Code layered management of interface testing based on RF framework

Catalog

One 、 Test environment

  Two 、 Basic operation

>>> stay RF Import custom library

3、 ... and 、 Code layered management

first floor ： Public level

>>> encapsulation http request

>>> management token

 >>> Others are RF Custom libraries used in

The second floor ： Request layer

  The third level ： Use case layer

>>> Initialize test case data

>>> Execute test case

>>> Test results assert

---

One 、 Test environment

Python	3.7.2
jsonpath	0.82
jsonschema	3.2.0
numpy	1.21.6
Pillow	9.1.1
PyMySQL	1.0.2
PyYAML	6.0
requests	2.26.0
robotframework	4.1
robotframework-requests	0.9.3
robotframework-ride	1.7.4.2
sqlparse	0.4.2
urllib3	1.26.9
wxPython	4.0.7

  Two 、 Basic operation

>>> stay RF Import custom library

To be in RF You need to import the gadgets written by yourself in Library, As shown in the figure below .

Mytools.py,AssertByDataBase.py Are custom libraries

  Of course, you can also import variables , Select import Variables, And then choose .py The file path is just .

  Variables can be written in classes or defined directly

# Variables.py
class BaseConfig:
    """ Basic environment variable """
    baseurl = "http://10.168.2.204:9095/api"
    #  Used to get Token, And the login user name used in the whole process of testing 、 password , The password needs to go through Base64 Encryption processing 
    login_username = "admin"
    login_password = "e10adc3949ba59abbe56e057f20f883e"
    #  Default request header 
    headers = {
        "Content-Type": "application/json",
        "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) "
                      "Chrome/95.0.4638.69 Safari/537.36"}

login_api = "/login"

stay RF Using variables in

comment     Use BaseConfig Variables in class baseurl
log    ${BaseConfig.baseurl}

comment     Using variables login_api
log    ${login_api}

3、 ... and 、 Code layered management

Here is the hierarchical management method I used in my work , Friends have better ways to share with me in the comment area to learn .（ At present, data-driven is not implemented, so only 3 layer , Wait for me to study RF How to realize data-driven re update ）

first floor ： Public level

Use at this level RobotFrameWork-requests The keywords provided by the module are encapsulated , Encapsulate the methods according to the methods required by the project .

>>> encapsulation http request

HttpRequest.txt Encapsulated get,post,delete,put Method is used to send the corresponding http request （ Add after method name u It's to prevent contact with RobotFrameWork-requests The provided original keywords conflict ）, When encapsulating, you can directly assert the response status code according to the project requirements .

getu
    [Arguments]    ${interface}    ${data}=
    ${token}    get_token
    &{headers}    MyTools.Get Headers    ${token}
    Create Session    addr    ${BaseConfig.baseurl}    headers=${headers}    timeout=5000
    ${url}    Catenate    SEPARATOR=    ${interface}    ?    ${data}
    ${res}    GET On Session    addr    ${url}
    Return From Keyword    ${res}

>>> management token

get_headers Function returns a headers be used for http request , First, in the Variables.BaseConfig Define a default request header , Can contain Contain-type,User-Agent Etc , When the user provides Token Data will be Token Join request header , Otherwise, a default request header is returned （ According to the design of different projects , Some login interfaces generally do not need Token value ）

get_token Keyword direct use post Method requests the login interface to get token value , Acquired token Values can be passed as parameters to get_headers So that Token Add to the request header

def get_headers(token=False):
    """
    HTTP Request header , Included by default Content-Type、User-Agent, If the project has other needs, you can edit it yourself headers The content in .
    :param token: The type is String or bool, if  String Then take its value as HTTP The request header Token value , Otherwise, the default request header content is returned .
    :return:  User defined request header content .
    """
    #  default HTTP Request header 
    headers = Variables.BaseConfig.headers
    #  Judge whether the user provides Token
    if token:
        #  Will be provided by the user Token Join in HTTP Request header 
        headers["Token"] = token
    #  Return to the final HTTP Request header 
    return headers

get_token
    log     Try to get TOKEN
    ${data}    Create Dictionary    username=${BaseConfig.login_username}    password=${BaseConfig.login_password}
    ${headers}    MyTools.Get Headers
    Create Session    addr    ${BaseConfig.baseurl}
    ${res}    POST On Session    addr    ${UserApi.apilogin}    json=${data}    headers=${headers}
    ${res_data}    Get From Dictionary    ${res.json()}    data
    ${token}    Get From Dictionary    ${res_data}    token
    Return From Keyword    ${token}

 >>> Others are RF Custom libraries used in

RF Frame pair For Circulation is really unfriendly , Therefore, I am still used to some complex logic python Encapsulated into keywords , Go again RF Call in , I will store these keywords in common/Mytools.py in , For example, database operations and some data operations .

The second floor ： Request layer

Send according to the characteristics of different interfaces in this level HTTP request , The purpose of its encapsulation is to call , After entering the test case parameters, the request results are returned and used in the third layer for assertion .（ At present, there are a lot of repetitive code in this layer , Considering how to further package ）

The third level ： Use case layer

In this level, we mainly achieve the following goals ：

>>> Initialize test case data

Most interfaces need to input certain parameters , Initialize test case data here

 Log in with the correct user name and password 
    [Documentation]     user name  = admin
    ...     password  = 123456 \ (base64 After transcoding, it is  ：e10adc3949ba59abbe56e057f20f883e)
    comment    ***** Initialize the test case ******  
    ${username}    Set Variable    admin
    ${password}    Set Variable    e10adc3949ba59abbe56e057f20f883e

>>> Execute test case

Input test data is performed through the keywords encapsulated in the request layer 、 Execute the test case and get the return result

comment    ***** Execute test case ********
${res}    login    ${username}    ${password}

>>> Test results assert

Execute the return result according to the test case And Compare the expected results To assert that

Comment    ***** Test results assert ********
${res_data}    Get From Dictionary    ${res}    data
${res_data_user}    Get From Dictionary    ${res_data}    user
${req_username}    Get From Dictionary    ${res_data_user}    username
${name}    Get From Dictionary    ${res_data_user}    name
Should Be Equal    ${req_username}    ${username}
Should Be Equal    ${name}     Administrators 

Common assertion keywords are ：

Should Be Equal	Parameters 1 == Parameters 2
Should Be Equal As String	str( Parameters 1) == str( Parameters 2)
Should Contain	Parameters 1 in Parameters 2

You can use it, of course python Write keyword setting assertion logic by yourself , Use assert To assert that , If the assertion fails, it returns AssertionError.