

1 . 4 . 2. The scope difference between class and function in scope

1 . 4 . 3. The difference between the scope of session and module in scope

1.5, fixture use case management

1 . 5 . 1. Skip use case skip /skipif

1 . 5 . 2. Mark failure xfail

1 . 5 . 3. Repeat repeat

1.6 . Assertion mechanism of fixture

1 . 6 . 1. Common assertions

1.7 . Use pytest.ficture decorator to carry parameters

1. 7. 1. Use the parameter params in the test firmware object

1 . 7 . 1. Parse the yaml file to transfer data

1. 8. mark custom mark

1 . 8 . 1. Customize mark test cases and execute test cases according to marks

Guess you like