WebDriver

The W3C WebDriver API is a platform and language-neutral interface and wire protocol allowing programs or scripts to control the behavior of a web browser.

Introduction

WebDriver API

WebDriver enables developers to create automated tests that simulate user interaction. This is different from JavaScript unit tests because WebDriver has access to functionality and information that JavaScript running in the browser doesn't, and it can more accurately simulate user events or OS-level events. WebDriver can also manage testing across multiple windows, tabs and webpages in a single test session.

WebDriver UDF

The WebDriver UDF allows to interact with any browser that supports the W3C WebDriver specifications. Supporting multiple browsers via the same code base is now possible with just a few configuration settings.

Requirements

(Last modified: 2021/06/18)

The following UDFs need to be installed - independent of the Browser you try to automate:

• JSON UDF (AutoIt) by Ward and Jos
• WinHTTP UDF (AutoIt) by trancexx or autoit-winhttp (GitHub)
• WebDriver UDF (GitHub) by Danp2

One of the following Drivers needs to be installed - depending on the Browser type and version you try to automate:

Browser	Download Link	Latest Version / Date	Comments
Chrome	Google	92.0.4515.43 / 2021.06.11	Follow this link to select the correct version depending on the Chrome version you run!
Edge	Microsoft	91.0.864.53 / 2021.06.17
Firefox	GitHub	0.29.1 / 2021.04.09	Firefox version ≥ 60 is recommended Note: You must still have the Microsoft Visual Studio redistributable runtime installed on your system for the binary to run. This is a known bug in version 0.26 which the authors weren't able fix for this release.
Opera	GitHub	91.0.4472.77 / 2021.06.09	The versioning of OperaDriver matches the Chromium version on which Opera browser is based on.

Limitations

(Last modified: 2020/01/28)
Not all WebDriver functions have been implemented by each browser. To check the status goto the corresponding website below:

• Chrome
• Edge
• Firefox
• Opera: "OperaChromiumDriver is a WebDriver implementation derived from ChromeDriver and adapted by Opera". That's why I think it has at least the same limitations as ChromeDriver.

Big Picture

How the browser independent and browser dependent parts fit together:

Used Terms

(Last modified: 2021/07/28)

You will find the following terms when using WebDriver. We try to shed some light onto this subject here:

CDP (Chrome DevTools Protocol)
Is a protocol that allows for tools to instrument, inspect, debug and profile Chromium, Chrome and other Blink-based browsers.

Marionette
Marionette is an automation driver for Mozilla’s Gecko engine. It can remotely control either the UI or the internal JavaScript of a Gecko platform, such as Firefox. It can control both the chrome (i.e. menus and functions) or the content (the webpage loaded inside the browsing context), giving a high level of control and ability to replicate user actions. In addition to performing actions on the browser, Marionette can also read the properties and attributes of the DOM.
Marionette consists of two parts: a server which takes requests and executes them in Gecko (the Marionette server ships with Firefox), and a client (the Marionette client ships with the GeckoDriver exe). The client sends commands to the server and the server executes the command inside the browser.
For details please visit this site.

ShadowRoot
The ShadowRoot interface of the Shadow DOM API is the root node of a DOM subtree that is rendered separately from a document's main DOM tree.
For details please visit this site.

Installation

(Last modified: 2020/09/27)

To automate your browser the following installation steps are needed:

• Download the files listed in section "Requirements"
• Move the UDFs to a directory where SciTE and Autoit can find them:
  • Json.au3 and BinaryCall.au3 from the JSON UDF
  • wd_Core.au3, wd_helper.au3 , wd_cdp.au3 , wd_capabilities.au3 from the WebDriver UDF
  • WinHttp.au3 and WinHttpConstants.au3 from the WinHttp UDF
• Move the browser dependent WebDriver to the same directory (with WD_Demo.au3):
  • chromedriver.exe (Chrome)
  • geckodriver.exe (Firefox)
  • msedgedriver.exe (Edge - Chromium) or MicrosoftWebDriver.exe (Edge - EdgeHTML)
• Run WD_Demo.au3 and select "DemoNavigation" to validate the installation.
  The result (for Firefox) displayed in the DOS window should be similar to the following:

