Selenium: using capabilities for different display resolution leads to error message

4 views
Skip to first unread message

arkto...@gmail.com

unread,
Dec 4, 2017, 10:04:45 AM12/4/17
to Extensive Testing Users
Dear Denis,

We're evaluating your very useful testing framework in our company.

For Selenium tests, there is an argument for doOpen()named browserProfile (of type object). Could you please give an example of how to define such a browserProfile object? What is its intended purpose?


We'd like to run our tests in e.g. Chrome browser with settings of an iPad (display resolution).


I've also tried to use information for creating a desired capabilities object directly in the code for the relevant unit test (cf. https://github.com/dmachard/extensive-testing/blob/a19589bfcef1e3ee44c59c79edd6caa770e6a03d/external-projects/selenium/selenium-3.7.0/selenium/webdriver/common/desired_capabilities.py).


Either if we just use:

capabilities = webdriver.DesiredCapabilities.IPAD.copy()


or if we complete this definition with the following lines:

capabilities['platform'] = "WIN10"
capabilities['appiumVersion'] = "1.5.1"
capabilities['deviceName'] = "iPad Retina"
capabilities['deviceOrientation'] = "portrait"
capabilities['browserName'] = "chrome"



or if we use a definition like the following:

capabilities = {
    "browserName": "chrome",
    "version": "",
    "deviceOrientation": "portrait",
    "platform": "WIN10",
}



and if we assign the capabilities object to webdriver:

driver = webdriver.Remote(desired_capabilities=capabilities, command_executor=selenium_url)



we're always getting the following error message, right after having started the unit test:

ERR_TE_500: string indices must be integers





Is there anything to be changed in definition or approach? Thanks very much for checking this issue and your answer.


Best regards,

Bernhard

Auto Generated Inline Image 1
Reply all
Reply to author
Forward
0 new messages