We've been using Selenium to automate some browser testing in Internet Explorer. On Windows 7 32-bit it works fine. However we're testing it on Windows 10 64-bit and it fails miserably.
Sometimes the browser doesn't even launch
If i change the driver versions, i can get the browser to launch but it hangs on the first page
The issue is only present when I use ForceCreateProcessApi. However I need to use ForceCreateProcessApi in order to use BrowserCommandLineArguments!
The exception is: The HTTP request to the remote WebDriver server for URL http://localhost:16639/session timed out after 60 seconds.
Here is the PowerShell code i use:
$seleniumOptions = New-Object OpenQA.Selenium.IE.InternetExplorerOptions
$seleniumOptions.InitialBrowserUrl = $SiteUrl
$seleniumOptions.ForceCreateProcessApi = $true
$seleniumOptions.BrowserCommandLineArguments = "-k"
$seleniumOptions.IgnoreZoomLevel = $true
New-Variable -Name IEDS -Value ([OpenQA.Selenium.IE.InternetExplorerDriverService]) -Force
$defaultservice = $IEDS::CreateDefaultService()
$seleniumDriver = New-Object OpenQA.Selenium.IE.InternetExplorerDriver -ArgumentList #($defaultservice, $seleniumOptions)
I've tried the following versions (x86 and x64 versions), and none of them work:
2.25.3
3.141
3.9.0
Can anybody advise on how to make this work? I've made sure that TabProcGrowth etc is set according to the documentation.
Thanks.
Try to use a 3.150.1 32-bit driver.
I'm not sure how it looks on PS - but I'm able to run IE with this driver config.
ie: { version: "3.150.1", arch: "ia32" }
Also I have a key for iexplore.exe here:
HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\Internet Explorer\Main\FeatureControl\FEATURE_BFCACHE
from this config link
Related
I am able to successfully execute a Powershell script with Selenium 4.x although i need to run chrome under a user profile (which will persist). The objective is to save the cookies. For that I believe I will need to add options as defined in Saving chrome cookies Selenium. I do see something similar here although I need it for Powershell.
So how do I implement options that can be used by this command. (This command is primarily for selenium 4.x).
$browser = Start-SeDriver -Browser Chrome
This is all I could get, but I am unsure how to add it all up:
$ChromeOptions.addArguments("c:/users/PsUser");
With Module https://github.com/adamdriscoll/selenium-powershell v4:
$browser = Start-SeDriver -Browser Chrome -Arguments "--user-data-dir=C:\Users\$($env:username)\AppData\Local\Google\Chrome\User Data"
Without Module:
When you instantiate the ChromeDriver object you can pass a ChromeOptions argument.
So basically:
# Your working directory
$workingPath = 'C:\selenium'
# Add the working directory to the environment path.
# This is required for the ChromeDriver to work.
if (($env:Path -split ';') -notcontains $workingPath) {
$env:Path += ";$workingPath"
}
# OPTION 1: Import Selenium to PowerShell using the Add-Type cmdlet.
Add-Type -Path "$($workingPath)\WebDriver.dll"
# Create a new ChromeDriver Object instance.
$ChromeOptions = New-Object OpenQA.Selenium.Chrome.ChromeOptions
$ChromeOptions.AddArgument("--user-data-dir=C:\Users\$($env:username)\AppData\Local\Google\Chrome\User Data")
$chrome = New-Object OpenQA.Selenium.Chrome.ChromeDriver($ChromeOptions)
# Launch a browser and go to URL
$ChromeDriver.Navigate().GoToURL('https://google.com')
# Cleanup
$ChromeDriver.Close()
$ChromeDriver.Quit()
I have been trying out some automation using Powershell and Selenium + Chrome, however I am struggling to be able to have the default download folder actually apply when Chrome is launched.