1577745813519   geckodriver     DEBUG   Listening on 127.0.0.1:4444
1577745813744   webdriver::server       DEBUG   -> POST /session {"capabilities": {"alwaysMatch": {"browserName": "firefox", "acceptInsecureCerts":true}}}
1577745813746   geckodriver::capabilities       DEBUG   Trying to read firefox version from ini files
1577745813747   geckodriver::capabilities       DEBUG   Found version 71.0
1577745813757   mozrunner::runner       INFO    Running command: "C:\\Program Files\\Mozilla Firefox\\firefox.exe" "-marionette" "-foreground" "-no-remote" "-profile" "C:\\ ...
1577745813783   geckodriver::marionette DEBUG   Waiting 60s to connect to browser on 127.0.0.1:55184
1577745817392   geckodriver::marionette DEBUG   Connection to Marionette established on 127.0.0.1:55184.
1577745817464   webdriver::server       DEBUG   <- 200 OK {"value":{"sessionId":"925641bf-6c5d-4fe2-a985-02de9b1c7c74","capabilities":"acceptInsecureCerts":true,"browserName":"firefox", ...

Function reference

(Last modified: 2021/12/09)

The functions are now documented in the CHM help file that comes with the UDF.

Capabilities

WebDriver capabilities are used to communicate the features supported by a given implementation. The local end may use capabilities to define which features it requires the remote end to satisfy when creating a new session. Likewise, the remote end uses capabilities to describe the full feature set for a session.
Details can be found on WebDriver Capabilities (sub page).

Browser related functionality

Google Chrome

ChromeDriver supports "Chrome DevTools Protocol" (CDP) commands (for an explanation of the term and related links please see the Used Terms section).
Details can be found in the CHM Help file that comes with the UDF.

Troubleshooting

Debug the WebDriver setup

(Last modified: 2020/07/06)

WinHTTP UDF

Make sure that you are running at least version 1.6.4.2 (currently unreleased, but can be obtained here).

Chrome

Chrome does not start and the DOS window for chromedriver does not get displayed

Problem	Solution	Reference
When running WD_Demo.au3 it does not start up Chrome and does not display the DOS window for chromedriver. When you manually run the chromedriver in a DOS window you get message "[0.023][SEVERE]: CreatePlatformSocket() returned an error: An invalid argument was supplied."	This could be caused by missing execution permission for the network drive. Please ask your IT admin for "Applocker" or "application directory whitelisting". Or run the chrome driver from a local HDD and call _WD_Option to set the location of the webdriver executable. Example: _WD_Option("Driver", "C:\Local\WebDriver\chromedriver.exe")	Stackoverflow

Firefox

Firefox does not start and the DOS window for geckodriver does not get displayed

Problem	Solution	Reference
When running WD_Demo.au3 it does not start up Firefox and does not display the DOS window for geckodriver. When you manually run the geckodriver in a DOS window you get message "geckodriver: error: An invalid argument was supplied. (os error 10022)"	This could be caused by missing execution permission for the network drive. Please ask your IT admin for "Applocker" or "application directory whitelisting". Or run the gecko driver from a local HDD and call _WD_Option to set the location of the webdriver executable. Example: _WD_Option("Driver", "C:\Local\WebDriver\geckodriver.exe")	Stackoverflow

FAQ

(Last modified: 2021/10/13)

1. How to connect to a running browser instance

Q: How can I connect to a running browser instance?
A: That's described (for Firefox, but should work similar for other browsers) in this post.

2. How to hide the webdriver console

Q: How can I hide the webdriver console?
A: The console can be completely hidden from the start by adding the following line near the beginning of your script:

$_WD_DEBUG = $_WD_DEBUG_None ; You could also use $_WD_DEBUG_Error

You can also control the visibility of the console with the function _WD_ConsoleVisible.

3. How to utilize an existing user profile

Q: Can I use an existing user profile instead of the default behavior of using a new one?
A: This is controlled by your "capabilities" declaration, with each browser using a different method to implement. Here are some examples:

Chrome

$sDesiredCapabilities = '{"capabilities": {"alwaysMatch": {"goog:chromeOptions": {"w3c": true, "args":["--user-data-dir=C:\\Users\\' & @UserName & '\\AppData\\Local\\Google\\Chrome\\User Data\\", "--profile-directory=Default"]}}}}'

Firefox

$sDesiredCapabilities = '{"capabilities":{"alwaysMatch": {"moz:firefoxOptions": {"args": ["-profile", "' & GetDefaultFFProfile() & '"],"log": {"level": "trace"}}}}}'

Func GetDefaultFFProfile()
	Local $sDefault, $sProfilePath = ''

