UI automation based on Selenium

Catalog

One 、Selenium brief introduction

1、Selenium Introduction to tool components

2、Selenium WebDriver Introduction and implementation principle

Two 、Selenium WebDriver be based on Python Implementation script

1、Selenium Environmental installation

2、Selenium Common operations for manipulating the browser

3、 Write a simple one Selenium Script

3、 ... and 、 Automated testing framework And Architecture design The choice of

1、 be based on Python A popular testing framework for languages

2、PO Design patterns

---

One 、Selenium brief introduction

        Selenium It's based on Web A collection of tools for automated testing of applications ,Selenium Test runs directly in browser , The essence is to drive the browser , Simulate user's operation . The main functions of the tool include ： Test compatibility with browser , Carry out system function test .

1、Selenium Introduction to tool components

• Selenium IDE： It's embedded in FireFox Browser plug-ins , Used in Firefox Record and playback on Selenium Script , Although only in Firefox Next use , But it can convert recorded scripts into various Selenium WebDriver Supported programming languages , Then it can be extended to a wider range of browser types ;

• Selenium RC： It mainly consists of two parts ：selenium Server+Client Libraries; among Selenium  Server Responsible for controlling browser behavior , and Client Libraries It is used to control when writing test cases for testers selenium Server The library of .Selenium RC Also known as Selenium 1 Later and WebDriver Merger brings Selenium 2; stay Selenium 3.0 In the version Selenium RC No longer exists .

• Selenium WebDriver： A set for operating the browser API; Can support multiple languages , Support various types of browsers , Cross operating system ,WebDriver Provides complete , Used to implement web Third party libraries for automated testing ;Selenium WebDriver There are currently two versions Selenium 2 and Selenium 3.

• Selenium Grid： Is a dynamic tool , It can be used in multiple physical / Distribute and run concurrently in the virtual machine Selenium test . It can significantly speed up the testing process across different browsers , And shorten the test cycle by providing fast and accurate results .

  

2、Selenium WebDriver Introduction and implementation principle

        Selenium WebDriver It is a general testing framework in the industry , Not only is web The standard of testing , At the same time, it is also the underlying core driver framework in the field of mobile testing

• characteristic ：

  • 1、 Support multi browser operation ：Chrome、IE、Firefox、Safaria etc.

  • 2、 Support platform ：Windows、Linux、Mac etc.

  • 3、 Support for multiple languages ：Python、Java、Ruby、C# etc.

• Selenium The process implemented by the back end when the script is executed ：

  • 1. For each Selenium Script instructions , Will create a HTTP Request and send to the browser driver

  • 2.WebDriver Launch browser driver , establish HTTP Server, To receive these http request

  • 3.HTTP Server After receiving the request, control the corresponding browser according to the request

  • 4. The browser performs specific test steps , Return the step execution result to HTTP Server

  • 5.HTTP Server And return the result to Selenium Script for , If it's wrong http Code, we will see the corresponding error message on the console

    

• WebDriver Is in accordance with the Client/Server Pattern design

  • Client： Programming language client ;

  • Server： Browser driver . It is used to receive requests from clients and drive browsers to perform operations , And then return the result ;

• WebDriver Based protocols ：JSON Wire protocol

  • JSON Wire protocol Is in http On the basis of the agreement , Yes http Request and response body Further specification of some data .body Part of the main transmission of specific data , And in WebDriver These data are based on JSON Existing and transmitted in the form of . So in Client and Server Between , As long as it's based on JSON Wire Protocol To transfer data , The driver of the same browser can handle scripts in different languages .

Two 、Selenium WebDriver be based on Python Implementation script

1、Selenium Environmental installation

• install selenium

  • pip install selenium

• install WebDriver The driver

  • The driver needs the corresponding browser version and selenium Version corresponding

    • Firefox  Browser driven ：Firefox

    • Chrome  Browser driven ：Chrome

    • IE  Browser driven ：IE

    • Edge  Browser driven ：Edge

    • PhantomJS  Browser driven ：PhantomJS

    • Opera  Browser driven ：Opera

  • WebDriver Drive installation path

    •  Win： Copy webdriver To Python Installation directory

    •  Mac： Copy webdriver To /usr/local/bin Under the table of contents

2、Selenium Common operations for manipulating the browser

• Page base operation

  • Get URL ：get(url)

  • window maximizing ：maximize_window()

  • Minimize the window ：minimize_window()

  • back off 、 Forward 、 Refresh ：back()、forword()、refresh()

  • Close the browser ：quit()

  • Webpage title ：title

  • Webpage URl：current_url

  • Window handle

    • Current window handle ：current.windows.handles

    • All window handles ：windows.handles

  • Current page source code ：page_source

  • Take a screenshot of the current page ：get_screenshot_as_file(path)

• Eight positioning methods of elements

  • id location ：find_element_by_id()

  • name location ：find_element_by_name()

  • class location ：find_element_by_class_name()

  • link location ：find_element_by_link_text()

  • partial link location ：find_element_by_partial_link_text()

  • tag location ：find_element_by_tag_name()

