Development

We are currently testing out DevOps but one thing that I can’t do is have MSTest hit our internal server (which is housed in Azure) that contains selenium and such.
DevOps of course is external.
I am no networking guru, but is there a way to build a connection between the two?
Or do we have to make our selenium server outward facing?

I have the following code

driver.find_element_by_id(“Element_ID”).send_keys(str(sheet.cell(row=r, column=2).value))

Values is taken from excel, but instead send_keys, I want to use document.getElementsByName script.

How can I adapt the getElementsByName script for my situation?

I’m wondering what some other options are for accessing the test automation framework dependencies from my test classes.
Right now I have a layered architecture:

Framework layer (contains all shared code for accessing deployed services)
Harness layer (builds the classes and dependencies within the framework. Ie handles DI)
Test layer (controls the flow of the tests via the objects created in the harness)

I’ve got a test base class in the harness layer that has access to the various API, database, and UI classes needed to run our tests. So using inheritance to provide access to these dependencies.
A sample test class would look like:
public class MyTestClass : TestBase
{
[Test]
public void sample_test()
{
APIs.MyAPI.GetSampleEndpoint(); // APIs is an inherited field in test base
}
}

I feel inheritance is the best way to go about the issue of dependencies within a test class. The downside is managing the wrapping class that the “APIs” field is attached to.
Are there better options to go about managing a large list of dependencies in a clean way?

Selenium webdriver is not able to pull up the chrome dev tools. From what it looks like puppeteer could do this, but will it give me a way to save the interactive timeline report? Can puppeteer be run in a headfull mode? in an incognito tab? I saw the developers page but no specific examples

I have a test scenario where the web application checks for the user’s leave intent i.e., mouse hovering from the page to the browser close and then a frame gets triggered. Is there any way to do it?

Edit: So the dev implementation is that when the user is moving to the top of the document frame gets triggered and it is not browser close but moving to corner of the document.