	Local $sProfilesPath = StringReplace(@AppDataDir, '\', '/') & "/Mozilla/Firefox/"
	Local $sFilename = $sProfilesPath & "profiles.ini"
	Local $aSections = IniReadSectionNames ($sFilename)

	If Not @error Then
		For $i = 1 To $aSections[0]
			$sDefault = IniRead($sFilename, $aSections[$i], 'Default', '0')

			If $sDefault = '1' Then
				$sProfilePath = $sProfilesPath & IniRead($sFilename, $aSections[$i], "Path", "")
				ExitLoop
			EndIf
		Next
	EndIf

	Return $sProfilePath
EndFunc

You will also likely need to specify the marionette port:

_WD_Option('DriverParams', '--marionette-port 2828')

4. How to specify location of browser executable

Q: Is it possible to launch a browser installed in a non-standard location?
A: This is controlled by your "capabilities" declaration. Here are some examples:

Chrome

$sDesiredCapabilities = '{"capabilities": {"alwaysMatch": {"goog:chromeOptions": {"w3c": true, "binary":"C:\\Path\\To\\Alternate\\Browser\\chrome.exe" }}}}'

Firefox

$sDesiredCapabilities = '{"desiredCapabilities":{"javascriptEnabled":true,"nativeEvents":true,"acceptInsecureCerts":true,"moz:firefoxOptions":{"binary":"C:\\Path\\To\\Alternate\\Browser\\firefox.exe"}}}'

Alternate Firefox method:

_WD_Option('DriverParams', '--binary "C:\Program Files\Mozilla Firefox\firefox.exe" --log trace ')

5. How to maximize the browser window

Q: Is it possible to maximize the browser window?
A: Simply call the following function:

_WD_Window($sSession, "Maximize")

Make sure to call _WD_Window after the session has been created with _WD_CreateSession.

6. How to specify location of WebDriver executable

Q: Is it possible to launch the WebDriver executable from a specific location?
A: This is controlled by function "_WD_Option". Example:

_WD_Option("Driver", "C:\local\WebDriver\WebDriver.exe")

7. How to retrieve the values of a drop-down list

Q: How to retrieve the values of a drop-down list (<Select> tag)?
A1: Here's a simple way to do it:

$sElement = _WD_FindElement($sSession, $_WD_LOCATOR_ByXPath, "//select[@name='xxx']")
$sText = _WD_ElementAction($sSession, $sElement, 'property', 'innerText')
$aOptions = StringSplit ( $sText, @LF,  $STR_NOCOUNT)
_ArrayDisplay($aOptions)

'xxx' is the name of the drop-down list.

A2: This can now be accomplished using the function _WD_ElementSelectAction:

$sElement = _WD_FindElement($sSession, $_WD_LOCATOR_ByXPath, "//select[@name='xxx']")
$aOptions = _WD_ElementSelectAction ($sSession, $sElement, 'options')
_ArrayDisplay($aOptions)

8. How to run the browser in headless mode

Q: How do I run the browser in "headless" mode?
A: This is controlled by the Capabilities string that is passed to _WD_CreateSession. Example:

$sDesiredCapabilities = '{"capabilities": {"alwaysMatch": {"goog:chromeOptions": {"w3c": true, "args": ["--headless", "--allow-running-insecure-content"] }}}}'

9. How to configure the UDF to call a user-defined Sleep function

Q: How to configure the UDF to call a user-defined Sleep function, and interact with _WD_WaitElement() and _WD_LoadWait() to make the script more responsive?
A: Try to use: _WD_Option("Sleep") . Example:

#include <ButtonConstants.au3>
#include <GuiComboBoxEx.au3>
#include <GUIConstantsEx.au3>
#include <MsgBoxConstants.au3>
#include <WindowsConstants.au3>
#include "wd_helper.au3"

Global $idAbortTest
Global $WD_SESSION
_Example()

Func _Example()
	SetupChrome()

	; Create a GUI with various controls.
	Local $hGUI = GUICreate("Example")
	Local $idTest = GUICtrlCreateButton("Test", 10, 370, 85, 25)
	$idAbortTest = GUICtrlCreateButton("Abort", 150, 370, 85, 25)

	; Display the GUI.
	GUISetState(@SW_SHOW, $hGUI)

	ConsoleWrite("- TESTING" & @CRLF)

	Local $sFilePath = _WriteTestHtml()