  • xpath location ：find_element_by_xpath()

  • css location ：find_element_by_css_selector()

• There are three ways to wait for elements

  • Mandatory waiting ： Commonly used for interaction between systems

    • time.sleep()

  • The recessive waiting ： Global settings , Timeout can be set ; Can only be used to wait for elements

    • implicitly_wait()

  • Explicit wait ： Wait for an element to be clicked 、 so 、 Whether the window is opened ; Each wait needs to be set separately

    • WebDriverWait(driver,timeout,poll_frequency=0.5,ignored_exceptions=None)

• Three switching modes

  • Window switch , Window handle

    • Get window handle ：windowhandles[-1]

    • switch_to_window()

  • iframe Switch

    • switch_to.iframe()

  • Pop up switch toast

    • switch_to_alert

• The mouse operation （ stay webdriver in Mouse operations are encapsulated in ActionChains class ）

  • Click on ：click()

  • perform() ： To perform all ActionChains Actions stored in

    • double-click ：double_click()

    • Drag and drop ：drag_and_drop()

    • hover ：move_to_element()

    • Right click ：content_click()

• Drop down box operation

  • Select()

• User input 、 Keyboard operation 、 File path

  • send_keys()

• send out js Instructions （ Mainly used to make up for selenium Specific page operations that cannot be done ）

  • Page scrolling （ for example ： Slide to the bottom of the page ）

  • Modify element properties

• System interactive upload

  • Win：pywinauto

  • Cross platform ：pyautogui

3、 Write a simple one Selenium Script

• Launch the browser

• get Open the specified page 、 Perform element positioning and operation

from selenium import webdriver​


#  Launch Google browser ​
driver = webdriver.Chrome(executable_path=r"/Users/bytedance/Desktop/chromedriver")​
#  Set hidden wait ​
driver.implicitly_wait(5)​

#  Open the web page CSDN​
driver.get("https://www.csdn.net/?spm=1001.2101.3001.4476")​
#  location " Input box "​
input_box = driver.find_element("xpath",'//*[@id="toolbar-search-input"]')​
#  Input text ："Selenium Automated testing framework "​
input_box.send_keys("Selenium Automated testing framework ")​
#  location " Search for " Function and click ​
driver.find_element("xpath",'//*[@id="toolbar-search-button"]/span').click()​


#  Close the browser ​
driver.quit()

• Use selenium The premise of locating page elements is to understand the basic HTML Page layout and various label meanings , adopt 【F12】 Developer tools , Locate and get the elements we need from the page source code ;

  • Shortcut ： After finding the element , You can right click. Copy Element positioning expression .    

3、 ... and 、 Automated testing framework And Architecture design The choice of

1、 be based on Python A popular testing framework for languages

         The encapsulation of public methods and the design of how to organize scripts and configuration files are called frameworks .

• Unittest: Python Bring their own , The most basic unit testing framework

• Nose: be based on Unittest Development , Easy to use , There are many plug-ins

• Pytest: Also based on Unittest Development , Easy to use , More details , There are many plug-ins

• Robot Framework： Based on a Python Keyword Driven Testing Framework for language , There is an interface , Functional perfection , Bring your own report and log Clear and beautiful

• On the whole ,Unittest More basic , Convenient secondary development , Suitable for experts ;Pytest/Nose More convenient and quick , More efficient , Suitable for Xiaobai and companies pursuing efficiency ;Robot Framework Due to the interface and beautiful report , Easier to use , The flexibility and customizability are slightly poor .      

  

2、PO Design patterns

         For a good framework , What is indispensable is the idea of hierarchy , And in the Web UI Automated testing ,PO Pattern （Page Object） Is a very mainstream technology ;PO Design pattern provides a pattern of separating page element positioning and business operation process . When page elements change , You only need to maintain the corresponding page Layer modification positioning , There is no need to modify the business logic code .

• PO The core idea is layering ： Realize script reusability , Maintainability , Readability , Extensibility ; There are three main layers ：

  • Object library layer ：Base（ Base class ）, Some common methods for encapsulating pages , Such as initialization method 、 Find element method 、 Click the element method 、 Input method 、 Get text method 、 Screenshot method, etc .

  • The operational layer ：page( Page object ), Encapsulates operations on elements , Encapsulate a page into an object .

  • The business layer ：business（ The business layer ）, Combine one or more page operations to complete a business function test . For example, login. ： Account number required 、 password 、 Click login to perform three operations .

• Based on the idea of layering and PO Design patterns , We can design a basic framework model ：

  • cases Test case layer ： Store all test cases

  • common Public level ： Store some public methods , Such as encapsulation page Page base class 、 Capture logs, etc

  • datas Test the data layer ： Store test data

  • logs Log layer ： Store all captured logs and error logs , It's easy to locate the problem

  • pages Page object layer ： Store all page objects , A page is encapsulated into an object

  • reports Test report layer ： Store the output test result data , Screenshot of failure

  • run Use case execution layer ： Store test execution files

  • requirements.txt： Record all dependent packages of the current project and their exact version numbers , For subsequent migration projects .