I have used this as a guide to set the default download directory in Powershell, and while it does boot launch Chrome with the directory set in "chrome://settings/?search=download", it isn't actually applied (unless you manually press the 'Change' button in settings, which defeats the point of automating it!) and still download to the default user downloads folder:
Chrome settings after PS launch
#WebDriver.dll - C# 4.10 - .NET 4.8
#ChromeDriver - 97.0.4692.71
$WorkingPath = 'C:\PSChromeSelenium'
[System.Reflection.Assembly]::LoadFrom("$($workingPath)\WebDriver.dll")
$ChromeOptions = New-Object OpenQA.Selenium.Chrome.ChromeOptions
$SavePath="C:\Temp"
$HashTable = #{}
$HashTable.Add("prompt_for_download", $true)
$HashTable.Add("default_directory", $SavePath)
$ChromeOptions.AddUserProfilePreference("download", $HashTable)
$ChromeDriver = New-Object OpenQA.Selenium.Chrome.ChromeDriver($ChromeOptions)
Downloading using 'Save / Save As' still opens to the default folder.
Any help would be appreciated! Thanks in advance.
I have been trying to get this to work via a game control panel TCAdmin.
$ModPg1 = Invoke-WebRequest "http://steamcommunity.com/sharedfiles/filedetails/?id=731604991"
$ModVer1 = ($ModPg1.ParsedHtml.getElementsByTagName('div') | Where{ $_.className -eq 'detailsStatRight' } ).innerText | Select -Last 1
If I run this cmdlet via a program like TCAdmin (or task scheduler), I get the following error....
Invoke-WebRequest : The response content cannot be parsed because the Internet Explorer engine is not available, or Internet Explorer's first-launch configuration is not complete. Specify the UseBasicParsing parameter and try again.
Explorer is installed, and set up. The script works just fine if I run it manually.
My guess is there is a way to get TCAdmin to run the scripts the same way I would as a windows User.
Cant find a way nearly as simple to scrape the info 'm looking for.
As for this...
get TCAdmin to run the scripts the same way I would as a windows User.
For any app to run as a user, that users profile must be used on the host where the code is to be run. You cannot natively run PoSH on a host as another user context. This is not a PoSH issue, it is a Windows User Principal security boundary. There are tools that let you do this. For example SysInternal PSExec and AutoIT. Yet as stated that error is pretty specific. The user profile for Internet Explorer has not been created and that only happens when you use IE at least once.
So, as Adam points, out, use the setting the error message states to use or use your code to start IE at least once.
$SomeUrl = 'https://stackoverflow.com'
$ie = New-Object -com internetexplorer.application
$ie.visible = $true
$ie.navigate($SomeUrl)
while ($ie.Busy -eq $true) { Start-Sleep -Seconds 1 } # Wait for IE to settle.
Again, if trying to run this in the context of another user, the two above tools will get you there, but you still have to fire up IE to have a profile for it.
I am trying to access the document of an internet explorer com object with windows 2012. The code works great in windows 2008 but as soon as I try to run it on windows 2012 (fresh install, tried on more than one server), the same code stops working. In other words, $ie.document.documentHtml returns as null.
Below is the code:
$ie = new-object -com "InternetExplorer.Application"
$ie.navigate2("http://www.example.com/")
while($ie.busy) {start-sleep 1}
$ie.document.documentHtml.innerhtml
Has the interexplorer com object changed in windows 2012? and if yes, how do I do I retrieve the document contents in windows 2012?
Thanks in advance
edit: Added a bounty to sweeten things up. Invoke-WebRequest is nice but it works only on windows 2012 but I need to use internet explorer and have it work both on windows 2008 and windows 2012. I have read somewhere that installing microsoft office solves the issue. It is not an option either.
edit2: as I need to remotely invoke the script on multiple windows server (both 2008 and 2012), I would prefer not to copy files manually
It's a know bug: http://connect.microsoft.com/PowerShell/feedback/details/764756/powershell-v3-internetexplorer-application-issue
An extract from the workaround:
So, here's a workaround:
Copy Microsoft.html.dll from a location where it is installed (eg: from C:\Program Files(x86)\Microsoft.NET\Primary Interop Assemblies to your script's location (can be a network drive)
Use the Load-Assembly.ps1 script (code provided below and at: http://sdrv.ms/U6j7Wn) to load the assembly types in memory
eg: .\Load-Assembly.ps1 -Path .\microsoft.mshtml.dll
Then proceed as usual to create the IE object etc. Warning: when dealing with the write() and writeln() methods use the backward compatible methods: IHTMLDocument2_write() and IHTMLDocument2_writeln().