	; Loop until the user exits.
	While 1
		Switch GUIGetMsg()
			Case $idTest
				_WD_Navigate($WD_SESSION, $sFilePath)
				_WD_WaitElement($WD_SESSION, $_WD_LOCATOR_ByXPath, '//a[contains(text(),"TEST")]', 100, 30 * 1000) ; timeout = 50 seconds
				ConsoleWrite("---> @error=" & @error & "  @extended=" & @extended & _
						" : after _WD_WaitElement()" & @CRLF)

			Case $GUI_EVENT_CLOSE
				ExitLoop

		EndSwitch
	WEnd

	; Delete the previous GUI and all controls.
	GUIDelete($hGUI)

EndFunc   ;==>_Example

Func _My_Sleep($iDelay)
	Local $hTimer = TimerInit() ; Begin the timer and store the handle in a variable.
	Do
		Switch GUIGetMsg()
			Case $GUI_EVENT_CLOSE
				ConsoleWrite("! USER EXIT" & @CRLF)
				Exit
			Case $idAbortTest
				Return SetError($_WD_ERROR_UserAbort)
		EndSwitch
	Until TimerDiff($hTimer) > $iDelay
EndFunc   ;==>_My_Sleep

Func _WriteTestHtml($sFilePath = @ScriptDir & "\TestFile.html")
	FileDelete($sFilePath)
	Local Const $sHtml = _
			'<html lang="en">' & @CRLF & _
			'    <head>' & @CRLF & _
			'        <meta charset="utf-8">' & @CRLF & _
			'        <title>TESTING</title>' & @CRLF & _
			'    </head>' & @CRLF & _
			'    <body>' & @CRLF & _
			'        <div id="MyLink">Waiting</div>' & @CRLF & _
			'    </body>' & @CRLF & _
			'    <script type="text/javascript">' & @CRLF & _
			'    setTimeout(function()' & @CRLF & _
			'    {' & @CRLF & _
			'        // Delayed code in here' & @CRLF & _
			'        document.getElementById("MyLink").innerHTML="<a>TESTING</a>";' & @CRLF & _
			'    }, 20000); // 20000 = 20 seconds' & @CRLF & _
			'    </script>' & @CRLF & _
			'</html>'
	FileWrite($sFilePath, $sHtml)
	Return "file:///" & StringReplace($sFilePath, "\", "/")
EndFunc   ;==>_WriteTestHtml

Func SetupChrome()
	_WD_Startup()
	_WD_Option('Driver', 'chromedriver.exe')
	_WD_Option('Port', 9515)
	_WD_Option('HTTPTimeouts', True)
	_WD_Option('DefaultTimeout', 40001)
	_WD_Option('DriverParams', '--verbose --log-path="' & @ScriptDir & '\chrome.log"')
	_WD_Option("Sleep", _My_Sleep)

	Local $sCapabilities = '{"capabilities": {"alwaysMatch": {"goog:chromeOptions": {"w3c": true, "excludeSwitches": [ "enable-automation"]}}}}'
	$WD_SESSION = _WD_CreateSession($sCapabilities)
	_WD_Timeouts($WD_SESSION, 40002)
EndFunc   ;==>SetupChrome

Tools

The following tools will help you to automate your browser:

• ChroPath plugin: Makes finding an element by XPath, ID or CSS incredibly easy (Chrome, Firefox, Opera)
• SelectorsHub plugin: Next Gen XPath tool to generate, write and verify the XPath and cssSelectors (All browsers)

References

(Last modified: 2020/03/01)

Further information sources:

• W3C: https://www.w3.org/TR/webdriver/
• AutoIt WebDriver threads:
  • Danp2's thread in the Example Scripts forum - WebDriver UDF (W3C compliant version)
  • Danp2's closed thread in the General Help and Support forum - WebDriver UDF - Help & Support (this is part I)
  • Danp2's closed thread in the General Help and Support II forum - WebDriver UDF - Help & Support (II)
  • Danp2's active thread in the General Help and Support III forum - WebDriver UDF - Help & Support (III)
  • Water's thread in the General Help and Support forum - Discussion about this Wiki Page
  • Water's thread in the Example Scripts forum - WebDriver example scripts collection
  • Danp2's thread in the AutoIt Projects and Collaboration forum - Webdriver, Websockets, and Chrome DevTools Protocol

• WebDriver Exe documentation:
  • Chrome
  • Edge
  • Edge Chromium
  • Firefox
  • Opera