$ie.document.documentHtml.innerhtml
The bigger question is how this ever could have worked. The Document property returns a reference to the IHTMLDocument interface, it does not have a "documentHtml" property. It is never that clear what you might get back when you use late binding as was done in this code. There is an old documentHtml property supported by the DHTML Editing control, that has been firmly put to the pasture. Admittedly rather a wild guess.
Anyhoo, correct syntax is to use, say, the body property:
$ie = new-object -com "InternetExplorer.Application"
$ie.navigate2("http://www.example.com/")
while($ie.busy) {start-sleep 1}
$txt = $ie.document.body.innerhtml
Write-Output $txt
If you still have problems, Powershell does treat null references rather undiagnosably, then try running this C# code on the machine. Ought to give you a better message:
using System;
class Program {
static void Main(string[] args) {
try {
var comType = Type.GetTypeFromProgID("InternetExplorer.Application");
dynamic browser = Activator.CreateInstance(comType);
browser.Navigate2("http://example.com");
while (browser.Busy) System.Threading.Thread.Sleep(1);
dynamic doc = browser.Document;
Console.WriteLine(doc.Body.InnerHtml);
}
catch (Exception ex) {
Console.WriteLine(ex.ToString());
}
Console.ReadLine();
}
}
As far as I can tell, on Windows Server 2012 to get the full html of a page:
$ie.document.documentElement.outerhtml
There is also an innerhtml property on the documentElement, which strips off the root <html> element.
Of course, if all you want to do is get the raw markup, consider using Invoke-WebRequest:
$doc = Invoke-WebRequest 'http://www.example.com'
$doc.Content
Get any PC with Office installed and copy Microsoft.mshtml.dll to your script location.
c:\program files (x86)\Microsoft.net\primary interop assemblies\Microsoft.mshtml.dll
add-Type -Path Microsoft.mshtml.dll
Script works.
I have installed the PowerShellPack on a Server 2008 R2 machine, but am unable to use the PSImageTools module. All image operations yield the error New-Object : Cannot load COM type Wia.ImageFile. I have tried installing the same package on my Windows 7 laptop, and there it works.
I figure this must be because of missing Dll:s on the server, but even copying the wia*.dll files to System32 from the laptop (there are no corresponding files already in place) helps. It is not possible to register the dlls (regsvr32 wiaaut.dll, for instance), that is replied to with "The file failed to load".
So, has anybody successfully used Wia on Windows Server 2008? We're going to run a scheduled powershell script to update the outlook image of our users, so it needs to be able to run on the server.
I get the same error on Windows Server 2008 x64. Tried both x64 and x86 version of PowerShell v2.
As an alternative you might be able to use System.Drawing.Image. Depends on what your script needs to do. Perhaps post some example code.
http://msdn.microsoft.com/en-us/library/system.drawing.image.aspx
Add-Type -AssemblyName System.Drawing
$image = [System.Drawing.Image]::FromFile("C:\pic.bmp")
EDIT: Here's how you can change an image's dimensions:
$pixWidth = 90
$pixHeight = 90
$image2 = New-Object System.Drawing.Bitmap -ArgumentList $image, $pixWidth, $pixHeight
$image2.Save('C:\new_pic.bmp', [System.Drawing.Imaging.ImageFormat]::Bmp)
Found this article after quite a bit of searching:
http://kb.winzip.com/kb/entry/207/
Basically, you need to enable and set to Automatic the "Windows Image Acquisition" service in Windows 2003, or in Windows 2008, install the Desktop Experience feature first, then enable the "Windows Image Acquisition" service and set to Automatic.